ENGIN. -CT05 iggi CORNELL UNIVERSITY LIBRARY Gift of the Family of John F. McManus Associate Dean Gbllege of Engineering _ Cornell University Library TA 403.T54T3 1894. A text-book of the materials of construe 3 1924 004 803 163 Cornell University Library The original of this book is in the Cornell University Library. There are no known copyright restrictions in the United States on the use of the text. http://www.archive.org/cletails/cu31924004803163 PRIVATE LIBRARY OF John R. XaldUondo. Works of M RoM. Hi Tliarston. Published by JOHN WILEY & SOqS, 53 E. Tenth Street, New York. SCATSRIAIiS OF ENGINEERINQ. A work designed for Engineers, Students, and lirtiaana in woo.-* metal, and stone. Aiso as a TEXT-BOOK in Soienlific Schools showl ing the properties of the subjects treated. By Pr( '. R. H Thurston Well illustrated. In three parts. Part I. THE NON-METALLIC MATERIALS C ? ENaiNEEH- ING AND METALLURGY. urxjsjijijt With Measures in British and Metric Units, and Me ic and Reduction 8vo, cloth, sa 00 3S0 Tables Fart II. IRON AND STEEL. The Ores of Iron ; Methods of Eeduotion ; Manuf aiuring Processes • Chemical and Physical Properties of Iron and Ste( ; Strength Duc- tility, Elasticity and Resistance ; Effects of Time, ' smperature and repeated Strain ; Methods of Test ; Specifications 8vo, cloth Part III. THE ALLOYS AND THEIR CONSTI' [JENTS. Copper, Tin, Zinc, Lead, Antimony, Bismuth, Nioke Aluminum, etc.; The Brasses, Bronzes ; Copper-Tin-Zino Alloys ; Other Valuable Alloys; Their Qualities, Peculiar Characteristics; Tses and Special Adaptations; Thurston's "Maximum Alloys"; trength of the Alloys as Commonly Made, and as Affected by Sioial Conditions: The Mechanical Treatment of Metals 8vo, cloth, "As intimated above, this work will form one of the aost complete as well as modem treatises upon the Materials used in al sorts of Building Constructions. As a whole it forms a very comprehei ve and practice book for Engineers, both Civil and Mechanical."— .Ama an Machinist. *' We regard this as a most useful book for reference i ita departments • it should be in every Engineer's library." — Mechanical i jineer. ' MATERIALS OP CONSTRUCTION. A Text-book for Technical Schools, condensed om Thurston's " Materials of Engineering." Treating of Iron anc teel, their ores manufacture, properties and uses ; the useful metel ,nd their alloys' especially brasses and bronzes, and their " kalcl as " : strength! ductility, resistance, and elasticity, effects of pre nged and oft- repeated loading, crystallization and granulation ; jculiar metals : Thurston's " maximum alloys " ; stone ; timber ; i iservative pro- cesses, etc., etc. By Prof. Robt. H. Thurston, of O lell University. Many illustrations lick 8vo, cloth, "Prof. Thurston has rendered a great service to the rofession by the publication of this thorough, yet comprehensive, text-1 k. . . . The book meets a long-felt want, and the well-known rcput ^n of its author is a sufficient guarantee for its accuracy and rhoroughnei '— Building. TREATISE ON PRICTION AND LOST WORK !! MACHIN- ERY AND MILL WORK. Contaiuing an explanation of the Theory of Friotio i,nd an account of the various Lubricants in general use, with a sord of various experiments to deduce the laws of Friction and Lul jated Surfaces, etc. By Prof. Robt. H. Thurston. Copiously illust ed..8vo, cloth, "It is not too high praise to say that the present trc le is exhaustive and a complete review of the whole subject. "—.Immco; ngineer. STATIONARY STEAM-ENGINES. Especially adapted to Electric Lighting Purposes, reating of the Development of Steam-engines— the principles of istruotion and Economy, with description of Moderate Speed anc igh Speed En- gines. By Prof. R. H. Thurston 12mo, cloth, " This work must prove to be of great interest to both uufacturers and users of ateam-eusi-mi"— Builder and Wood'wor/cer, Z 50 5 00 3 00 150 Copyright, 1884. Bt ROBERT H. THURSTON. Copyright, 1890, By ROBERT H. THURSTOM Press of J, J. Little Sk Co., Nos. 10 to 20 Astor Place, New York, PREFACE. The volume which is here presented to the profession is an abridgment of the larger work of the author, " Materials of Engineering." It is prepared for the purpose of supplying to students and instructors a text-book, for use in private reading, and in those colleges and technical schools in which the time which would be demanded for the larger work can- not be given to the subject. It is only in special schools of engineering, and rarely, even in them, that more than a small part of the matter contained in the three volumes of the larger treatise can be taken in the class-room. The Author has, in his own course, used a part of the first and second volumes, and has summarized in lectures the principal facts contained in the third ; but even a special school of mechani- cal engineering cannot afiford time for more than this, or, indeed, perhaps, as much. An abridgment has, therefore, been prepared, in which the essential parts only of the earlier treatise have been retained. The volume thus constructed is of manageable size, is of such comparatively low cost that it can probably be brought into use in any technical school ; and yet it contains enough of the matter collected in the extended work to furnish a basis for the course in construction and machine-design taught in the most elaborate course yet adopted in any such schools. The origin, nature, method of preparation, and the useful properties of all the common, and so-called useful, metals, and their strength, elasticity, and other qualities essential to their introduction into the various constructions which the engineer is called upon to build or to inspect, are IV PREFACE. treated of at considerable length, and the influence of the more common conditions affecting them is studied. The chapters on the reduction of the ores of the metals are substantially as complete as in the unabridged work ; those treating of the properties and uses of those metals are but slightly condensed ; and the portion of the treatise relat- ing to the alloys retains the more essential facts. In the con- densation of the matter found in the original, the effort has been made to select for excision, mainly, the parts which give at great length the details of the less important proc- esses, and the less essential data obtained by experiment ; while the general, and the average, results are retained, as being essential information for the designing engineer. In the text, " Materials of Engineering," refers to the larger work. CONTENTS. CHAPTER I. QUALITIES OF THE METALS. PAGE Early Metallurgy 3 Methods of Production 5 Fluxes 8 Fuels 9 Mechanical Processes 9 Early Working of Metals 10 Metal defined 12 Useful Metals 12 Distribution of Ores 13 Requirements of the Engineer 13 Special Qualities of the Metals 14 Relative Tenacities 15 Hardness 16 Conductivity 17 Density and Lustre *. 20 Ductility and Malleability 23 Odor and Taste 24 Crystallization 26 Specific Heats of Metals 27 Expansion of Metals 3° Application of Principles 3^ Force of Expansion 31 Fusibility 34 Latent Heat '• 3^ Chemical Character 37 Alloys defined 37 VI CONTENTS. CHAPTER II. IRON AND ITS ORES. PAGE Iron ; its properties 39 Wrought Iron of Ancient Times 39 Prehistoric Metallurgy 4° Greek " 4i Egyptian " 42 Early Methods 43 Direct Process 44 Puddling 48 Six Epochs of Metallurgy 49 American Iron-making 49 Ores Classified 5° Meteoric Iron S^ CHAPTER III. REDUCTION OF ORES ; PRODUCTION OF CAST IRON. Preliminary Operations 53 ; Grading the Ores 53 Calcination and Roasting 53 Methods of Roasting 54 • Roasting in Heaps 55 ? " " Kilns 56 ' Making up the Furnace Charge 57 \ Character of Ore desirable .• 57 Composition of Charges 58 Form and Proportions ©f Blast Furnaces 58 Shape of the Furnace 62 Putting Furnace in Blast 63 Chemistry of Reduction 64 Bell's, Akerman's, other Researches 65 Charges in the Furnace 67 ' Densities and Specific Heats 67 Size of Blast Furnace 68 Height " " 69 Temperature of Furnace 69 " Blast 70 CONTENTS. Vir PAGE Hot-blast Stoves 71 Whitwell's " 75 Blowing Engines 76 Hoists 77 Water Supply 81 Steam Boilers 81 Grades of Cast Iron 82 Bloomary ; Catalan Furnaces 83 American " " 84 Siemen's Process 85 CHAPTER IV. MANUFACTURE OF WROUGHT IRON. Characteristics of Wrought Iron 87 Decarbonizing Processes 87 Forge Process • 88 Refineries •• 89 Puddling Processes 90 Theory of Puddling 94 New Methods ' 97 Rotating Puddling Machines 98 The Puddle Ball 104 Mill-Work 105 Hammering i°7 Squeezing i°9 Rolling I" Rolls "I Roll-trains ii4 Rail and Beam Mills 1 15 Continuous Mills "S Universal " "5 Rate of Cooling 116 Rolled Iron "^ Forms of Rolled Iron "7 Wire-drawing ^2° Draw-plates ^^° Resistance in Wire-drawing 121 VliI CONTENTS. PAGE Wire-gauges 123 Peculiar Shapes , . , , 125 Designing 126 Malleableized Cast Iron 127 Tin-plate , 128 Russian Sheet Iron , , . 131 CHAPTER V. MANUFACTURE OF STEEL. Steel ; Definitions ,. 132 Steel-making Processes 136 Steel of Cementation 137 Cast Steel 140 Crucible Steel 141 Open-hearth Steels 145 Siemen's Direct Process 149 Fluxing Steels 151 Pneumatic or Bessemer Steel'. 151 Plant and Apparatus 152 Operation of the Bessemer Process 160 Properties of Bessemer Steel 164 Dephosphorization 165 Qualities of Steel 167 Decarbonization of Cast Iron 168 CHAPTER VI. CHEMICAL AND PHYSICAL PROPERTIES OF IRON AND STEEL. Pure and Commercial Iron 170 Effect of the Elements on Iron 171 Chemical Composition of Irons 178 Analyses of Cast Irons 180 Cold and Hot-blast Irons 181 Charcoal, Coke, and Anthracite Irons 181 Analysis of Wrought Iron 182 Chemistry of Cementation 190 Composition of Crucible Steels 191 " " Bessemer and Siemen's Steels 194 CONTENTS. ix PAGE Physical Properties of Iron ic,6 " " Cast Iron 198 Steel 199 Working Steel 200 Tempering. 205 Theory of Hardening and Tempering 209 Hardening by Compression 210 Corrosion of Iron and Steel 210 Durability " " 213 Preservation " " 214 CHAPTER VII. THE NON-FERROUS METALS. Copper 215 " Ores... 216 " Reduction 219 " Properties 226 " Commercial 227 Tin 234 " Ores 234 " Commercial 236 Zinc 239 " Ores 240 " Commercial 243 Lead , 247 " Ores 348 " Reduction 249 " Commercial 251 Antimony 252 Bismuth 253 Nickel 254 Aluminium 256 Mercury. 258 Platinum 260 Magnesium 262 Arsenic 263 Iridium , 264 X CONTENTS. PAGE Manganese 265 Rare Metals 266 Costs of Metals 268 CHAPTER VIII. THE BRONZES. Alloys of Copper 270 " " " History 271 " " " and Tin 274 Properties 276 Principal Bronzes 277 Phosphor-bronze ; Manganese bronze 281 Table of Properties of Alloys 282 CHAPTER IX. THE BRASSES. Brass Defined 288 " its uses 289 Muntz Metal 290 Special Properties of Brasses 291 Applications in the Arts 292 Working Brass 293 Table of Properties of Brasses 295 ' CHAPTER X. KALCHOIDS AND MISCELLANEOUS ALLOYS. The Kalchoids 300 Copper, Zinc, and Iron 302 " Tin, and Iron 302 Manganese Bronze 303* Aluminium Bronze 306 Copper and Nickel 309 German Silver 310 Tin and Zinc 311 CONTENTS. xi PAGE Antimony Alloys 312 Pewter and Britannia Metal 312 Iron and Manganese 313 Platinum and Iridium 313 Spence's " Metal" 314 CHAPTER XI. MANUFACTURE OF ALLOYS. Brass Working 315 " Founding. 317 Melting and Casting 317 Furnace Management 319 Preparation of Alloys ■ 320 Art Castings 322 Stereotype Metal 324 German Silver 325 Babbitt Metal 325 Solders 326 Standard Alloys 329 Bronzing 335 CHAPTER XII. STRENGTH, ELASTICITY, AND DUCTILITY OF METALS. Resistances Classed 340 Factors of Safety 342 Measures of Resistance , 344 Method of Resistance 345 Equations of Resistance 346 Series of Elastic Limits 347 Shock and Resilience 349 Elastic and Total Resilience 350 Variation of Form of Test-piece 352 Use of Testing Machines 358 'Method of Record 359 Records of Test 362 Xll CONTENTS. PAGE Variation of Tenacity with Size 365 Tests of Long Bars 368 Piling and Reworking 368 Tenacity of Ingot Irons and Steels 369 Elongations 370 Boiler and Bridge Steels 373 Strain- diagrams of Cast Iron 378 Stays 379 Cylindrical Boiler Shells 382 Strength of Cast-iron Cylinders 384 Cast Iron in Compression 390 Long Bars in Compression 391 Wrought Iron in Compression 392 Flues and Cylinders under External Pressure 393 Columns, Posts or Striits 394 Flexure of Columns 395 Columns of great length 396 Formulas for strengthening Columns 398 Shearing 405 Riveted Work 406 Shafts 408 Metals and their Strain-diagrams 411 Strain-diagrams of Forged Iron 413 Inspection of fractured Test-pieces 417 Strain-diagrams of Steels .' 419 Strain-diagrams of Tool-steels 422 Inspection of Steel Test-pieces 424 Fractured Surfaces . . . ? 426 Tenacity of Copper 429 Copper in Compression 430 Copper under Transverse Stress 432 Modulus of Elasticity of Copper 434 Copper in Torsion 435 Tests of Copper by the U. S. Board 435 , Strength of Tin '. 436 Modulus of Elasticity of Tin 440 Strength of Zinc 442 Strength of Lead 444 CONTENTS. xiii PAGE Wertheim's Tests of Metals 446 Bishop's Method of Test 447 CHAPTER XIII. STRENGTH OF BRONZES. Classification of Bronzes 448 pun Bronze 448 Manganese Bronze 450 Copper-Tin Alloys and U. S. Board Tests 451 Strain-diagrams of Bronzes 454 Tenacities of the best Alloys 454 CHAPTER XIV. STRENGTH OF BRASSES. Classification of Brasses 456 Early Experiments • • ■ ■ 45 7 Sterro-Metal 458 Moduli of Elasticity. 459 Copper-Zinc Alloys 459 " and U. S. Board Tests 459 Results of Tests by the U. S. Board 463 CHAPTER XV. STRENGTH OF KALCHOIDS AND OTHER COPPER-TIN-ZINC ALLOYS. Kalchoids defined 464 Sterro-metal 465 Copper-Tin-Zinc Alloys 466 Plan of Research ■ 466 Alloys selected 468 Method of Exhibition and Record of Results 469 Deductions 471 Strain'diagrams' ■ • • • 473 Tenacities of Strong Alloys 474 XIV contents: PAGE Ductilities 47^ Improvement , 4^1 Maximum Bronzes 4°^ Results of Tests 484 Conclusions as to the Strongest Bronzes 4^6 CHAPTER XVI. CONDITIONS AFFECTING STRENGTH. ' Effect of Heat 489 Conclusions from Experiment 49^ Effect of Age and Exposure 49^ Crystallization 5°° Flow of Metals 5°? Relief of Internal Stress by Rest 5^° Effect of Time S12 Velocity of Rupture 5^4 Rate of Set of Metals 516 Variation of Normal Elastic Limits 5^° Evidence of Overstrain 5^2 Effect of Orthogonal Strains 526 CHAPTER XVII. CONDITIONS AFFECTING STRENGTH OF NON-FERROUS METALS. Variation of Strength of Copper with Temperature 528 Effect of Heat on Bronze 529 " " " "other Metals 532 " , " " " Moduli of Elasticity 532 Stress produced by Change of Temperature 533 Effect of Sjidden Variations of Temperature 534 Chillrcasting. 535 Tempering and AnneaUng \ 536 Effect of Time, of Stress, and of Velocity of Rupture 539 Effect of Prolonged Stress on Tin 542 " ," " " " Bronzes 545 Fluctuation of Resistance with Time 546 CON-TENTS. XV PAGE Decrease of Resistance with Time of Stress 548 Increase of Deflection under Static Load 550 Depression of Elastic Limits , 556 Effect of Variable Stress 560 " " Repeated Strain 561 CHAPTER XVIII. STONES AND CEMENTS. Use by Engineers 563 Classes of Stones 564 Silicious Stones 564 Calcareous Stones 569 Argillaceous Stones 571 " Firestones " 572 Hardness of Stones 573 Crushing Resistance of Stones 573 Transverse Strength 575 Durability of Stones 575 Effect of Heat on Stones 577 Artificial Stones 578 Brick 578 Firebrick 581 Mortars and Cements 582 Concrete 585 Bdton Coignet 586 Strength of Mortars and Cements 587 Bituminous Cements 588 Masonry 588 Brickwork 594 Cost of Masonry 595 CHAPTER XIX. TIMBER. Characteristics of Timber ; Definitions 597 Felling Timber 599 XVI CONTENTS. PACE Seasoning Timber 599 Shrinkage of Timber ' 6oi Nomenclature 6°^ Characteristics of good Timber 603 Influence of Climate and Soil 604 Decay of Timber 605 Wet-rot and Dry-rot 606 Marine Destroyers of Timber 606 Varieties of Timber 607 White Pine 608 Red Pine 608 Yellow Pine 609 Southern Pine 609 Pitch Pine 610 Northern Pine of Europe 611 Cypress 611 Qualities of Pine Timber 612 Norway Fir 612 Black Spruce 613 Hemlock 613 Red Spruce 614 Larch 614 Linden 615 Cedar 615 Tar, Pitch, Turpentine 617 Oaks 618 Live Oak 619 White Oak ^. 619 Post Oak 623 Various Oaks 624 Beech 625 Chestnut 625 Ash 627 Elm 627 Locust 628 Hickory 628 Black Walnut 628 Cherry, Plum, Holly 629 CONTENTS, xvii PAGE Maples 630 Dogwood 630 Mahogany 6,1 Lignum Vitae 632 Spanish Cedar '. 632 Teak 632 Camphor-AVood, Boxwood, Ebony 633 Lancewood, Greenheart, Rosewood 634 Measuring Timber 634 CHAPTER XX. STRENGTH AND USES OF TIMBER. Conditions Determining Strength 635 Coefficients of Elasticity 636 Factors of Safety 636 Tenacity of Timber 636 Crushing Resistance 638 Long Pillars 639 Hodgkinson's, Rankine's, and Gordon's Rules 639 Columns for Mills 642 Resistance to Shearing 644 Transverse Strength 646 Stiffness of Beams 650 Formula for Flexure 650 Working Loads for Floor-beams 652 Effect of Prolonged Stress 657 Woods having Commercial Value 659 Markings of Wood 661 Coloring of Wood 662 Carpentry 663 Joints 664 Pins 666 Glues 669 Preservation of Timber 670 Paints and Varnishes 671 Charring Timber 675 Saturation of Timber 676 XVUl CONTENTS. CHAPTER XXI. MISCELLANEOUS MATERIALS. F^GE Leather ; Belts 687 Friction of Belts 689 Paper 692 India Rubber and Gutta Percha 693 Cordage ; Knots 695 MATERIALS OF CONSTRUCTION. THE USEFUL METALS. CHAPTER I, HISTORY AND CHARACTERISTICS OF THE METALS AND THEIR ALLOYS. !^ The knowledge of metals possessed by the early races of mankind was of the most inexact and unsatisfactory character. They were probably led to seek a method of utilizing them, first, by the demands of their fighting classes. Their structures, their implements of agriculture and war, and their domestic utensils were, in the earliest stages of their race- history, of wood, bone, and stone. All races are found to have advanced to their present condition of civilization from a primitive state of barbarism, in which they were entirely ignorant of the use of metals, and knew nothing of even the simplest processes of reduction. The weapons of mankind, in prehistoric times, were at first made of hard wood, of bone, or of stone, fashioned with long and patient labor into rude and lAefficient forms. As the race advanced in knowledge and intelligence, they acquired, by some fortunate circumstance, a knowledge of the methods of reducing from the ores the more easily deoxidized metals, and, still later, those which cling with tenacity to oxygen, and require considerable knowledge and skill, and special apparatus for their reduction to the metallic state ; and at a still very early period, they applied the more common and more generally useful metals in their rude manufactures. 4 MATERIALS OF CONSTRUCTION— THE USEFUL METALS. It has thus happened, that mankind has passed through what are designated by the geologists as the ages of stone, of bronze, and of iron, and may be considered as having just entered upon an age of steel. The ancients, at .the commencement, and immediately before the Christian era, were familiar with but seven metals. The eariiest of historical records indicate that, long pre- vious to their date, some metals were worked, although with rude apparatus, and in an exceedingly unintelligent manner. Tubal Cain was an artificer in brass and in iron ; and several sacred writers refer to the use of these metals and of gold and silver, in very early times. Profane writers also present similar testimony ; and the discovery of implements of metal among the ruins of the ancient cities of Asia and Africa, and in the copper mines and other localities of North America, indicate that some knowledge of metallurgy was acquired many centuries before our era. The Hebrews were familiar with gold, silver, brass (bronze ?), iron, tin, and lead, and possibly copper and other metals. Bronze and brass were not always distinguished by ancient writers, but both alloys were known at a very early date. Phillips gives analyses * of a number, of samples of the latter dating from B.C. 20 to B.C. 165, and bronze was certainly made much earlier. Zinc was known in the metallic state at some early date, while tin was known in the earliest his- toric times. The Chinese, at a time far back of even their oldest his- torical records and traditions, seem to have been workers in iron and in bronze. Evidence has been found, in Hindostan, that the inhabit- ants of the Indian peninsula, at an era of their history of which we have lost every trace, were able not only to reduce these metals from their ores by rude metallurgical process^, but that they actually constructed in metal, works which are looked upon as remarkable for their magnitude. The Chaldeans, four thousand years ago, the Persians, the * Metallurgy, 1874, p. 6. HISTORY OF THE METALS AND THEIR ALLOYS. 5 Egyptians, and the Aztec inhabitants of America, if not an earlier race, had some knowledge of the reduction and of the manufacture of metals. The " Bronze Age," in Europe, is supposed to have origi- nated in the south of England, and to have gradually spread over Europe, a knowledge of the methods of working coppet and bronze finally becoming very general. The bronze age of Central America antedated that of Northern America, where the contemporaneous age was that of copper. It is probable that copper may have been the first metal worked by these early metallurgists, and that tin was next discovered and used to harden the copper, as is done at the present time. In the manufacture of bronze, the ancients became very skilful; probably long before the discovery and use of iron. The bronze implements discovered on both con- tinents have sometimes nearly the hardness and sharpness of our steel tools. It is only within a comparatively recent period, however, that metallurgy has become well understood. To insure its rapid and uninterrupted progress, it was necessary that the science of chemistry should be first placed upon a solid basis, and this was only done when, about a century ago, Lavoi- sier introduced the use of the balance, and by his example led his brother chemists to employ exact methods of re- search. The valuable qualities of the metals used in con- struction are very greatly influenced by the presence of impurities, and by their union with exceedingly minute quantities of the other elements, both metallic and non- metallic. In the processes by which the metals are reduced from their ores and prepared for the market, there is always greater or less liability of producing variations of quality and differences of grade, in consequence of the impossibility of always avoiding contamination by contact with injurious ele- ments during these operations, even where the ore was origi- nally pure. In the time of Lavoisier, but seventeen substances were 6 MATERIALS OF CONSTRUCTION— THE USEFUL METALS. classed as metals, and of these the characteristics upon which the classification was based were principally physical, and the place of newly discovered elements was long uncertain ; potassium and sodium were at first (1807) classed as non- metals. The distinction between metals and metalloids remains somewhat indefinite, and the type metal is considered, neces- sarily, ideal. The metals are usually solid, mercury being an exception ; they are usually liquefiable by heat, but arsenic is volatile without fusing; they are generally opaque, but gold is, in very thin leaves, translucent ; they are nearly all malleable and ductile, but in very variable degrees. The metals are good conductors ; the metalloids are not. The metals are electro-positivfe, as a rule ; the metalloids electro- negative. Metallurgy is the art of separating the metals from the chemical combinations in which they are met in nature, freeing them from impurities with which they may be mechanically mingled, and reducing them to the state in which they are found in our markets, and in which they are adapted for application in construction. The chemical combinations from which the useful metals are obtained, are usually either the sulphides or the oxides. The common ores of iron are peroxides, either hydrated or anhydrous, and copper is generally, except in the Lake Su- perior mining region of the United States, reduced from the state of sulphide. Lead is usually found combined with sulphur, forming a sulphide known as galena. Zinc is found and mined as an oxide, as a sulphide, and also as carbonate and silicate. The sulphide of iron is rarely or never mined as an ore of iron, although abundantly distributed in the form of pyrites. , The following table * illustrates the general character of the chief chemical processes employed for the purpose of reducing metals of ordinary occurrence from their ores. * Metals and Applications, G. A. Wright, London, 1878, HISTORY OF THE METALS AND THEIR ALLOYS. REDUCTION PROCESSES IN USE. \.— NATIVE METALS. By mechanical means e.g. gold washing. By simple fusion (liquefaction) .... e.g. bismuth. By solution in mercury e.g. gold-quartz. By solution in aqueous chemicals. . e.g. gold-quartz. II. — SIMPLE ORES ; i. c. , containing only one metal. A. — Oxides. Analytic By simple heating e.g. mercury, silver. (By heating in hydrogen eg. nickel, iron. By heating in carbon oxide e.g. iron (blast furnace). By heating with carbon (coal, I ( tin, arsenic, zinc, iron, coke, etc.) ) '•=■ ' j antimony. Analytic . Single decom- position . . . B. — Chlorides, Fluorides, etc. By heating alone By heating in hydrogen By action of cheaper metal, etc. By (a) wet processes By (i) dry processes . . _ By (t) amalgamation processes e.g. silver. e.g. platinum, gold. e.g. silver. e.g. copper, gold. e.g. magnesium, aluminium. C— Sulphides. Single decom- position . . . Double de- composition followed by single de- composition By heating with air e.g. mercury, copper, lead. By heating with cheaper metal, etc. e.g. mercury, antimony, lead. By roasting to oxide and reducing \ • as above f '*' ' By converting into chloride and \ ^ silver treating as above ) * ' zinc, antimony. Single decom- position . . . Double de- composition followed by single de- composition D .—Carbonates. • By heating with carbon. e.g. zinc, sodium, potassium. By roasting to oxide and reducing ) ^ j^.^^^ as above S By converting into chloride and ) ^ copper. treating as above ) '* ' ^'^ III. — COMPLEX ORES; i. e. , containing more than one metal. spi6 I. Alloy extracted by some or ) ( silver-lead alloy, spi other process, as above. ... i ■*" ( geleisen. II. Special processes adopted for ) extraction of metals sepa- ^ ^.^. cupriferous pyrites, rately ) B MATERIALS OF CONSTRUCTION— THE USEFUL METALS Descriptions of some of these methods are given in those chapters relating to the several metals. Fluxes are used in nearly all of the metallurgical processes, and their characteristics are determined by the special requirements of each case. Fluxes are, as the name (from fluo, to flow) indicates, sub- stances which assist in reducing the solid materials in the smelting furnace to the liquid state, forming a compound known as slag, or sometimes as cinder. It frequently happens that two substances, having a pow- erful affinity for each other, will unite chemically, when brought in contact, and fuse into a new compound at a much lower temperature than that at which either will melt alone. Silica fuses only at an extremely high temperature, if iso- lated, or if heated in contact with bodies for which it has no affinity ; but, if mixed with an alkali, as potash, soda, or lime, the mixture fuses readily. The two first-named alkalies are too expensive for general use in metallurgy; but thelat- ter is plentifully distributed, as a carbonate, and it is, there- fore, the flux generally used in removing silica from ores, by fusion. Borax similarly unites with oxide of iron to produce a readily fusible glass ; and it is, therefore, often used by the blacksmith as a flux when welding iron. Quartz sand is also used by the blacksmith for precisely the same purpose. Being composed almost purely of silicic acid, it forms a readily fusible silicate with the oxides of iron, and it is used wherever the mass of iron is of considerable size, and is capable of bearing, without injury, the high temperature necessary for its fusion. Fluor-spar, a native fluoride of calcium, has been fre- quently and extensively used as a flux. Its name was given to it in consequence of that fact. It is a very valuable fluxing material, and is used where the expense of obtaining it does not forbid its application. It has special advantages arising from the fact that it is composed of two elements, both of which perform an active and a useful part in the removal of the non-metallic constituents of ores. In the removal of sulphur HISTORY OF THE METALS AND THEIR ALLOYS. 9 and phosphorus from iron, it also possesses the great advan- tage that the resulting compounds produced by its union with those elements are gaseous, and pass off up the chimney, in- stead of remaining either solid or liquid in the furnace and contaminating the iron by their contact. Since the aim, in selecting a flux, is usually to form, with the impurities to be removed, a readily fusible glass, such materials are selected, in each case, as are found, by analysis or by trial, to unite in those proportions which produce such a compound. The "j/i3^" thus formed should usually be a compound silicate of lime and alumina, as free as possible from refractory substances, like magnesia, and from the oxides of the metal treated. The flux used, therefore, where an ore contains excess of silex, is a mixture of lime and alumina — as, for example, limestone and clay. Where the ore already contains alumina, limestone only may be needed. In the reduction of iron ores, limestone is very generally the only material added as a flux. The Fuels used in engineering and metallurgy are con- sidered very fully in special treatises (see Mat. of Eng.). Mechanical Processes. — Metallurgy includes both mechanical and chemical processes. The former consist in crushing and washing ores, or the gangue with which they are associated, to render the processes of reduction or of separation more easy, complete, and economical. The " stone- breaker," or " rock-crusher,'' is the form of crushing apparatus used for breaking rock into pieces of fixed size. It often consists of an arrangement of vibrating jaw, J (Fig. i), hung from the centre, K, and operated by a knee-joint, GEG, the connecting-rod of which, E, is raised and depressed by a crank, C, driven by a steam engine. A fly-wheel, B, gives regularity of motion, and stores energy needed at the instant when the squeeze occurs. Steel or cast-iron faces, PP, receive the wear. The breadth of opening at /, which de- termines the maximum size of pieces crushed, is adjusted by a wedge at OW, set by a screw at N. The jaw is pulled back 10 MATERIALS OF CONSTRUCTION— THE USEFUL METALS. Fig. I. — Stone-Crusher. by a spring R. Many modifications of this, the Blake crusher, are no^ made. Stamps consist of heavy weights carried at the ends of vertical rods, which are lifted either by cams on a continuously revolving shaft, or by the ac- tion of a steam piston. The former are the older, and for many kinds of work the most effective style ; the latter are, however, found vastly more economical for other cases, as in the crush- ing of some of the copper- bearing rock of the Lake Superior district. Washing machinery is largely used in silver mining and reduction, and less generally in working the ores of the " use- ful " metals. It takes many forms, according to the kind of work to be done ; this is usually the washing of earthy matter from harder ores or the separation of heavy masses from an earthy mass in which it is imbedded. The Working of Metals, as an art, antedated, un- questionably, the very earliest historic periods, and introduced the " age of bronze." The first metal-work was done in gold, silver, copper, bronze, brass, lead, and iron, and possibly tin. The East Indians, the Egyptians, the early Greeks, and per- haps other nations, were familiar with methods of working these metals and alloys, and are said to have been conversant with a now unknown art of hardening and tempering bronze, to give cutting edges on knives and weapons, which were only equalled by those of steel. Copper was much used during the Middle Ages, and from A.D. iicxD to 1500 espe- cially, for a great variety of objects. Bronze was the most common material for works in art among the older nations. ■ The metals were worked both by casting and by the " repouss^" method. The earliest castings were solid, and the art of economizing cost and weight by " coring out " the inner portions was one of later introduction. The first HISTORY OF THE METALS AND THEIR ALLOYS. 1 1' " cores " in bronzes were of iron, and were left in in the cast- ing ; still later, removable clay and wax cores were used. The finest Greek art-castings and those of the Romans, and later the Italian artists, were made by the method called, by French workers in bronze, that " h cire perdue" The statue or other object was first roughly modelled in clay, and in size slightly less than that proposed for the finished piece. On this clay model was laid a coating of wax, which was worked to exactly the intended finished size and form, and was frequently even given the smoothness of surface desired in the finished casting ; this formed a thin skin over the clay. A clay, or earthy, wash was next applied, covering the wax surface, and over this was placed a thick and strong mass of clay, worked on in soft state and allowed to dry and set. The whole was then baked slowly; the wax melted and flowed out from between the two masses of clay, leaving a space into which molten bronze was finally poured to form the casting. The two parts of the clay mould were secured together by stays of bronze which were built, or afterward driven, into both parts, and thus connected them together. When the casting had cooled, the clay was torn away from the outside and removed from the interior of the bronze ; the surface was finished up as required, and the work was done. The finest antique bronzes were thus made. The hammered, or " repoussi," work of the Greeks was wonderfully perfect at a date which is supposed to have been earlier than that of their large castings. The first efforts in this direction were rude ; the sheet metal was hammered into shape over blocks of wood, which had been roughly given the desired form. Later, a bed of pitch, or of soft kinds of cement, was prepared, and the sheets hammered into form by striking them on the back side, the bed yielding to the blow and thus allowing the metal to assume the desired shape without being broken by the hammer or by the punch used. The work was often reversed and the final finish given on the front side. This method produced some of the largest and the finest of the ancient Asiatic bronzes, and fine work in gold, silver, and copper. The Greeks excelled in this 12 MATERIALS OF CONSTRUCTION— TffE USEFUL METALS. method of metal-working. In many cases, the thickness of the metal was reduced nearly to that of paper, without injury to its surface. The Siris bronzes of about B.C. 400 are of this kind. Tin was probably worked into vessels for domestic use by the natives of Cornwall before the settlement of the country by the Romans. Lead was used throughout Europe, in the mediaeval period, in sheets for roof-coverings, and cast into objects of complicated form. Specimens remain of the former, exhibiting its great durability when exposed to the weather. Like the modern Chinese and Japanese artists, the ancient workers in metal used gold and silver to adorn and give relief to their castings in bronze. Mirrors, of fine surface and thus ornamented, are common among collections of the products of Greek art. The bronzes of the Italian artists of the Middle Ages are remarkable for their beauty as art work in metal, as well as for their beauty of design ; even their work in iron is famous for its unexcelled beauty and the skill exhibited in forging it. Modern work has not equalled that of the Middle Ages, or even that of the early Greeks. Metal is the name applied to above fifty of the chem- ical elements. The larger number of the metals are but httle known, and many are found in such extremely minute quantities, that we are not well acquainted with either their chemical or their physical characteristics. Some approach the non-metallic elements so nearly in their properties, that they are placed, sometimes in the one class, and sometimes in the other. Very few of the metals are well fitted for use in construction ; but, fortunately, those few are comparatively widely distributed, and are readily reduced from their oxides or sulphides, in which states of combination they are almost invariably found in nature. The " Useful Metals " are iron— in its various forms of cast iron, malleable or wrought iron, and steel — copper, lead, tin, zinc, antimony, bismuth and nickel, and occasionally aluminium and rarer metals are used for similar purposes. From this list of metals, and from their alloys, the engi- HISTORY OF THE METALS AND THEIR ALLOYS. 13 neer can almost invariably obtain precisely the quality of material which he requires in construction. He finds here substances that exceed the stones in strength, in durability under the ordinary conditions of mechanical wear, and in the readiness and firmness with which they maybe united. They are superior to timber of the best varieties in strength, hard- ness, elasticity and resilience, and have, in addition, the im- portant advantages, that they may be given any desired form without sacrificing strength, and may be united readily and firmly to resist any kind of stress. By proper selection or combination, the engineer may secure any desired strength, from that of lead, at the lower, to the immense tenacity of tempered steel, at the upper limit. He obtains any degree of hardness, or fusibility, and almost any desired immunity from injury by natural destroy- ing agencies. Elasticity, toughness, density, resonance, and varying shades of color, smoothness, or lustre, may also be secured. The Laws Governing Distribution of the Ores of the metals are comprehended in the science of geology. The detection of their presence in any locality, and bringing them to the surface of the ground, free from the foreign earthy substances which accompany them, is the work of the min- ing engineer, and of the miner. The "reduction" of the metals from ores, by chemical and mechanical processes, con- stitutes the business of the metallurgist. The engineer takes the metals as they are brought into the market, and makes ■use of them in the construction of permanent or movable structures. The Requirements of the Engineer include some acquaintance with the general principles, and with the ex- perimental knowledge, which are to be obtained by the study of geology, of mining, and of metallurgy, to aid him in select- ing the metals used in his constructions ; since their quali- ties cannot always be determined by simple inspection, and it. is not always possible to subject them to such tests as he may consider desirable before purchasing. In such cases, a knowledge of the localities whence the ores were obtained. 14 MATERIALS OF CONSTRUCTION— THE USEFUL METALS. familiarity with the processes of manufacture, and with the nature of the materials employed by the metallurgist, coupled with a knowledge of the effects of various foreign substances upon the quality of the metal, may enable the engineer to judge with some accuracy what metal will best suit his pur- poses, and what will be likely to prove valueless. He is also thus enabled to judge, should the purchased materia'l prove defective, where the defect in quality originated, and to place the responsibility where it belongs. The student will seek this knowledge in special works on geology and metallurgy. But brief reference can be made to these subjects here. All the metals possess, as a whole, a number of properties which define the class, although few of these properties are common to all. The metals all unite chemically with oxygen to form basic oxides, and some of them take higher proportions of oxygen, forming acids. All metals are capable of similarly uniting with chlorine. All are capable of fusion and lique- faction at certain temperatures, fixed for each, which are usually high. Mercury, however, is liquid at ordinary tem- peratures. The metals are also capable of vaporization, and their vapors have some physical characteristics quite different from those of the solid metal. Thus, silver, white when solid or liquid, becomes blue as a vapor ; mercury vapor is color- less, potassium is green. All are opaque, except in exceed- ingly thin films, when some become apparently translucent. Gold transmits green light, mercury blue, and silver remains opaque in the thinnest leaf yet made. The Special Qualities of the Useful Metals which give them their importance as materials of construction are : their strength, hardness, density, ductility, inalleability, fusibil- ity, lustre, and conductivity. Strength, or the resistance offered to distortion and fract- ure, is their most valuable quality. The strength of metals and alloys in general use has been very carefully determined by experiment, and will be given hereafter. Of the metals in our list, lead is the least tenacious, and steel is the strongest. QUALITIES OF THE METALS AND THEIR ALLOYS IS The Non-Ferrous Metals, which are to-day of com- paratively little importance to the engineer in the construction of machmes or of structures, and which have been so generally superseded by iron and steel in every department of art, were, m earlier times, in some cases, as copper, tin, lead, the most common materials of construction. The three just mentioned were known in prehistoric times, and the Greeks were also familiar with mercury, as well as with iron. Valentinus dis- covered and described antimony in the 15th century, and bismuth and zinc became known at about the same time or a little later. Brande discovered arsenic and cobalt about the middle of the i8th century, and Ward discovered cobalt.* Cronstedt discovered nickel and Scheele manganese in 1774, and tungsten was prepared in 1783 by the brothers D'Elhu- jart. Palladium, rhodium, iridium and osmium were isolated and described by Wollaston and others in 1803. The alkaline earths, recognized as oxides by Davy in 1807-8, were soon after deoxidized, and potassium and sodium became known. Aluminium and magnesium were separated in 1828 and 1829, respectively by Wohler and by Bussy, and cadmium had already been discovered by Stromeyer in 1818. The rarer and more unfamiliar metallic elements were found later. The properties of these metals have been referred to in a general way in an abridged account of them given in Part II. of this work. A more detailed account of those' used, in con- struction will occupy the greater part of this volume. The following is a resumi of the general characteristics of these metals. The Relative Tenacities are approximately as below, lead being taken as the standard. RELATIVE TENACITIES OF METALS. Lead i.o Tin 1.3 Zinc 2.0 Worked copper 12 to 20 Cast iron 7 to 12 Wrought iron 20 to 40 Steel 40 to 100 * Encyclopsedia Britannica, 1883, art. Metals. l6 MATERIALS OF CONSTRUCTION— THE USEFUL METALS. No two pieces of metal, even nominally of the same grade, have precisely the same strength. The figures can therefore only represent approximate ratios, as every variation of purity, structure, or even of temperature, is found to affect their strength. Cast metal is usually weaker than the same metal after having passed through the rolls or under the hammer ; those which can be drawn into wire are still more considerably strengthened by that process. Metals are stronger at ordi- nary temperatures than when highly heated, and " annealing " is found to reduce the strength of iron and steel, although frequently increasing their ductility, and produces an op- posite effect on copper and its alloys. " Hardening," pro- duces the contrary effect. The presence of impurities and the formation of alloys produce changes of strength, some- times increasing, sometimes diminishing it. Copper alloyed with tin or zinc, in certain proportions, is strengthened ; and the addition of a small percentage of phosphorus to the alloy has a marked effect in increasing its tenacity and ductility. Hardness varies in the metals as considerafjly as their tenacity, and, like the latter quality, is greatly influenced in the same metal by very slight changes, either physical or chemical. Thus metals are hardened by cold hammering and softened by sudden change of temperature. The addition of scarcely more than a trace of impurity often produces a marked change in the degree of hardness of metals. The scale of hardness, according to GoUner,* is as follows : Soft lead i Tin 2 Hard lead 3 Copper 4-5 Alloy for bearings (C, 85 ; T., 10 ; Z., 5). 6 Soft cast iron 7 Wrought iron 8 Cast iron lo-i I Mild steel 12-13 Tool " blue 14 " " violet 15 " " straw 16 Hard bearings (C, 83;Z., 17) 17 Very hard steel 18 * Tech. Blaetter; London Engineermg, June i, 1883, p, 519, QUALITIES OF THE METALS AND THEIR ALLOYS. I7 The hardness of metals, as determined by Dumas, is exhibited in the following table of their order. HARDNESS OF THE METALS. Titanium Manganese Platinum Palladium Copper Gold Silver Tellurium Bismuth Cadmium Tin Scratch steel. Scratched by Calc Spar. Chromium Rhodium Nickel Cobalt Iron Antimony Zinc Lead Potassium Sodium Mercury, Scratch glass. ■ Scratched by glass. Scratched by the nail. Soft as wax. Liquid, Conductivity, or their power of transmitting molecu- lar vibrations of either heat or electricity, is another property of the metals, upon which is founded many useful applications. Of the " useful " metals, copper has by far the highest conductivity, and is only second in this respect to gold and silver, the best known conductors. Its conductivity is greatly reduced by the presence of foreign substances. The powers of conduction for heat and electricity seem to have very similar relative values. Conductivity is reduced by increase of temperature and by presence of impurities. The following table of relative conductivities was deter- mined by the experiments of Despretz, and very closely con^ firmed by Forbes. RELATIVE THERMAL CONDUCTIVITIES OF METALS. Gold 1,000 Silver 973 Copper 878 Iron 374 Zinc 360 Tin 304 Lead 180 Marble 25 The electric conductivities obtained by Becquerel, and the 2 1 8 MATERIALS OF CONSTRUCTION— THE USEFUL METALS. thermal conductivities given by Wiedmann and Franz, are as below : * CONDUCTIVITIES OF METALS. , Electric. Thermal. In Vacuo. In Air. Silver 1,000 915 649 140 79-3 82.7 1,000 ■748 548 240 154 84 79 1,000 736 Gold 532 236 Tin 145 840 Lead 85 Bismuth 18 The resistance to the voltaic current has been found by Mr. K. Hedges t as follows, wire and foil being used, and strength of the current so adjusted that on increasing it 20 per cent, the metal would fuse. The experiments continued 24 hours and the temperature was 69° F. (21° C.) RESISTANCES OF METALS TO ELECTRIC CURRENTS. Metal. 1. Commercial tin, wire 2. Lead, soft 3. Copper, soft 4. Tin-foil, pure 5. Tin and lead 6. Aluminium (" Albo ") alloy, foil 7. Aluminium and tin Resistances as Measured. Before Heating. 0.815 Ohms. 0.835 " 0.810 0.860 0.800 " 0.835 " 0.820 " Change in 24 Hours, — 0.003 — 0.005 + 0.000 + 0.000 , — 0.160 + 0.000 + 0.0008 * Part II., p. 8, § 10. t Brit. Assoc. Reports, 1883, Sec. G, QUALITIES OF THE METALS AND THEIR ALLOYS. 1 9 Commercial copper (Rio Tinto), has been found to have, in some cases, but one-seventh the conductivity of pure copper. Conductivity is reduced by increase of temperature, ac- cording to Forbes, and at rates varying with the character of the metal. M. Benoit has measured the electrical resistance of various metals at temperatures from 0° to 860° C. The mean of the figures obtained is given in the following table, the second column giving the resistance in ohms of a wire 39.37 inches (i metre) long, and having a cross section of 0.03 inch (0.2 sq. cm.), and column three the same quantity in Siemens units. Column four gives the conductivity compared with silver : Metal. Silver, A Copper, A Silver, A.(i) Gold, A Aluminium, A Magnesium, H Zinc, A., at 350° Zinc, H Cadmium, H Brass, A. (2) Steel, A Tin Aluminium bronze, A. (3) Iron, A Palladium, A Platinum, A Thallium Lead German silver, A. (4) ... . Mercury Ohms. Siemens. .0154 .0161 100 .0171 .0179 90 .0193 .0201 80 .0217 .0227 71 .0309 •.0324 49-7 .0423 •0443 36.4 .0565 .0591 27-5 .0594 .0621 25-9 .0685 .0716 22.5 .0691 .0723 22.3 .1099 .1149 14 .1161 .1214 13.3 .11S9 .1243 13 .1216 .1272 12.7 .1384 .1447 . J^i-i ■ 1575 .1647 9-77 .1831 .1914 8.41 .1985 .2075 77.60 .2654 ■ 277s 5.80 ■ 9564 I. 0000 1. 61 A, annealed; H. hardened; (i) silver .75; (2) copper 64.2, zinc 33.1, lead 0.4, tin 0.4; (3) copper 90, aluminium 10 ; (4) copper 50, nicliel 23, zinc 25. These results, are all taken at 0° C, and agree closely with those obtained by other observers. The resistance increases regularly for all metals up to their points of fusion. This 20 MATERIALS OF CONSTRUCTION— THE USEFUL METALS. increase, however, differs for different metals. Tin, thallium, cadmium, zinc, lead, are found to vary similarly ; at 200° to 230° their resistance has doubled. The resistance of iron and steel doubles at 180°, quadruples at 430°, and at 860° is about nine times that at 0°. Palladium and platinum increase much less, their resistance becoming twice that at 0° C, at 400° to 450°. Gold, copper, and silver form an intermediate group. In general conductibility decreases more rapidly the lower its point of fusion. Iron and steel are exceptions to this rule. In alloys the variation is less than in their constituents, and this is especially the case with German silver. The thermal conductivity of brass was found by Isher- wood to be 556.8 thermal units (British) per hour per square foot and per 1° Fahr., and to vary at the difference of tem- perature. Silicon-bronze may be given a conductivity but little less than that of copper, but its tenacity then diminishes con- siderably; that having 95 per cent, the conductivity of copper, has but one half the strength of that of which the conductivity is 25 per cent. The Lustre of these metals is measured by their power of reflecting light. Thus, according to Jamin, silver may reflect 0.9 of the light sent between surfaces of mirrors made of that metal ; after ten normal reflections it yields from 0.24 to 0.48, the former figure being that for violet, and the latter for red light. The figures for speculum metal are 0.6 to 0.7, O.CX36 a^d 0.035 ; those for steel, 0.6, 0.006, and 0.007. Estimating weights of metal in various forms as used by the engineer is a simple operation. Thus : if d = diameter of a circular section, or the minor diameter of an ellipse ; d' = major diameter of ellipse ; / = length of piece, section uniform ; b = breadth ; /5 = a constant ; IV= total weight. QUALITIES OF THE METALS AND THEIR ALLOYS. 21 The weight of any piece of uniform section is W— kd'l for cylindrical bars ; = kdd'l " elliptical sections ; = Mdl " rectangular sections. The values of k when / is in feet, other dimensions in inches and W in pounds, are VALUES OF i IN IV = kbdl. Brass, sheet . . Iron, wrought Lead , sheet . . Steel, soft. . . . 3 -70° 3-333 4.950 3-400 For pipes, W= k{d^ — d^") when d^d^ represent the inner and outside diameters in inches. To obtain weights in kilogrammes when measures are in centimeters, multiply the above by 0.00241. The " metallic lustre " is a property of the metals almost peculiar to them, and constitutes one of their marked charac- teristics. Polished steel, and an alloy of copper and tin known as speculum metal, burnished copper and aluminium, as well as the precious metals, gold and silver, exhibit this beautiful and peculiar lustre very strikingly. Tin, lead, and zinc, are lustrous, but they are not capable of taking a sufficiently high polish to exhibit this quality in such a degree as the metals first named. The Specific Gravities of the commercial metals are as follows : The densities of pure metals according to Fownes,* are * Chemistry, loth ed., p. 297. 22 MATERIALS OF CONSTRUCTION— THE USEFUL METALS. SPECIFIC GRAVITIES OF PURE METALS. (Water at 60° F. (13.5 C.) = i.) Platinum 21 . 50 Iridium 21.15 Gold 19.50 Tungsten 17 . 60 Mercury 13 • 59 Palladium 11 . 80 Lead 11.45 Silver 10. 50 Bismuth 9 . 90 Copper ... V 8 . g6 Nickel 8.80 Cadmium 8 . 70 Molybdenum 8 . 63 Cobalt 8.54 Manganese 8.00 Iron 7-79 Tin 7-29 Zinc 7-10 Antimony 6 . 80 Tellurium 6.11 Arsenic 5-^8 Aluminium 2.67 Magnesium r.75 Sodium o.g7 Potassium 0.87 Lithium 0.59 For the purposes of the engineer, the densities and the weights per unit of volume of commercial materials are the data desired. The following table gives such a set of figures. As is seen by comparing the tables, authorities differ some- what in these figures. WEIGHTS AND DENSITIES OF COMMERCIAL METALS. Aluminium, cast. . ' ' sheet Antimony, cast . . . Bismuth, " ... Brass,* cast sheet " wire Bronze * (ordinary) Copper,* bolts " cast " sheet . . . " wire . . . . Gold, hammered . . " standard.... Gun metal (bronze) 2.56 2.67 6.7 9.8 8.4 8.5 8.54 8.4 8.8s 8.60 8. 88 8.88 19.4 17-65 8.153 LBS. IN CU. FT. KILOGS IN CU. M. 160 167 418 614 525 532 533 524, 548 537 549 550 1,205 1,103 510 2,560 2,670 6,700 9,800 8,400 8,500 8,540 8,400 8,850 8,600 8,800 8,800 19,400 17.650 8,153 QUALITIES OF THE METALS AND THEIR ALLOYS. 23 Iron, cast, from " " to " " average . . . , " wrought, from. . . " " to " " average Lead, cast " sheet Mercury, fluid ' ' solid Nickel, cast Pevifter Platinum, mass " sheet Silver, mass " standard Steel, hard , " soft Tin,* cast Type metal, cast Zinc,* cast " sheet 6.955 7.295 7.125 7.560 7.800 7.680 11.352 II. 4 13.6 15.632 7.807 11.600 19-550 20.337 10.5 10.534 7.82 7.834 7-3 10.450 7.03 7.29 LBS. IN kilog's CU. FT. IN CU. M, 435 6,955 456 7.295 445 7.125 473 7,560 488 7,800 480 7,680 710 11,352 712 11,400 848 13,600 977 15,632 488 7,807 725 11,600 1,219 19,500 1,271 20,337 655 10,500 658 10,534 496 7,820 491 7,834 456 7.300 653 10,450 439 7.030 456 7,290 Ductility and Malleability are properties of the met- als scarcely less important to the engineer than that of tenacity. The ductility of a metal or an alloy is its capacity for being drawn out into wire, by being pulled through holes in the wire-drawers' plates, each hole being slightly smaller than the preceding, until the wire reaches a limit of fineness which is determined by the degree of its ductility, and, as well, by the skill of the workman. Great tenacity, in proportion to the degree of hardness, or high tenacity, a low elastic limit and a certain viscosity, is the combination of qualities required to insure dura- bility. Gold has been drawn until the wire measured but -^^ inch in diameter, and silver and platinum are nearly as duc- tile. Iron and copper are the most ductile of the common metals. * See text later. 24 MATERIALS OF CONSTRUCTION— THE USEFUL METALS. The malleability of a metal, or the power which it pos- sesses of being rolled into sheets without tearing or breaking, is determined by its relative tenacity and softness. The malleability of the non-ferrous metals is determined by their plasticity simply, and this quality is observable in all metals having no defined elastic limit. It is also often determined to some extent by the physical condition of the metal; thus zinc, brittle in the ingot, is malleable at the boiling temperature of water, and, if worked at that tempera- ture, becomes permanently malleable in the sheet or the bar. Hardening and tempering are operations which can be per- formed on many metals with the effect of modifying their malleability and other properties ; but while sudden cooling from high temperature hardens steel, it softens copper and the bronzes and brasses. Ductility, being dependent upon tenacity largely, is not as generally observed as malleability. Gold is the most malleable of all metals, and has been beaten into sheets of which it would require 300,0CX) to make up a thickness of one inch. "Wrought iron of good quality, and the softer grades of steel, are very malleable ; the former has been rolled to less than YB^jj^ of an inch (0.00254 centimetre) thickness. Cast iron and hard steels are neither malleable nor ductile. Copper is very malleable, as well as ductile, if kept soft by frequent annealing ; tin possesses this property, also ; and zinc, although quite brittle when cold, becomes malleable at a temperature somewhat exceeding the boiling point of water ; its temperature being still further elevated, it again becomes brittle, so much so that it may be powdered in a mortar. Some of the copper-tin alloys exhibit the same peculiarity. Odor and Taste characterize many metals. Brass, for example, possesses a very marked taste and perceptible odor when applied to the tongue and when rubbed. These qual- ities may indicate solubility and volatility, but no direct ex- periment has revealed their precise nature. Many of the lighter metals are quite volatile at moderately high tempera- ture. QUALITIES OF THE METALS ANI> THEIR ALLOYS. 2% Lead can be rolled into quite thin sheets, but it is less malleable than either copper, tin, or the precious metals. The following is a table of the relative ductility of metals : ORDER OF DUCTILITY OF METALS. 1. Gold, 4. Iron, 7. Zinc, 2. Silver, 5. Copper, 8. Tin, 3. Platinum, 6. Aluminium, 9. Lead. In the following list, the metals named are placed in the order of their malleability. ORDER OF MALLEABILITY OF METALS. I. Gold, 4. Tin, 7. Zinc, 2. Silver, 5. Platinum, 8. Iron, 3- Copper, 6. Lead, 9. Nickel. Prechtl gives the following as the order in which the metals stand in this respect :* MALLEABILITY. DUCTILITY. Hammered. Rolled. Wire-drawn. 1. Lead, 2. Tin, 3. Gold, 4. Zinc, 5. Silver, 6. Copper, 7. Platinum, 8. Iron. Gold, Silver, Copper, Tin, Lead, Zinc, platinum. Iron. Platinum, Silver, Iron, Copper, Gold, Zinc, Tin, Lead. Authorities differ, however, in their statements in regard to the order of the metals in these respects, and the preceding figures as given in tables are often quoted from Regnault.f * Encyclopaedia Britannica. \ Regnault's Chemistry. 2b MATERIALS OF CONSTRUCTION— THE USEFUL METALS. The following table of the principal metals and then properties is extracted from Watts :* CHARACTERISTICS OF METALS. NAME. Ss 1 741 1803 NAME OP DISCOVERER. S. G. SP. HEAT. MELTING- POINT. CONDUC- TIVITY. Water - i. Ther- maJ. Elect. Platinum . . . Wood Descotils 21.5 21.15 19.26 15.60 11.80 11.33 10.57 9.80 8.94 8.82 8.02 7.84 7.30 7.13 6.72 2.56 1.74 0.0324 0.0326 0.0324 0.0319 0.0593 0.0314 0.0570 0.0308 0.0952 0.1086 O.I217 O.II38 0.0562 0.0955 0.0508 0.2143 0.2499 8.4 tR Iridium .... Gold 1200° C. (?) — "^q" C. . . 53.2 78. Mercury. . . . Palladium . . 1803 WcUaston 6.3 8.5 100 1.8 73.5 18.4 8.3 100 Lead 332° c... 1000° c. . . 270° c . . . 1200° C. (?) Silver Bismuth .... 1.2 Copper 99.9 13. 1 Nickel 1751 1774 Cronstedt Gahn; Scheele. Manganese. . 2000° C. (?) II. 9 14.5 16.3 Tin 12.4 29. 4.6 56.1 41.2 Zinc ....... A.%'J^ C &.vy C. . . 1828 1829 WOhler Bussey Magnesium . 1234S.G.789IUll]2fU PRODUCTION OF CAST IRON. 63 working condition. Scaffolding occurs whenever any part of the furnace-wall within the zone of fusion becomes cooled below the temperature of fusion. The stock then adheres to the furnace wall and prevents the material above from descending regularly, by forming an abutment to an arch, which is naturally formed above it, and thus sustaining the whole mass, until, the supports giving way suddenly, it lets down the stock ; and then the process repeats itself. This is usually checked by increasing the tem- perature of the blast ; but it may be less effectively overcome by other means. As coke is more bulky than anthracite, a furnace running on coke will carry but about half the stock of an anthracite furnace. Coke introduced into an anthracite furnace increases the temperature and elevates the zone of fusion. Anthracite added to coke raises the pressure by its closer packing and its pasty character at high temperatures. Tuyeres are best placed high, say 5 feet (1.5 metres) or more above the hearth, and the cinder-notch 2 feet (0.6 metres) lower, to keep the cinder well under the streams of entering air. Putting the Furnace in Blast is an operation which requires great care and considerable time ; since rapid eleva- tion of temperatures, caused by irregular expansion, would be certain to crack the lining of the furnace, and might pro- duce serious damage. The dam-stone is left out of place until the furnace is heated up. A small fire is first made in the crucible, and a gentle heat gradually dries the masonry of the interior, and warms up the furnace walls. This fire is kept up some days, and is then very gradually increased by adding fuel and small quantities of ore and flux, until, after several weeks, the furnace is filled to the mouth. As the supply of fuel is in- creased, and the furnace becomes hot enough to reduce the ore, and to melt the cinder, ore and flux are added in larger proportions, and the blast is finally turned on, and the opera- tion of smelting is thep fairly commenced. The furnace is very gradually supplied with a larger and larger proportion of ore and flux, and its " burden " is thus increased, until, 64 MATERIALS OF CONSTRUCTION— IRON AND STEEL. after some weeks, it is producing a maximum amount of iron. In some cases, the furnace is filled at once to the top, and, with a low charge of ore, started at once making iron. When the furnace goes out of blast, as it must after a run which may be a few months or may be eight or ten years, the same care is taken in cooling it down. In this case, the pro- portion of ore in the charge is gradually reduced and that of limestone increased, until, finally, the fire burns out, and the furnace stands full of burned lime. It is then left to cool, and is not opened until it has become quite cold. The Chemistry of the Process of ore reduction in the blast furnace has been carefully studied, and is becoming well understood. The reactions are too complicated and numerous for description here. They are given in works on metallurgy. The principal chemical changes' may, however be briefly stated. The charges, entering at the bell, slowly descend toward the hearth ; the air, forced into the furnace through the tuyeres, rises through the mass of material filling the shaft, and, meeting with fuel at a temperature much higher than that required to produce combustion, the oxygen unites with the carbon of the fuel to form carbonic oxide and carbonic acid. The carbon dioxide at once meets with other fuel, and surrenders to it one atom of oxygen, and two molecules of carbon monoxide are produced, and this gas rises through the superincumbent material, accompanied by all the nitrogen of the air. , Below the zone of incandescent carbon, the metal present is deoxidized, and to some extent carbonized. Above this zone the rising carbon monoxide meets the unchanged ore, and at a temperature which, while permitting deoxidation, does not fuse the iron. Here a portion of the gas takes up another atom of oxygen, thus becoming carbonic acid, and in that state passes out of the top of the furnace. The issuing gas is not entirely free from carbonic oxide. Much of the carbon monoxide escapes complete oxidation, and the furnace exhibits a gradual decrease in the pro- portions of carbonic oxide, and increase of carbonic acid, PRODUCTION OF CAST IRON. 65 from the bottom to the top. In the issuing gas, in cases cited by Percy and other authorities, the proportion of car- bonic oxide falls from about 35 or 40 per cent, at the tuyeres to 25 per cent, at the top, while the proportion of carb"c3nic acid is still more variable, but usually reaches about 12 or 15 per cent, at the furnace mouth. The proportions in a typical case were : Volumes. Nitrogen 55 Carbonic oxide 25 Carbonic acid 10 Hydrogen 6 Marsh gas 3 defiant gas I 100 The total distribution of all materials in the furnace, in- cluding gases, is given by Kent for the case already quoted, in the article above on ore-mixtures, in the table page 66. Investigations made by Akerman, Bell, Gruner, Schinz and Tunner, have yielded some valuable results. Schinz, of Strasburg, first showed it to be essential that the analyses of the waste gases should be made a basis of all conclusions as to the character and succession of phenomena of reduction. He showed experimentally that the influence of temperature, quantity of gases, proportion of carbonic oxide present, time given, and the quality of material, were all to be carefully observed, and that each had an important influence in determining reactions. He indicated that, when the precise character of the charge is known, it is possible to calculate, by analyzing the waste gases, the quantity of car- bon not burned at the tuyeres. Bell made a series of analyses of escaping gases, and con- cluded that, in the cases examined — the reduction of calcined argillaceous ores, with coke as fuel — the reduction of the ore was completed at a very low temperature, and the size and form of the pieces of ore modified the position in the furnace, and the temperature, at which the change occurs. He con- cludes that, with sufficient time to permit complete perme- ation of the ore by the reducing gases, a temperature of 637" ^ 5 66 MATERIALS OF CONSTRUCTION— IRON AND STEEL. •aaixo „ „ tn SnONVONVW " "• ■aiov ■? • : 5- ; ^ fO OINVili «■ •VNiwaiv m O i ? : a ^ fo i ^ *6 ^ C7- VC V o- m r>. M_ t^ •VISaNDVW »o i '^ •O i m fO « M t^ ^. •awn CO M M ' c t^ " " O- • e H n m m bsC • V txC CO H CO ON \0 o ; -^ xo o t- CO •vDnis fO CO *." • « * H ti 2 "s M S CA 11 PI en : tn •snaoHdsoHJt t p W ^ •anHdios y- •aaxivw r« w t^ M H ^ aiixinoA •^ Ov \0 'l T>- Ct QNV aaiVM c^ " CO c 6\ ^' •) ■ V w r> ( c o\ •waooaxiN ( , ; s 1 • £ 00 1 m f . '^ -d- ->_ ON <0 -g o t^CT. ■ o •N30AXO (^ n 1 ■ c •4- -4- c ' o> I -1 cc o 1 , . 00 " \n o »n o o Oi • ^ & in •NoaavD M ro •t^ CC jn H *" If H M '. ^D V) H X ■rf>< OO V£j t> oo q O •Noai ^ • tn ■* • w o ' s '. a OS '& a; ■*■ VC N N 8 \o «: c 'O ■> VO « VO ■saT 1 g, i o 1 C ; 8 j >2 1 * a u • H ; : & s. O ft 1 • : S! 1 5 ; u o o 4 : .9 1 1 . 3 1 § "3 s 1 1 4 ■d • 1 1 A 8 i 6 ;i3 & 5 I Q a 6o o + u 11 II II I I ou PRODUCTION OF CAST IRON. 67 to 842° Fahr. (336° to 450° Cent.) is sufficient for insuring complete reduction. Other authorities place the figures nauch higher. The Changes in the Furnace, other than those above described, are, principally, incidental reductions of compounds of silicon, sulphur, phosphorus, and other elements, the for- mation and fusion of slag, and the fusion of the pig iron pro- duced. The silicon is principally taken away as silica in the slag, with the lime, magnesia, and a small proportion of sulphur and phosphorus possibly. The remainder passes into the iron with some carbon, the greater part of the sulphur, and nearly all of the phosphorus, and with small proportions or traces of every metallic element present in the furnace charge. The issuing cinder is a silicate of lime and magnesia, with small proportions of other elements ; and the metal tapped from the furnace contains from 3 to 6 per cent, carbon, from i to 3 per cent, silicon, and minute quantities of other elements. Its precise constitution will be treated of at some length hereafter. The Specific Gravities, and Specific Heats at the boiling point of water, for materials charged into the blast furnace, are as follows : SPECIFIC HEATS AND SPECIFIC GRAVITIES. WEIGHT. MATERIAL. GRAVITY. HEAT. PER CUBIC METRE IN KILOGRAMMES. PER CUBIC FOOT IN POUNDS. Anthracite coal 1.27 to 1.92 0.201" 1,270 to 1,920 7g to 120 Bituminous coal 1.23 to 1.36 0.2009 1,230 to 1,360 77 to 85. Coke 0.76 to o.8z O.I57I 760 to 820 47 to 50 Soft charcoals 0.38 to 0.40. 0.2415 380 to 400 24 to 25 Hard charcoals 0.45 to 0.48 0.2415 450 to 480 28 to 30 Magnetic ore 5-3 to 6.0 0.1667 5,300 to 6,000 331 to 374 Red hematite 4-7 to 5.3 0.172 4,700 to 5,300 2g3 to 331 Brown hematite 3-9 to 4.0 0.154 3, goo to 4,730 243 to 250 Spathic ores, raw. . . . 3.6 to 3.9 O.I16 3,600 to 3, goo 225 to 243 Spathic ores, roasted. 4.61 to 4.73 0.16 4,610 to 4,730 288 to 295 Limestone 2.25 to 2.84 0. 1666 2,250 to 2,840 140 to 177 Lime 2.00 to 3.08 0.217 2,000 to 3,080 125 to 192 68 MATERIALS OF CONSTRUCTION— IRON AND STEEL. Knowing these quantities, it is easy to estimate tem- peratures and quantities of heat wherever definite conditions of operation can be stated. The specific heats increase slowly with rise of tempera- ture, approximately doubling with an increase of 1,652° Fahr. (800° Cent.) for coke, becoming increased four times in the same range with limestone, and increasing 50 per cent, with lime and with hard ores, and 10 per cent, with charcoal. At 3,632° Fahr. (2,000° Cent.) charcoal has a specific heat of 0.30, coke 0.50, pig metal 0.167, and slag 0.4. The Size of the Blast Furnace has an important in- fluence in determining the cost of production and the yield, as shown by Bell. The ordinary sizes of furnaces using dif- ferent fuels and hot and cold blast have already been given. The direction of change has been, for many years, in that of enlarged stacks. The largest furnaces in the world are those in the Cleve- land district, in the North Riding of Yorkshire, England, the largest having been 30 feet (9 metres) in diameter and the highest exceeding loO feet (30 metres) in altitude. It is found that neither the economy nor the yield of the furnace increases to any important extent with the increase in ca- pacity over these extreme dimensions. The largest furnace of 30 feet (10 metres) bosh, has been reduced in diameter to 27 feet (9 metres) by lining it. With the same class of fur- nace, 200 cubic feet (5.7 cubic metres) of capacity is de- manded per ton^(i,oi6 kilogrammes) of iron made per day in furnaces of 5,000 cubic feet (141.6 cubic metres) contents, while 300 cubic feet (8.5 cubic metres) are required per ton (1,016 kilogrammes) with furnaces of double this size, and 500 (14.16 cubic metres) in furnaces of the largest size named. The product of iron in the larger sizes is, at best, about equal to the weight of fuel charged. Many small furnaces use fifty per cent, more fuel. The difference in this respect is not marked between the very largest furnaces of Cleveland and those of one-half their size. In the locality named, sizes of furnaces have been reduced to about 75 feet (22.5 metres) high, and 27 feet (8.1 metres) PRODUCTION OF CAST IRON. 69 in diameter as maxima, and in other places to considerably smaller dimensions. The Height of Furnace is generally limited by the power of the materials charged to resist the pressure of the superin- cumbent mass as they approach the lower part of the furnace ; and this limitation of height also limits the diameter of the shaft, as an ex- cess in the latter dimension intro- duces a difficulty in securing a proper distribution of the ascend- ing currents of reducing gases. The proper ratio of height to maximum diameter is fixed, by usual practice, at 3 in coke furnaces and at about 4 in anthracite furnaces. After reaching a certain altitude, also, no useful gain is secured by this transfer of heat from the gases to the material in the upper part of the stack. Temperature of Furnace. — Bell presents the adjacent figure as illustrating the distribution of work and adjustment of tempera- tures in the blast furnace ; the tem- perature falling as the rising gases flow through the successive zones of fusion of the reduced metal, of absorption of carbon, calcining of limestone, and of reduction of ore, from a white to a dull red heat, and finally issue still hot and pass off to the stove. The gases issuing from the top of furnaces having an exceptionally high temperature of blast, are cooler than those issuing from furnaces having a colder blast. This fact also Fig. 6. — Temper atukes of Furnace. 7° MATERIALS OF CONSTRUCTION— IRON AND STEEL. assists in producing a limit of temperature. The fact is due to the reception of the larger proportion of heat in the former furnace, from its blast, and the less proportion from combus- tion. Combustion also involves less gas, and the smaller vol- ume is more completely cooled in rising through the furnace. The minimum temperature of gases tends to remain constant at about 392° Fahr. (200° Cent.), according to Gruner, in con- sequence of the regulating effect of the dissociation of the carbonic oxide, which can only occur above a fixed limit. The maximum temperature of blast, with even fire-brick stoves, may be taken at about the higher figure above given. The accompanying graphic representation. Fig. 7, is Economy of Fuel from increased Temperature of Blast. 1200 1300 UOO 1500 ICOO Temperature of Blast Fig. 7.— Economy of High Temperatures. given by Bell, illustrating the gradual approxirhation to a maximum of the benefit derivable from increasing tempera- ture of blast. It represents the consumption of coke per ton (1,016 kilo- grammes) of iron made in a furnace of 25,624 cubic feet (724.7 cubic metres) capacity. The figures corresponding to 800°, 925", 117s", 1275°, and 1425° Fahr. (450°, 490°, 591°, 630°, and 768° Cent.) are observed ; all others are defer- mined by plotting the curve. The benefit of increasing the temperature of blast is least in large furnaces running on good ores. The Iron tapped from the Furnace is led by properly PRODUCTION OF CAST IRON. /I arranged channels to the " pig bed." This is usually a con- siderable area of sand leveled off and scored longitudinally and transversely to form moulds. The main channel, called the " sow," of each section, has on each side smaller channels of about four feet (1.22 metres) length, in each of which a " pig " is cast. The whole arrangement resembles somewhat that of a gridiron. When forge iron of high grade is to be made, the pig metal is often cast in iron moulds instead of in sand, to avoid the introduction of silicon. The pig bed is covered by a roof to protect it from the weather. This " casting-house " is built close against the stack of the furnace. When the metal has cooled in the pig bed, the pigs are broken from the sow and are stacked in the yard, or are sent off to market after they have been graded, numbered, and distributed into lots of similar quality. The sow is some- times also sent to market after it has been broken into pieces of proper size to handle ; in other cases it is charged into the furnace and remelted. The iron having been removed, the pig bed is made up anew ready for the next cast. The casting-house is usually built of substantial and fire- proof materials. Brick walls and an iron roof are adopted when the expense can be met without serious inconvenience. The gases are taken from the top of the furnace through a sheet-iron pipe leading out at the side, under the charging floor, and led to the hot-blast stoves and to the steam boilers. The gas main is lined with fire-brick to prevent escape of heat and to prevent rapid oxidation of the iron pipe at the comparatively high temperature which would be given it by direct contact with the heated gases. The gas main is sometimes carried some distance a little above or below ground to the stoves. The Hot-blast Stoves or ovens are of either iron or brick. The former consist of sets of cast-iron pipes of various forms and variously arranged in different cases, inclosed in large chambers lined with fire-brick. The blast is driven through the pipes, which are kept heated by the burning 72 MA TERIALS OF CONSTRUCTION— IRON AND STEEL. gases from the furnace, which latter are inflamed in the in- closing chamber. In some forms of stove the gases are burned in a " com- bustion chamber," and the heated products of combustion rise into an upper chamber containing the set of pipes carrying the blast. This system is claimed to possess the advantage of giving greater uniformity of heat and less danger of crack- ing the pipes by sudden and great changes of temperature. In other forms, the gas is burned in contact with the pipes, and but one chamber is used. The forms of pipe adopted are very numerous. In some ovens, mains are led across the chamber parallel with its sides. The blast enters one main and passes over into the other through a set of J^-shaped pipes, emerging from the second main, whence it is conducted to the tuyeres. The intermediate pipes are usually of elliptical or oval section, having the longer diameter in line with the mains. This form permits expansion and contraction with change of tem- perature to take place with little danger of frequent fracture of the pipes. In other forms of cast-iron pipes, they are divided by a diaphragm into two parts. The air is led into the main, rises into one of these chambers, returns through the other to the main, which it traverses until reaching the next pipe, it enters that, and thus passes from pipe to pipe until it emerges at the extremity of 'the main fully heated. In such stoves, several mains are laid down, each canying a set of these double pipes. A modification of this form of pipe which has been found Fig. 8.—" Pistol-Pipes." PRODUCTION OF CAST IRON. 73 to work well is that known as the " pistol-pipe," in which the upper extremity is enlarged, giving the pipe the shape of a pistol erected muzzle downward, the well at the top of the pipe representing the butt, Fig. 8. This form permits a reduced velocity of gas at the top of the pipe, and gives an increased area of heating surface ; opposite pipes have their tops turned toward each other, forming an arch above the chamber through which the hot products of combustion are rising, and the whole makes a very efficient form of stove. A recent form of stove is fitted with iron pipes suspended from above, instead of being supported from beneath. It is stated that this modification increases the durability of iron pipes very greatly. The least area of heating surface required to give the maxi- mum temperature permanently sustainable by cast-iron pipes is considered by some engineers to be from i,ooo to 1,200 square feet per 1,000 cubic feet (3.26 to 3.9 square metres' per cubic metre) of air passing per minute ; the proportion is often much less. A pair of furnaces having three blowing engines of 84 inches (213 centimetres) diameter of blowing cylinders and 5 feet (4.57 metres) stroke of piston and making 20 to 30 double strokes per minute, when fitted up with twelve stoves, each containing 14 double pipes 16 feet (14.16 metres) long, 19 inches (48.26 centimetres) wide, and 5 inches (12.7 centimetres) deep, received the blast at a temperature of 1,130° Fahr. (610° Cent.). In this example each stove had a separate chimney. The limit of temperature with iron stoves is so low that, at many furnaces, stoves are now built of fire-brick through- out, including heating surfaces. These are comparatively expensive, but they have been used with a blast heated to 1,382° Fahr. (750° Cent.), and even 1,742 Fahr. (950° Cent.) has been attained at times. When these stoves are con- structed in such form that they are not liable to become choked with the dust carried over with the combustible gases, they are found to give excellent results. These stoves are usually constructed upon the principlie of the regenerative, or the fire-brick furnace, as, for example, in the stoves of 74 MATERIALS OF CONSTRUCTION— IRON AND STEEL. Cowper and Whitwell. In the Sellers regenerator, the action is continuous, as in the cast-iron stoves, and the structure is, like them, composed of pipes. The material of the pipe is a refractory clay. g. — Hot-Blast Stovk Elevation in Section. jSW ' H» ' it* ' - <'i <(Hf il i *tai?S^P Arrows denote direction of Gas *^" ^ - - jf* T* V^ ■' and Air for combustion wliea being heated. /^_. Ciifijorf Position of the Valves show the uas i^uivert. course of the Air when on blast. Chimney Flv*. The fire-brick stoves must usually be given two or three times greater area of heating surface than cast-iron stoves. The weight of brick used is about one ton (i,io6 kilogrammes) PRODUCTION OF CAST IRON: 75 for each 20 square feet (1.8 square metres) of heating sur- face. The Whitwell Stove (Figs. 9, 10) is a modification of the regenerative apparatus used in gas furnaces, and its Fig. 10.— Anthracite Blast Furnace, Gas Tubing, Hot-Blast Main, AND Stoves. method of operation is very similar. They are stated to bear the highest attainable temperature, to be free from lia- ;6 MATERIALS OF CONSTRUCTION— IRON AND STEEL. bility to wear and tear, easily cleaned even y/hen hot, free from leakage, to produce no loss of pressure of blast, to be very efficient regulators of temperatures, and to secure great economy over iron stoves. For a production of 600 tons (609,600 kilogrammes) per week, two of these stoves are fitted, each of about 3,000 cubic feet (84.945 metres) volume and of 2,000 to 2,500 square feet (185.8 to 232.25 square metres) of heating surface. Blowing Machinery. — The blast is forced into the blast-mains by blowing machinery driven either by water or steam ; the latter is the usual motor. The steam cylinder and blowing cylinder are generally parts of one machine, which is called the blowing-engine. In the most common, and in some respects best, form the cylin- ders are vertical, the air cylinder above the steam cylinder, and the pistons have a common rod. A large fly-wheel is used to insure a uniform motion. The ratio of piston area is determined by the relative mean pressures on the two pis- tons. A blowing-engine having steam cylinders 45 inches (143 centimetres) diameter and air cylinders 72 inches (183 centimetres) diameter, making 14 revolutions per minute, took in about 200,000 cubic feet (5,632 metres) of air per hour and delivered it under a pressure, at the tuyeres, of 4 pounds per square inch (0.28 kilogrammes per square centi. metre), drawing steam from four plain cylinder steam boilers 5 feet (1.5 metres) in diameter each and 36 feet (10.8 metres) long. The furnace was 66 feet (20 metres) high, 17 feet (5.1 metres) in diameter of bosh, had 5 tuyeres 6 inches (15 centi- metres) diameter inside supplied with a blast heated to about 932° Fahr., (500° Cent.). The stoves were of cast iron, three in number, and contained, each, 24 pipes 14 feet (4.2 metres) long, and 9 inches (22^^ centimetres) internal diameter. The speed of piston is usually, in Pennsylvania, not far from 225 feet (67.5 metres) per minute but rises to 300. Blowing engines are often built with a beam mounted on columns and linked to the piston of a vertical steam cylinder at one end, and to that of a blowing cylinder at the other. Horizontal blowing engines are also sometimes built.\V^ «/^ PRODUCTION OF CAST IRON. 77 The Winding-Drum, when used, is usually turned by a pair of small, quick-working steam-engines. The rope is about y^ inch (1.9 centimetres) in diameter, of steel or iron wire. A brake controls the drum and gives the attendant control of the platform when descending. All parts should be given a large factor of safety. An inclined plane was formerly very frequently built for the hoistway, and the charges were raised in wagons running on rails. The Water-Bucket Hoist consists of a set of timber sup- ports and guides, at the top of which are pulleys carrying the rope or chain, which, at one end, is attached to the traveling platform or cage of the hoist, and at the other is secured to a large " bucket." Both bucket and platform are guided by the timber frame of the hoist. The bucket is alternately filled and emptied, receiving its water from a reservoir at the level of the charging floor and discharging its contents into another reservoir, or into a tail-race at the bottom of the hoist. The motion of bucket and cage is controlled by a brake, and latches at top and bottom hold them when they are to be kept stationary. The bucket and cage may be made of either wood or iron, the latter being preferred. Water is supplied from some natural or artificial source, by gravity or by force pumps, or through hydrant pipes. The supply-cock of the reservoir and the discharging-cock of the bucket are usually opened and closed by hand. When the bucket is at the top, the cage is at the bottom of the hoist. When filled with water, the former overbalances the weight of the latter and its load, and, being unlatched, descends, pulling up the loaded platform with which it is connected by the rope or chain passing over the pulleys at the top. When the charges have been thrown into the furnace, the unloaded barrows are wheeled back upon the platform. Meantime the bucket has been emptied, and the weight of the cage now preponderating, it descends, raising the empty bucket to the top. This apparatus is simple in construction and durable, but 78 MATERIALS OF CONSTRUCTION— IRON AND STEEL. it is heavy, slow in operation, and quite bulky. It is much less used than formerly. From 20 to 50 per cent, of the whole energy of the water supply is wasted by friction and loss of head. The Plunger Hoist, or water-pressure hoist, is a hydraulic press which either carries the platform directly or raises it through the intervention of pulleys and tackles. In the first, a strong hollow iron cylinder, of a length somewhat exceed- ing the height through which the platform is to be raised, is sunk into the earth in a vertical position. Water, under a pressure of sometimes 300 or 400 pounds per square inch (21 to 28 kilogrammes per square centimetre), is led from forcing pumps, or from an " accumulator," into this cylinder, and forces up a " plunger," which is fitted to the latter, and which carries the platform. The lower end of the cylinder is closed, the upper end is fitted with a stuffing-box, or a collar packing which prevents leakage as the plunger slides vertically through it. The load at the plunger is equal to the weight of useful maximum load added to the weight of plunger and platform plus the frictional resistance to sliding, which varies somewhat, but which may be taken at one-tenth. Calling W = the maximum load, and W = the weight of the moving parts of the hoist in pounds, d — the diameter of the plunger in inches, and/ = the available pressure of water in the cylin- der of the press in pounds per square inch, d= \ i.i -— r-- — = 1-2 \/ . nearly. The accumulator consists of a heavily weighted plunger of considerable volume rising and falling under its load in a cylinder like that of the hydraulic press. A set of pumps driven continuously forces water into the accumulator, while it is drawn out intermittently by the working apparatus to which the water is supplied. This accumulator, or store- cylinder, must have such volume that it shall not be exhausted completely at any time, and its plunger must be loaded with the weight needed to preserve the maximum pressure desired. PRODUCTION OF CAST IRON. 79 By its use, small pumps and a small prime-motor acting con- tinuously are enabled to supply water, which is drawn by the hoist in comparatively large volumes and intermittently, thus securing economy of maintenance, and, usually, of first cost. The accumulator is also useful as a safety-valve. The connecting pipes should be made as large as is con- sistent with economy of cost, to reduce frictional losses. The velocity of the hoist is variable ; one foot (0.3 metre) per second is a speed sometimes adopted. The expenditure of power at the pumps is frequently one-half greater than that usefully applied by the hoist. An empirical formula for thickness of water pipes, used by some engineers, is /= o.oooo^'iHd + 0.5, in which d is the diameter, t the thickness in inches, and H is the total maximum head of water in feet ; for thickness and diameter in centimetres and head in metres we have t= o.OOiSHd + 1.5, nearly. Pipes with sockets are generally used, although flanged pipes are common. The thickness of hydraulic press cylinders, and of pipes, also, may be taken by Barlow's formula, which gives an excess of strength : /-/ in which ( = the thickness in inches, r = the internal radius in inches, / = the assumed bursting pressure in pounds per square inch, and/= the tenacity of the material in pounds per square inch. For radius and thickness in centimetres, strength and pressure in kilogrammes per square centimetre, we have / - ^ -^ 8o MA TERIALS OF CONST RUCTION— IRON AND STEEL. The form of hydraulic hoist with which pulleys are used requires a shorter plunger, and one of larger area than that just described, the two dimensions varying in inverse ratio and proportionally with the velocity-ratio of the pulley com- bination. A small additional frictional resistance must be allowed for. This form is of less first coat. The pressures used are limited by circumstances. The maximum, as determined by the ultimate safe pressures for the material used, are, in hydraulic presses, about 4 tons (630 kilogrammes per square centimetre) for ordinary cast iron, 6 tons (940 kilogrammes per square centimetre) when lined with copper, and from 7 to lotons (1,090 to 1,560 kilogrammes per square centimetre) when made of wrought iron or steel. The Pneumatic, or Air Hoist, consists of a cylinder of proper size traversed by a piston, which is connected by ropes, carried over large pulleys, to the platform of the hoist. By a pressure exceeding that of the atmosphere, on the one side, or sometimes by the creation of a partial vacuum on the other side, the piston is caused to move through the cylinder, raising the load. The cylinder must be nicely bored, and the piston well fitted and carefully packed. The pressure adopted is usually that of the blast of the furnace, and the air is, in such cases, supplied by the main blowing engines. The diameter of the piston may be calculated as for the hydraulic hoist. The efficiency of this hoist is greater. In one of the best designs of this form of hoist the working cylinder stands in the middle of the hoistway, and there is a platform on each side, both of which move together. In other forms the cylinder has the platform built around it. In some cases, hoists are constructed having two platforms or cages so arranged that, while one is ascending the other is descending, and vice versa, balancing each other. For a furnace using i,ooo tons (1,016,000 kilogrammes) of material per week, the elevator, or hoist, would be calculated to carry two barrows of 500 pounds (227 kilogrammes) each, or about 1,000 pounds (454 kilogrammes) total, and the cylinder would be about 30 inches (762 centimetres) in diameter, as usually designed. The platform may be calcu- PRODUCTION OF CAST IRON. 8 1 lated to rise with greater velocity than that of the hydraulic hoist. This is the most generally approved form of furnace hoist. The Steam Hoist is of similar form, and is worked by steam taken from the boilers supplying the blowing and pumping engines. In this class the piston often forms a counterbalance to the platform, and is, if necessary, weighted. The Water Supply of the blast-furnace is an im- portant detail. Water is required for the tuyeres, for the steam boilers and the condenser, and, frequently, for the hydraulic hoist and other minor accessories. It is necessary to secure such a supply that the furnace may not be interrupted in the dryest seasons. The required head is sometimes secured by a natural fall, sometimes by direct pumping, and sometimes by means of a large reservoir, at the required elevation, which is kept filled by forcing pumps. The water should be as pure as possible to avoid injury to boilers and to tuyeres by the formation of incrustation. Salts of lime are the most common impurities. They are removed to a greater or less extent, frequently, by heating, or by the use of chemicals, which render the precipitate pul- verulent and readily removed, or which produce solutions which may be removed by occasional " blowing out " of the boiler, and which do not precipitate insoluble " scale." The Steam Boilers are placed as near the engines as possible. The type may be, to some extent, a matter of choice, but they are usually of plain cylindrical form, set in brick-work, and fitted both with grates for use with solid fuel and with chambers and supply-conduits for gas from the furnace-top. The extent of heating surface is determined by the quan- tity of steam required by the pumps and blowing engines. This amount is variable, but may be taken with ordinarily good machinery, as equivalent to about 35 pounds of water evaporated per hour per horse-power (15.64 kilogrammes per cheval). As a minimum, a square foot of heating surface for each six pounds of water (or about one square metre for each 30 kilogrammes) per hour may be given for solid fuel. With 6 82 MATERIALS OF CONSTRUCTION— IRON AND STEEL. gas, the lower temperature of fire usually compels the use of more boiler surface. One square foot of heating-surface to each two pounds (one square metre for each lo kilogrammes) of water per hour is not unusual. S/^ ¥^ Cast Iron is the name given to the product of the blast furnace. It consists of metallic iron chemically united with carbon in proportions varying from two to nearly six per cent., silicon to the amount of sometimes five per cent., and usually manganese, phosphorus, and sulphur, in smaller proportions. Foundry irons also contain carbon in two forms, chemically united, forming a carbide, and mechanic- ally mingled with the metal in the form of graphite. Minute quantities of calcium and other substances are also found in it. Analyses will be given hereafter. The cast iron, when removed from the casting house, is assorted and sent to market in several grades. The darkest kinds of metal, which contain most carbon, are called foundry- pig, and the lighter grades forge-pig. The classification is usually as No. i Foundry, No. 2 Foundry, No. 3, or Gray Forge Iron, Mottled Iron, and White Iron. The darker grades are used for castings, and the lighter for the manufacture of wrought iron. The characteristics of the several grades of iron are thus summarized : Foundry Irons. — No. i (Dark Gray Iron) : Fracture dark gray, with high metallic lustre. Crystals large, with lustre resembling surface of fresh-cut lead. Makes fine castings ; fuses easily ; flows freely, is soft and rather ductile. . No. 2 (Gray Iron) : Fracture gray ; lustre clearly metallic ; crystals smaller than preceding; a freely melting, free flow- ing, and moderately strong iron. No. 3 (Light Gray Iron): Fracture light gray; crystals small ; lustre dull ; crystals larger and brighter near cent^ ; makes best material for large castings. Forge Irons. — No. 4 (Bright Iron) : Fracture light gray ; crystals small ; lustre slight ; too infusible and pasty for foundry use ; makes good mill iron. No. 5 (Mottled Iron); Fracture dull, silvery white; line REDUCTION OF IRON ORES. 83 of whiter iron around edge of fracture ; speckled with gray within ; hard, brittle, but sometimes a good forge iron. No. 6 (White) : Fracture silver white ; often bright ; granu- lated texture, with radiating lines of crystallization ; extremely hard and brittle ; useless except for low grade puddled iron. The properties of cast iron will form the subject of a suc- ceeding chapter. The Bloomary, or Catalan Process. — The reduction of ores of iron is sometimes practiced by other methods than that already described. The most common is that known as the Bloomary, or the Catalan Forge Process. It is practiced in Spain, as indicated by its name, in Sweden and Germany, and perhaps other parts of Europe, and, in a rude way, in Asia and Africa, as already described page 46. The product is wrought iron or steel in masses called " blooms," which vary in size with the size of furnace, up to 300 or 350 lbs. (136 to 159 kilos) in weight. The process can be made commercially successful in dis- tricts in which very rich ores and abundance of wood for charcoal can be obtained and at low prices. , The furnace usually consists of an open hearth of about 28 inches (70 centimetres) depth to rear wall, and 30 inches (76 centimetres) width, and with tuyeres inserted 2 feet (0.6 metres) below the level of the top of the mass of fuel. The casing is of cast iron, double and supplied with water, to keep it from becoming overheated. Above the hearth a stack is built to carry away the products of combustion. The hearth is open at the front like an ordinary open fire- place. The blast is supplied under a pressure of from l^ to 2 pounds per square inch (o.i i to 0.14 kilogrammes per square centimetre), and heated to a temperature rarely if ever meas- ured, but generally supposed to be 600° to 800° Fahr. (316° to 426° Cent.). The heating pipes are siphon tubes placed in the stack. The tuyeres are either pointed horizontally or slightly inclined downward, and have a segmental opening for the better distribution of the blast. In working ores by this process, the furnace is filled with 84 MATERIALS OF CONSTRUCTION -IRON AND STEEL. charcoal, the fire lighted, and the blast turned on. When the whole is well ignited, the ore, calcined and coarsely pulver- ized under stamps or breakers, is sprinkled with a shovel over the surface of the mass of fuel in small quantities, and at short intervals basketfuls of charcoal are added as the fire burns down. The ore is deoxidized by the carbon of the fuel as it works downward, and the metal. finally aggregates in an unfused pasty mass of agglutinated grains at the bottom of the hearth, like a great sponge. The cinder fills its pores and surrounds it as a liquid bath, and is tapped off occasionally at the front. A "loup " weighing about 300 pounds (136 kilogrammes) is formed in about three hours. This is lifted out from under the mass of fuel and is worked under a hammer of about 5,000 pounds {2,272 kilogrammes) weight into a billet or bloom, being reheated, when necessary, at the bloomary fire. The men work in two " shifts " of twelve hours each, and each fire is expected to yield from one ton to 2,500 pounds (1,136 kilogrammes) per day. The amount of fuel used varies from 200 to 300 bushels (5,664 to 8,496 litres), 3,500 to 5,000 lbs., per ton (i^^ to 2^ kilogrammes per kilogramme), according to the skill of the iron-maker and the quality of ore and fuel. The total cost of the bloom is not far from that of puddling iron. One ton (1,016 kilogrammes) of finished metal requires from i^ to i^ tons (1,524 to 1,778 kilo- grammes) of selected ore, which is equivalent to from 2}^ to 4 tons (2,540 to 4,064 kilogrammes) of ore as mined. The American Bloomary Process, for making iron direct from the ore, is a modification of the old German proc- ess, although it is in many places incorrectly spoken of as the Catalan Process. Like the Catalan, it is adapted to rich ores of iron that are free from all impurities save gangue, which, before entering the furnace, must be removed as far as pos- sible. Ores to be profitably worked by this process usually contain above 90 per cent, of magnetic oxide of iron. The ore is by this method roasted, crushed, and then subjected to this process. The furnace is composed of cast-iron plates 2 or 3 inches (s to 7.5 centimetres) thick joined together, form- PRODUCTION OF CAST IRON. 85 ing an open box, which, at the base, is from 24 to 30 inches in its length (61 to 76 centimetres) at right angles to the tuyere, while the dimensions are 27 to 32 inches (68 to 81 cm.) laterally. In the rear, parallel to the tuyere, it is from 28 to 36 inches (71 to 91 cm.) deep. In front, however, it is from 15 to 19 inches (38 to 48 centimetres), to make room for the " fore plate." This rectangular space is known as the " fire box," and it is here that the reduction takes place. The air pressure is from i^ to 2 pounds per square inch (o.ii too.14 kilogramme per square centimetre). The stack, through which the products of combustion and gases pass, is of rec- tangular section and of sufficient size to receive the whole furnace under it. This stack is about 20 feet high (6.1 metres). The higher the temperature of the blast the less fuel con- sumed. It ordinarily varies from 600° to 800° Fahr. (316° to 426° Cent.), but it has been found that by raising the tempera- ture of the blast, the tendency for the impurities present to enter into the iron is'increased. One " P "-shaped tuyere is used, made of J^-inch (1.27 centimetres) wrought iron. The nozzle is about a foot (18 centimetres) long, and is inclined at an angle of about 15". If this angle is too low, the capacity of the furnace is dimin- ished by the coal forming a crust on the bottom ; if it is too high the blast cuts through the loup. The ordinary cost of a furnace such as is described here is about $600. The remainder of the process closely resembles the Catalan. The amount of fuel used is from 300 to 350 bushels (8,496 to 9,912 litres) of charcoal per ton of iron. A ton is said to have been produced with an expenditure of 240 bushels (6,796 litres). The production of a furnace of the size described averages i ton (1,016 kilogrammes) in 24 hours, or about 300 tons (304,800 kilogrammes) in a year. The Siemens Process of reduction of ore, or " Di- rect Process," as this method is termed, is one which has attracted much attention, but one which is not yet generally introduced. In this process the ore and flux are fused together in the reducing flame of the regenerative furnace, 86 MA TERIALS OF CONSTRUCTION— IRON AND STEEL. and the . cinder is tapped off at intervals, leaving, finally, the molten iron on the hearth, to be drawn off into ingot moulds. The process occupies four or five hours, and the product con- sists of four or five tons of wrought iron or steel. In the latest modification of this process Mr. Siemens avoids the serious difficulty attending the reduction of an ore on the hearth of the reverberatory furnace by effecting the change in a rotating cylinder similar to the rotating puddling furnace, to be hereafter described, and by adopting a peculiar composition for the lining, y^ CHAPTER IV. THE MANUFACTURE OF WROUGHT OR MALLEABLE IRON. Wrought Iron is distinguished from cast iron, chem- ically, by its comparative freedom from carbon, silicon, and other elements which enter into the composition of the product of the blast furnace to such an extent as to form an important part of the latter material, and by its greater strength, ductility, and homogeneousness. It has immensely greater value as a material of construction. Its peculiar prop- erties will be considered at length in a succeeding chapter devoted to that subject. It may be manufactured by the direct reduction of the ore, as in the bloomary, the Siemens and the other " direct " proc- esses already described ; but by far the greater part of the wrought iron which appears in the market is made from cast iron by the removal of carbon, silicon, and impurities by the process of refining and puddling, and is worked into marketable shape by rolling or by hammering. Very large quantities of a metal which resembles wrought iron closely in chemical composition and in mechanical prop- erties — and which is properly classed with malleable or wrought iron — is made by the pneumatic, or Bessemer proc- ess, and by the Siemens-Martin process, and sold in the market as " low steel," " Bessemer steel," or " Siemens-Martin steel," or, as lately proposed, under the name of " ingot iron." These processes of manufacture will be described in a chapter on the manufacture of steel. The Decarbonizing Process consists in the subject- ing of molten cast iron to the oxidizing flame of a reverber- atory furnace until the carbon has been burned out and the metal is sufficiently pure to become pasty, and to cohere in spongy masses at the maximum temperature of the furnace. 00 MATERIALS OF CONSTRUCTION— IRON AND STEEL. For these processes, the lighter grades of cast iron are selected as containing least carbon, and therefore demanding less labor, and as they are cheaper than the dark, foundry grades. Cast irons containing sulphur and phosphorus are less val- uable than irons free from these elements, as the former yield a malleable iron which is brittle and difficult to work and to weld at high temperatures, and the latter make the product brittle and non-ductile when cold. Manganese, from its chemical relations as an antidote to sulphur, is a desirable ingredient. All other foreign substances are undesirable. The carbon and silicon are removed during the process of con- version ; the sulphur is partly driven off, as is manganese ; the phosphorus is retained in the iron. The earliest processes of making wrought iron were, as already stated, direct processes. The earliest process of re- duction of cast iron, and that which was practiced at the time of the invention of puddling by Cort, is still practiced, and is known as the Refinery or Forge Process. The Forge Process is, in method of working, simi- lar to the bloomary process, and the forge fire is constructed very much as is the bloomary. Instead, however, of reducing ore by expelling its oxygen in presence of an excess of carbon, the forge process burns out carbon from cast iron in presence of an excess of oxygen. As practiced in the United States, where the process is adopted to a limited extent in making blooms and billets to be worked into boiler-plate, it consists in melting down pig- iron on a shallow hearth under a blast until about 250 pounds (113 kilogrammes) is collected under the tuyeres. The molten mass is stirred with an iron rabble and the blast kept on it until, the carbon having been burned out, the iron becomes pasty and adherent, and can be v/orked into a ball. The cinder which collects as the impurities are worked out, is now and then tapped off. When steel is made by this process the cinder is retained, and the ball is worked in a bath of the molten slag. When ready to " ball up," the temperature of the fire is raised, the metal worked over to free it from cin- WROUGHT OR MALLEABLE IRON. 89 ders, and then balled up and removed from the fire to be worked into billets or blooms. One fire worked twelve or thirteen hours per day by a single shift of hands, produces five or six "loups" weighing about 200 pounds (91 kilogrammes) each. If the iron has been refined previously, as described in the succeeding article, the production is sometimes doubled. The consumption of pig-iron and of charcoal in this process is about 1,800 and 2,400 pounds (817 and 1,090 kilos) respectively. The greater number of forges of this kind in the United States are in Pennsylvania. The process is adopted, to some extent, in Sweden, Germany, and other parts of Europe. In a large number of cases, establishments started as bloomaries have been changed into forges for the reduction of malleable from cast iron. Refineries are forges in which the process just de- scribed is interrupted when but a portion of the carbon and other oxidizable substances are removed from the cast iron. The refinery is usually larger than the forge above de- scribed, measuring 3^ feet (1.06 metres) wide to the back, 5^ feet (1.67 metres) long, and its hearth has a depth of a foot or eighteen inches (0.304 or 0.456 metre). From one to two tons (1,016 to 2,032 kilogrammes) of pig-iron can be melted down and retained in it. The large size of the hearth compels the use of four or more tuyeres. The metal is sometimes run into the refinery from the blast furnace, and sometimes charged in pigs and melted in the forge. The metal, subjected to the action of the blast, " boils," and gradually loses carbon, and is finally tapped off on the casting-floor, or into moulds, in which it assumes the form of flat plates about ten feet (3.04 metres) long, three feet (.91 metre) wide, and two inches (5.08 centimetres) thick. These plates are broken up and used in the forge or in the puddling furnace. Each charge requires about two hours for complete re- fining. The loss of iron is from S to 20 per cent., according to the skill of the workmen. The usual loss is about ten per cent. 90 MA TERIALS OF CONSTRUCTION— IRON AND STEEL. The expenditure of fuel is about one part to five parts of iron in good work. This " finery furnace " is also called a " run- ning-out fire." One fire will refine from lo to 20 tons (10,160 to 20,320 kilogrammes) of metal per day. Either coke or charcoal may be used as fuel. The " fine metal " is a white cast iron, from which nearly all silicon, a large part of its carbon, and much of its manga- nese and sulphur, as well as some phosphorus, have been removed. The slag, which contains those substances which are not volatilized, is a silicate of iron containing about 53j^ per cent, iron oxide, and 14}^ per cent, silicon, and 32 per cent, oxygen, the formation of the silicate involving a serious loss of iron. Puddling and Boiling are modifications of the same process, and in both the refining, as already described, is car- ried on until the character of the metal is entirely changed, and the product is obtained which is known as malleable or wrought iron. In the manufacture of malleable iron by these processes, the metal is melted as in refining, but the fusion takes place on the hearth of a reverberatory furnace in which the metal only comes in contact with the gaseous products of combus- tion, and is thus less exposed to contamination by the deleterious elements found in fuel, and the necessity of a powerful blast to secure a supply of air is avoided. Draught is secured by a moderately tall chimney, or by a fan-blower, and controlled in the former case by a damper on the chim- ney-top. While the metal lies in the molten state on the hearth the puddler stirs it with an iron rabble, and thus brings every portion in contact with the decarbonizing flame. This- oxi- dizing action of the air is seconded by the presence of fluxes rich in oxygen, such as magnetic hematite ores, or scales from the blacksmith's forge. Slag and cinder, rich in iron oxidfe, are sometimes used where they can be obtained free from sulphur and phosphorus, or where the puddled iron is of a cheap grade. In " dry-puddling " the puddler relies upon the action of WROUGHT OR MALLEABLE IRON. 9» Fig. II. — Puddling Furnace — Vertical Section. the air principally ; in " wet-puddling " the work is largely done by the oxides used in " fettling." The first method is usually called, simply, puddling, the latter is often known as " boiling," or as " pig boiling." The form of reverberatory furnace ordinarily used in the puddling process, is illustrated by Fig. ii. The hearth, A, is made, usually, of plates of cast iron, carried on brick walls or on short iron pillars, bb. It is usually about five or six feet (1.5 to 1.8 metres) in length, and four feet in width, oppo- site the charging door. This hearth is covered thoroughly with slag, or with a " fettling " of iron oxide, which is melted down upon it to protect the plates from the solvent and cor- roding action of the charge. The fire is built, usually, of bituminous coal, in the fire- place, B. The grates are generally single square rods of iron separately detachable. They can be removed singly to clean any part of the fire, or replaced when any, one of them is burned out or droops under excessive heat. The area of grate is usually six square feet (0.56 square metre), and sometimes ten feet (0.93 square metre), or more, its precise dimensions being determined by the character of the metal used, of draught, and of fuel. Between the grate and the hearth is a fire-brick wall, or " bridge-wall," C, ex- tending from side to side, rising sufficiently high to prevent any portion of the fuel passing over on the hearth or any molten iron falling over upon the fuel. Resting upon and extending around the sides of the hearth is a box, or a double wall of cast iron, eight or ten inches (20 to 25 centimetres) high, through which hollow box 92 MA TERIALS OF CONSTRUCTION— IRON AND STEEL. water circulates, preventing its fusion. The iron bottom is also sometimes double and similarly kept cool. Air is some- times substituted for water as a cooling medium. The sides, like the bottom, are carefully protected by a coating of slag and ore laid on under so high a temperature that it may be readily moulded. At the end nearest the chimney, the hearth terminates at a second bridge or " altar," D, which prevents the overflow of the molten metal at that end. Beyond this bridge-wall, the furnace flue inclines downward and terminates at the chim- ney flue, E, the cross-section of which is usually given 20 per cent, of the area of the fire-grate. At the foot of the incline is an opening out of which the molten slag passes after leav- ing the hearth and overflowing at the altar. On the side of the furnace, opposite the middle of the hearth, is the working door. It is about 20 inches (51 centimetres) square, and is closed by a slide lined with fire-brick, which is arranged to rise and fall vertically, and is counterbalanced. A small opefting at the bottom of the door, large enough to admit the puddler's rabble, permits the workman to stir the charge without serious discomfort, and without admitting cold air to chill the furnace and check the draught. The roof of the furnace is an arch of fire-brick, about two feet (0.6 metre) high at the fire-place, and sloping gradually, until, at the chimney, it is less than a foot above the bottom of the flue. Outside the fuS-nace, at the opening at the foot of the chimney^the floss-hole — a fire is maintained to keep the escaping cinder fluid until it has fairly left the furnace. In some furnaces a charging-door is placed near the altar, through which the pig-metal is placed " on the bank" to be melted down. The original process, as practiced by Cort, the inventor, was that of dry puddling. He made up his furnace-bottom with quartz sand, and used iron containing little carbon, removing that remaining by the oxidizing action of the flame alone. Refined iron, and the white and mottled grades, are generally used in this process, as they promptly become WROUGHT OR MALLEABLE IRON. 93 decarbonized, assume a pasty condition, and can then be balled up. The cinder-bath is produced by the union of the silica present with iron-oxide. The loss of metal is variable, but may be taken at an average of nearly ten per cent. In the time of Cort a furnace could be made to produce but about ten tons (10,160 kilogrammes) per week. Subsequent improvements, including the ventilated bottom, doubled this quantity. In the Boiling Process, or that of " wet-puddling," the furnace is made rather deeper than for puddling. Rich cin- der is used for fettling, and is charged into the furnace, and mingles with the pig-iron, its oxygen taking up the carbon of the iron, and thus hastening the process of decarbonization, the melting and molten iron lying in a bath of fluid cinder. Unrefined iron can be worked by this method, and instead of losing iron, if the process is well managed, some gain of weight is made by reducing the oxide charged, and the fet- tling of the furnace. The first process, often called simply puddling, is in use in making low grades of iron from white pig-metal. In the boiling process. No. 3 pig-metal is gener- ally used, but sometimes gray pig, and in other cases, refined iron. The charge of the puddling furnace is about five hundred pounds (227 kilogrammes) of pig-metal, which is laid care- fully on the bed of the furnace, or broken up and piled on the bank and around the sides. In the boiling process, the necessary amount, 100 pounds (45 kilogrammes), more or less, of cinder, or ore, and hammer-scale is added. The door is then closed, and the damper opened wide, and the charge is melted down, the puddler moving the pieces among each other to secure a regular and not too rapid fusion, and to give the flame free access to the metal. Fusion commences in fifteen or twenty minutes, and in a half hour the charge lies in a molten pool on the hearth, and assumes a pasty condition. The puddler stirs the fluid mass with his rabble, checking the draught to give more time for completing the chemical reactions, and even chills the metal by throwing water upon it. The heat is again increased, and <"4 MATERIALS OF CONSTRUCTION— IRON AND STEEL. the intermingling of iron and cinder produces a rapid union of oxygen and carbon, and this evolution of carbonic acid and oxide produces rapid " boiling " and frothing. The lin- ing of the furnace yields oxygen also by the reduction of the oxide of which it is composed. The boiling soon ceases, and small masses of reduced iron appear here and there. In an hour or more from the com- mencement, the whole mass is an aggregation of pasty grains, and the puddler, raising the temperature of the fur- nace to its maximum, works the iron which has been "brought to nature " into a half dozen spongy masses of con- venient size, weighing about sixty to eighty pounds (27 to 36 kilogrammes) or more each, meantime working his fire until a smoky flame appears, and thus he secures the now nearly pure iron from oxidation. The balls thus made are heated up to welding temperature, well worked and compacted, and finally removed from the furnace. Using gray iron, six heats are made in twelve hours ; with white iron seven can be made. The loss of weight of pig amounts to under ten per cent., and the fuel used amounts to about one ton or kilogramme of coal per ton or kilogramme of iron made by the process of boiling. With refined iron, the consumption of metal is about 2,300 pounds per ton (1,044 kilogrammes per 1,016 kilogrammes), and of fuel about ^ton (762 kilogrammes). Two men are employed at each furnace, the puddler requiring an assistant to manage the fire, and to aid, in lifting the iron into and out of the fur- nace. The best work is done when the iron is puddled in small quantities. The Principles and Theory of Puddling are evi- dently very simple. Urbin divides the process into five periods : that of fusion ; that of purification ; that of refining to produce grain ; that of carbonizing the grain, and that ^ to 5 centimetres) greater diameter and at lower speeds, reaching a minimum of 150 feet (45.6 metres) per minute. Smaller sizes are drawn on drums of 22 inch (56 centimetres) or less diameter, and at speeds running up to 500 feet (152 metres) per minute. The Resistance offered by Wire in passing through the draw-plate varies with the size of wire, character of metal, and arrangement, proportions, and management of the wire- 122 MATERIALS OF CONSTRUCTION— IRON AND STEEL. blocks. Good metal, under ordinarily good conditions, requires, to reduce it one size, from lOO pounds (45-4 kilo- grammes) with the finer, to i,ooo pounds (454 kilogrammes) of the larger grades. The ratio of reduction of area of section is usually about i^ to i. An approximate value for good wire is obtained by the empirical equation P= 30 . 0.250 — a?' m metric measures. 14 *" 0.25 — o.4//„' in which P is the pull and d is the diameter of the wire. The velocity of drawing is, customarily, in feet per minute, nearly where N is the number of the wire on the Birmingham gauge. The power demanded is, therefore, at the draw plate, in British measures, PV _ 7SoN 33,000 33,000(0.250-^)" In drawing down billets, the heaviest work and greatest reduction of size take place in the " roughing " or " nipping" blocks, and no special attention is paid to the size or to gauging. The last drawing is done in the " finishing blocks," and the wire is carefully drawn precisely to gauge. * In " wet-drawing " the metal is drawn directly from the lees-tub in which it receives the alkaline coating, and the wire is thus preserved from oxidation, as is also the draw-plate, and is, at the same time, lubricated. Lime- WROUGHT OR MALLEABLE IRON. 123 coated wire is drawn through grease. Bright wire is drawn dry. Wire is often " coppered " by drawing it through a bath of solution of copper sulphate, or is tinned or " galvanized " by leading it through a bath of tin or of zinc kept at a tem- perature slightly above the melting point, to the finishing block. When finished, sizes o to 20 are made up into " bundles " weighing 63 lbs. (28.6 kilos) each, and smaller sizes into "stones" of 12 lbs. (5.4 kilos) each. The smallest size ordi- narily met with is No. 36 (0.004 inch, 0.102 centimetre di- ameter), but No. 40 (0.003 inch, 0.008 centimetre diameter) has been made. Sizes. — Wire is gauged by the " Birmingham Wire Gauge " in Great Britain, and by the " American Gauge " sometimes, but not 'usually, in the United States. The table on page 124 gives the sizes of the standard numbers. The Processes of Rolling and of Wire-Drawing-, greatly increase the strength of iron. Good iron, which, in round bars, 2 inches (5.08 centimetres) in diameter, has a tenacity of 54,000 pounds (3,780 kilogrammes per square cen- timetre) per square inch, when rolled into one inch rods often attains a strength of 60,000 pounds (4,200 kilogrammes). When drawn into No. 10 wire (0.134 in., 0.34 centimetre), its strength becomes about 90,000 (6,300 kilogrammes), and Nos. 15 and 20 (0.072 and 0.035 in., 1.8288 and 0.88899 milli- metres), respectively, have a tenacity of about 100,000 and 111,000 pounds per square inch (7,030 and 7,733 kilogrammes per square centimetre). A wire yi inch (0.31 centimetre) in diameter is ordinarily expected to sustain 1,000 pounds (454 kilogrammes), and one of -jV inch (0.079 centimetre) diam- eter, should carry 100 pounds (45.4 kilogrammes). In wire mills, great skill and judgment are necessary in choosing good metal, and in preserving its excellent qualities throughout the processes of reduction in the rolling mill and in drawing. Good iron for fine wire must be pure, free from cinder, strong and ductile, and probably must have a comparatively 124 MATERIALS OF CONSTRUCTION— IRON AND STEEL. GAUGE OF WIRE. D lAMETER, BIRMINGHAM DIAMETER, AMERICAN GAUGE. NUMBER OF GAUGE. GAUGE. ] •^CH. MILLIMETRES. INCH. MILLIMETRES. oooo 454 11.532 .46 11.684 ooo 425 10.795 .40694 10.336 00 38 9.6519 -3648 9 266 o 34 8.6359 .32486 8.2511 I 3 7.6199 .2893 7.3481 2 284 7-2135 .25763 6.5437 3 259 6.5785 .22942 5,8272 4 238 6.0451 .20431 5 • 1894 5 22 5.588 .18194 4.6212 6 203 5-1562 . 16202 4-II53 7 18 4.572 . 14428 3-6647 8 165 4-191 .12849 3.2636 9 148 3-7593 .11443 2.9065 10 134 3-4036 .10189 2.588 ir 12 3-048 .090742 2.3048 12 log 2.7686 .080808 2.0525 13 095 2.413 .071961 1.8278 14 0S3 2.1082 .064084 1.6277 15 072 1.8288 .057068 1-4495 i6 065 I-65I .05082 1.2908 17 058 1-4732 -045257 I . 1495 i8 049 I . 2446 . 040305 1.0237 19 042 1.0668 -03589 .9116 20 035 .88899 .031961 .S118 21 032 .81279 .0284C2 .7229 22 028 .71119 -025347 -643S1 23 025 .63646 .022571 -5733 24 022 -55879 .0201 -51054 25 02 • .508 -0179 •45466 25 018 .4572 -01594 .40487 27 016 .4064 -OI4I95 •36055 28 014 •3556 .012641 .3210S 29 013 .33096 .011257 •28593 30 012 .3048 .010025 -25463 31 01 -254 .008928 .22677 32 oog .2286 .00795 .20193 33 008 .2032 .00708 -17983 34 007 .1778 . 006304 .16012 , 35 .005 .127 .000614 -14259 36 .004 .1016 .005 .127 37 . . . . -004453 .11311 38 .003965 . 10071 39 .003531 .0S969 40' .003144 .07986 WROUGHT OR MALLEABLE LRON. 1 25 low elastic limit. It has been observed, that of two irons selected, both having great strength and ductility, that which stood lowest in both respects would draw into the finest wire. The only peculiarity detected by the Author, in this case, was a comparatively low elastic limit, as shown on the auto- matically produced strain-diagram. Charcoal-bloom iron is usually found best adapted for wire-drawing if free from cinder and coal. Very excellent metal has been made by the Bessemer and Siemens-Martin process for wire of sizes exceeding No. 10. Irregular and Peculiar Shapes, which cannot be produced in the rolling mill, are given to wrought iron by the process of forging. This constitutes a trade, the methods and principles of which are properly the subject of a special treatise. The art consists principally in working simple forms, as bar or plate iron, into more complicated shapes, by hammer- ing at a bright red heat, and in making up larger masses and forming them as desired, by welding together smaller pieces. This work is sometimes done by the common hammer and sledge, and on the anvil and in formers by the blacksmith, sometimes under the steam or trip-hammer, or under the drop-press. It is also sometimes done under the hydraulic press. The practice of special methods, and the use of special tools for peculiar forms and sizes, is an important branch of the art. The conditions of success in doing this class of work are : the choice of iron free from sulphur and phosphorus, but containing some silicon where it is to be welded ; working at a bright red heat, and welding at a white heat, or " welding-heat," with thorough fluxing ; working rapidly, and with the least possible number of heats ; and keeping the direction of " grain " as nearly as possible in the line of strain anticipated, and without breaking the fibre. Irregular shapes and peculiar forms are made in cast iron by the processes of moulding and founding, which must also be described in detail in special treatises on that trade, on designing work to be made in cast iron, and on pattern- making. 126' MATERIALS OF CONSTRUCTION'— IRON AND STEEL. The method, briefly described, consists in the preparation of a pattern, or mould, usually of wood, but sometimes of plaster or of metal — of the shape of the piece to be made, modified in some respects to avoid difficulties arising in moulding. This pattern is imbedded in moulding sand, and, when removed, leaves an impression which has the shape of the casting which is to be made. Molten metal, poured into the mould thus made, solidifies in the desired shape. Cavities in the casting are produced, sometimes, by making similar cavi- ties in the pattern, and moulding within them masses of sand, which displace metal in the mould so as to produce the proper form of cavity in the casting. Often er, however, a projection is made on the exterior of the pattern, which cor- responds in location and in its cross-section with the mouth of the opening desired. These " core-prints " leave in the mould impressions, into which the " cores " fit, and by which the latter are firmly held. The cores are masses of sand, which have been moulded in " core-boxes," and given pre- cisely the shape of the cavity, but with extensions at those points at which the cavity comes to the exterior surface of the castings. These extensions fit into the impressions made by the core-prints, and the core is thus held in place while the molten metal is flowing around it. The Work of Designing metal parts of machinery involves the intelligent consideration of the cheapest and most satisfactory^methods of moulding those which are to be cast, as well as of forging parts made of wrought iron and steel. The pattern-maker must also know how to prepare the pattern, so as to avoid the difficulties frequently met with in moulding. The moulder is required to know how to mould the piece in order to secure sound castings ; and the founder must understand the mixing and melting of metals in such a man- ner as will give castings of the required quality. The engi- neer should know what forms can be cheaply made in cast metal, and what cannot be cast without difficulty, or without liability to conie from the mould unsound. He should be WROUGHT OR MALLEABLE IRON. 1 27 able to instruct the pattern-maker in regard to the form to be given the pattern in order to make the moulder's work easy and satisfactory, to tell the moulder how to mould the pattern, with what to fill his flask, and how to introduce the molten metal, and to provide for the escape of air, gas, and vapor, and he should be able to specify to the founder the brands and mixtures of iron to be chosen. Malleableized Cast Iron, or malleable cast iron, are steel castings, i.e., castings made originally of ordinary cast iron, which have been subjected to a process of decarboniza- tion which results in the production of a crude wrought iron. The process is most conveniently applicable to small cast- ings, although pieces of very large size are sometimes thus treated. Handles, latches, and other similar articles, cheap harness mountings, ploughshares, iron handles for tools, wheels and pinions, and many small parts of machinery are made of malleable cast iron, or as steel castings. For such pieces, charcoal cast iron of the best quality should be selected in order to insure the greatest possible purity in the malleable product. White iron must be used. The castings are made in the usual way, and are then im- bedded in oxide of iron — in the form, usually, of hematite ore — or in peroxide of manganese, and exposed to the tem- perature of a full red heat for a sufficient length of time to insure the nearly complete removal of the carbon. The process, with large pieces, requires many days. If the iron is carefully selected, and the decarbonization is thoroughly performed, the castings are nearly as strong, and sometimes hardly less malleable, than fairly good wrought iron, and they can be worked like that metal. They will not weld, however. The pig-iron should be very free from sulphur and phos- phorus. The best makers for small work melt the metal in crucibles having a capacity of 50 to 75 pounds (22 to 34 kilo- grammes), keeping it carefully covered to exclude cinder and other foreign matter. The furnace is similar to that of the brass foundry, from 2 to 2>^ feet {^.6 to 0.75 metre) square, and the fire is kept up 128 MATERIALS OF CONSTRUCTION— IRON AND STEEL. by natural draught. The temperature is determined with sufficient accuracy for the practical purposes of the foundef by withdrawing a portion on an iron bar. If hot enough, the drop burns on exposure to the air. For large work weigh- ing, as now often happens, many tons, the open-hearth fur- nace is employed successfully. The " cementation," or decarbonization, is conducted in cast-iron boxes, in which the articles, if small, are packed in alternate layers with the decarbonizing material. As a maxi- mum, about 800 or 1,000 pounds of castings are treated at once. The largest pieces require the longest time. The fire is quickly raised to the maximum temperature, but at the close of the process the furnace is cooled very slowly. The oper- ation requires from three to five days with ordinary small castings, and may take some weeks for large pieces. This process was invented in 1759. Decarbonization is often performed, in the production of steel castings, by a process of dilution accompanied with, possibly, some " dissociation." By the preceding method the carbon takes oxygen from the surrounding oxides, and passes off as carbon monoxide (carbonic oxide) ; in the process now referred to the carbon of the cast iron is shared between the latter and the wrought iron mixed with it in the melting pot, and a small portion may possibly pass off oxi- dized. The latter method has been practiced to some ex- tent for a century. Selected cas^ iron and good wrought iron are melted down together in the crucible and cast in moulds like cast iron. The metal thus produced contains a percentage of carbon, which is determined by the proportions of cast and wrought iron in the mixture. The amount is so small, fre- quently, that the castings made can be forged like wrought iron. The process is properly a steel-making process, and will be considered at greater length when treating of the manufacture of steel. Tin-Plate is sheet iron coated on both sides with a very thin layer of tin. The market is supplied with two kinds : charcoal plate, and coke plate. The blooms intended for manufacture into tin-plate are WROUGHT OR MALLEABLE IRON. 1 29 prepared in Wales, whence the greater part of the tin-plate in the market is received, by refining with charcoal and re- working after reheating in a coke fire. The metal selected is usually perceptibly red-short, the effect of sulphur, although deleterious at a high temperature, being to confer upon the metal exceptional toughness when cold. The pig-iron loses about 25 per cent, of weight in refining and conversion into bar. The bar is usually about 30 feet (9.1 metres) long, 6 inches (15.2 centimetres) wide, and i^ or I ^ inches (3.8 to 4.4 centimetres) thick. The bars are cut up into pieces a foot long, piled, reheated, and rolled into finished bars, losing again about 25 per cent, in weight, and taking about ^ or % ton (635 to 762 kilo- grammes) of coke per ton (i,oi6 kilogs). The short finished bars are given such size and proportions as best fit them to be worked into plate of the thickness and other dimensions pro- posed. The bars, cut to proper length, are taken to the rolling mill, where they are reheated in a reverberatory furnace, rolled, doubled, and reheated, and again rolled, the rolling being re- peated from four to six times, and the bar gradually assumes the form of a small, rectangular sheet, which, after being sheared to gauge, is ready for the operations preparatory to tinning. In rolling the bar, it is passed through the rolls with its axis parallel to that of the roll. Throughout these processes the greatest care is taken to keep the metal, when heating it, under a deoxidizing flame, and to avoid every cause of injury of surface, as by the formation of scales. The pile of plates, brought finally from the rolls, is "opened," the sheets separated, each bar having made, usually, eight or sixteen sheets. The more frequently they are doubled, the greater the waste in rolling. The plates are next " pickled," in a bath of dilute sul- phuric or hydrochloric acid in leaden vessels heated by a fire beneath, are then washed thoroughly two or three times, and finally dried and annealed by heating in tight boxes to a bright red heat, and slowly cooled. The annealed plates are " cold rolled " between very 9 130 MATERIALS OF CONSTRUCTION— IJiON AND STEEL. smooth and accurately turned rolls, again annealed at a moderate temperature, and pickled and washed again. Those which are not well cleaned and smooth are scoured with fine sand, and all are singly examined and are handed to the tin- ner as nearly as possible absolutely clean. The Tinning is done by a " gang " consisting of the " tinman," the " wash-man," the "grease-boy," and the " list- boy." The first receives and inspects the plates and places them in a trough of clean water, whence they are taken as required and immersed in a vessel containing warm melted grease. When well coated with grease they are put into the " tin-pot " and submerged in molten tin, the surface of which is flooded with grease, and kept clean by plunging into it wooden sticks, the gas and vapor evolved from which carry impurities to the surface, and check oxidation. This tinning is repeated in the " wash-pot," which contains tin of better grade, and at a lower temperature, and the plates are again carefully brushed and cleaned, and, finally, are dipped into another compartment of the wash-pot, in which they take a coating of the best quality of tin. The plates are then removed to the " grease-pot," in which, under grease, a small quantity of tin is kept at a temperature exceeding the melting-point, the superfluous metal is drained off the plates. On removal, they are cooled in the " cold pot." The " list pot " contains molten tin in a pool at the bot- tom, only about a quarter of an inch deep. The line of tin of excessive thickn^s, which forms at the lower edge of the plate when draining and cooling, is here melted off, and the plates are scoured with bran and woolen cloths, and are ready for inspection, classification, and packing for market. A Box of " IC " plates contains 225 sheets, each 13^x10 inches, and, if standard, weighs 1 12 pounds. If not less than 109, nor more than 115, the box would pass in the market. A box of "HC" tin weighs 119 pounds; one*of " IX " weighs 140 ; one of " IXXXXXX " weighs 245. " D " plates are packed 100 in a box, are i6^xi2j4, and weigh from " DC," 98, to "DXXXX," 189 pounds. In making coke plates, the cast iron is refined, and is WROUGHT OR MALLEABLE IRON. 1 31 usually puddled instead of being decarbonized in the refinery. The. loss of metal is about the same. " Terne-plates " are tinned with an " alloy " of tin and lead ; the proportion of tin varies from one-third to two-thirds. These plates are largely used for roofing. Russian Sheet-Iron is a thin sheet iron used for purposes requiring a smooth polished surface, which is not likely to oxidize readily, as for stove-pipe. The bright sur- face coating consists of iron oxide. It is made of a fine quality of wrought iron, rolled and annealed. The polish is given by hammering packages of the sheets with charcoal in- terposed. The best is imported from Russia ; but less excel- lent iron of this class is made in other countries. In Russian works, selected iron is hammered into slabs of the right size to make each a finished sheet. The slab is passed through the rolls three or four times, and subsequently hammered again. Several sheets are then heated to a full red heat, covered with charcoal shaken on them from a bag made of coarse linen, and piled with covering sheets of heav- ier iron, top and bottom. The pile is then worked down under a heavy hammer, nearly to the finished size. When cool, the hammering ceases, the plates are separated, re- heated, and piled again with cold plates interposed, the hot and cold sheets alternating in the pile, and hammering them until cool, they are finished. They are then separated, cut to size, weighed and assorted for the market. The loss of metal in the manufacture is sometimes 30 per cent. The sheets are usually about 5 feet by 2j4 (1.6 by 0.8 metres) in size, and weigh from 6 to 12 pounds (2.7 to 5.4 kilogs), exceptionally heavy plates being made, however, weighing 30 pounds (13.6 kilogs). The rolls make 75 or 80 revolutions per minute, and re- quire driving power of about 40 horse-power. The hammers have very broad faces. The finished sheets should be capable of being bent from four to six times without cracking. CHAPTER V. THE MANUFACTURE OF STEEL. Steel is variously defined by acknowledged authori- ties, and the metals known in the market and to the trade as steel cannot be completely and satisfactorily classed under any definitions yet proposed. The term includes, as formerly accepted, all impure irons which, in consequence of the presence of other elements, have the property of hardening by sudden cooling from a high temperature, and of taking a definite " temper," or de- gree of hardness, by a definite modification of temperature, and which may also be forged. It has been more recently proposed to define steel as a compound consisting principally of iron, which has been ren- dered homogeneous by fusion ; still another definition is " iron recarbonized. " The first definition is based upon composition and properties ; the others upon the method of manufacture. The latter compels the engineer to ascertain the history of the metal before he can give it a name. The trade has prac- tically adopted the last method of nomenclature. An international committee, appointed at the instance of the American Institute of Mining Engineers, in the year 1876, recommended the following nomenclature : I. That all malleable compounds of iron with its ordinary ingredients, which are aggregated from pasty masses, or from piles, or from any forms of iron not in a fluid state, and which will not sensibly harden and temper, and which gener- ally resemble what is called " wrought iron," shall be called Weld-Iron (German, Schweisseisen ; ¥ rench, fer soud^). II. That such compounds, when they will, from any cause, harden and temper, and which resemble what is now called Forgeable ; difficult to melt : Forgeable iron Iron. MANUFACTURE OF STEEL. ^33 " puddled-steel," shall be called Weld-Steel (German, Schweiss-stahl ; French, acier soud^). III. That all compounds of iron with its ordinary ingre- dients, which have been cast from a fluid state into malleable masses, and which will not sensibly harden by being quenched in water, while at a red heat, shall be called Ingot-Iron (German, Flusseisen ; Yvench., fer fondu). IV. That all such compounds, when they will from any cause so harden, shall be called Ingot-Steel (German, Fluss- stahl ; French, acier fondu). As arranged by Wedding, the following is the scheme of the system : Obtained in a fluid state : |j_ -^ngotTtefi; A. Ingot iron, j Not hardening: Obtained in a non-fluid V xj j statp ■ Hardening : ^"^^"^^ • „„,,,. J 3. Weld steel. B. Weld iron. -^^ Not hardening: (_4. Weld iron. Not forgeable ; easy i ^;,^ amorphic carbon only: C. White pig. Pi iron ' ( "^''^ graphite : D. Gray pig. There may be added, called as formerly, Rcmelted pig iron — cast iron ; remelted steel of every description — cast steel. Here the division following the fluid or non-fluid state keeps the principal place ; the hardening only the second. The grades are practically distinguished quite readily, thus: I. and II. by their capabiHty of being readily forged, or otherwise ; A and B by their characteristic fracture ; C and D by their color. Or, by chemical analysis, according to Wedding, the percentage of carbon will be : (L contains from 0.0 to 0.3 per cent. ( i from 0.6 to 3 percent. C. Iron. -^ 2 and 4 contain from o to o. 6 per cent. Steel. \ ^^^^ o. 6 to 3 per cent. C. ( II. contains 2 to 3i per cent. The higher figure for steel is unusually large. The American Institute of Mining Engineers, discussing the report of the International Committee on Nomenclature, 134 MATERIALS OF CONSTRUCTION^— IRON AND STEEL. assented to the following as a correct Commercial Nomencla- ture of iron and steel, while recommending the use of the pro- posed nomenclature in papers written by members. Iron. Wrought. - Bloom . Puddled . , Steel . (Catalan Finery. ( Bars, -j Plates. (Beams, etc. f Blister. J German. ] Shear. [.Puddled. Cast., "Not malle- able. . . . Malleable. (Pigii (All 01 iron, rdinary castings. Castings, annealed and decarbonized in oxides. ' Castings not highly carbonized. Crucible. Steel. ■< Bessemer, or pneumatic. ( Siemens-Martin. Open hearth . . \ Siemens by pig iron ( and ore process. The metals called Steel grade into each other by imper- ceptible variations: Hand-puddled iron has the properties of crucible steel and ingot metals, which are considered by the trade as indisputably steel ; while the product of the Bes- semer and other processes yielding ingot metal is, when con- taining very little carbon, sometimes as fibrous and silky in texture as common wrought iron. Steel, made at one of the largest works in France, is classified into three divisions. A, B, and C ; of which A covers all cheaper grades of steel, such as are produced by the Bes- semer and the Siemens-Martin processes, and the low grade crucible steels ; division B includes steels of ordinarily good quality ; and C includes the purest and best metals, such as are made from the best Dannemara, or similar Swedish or%s, from charcoal pig and by the crucible process. Each division thus designated by the purity of the metal is subdivided, according to "temper" or hardness, and the several grades of temper are determined by mechanical test MANUFACTURE OF STEEL. 135 These grades are selected by the eye, all ingots being " topped " — /. e., having the top broken off — before being rolled, inspected, and assorted into these grades. Although the proportion of carbon mainly determines the grade or temper of a steel, other elements frequently, and sometimes greatly, modify its quality. Silicon, manga- nese, chromium, sulphur, and phosphorus, are the most com- mon of these modifying ingredients. Steels made by alloy- ing iron with tungsten, chromium, and titanium, are some- times called "compound " steels. The distinction which is often made between irons and steels according to method of manufacture, and which classes metal made by any process involving welding, and metal made by melting and casting into ingots as steel, is sometimes made the basis of a double classification of steels, thus : CARBON, PER CENT. WELDED METAL. ("IRONS.") CAST METAL. (" STEELS.") o to 0.25 0.15 0.45 to to 0.45 0.55 0.55 to 1.50 Common iron. Granular iron. Steely iron ; puddled steel. Cemented iron or steel. {Very soft steel. ' ' Homogeneous metal. ' Soft steel. Semi-soft steel, j Hard steel. ( Tool steel. One of the largest makers of Europe divides all steel into four classes : 1st Class. — Extra mild steels. Carbon, 0.05 to 0.20 per cent. Tensile strength, 25 to 32 tons per square inch. Ex- tension, 20 to 27 per cent, in 8 inches of length. These steels weld and do not temper. Used for boiler-plates, ship-plates, girder-plates, nails, wire, etc. 2d Class.— Mild steel. Carbon, 0.20 to 0.35 per cent. Ten- sile strength, 32 to 38 tons per square inch. Extension, 15 to 20 per cent. Scarcely weldable, and hardens a little. Used for railway axles, tires, rails, guns, and other pieces exposed to heavy strains. 3d Class. — Hard steel. Carbon. C.35 to 0.50 per cent. 136 MATERIALS OF CONSTRUCTION— IRON AND STEEL. Tensile strength, 38 to 46 tons per square inch. Extension, 15 to 20 per cent. Do not weld, but may be tempered. Used for rails, special tires, springs, guide-bars of steam-engines, pieces subject to friction, spindles, hammers, pumpers. 4th Class. — Extra hard steel. Carbon, 0.50 to 0.65 per cent-. Tensile strength, 46 to 51 tons per square inch. Ex- tension, 5 to 10 per cent. Do not weld, but may be strongly tempered. Used for delicate springs, files, saws, and various cutting tools. The peculiarities and the characteristics of the several grades of carbon steels, and the differences produced by the introduction of the various metallic and non-metallic ele- ments found in manufactured steels, and the modification of quality produced by special treatment, will be described at length in a chapter on the properties of steel. The International Committee's classification may be put in the following convenient form : I. CANNOT HARDEN — IRON. II. CAN HARDEN — STEEL. Puddled iron. ) , , . , ) tur i j ( A, has not ( Blister steel. Bloomary iron. L '^^ ^^' "°' >^^" \ ^f°- \ been fused - \ PuddUd steel. Malleable castings. ) f'^s^d-a/./rf tron. ^ Metal. 1 ^^^ ^^^^^_ ] ^j.^^^ ^^^^^_ Bessemer iron. \ '\ I T? h h {Bessemer steel. Siemens-Martin I B, has been ! Ingot) , 'a _ • „■ t } Siemens- Mar- iron. ( fused — in.^of iron. ( Metal, "l j^° /" •■'•■&"' -j ^^ ^f^^j_ Crucible iron. I J L ' I Crucible steel. The Steel-Making Processes may be divided into three classes : (i.) That which includes those steels made of malleable or wrought iron carburetted ; (2.) That which comprehends all processes in which metal rich in carbon, as common cast iron, is partially decarbonized ; (3.) That in which highly carburetted iron is first completely decarbonized and then recarbonized to the proper degree in a single process. The common " Crucible Process " is of the first class ; one form of the pneumatic process, and the ordinarj'- methods' of making "Puddled Steel," belong to the second; while the now generally practiced pneumatic method known as the Bessemer process, and the "Siemens-Martin Process," are examples of the third class. MANUFACTURE OF STEEL. 137 Of these three processes, the first supplies the fine steel of the cutlers, the last the softer metal of construction. Steel of Cementation.— A very usual method of making steel consists in the carburization of bar iron by heating in charcoal, and subsequently working under the hammer or melting in crucibles. The first step in the proc- ess is that of conversion or cementation. The " converting furnace," Fig. 34, as built by the best steel-makers, is a structure of brick-work inclosing a pair of fire-brick boxes, troughs, chests, or pots, as they are variously termed, in which the bars are placed in a bed of charcoal. Beneath these chests is a fire-place, the flames from which envelop the chests while passing to the chimney. These chests are open, and a fire-brick arch is turned over them. The ends of this arched roof are closed in ; but openings are left, through which a workman can enter to fill the chests. Flues from the fire-place are led up between and around the chests, and the flames, after en- veloping the latter and filling the arch, pass out on either side, entering low chimneys, whence they issue into a tall, open-topped, pyramidal covering of brick-work, which constitutes the main and external portion of the whole structure. The size of furnace varies somewhat with different makers. The usual size of chest, as adopted by Sheffield makers, is from 8 to 15 feet (3.4 to 4.6 metres) long, and 2 to 3 feet (0.61 to 0.91 metre) wide. The height of the pyramidal stack is usually 30 or 40 feet (9.1 to 12.2 metres) ; its base has a length of about three times that of the chest, and is twice the width of the pair of chests inclosed. Sometimes two pairs of chests are placed side by side, and the width of base is then about two-thirds its length. Fig. 34. — Converting Furnace. 138 MATERIALS OF CONSTRUCTION— IRON AND STEEL. The bars selected for cementation are -^^ to f-inch (0.8 to 2 centimetres) thick, 3 inches (7.6 centimetres) wide, and of such length that they may be conveniently packed in the converting furnace. A thin layer of coarsely ground charcoal is spread over the bottom of the chest, and on this the bars are laid with spaces between them, which are filled with ground charcoal, while another layer covers the iron ; this last layer is, in turn, made the bed for another set of bars. Alternate layers of iron and coal are thus laid down until the chest is nearly filled, leaving room for a thicker layer of charcoal at the top. The whole is finally covered with fine, dry sand, or clay, or is plastered over with " wheelswarf " — the sand from grind- stones. Care is taken to exclude the air very thoroughly. " Trial bars " are placed where they may be withdrawn, through openings left in the chest, for the purpose of occa- sionally determining, by their examination, the condition of the steel. These openings are plugged with clay, and the manholes in the heads of the arch are closed by bricks. The powdered charcoal, used as the " cement," is usually made from hard wood, and is sometimes mixed with a small proportion of salt and of wood-ash. The last-named mate- rials are expected to flux the silica contained in the charcoal, and thus to prevent injury of the steel by the absorption of silicon. The iron used is very carefully selected, if tool-steel is to be made, and is»the purest known in the market. Swedish iron is used almost exclusively by British steel-makers, and largely by makers in the United States. The latter also use a few well-tried brands of iron, which are generally made from Lake Champlain or from Lake Superior ores. For machinery steel and cheaper grades, other less costly and less pure ores and irons are used. The troughs having been charged and closed up, the fire is started and the furnace is slowly heated up, attaining, in two or three days, a temperature of about 2,000° Fahr. (1,095° Cent.), at, or above, which temperature it is held for several days. Steel for tools requiring a considerable degree of car- MANUFACTURE OF STEEL 139 bonization, and made from the heavier sizes of iron, is a week in acquiring the necessary temperature, is retained a week or ten days at maximum heat, and occupies nearly a week cool- ing down ; thus, three weeks' time is needed to convert each charge. Each furnace makes sixteen charges per annum. With thinner metal, and a lower degree of carbonization, less time is required ; a week of maximum heat answers for shear steel, and four or five days for spring steel. During the latter portion of this period, the trial bars are occasionally examined, and the gradual change of texture, which indicates the gradual introduction of the carbon as it penetrates the metal, is observed. When the carburization has become satisfactorily complete, the furnace is cooled down and the steel removed. The bars are then found to have become somewhat in- creased in dimensions, with a corresponding decrease of density, and are seen to be " blistered " in many places, by the bursting off of a pellicle of surface metal where the carbon oxides have forced their way out. The metal has become hard and elastic, with the granular fracture and all the char- acteristics of steel. The proportion of carbon is a maximum at the surface, and regularly decreases toward the centre of the bar, the carbon necessarily penetrating the metal, under a gradual decrease in " head," by a slowly progressing flow from the surface. The texture is usually irregular and crystalline, the color white ; the grain is finest toward the centre and coarsest toward the surface. Case-hardening is a modification of this process, in which cementation is only carried so far as to give a steely charac- ter to a thin surface layer. The " cement " used contains less carbon, apd often con- sists largely of nitrogenous matter and hydrocarbons, such as are found in scraps of leather. A common mixture consists of about ninety per cent, carbon, and ten per cent, of car- bonate of lime or of potash. The prussiate of potash — potas- sium ferrocyanide — is often added. Blister steel, which is the product of conversion by the usual method, is, in consequence of its irregular constitution 140 MATERIALS OF CONSTRUCTION— IRON AND STEEL. and structure, unfit for general use, although sometimes made into cheap grades of tools. It is largely used only for conversion into " shear steel " when containing so little car- bon as to weld readily, and into cast steel when containing too much carbon to permit welding. From one to two parts of fuel are consumed, according to degree of carboniza- tion, per part of steel made. Shear steel is made from blister steel by shearing the bars into short lengths, piling, reheating, and drawing down at a good welding heat, using a flux to insure thorough union into a sound bar. A common method consists in piling five bars of blister steel, of which one is longer than the others, and serves as a handle by which the mass is manipulated under a tilt-ham- mer. The bundle is secured by wrapping with wire ; the flux is clean sand. As soon as the pile is compacted suffi- ciently by a few blows of the hammer, the binding wire is knocked off, and the pile is reheated and drawn down to the desired finished size. This process of piling and drawing down greatly improves the metal ; the bar of tilted steel is much superior in strength, ductility, and homogeneousness, to the blister steel from which it is made. A repetition of this process gives " double shear " steel, and still further improves it. Double shear steel is used for cheap edge tools and some other instruments, but cannot be used for fine work. The hammer used for tilting the steel is light and quick working, making 300 or 400 blows per minute, and capable •of regulation by the workman. Tilted steel is usually considered better than rolled ; the hammer is almost invariably used in working shear steel. " Cast Steel " is produced whenever fused steel is cast into ingots, or other forms, for the market. It is made by all methods which involve fusion, either in the operation of steel-making or subsequently. The tool steels and other fine grades are all cast steels. The finest cast steels in the market are usually produced either by melting in crucibles and casting blister steels, or by MANUFACTURE OF STEEL. 141 fusing together, in crucibles, wrought iron, carbon, and flux, and casting in ingots after thorough fusion. The products of these methods are both known as " cru- cible steel," and constitute the greater part of all steel used for cutlery, fine tools, and every kind of work for which metal of the greatest possible purity and uniformity of composition and character is demanded. In some few cases, these steels have now been displaced by the product of the Siemens and the Bessemer processes, which latter have the advantage of greater cheapness. Crucible Cast Steel, as made by melting blister or cemented steel to give it a homogeneous character, is the standard steel for fine tools. This process was introduced in Great Britain, and was probably invented, by Benjamin Huntsman, about the year 1770. He was then living near Sheffield. The " crucibles " or pots (Figs. 35, 36, 37) in which steel is melted, as used at Sheffield and by many American makers, are composed of a fine and very refractory clay. They are about 16 inches (40 centimetres) high, 7 or 8 inches (17 to 20 cen- timetres) in greatest di- ameter, and weigh about 25 pounds (11.3 kilo- grammes). To make each pot, 20 or 22 pounds (9.1 to 9.9 kilogrammes) of new clay, i or 2 pounds (^ to i kilo- gramme) of cinder, and 2 or 3 pounds (i to 1.3 kilogrammes) of old pot material are ground together very thoroughly, and 5 or 10 per cent, of ground coke-dust is often added. The material is mixed and kneaded by treading under foot on the " treading floor " 8 or 10 hours, and is then ready for use. The pots are then formed by hand, and are allowed to dry a week or more before " annealing " them by slowly and steadily raising them to a bright red heat, and as slowly and steadily cooling them. Fig. 36. Fig. 37. 142 MA TERIALS OF CONSTRUCTION— IRON AND STEEL. In the United States, crucibles are very extensively used in which the clay is mixed with a considerable proportion of graphite. These pots wear better than those of clay. They yield some carbon to the metal, and this is compensated by a corresponding reduction of the charge of carbon introduced. These crucibles are often larger than those first described, and carry from 40 to 70, or even 80 pounds (18 to 36 kilo- grammes) of steel. Steel-melting furnaces (Fig. 38) are usually plain fire-brick structures with rectangular chambers, each of a size sufficient to permit the introduction of two crucibles. These chambers are arranged side by side before a common flue, and with their tops level with the floor. The ash pits are reached from a trench or " cave," extend- ing along the front of the row. A high chimney gives the sharp draught which is essential to secure the in- tense heat demanded in steel melting. Coke is the usual fuel. The Siemens gas fur- nace with its regenerative system is now mainly employed, in large establishments, for heating steel melting-pots. The higher temperature, purer flame and lower cost, in fuel, breakage of crucibles, give great advantages. The use of manganese was first practiced by Josiah M. Heath, who patented the process in 1839; but it is probable that manufacturers sometimes used fluxes containing manga- nese at an earlier date, without, however, knowing the reason of their efficiency. The precise action of manganese is not fully determined. Without it, only the purest known irons could be used in making steel, and even those" were not usually capable of Crucible Furnace. MANUFACTURE OF STEEL. 143 being converted into a thoroughly malleable product. By the addition of a small quantity of manganese, which rarely exceeds in amount, in the cast steel as finally produced, one eighth of one per cent., the metal becomes easy to work, and in every way improved. It acts as a corrective of red shortness and an " antidote " to sulphur ; but this action takes place, however, without removal of the sulphur. Its effect is to pro- duce hardness without great brittleness, while giving the steel greater malleability at high heats. Manganese also assists by reducing liability to the formation of blow holes in the ingot by absorption of oxygen. The pots are set in the furnace upon stands made of clay, which sustain them clear of the grate bars. They are charged with the broken blister steel and manganese, and sometimes with a flux of broken bottle-glass, are carefully covered with a lid of crucible clay, and allowed to remain at a temperature exceeding, probably, 3,600° Fahr. (about 2,000 Cent.) an hour or an hour and a half, the furnace being, meantime, kept well supplied with fuel and at the highest temperature possible. When the steel, at the end of this time, is found to be thoroughly melted, it; is "teemed," /. e., poured off. The "puller out," protecting his clothing and person by coarse bagging or cloth saturated with water, stands directly on the top of the furnace and raises the crucible from its bed with a pair of large tongs fitted with handles 4 feet (1.2 metres) or more in length, and swings it out into the "teeming hole," as it is called, a square, iron-lined hole in the floor, large enough and deep enough to take it in. The melter next takes the pot in the " teeming tongs," and pours the molten metal into the ingot mould, a helper standing by with a "flux stick," to prevent the escape of flux with the steel. The ingot has either an octagonal section or a section like a square with the corners removed ; it tapers slightly from end to end. The mould is in two halves, and is coated within with a preparation containing soot or other form of carbon. The ingot, when removed from the mould, is " topped," i. e., a small piece is knocked from the upper end 144 MATERIALS OF CONSTRUCTION—IRON AND STEEL. with a heavy hammer or sledge, and inspected by an experi- enced judge of the metal, who assorts the several qualities into standard lots, each of which is of a definite "temper." With care and skill, differences of less than one tenth of one per cent, carbon are noted, and, in some cases, selected tool steels of high grades have been distinguished by the eye when dif- fering one twentieth of one per cent, in proportion of car- bon. The ingots are finally hammered and rolled into forms suitable for the market. The " mixture " introduced into the crucible has a compo- sition which is determined by the character of the available materials and of the product demanded. It usually consists of either blister steel produced by the cementation process, with a small percentage of carburet, or of peroxide, or of other compounds of manganese, with some suitable flux ; or it is chiefly selected iron of fine quality, with a definite proportion of charcoal and of manganese and flux. Chromium, titanium, tungsten, and other elements, are sometimes introduced, as in the production of chrome steel or of titanium or tungsten steels, so called. The principal properties of the crucible steels and of car- bon steels generally will be discussed in a later chapter. In structure they are usually very uniform and homogeneous, granular, compact and dense, and are generally found better fitted for purposes demanding fine quality than are other kinds of steel. Although each crucible contains but a small quantity of metal, immensely large castings are made by fus- ing the metal all at one time in a large number of crucibles, which are systematically emptied into a common reservoir or mould. Castings weighing many tons are sometimes thus made. Crucible steels are principally used for cutlery and other tools ; but they are sometimes made with so low a proportion of carbon as to be properly denominated "homogeneous iron," and are then used for boiler-plates, axles, lining tubes for ordnance, and for parts of machinery in which homo- MANUFACTURE OF STEEL. 145 geneousness, ductility and strength are desired in combina- tion. Wootz, or Indian steel, is made by the second of these crucible processes, as already described. Cast steels owe their excellence to the care taken in the selection of the raw material used, to the extraordinary skill acquired in assorting it, and to the great uniformity of quality and texture insured by fusion. The harder kinds of cast steel made from cement steel require the addition of carbon in the crucible when the process of cementation is interrupted before the degree of carburization demanded for the cast steel is fully attained. As steel in fusion absorbs oxygen freely, the exclusion of the air from the melting pot is important. The higher the temperature, and the less the proportion of carbon, the more freely is gas absorbed. The crucibles, when of purest clay, can only be used a few times. Those of mixed clay and graphite are more durable. In pouring, the temperature should be as low as possible, without incurring danger of premature solidification. The moulds should be warmed be- fore they are used, should be kept hot until filled, and should then be promptly and carefully covered. Puddled steel is often remelted for large castings. The hardness, the lack of ductility, and the diminished welding power of the steels make it necessary to observe greater pre- caution in working them than in the working of iron, and their susceptibility to injury by prolonged high temperature compels greater care in heating them. The ingots are worked at from a moderately high red heat to a low yellow heat ; at too high a temperature they burn, scintillate, and lose carbon ; at tod low temperatures they crack, and refuse to weld. "Open Hearth Cast Steel" is produced by melt- ing in large quantities on the hearth of a reverberatory fur- nace. The class includes steels produced by the Siemens- Martin process, to be described later. The Siemens furnace is used as a melting furnace for steel, both when melting in crucibles, and when melting on the hearth in masses of foul 14^ MATERIALS OF CONSTRUCTION— IRON AND STEEL. or five tons (4,064 to 5,080 kilogrammes). This latter is the most practicable method of producing large ingots and cast- ings. By the old method the melting of a ton (1,016 kilo- grammes) of steel requires often three to four tons (3,048 to 4,068 kilogrammes) of coke ; by the Siemens furnace the work is done readily with a ton (1,016 kilogrammes) or less of the cheapest fuel, the breakage of crucibles is greatly reduced, and the lining of the furnace is better preserved. Steel, melted on the Siemens open hearth, requires the consumption of three-quarters of a ton (762 kilogrammes) or less of fuel per ton of metal, and the rapidity of melting is very much greater than in pots. The high temperature required for steel melting has pre- cluded the use, to any considerable extent, of the ordinary form of reverberatory furnace, and the Siemens furnace is the only one which has, up to the present time, been adopted for this work with satisfactory and general success. Steel making on the open hearth of the Siemens furnace, as proposed by C. W. Siemens, was first practiced in France with commercial success, by the Messrs. Martin, in 1865. The process is generally known as the Siemens-Martin process. This method is sometimes regarded as one of decarburi- zation of cast iron by the addition of uncarburized metal ; it is perhaps more correctly a method of imparting a definite proportion of carbon to wrought iron by mixture, in fusion, with cast iron. It is extensively practiced, and is the princi- pal, and, practically, as yet, the only competitor with the pneumatic process in the production of soft and low grade steels. It may be used for the production of tool-steels, but it is seldom so applied. For this process the bed of the reverberatory furnace is given a form somewhat resembling that of the puddling fur- nace, and the lining, or bed, is made of selected clean silici- ous sand, containing some alumina or magnesia. Bauxife, containing 65 per cent, alumina, 15 per cent, silex, 5 percent, iron-oxide, and some water in combination, makes a good lining also, so long as it can be retained in place. A charge of scrap iron, or old rail-ends, either iron or MANUFACTURE OF STEEL. I47 steel, is introduced after the furnace has been brought up to a full white heat, and to this is added the required amount of cast iron. The weight of a charge is made up from 4 or 5 to 20 tons (4,064 to 20,320 kilogrammes). The following represents a fair charge : Pounds. Kilogrammes. Pig-metal ; No. 3, or white iron 5,000 2,272 Wrought iron scrap, or puddle bar 4,500 2,043 Spiegeleisen 500 227 10,000 4,542 A half ton (508 kilogrammes) of good coal is here required to make a ton (1,016 kilogrammes) of steel. As practiced in some French establishments, the follow- ing are the details of the process as given by Kohn : The pig and scrap iron and steel are heated separately be- fore charging into the steel furnace. A charge of about 2,000 pounds (900 kilogrammes) of the heated pig metal is first placed on the hearth and melted down. The scrap steel, and wrought iron, heated to a white heat, are next added in charges of 440 pounds (about 200 kilogrammes) at intervals of a half hour, each charge being melted down and thor- oughly incorporated in the bath before the next is added. Decarbonization becomes complete in six or seven hours, and the mass becomes pasty, as in puddling, after the fusion of 4,840 to 5,280 pounds (2,200 to 2,400 kilogrammes) of wrought iron. Recarburization is then effected to the customary extent by the addition of cast iron, in, usually, four charges of 440 pounds (200 kilogrammes) each, this metal being similar to that first charged, and generally containing some manganese. The bath is covered with a slag of blast-furnace cinder, rendered more silicious by the addition of sand, and the metal is kept fused any desired lengtn of time while adjust- ing its quahty. The total weight of metal charged, as above, is usually about 8,800 pounds (4,000 kilogrammes), and the product is not far from 8,140 pounds (3,700 kilogrammes). The exact proportions will vary with the composition of 148 MATERIALS OF CONSTRUCTION— IRON AND STEEL. the materials used, and with the character of product de- manded. The charge is melted down under an oxidizing flame, and the carbon is thus partly removed by burning out, and the exact proportion required is then secured by dilution with malleable metal. The molten cast metal forms a liquid bath on the hearth, into which the wrought iron gradually dis- solves. The metals having been thoroughly fused, the proc- ess is continued until samples taken from the furnace and tested exhibit the desired quality, or until they indicate complete decarburization. In the latter case, the spiegeleisen, or other manganese-bearing material is added, and tests of samples are again taken to determine quality. If the metal is not now of precisely the quality wanted, the addition of cast iron or spiegeleisen, or of wrought iron, is continued, as required, until the steel is found to be of the exact character demanded ; and it is then tapped off into ingot moulds. The cast iron should be carefully selected, and especially free from phosphorus when steel containing considerable carbon is to be made. The wrought iron should be selected with similar care, when fine steels are to be produced, and the spiegeleisen should always be very free from either phos- phorus or sulphur ; it usually contains ten per cent, or more of metallic manganese. The softer the grade of steel made, the richer should the spiegeleisen be made in manganese, and the lower its proportion of carbon. For steels containing a very small proportion of carbon, a comparatively high percentage of phosphorus, or other hard' ening element, is admissible. The cost of the steel will vary greatly with the cost oi scrap metal and of cast iron, but is usually, in gross, not fal from that of making steel by the pneumatic process. , Each furnace requires three furnace-men, and outside labor to handle the product. The cost of repairs varies greatly with the management. It should be a small item. The total cost of steel per ton (1,016 kilogrammes) should not exceed the value of ten days' laborers' work. MANUFACTURE OF STEEL. I45 This process possesses some peculiar and important ad- vantages. The steel lying on the open hearth, under a flame which may be made oxidizing, deoxidizing, or neutral at pleasure, may be sampled at convenience, retained in fusion any desired length of time, and treated in any way that may be necessary in the modification of its quality. The plant is simple, inexpensive, and can be, from time to time, enlarged, as may be considered expedient, without limit. The range of quality of material available for use is less restricted than in some other processes, and in making "mild steels" — "ingot irons" — as little as o.io per cent, carbon can easily be reached. The " Direct Process " of steel making, as practiced by Siemens, is similar to that already described as producing malleable iron, except that the final step in the process is the addition of spiegeleisen, in the manner described above, to recarburize the iron to the desired degree. After the addition of this spiegeleisen, the metal is sam- pled and tested, and, if found of proper quality, is tapped off. If it requires an additional dose of carbon or manga- nese, more pig-metal or more spiegeleisen is added, and, if the carbon or the manganese is in excess, the bath is modi- fied by addition of scrap iron, or of ore, and by exposure to the oxidizing flame. When found to be precisely of the char- acter demanded, the steel is tapped off into ingot moulds, and finally sent to the rolling mill, or shaped under the hammer. The process above described is, in greater detail, given by its inventor, in the larger work. In the Bessemer process, carbon, silicon, and manganese ap- pear to be eliminated uniformly. In the open-hearth process, the degree and the time of elimination are quite different. During the time the charge is passing into the fluid state, car- bon, silicon, and manganese are all more or less oxidized, leav- ing usually about 50 per cent, of the total amount contained in the charge, this quantity varying , slightly with the temperature of the furnace. As soon as the whole of the charge is fluid, the carbon remains almost, if not entirely, unchanged, until the whole of the silicon and manganese are oxidized. ISO MATERIALS OF CONSTRUCTION— IRON AND STEEL. The pig iron most suitable for the open-hearth process — the sulphur and phosphorus being low — is that containing the least carbon and silicon. In the first place, it con- tains a higher percentage of iron, and, in the second, it does not require to be so long in the melting furnace before the metal is completely decarbonized. Moreover, pig iron containing a large percentage of silicon, although it is all oxidized, invariably yields inferior steel. More than 0.50 per cent, of manganese is objectionable, not only on account of the delay it causes, but because of the destruction of the silica bottom by the formation of a fusible silicate of manga- nese. From long experience, it has been found that steels from different brands of hematite pig iron, chemically the same and made from the same ores, not only act differently in the furnace, taking more time, cutting the bottom, etc., but in their finished state show a marked difference in their ten- sile and other tests. When first noticed, this was attrib- uted to some defect in the mode of analysis, which failed to detect minute traces of elements possibly derived from the coke or limestone used in their manufacture ; but it was found that two cargoes of pig iron, of different brands, both of which worked in a most unsatisfactory manner by themselves, gave, when mixed in equal proportions, results which were everything that could be desired. Others invariably gave good results per se, and by mixing as many brands as possible uniform results may be obtained. Experiments jnade at Landore show that no metal added to the bath of steel has the slightest effect, so far as the elimination of sulphur is concerned, and manganese is the only metal that will counteract it. Manganese is indispensable in steel made by an oxidizing process. An ingot from a charge composed of Swedish pig iron and puddled bar made from the best hematite pig containing no manganese, will break into pieces at the first blow of a hammer, whilst a similar ingot containing 0.08 per cent, man- ganese will forge. Tungsten, alloyed with steel in small amount appears to harden it without detracting from its toughness. MANUFACTURE OF STEEL. 15 1 The Fluxing of the Steel, more especially with a view to the removal of phosphorus, when that element is present in objectionable quantity, is a subject which has attracted much attention from metallurgists. Henderson's process, applied to steel making, is an exam- ple of such an effort. In this process, fluxing is effected by the use of the fluorides and oxides, fluor-spar and iron oxide, as in the puddling process introduced by the same metal- lurgist. These materials are usually applied as a fettling on the bottom and the sides of the furnace, but may be injected in a finely divided state into the bath of molten metal. Oxide of manganese is added with oxide of iron ; and any lime present, or added in the process, assists in the formation of cinder. The character of the steels made by the above described methods is very variable, and depends upon the character of the materials found available, and upon the skill with which the work is done. The steels thus made, and especially those produced by the open-hearth process in general use, are generally " low " steels, such as are best adapted to those purposes for which iron was formerly exclusively used, and should be classed as "ingot" iron, rather than as ingot steel. The Siemens-Martin process is peculiarly well adapted to making fine grades of such metal. It is more or less economical than the Bessemer process, according to the value of the scrap wrought iron, although the continually in- creasing production of the standard plant in the latter branch of manufacture is making it more difficult to compete when extreme exactness in securing the specified grade is not de- manded. The Pneumatic Method of Steel Making:, generally known as the Bessemer Process, is the most extensively prac- ticed and the most productive, by far, of all known methods of making ingot metal. It has been known since the time of Cort that the agita- tion of molten cast iron, in presence of oxygen, will produce combustion and removal of carbon, and the reduction of the cast iron to the state of malleable iron or of steel. The pneumatic process secures such an agitation, and a 152 MATERIALS OF CONSTRUCTION— IRON AND STEEL. very thorough intermixture of the fluid iron with the oxidiz- ing atmosphere, by causing the latter to stream up through the molten mass in innumerable minute bubbles ; the rapid combustion thus secured is sufficient to supply all heat needed, not only to retain the metal in a fused condition, but, also, so rapidly and so greatly to elevate its temperature during the operation, that the product, even when entirely deprived of carbon, remains a perfectly fluid wrought iron in the converting vessel. The process was invented independently by Henry Besse- mer, in Great Britain, and by William Kelley, in the United States. Patents were issued to both by the U. S. Patent Office, and their interests were combined when the manufac- ture was established. Bessemer's patent dates from Novem- ber, 1856, and Kelley's from January, 1857, the latter having been granted after a declaration of interference. Kelley at first forced air under high pressure downward into the mass of molten metal in comparatively few and large streams ; Bessemer began in a somewhat similar way, treating steel in crucibles. In both cases, the metal was converted so slowly that chilling took place before the work was completed. When the same operation was conducted in the more effect- ive method now familiar to engineers, it became at once a practically useful process. Since the date of the early patents, this manufacture has grown to enormous proportions. It is the source of the greater part of the " steel " rails now used ; it furnishes a large amount of " steel boiler-plates." The Plant and Apparatus standard in the United States may be taken as illustrative of a highly efficient arrangement. It owes its excellence largely to the skill and intelligence of the men who have developed it in this country, mainly to the late Mr. A. L. Holley, who designed the works and nearly all peculiar details observable either in arrange- ment of plant or form of apparatus. It usually consists of a pair of " lo-ton converters," with accessory apparatus. A pair of such converting vessels was originally expected to make about 150 tons (81,280 kilo- grammes) per day, or 25,000 tons (25,480,000 kilogrammes) MANUFACTURE OF STEEL. 153 per annum. The charge has been gradually increased, and the number of heats as well, until 10 tons (10,160 kilogrammes) per charge, and 80 heats, or even 120, per day, giving a pro- duction of 300,000 tons (304,800,000 kilogrammes) per year, has been obtained from this plant. Fig. 39. — ^American 5-T0N Bessemer Plant. The general arrangement of plant is shown in the accom- panying drawings. Fig. 39 represents the ground plan as designed by Holley, and Fig. 40 is a section laterally I 1 - on the centre line of the section on h.k. pit surrounding the con- verter. The cast iron is melted in cupolas, A, A, A, A, Fig. 39 in plan, and seen in ele- y\g. 40.— Section of Bessemer Works. vation in the section rest- ing upon the second floor of the converting house, at the right of the converters, C, C. Materials are hoisted from the lower levels by hydraulic 154 MATERIALS OF CONSTRUCTION—IRON AND STEEL. elevators placed at each end of the charging floor, the one for fuel, the other for metal. The barrows on which the charge is transported are of iron, and carry about a ton (1.016 kilogrammes) of iron each, or ^-ton (772 kilogrammes) of coal. One barrow for each 8 or 10 tons of steel made per day is sufficient. The metal is charged with the limestone required for the flux, and with about one sixth or one eighth its weight of fuel. This small consumption of fuel is one of the sources of economy secured by this plant. Reverberatory furnaces, often used elsewhere for melting, are less liable to injure the product by the introduction of sulphur and phosphorus when these elements are present in fuel and flux ; but they are vastly more expensive in operation. With good fuel and pure limestone the cupola gives good economical results, however, without injury to the iron. The cupola furnaces (Fig. 41) used are made especially for this work, and, as designed originally by Mr. J. B. Pearse, are made with much greater depth of hearth, greater tuyere area, and straighter boshes than the foundry cupola, in order to fit them to carry piore iron and to melt it more rapidly. A usual form is of elliptical transverse section, 3.5 by 6.5 feet diameter (i.i X 2 metres) and 13 or 15 feet (4 or 4.6 metres) high, and with a total cross- section of all tuyeres of about 200 square inches (0.13 square metre). In charging these cupolas a bed is first laid of about 2yz tons (2,540 kilogrammes) of fuel, above which is placed IJ^ tons (1,245 kilo- grammes) of pig iron, then a half ton (508 '■A^ kilogrammes) of fuel, and above that ij^ tons (1,245 kilogrammes) of iron, and so dn until the cupola is full to the charging door. A little limestone is now and then added as a flux. Such cupolas will melt 50 tons (50,800 kilogrammes) of iron in eight hours. From Fig. 41. MANUFACTURE OF STEEL. I5S the cupolas the iron, when ready, is tapped into 12-ton (1,393 kilogrammes) ladles standing on balanced scales, where it is weighed and where it remains until the converter is ready to receive it. This arrangement permits the determination to be made of the amount of spiegeleisen needed, and the great capacity of the ladles permits the manager to use them as reservoirs into which the molten iron from the cupola can be run instead of lying in the cupola hearths and interrupting the melting process when they become too full. Air is supplied to the cupolas by fan-blowers, at a pressure of nearly or quite one pound per square inch (0.07 kilo- gramme per square centimetre). The cinder and other material dumped from the cupola slides down the inclined plane B, and are deposited near the cinder-mill, in which they are ground ; they are then assorted, and any iron found in the mass is saved and remelted. The cinder is cooled and broken up by a stream of water from a hose. From the ladles, Z, L, the molten iron is poured into troughs, D, E, or runners, of which an upper movable section, D, is attached at one end to the ladle L, while the other end is- carried on rollers, thus being given a power of self-adjust- ment as the ladle turns, and while discharging the stream of molten iron into a single lower fixed section, E, which re- ceives it from both ladles. The latter has two branches, each leading to a converter, so that the metal can be charged into either, as required. In this " melting department," as this portion of the steel- works is called, are also placed the furnaces, D, D, in which the spiegeleisen is melt- ed. These are usually reverberatory furnaces ; cupolas, although some- times used, are found less well adapted to this work, since the metal is often kept in the molten Fig. 42. condition for so long a time, that it might, in the comparatively cool and unheated 156 MATERIALS OF CONSTRUCTION— IRON AND STEEL. hearth of a cupola, get chilled. Where, as in the more pro- ductive works, the demand for spiegeleisen very frequently recurs, cupolas give good results ; they are in all cases vastly more economical of fuel than the other furnaces. In American works the reverberatory furnace, Fig. 42, is usually built with a slope, on which the metal is charged, near the flue, where the flame impinges most violently, and where it will be most rapidly melted, although it is then most liable to injury by oxidation of its manganese, which element has an exceedingly strong affinity for oxygen. In Europe, the furnace. Fig. 43, is often so built that the metal may be charged upon the slope of the bed, near the bridge- wall, where, although melted less quickly, it is less liable to oxidation. The space below the furnaces, as shown in the figures, is large enough to permit storage of raw materials for linings and repairs, and all common stores. A " crusher " and set of rolls on the cupola floor are used for crush- ing the materials to be mixed for linings. A wide gangway is carried entirely through this building on the ground level, and gives a roadway for the carriage of material. The " converting department " is placed in front of the melting department, and here two converters are mounted side by side, above the general ground level. In European works. Figs. 44, 45, they are usually placed in a pit sunk in the floor, and opposite each other, an arrangement which is less convenient, and causes the workmen to suffer more from heat than where everything is above ground. In front of the converters. A, is the casting pit, in the centre of which is a hydraulic crane, C, and around which are three other cranes, by which the ingots and ingot-moulds, D, are handled. Nearly on a level with the trunnions of the converters is another Fig. 43. MANUFACTURE OF STEEL. 157 floor, on which workmen may stand, and where materials can be placed while repairs are going on ; it is reached from the hoists, which open on this level, and the workmen enter upon it from the melting house by a side passage, when firing or repairing the converters, or when inspecting the converter bottoms. The cranes, K, swing completely around, the middle one handling all ladles and ingots in the pit, and transferring its load within reach of the other three, which latter convey it to the stor- , „ age space, near the pit, or to a carriage, which traverses a rail- way of 30-inch (76.2 centime- tres) gauge, passing over the scale, on which the weighing is done. At the left of the latter is the weigh- house, where, also, are ram- med the con- verter - bottom linings. The build- ings at the ex- treme left con- tain the boilers, engines, and pumps. The engines used for blowing the air through the converters are of various forms. The most usual arrange- IS8 MATERIALS OF CONSTRUCTION— IRON AND STEEL. ment includes two independent engines, with steam and air cylinders and fly-wheels complete, in order to avoid the entire cessation of work that must otherwise follow upon the break-down of a coupled engine. Condensing engines 42 to 44 inches (1.07 to 1. 12 metres) in diameter of steam cylinders, and about 54 inches (1.37 metres) in diameter of air-cylin- ders, with a stroke of piston of 5 feet (1.5 metres), and work- ing steam of 60 to 70 pounds per square inch (4.2 to 4.9 kilogrammes per square centimetre), are adopted with a " S-ton (5,080 kilogrammes) plant." The water sup- ply for the hy- draulic cranes, and for the ap- paratus handling the converters and the hoists, is obtained from di- f^^g^ rect-acting steam -r^~' pumps forcing water under a pressure, at the accumulator, of 350 to 400 pounds per square inch (24.6 to 28 kilogrammes per square centimetre). The converter is the only part of the plant which de- mands more mi- nute description. « In the sketch. Fig. 47, is seen the lower part of this vessel, the general outline of which is seen in the above illustra- tions already de- scribed. The jr,Q. ^g, bottom of the converter consists of a hollow detachable box, into which the Fig. 45. MANUFACTURE OF STEEL. 159 blast, which enters at the trunnion, is conveyed by side-pipes, and from which it rises through a large number of small holes, which traverse the lining mass. A, which protects the bot- tom from injury. The sides of the converter are also covered by the thick, infusible lining, B,B, of ganister or of ground siliceous stone, mixed with 10 to 12 per cent, fire-clay. The composition of ganis- ter, and of the artificial substitute in use, is about : silica, 93 ; alu- mina, 4, with 3 per cent, of other substances. The mass, v4, is mould- ed upon the top of the blast-box, as seen below, and the im- bedded tuyeres are firmly held by adhesion in the mass, and steadied also by the metal plate, which forms the top of the blast-box. When the lining is worn down, and the tuyeres become too short for further safe working, the bottom is removed, and another, which has been meantime prepared, is put in its place as shown in the sketch. An open space, C, D, is left in order that no impediment may arise from too close a fit, and this space is tamped full by driving through an annular space, D, which is left open by the construction adopted. This space is three or four inches (7 to 10 centimetres) wide, and the filling is a mixture of ganister and fire-clay worked moist and driven snugly in place. This is done while the converter lining is still red hot, and occupies from three quarters to one hour after the opening has been trimmed and smoothed out. The bottom requires frequent replacement ; the con- 48. — Converter Bottom. l6o MATERIALS OF CONSTRUCTION— IRON AND STEEL. verter itself requires less frequent relining. The latter is divided into three separate sections, which can be lined inde- pendently. The relining of the upper and lower parts is usually done in the weigh-house. The middle part, being fixed by the trunnions, is relined in place. These linings are worn away by both mechanical wear and by chemical action. The replacement of the converter bottom, when worn, does not cause delay, nor does it reduce production ; but the relining of the converter itself often causes serious loss . by stopping all work. During its work- ing period, however, 8o or 90 charges per day of 24 hours can be made into steel, and 75 charges have been made in 12 hours. In Holley's latest form of converter, the trunnion ring is detachable, and the whole converter can be removed for re- lining, another being kept ready to take its place at the removal of the first. Of the pig metal melted in the cupola, about 85 per cent, appears in the form of ingot steel, the remaining 15 per cent, is lost by oxidation and by the formation of " skulls " and slag. A nominally 5-ton (5,080 kilogrammes) plant produces, usually, 7,000 to 8,000 tons (7,112,000 to 8,128,000 kilo- grammes) per month, and sometimes greatly exceeds this figure. This rate of production has been steadily growing from the beginning, and is still increasing. Operation. — The following is, in brief, the method of operation : The cast iron is melted in the cupola — or, if fears are entertained of the introduction of injurious elements from the fuel, in a reverberatory furnace, and after complete fusion, it is transferred to the converting vessel. Before the iron is run into the converting vessel a fire of charcoal or coke is started within it, and a gentle blast turned on until the interior has been raised to a white heat. It is then ready for the charge, which has meantime been melted in a cupola or air furnace. . To fill the converting vessel, it is turned on its trunnions until the charge may be run into it, and so far that the molten metal shall not fill the tuyeres. The blast then is turned on at a pressure of from 15 to 25 pounds, per square MANUFACTURE OF STEEL. i6l inch (1.4 to 1.8 kilogrammes per square centimetre) accord- ing to the depth of the charge ; it enters at the bottom in a multitude of fine jets, through orifices about a quarter of an inch (0.6 centimetre) diameter. The vessel is then turned into the vertical position, and the blast, permeating every portion of the liquid metal, seizes upon the oxidizable ele- ments present, burning them out. The operator watches the process with great care, observ- ing the indications of the pressure gauge, the sound issuing from the converting vessel, the character of the flame, sparks and smoke issuing from the nozzle, and noting the duration of the phenomena exhibited as the operation proceeds. At the instant that the blast commences passing through the metal, oxidation begins. The air, expanding violently as it rises, dividing into large globules or minute bubbles, seizes, as it goes, first upon the silicon. As the current passes from the converter to the chimney, it exhibits but little smoke, and carries with it large and brilliant sparks. The metal, instead of being cooled by the great volumes of cold air forced through it, grows hotter and more liquid as combustion proceeds at the surfaces of the innumerable continually rising air-bubbles. The whole mass becomes agitated until the vessel, and often its foundation, trembles. A regular mufHed clapping sound is heard as the iron thrown up by the blast falls back again, and in six or eight minutes from the commencement of the process the sparks diminish suddenly in number, and a flame appears, first dull and red, but soon changing to a long tongue of fire, as the air, finding no more silicon with which to combine, seizes the carbon. The silica formed in the first stage of the process com- bines with any oxide of iron then or subsequently present, and with it makes a glassy cinder that covers the surface and assists in retaining the heat of the iron. As combustion progresses, the flame becomes, for a short time, partially obscured by smoke ; then it clears again, and a voluminous clear white flame indicates that the graphite has all been burned away, that the combined carbon has l62 MATERIALS OF CONSTRUCTION— IRON AND STEEL. begun to leave the iron, and that if the process is not checked at the proper moment, the iron itself will soon begin to burn. As the end of the operation approaches, some loss of iron is unavoidable. The temperature has become much higher than the melting point of cast iron ; for the converter now contains a mass of nearly pure wrought iron, perfectly fluid, its whole mass pervaded by minute globules of air, which no longer finds sufficient combustible matter to satisfy its affinity unless it takes the iron itself. After the appearance of the white flame the detonations gradually cease, and after about fifteen minutes from the beginning the flame grows irregular and fitful, and then suddenly disappears. The process is completed ; the converting vessel is rapidly turned down, and the blast shut off. The iron is found, if then examined, to be almost perfectly pure malleable iron, which may be run into ingots and passed through the rolls or worked under the hammer. As a low steel is, for most purposes, far more valuable than any wrought iron, and as the iron in its present state is " short," in consequence of the presence of iron oxide and of gas, the charge is next recarbonized to a certain extent before being removed from the converter. For this purpose some frank- linite, spiegeleisen, or other manganiferous cast iron is melted in a cupola before the conversion of the charge is com- menced. As soon as the process of decarbonization has ceased, a quantity of the recarbonizing material, in such pro- portion, usually S or 8 per cent., as is determined by the kind of steel required, is added to the purified iron. The materials alloying perfectly, the carbon, manganese and silicon arediffused uniformly in proper proportion through- out the charge, and thorough intermixture is insured by blow- ing it in the converter when necessary. The whole time ' occupied in changing into steel a charge of twenty thousand pounds (about io,ooo kilogrammes) of cast iron and forming it into ingots is less than half an hour. The operation might be shortened, and the recarbonization avoided, by stopping the process before the carbon is all con- MANUFACTURE OF STEEL. 1 63 sumed, and considerable quantities of cheap steel have been sometimes made in this way. But the rapidity with which combustion proceeds renders the method unreliable and the product variable in quality, and it is found far more satisfactory to complete the decarbonization and then to add a known quantity of carbon by the method already described. The steel from the converter is run into moulds, where the ingots rapidly cool, and in 30 or 35 minutes the largest masses are taken out and placed in a heating furnace or a " soaking pit," and are kept at a sufficiently high temperature to work freely while the interior cools to such an extent that the in- got may be safely carried through the rolls or forged to the desired shape under the hammer. The capacity of the converting vessels has in Europe been gradually increased until some English manufacturers are con- verting 12 tons (12,192 kilogrammes) at a single charge; but converters of 5 tons (5,080 kilogrammes) and of 7^ (7,620 kilogrammes) tons capacity are most usually employed, and larger vessels are not called for in the United States. The result of a moderately productive week's work of one pair of 5 ton (5,080 kilogrammes) converters gave in one instance a product of nearly 4,000 tons (4,064,000 kilo- grammes), and produced 325^ (330,538 kilogrammes) tons waste scrap. The product of an American plant, as here described, has in one case been reported as often above 13,000 tons (1,320,000 kilogrammes) of ingots for one month's work, and yielding 11,000 (11,176,000 kilogrammes) tons of finished product (rails). The steel, when completely recarbonized, is, as above stated, poured off into a ladle, which distributes it to the ingot-moulds, which are made of gray iron of open texture, washed within with clay or plumbago, and set in a circle about the outer circumference of the ingot pit. The ingots so produced are a foot (0.3 metre) or more square at their lower ends, tapering to a cross-section ten per cent, less at the top, and each ingot is usually so propor- tioned as to furnish material for either two or three rails, 164 MATERIALS OF CONSTRUCTION— IRON AND STEEL. where railroad metal is to be made, in order to save in the weight of rail ends returned as scrap. The large ingots also furnish better steel, as they are worked more in the process of rolling. They are 3^ to 4^ feet (1.06 to 1.37 metres, nearly) in length, and usually weigh from 1,300 to 1,600 pounds (590 to 726 kilogrammes). Even heavier ingots are sometimes made. The ingots are some- times hammered, but oftener rolled, into blooms of about one-fourth their original section, and are then sent to the rail-train. Where the blooms are rolled, a " three-high mill " is gen- erally employed, which is somewhat similar in general plan to those already described, but which are specially adapted to their work by ingenious and important details designed by their builders or the engineers of the works. In some of these mills, as designed by Mr. George Fritz, the move- ments of the bloom on the table, including the entering of it into the rolls, the lateral shifting, and even the turning of the piece, are all done by steam power. In the United States the rail-train is usually made in three lengths of 21-inch (53 centimetres) rolls, or of two lengths of 24-inch (61 centimetres) rolls; they are "three- high," and are very strongly built; they can deliver 150 tons (152,400 kilogrammes) or more of finished rails per day of 24 hours. The Steel made by this Process is principally used in the manufacture of steel rails ; a considerable amount is employed for boiler-plates, and also for axles and for running parts of machinery. Good steel for either of these purposes is so low (0.20 or 0.15 per cent.) in carbon that it should properly be classed as ingot-iron. The desired strength, toughness, and freedom from liability to harden with changing temperature whgn placed under such conditions, for example, as are met with in boiler-construction and working, can only be secured by care in the selection of the best of ores, and choosing the best of pig-metal for the preparatory processes, and avoiding the introduction of excess of any of the hardening elements. MANUFACTURE OF STEEL. 165 The pig-iron should not contain more than 2 per cent, silicon, usually, although 2j^ per cent, is often allowed, and i>^ per cent, is considered a minimum ; less would cause chilling, or working too cold in the converter. Too much silicon causes rapid wear of linings by producing excessive heat and from waste of iron, and makes the product too hard when the steel retains o.i per cent, silicon or more. The pig-iron should contain less than o.i per cent, sul- phur, and less than one-half that proportion of phosphorus is desirable. Equal care should be observed in selecting fuel and flux. The purer the iron, the higher is the percentage of carbon admissible in the finished steel. Phosphorus is never desirable ; but very good ingot- iron has been made from ores and pig-iron containing a con- siderable amount of that element. Many attempts have been made to produce good mild steels by dephosphorizing iron " high " in phosphorus. Of such processes, those of Heaton and of Henderson are exam- ples. The Ponsard Furnace, with blast nozzles like Berard's, has been used with some success in such attempts, and the use of basic linings in the Bessemer converter with H el- ley's improvements is one of the latest methods. This latter is due to Snelus, and to Thomas and Gilchrist, the former using the magnesian lime obtained from dolomite, and the two last-named chemists using a mixture of lime and silicate of soda ; the same plan has been invented in the United States by Jacob Reese, and at about the same time and independ- ently of the foreign experimenters.* Jeans gives the following diagram, as prepared by Richards, to illustrate the action in such cases.f In this case, the converter was charged with about 13,200 pounds (6,000 kilogrammes) pig-iron and about "jyi, per cent, of lime, and a blast of 25 pounds per square inch (0.176 kilo- gramme on the square centimetre) was applied. The phos- phorus, carbon, and silicon were burned out, as seen in the * The Basic Dephosphorizing Process ; Trans. Eng'r's. Soc, of West. Penn- sylvania, Dec. 2lst, 1880. f Steel ; its History, Manufacture, &=£., London, 1880. l66 MATERIALS OF CONSTRUCTION— IRON AND STEEL. diagram, the silicon going first at the rate of about one-fourth per cent, per minute for nine minutes, and then more and more slowly until, at the end of \T% minutes, it had prac- tically all gone. The" carbon commenced burning at the end of three minutes at the rate of one-fifth per cent, per minute, and was all gone at about the same time that the last of the silicon disappeared, at the end of the blow. 3.5 C \ X_ .Si \, \ s \ £.25 2.0 V \ ^N *\ \ \^ 3 1.5 J.25 1.0 P ^\ s3 \ s \. N s *. \ ^ \ c5 ' -, \^ I, - -. ■--, \ ^ '■^ -- ^ Mi& \% ^ 1 i 3 t 5 6 7 8 10 1! 12 13 U 15 10 17 18 la SO 81 TimA Fig. 49. — Elimination of Elements. Six minutes from the beginning of the blow a mixture of one-third oxide of iron and two-thirds lime, to the amount of 13 per cent, of the charge, was introduced. It was only at the end of 12 minutes that the phosphorus, which amounted to J. 5 per cent, iff the pig-metal, began to leave it, at first slowly then more and more rapidly, until at the end of 16 minutes it was burning at the rate of one-fourth per cent, per minute, leaving almost none at the end of the blow. This charge was overblown — i. e., blown after the carbon had gone — three minutes, at the end of which period, 20)^ minutes from the beginning, the phosphorus had very nearly all bee^ removed. The rails made from this charge are said to have been of excellent quality. Where phosphorus exists in objectionable quantity, and the attempt is thus made to remove it by the use of lime and by substitution of basic for the acid silica linings usually adopted, MANUFACTURE OF STEEL. 167 these basic linings are found far less durable than the acid lining, and the necessity consequently arises for the adop- tion of some method of quickly and cheaply replacing them. By the plan adopted by Holley, this requirement is met. The whole converter, A, is made detachable from a supporting ring, B, which is carried by the trun- nions — as seen in the figure — which latter and the ring are left in place when the converter is taken away for relining. Meantime another converting vessel, which has been previously prepared, is put in place, and the work goes on, while the first is taken away and relined at leisure. The delay for repairs, which would ordinarily reduce the output of American works one- half, is thus avoided, and the use of iron containing a small propor- tion of phosphorus becomes less objectionable. Dolomitic or magnesian limestones are used in the pro- duction of the lime, and this lime, formed into bricks, and baked, is fitted into the converter as a lining. The cost of production is greatly reduced when, as is sometimes the case, the pig-iron is taken to the converter directly from the blast furnace. This has been done even when they are two miles or more apart. The Quality of the Steel produced, which is prima- rily dependent upon the nature of the materials employed, may be improved by either or all of several chemical and physical processes. The use of manganese is customary, in all standard meth- ods of steel making, for the purpose of reducing the porosity of ingots ; it acts by absorbing oxygen dissolved in or mechan- ically mixed with the molten metal, and by antagonizing the ill effects of any sulphur present. Where the metal is made very " low " in carbon, other expedients have been adopted. Fig. 50. — Converter. l68 MATERIALS OF CONSTRUCTION— IRON AND STEEL. as the use of silicon, which may be introduced as a silicate. This element is allowable in small proportion in ingot-iron, and if added in such quantities as to absorb all free oxygen, without combining to a serious extent with the iron, it is found to be of very decided value. The presence of minute quantities of silicon in iron intended to be forged and welded is of advantage, since it acts as an efficient flux. Ingot-iron, containing but small quantities of carbon, may be made of good quality from ores containing moderate quan- tities of phosphorus, Avhich latter element hardens the metal and strengthens it. If present in the iron with but a small total proportion of manganese and carbon, phosphorus does not seriously reduce the ductility. The proportion permissi- ble varies for the usual applications, from O, where the carbon is present in proportion above 0.5 per cent., to 0.25 per cent, where a minimum of carbon is present. Generally, a maxi- mum of 0.15 per cent, is accepted. The sum of all the harden- ing elements is usually specified not to exceed a certain fixed amount. With iron of great purity a high percentage of car- bon may be permitted without incurring danger of seriously injuring the steel by producing cold-shortness or brittleness. Boiler-plate " steel " usually contains less than 0.25 per cent, carbon, and is properly iron ; the same may be said of rail " steel," and of that used for all purposes for which iron was formerly employed. Physical treatment, usually compression, either hot or cold, is, in some cages, practiced, for the purpose of improv- ing the product. Capt. W. R. Jones's method of compression consists in admitting steam at high pressure into a space left above the ingot, the mould being temporarily covered by a steam-tight cap secured by a clamp. The pressure of the steam upon the molten ingot closes up all air-cells, and the thus solidified ingot is found to be homogeneous, and to pos- sess greater strength and toughness, as well as greater den- sity, than uncompressed ingots. The Whitworth method consists in the subjection of the metal, while cooling and solidifying, to the pressure of a pow- erful hydraulic press, and the effect, as will be seen hereafter, is as satisfactory as it is remarkable. MANUFACTURE OF STEEL. 169 Both of these methods of securing solidity and homogene- ousness of the ingot steel are now familiar processes, and are generally considered as standard modifications of the Bes- semer process. Another recent improvement consists in the use of the " soaking pit " which is simply an oven, set in the floor of the mill and lined with good fire-brick, in which the ingots may be placed as soon as cast, and there allowed to attain uniform temperature. This is an economical substi- tute for the reheating furnace. The " Basic Process," by which phosphuretted ores may be used in the manufacture of good grades of steel, the processes of compression just described, and the " soaking-pit " are the most important of the later inventions in connection with Bessemer steel making. Steel Castings are now extensively made by all the steel- making processes, and with a great variety of composition. Their use has been seriously retarded by the difficulties met with in the attempt to insure soundness and freedom from " blowholes," uniformity of product, and homogeneousness in quality. The use of now familiar precautions in melting and pouring, and of various fluxes, have resulted in very great suc- cess in this direction. " Mitis metal " is iron or steel castings fluxed with aluminium. Castings are often made of great size, sometimes approximating 50 tons. The following are figures from tests reported for the U. S. Navy on a casting weighing about 7 tons, "open-hearth " steel: Tenacity, lbs. per sq. in. Elastic limit Elongation Contraction of section. . . . Aluminium in Iron is found to increase its strength and to insure greater soundness. The addition of one-thirtieth of one per cent., three-quarters of a pound to the ton, gives good results with open hearth or Bessemer steel castings. CHAPTER VI. CHEMICAL AND PHYSICAL PROPERTIES OF IRON AND STEEL. Chemically and absolutely pure Iron is probably an unknown substance. Nearly pure iron is made by several methods of manufacture, or it may be deposited by electrol- ysis. Such metal has an exceedingly strong affinity for oxy- gen, sulphur, phosphorus, and some other elements, and alloys readily with many metals which are always present in its ores, or in fuel or flux, in small proportions. As it ap- proximates to the chemically pure condition it assumes more perfectly the character of a silvery white and very lustrous metal, soft, ductile, and malleable, and, though tough, not remarkably strong ; it is very heavy, its density, as deposited by electrolysis, being 8.14; is very easily oxidized; and it is an excellent conductor of heat and electricity. In conse- quence of its greed for oxygen, even ordinary and impure merchant iron requires to be alloyed with some silicon or other fluxing element, to make it weld easily; but, thus alloyed, it welds readily at a bright red or a white heat, at which latter temperature it is in a pasty state. Its melting point is unknown, but is higher than that of commercial wrought iron, for which the melting point is given by Pouillet at about 2,910° Fahr. (1,599° Cent.), and higher than for hard steel (2,533° Fahr., 1,389° Cent.). Alloyed with minute quan- tities of those elements which are usually present in its ores, it becomes harder, stronger, less ductile, and of less density, and is, for purposes of commerce and construction, thus madfi more valuable. The elements which almost invariably con- taminate it or alloy with it are carbon, silicon, sulphur, and phosphorus in small quantities, and traces are found of alumi- num, calcium, titanium, tungsten, and other metals. The PROPERTIES OF IRON AND STEEL. 1 71 purest ores only are used in the manufacture of the best quali- ties of iron and steel. The familiar grades of iron — wrought iron, steel, cast iron — vary in chemical composition from approximate purity to compounds containing 5 per cent, or more of foreign elements, principally carbon and silicon. Commercial wrought iron of good quality contains from 0.05 per cent, of foreign elements, in the softer and purer grades, to 0.30 per cent, or more in the harder irons. Its text- ure is more or less fibrous if made by the processes pro- ducing weld-iron, and it often exhibits some fibre, although from a different cause, even when worked into shape from ingots. In the former case the fibre is produced by the draw- ing out of masses of cinder, inclosed in the sponge, into long lines of non-coherent substance, as the iron is worked under the hammer or in the rolls ; in the latter case each line is the trace of an air-cell, originally of spherical form, in the ingot, but which has been similarly extended in working. ' The fine- ness and silkiness of this fibre, and the general texture of the iron, are gauges of its quality. Magnetism is readily induced in irons, and is stronger as the iron is purer; but it is more permanent as the iron con- tains more carbon or other steel-making elements ; and this property affords another means of determining its quality, and gives some idea of its composition. Oxidation usually occurs less readily as the iron becomes more complex in composition ; but it does not occur at ordi- nary temperatures, even with the best wrought iron, except in the presence of both carbonic acid and moisture ; rust once appearing on polished surfaces, accelerates oxidation. The scales of oxide formed on iron when highly heated differ in composition and physical character from the rust formed under more usual conditions ; it is magnetic, while rust is the peroxide of iron. The former is extremely hard, smooth, bluish-black in color, and is elastic and durable. The Influence of the Elements found in iron and steel is determined, not only by their own character, but, as has already been stated, by their mutual interactions. 172 MATERIALS OF CONSTRUCTION— IRON AND STEEL. Carbon is the most important of all these substances. When added to pure iron, it hardens and strengthens, while reducing ductility and ultimate resilience. It also, in a pro- portion exceeding about one-half per cent, (or less, in pres- ence of other hardening elements), confers upon the steel the property of hardening when suddenly cooled, and of regain- ing its original softness by slow reduction of temperature from red heat — in other words, the property of " taking a temper." Below this limit, as in " boiler steels," wrongly so called, containing* 0.20 to 0.15 per cent, carbon, the metal should soften when suddenly cooled, and this fact furnishes a " test" for such metal. Between ^ per cent, or something above, and i ^ to 2 per cent, carbon, the quality of the steel varies from that of the softest of the " mild " steels, to the hard- est of the tool-steels; passing the upper limit, the metal be- comes too unmanageable and brittle for use even in the harder kinds of tools. Through this whole range the metal is capable of being forged, and can sometimes be welded even when containing as much as one per cent, carbon. The presence of silicon renders welding less difficult. Iron alloyed with over two per cent, carbon is rarely made in the refined state ; all such metal is found in the market in the state of cast iron, which contains all those elements which crude iron brings with it from the ore, and from fluxes and fuels used in its re- duction. Throughout the whole series, the amount of man- ganese, silicon, phosphorus and other " hardening elements " present, have an important influence in determining the char- acter of the steel, and the effect of carbon as well. The steels in the market are usually distinguished from the irons, malleable and cast, by their freedom from phos- phorus, and the completeness with which they have been refined, as well as by the proportion of carbon contained in them. The cast irons grow harder with increase in the propor^ tion of combined carbon ; but, passing a certain limit, they begin to exhibit the influence of graphitic carbon, and become softer and weaker, until, when containing five or six per cent, carbon, one-half of which is often in the graph- PROPERTIES OF IRON AND STEEL. l^l itic State, they become very soft and easily cut, of low den- sity, and of little strength. The value of steel in the market approaches a maximum as it approximates to 0.8 or i per cent, carbon, with freedom from any other elements except manganese and silicon, which should be present in very small quantities. Manganese hardens iron"«nd steel, and, at the same time, usually diminishes its malleability and ductility to a less extent than does carbon. If, however, but little carbon is present, the effect of manganese is quite similar to that of carbon alone, and a steel can be made of very great tenacity, combined with great ductility and resilience. Its effect on metallic iron, in presence of other elements, is not fully deter- mined. It is even considered by some chemists as simply an antidote to the other more injurious elements present, as sul- phur and oxygen, while it is itself a lesser evil. It has usually been found that iron ores containing manganese make excel, lent steel. This element is of value when the steel ils to be forged or welded, and in the various processes of steel making, as a pre- ventive of the formation of oxides that would impede in the former case, and by preventing that porosity which steel cast in ingots would always exhibit, in consequence of their ab- sorption of oxygen, if manganese were absent. Manganese is very effective as a preventive of the hot shortness caused by sulphur, which latter element it counteracts very completely when present in the crude steel in moderate amount. Ingot- iron and steel low in carbon, and especially very low in phos- phorus, is increased in strength and ductility also, by the addi- tion of small doses of manganese, and the degree of hardening on tempering the steels is increased by its presence. Steel containing considerable carbon may take up nearly i per cent, manganese, if otherwise pure, and yet lose little ductility, while gaining considerably in tenacity and in tempering quality. Mushet, pne of the oldest and best authorities, considers that manganese should only be added as an antidote to sul- phur, silicon, and oxygen, and that all these elements, as well as phosphorus, should be kept out of the steel to the utmost 174 MATERIALS OF CONSTRUCTION— IRON AND STEEL. possible extent. In this opinion, Siemens and other later authorities agree, and all unite in stating that toughness can best be secured by insuring purity of metal, whether iron or steel. Recent practice does not, however, follow this view, as the manufacture of a " manganese steel," containing little car- bon, has become a well-established branch of steel making. Its presence in spring-steel has been found objectionable. Phosphorus is the most injurious of all the elements which usually contaminate iron and steel. It confers hardness; but that quality can be far more satisfactorily obtained when desired, by the addition of other elements. It greatly reduces ductility, and causes a serious degree of cold- shortness in both iron and steel ; even in foundry grades of cast iron it should usually not be admitted in higher propor- tion than yi per cent,, except when fluidity is desired even at the expense of considerable loss of strength. Cast-iron containing phosphorus is peculiarly liable to break under shock, and the same is true of wrought iron and steel in which it may exist in measurable amount, unless it be the only hardening element present in any notable quantity. Good tool steel should not contain more than o.oio or 0.015 Per- cent., but ingot-iron and mild steel may contain, if otherwise pure, as much as o.i per cent., which amount has actually been found in some spring-steel. When nearly free from all other hardening elements, iron sometimes contains 0.35 to 0.40 per cent, phosphorus, and yet exhibits fair quality. The effect of this element is to increase elasticity, to ele- vate the elastic limit, and to increase slightly the modulus of elasticity. Wedding exhibits the method of variation of the maxi- mum allowable percentage of phosphorus in iron and steel, varying in proportion of carbon, by a curve shown in full line in the figure. One-fourth the proportions thus give», as exhibited by the dotted curve, T, introduced by the Author, may be taken as the usual limit for metals of good reputation in our markets. Some few cast irons, having a reputation for fluidity when melted, contain higher proper- PROPERTIES OF IRON AND STEEL. 175 Wrought Iron «o/r /ot tions. Wade's experiments on metal for ordnance indicate y^ per cent, phosphorus to be the maximum allowable, and y^ per cent, to be a usual proportion in such cast iron as is used for guns ; such iron con- - casTjior^ tains about 3 per cent, carbon. Sulphur caus- es brittleness at high tempera- tures in all grades of iron and steel. Its effect is most marked in the absence of other impurities, and it is therefore -^ \ \w \ ^T \ "X \ ^ "---, _^ yi Phosphorus vs.Carbon Fig. 51. most objection- o/,c able in the finer ' grades of tool steel, and in iron or mild steel, which is to be welded. Its antidote is manganese, which greatly reduces its ill effect. Iron and steel containing 0.2 per cent, are not easily welded, and this figure maybe taken as the extreme maximum allow- able in any malleable iron or steel. When bar-iron, containing a considerable amount of sul- phur, is converted into steel by the method of cementation, a considerable amount of that element may be eliminated as bisulphide of carbon, and the steel made may be of fair quality. Using iron of a good degree of purity, the steel be- comes almost perfectly free from sulphur, and particularly is this the case where manganese is present. Since the effect of sulphur is to produce hot-shortness only, its presence in moderate quantity is not objectionable in either iron or steel castings. It is stated by some authorities that iron and steel castings are stronger and tougher when containing sulphur, and it has even been added in making iron castings, both for 176' MATERIALS OF CONSTRUCTION— IRON AND STEEL. this reason and because it gives greater fluidity to the molten metal ; ordnance iron has sometimes been thus treated. Sulphur is eliminated to some extent in the Bessemer converter when the basic process is adopted. Silicon is a hardening element, and its influence is the greater as the metal is otherwise purer. In weld-iron and in soft steels it is of value in small quantity where the metal is to be welded, either in the process of manufacture or in sub- sequent forging. In the various processes of making ingot- steel it is of use, like manganese, as a preventive of injury by oxidation, or by the formation of air-cells and production of porous ingots, and, consequently, of non-homogeneous blooms, bars, or rails. In the pneumatic process its presence has been seen to be of direct advantage in supplying fuel needed to elevate the temperature of the molten mass in the converter to a higher point than could be obtained by com- bustion of carbon alone. It is of further use, in some cases, if added after the blow, as it gives a sound ingot and a weld- able steel. One-twentieth of one per cent, is said to be suffi- cient to produce a decided beneficial effect in this latter case. In its hardening power silicon resembles carbon, which element it may replace to a limited extent. Good irons and steels usually contain from o.io to 0.25 per cent, of sili- con ; while cases are reported in which* in the absence of other hardening elements, between one and two per cent, of silicon has been found in iron of apparently good quality. Bessemer cast irons are expected to contain from 2 to 2)^ per cent, silicon, but foundry irons should contain very little ; when the latter contain an excess they become weak, brittle, and peculiarly liable to crack in large castings ; this latter defect is exaggerated by the presence of phosphorus. Two per cent, may be taken as the usual limit for silicon in cast iron, one-half per cent, for the steels, and one-fourth or one- fifth per cent, for good wrought irons. Nitrogen is said by many chemists to be invariably a con- stituent of all grades of metal which exhibit steel-like proper- ties, and it is by some considered an essential constituent ; it has been found in all grades of both iron and steel in propor- PROPERTIES OF IRON AND STEEL. 1 77 tions varying from o.oi to 0.20 per cent., the usual proportion being about 0.05 per cent. All methods of steel-making include the exposure of the iron used to the action of nitrogenous compounds or of nitro- gen in the gaseous condition; but the element has been found in wrought iron in higher proportion than in steel, and it seems not to be, in itself, a steel-making element. All the familiar metals have been alloyed with iron, but the resulting product has in no case become commercially in- troduced. Nickel has been alloyed with iron, and iron has been added to bronze and to brass ; in each case the allloy has exhibited some peculiar properties, but has not come into general use. So-called titanium and silicon steels have been made ; but analysis has not shown them to contain either metal in any notable quantity, and such peculiar qualities as they may have exhibited have been supposed to be due to the presence of other elements. Tin alloys freely with iron, but no use has been made of the product. Nickel and cobalt alloy with iron and steel, giving them a whiteness and lustre, and without seriously im- pairing ductility when added in small doses. Gold and pla- tinum will unite with iron and steel, and the latter was found by Faraday to add strength and to give fineness of grain when not exceeding in amount one per cent, of the whole. Antimony injures iron by making it brittle and difficult to work. Copper has been found by the Author to strengthen and toughen steel, when added in very small quantity, and Tred- gold* states that it has a similar effect upon cast iron. Wrought iron containing some tenths per cent, of copper, is red-short ; in some of the best irons from Siberia were found from 0.01 to 0-03 per cent, of copper, and in some specimens of steel 0.2 per cent. ; this steel was not brittle, and had been used with success for manufacturing steel axles. The presence of copper was noted in several specimens of cast iron coming j2 '" Tredgold on Cast Iron. 178 MATERIALS OF CONSTRUCTION— IRON AND STEEL. from blast-furnaces of the South Oural mountains. These specimens, when examined and analyzed, showed that the presence of copper in cast iron may amount to a higher per- centage than in steel or iron without injuring the quality of the metal. The specimen examined was used for castings ; it filled up the moulds well, and had a fine appearance ; when freshly cut it had a dark-gray color. Under the microscope small grains of copper were easily seen in the mass of the metal. This cast iron had the following average composition : * Per Cent. Iron 83.514 Copper 8 . 123 Tin 1.252 Cobalt o. 501 Silicium 0.952 Tungsten 0.125 Carbon 3 . oor Manganese 2.312 99.780 The case illustrates well both the variety and extent to which other elements may enter into union with iron. The Chemical Composition of the various classes of iron and steel has already been indicated as determining their nomenclature and uses. Beginning with the most impure of the cast irons and passing through the several grades of cast iron, steel, and malleable or wrought iron, it has been seen that their qualities and their applications are generally deter- mined primarily by the proportions of carbon present, and secondarily by the effect of sulphur, phosphorus, chromium, manganese, silicon, and the other less usual ingredients which give them peculiar characteristics. Of the Cast Irons, " No. i Foundry Iron " is the softest grade, is richest in carbon, and is the darkest in color of §11 the irons ; it is weak, moderately tough, of low density, and is quite fluid when molten. It is used principally for mixing with harder grades or for purposes which compel repeated or * Chem. News. PROPERTIES OF IRON AND STEEL. 1/9 prolonged fusion. It is the most -expensive of all grades of cast iron. It is to a very slight extent malleable, ductile, and somewhat flexible, is very easily worked by the file and by cutting tools. Its fracture is bright and granular, and of a bluish-gray color. Its texture is finer and more close-grained as its color is lighter. When melted it has much more flu- idity than the lighter grades, flows smoothly, and fills the moulds well, taking a good impression from the minutest lines of the mould, and rarely causes trouble by " cold shuts " or "blow-holes." The best qualities have a medium fineness and closeness of texture; a clear, dark-gray color; a clean, brilliant fracture, with sharp edges ; and a density that is not far from 7.2. Coarseness of grain, a dull color, and irregu- lar structure, indicate an inferior iron. When annealed, gray cast iron is softened, weakened, and reduced in den- sity. " No. 2 Foundry Iron " contains less carbon than No. i, is harder, stronger and denser, and has a finer, closer grain ; it is of more frequent use than either of the other foundry grades, and is the iron most called for in all ordinary kinds of work. Its specific gravity is about 7.3. " No. 3 Foundry Iron " is still lower in carbon, is whiter, stronger, denser, and finer in grain. ^ It is too hard and brittle for general use, and is purchased to mix with softer irons. It often has a slightly mottled surface of fracture. The Forge Irons are numbered 4, 5, and 6 by many mak- ers ; of these the first is often called " Bright Iron," the second " Mottled," and the third " White." White iron is very hard and brittle, but strong and dense, attaining a spe- cific gravity of 7.5 or more. It cannot be easily filed, but takes a very high polish if ground. Its fracture is clean, bright, and silvery white, usually granular, but sometimes plainly crystalline. It burns with bright scintillation at the melting temperature. Anneahng reduces its density, but increases its strength. The forge irons are generally converted into wrought irons by the puddling process. The rich gray irons containing sili- con are converted by the pneumatic method. The carbon l8o MATERIALS OF CONSTRUCTION— IRON AND STEEL. contained in white iron is all combined, as it is in steel ; that in the gray irons is, to a considerable extent, graphitic. The appearance of the fracture and the physical qualities of the darker grades of iron are considerably modified by vari- ations in the size of the castings made of them, and in the rate of cooling. A large casting cooling slowly retains much of its carbon in the graphitic condition, while a small piece rapidly cooled will become whiter, and will exhibit a " chill," produced by more complete combination of its carbon with the iron. This change affects the superficial portions of iron cast in contact with " chills," or masses of iron set in the mould for the purpose, to an extent which is determined by the grade of the iron and by the amount of silicon and carbon present. For example, a well known brand of " chilling iron " is thus numbered : No. I. — Soft; does not chill. No. 2. — Harder ; does not chill. No. 3. — Still harder ; does not chill. No. 3^. — Just shows a chill on the surface. No. 4. — Chills to the depth of 5^ to ^ inch. No. 4j^. — Chills to the depth of i to 2 inches. No. 5. — Mottled iron. No. 6.— White iron, " all chill." The depth of chill is determined in each grade by the facility with which the carbon may be changed from the graphitic to the combined state. Analyses of the several Grades of Cast Iron have been made in great numbers and with extreme accuracy. Examples of good foundry irons are the following, made from magnetic ores : Carbon.... 4.81 3-94 Silicon 1.18 2.43 * Sulphur trace 0.04 Phosphorus 0.12 0.04 Manganese 0-99 o. 1 1 Iron and loss 92.90 93-44 100.00 100.00 FKOPERJIES OF IRON AND STEEL. l8l Of irons made from red hematite, Abel* gives the fol- lowing : ELEMENTS. Carbon, combined Carbon, graphitic. Silicon Sulphur Phosphorus Manganese Arsenic Copper o. 3.22 3.02 o. 0.06 O.IX trace trace 2. trace 2.24 2.77 CO! 0.05 0.07 trace trace trace 2.30 2.72 0.05 0.05 trace trace trace 0.3S 1.86 2.63 o.io 0.03 0.07 trace trace Cold and Hot Blast Irons differ considerably in quality, and this difference is so marked and so generally well understood that the market prices of iron made by the two methods differ greatly. The higher temperature of furnaces having a hot blast causes a more complete deoxidation of the ores and the reduction of elements which are less readily deoxidized at the lower temperatures of cold-blast furnaces. The effect of heating the blast is, therefore, to cause loss of quality by increasing the proportion of deleterious elements reduced, and which combine with the iron, while greatly increasing the yield of the furnace, and decreasing the cost of fuel. When the finest quality of iron is demanded, pure ores, fuel free from sulphur and phosphorus, and flux equally pure, must be used. Hence " Cold Blast Charcoal Iron " is demanded in many cases, to the exclusion of other grades. It has been stated by some writers that the amount of phosphorus is greater in hot than in cold-blast iron. This is considered by Percy and other chemists, and by experienced furnace-men a mistake, as it is found that all phosphorus goes into the iron in any case. Charcoal, Coke, and Anthracite Irons differ in value for the same reason that cold-blast and hot-blast irons differ. Coal, as mined, usually contains some impurities, and some kinds of bituminous coal are very seriously contami- nated by sulphur. Anthracite, used as fuel in the blast fur- nace, while cheap in certain localities, and convenient to handle, and while giving intense heat, has some objectionable * "Cast Iron Experiments." i82 MATERIALS OF CONSTRUCTION— IRON AND STEEL. qualities ; the " anthracite irons " are, therefore, often found to be of unsatisfactory character. Bituminous coals are sometimes used " raw " in the furnace, and the " raw coal iron " thus made is often very hot-short, in consequence of the presence of an excessive amount of sulphur. To secure immunity from this injury, the bituminous coals are usually coked, and the iron made with coke is, usually, if the flux is free from phosphorus, of good quality. Charcoal has the least proportion of injurious elements of all the fuels used in making iron in the blast furnace, and the charcoal irons are, therefore, of better quality, other things being equal, than the other kinds of cast iron. Charcoal fur- naces are also usually small, as this fuel is too weak to carry a heavy burden, and the temperature attained within them is less likely to become excessive. They are usually supplied with a blast that is either cold or very moderately warmed — a circumstance which aids in securing excellence of quality of product. The Chemical Change produced by Puddling, and usually in the manufacture of either weld or ingot iron, has been described as a removal of the impurities contained in cast iron, the principal of which are carbon and silicon. This process is traced by Dr. Hartmann* in a series of analyses which are, in part, given below. Beginning with a fair quality of No. 3 cold-blast gray pig- iron made with coke, and having the composition : 12 M. FIRST ANALYSIS. Per cent. SECOND ANALYSIS. Per cent. AVERAGE. Per cent. Carbon Silicon 2.320 2.770 0.580 . 0.318 traces 94.059 2.230 2.670 0.710 0.228 traces 94.059 2.274 2.720 0.645 0.302 traces ^ 94.059 Sulphur Manganese and Aluminum Iron Total 100.047 99-957 * Waltz und Puddle Meister. i PROPERTIES OF IRON AND STEEL. 183 The charge of 200 pounds (100 kilogrammes) became fully melted in forty minutes. The silicon had already begun to burn out very rapidly, and the carbon had remained unchanged. One hour from the time of charging the sample taken out contained Carbon. Silicon. Original pig 2.274 2.720 First sample, 12.40 p.m 2.726 0.915 Second sample, i.oo p.m 2.905 0.197 The carbon had increased 0.629, whilst the silicon had diminished over 90 per cent. A second sample contained I P.M. First Analysis. Second Analysis. Average. Carbon 2.910 2.900 2.905 Silicon 0.226 0.168 0.197 The first differed from No. i in being slightly malleable when hot, whilst No. i was brittle. The cinder remained after cooling on the surface, and not mixed with the metallic iron, as in the second analysis of sample No. 3, which was taken out five minutes later. Boiling soon commenced, and the sample next analyzed contained a considerable amount of cinder, which was with difificulty removed. It contained i.io P.M. First Analysis. Second Analysis. Average. Carbon 2.468 2.421 2.444 Silicon 0.188 0.200 0.194 This sample consisted largely of minute globular grains adhering to each other and to the slag with which they were mingled with a strong glutinous adhesion, even when very hot. The grains were black, lustrous, and very brittle. Eighty minutes from the beginning, a sample taken out for analysis somewhat resembled the last. While cooling it, little jets of blue flame were seen to burst out from it. The grains were finer than before, and their coherence was so slight that the mass was easily broken up. Their color was a lustrous black, their fracture silvery white, and the metal 1 84 MATERIALS OF CONSTRUCTION— IRON AND STEEL. was as brittle as glass. The carbon and silicon contained were as below : 1. 20 P.M. Carbon Silicon First Analysis. Second Analysis. Average. 2.335 2.376 2-355 0.187 0.178 0.182 The boiling was in full operation when the above sample was taken, the heat had been reduced, and the puddler was working the bath with his rabble. The molten mass had swollen to four times its original volume. The silicon was still passing off, and the carbon had become somewhat reduced in amount. At 1.35 P.M., ninety-five minutes from the beginning, boil- ing ceased, and the mass began to shrink in volume. The carbon had ceased to burn rapidly, and the bubbling carbon monoxide no longer puffed up the semi-fluid iron. The damper had been nearly closed, the flame had become strongly charged with smoke, and the puddler had begun to " ball up " the sponge. The sample now taken contained : 1.33 P.M. Carbon . Silicon . First Analysis. Second Analysis. Average. 1. 614 1. 681 1.647 0.188 0.178 0.185 It had lost a large amount of carbon since the preceding sample was taken, but the silicon had been but slightly changed. The sample was somewhat malleable, and could be beaten flat and smooth by the hammer. In five minutes more a sample taken out gave : 1.40 P.M. Carbon Silicon First Analysis. Second Analysis. Average. 1.253 I.l6o 1.206 0.167 0.160 0.163 In this case the appearance was similar to the last. Blue flames of carbonic oxide appeared while it was cooling; its grains had increased in size and could be welded together. The slag was more easily separated than before. Just at this time the mass in the furnace was divided, the metal being PROPERTIES OF IRON AND STEEL. 185 separated from the great mass of cinder by the puddler, as he made up his ball. A sample taken out at 1.45, after 105 minutes' working, contained still larger grains, and analysis showed the rapid loss of carbon, already noted, to be still going on. The grains were decidedly more malleable than before. This contained : 1.45 P.M. First Analysis. Second Analysis. Average. Carbon 1. 000 0.927 0.963 Silicon 0.160 0.167 0.163 Five minutes later the metal contained : 1:50 P.M. First Analysis. Second Analysis. Average. Carbon 0.771 0.773 0.772 Silicon 0.170 0.167 0.168 The grains were still larger, more coherent, and more malleable and tenacious. Each grain was coated with slag, and was thus perfectly protected against oxidation, as was shown by the fact that portions of the sample remained in the laboratory unoxidized many months. The puddle-ball was next hammered, and rolled into a bar, which was found to contain : First Analysis. Second Analysis. Average. Carbon 0.291 0.301 0.296 Silicon 0.130 o.iio t..l20 Sulphur 0.132 0.126 0.134 Phosphorus 0.139 •••• 30 12.68 3-8o 3.26 3-26 1-95 1. 00 -36 •25 .22 ■15 .10 .06 .OlJ I1.O3 .96 -52 ■ 43 -43 ■27 .24 .12 .10 ■ -05 .036 .028 .008 .007 .007 .0043 T. T. Uranium Gold Fused Tellurium .... Chromium Platinum Manganese Molybdenum T. T Magnesium Wire and tape Globules T. T. Silver. Aluminium* Cobalt Bar Cubes S. S. T. T Nickel Cadmium Sodium T Crude S. T Tin Arsenic Zinc Lead Iron The prices of many may be considered also as " fancy prices," and a whole pound of some of the metals named could hardly tie obtained at even these figures. In compiling the table, the prices of the rarer metals are obtained from Trommsdorff's and Schuchardt's price lists ; the avoirdupois pound is taken as equal to 453 grammes, and the mark as equal to 24 cents gold. It is evident that the prices of the metals bear no relation to the rarity of the bodies whence they may be derived ; for calcium, the third in the list, is one of the most abundant elements. Even indium, the most recently discovered ele- ment, stands tenth in the list, below strontium. * Since 1884 reduced by new processes to between $0.50 to %\ per pound. CHAPTER VIII. THE BRONZES AND OTHER COPPER-TIN ALLOYS. The Alloys of Copper, with smaller quantities of the more common metals, are the most valuable and the most common, and the most extensively used of all compounds or mixtures known to the engineer and the metallurgist. Those which are produced by the union of copper and tin are generally classed as the " Bronzes." When copper is alloyed with zinc, the composition is known as " Brass." These terms are not exclusively so applied, however, and the term brass is not infrequently used to cover the whole series of alloys com- posed, wholly or in part, of alloys of copper and tin, copper and zinc, or combinations of brass and of bronze with each other or with less quantities of other metals. Bronzes are here sup- posed to contain principally copper and tin. These alloys are produced by the union, either chemically or by solution, when molten, of two or more metals. Nearly all metals can unite with nearly all other metals in this manner, and the number of possible combinations is infinite ; nevertheless, but few alloys are found to be very generally used in the arts. It is consid- ered probable that the metals may combine chemically in definite proportions, but the compounds thus produced usually dissolve in all proportions in either of the constituents, and it is rarely possible to separate the chemically united portions. In some cases the affinity is very slight, as between lead and zinc, either of which will take up but about one and a half per cent, of the other. The alloys are usually the more stable as their constituents are the more dissimilar, and, when this dif- ference is chemically great, the compound becomes brittle. Occasionally, an alloy is formed which gives evidence of the occurrence of chemical union, by the production of heat ; this is seen in some copper-zinc alloys. BRONZES AND OTHER COPPER TIN-ALLOYS. 271 Copper alloys are formed with nearly all metals with great facility, and with no other precaution than that of either preventing access of oxygen to the molten mass, or of thor- oughly fluxing the alloy, to take up such as may have com- bined with it. Many of these alloys were once considered chemical compounds ; but the view which seems most gener- ally accepted, at the present time, is that they are almost in- variably either mere mixtures, or that a species of solution of the one metal in the other takes place. The most minute trace of foreign element often produces an observable, or even an important, alteration of the proper- ties of copper. This is especially true of its conductivity for electricity, which is reduced greatly by an exceedingly minute proportion of iron or lead. History. — The alloys of these metals were used ex- tensively by the ancients for coins, weapons, tools and orna- ments, and the composition of their bronzes, as shown -by recent analyses, indicates that they were as skilful in brass- founding as the modern workman. Thus, Phillips gives the following as the results of his own examinations and as showing the proportions of the con-stit- uents employed in the manufacture of brass, at times both preceding and closely following the Christian era : DATE. i 8 ZINC. TIN. LEAD. IRON. Large brass of the Cassia family. . " Nero " .. " Titus " .. " Hadrian " .. Faustina " .. B. C. 20 A.D. 60 " 79 " 120 " 165 82.26 81.07 83.04 85.67 79-14 17-31 17.81 15-84 10.85 6.27 1.05 1. 14 4-97 1.73 9.18 • 35 -50 -74 -23 Thus, copper and zinc were the essential constituents of the alloys examined ; but then lead was sometimes present in considerable quantities, together with tin and iron. Although zinc occurs in such considerable quantities in these alloys, it 272 MATERIALS OF CONSTRUCTION— NON-FERROUS METALS. was not known in the metallic state until about the thirteenth century, when it was described by Albert of BoUstadt. Many analyses of ancient articles of bronze have been made, and our knowledge of this very old alloy is consider- ably greater than that of the alloys of zinc. The proportion of the constituent metals was varied according to the purpose to which the alloy was to be applied, as will be seen from the following analyses, the hardness being modified according to the proportion of tin present. The alloys containing the largest amount of tin were used for mirrors, while those of medium hardness were used for sword-blades and other cut- ting instruments : 1. Chisel, from ancient Egyptian quarry 2. Bowl, from Nimroud 3. Bfonze overlaying iron 4. Sword-blade, Chertsey, Thames 5. Axe-head 6. Celt 7. Roman As, b.c. 500 8. Julius Caesar , COPPER. TIN. LEAD- IRON. COBALT. 94.00 5.90 .10 89-57 10.43 88.37 ".33 89.69 o.ss •33 88.05 11.12 .78 81.19 18.31 69.69 7.1b 21.82 •47 -57 79-13 8.00 12.81 Wilkenson. Dr. Percy. J. A. Phillips. Prof. Wilson. J. A. Phillips. The third specimen was analyzed by Dr. Percy, who de- scribes it as a small casting in the shape of the foreleg of a bull, forming the foot of a stand, consisting of a ring of iron supported upon three bronze feet. A longitudinal section disclosed a central core of iron, around which the bronze had been cast. Some writers, to account for the immense masses of hard stone wrought by the Egyptians and ancient Americans, sup- pose that they possessed means of hardening bronze to a degree equal to that of our steel ; this requires confirmation, since no remains of bronze of such a hard variety have ever been discovered. The bronze weapons discovered by Dr. Schliemann amdhg the ruins excavated by him at or near the site of ancient Troy* were often of nearly the composition of modem gun- bronze ; they contained copper 90 to 96, tin 8.6 to 4. The date, * " Troy and its Remains ;" London and New York, 1875 ; p. 361. BRONZES AND OTHER COPPER-TIN ALLOYS. 273 archaeologlcally, is at the beginning of the " bronze age," and immediately at the close of the "stone age." Sir John Lubbock finds the bronze implements and ornaments of the bronze age as remarkable for their beauty and variety as for their utility.* They consisted of axes, arrow-heads, knives, swords, lances, sickles, ear-rings, bracelets, rings, etc., etc. The bronze used by the prehistoric nations contained no lead; that of the Romans and post-Romans was rarely of pure copper and tin, but were usually more or less alloyed with lead. Silver, zinc, and lead was not known in the bronze age. The prehistoric bronzes were cast, sometimes in metal or in stone, and sometimes in sand, moulds. A more common method was by wax models, or " patterns," which were used to make the desired cavity in an earthen or sand mould, the wax being melted out afterward. According to Charnay,f the Aztecs discovered a means of tempering copper, and of giving to it a considerable degree of hardness, by alloying it with tin. Copper hatchets were known among them ; since Bernal Diaz states in the narrative of his first expedition to Tobasco, that the Spaniards bartered glass-ware for a quantity of hatchets of copper, which at first they supposed to be gold. Copper abounded in Venezuela, and we still find there in great numbers trinkets of copper mixed with gold, or of pure copper, representing crocodiles, lizards, frogs and the like. In cutting down trees, they employed copper axes like our own, except that, instead of having a socket for the haft, the latter was split, and the head of the axe secured in the cleft. The hatchet described seems to have been a piece of native copper wrought and fashioned with a stone ham- mer. The Aztecs made good bronze chisels, as described by Sefior Mendoza, director of the National Museum of Mexico. He describes certain specimens of bronze chisels belonging to the collection in that museum. When freed from oxide the bronze presents the following characteristics : In color it resembles gold ; its density is 8.875 ! i* is malle- * " Prehistoric Times ; " London and New York, 1872. f N. A. Review, 1875 ; Ruins of Central America. 18 2/4 MA TERIALS OF CONSTRUCTION— NON-FERROUS METALS. able, but unlike pure copper, is hard, and breaks under strong tension or torsion ; the fracture presents a fine granulation like that of steel ; in hardness, it is inferior to iron, but it is sufficiently hard to serve the purpose for which it was in- tended. One of these chisels was found to consist of copper 97-87 per cent., tin 2-13 per cent., with traces of gold and zinc. The bronzes were used by the ancients in the manufacture of weapons and of tools. The use of phosphorus increases the purity and adds strength and hardness to these alloys, and the remarkable hardness of ancient bronze weapons is found by Dr. Reyer to be due, in part at least, to the presence of phosphorus, probably introduced with the flux used in melt- ing. The proportion of tin varied up to 20 per cent. The Alloys of Copper and Tin have many uses in the arts. The two metals will unite to form a homogeneous alloy in a wide range of proportions. As tin is added to pure cop- per, the color of the alloy gradually changes, becoming decidedly yellow at 10 per cent, tin and turning to gray as the proportion approaches 30 per cent. In the researches conducted by the Author, it was found that good alloys may contain as much as 20 per cent. tin. When the color changes from golden yellow to gray and white, the strength as suddenly diminishes ; and alloys containing 25 per cent, tin are valueless to the engineer ; nevertheless, this alloy and those contain- ing up to 30 per cent, show compressive resistances increas- ing to a maximum. The tensile and compressive resistances have no known relation ; the torsional resistance is more closely related to tenacity. A small loss of each constituent occurs in melting, the loss often being highest with the metal present in the lowest proportion ; this loss rarely exceeds one per cent., except when the fusion has taken place slowly with exposure to the air, when considerable copper-oxide is liable to form. The specifi'c gravities of these alloys do not differ much from 8.95. Under 17.5 per cent, tin, the elastic limit lies between 50 and 60 per cent, of the ultimate strength ; beyond this limit the proportion rises, and at 25 per cent, tin the elastic limit BRONZES AND OTHER COPPER-TIN ALLOYS. 275 and breaking point coincide. Passing 40 per cent, tin, this change is reversed and the elastic limit, although indefinite, is lowered until pure tin is reached and a minimum at about 30 per cent. The modulus of elasticity of all the bronzes lies between ten and twelve millions. Riche states that tempering produces on steel, forged or annealed, an inverse effect to that which it produces on bronzes rich in tin ; it diminishes its density instead of in creasing it, from which it may be seen that tempering diminishes the density of annealed steel and makes it hard, while tempering increases the density of annealed bronze and makes it soft. There is always an increase in density, whether the bronzes rich in tin be tempered, or slowly cooled, after compression. These experiments confirm most clearly the fact affirmed by D'Arcet, that tempering softens the bronzes, rich in tin, for we can flatten in the press the tempered bronzes, while it is impossible to do this with steel. It is evident from his experiments that tempering aug- ments considerably the density of bronze rich in tin, and that annealing evidently diminishes the density of tempered bronze. Still the effect of slow cooling by no means destroys the effect of tempering, for the density continues to increase till it becomes remarkable. While all mechanical action increases the density of the annealed bronze, it very slightly, but still sensibly, diminishes the density of annealed steel, and, on the whole, tempering and shock increase the density of annealed bronze, while they diminish the density of annealed steel. But the variations are very decided for bronze and very slight for steel. Bronze of 96 and 97 parts copper may be employed to great advantage, and with no serious inconvenience, in the manufacture of medals. Its hardness, much less than that of the alloy of M. de Puymaurin, does not much exceed that of copper ; it possesses a certain sonority and casts well, rolls evenly, and its color is more artistic than that of copper. 276 MA TERIALS OF CONSTRUCTION— NON-FERROUS METALS. The action of the press and of heat modify its density but little. Properties. — Copper and tin alloy in all proportions, and the most useful compounds known to the engineer are the " bronzes," as these alloys are called. They include gun- metal, bell-metal and speculum alloys. The following is Mallet's list of these alloys and table of their properties.* PROPERTIES OF COPPER- TIN ALLOYS. At. wt. : Cu. —31.6 ; Sn = 38.9. AT. COMP. COPPER. b. Lf. COLOR. FRACT. TENACITY. MALL. HARD. FUS. Cu Sn per ct. Tons per sq. in. I 100. , 84.29 8.607 8.561 red-yellow 24,6 I 2 10 8 16 a 10 fine graia 16 I 15 b g 82.81 8.462 yellow-red " 15 2 3 5 14 c 8 Si.io 8.459 " 41 17 7 4 4 13 d 7 78.97 S.728 pale red vitreous 13 6 5 3 12 e 6 76.29 8.750 " " 9 7 brittle 2 II f S 72.80 8.57s asli gray conchoid. 4 9 " I 10 i 4 6B.21 8.400 dark gray " 7 friable 6 9 A 3 61.69 8.539 white gray *^ 5 " 7 8 I 2 51 -73 8.416 white lam. g^aiu I 7 brittle 9 7 J I 34.92 8.056 li vitreous I 4 '^ II 6 k I 21. IS 7.387 *' lam. grain 3 9 " 12 S / I 15.17 7.447 " (t 3 I 8 tough 13 4 m I 11.82 7.472 *■' *' 3 I 6 '" 14 3 n I 9.68 7-442 " earthy 2 5 7 15 2 0. 7.291 2 7 16 I tty ^, c are gun-metals ; dy hard brass for pins ; c, y, ^, A, r, bell-metal ; j^ k, for small bells ; /, 7n, n^ o^ are speculum alloys. The addition of a small quantity of tin to copper causes it to become brittle under the hammer, according to Karsten, and the ductility is restored only by heating to a red heat and suddenly cooling. Mushet finds that the alloy, copper 97, tin 2, makes good sheathing, as it is not readily dissolved in hydrochloric acid. The best gun-metal is from copper go, tin 10, to copper 91, tin 9 ; if richer in copper, it is especially liable to liquation, which action is detrimental to all these alloys. Bell-metal, copper 80, tin 20, to copper 84, tin 16, is sonorous and makes good castings, but is hard, difficult to * Dinglet's Journal, Ixxxv., p, 378 ; Watts's Diet, ii., p. 43. BRONZES AND OTHER COPPER-TIN ALLOYS. 277 work and quite brittle. Suddenly cooling it from a high temperature reduces its brittleness, while slow cooling re- stores its hardness and brittleness. It is malleable at low red heat and can be forged by careful management. Speculum-metal, copper 75, tin 25, is harder, whiter, more brittle and more troublesome to work than bell-metal. Old flexible bronzes contain about ^ ounce of tin to the pound of copper, or copper 95, tin 5, as stated by Ure. Ancient tools and weapons, as shown elsewhere, contain from 8 to 15 per cent, tin; medals from 8 to 12 per cent., with often 2 per cent, zinc to give a better color. Mirrors con- tained from 20 to 30 per cent, tin. The metals mix in all proportions, and the alloys are, to a certain extent, independ- ent of their chemical proportionality. The occurrence of hard, brittle, elastic alloys between the extremes of a series having soft tin and ductile copper at either end, both of which metals are inelastic, is probably a proof that these alloys are sometimes chemical compounds. They are proba- bly, usually, compounds in which are dissolved an excess of one or the others of the components. The Principal Bronzes are those used in coinage, in ordnance, in statuary, in bells, and musical instruments, and in mirrors and the specula of telescopes. These alloys oxid- ize less rapidly than copper, are all harder, and often stronger and denser. Coin bronze, as made by the Greeks and Romans, con- tained from copper 96, tin 4, to copper 98, tin 2, and Chaudet has shown that the first of these alloys can be used for fine work, obtaining medals of this composition of very perfect polish while sufficiently hard to wear well. Puymaurin succeeded well with alloys of copper 93.5, tin 6.5, to copper 90, tin 10; and Dumas found the range of good alloys for this purpose quite large, varying from 96 copper, 4 tin, to 86 copper, 14 tin, but the best falling near the middle of this range. Gun bronze has various compositions in different countriesi The most common proportion would seem to be copper 90, tin 10, or copper 89, tin 11. Well made, it is solid, yellowish, denser than the mean of its constituents, and much harder, 278 MATERIALS OF CONSTRUCTION— NON-FERROUS METALS. •stronger, and more fusible than commercial copper; it is somewhat malleable when hot, much less so when cold. It is subject to some liquation, and should therefore be quickly chilled in the mould; it loses some tin when per- mitted to stand at a temperature of 400° to 500° Fahr. (200° to 260° C). This liquation gives rise to light-colored spots throughout the metal. This bronze does not readily oxidize at ordinary temperatures, but is quickly attacked when hot; it usually becomes greenish when exposed to the weather, by the formation of the hydrated carbonate ; thus " patina " is ob- served on all unpolished old bronze guns or old statues. Statuary bronze is usually of nearly the same composition as gun-bronze. It should be rapidly melted, poured at high temperature, and quickly cooled to prevent liquation. Bell-metal is richer in tin than the preceding, and varies in composition somewhat with the size of bell. The propor- tion, "JJ copper, 23 tin, is said to be a good one for large bells ; it shrinks 0.015 i^i the mould while solidifying. The range of good practice is found to be from 18 to 30 per cent, tin, 82 to 70 per cent, copper; the largest proportions of tin are used for the smallest bells, and an excess is added to meet the liability to oxidation and liquation ; copper 78-82, tin 22-18, is a very usual composition. When made of scrap metal, as is not uncommon, serious loss of quality is liable to occur by the introduction of lead and other metals deficient in sonorousness. When properly made, this, alloy is dense and homogeneous, fine»grained, malleable if quickly cooled in the mould, rather more fusible than gun-bronze, but otherwisequite similar; excelling, however, in hardness, elasticity and sonority. These bronzes become quite malleable when tempered by sudden cooling, and this treatment is resorted to when they are to be subjected to prolonged working or to a succession of processes. Chinese gongs are made of copper 78 to 80, tin 22 to 20, and are beaten into shape with the hammer, the metal being softened at frequent intervals by heating to a low red heat and plunging into cold water. The tone desired is obtained by hammering the instrument until the proper degree of hardness is obtained. Tempering not BRONZES AND OTHER COPPER-TIN ALLOYS. 2jg only increases the ductility and malleability of these alloys, but also, it is claimed, their strength, while decreasing their hardness and density, when they are made into thin sheets; thick plates are less affected ; annealing by slow cooling pro duces an opposite effect. Specuhcm-metal contains, often, as much as 33 per cent, tin ; it is steely, almost silvery white, extremely hard and brittle, and capable of taking a very perfect polish. The most suitable proportion of tin varies slightly with the character of the copper, some kinds requiring more and some less to give the degree of whiteness and the perfection of polish required. An excess of tin injures the color and reduces the lustre of the mirror. The finest speculum metal is perfectly white, without a shade of yellow, sound, uniform, and tough enough to bear the grinding and polishing without danger of disintegration. The specula made by Mudge were twice fused, and con- tained from 32 parts copper and 16 tin to 32 copper and 14.5 tin. A little tin is lost in fusion. According to David Ross, the best proportions are: copper, 126.4; tin, 58.9, i.e., atomic proportions. He adds the molten tin to the fused copper at the lowest safe temperature, stirring carefully, and secur- ing a uniform alloy by remelting, as is often done in making ordnance bronze. Bronze for bearings and pieces subject to severe friction, as in machinery, is made of many proportions. Gun-bronze is one of the best ; the Author has known of one case in which the bronze was made of ingot copper 90, ingot tin 10, and used in the main crank-shaft journal of a steam vessel for ten years without appreciable wear, although the area was not unusually large for the load and the velocity of rubbing was high, as is usual in screw engines. The proportions given in several cases will be found elsewhere; they vary in practice from 88 to 96 per cent, copper, as more or less hardness is required. Bronze for steam engine packing rings is some- times made of 92 to 94 copper, 7 to 9 parts tin, i part zinc. 2?,0 MATERIALS OF CONSTRUCTION— NON-FESROUS METALS. The fracture is of a uniform yellow color, with an even grain. The specific gravity of bronze is about 8.7, being greater than the mean of the specific gravities of copper and tin. Copper proposed to be used in ordnance bronze should be condemned for the manufacture of guns, if it contains sulphur in an appreciable quantity ; more than one-thou- sandth of arsenic and antimony united ; more than about three-thousandths of lead, iron, or oxygen ; if it contain more than about five-thousandths of foreign substances altogether ; or if, near these limits, it give bad results when subjected to the mechanical tests of hammering, rolling, and wire-drawing. It is also stated that tin offered should be rejected if, when run into elongated drops, it have not a smooth and re- flecting surface, without any considerable sign of rough spots ; if, when analyzed, it contain more than about one-thousandth of arsenic and antimony united ; more than about three- thousandths of lead or iron ; or more than four-thousandths of foreign substances. All bronze ought to be rejected which contains sulphur in an appreciable amount ; which contains more than about one- thousandth of arsenic and antimony united; more than about three-thousandths of lead, iron, or zinc; or, in all, more than about five-thousandths of foreign substances. Notice should be taken of the appearance of the fracture of specimens ; it sometimes gives indications sufficient to authorize the rejection of certain bronzes full of sulphur or oxides. Gun-metal, wh'en broken, should present a fine, close- grained fracture, of a uniform, beautiful golden color; it should be ductile, although finely granular and possibly crystalline. Bronze guns often exhibit, when burst, a decidedly crystal- line surface, the axes of the crystals lying radially to the bore. According to the practice of the Navy Department, the bronze used for rifled howitzers is composed of Lake Superiol- copper 9 parts, tin i part. This is used when the casting is made in a sand mould. When a chill mould is used, which is the method now adopted for such castings, the proportion is changed to 10 to i. BRONZES AND OTHER COPPER-TIN ALLOYS. 281 The copper is melted in a reverberatory furnace, and three hours after the fires are started, when the copper is in perfect fusion, the tin is stirred in ; half-an-hour after, the bronze is run off into the moulds. The casting cools nat- 'urally, and is taken out of the mould about twenty-four hours after the metal is run in. The chill mould is warmed suf- ficiently to drive out the moisture. Phosphor-Bronze and Manganese Bronzes are alloys which are now so well known and have become so important in the arts as to demand special notice. Phosphor bronze has been known many years. It consists simply of any alloy of bronze or brass or any ternary alloy of copper, tin and zinc which has been given exceptional purity and excellence by skilful fluxing with phosphorus. It is also supposed that the presence of phosphorus is useful in giving the tin a crystalline character which enables it to alloy itself more completely and strongly with the copper. Phosphor- bronze will bear remelting with less injury than will common bronze. The phosphor bronzes greatly excel the unphos- phuretted alloy in every valuable commercial quality, and they are very extensively used for every purpose for which such alloys are fitted. The following are Kirkaldy's figures for tenacity and ductility of phosphor-bronze wire of No. 16 Birmingham gauge : PHOSPHOR-BRONZE WIRE, NO. l6, B. W. G. MATERIALS. LOAD Al FRACTURE • ii No. twists be- g bo fore breaking- Unannealed. Annealed. S^ Per sq. Per sq . Per sq. Per sq. Per Unan- An- mm. in. mm. m. cent. nealed. nealed. ' 72.3 kil. 46 T. 34.7 kil. 22 T. 37-5 6.7 80 Phosphor-bronze of several pro- ■ portions, 85.1 85.2 97-7 54 54-1 62.1 33.6 37-5 42.8 21.3 23.8 27.2 34-1 42.4 44.9 22.3 13.0 17-3 52 124 53 112.2 71.2 41.7 26.5 46.6 13-3 66 106,3 67.6 45-4 28.9 42.8 15.0 60 282 MA TERIALS OF CONSTRUCTION— NON-FERROUS METALS, 1 u u bo toga a Cast copper. Sheet copper. Mean of 9 samples. Defective bar. Canbeforged like copper. Ramrods for guns. Defective bar. Resists action of hydro- chloric acid. Annealed and com- pressed. Hard malleable. Pieces of machines. Specific gravity after re- peated tempering. •Xlijoqjny fQ . I • 03 03 P3 •oox = jaA -[IS *j{3pu)03i3 joj XjjAiiDnpao3 •■o S so '• : : S ' * a ooi = j3Airs 'jBaq joj X}iAiiohpuo3 CO •0 Alqiqisnj JO japJQ ■o ■(13IIEW) Xjt -IjqBOiiEai JO japjQ « ■(uosnqof ptre )J3A1E3 pUE M •(nojsjnqx) •Xjinpnp 3AiiEi3-a a 8 : : g. -(lailEM) Xi!i!)3np JO japjo : " •qDUi ajEnbs jad spunod *XjpEU3x 1 5- t^ to M 1 3 ^ £ X. Si id E w 3 > Vesicular i ■a Red lii i "o : : ; ^ : : : •jCjiaejS Dgpads 00*00 CO 00 HI 00 00*00 00' 00 00 ■ 00' CO • 00 < i'S| 3. ? 3 : ^ Composi- tion of original mixture. d en o o.oo o.oo o.oo o.co 0,00 o.oo o.oo o.oo 1. 41 1.90 1.96 2.00 2.50 3.03 3-73 4.00 5.00 5-9° 6.00 6.02 3 u 100 100.00 100.00 100.00 100.00 100.00 100,00 100.00 100.00 98.10 98.04 98.00 97' 50 96.97 96.27 96.00 9S-00 94,10 94.00 93. gs "einouoj oioio^y U : y -J9C mn>i C^ [1^ r u rno t ^OC C hC '' n-t ';? 's: > c K « « BRONZES AND OTHER COPPER-TIN ALLOYS. 283 ri ft) CQwWOoi 5«Cc a P u oil O'-i o-z! o S a ocj-," o** "1 B c a 8 a H * a HO-Ofc. U . 1 1 1 i 1 1 ill H oc 10 00 C. CO 00 CO 4I? 00* CO* 00* 00 00' CO CO 0* 8, 10 M 00 \o t^ tNoo' CO CO 00 CO CO OS 6\ 6>6^ o\ 6 o 00*0 o SSSKSmmSShS JT ff ? M"3"Ji'Jl?J? CO N N N —' M M H M M O O O O d O o O O' 0» o^ 0« o O^o5 00 00 CO tN tvVD VO VO "O >nv)«rl--i-->t 0{oCTiC'0NOlO>0VCTvCft0^Oi0\Ch0% On Ov OlOO 00 OOaOCOCOa-MOOOOOOQOQO CO 00 CO C00000000» LJU B a tJ : e : ■■do. ^ 3 : 3 u •u ; a W ■ .«] 1LO ^ lO to 10V}"1 MV> 284 MA TERIALS OF CONSTRUCTION— NON-FERROUS METALS. 2 1 Axle-bearings. Jewelers' punches. Strongest of series. Annealed and tempered. 1 J; OjS'O rl ll a i 1 i i c 11 C/3 •Xjuoipny _ra ... M . . ^ . . • ^^' . - . . - ■001 = jaA -|IS 'Xipu^03|3 joj X;iAp"Dnpuo3 ■001 — JSAIJS 'jraij joj XjiAiionpuo3 8 ■(KIIEM) Xjiljqisnj JO JspjQ : Tt- • 1 ■ ro j s s -IiqrauEnijojapjo ■ tx • • • • M M '(uosuqcf pUE waAi^o pUB ■ISIIEK) ssanpjEH t m ■ ■ • ■* ' m « U ■(uojsjniix) X;iI!Pnp 3AIJEPH - ■ • « ■ 0' d •(jail^W) Xjijiionp JO jspjQ • m j j • ■* ■ »o ■■apm sjBnbs jad spunbd 'jtipenax ill,: : X% \ .00 ■ is : ii S,2 N £ ;e ■ : w : s s > if 1^ i > • :-S i 1 1 i 1 4) 1 1 ■ijiABjS ogpsds ; ■ -*■ f. OiOO -^ C7> t>« 00 00 00 OD 00* 00 H^o M in'—' Oi \n o> t^ M 00 00 00 CO 0* m 0000 CTi 000000 6-1 s. 00 5 : [^ Composi- tion of original mixture. eg vo vo rl tNoo 00 00 \ 8 8 88 85'S,S'8 8,§>g'K.,g ,g 8 8'» 8 888 88a8aS5'5'5'i? S-ffS-S" a 8 S>^ ass's : ■* (^ M N M M M t 00 CO 00 00 00 00 00 ■ 8 888 88SS8S,S,S,a^8 S8a8. B sss. s ss.i.i.ik'i.i.'i'i. ssia. •Bimtuoj 3irao)v * 4 i> : ju : ^& : ; ^• A. ; ' « ■■.33 ] -.853 ■ tnui '■ '■&&$ •XX. iranN \3 SS-J 'S S^S S' i N t Si . E :; tt 51 u ;'g,g,i li ii « r ) « 0" }0( 30 ?S BRONZES AND OTHER COPPER.TIN ALLOYS. 285 s 1 a I s •3 s :2i •c ca .MM gmuJMg b5 i«"te:S«tn -S'c/i gffltnuicn pjiis -cng 3^tnu5|2u5cntnM"«uii2"^(jcng^(n i : ■ • 00 . 5- ■ • ■ Ch ■ 00 ■ t% i'o to ■ • ■ "^^ \ "^ :;r • o- H • l> ■ :| ■ ts m : S M ; = n d 6 d 8.S 0* 6 d §.P. d d d d 8" 87 d d ?l8 d d d d d 8 d 3 d \ \ ; : ° : *^ : ° vo m^ H m" R cT o^g'g, M C« H CO inoo ti. ci CO 10 t^ M M 5,0' rocTco II MIO cT 'rt'cd ■s-s §§ 1 1 & 1 d ■§•3 "b -Su f •a 1 •a s III ll II Is go II « 1 ll 11 1 t a; 1 'I- 1 si RffS. in cjOi 000000 CO 00' M 00 GO o«o> ovoio* in OOOOCO OOCOOOOOOO N vo N t-»« f. ■'f -* m en "<»■ CT) MM N « mvo 00 00 CO ^00 00 00 8^ if, 10 vo & s ■8, vo' VO m vo ■* i ft 00 tir^d hmh'mi-*m c< romm t«.o3 0000 ooowrot^ co E-gS8 8S'S,RSS,^8.gS.oS vo VOVOVO mm K 8 o> mvo vo mt-o cjioovov M H m' d d t^ IS. 4 -^ N "-' d crv (^ ■* ■*• met mo • cofoSn i ;o5<35 : 3 p 3 P UUU CJ 5 5c 9,5 '3 3*3 3* : 3 3 'a 'p • J -' a '.c a : a :a :cac ri O M « ro ^ invQ t*,co a\ i ^oooo 00000 gi I m ■<• mvo tsoo o- M N m-*»ovo r^oo ov o h iHHHHHHHnpincinonnPiNcrirfl IMHHHMHHMHHHHHHHHMHH 286 MA TERIALS OF CONSTRUCTION— NON-FERROUS METALS. O 1 1 •1 1 i 1 •-fjuoqjnv •OOI = J3A IIS 'XIIDU1D3I3 JOJ A^lATpnpuo^ ci •OOI = JSAflS '5B3H jo; X?!A!p'npuo3 i ^ :i d Xiniqren; JO jspjQ \r, :*; ', f* . -liqBailEoijojgpjo CO : "^ : "^ ^ •(nosuiior POB 3J9AIB3 pUB H ■ m ■ ' 3 . 00 to •(uojsjnqx) HOO • SB : o' '• 0* g : d 1 8, d 4 to ■ ■(131I=W) jCl!l!lonp JO Jap-iQ • : '^ : i !" jo ■HDin 3.iBnbs jad spunbd 'AipBuax K 8 P^ o \o . m o\ t* t 1 i u a u 3 1 : o "1 ; 5« ■1 • 3 :0 o U 1 s| 5 "cS .5 1 ? : :§ ■; 1; 1 1 til ■ : 2 ■s- t 1 •jtjiAui3 OTjTDadg o> osoo CO h* m tx t^ ^^ rn^o \o ■ t^ t^ t>. cs t^ ts r> t^ t^ t-. t^ pn . vn ^% b* moo « « t^ to o ei t^ f^p »n-*J--«-inioo t^M «oo ^bvw t^t^[<^C^t^t^totCt^t>.tC.t^C^t^t^ (J "^ p c^' s. 1: In tN ■3 CO . to . ■ "^ ■ S : 6 00 E? ; ? - « • d :? i joo j Composi- tion of original mixlurc. a •■* « »o w> t>.oo OOCOOOOO M N roTf-^**-* r^oo oooooo d QO QQn tot^t^r^i^r^ t^c>rv t^co oocooocooooooocooooooo 0\0«o^o>o^ot u •Binaijoj Ditnojv i i 4: in . 33P3 :333333 3 3 uuuu :uuouuuuo •jac jrnnM u 3r S ?ff ?. ? 1- t ?5 ■* M 3^1 •■3=° ? ? H M 1\ J to M M BRONZES AND OTHER COPPER-TIN ALLOYS. 287 Ma. U. S. B. U. S. B. U. S. B. U. S. B. U. S. B. Ml. ■Wa. Ma. Cr. C.J. ■We. ?: : : : H • • • • »n - ■ ■ e • > c 1 « ! ■ ' t * 1 M i M i : i" • ■ ; ; 1 j'O - ; 1 t^ Oi Oioo 00 - ' • vo M meg c> • - ; ; ; • ■ 1 t^ • •CO Q lOt^O -^W t^\6 \0. Th in H . 'f in ro -^ mvo" eS Granular Fibrous u Grayish white White .S^g|5SvS;|S; CO d 0* tSM 0N-»- *o-*MMdooooooo I £,2 M M H H M H H H M M M ■^a -*-» 8 .3 ^ 5 .^ f '^ »; ^ ^5 .p. a o 3 S ■^ o g r ■-G 3 ^ IS - ^ 1-1 . ■J" S •S i ■ '3 « ^Y I. I. i'7 ca uu S ►«! >^ ^ 5 - «J , S> P C a ., <• ■" ^ „ a >v .j-d H .5^ "ij o "^ 00 rs .- "^ ^ %i 3 r ■§ i; R R ■« f« t ■5 I -2 <» Is ^ ~ r I u "^^ i ; s I I ^ " S F^ •? ^ <». ^« ."^ I ^ ^ S "O .g 8 ■" 2 '2 a 1 1 .• I i I I I I. "? ". [5 ^ i- !5 § .i " 7 It) t<_ ■5: d •§,5 ■S S « . Q ■g . . 3 -w M £ S «' i3 s i; i i «• !5 § .i "^ "I I. 2 S CHAPTER IX. THE BRASSES AND OTHER COPPER-ZINC ALLOYS. Brass is a term which is applied by many, and espe- cially older, authors indifferently to all alloys composed princi- pally of copper, combined with either tin or zinc. The alloy of copper and tin and its minor modifications are now becom- ing better known as bronze, and the name brass is generally restricted to the designation of alloys consisting mainly of copper and zinc. " Brass " ordnance is properly called bronze ordnance, and the compositions used in the bearings of machinery, which are usually of somewhat similar compo- sition, are also properly called bronzes. The alloys of copper, tin and zinc, which occupy intermediate positions between the bronzes and the brasses, are as often known by the one name as by the other. Copper and Zinc together form " Brass," which is usu- ally made nearly in the proportion, copper, 66^, zinc 33^. Brasses of certain other proportions have specific names, as Tourbac, Pinchbeck. The mixture and fusion of the metals must be so conducted that the loss of zinc by volatilization may be the least ^possible ; there is always some loss, and it may not only be serious as a matter of cost, but the introduc- tion of oxides into the alloy is exceedingly injurious to its quality. The fusion is generally performed in crucibles heated in air-furnaces. The change of color and of other qualities with the intro- duction of zinc is gradual and very similar in character to that produced by the admixture of tin ; but the quantity *of zinc demanded to produce the same modification is about twice as much as of tin. On adding zinc, the deep red color of copper is changed at once, becoming lighter and lighter, BRASSES AND OTHER COPPER-ZINC ALLOYS. 289 and finally shading into a grayish white" and then assuming more of the color of zinc. The alloy generally increases in hardness and loses ductility as the percentage of zinc is in- creased, up to a maximum, which being passed, ductility increases again. The most ductile are, however, those which contain 70 to 85 per cent, copper, 30 to 15 of zinc, the iirst being called " tombac," the latter " brass." Mallet's Classification.— The following is Mallet's table of the copper-zinc alloys : PROPERTIES OF COPPER-ZINC ALLOYS. ORDER OP AT. COMP. COPPER S. G. COLOR. FRACT TENACITV. Mall. Hard Fus. by anal. Cu Zn per ct. Tons persq. in. I : 100. 8.66r red 24.6 12.1 8 6 JO ; I 98.80 8 -605 red -yellow coarse 22 21 15 14 13 X2 9 : I 90.72 8 .607 *' fine ii-S 4 2a 8 : I 88.60 8 ■633 '* " 12.8 2 \l I '■ ' 87.30 8 587 (( (( 13.2 II 6 : I 85.40 8 591 yellow-red fine fibre II. 1 5 17 ID 5 : I 83.02 8 41S " " 13-7 IT i t 4 : I 79-65 8 448 '* " 14.7 7 15 3 : I 74.58 8 397 pale yellow t( 13.1 10 14 7 2 : I 66.18 8 299 deep " *' 12.5 3 4 6 I : I 49-47 8 230 " il coarse 9.2 12 12 6 I : 2 32-85 8 263 dark " " »9-3 I 10 6 8 : 17 31-52 7 721 silver white 41 2.1 very brittle s S 5 8 : 18 30.36 7 836 silver white *' 2.2 6 8 : 19 29.17 7 019 light gray (I 0.7 ** 7 S : 20 28.12 7 603 ash " vitreous 3-2 brittle 3 s 8 : 21 27.10 8 058 light ;; coarse 0.9 *' 9 5 8 : 22 26.24 7 882 *^ 0.8 " I 5 8 : 23 25-39 7- 443 ash " fine 5.9 slight duct, brittle I s I : 3 24.50 7 449 ** " " S'l 2 4 1 : 4 19.65 7- 371 4< *i '* 1.9 *' 4 3 ' ■' 5 16.36 6. 605 dark " (I 1.8 " II 2 : I 0. 6. 89s IS-2 23 la the above table, the minimum of hardness and fusibility is denoted by i. The conclusion of Storer * that these alloys are mixtures rather than true compounds, is accepted by Watts and other authorities. Uses of Brass. — Brass is the alloy commonly em- ployed in the arts in the construction of scientific apparatus, 19 * Mem. Am. Acad., N. S., vol. viii, p. 97, 290 MA TERIALS OF CONSTRUCTION— NON-FERROUS METALS. mathematical instruments, and small parts of machinery. It is cast into parts of irregular shape, drawn into wire, or rolled into rods and sheets. It is harder than copper, very malle- able and ductile, and can be "struck up" in dies, formed in moulds, or " spun " into vessels of a wide variety of forms if handled cold or slightly warm ; it is brittle at a high tempera- ture. A common proportion for making brass is copper 66, zinc 34. This alloy is a much slower conductor of electricity and of heat than copper, is more fusible, oxidizes very slowly at low temperatures, but rapidly at a high heat. The brass of Romilly, which works remarkably well under the hammer, is composed of copper 70, zinc 30 ; English brass is often given 33 per cent, zinc, and for rolled brass 40 per cent. This constitutes "Muntz sheathing metal," as patented by G. F. Muntz in 1832. The proportion of zinc ranges, however, for such purposes, from 37 to 50 per cent copper 63 to 50. Muntz Metal is thus described by its inventor : — " I take that quality of copper known in the trade by the ap- pellation of ' best selected copper,' and that quality of zinc, known in England as ' foreign zinc,' and melt them together in the usual manner in any proportion between 50 per cent, of copper to 50 per cent, of zinc, and 63 per cent, of copper to 37 per cent, of zinc, both of which extremes, and all intermediate proportions, will roll and work at a red heat ; but as too large a proportion of copper increases the diffi- culty of working »the metal, and too large a proportion of zinc renders the metal too hard when cold, I prefer the alloy to consist of about 60 per cent, of copper to 40 per cent, of zinc. This compound I cast into ingots of any con- venient weight, and then heat them to a red heat, and roll or work them while at that heat into bolts and other like ship's fastenings, in the same manner as copper is rolled or workq^d, but only taking care not to overheat the metal so as to pro- duce fusion, and not to put it through the rolls or work it after the heat has left it too much, say, when the red heat goes off." This alloy is cast into ingots, and rolled, hot, into sheets, BRASSES AND OTHER COPPER-ZINC ALLOYS. 2gi which are cleaned by pickling and washed before they are sent into the market. As this alloy is cheaper and more du- rable than copper sheathing, and equally effective, it has dis- placed the latter almost entirely in the protection of wooden ships. When made on a large scale, the alloy is melted in a reverberatory furnace. Special Properties.— Farmer has deposited brass by electrolysis and obtained an alloy containing copper 75, zinc 25, as ductile and malleable as rolled brass. The brasses, or copper-zinc alloys, although probably of more extended use than the bronzes or copper-tin alloys, are not as well studied as the latter. The metals, as already stated, p. 288, mix in all propor- tions, and produce alloys of which the general character has been shown in the introductory chapter of this part of the work and in the earlier paragraphs of this chapter. The red color of copper, in this series, fades into yellow very gradually, and becomes golden-yellow at about 40 per cent, zinc ; the color then becomes lighter, and at 60 per cent, zinc is bluish-white or silvery. With the change of color occurs the same change of strength and ductility noted with the copper-tin alloys, but it requires about twice as much zinc as tin to produce it. The white metals richest in copper are, like those of the bronze class, too brittle to be of use in engineering construction, but the yellow metals ob- tained with from 40 to 50 per cent, zinc are very valuable. Brass has a high coefficient of expansion, 0.000054 to 0.000056 per Cent, degree (0.00003 to 0.000033 per degree F.).* Yellow brass fuses at from 1,870° F. (1,021'' C), and other compositions from 1,000° F. (550° C, nearly) to 2,000° F. (1,100° C, nearly), and loses strength and ductility as its tem- perature rises. The composition of the several most useful brasses is given elsewhere. Brass for fine work is often made of copper, 80; zinc, 17; tin, 3 ; "fine brass" of 2 copper, i of zinc ; sheet brass of 3 copper, i zinc. A hard solder is made of 3 parts brass to i of zinc, etc., etc. Castings shrink in cooling j\ inch to the foot (0.015). * Viae Chapter I. 292 MA TERIALS OF CONSTRUCTION— NON-FERROUS METALS. Hydrochloric acid reddens brass by dissolving its zinc; ammonia whitens it by taking up the copper. Brass may be made tough and soft, hard and brittle, strong or weak, elastic or inelastic, dull of surface or lustrous as a mirror, friable or nearly as malleable and ductile as lead, as may be desired, by varying its composition. No known ma- terial, perhaps not even excepting iron, can be given so wide a range of quality or so wonderful a variety of uses. All the common varieties are composed of 6^ to 70 parts copper and 33 to 30 of zinc. A little lead is often added to soften and cheapen it and tin in small proportion to strengthen it. Brass is subject to flow under stress, like all other metals of what the Author has called the " tin class," and it is not safe to leave heavy loads upon it. Weights should not usually be hung upon brass chains, or upon brass tie-rods. The alloy is capable of being considerably hardened by compression, as when rolled into sheets, or by wire-drawing, and becomes much stronger and is less liable to permanent change under load. Some compositions are very elastic and make good springs for intermittent and occasional use. The thin sheet brass used for metallic cartridges and other purposes requiring a metal in this form of great strength combined with ductility, is subject, frequently, to a singular deterioration with age which seem to be partly a physical and partly a chemical change. It results, sometimes in a very brief interval, in the entire destruction of the essential proper- ties of such forms of this alloy. This has been studied by Egleston,but the results of investigation are not yet fully known. Weems has found * that a pressure of 4,000 tons (or ton- nes) being applied to brass, in the endeavor to produce brass tubes by " squirting " as is usual with lead, causes a separa- tion of the zinc, which issues as a zinc pipe, leaving the cop- per behind. This is considered a proof that this alloy is a mixture rather than a chemical compound. Applications in the Arts. — Bronze and brass have in- numerable uses in the arts : locks, keys, shields, escutcheons, * Land. Engineer, 1883. BRASSES AND OTHER COPPER-ZINC ALLOYS., 293 hinges, journal-bearings, pump-plungers, screw propellers, all small parts of machinery, optical and other philosophical instruments, cabinet-makers' fittings, sheathing of ships. Even so-called copper castings usually contain a small amount of zinc — 2 to 5 per cent., to give them soundness. The copper and brass manufactures of the United States are very extensive and of excellent character, both as to ma- terial and workmanship, and in those departments which are purely mechanical, are probably unequalled elsewhere. The purest copper is at their doors and the best of zinc ; while tin is likely, in time, to be largely produced in this country also. Brass to be used in the rolling mill in the manufacture of sheet metal, is cast between marble blocks which are separ- ated to a distance which determines the thickness of the ingot or slab. The marble is coated with a thin layer of loam prepared for the purpose ; the sides are closed with moulding sand. The slabs, when cast, are rolled, several " passes " being necessary, and the sheets are annealed at intervals, and when finally finished are "pickled " to, give them a good surface. For fine work, the surfaces must sometimes be repeatedly scraped during the process of rolling to remove surface impurities and defects. Wire brass is similarly treated, and the plates are then slit into rods in the " slitting mill," rolled to give them a section which can be handled in the wire-mill, and the rods are then drawn as in making iron wire.* Brass tubes for steam boilers, condensers and other pur- poses, are usually drawn, as are many other forms of section. Working Brass.— Yellow brass, and several composi- tions of similar character, are so easily worked cold that many articles are made by " striking up " in a die, under a press or a drop-hammer. Where a considerable change of form is necessary, the work is done by a succession of opera- tions alternating with annealing. Rolls may often be used to form brass into the desired shape and they are still oftener employed to impress a pattern on the sheet. * See account of methods of wire-drawing. 294 MA TERIALS OF CONSTRUCTION— NON-FERROUS METALS. " Spinning " is a peculiar and very interesting, as well as useful process. It is employed in altering the shape of a disk or of a cylinder which can be " chucked " and held in a lathe, while the tool of the workman, pressing on the edge, turns it over and forces it into a new shape. Spinning brass often consists merely in forming a flat sheet, turning in the lathe, by the pressure of a smooth burnishing tool. Chasing is done with a graver, and matting and emboss- ing with formers and hammers. In burnishing to give high lustre, the metal is kept wet with sour beer, while the burnisher by a steady friction produces the polish. " Burnishing " consists in giving a fine lustrous surface by the pressure and friction of a smooth, highly polished steel tool, lubricated well, as above. The surface is first prepared by giving it a good polish by the usual methods. The "burnishers" are made of fine steel, carefully polished with crocus and oil, and kept in the most perfect possible con- dition. The working of brass in the lathe requires especial care, not only in the handling, but also in the form of the tool. The cutting edge is given a much larger angle than in cutting iron and steel ; hand-tools require to be given precisely the right inclination and a constant rotation ; the velocity of cutting greatly exceeds that usual with iron. Brass tubes are sometimes made by simply rolling sheet- brass, cut to exact size, upon a mandrel and brazing or solder- ing the joint ; but they are more usually " drawn." The roll and its mandrel are sent through the draw-plate together and the tube is thus drawn to size and the soldered lap becomes distinguishable only by the color of the joint. Locomotive tubes, and others required to bear very high temperatures and pressures, are drawn solid and seamless. Brass condenser tubes should be made of copper 70, zinc 30, as prescribed by the British Admiralty. Boiler tube3 are made of copper 18, zinc 32. The metals should be pure. In many cases peculiar and ornamental shapes are given by modification of the form of mandrel or of draw-plates. Patterned sheets are produced by the use of rolls with BRASSES AND OTHER COPPER-ZINC ALLOYS. 295 properly cut surfaces. The " die " in which the metal is given shape under the blows of a " drop," or of a heavy hammer, is very extensively used in working brass. The Properties of Brasses, as described by the best authorities, are exhibited in the most concise manner in the following table, which was originally collated for the Com- mittee on Alloys of the U. S. Board,* as was that already given for the bronzes. It includes the results of work done for that board. A more complete exhibit of the mechanical properties of the bronzes and brasses will be given in succeeding chapters describing investigations, usually conducted by the Author, as above. * Report, vol. ii, 1881, p. 67. 296 MA TERIALS OF CONSTR UCTION—NON-FERRO US ME TALS. a i I 1 1 .1 III w SI'S ffl til -Is 1 00 f •3 i a •X;!Joqjnv « . .^. . . . .«::2 rt'~*a « 4" !3 •001 =J3A1IS 'jCjpiJiDSia JOJ XjIAIJDnpUO^ » i ■001 = J3AIIS 'jraq JOJ i£}iAfiDnpao3 ; H 00 Xjinqisnj JO japjo ' V, * fO s •(oosnqof pnE 1I3A1E3 pUE \ ^ i « g s -nqESiiEm JO JspjQ :" M H o •(noisjnqx) 00 i ft ^ * Xinjjonp JO jspjo '■CO VO ■* C4 •qom sjEnbs jad 8pun6d *i!jiDEU3x 3- 1 t ■ s 1 1 1 1 o 1 ill 1 1 1 1 1 1 1 1 1 1 1 ■jfjiABjS Dgraadg 00* • do' ? 1 00 "1 00 m^-co O m^s \6oo tN 00 00 00 1 00* oa B t3 .3 CO OO s ti .8* 5 { • Composi- tion of original mixture. .5 OOOOOOOOMM N V)\o tC t^OD Ot 0> On O O* h H H H. ■- M ''I S3 5 88888888J&& & SJS!S.iS.3.&8.S.g .si c ^ c a d N N 5 5 BRASSES AND OTHER COPPER-ZINC ALLOYS. 297 is o;g 'til ^1 Tombac of Oker. Gold leaf. Tombac for butto Bronze powder. Bath metal. 1 1 1 1 1| 3 a "o So. J .s . C (U Mil ba .s o ea M .B . .pq . .B .« . . m « n -^K.*' o o or:-i o,«" 000 orr'~»o ■no ■2'~>.j o . o ■ oa •"* o -ood -fljo* -do' d— iL*~^ & & tJ D p__p __ ^ D D i 0. • H 00 •2 • 00 ■ M «3 s Oi I ■00 • tN ' \o CO t^ : R si m C 01 CO « • cn -t- \o i ro 00 CO ? Ov V) K ■ t^ ro CO -8 1 1 co' i 1 a M 1 ^ Is 1 i u 1 E I- >x 1 1 i 3 •d 1 .1 1 •53 >< 1 1' Si 1 1 1 1 00 coco 00 00 CO co' 00* Oco 0000" s oo co' 00' 00 CO 00' % a;&f s "4- 6 leg i ■8 CO a ^ s i 00 n -s i ro h- i ■0 n IH *n c a i\£ •^ 1^ ■< 00 « r -•! .1 !■: >■ « « focr>-*-*--.j--*ifioin»nic rj.^ h «" m" m" m" 8 00 M N T^miniO ^' S-S-'S'8g'8,aS>S,8,SSSSS s,8,.s-'s, «*Sn -osassis^'s.ro mvD M NCO ■* ro ■* ■* oo^oso iiSS S i£'o?. t^ (> t^ S: K.g.K.S.8.g.-B-vE->S->B-!S!8g!g!8 c c= c ■3 S r>9is) NN NNN N N 3 s 3 S 3 3 3 .=? .H UU uo uuu u 298 MATERIALS OF CONSTRUCTION— NON-FERROUS METALS. ! / r Suitable for forging. Sp. gr. of powder, 8.390. Good brass wire. Mosaic gold. Suitable for forging. Strong solder for brass. Bristol metal. Suitable for forging. Muntz metat. Ship-sheathing. sp. gr. of powder^ 8,329. Suitable for forging. Bath metal. I' 1 fc-' > I c Sp. gr. of ingot, 8.263. Escutcheons of locks. i 00 d •^uoipny .« .n . .E3B . . can .cnP3 _m m •00: =.J3A "lis *X5I0U1DSI3 JOJ XiiAijonpuo^ •ool=J3ATIS*lB3q JOJ ijiAijD'npuo^ 00 -ttSUEW) AjiIjqTsnj JO japjQ : i\o ■ ■ '(uosuqof PUE 5J3AIE3 pUE 'J31IEJM) ssanpjBH •(lailEK) .41! -[iqrailEinjojapjo - . ■ 10 •(noisjnqx) itl![!Jonp SAiiEia^ 00 H w 0" d S3 : d d j 8 d 1 0* •CWllEK) AiiniDnp JO japjo i :H •qDui sjBnbs jad spunbd 'XiiDEoax 1 113 as, H mm : 1 t^ >* t j: 1 n u " i i =1 1 8 g g ■s g i Color. < "3 ■s » Pi 1 1 1 > in s > CO •X^tabjS 3y[33dg s 00 00* 00' GO 00 00' 00 00' 00 CO 00 t^OO 00 • ooooco t^t-< r^ CO 2 % g ■a- 3: 00 •<^ ■4-d :S„g'5 •s,s,s, .■ai^i ii 3: ;.8 "00' \i% is? i!?i Composi- tion of original mixture. .d N \o m M ■+ '*■ tv rooo ^ fo « 0^oi>. r^iomt-nt^ wivo r-. ts. w m m lo^o t^ m v\ r* ir , h U 00 mvo vo\om t»Mc"*fo wimwooM iotn«m»n o> »o -^roc^io m w tr.00 A Eg i" ^ 3 ^^ 3 3^ ■1 3 3 1 3 J BRASSES AND OTHER COPPER-ZINC ALLOYS. 299 ^32 ■c ■a u 11 C S «j 1.§1 . C " o a 9 Si" 5S .9 .9 fis, « . .« uiggtn .m . .m« . . . m UP DP &S5&& • >n\o . txfo ■ »co " -sg o M- j: f - ,£3- o - c-r fl ty o> o > > t/3 h4w H \o M o^ mvo o • CO «o ■<*-so c, , , „ „ „ „ • « 00 o oi tv.00 CO o v5 r» o 00 • 00 tsoo tN ti ti (Coo t^ t».oo t^ •2 ' §^& . ro ■^ f- w ■*\0 t- '^ ■* "* tC. ti bs (C t«. t^ ) CI O n ro o Ct2 (1,(1, sfc > J3- >-> o 00 'o >n O CO o « ^' be? art bib ^ ^ H 10 ro ^o "1 M nco m ro h oa o ro moo ro -^CO (N>OTf.H N fON H H OHOO M JO lo^o ■* CO o moo o « M vo m moo ■* o ,_co ^MNiTim.^[^ 0003100* (^^ "o fo m »o ro -"i- ■+ mvo t^ ^^ •^. ^ ^ •O t^ t^ t^ 1^00 OiO O M N «' rn-*'"l-tn»n>nt«.0 O O O O N rorom ^ 0^ O n m b«. >o\o>0'0^ovo'fi ^*^-.^s^s^^tNr«.tN^>.t>.^. csoo cooococooocoooooooco o^o^o^d •* m ^^o CI o t<> M o ^ oi ■* m 1^ N "S o J3 ,_ ,- cnoo t^^D inmro HH»OMci'*ro *a»nio vommmcoro moo m m en « ci cj « H d o <>oo' i^ »^«3 lAinm-i-Tt-pi o o oid^cJ* i^^o \oio«go^>«oN commmrorrrcotncj m w n « n«« n« « n «mm mmm hh h mm il.1 "14 44 41 4 ,li i.5 1 .11 I -aas z^ 5a s>s> 5 aa ^^l s>5> a 00000 w N H "j* P< M w k K P CO S ^ o ft * H ^ 00 ms - Si >■» ?3l o-2i •5 s jK >»S-« i ho (ao|SSB!nif mu(jgi5PP CHAPTER X. THE KALCHOIDS AND MISCELLANEOUS ALLOYS. Other Alloys than Bronzes and Brasses exist in im- mense variety and have numerous applications in the Arts, although of far less common application than the classes of alloys already described. Of these alloys, the most important are those which most closely resemble the true bronzes and brasses in composition, as alloys consisting of bronze or brass with which are united smaller proportions of lead, iron, nickel, antimony, bismuth, and other common metals. In this class also fall the "Kal- choids," as the Author would call them, or the copper-tin-zinc alloys which are usually called brass or bronze accordingly as zinc or tin predominates. The white " anti-friction " or "anti-attrition" metals, the fusible alloys, and type and stereotype metals, all come within this classification. The Kalchoids (Gr. Kalchos), or Copper-Tin-Zinc Alloys, are of great value and include the strongest, and probably the hardest, possible combinations of these metals. They are, in most Respects, usually, intermediate between the brasses and the bronzes obtained by uniting two metals. According to Margraff, these alloys are often very valu- able and have the character as per table on next page. Mackensie finds the alloy, copper 58, zinc 25, tin 17, excellent for castings and a good statuary bronze ; and pro- poses copper 50, zinc 22, tin 28, for mirrors for telescopes ; it is slightly yellow and takes a very fine polish. Bronzes in which equal parts tin and zinc are used are of common use for very small articles — as " brass " buttons. Knives for cotton printers' rolls are often made of copper 82, zinc 10, tin 8. Depretz' " chrisocalle " is of copper 92, tin 6, zinc 6, KALCHOJDS AND MISCELLANEOUS ALLOYS. 301 it has a beautiful golden color. Another composition imitat' ing gold is, copper 81.5, zinc 8, tin 0.5 ; and still another, which retains its lustre well, is of copper 80, zinc 17, tin 3; it is used for the small parts of ornamented pistols, etc. Alloys containing these metals are used for bronze medals, the zinc and tin being introduced to the extent of from 2 to 8 per cent, and the total of both being usually 10 per cent, or less. The percentage of zinc is usually kept under 3 or 4 in ordnance metal. COPPER-TIN-ZINC ALLOYS. NO. COPPER. TIN. ZINC. REMARKS. I 100 100 100 Very white, brittle, subject to liquation. 2 100 50 50 " but finer grain. 3 100 25 50 Yellowish tint, hard, fine, not malleable. 4 100 25 25 Brittle. 5 100 20 20 Brittle, hard, yellow. 6 100 16 16 " " " close grained. 7 100 14 14 Yellow, slightly malleable. 8 100 12.5 12.5 " more malleable. 9 100 II II " (( 11 10 100 10 10 Fine yellow, fine grain, malleable. II 100 8 8 Yellow, softer, more malleable. 12 ICO 7 7 Golden, malleable, soft. 13 100 6 6 it it tt The use of 8 to 15 per cent, of tin and 2 per cent, zinc in alloy with copper is probably as common as the employment of the bronzes without zinc ; the latter is added to improve the color. Alloys of copper containing from 3 to 8 or 10 per cent, zinc and from 8 to 1$ per cent, tin are used in engineer- ing very extensively, the softer alloys for pump-work, the harder for turned work and for nuts and bearings. An alloy of 5 per cent, tin, 5 zinc and 90 copper is cast into ingots and remelted for general purposes. It is tough, strong and sound. Copper 75, tin 12, zinc 3 makes a good mixture for heavy journal-bearings. Copper 76, tin 12, zinc 12, is as hard as tempered steel and was made into a razor-blade by its 302 MA TERIALS OF CONSTRUCTION— NON-FERROUS METALS. discoverer, Sir F. Chantrey.* When copper and brass are mixed in equal proportions and their sum is equal to the weight of tin, the alloy constitutes a solder. Copper, Zinc and Iron unite with some difficulty, and the presence of iron is thought to make brass harder, to weaken it, and to increase its liability to tarnish. A ternary alloy of this character was introduced in England as early as 1822 and was claimed to be stronger and better for the pres- ence of the iron. An alloy of i per cent, brass with 99 of iron was advised for castings exposed to corrosion, and Kars- ten found that it was harder than the cast iron, and considered it well adapted for use in steam engine cylinders and heavily loaded journal bearings. Herve -found the zinc less desira- ble in copper-iron alloys than tin. He states that alloys containing 1.33 to 4 per cent, copper and 0.65 to 3 per cent, zinc were stronger than the cast iron with which they were alloyed. Sterro-metal, elsewhere described, is a metal of this kind, containing also a small amount of tin. Copper, Tin and Iron may be alloyed to make a ferrous bronze of great value. The introduction of cast-iron into gun-bronze (copper 89, tin 11, or copper 90, tin 10) is not only useful, in small amounts as a flux, but this ferrous alloy is harder and stronger than the bronze alone. This alloy was made in Russian arsenals about 1820-5, and used for ordnance. The maximum proportion of iron was from 12 to 25 per cent., according to the use intended. The guns made of these alloys were found, according to Depretz, to excel good gun-bronze ordnance in strength and endurance. Similar alloys were made in France by the Messrs. Darcet f and by M. Dussaussoy, of the artillery, and on a large scale, in the government foundry at Douai. The latter experiments were made with alloys containing : Copper. Tin. . Iron. Copper. Tin. Iron* 90 10 6 90 10 4 90 10 3 * Holtzapffel. f Alliages M^talliques, p, 333. KALCHOIDS AND MISCELLANEOUS ALLOYS. 303 The results were not such as to lead to the adoption of these alloys in making field guns. Wrought iron was introduced into standard gun-bronze by Dussaussoy as early as 18 17, using tin-plate for the pur- pose. When the proportion of iron exceeded 2 per cent, the result was not satisfactory. For small articles, the ferrous bronze was found an improvement, it being stronger, harder and less fusible. Gen. Goguel, of the Russian Army, added 12 per cent, of wrought iron to gun-bronze, and reported that the ordnance made of this alloy proved much superior to that made of common gun-bronze. Subsequently, an extended investigation was made by the order of the French govern- ment by MM, Gay Dussac and Darcet, and later by M. Herve of the French Artillery. The former research led to no result ; the last named investigator concluded that the use of tin in securing an alloy of iron with copper is of ad- vantage and that re-fusion is advisable to secure the best results. Manganese Bronze is said to have qualities resem- bling those of phosphor-bronze, the introduction of man- ganese increasing the strength, ductility and homogeneous- ness of the alloy. The manganese alloys are usually white tinged with red, ductile, hard and tenacious. They are often known as white brass, white bronze or white alloys ; they take a fine polish ; those richest in copper have a decided rose hue. These alloys, as well as the phosphor- bronzes, are in somewhat extensive use, especially in Great Britain. Copper and manganese alloy easily, or with difficulty, under different conditions, making a metal of considerable mallea- bility, red in color, turning green when weather stained. It is less fusible than copper, lighter in color, and more liable to tarnish ; it may be made by fusing together copper and the black oxide of manganese. Manganese bronze contains iron, also, and is made by melting together copper and spie- geleisen or " ferro-manganese." When containing 10 per cent, manganese, the alloy of copper and this metal is dense, grayish-white with a tinge of red, very ductile and malleable, 304 MA TERIALS OF CONSTRUCTION— NOK-FERROUS METALS. and of rather a short fracture ; with 20 per cent, manganese, the color is silver-white to tin white, strong and ductile, with a fine lustre ; with 30 per cent, manganese, the properties remain little altered; with 40 per cent., the alloy becomes iron-gray, malleable and ductile, very strong, fracture inclined to fibrous. Thus, according to Berthier, all these alloys are ductile, strong, compact and homogeneous. Manganese-bronze is very similar in its general character- istics to phosphor-bronze ; but is a white alloy and differs in being a triple compound of the metals, copper, tin and man- ganese, instead of an alloy of copper and tin fluxed with a metalloid. It possesses some peculiarities which give very great value to this metal as a material of construction. It is remarkably hard, tough and elastic, has rather a high elastic limit, as compared with ordinary bronze, and is found to be very durable when used for bearings of machinery. A common pro- portion of its constituents is, copper, 88, tin, 10, manganese, 2. Preparation and Uses of Manganese-Bronze. — As described by the inventor, Mr. P. M. Parsons, white bronze, or manganese-bronze, is prepared by combining ferro-man- ganese, in different proportions, with various bronze alloys, thus producing qualities suited to various uses. The ferro- manganese is first subjected to a refining process, by which the silicon is eliminated, and the proportion between the iron and manganese adjusted in various degrees, for use according to the quality of bronze to be produced. To effect this com- bination, the temperature of the copper must be brought up to the melting point of the ferro-manganese, which is melted separately and then added in a fluid state. The effect of this combination is similar to that produced by the addition of ferro-m.anganese to decarbonized iron in the Bessemer converter. The manganese in its metallic state having a strong affinity for oxygen, cleanses the copper of oxides, and renders the metal more dense and homogeneofts. A portion of the manganese is utilized in this manner, while the remainder, with the iron, becomes permanently combined with the copper, and plays an important part in improving and modifying the quality of the bronzes prepared from the KALCHOIDS AND MISCELLANEOUS ALLOYS. 30$ copper thus treated, the effect being to increase their strength, hardness, toughness in various degrees, according to the quality and quantity of the ferro-manganese employed. Manganese, when once incorporated with the copper, is not driven off by remelting ; the quality of the naanganese-bronze is improved by remelting. Manganese-bronze, as is stated, when forged, is remarkable for its strength and toughness, having an average tensile strength equal to mild steel, and elongating as much before breaking. It is suitable for forgings of all kinds, for bolts and nuts for engine and machine work, for ships' bolts, rud- der and other fittings, screws, pins, nails, pump-rods, wire, and for all purposes for which yellow metal, brass, and cop- per are employed. In forging this metal, it should be heated to a clear cherry red (not bright), when it may be hammered, rolled, pressed, or worked in any way as long as it retains any color. It should not be worked at a black heat, but when the color is just fading it should be reheated. In rolled sheets and plates it can be worked both hot and cold. In working hot, the instructions given for forgings should be followed. The metal can be rolled, stamped, pressed, and worked cold like brass or copper, being annealed as required. It is stronger, stiffer, and harder than copper, brass, or yellow metal, for which it can be substituted for purposes to which these are applied. The rods, plates, sheets and angles are supplied of mild, medium, or high qualities, as required. The mild and medium qualities have a tensile strength of 28 tons per square inch (4,410 kgs. per sq. cm.), with an elastic limit at 40 per cent, and stretch from 28 to 45 per cent, before breaking. These qualities can be worked and riveted up cold, and are claimed to be greatly superior to yellow metal or gun metal. When ships' screws are made of this material, they are given less thickness than when made of mild steel or of com- mon bronze; it is not subject to alteration of form when taken from the mould or by the annealing which must be done with steel castings ; it retains a clean surface remarkably well, but its cost is considerable. 20 3o6 MA TERIALS OF CONSTRUCTION— NON-FERROUS METALS. The ferro-manganese used to mix with gun metal con- tains from lo to 40 per cent, of metallic manganese ; with brass alloys, 5 to 20 per cent., and with bronze alloys, the proportion lies between the above, according to the propor- tions of tin and zinc employed. To prepare ferro-manganese containing a given amount of metallic manganese, the invent- or melts rich ferro-manganese, containing up to 70 per cent., in a crucible under powdered charcoal, and with a quantity of the purest wrought-iron scrap. If it is desired to employ a ferro-manganese to mix with any of the alloys containing 20 per cent, of manganese, a ferro-manganese containing 60 per cent, of metallic manganese, and, say, i per cent, of silicon, is melted with wrought-iron scrap, in the proportion of lOO of ferro-manganese to 300 scrap. Then a ferro-manganese containing 20 per cent, of metallic manganese will be ob- tained, in which there is only one-third of i per cent, of silicon. Dry sand or loam moulds are recommended for casting. Metal moulds render the alloy somewhat harder and closer in texture. Manganese-bronze is said to be much less subject to cor- rosion in salt water than is pure copper. Alloys containing from 75 to 85 per cent, copper are most usually adopted for machinery. Zinc often forms a constituent of these alloys, in the proportion of from 2 to 10 per cent. The addition of manganese to bronzes and brasses gives them much lighter color, greater hardness and tenacity, with- out proportionally decreasing ductility and resilience. Cop- per and manganese alone form white alloys of great hardness, strength and ductility. Some of these alloys forge well and can be rolled with ease. They are somewhat susceptible to the action of the atmosphere at high temperature, and should be worked as little and at as low temperature as pos- sible. Aluminium-Bronze. — Aluminium is added to copper and to the bronzes and brasses with good results. The alloy, copper 90, aluminium 10, may be worked cold or hot like wrought iron, but not welded. Its tenacity is sometimes KALCHOIDS AND MISCELLANEOUS ALLOYS. 307 nearly 100,000 pounds per square inch (7,030 kilos per square mm.), and its average is not far from three-fourths as great. It is hard and stiff and very homogeneous. Wire has been given a tenacity exceeding 125,000 pounds per square inch (8,776 kilos per square mm.). Its specific gravity is •j.'j. In compression this alloy has been found capable of sustaining a little more than in tension (130,000 pounds per square inch, 9,139 kilos per square mm.), and its ductility and toughness were such that it did not even crack when distorted by this load. It is so ductile and malleable that it can be drawn down under the hammer to the fineness of a cambric needle. Measuring its stiffness, the Messrs. Simms found * that it had three times that of gun-bronze and 44 times that of brass. It works well, casts well, holds a fine surface under the tool, and when exposed to the weather ; and it is, in every respect, considered the best bronze yet known. Its high cost alone prevents its extensive use in the arts. {See note, p. 314.) Alloying 2 to 8 per cent, copper with aluminium increases its strength 65 to 90 per cent., making it stronger, weight for weight, than machinery steel. f The density of aluminium-bronze has been determined by M. Riche,:j: with the following results: BRONZE CONTAINING TEN PER CENT. OF ALUMINIUM. ~ DENSITY. I. II. WT. = 1 208'. 568. WT. = 120^' .275. 7-705 7-704 7.706 7-704 7.706 7 -70s 7.707 7-707 7-703 7-704 7-703 7.702 7.701 7.702 7.699 7-703 After casting. . . After tempering After annealing . After tempering After annealing , After impact. . . After tempering. After impact. . . * Ure's Diet. , Art. Aluminium. f Ry Rev., Jan. 17, 1891. X Ann. de Chimie, vol. xxx., 1873, pp. 351-419. Appendix. 308 MATERIALS OF CONSTRUCTION— NON-FERROUS METALS. BRONZE CONTAINING FIVE PER CENT. OF ALUMINIUM. After casting. . . After tempering After annealing After tempering After annealing After impact. . . After tempering After impact. . . 8.262 8.259 8.262 8.262 8.262 8.264 8.264 8.265 Tempering, annealing, and mechanical action produce no noticeable variation in the volume. These alloys are very uniform in character and work regu- larly and smoothly. Uses of Aluminium-Bronze — Aluminium-bronze, composed of 9 parts copper and i part aluminium, was pro- posed in 1864 as a material for small coins, and with this ob- ject in view the assayer of the United States mint made a number of careful experiments with it. The assayer states that aluminium-bronze possesses much greater hardness than copper alone, but less malleability and ductility. When rolled into sheets, it requires annealing at every third pas- sage through the rolls ; when drawn into wire it must also be frequently annealed. To strike a coin of this bronze required unusual force. It .tarnishes quite readily, but not more so than copper. '^- Aluminium-bronze containing 7J^ per cent, of aluminium is greenish in color, according to Morin, while other compo- sitions on either side are golden. Even i per cent, added to copper causes a considerable increase in ductility and fusi- bility, and enables it to be used satisfactorily in making castings. Two per cent, gives a mixture used for castings which are to be worked with a chisel. The standard alumin- ium-bronze — 10 per cent, aluminium — is brittle after the first fusion, but becomes more ductile as well as stronger by repeated refusion. It makes good castings, is easily worked, KALCHOIDS AND MISCELLANEOUS ALLOYS. 309 and may be forged at a red heat, and is fairly ductile under the hammer even when cold. It is softened by sudden cool- ing from a red heat. It takes a fine polish, is a half stronger than good wrought iron in tension, but has less strength in compression. Its coefficient of expansion is small at ordinary temperatures. Its liability to crack in large masses makes it difficult to get large castings. It has great elasticity when made into springs ; it is found useful for watches, and has the decided advantage over steel of being but little liable to oxi- dation ; the addition of 5 per cent, silver is advised to pure aluminium to make springs. Kettles of this alloy have been used in making fruit syrup and preserves. The alloy of aluminium with 4 to 5 per cent, silver is used in making balances for chemists. The introduction of a very minute proportion of bismuth makes this metal very brittle. Steel containing but 0.08 per cent, aluminium is said to be greatly improved by its presence. An alloy of 2 or 3 copper and 97 or 98 aluminium is found useful in making ornamental silver-colored castings which are to be chased and engraved. The alloys of aluminium and copper may be made by fus- ing together the oxides with metallic copper and enough car- bon and flux to reduce them. The oxides, as well as the other materials, should be as finely divided as possible, and the carbon introduced in excess. Copper and Nickel are quite easily alloyed, giving a metal of usually white color, hard, rather brittle, and quite easily oxidized. When the nickel forms 30 per cent, of the whole, the alloy is easily fused, strong, and tough, of a silvery- gray color, and slightly magnetic. White copper and Ger- man silver consist wholly or partly of this alloy. Copper and nickel unite in a wide range of proportions. In color they range from the red of copper to the blue-white of nickel, according to their proportions. Adding nickel in the proportion of o.io, the alloy is very ductile, light copper- red in color, and moderately strong; with 0.15 nickel, the color becomes very light red and the ductility is still great ; 0.25 nickel gives an alloy nearly white ; 0.30 nickel produces 'ilOMA TERIALS OF CONSTR UCTION—NON-FERRO US ME TALS. a silver-white metal. Berthier's alloy, copper 0.682, nickel 0.318, is fusible, ductile, strong, bluish-white, slightly mag- netic and somewhat crystalline near the surface. " White copper," so-called, is such an alloy, usually con- taining slight quantities of iron and silicon. Nickel coinage is now used by several nations; it was first privately coined by Feuchtwanger, of New York City, in 1837 ; Switzerland began using it in 1850, the United States in 1857, and Belgium in i860. The U. S. coins now contain copper 75, nickel 25. German Silver. — Copper, zinc, and nickel alloy readily. These compositions were used at a very early date in China, and have been known as packfong, tutenag, and white copper. The East Indian or Chinese tutenag is a grayish-white alloy, somewhat sonorous, and brittle. Its composition has been given as copper 44, zinc 40, nickel 16. The other alloys above named are nearly silver-v/hite, malle- able hot or cold, have a beautiful lustre, and very sonorous. The specific gravity is 8.5. Alloys of European manufacture, of similar characteristics, are now common. Viennese alloys have been found by Gersdorff to contain : — Table utensils ; copper, 50>; zinc, 25 ; nickel, 25. Ornaments " 55; " 25; " 20. Sheet metal " 60 ; " 20 ; " 20. Frick's alloys contain copper, 50 to 55; zinc, 30 to 31; nickel, \^ to 19. These are white and hard but ductile, and have a specific gravity from 8.5 to 8.6; they are used in making table utensils and ornamental objects. The alloy, copper 56, zinc 5, and nickel 39, makes a fine white metal of the same class with the preceding. German silver, as made by good makers, consists usually of Copper 60 Zinc 20 Nickel , 20 THE KALCHOIDS AND MISCELLANEOUS ALLOYS. 3" Tin and lead in equal parts make an alloy used for organ pipes. It is cast in sheets on a table; these sheets afe beaten smooth with a " planer," trimmed to size, rolled into shape and soldered together at the abutting edges. Tin and Zinc unite, in all proportions, readily and uniformly, the quality varying less with variation of propor- tions than in alloys generally, as may be seen by studying the change of strength exhibited by the map and model shown in the chapter on the ternary alloys. The introduction of zinc increases the hardness of tin, and rather increases its whiteness, when in small proportion ; in larger quantities it reduces ductility perceptibly. The alloy is of granular, some- times crystalline, structure, as revealed by fracture, and is somewhat sonorous. With equal parts tin and zinc the alloy is rather hard, moderately ductile, and of a very brilliant lustre. According to Koechl, the following are melting-points of these alloys : TUSION OF TIN-ZINC ALLOYS. ZINC. TEMPERATURE OF FUSION. REMARKS. Deg. Fahr. Deg. Cent. 1 2 3 I I 3 4 2 I I 500-572 572-662 428-680 472-662 680-932 260-300 300-350 220^360 250-350 460-500 Pure metals. ft (C Commercial. Pure metals. The alloy of equal parts of tin and zinc is said by some authorities to be nearly as strong as brass, to be much cheaper, and a better anti-friction metal ; but it is necessary that the zinc should be very pure. This alloy has been used in the form of roofing sheets. The alloy tin 75, zinc 25, makes ex- cellent metal patterns, the alloy flowing freely, running " sharp" and expanding slightly when solidifying ; it should not be overheated, and should be constantly stirred while pouring, 312^^ TERJALS OF CONSTR UCTION—NON-FERRO US ME TALS. to insure uniformity. This metal works easily, turns well in the lathe, and does not clog the file. Antimony, Bismuth, and Lead unite to form an alloy which expands on cooling, and which is therefore used for type-metal. Mackensie's alloy is antimony i6, bismuth i6, lead 68. Stereotype plates of good quality may be made of this composition. Antimony, Tin, and Lead are alloyed in the pro- portion of antimony 17, tin 13, lead 70, to form another Mac- kensie metal for stereotype plates and other printers' work. Sheets of this, or a similar alloy, are used in engraving music for printing ; a composition reported by Berthier is antimony 5, tin 60, lead 35. Antimony, Tin, and Zinc, in the proportions anti- mony 12, tin 44, zinc 44, make an alloy considered excellent for lining pump-barrels. Antimony, Bismuth, Tin, and Lead, in the propor- tions tin ^6, bismuth 8, antimony 8, lead 8, form the " Queen's Metal," which is one of the " pewter " alloys of greatest beauty and durability. Pewter and Britannia Metal. — Pewter has a wide range of composition, from tin 20, copper i, to tin 2, copper I. The alloy is often mixed with lead, of which the Pewterers' Company in 1772* permitted enough to bring the density of the pewter from ^f|^ to \\\% that of tin. The best Britannia, a metal of this class, is said to be tin "jj, antimony 15, copper 7, zinc 2 ; the alloy is cast in flat ingots and rolled into sheets. Britannia wares, made in Sheffield, are often composed of 3j^ parts block tin, 28 parts antimony, 8 of copper, and 8 of brass. The tin is melted and kept at a red heat while the antimony, the copper, and the brass are successively added, molten. The liquid alloy is ladled into the ingot moulds, which are slab-shaped cast-iron boxes, and the slabs thus made are subsequently rolled into sheets or recast into the form desired, or into such shapes as may be easily modified to the necessary extent. Spherical vessels are usually " spun up " in halves, which are then united by soldering. The * British Industries. Bevan, 1S71, THE KALCHOIDS AND MISCELLANEOUS ALLO YS. 313 solder is any very fusible composition of this class, and is often made of tin 75, lead 25. The fusibility of the metal is such that it requires some dexterity and great care to prevent its injury in the process of soldering. Britannia is easily shaped by all the familiar processes ; it may be cast, rolled and ham- mered, and cut in the lathe or by hand tools with equal facility. Iron and Manganese have a strong affinity. In small proportions manganese confers whiteness upon iron, and the alloy called " ferro-manganese " is considerably used in making steels containing very Httle carbon ; the carbide of this alloy, known as " spiegeleisen," or simply " spiegel " in the trade, is used in carburetting iron to produce steels " higher " in carbon. A small proportion of manganese renders iron less fusible, and is said to increase its tenacity. Many of the ingot-irons in the market, called " mild " or " low " steels, contain more manganese than carbon and are very strong and ductile, and make excellent material for use where great changes of tem- perature are not met ; this alloy is not considered suitable for springs, however. In large doses, manganese dees not re- duce the ductility and malleability of iron to the extent ob- served with the introduction of carbon. Karsten found that nearly 2 per cent, manganese improved iron. Mushet found that the alloy iron 71, manganese 29, was not magnetic, and concluded that the maximum attainable in iron was 40 per cent, manganese. As the percentage of manganese increases, the alloy becomes whiter, harder, more infusible, and more brittle if the manganese is present in considerable amount ; it is more subject to oxidation also. Platinum and Iridium alloy to form a composition, according to Matthey,* which is homogeneous and is capa- ble of being forged. Its density is 21.5 when of the com- position, platinum 98.5, iridium 12.5 by mixture, and platinum 90, iridium 10 by analysis. The density of the iridium was 22.38. The coefficient of expansion was from 0° to 16° C. (32° to 41° F.), 0.0000254. * Proc. Royal Society, 1878. 314 MATERIALS OF CONSTRUCTION— NON'-FERROUS METALS. Spence's " Metal " is not, strictly speaking, a metal, but is a compound obtained by dissolving metallic sulphides in molten sulphur,* which is found capable of receiving into solution nearly all known compounds of sulphur and the use- ful metals. It was discovered by J. B. Spence in the year 1879. The solution, on cooling, solidifies, forming a homo- geneous, tenacious mass of the specific gravity 3.37 to 3.7 at 0° C. (32° F.). According to Dr. Hodgkinson, when finely powdered, it is acted upon slowly by concentrated HCl and NOjHO in the cold ; in large lumps, little or no action takes place ; the expansion coefficient appears to be small. The fracture is not conchoidal, but somewhat like that of cast iron. It is said to be exceedingly useful in the laboratory for making the air-tight connections between glass tubes by means of caoutchouc, and a water or mercury jacket, where rigidity is no disadvantage ; the fusing point is so low that it may be run into the outer tube on to the caoutchouc, which it grips on cooling, like a vice, and makes it perfectly tight. It melts at 320° F. (160° C), expands on cooling, is claimed to be capable of resisting well the disintegrating action of the atmosphere, is attacked by but few acids and by them but slowly, or by alkalies, and is insoluble in water, and may re- ceive a high polish ; it makes clear, full castings, taking very perfect impressions ; it is cheap and easily worked. It has been used as a solder for gas-pipes, and as a joint-material in place of lead. Note. — Large quantities of aluminium bronze are now made in the United States and Europe by the Cowles process of smelting with the electric current. * Jour. Society of Arts. London, 1879. CHAPTER XI. MANUFACTURE AND WORKING OF ALLOYS. Alloys of General Application; Brass Working. — Of the alloys described in the preceding chapter but a few are employed by the engineer in his professional work, and still fewer are familiar and in common use. Of all the known alloys, the bronzes and the brasses, the coin alloys and a few compounds of tin, lead, zinc, antimony and bismuth, only, are so well known as to be properly classed among the ma- terials of constructive engineering. All the others are of use only in a restricted range of application and for a few special purposes. The methods of preparation are practically the same for all, and the " brass foundry " is usually resorted to in making them all. Brass work is divided into several branches, which, accord- ing to Aitken, are : 1. Brass casting, or ordinary foundry work; 2. Bell and cabinet-ware casting ; 3. Pot-metal and plumbing work ; 4. Stamped brass-work ; 5. Rolled brass ; wire-work ; sheathing ; 6. Tube making ; 7. Lamp making; 8. Gas fitting ; 9. Naval brass-founding. Several of these lines of work may often be carried on together, but it is usual to combine those most nearly re- lated — as those involving casting, those in which the metal is rolled or wire-drawn, stamping, tube-making and brass finishing. il^MA TERIALS OF CONSTRUCTION— NON-FERROUS METALS. Casting is described at length in pp. 317-319, on the brass foundry. Sheet-rolling is a very important branch of brass-making, employing a large number of work-people and sustaining a host of minor trades. The ingot brass for sheet-brass rolling is cast in broad, shallow, iron ingot-moulds, or when larger masses are to be used, in stone moulds, cut out of the solid block. They are well oiled and powdered with charcoal before filling them. The cast ingots of brass are called " strips," and are rolled, cold, by several successive " passes " through heavy rolls, with occasional annealing as they become hardened by the operation of the rolling-mill. When the surface of the sheet is found to be irregular and to contain spots of im- purity, the hand-scraper, or a scraping machine, is employed to remove them, and thus to prevent liability to cracking and raggedness of surface or edges. When rolled nearly to gauge, the sheet is " pickled," to remove the oxidized surface, and is then passed through the finishing rolls, which are finely polished and give the sheet its final finish. Muntz metals can be rolled hot, and therefore much more cheaply than other brass. Wire-drawing is conducted as in the drawing of iron and steel wire ; but the rods to be drawn are cut, by a slitting- mill, from sheet-brass. Like iron wire, brass must be occa- sionally annealed, in passing from wire-block to wire-block. Stamping in dies can be practised with any of the soft and ductile brasses, or other alloys. It is by this process that a larg'e proportion of the cheap brass ornaments are made, as well as many parts of various utensils, as lamps, door-fixtures and kitchen utensils. The die on the anvil is made of the desired form, and the metal is " struck " into it by the blow of a " drop-hammer" carrying a companion die, the drop falling from one to five feet according to weight and power. Heavy drops are always worked by steam power. The " force," or die carried by the drop, is usually of soft metal ; the die on the anvil is of steel. For fine and small in- tricate work, several blows are struck. This kind of work MANUFACTURE AND WORKING OF ALLOYS. 317 does not compare favorably with cast brass, or bronze, in clearness and fineness of lines. Brass Tubes are made by either of several methods. Sheet-brass is rolled, over a form, into a tube, and the edges soldered together, or they are rolled into cylindrical shape and soldered. For exact sizing, a mandrel is placed within the tube and on this it is rolled to gauge. Seamless tubes, such as are used in steam boilers and elsewhere under pres sure, are made by rolling, or by drawing down cast cylinders in a mill consisting of several sets of steel rolls. Brass-finishing includes lacquering, bronzing, dipping and burnishing and other methods of giving a surface finish, described at the end of this chapter. The Brass Foundry is usually an adjunct to large manufacturing establishments. It is generally small, and the moulding room and casting room are in one. A drying room, or core-oven, is conveniently located at the moulding room side ; it may be heated by either steam or by stoves, the for- mer being the better plan. A cleaning room and, beyond it, a finishing or dressing room, should be attached to the foun- dry, and, for fine work, a lacquering room is also required. The " patterns " are of wood or iron, as in iron founding, or they may be of stucco and pipe-clay. Patterns for brass castings must be larger than for iron, as shrinkage is one-half greater, i.e., i^th inch to the foot, or about 20 cm. per metre. The " shrink-rule" used for iron will not apply for brass-work. The flasks, and all details of apparatus, tools, and work are very similar to those used in an iron foundry, and the meth- ods are the same in the main. Castings are cooled rapidly, often with water, to soften and toughen them. Melting and Casting. — In the melting of the ma- terials in the making of alloys in the foundry, two general methods of procedure are practised ; in the one, all the con- stituents are fused at the same time in the same crucible or melting pot ; in the other they are fused one after another in a definite order, which is determined by their relative fusibility, volatility, and liability to oxidation, or to absorb oxygen and other gases. The first of these methods is, perhaps, the most 3l8 MATERIALS OF CON^STRUCTFON— NON-FERROUS METALS. common, but the second is by far the better; thus in making the most common ternary alloys, those of copper, tin, and zinc, the copper is best melted first, the tin should be next introduced, and the zinc, which is volatile and oxidizable, is added last. If lead is to be introduced into such an alloy, it is found best to put it into the crucible last. Other things being equal, the metals should be added in the order of their non- volatility ; the next controlling quality is infusibility ; the least fusible should generally be melted first. The casting and cooling of the alloy is hardly less a mat- ter of importance than the methods of fusion. Liquation is very liable to occur in certain cases, as in many alloys of cop- per with tin, and to prevent it the most prompt cooling pos- sible is resorted to ; the use of " chills," or metal moulds, is sometimes found essential to success. In these cases, it is not advisable to pour the alloy " cold," as liquation may have al- ready commenced ; they should be poured hot — " sharp," as the term is often used in the foundry — and yet compelled to chill quickly, if possible. The apparatus of the foundry, in which alloys are mixed and cast, consists of an air, or wind, furnace, sufficiently large to receive the crucibles in which the metals are melted, or, sometimes, when the masses handled are very large, a rever- beratory " open hearth " furnace, which is preferably heated with gas or liquid fuel ; of a collection of crucibles, which may be iron melting-pots for lead and alloys which melt at a low heat and have no affinity for iron, but which are usually of clay, of graphite, or of graphite mixed with clay ; and utensils for weighing and handling the metals, fuels, and crucibles. In some cases platinum and silver crucibles are used, as in lab- oratory work, but these are rarely needed. The crucibles should be carefully selected, since the cost of these vessels is often an important item of the expense account. In melting, the constituents of the charge being intro- duced in the order decided to be, on the whole, best, the liquid metal should be carefully stirred after each addition, and in such a manner as to secure most complete intermixture without liability to injure it by exposure to an oxidizing MANUFACTURE AND WORKING OF ALLOYS. 319 atmosphere. When the alloy is not homogeneous and sound, it may sometimes be greatly improved by refusion. In mak- ing large castings, the several metals to be alloyed are usually melted separately and all finally poured together into a reser- voir in which they are thoroughly mixed before " pouring the casting." Where a reverberatory furnace is used, the process may be conducted as in crucibles ; in this case, especial pre- cautions must be observed to preserve a deoxidizing flame within the furnace. When they are used in making bronzes, great care is taken to keep the mass constantly stirred to pre- vent liquation and the floating of the tin to the top. The fuel used in the mint-furnace is generally coke, which should be dense, hard, bright, and should ring when struck. In large establishments, and especially in melting bronzes, the open-hearth reverberatory is very generally used. Bell founders use a peculiar dome-topped furnace, melting at more moderate heat. In " pouring," the small castings are made first and the heavier are poured with the cooler metal. Sheet-brass is first cast in plates between heavy marble blocks washed with loam and well dried, or, often in ingots. They are rolled in heavy plate-mills and occasionally annealed as they become hard and unmalleable in the rolls. In making brass-plates and sheet-brass, the surface is pickled, after the sheet is reduced nearly to size, in order to give it a clean surface, and then, when a finish is demanded, it is given by a set of poHshed rolls. Wire-brass is cast and rolled into plates, which are cut into narrow strips in a " slitting-mill " by narrow interlocking roll-collars. These strips are rolled to a conveniently small size, and are then sent to the wire-mill to be finished in the draw-plates. Furnace Manipulation. — In filling the furnaces, the crucibles are slowly heated to avoid danger of breaking; they are at first set bottom upward. When well heated, they are set mouth upward and charged with the broken copper. The tin or zinc is heated at the mouth of the furnace and is added gradually to the copper as the latter becomes fluid. 320 MA TERIALS OF CONSTRUCTION— NON-FERROUS METALS. The zinc is liable to volatilization, and is, therefore, when introduced, plunged well below the surface of the molten copper. The Author has sometimes had it wrapped in dry paper or other protecting material to secure protection from loss by volatilization and oxidation. Great care is needed to prevent the introduction of cold and especially of damp metal ; seriously dangerous explosions are sure to take place if this should happen. The fumes arising from the molten alloys when poured are unhealthy, and a form of intermittent fever known as the " brass ague " is often produced by them where proper pre- cautions in handling and in securing ventilation are not observed. The brass-founder's furnace consists of a vertical cast- iron cylinder or other casing — often a brick structure — lined with fire-brick to a diameter of lo to 15 inches. The flue is led off at one side at the top, and the whole is covered with a plate having an opening of sufficient size to permit the crucible to enter and fitted with a cover plate. The grate is usually composed of loose bars which can be easily and in- dependently withdrawn or inserted. Each furnace contains one crucible, and large castings are only made where several furnaces are in use or where the alloy can be melted in a reverberatory furnace. Tuyeres are sometimes fitted for the purpose of increasing the rapidity of melting, and the crucibles are then, when large castings are to be made, emptied as fast as ready into a ladle which serves as a collecting reservoir from which the mould is filled. The fuel is usually either coke or charcoal. The Preparation of the Alloys involves considerable knowledge of the behavior of the mixture and its constituents while fusing and while the alloy is being formed, and can only be successful when the skill and judgment of an ex- perienced founder aid in the work of melting and casting. There are two methods of making alloys : the one is that of weighing out the constituents in proper proportions and mixing and melting all together ; the other is that of mixing and melting the constituents suCpessively and in an order MANUFACTURE AND WORKING OF ALLOYS. 321 dependent upon the character of the metals and the alloy made of them. The first is the usual method and is the least troublesome and expensive ; but it does not usually give as sound, uniform, and strong castings of the alloy as the second. In the latter case, the metal of highest melting point is usually first fused and the others are added in the order of fusibility or volatility. The order of introduction into the crucible or melting-pot has an appreciable effect on the quality of the alloy. After the alloy has been made and poured into the ingot, or other mould, it may change in composition by the process of separation or " liquation," to which reference is elsewhere made, either by the denser metal settling out or by the change due to formation of other definite alloys of greater stability at various points in the mass, thus altering the com- position of the metal all around those points. Thus in gun- metal (bronze) the surface of fracture often has a variegated color due to separation of the tin and the production of a variable composition of alloy. This will be noted in the description of the behavior of many alloys made by the Author. It will be seen that the rapid cooling secured by the use of metal moulds is the best means of preventing this liquation. Slow cooling, affording ample time for the separa- tion and reconcentration of the constituents, and for the pro- duction of crystals, permits, often, very serious loss of quality. It will be noted that the best alloys are "usually made most successfully when the molten metal is poured " sharp," i. e., hot, and then rapidly cooled to the point of solidification. The process of liquation is sometimes usefully applied, as in the Pattinson process of separating the metals in argentif- erous galena, or in cupriferous ores of lead. The desired alloy is very rarely made of its essential con- stituents only. A simple binary alloy of copper and tin is, for example, rarely made in commercial work. Lead is often added to give color, zinc to cheapen it or to Hux it, and some- times other metals to give special qualities. Statuary bronze usually contains some lead and zinc to give it its peculiar "patina"; bronze used in "hardware" and architectural 322 MATERIALS OF CONSTRUCTION— NON-FERROUS METALS. work, in bearings, etc., contains lead to give color and to make it wear well ; brass is hardened greatly, and strengthened, by the addition of one per cent, tin, or more, as in the " maxi- mum alloys " discovered by the Author. In such cases, the metal is added in small quantity to the mixture, after the latter is in fusion and alloyed. Minute Quantities of Alloy often greatly influence the properties and quality of metals. Thus, it is stated * that lead alloyed with 0.003 per cent, of antimony turns percep- tibly freer than pure lead ; that an addition of 0.007 P^'' cent, copper unfit leads for use in the manufacture of white lead ; that gold containing 0.05 per cent, of lead exhibits greatly decreased ductility ; that copper containing 0.5 per cent, iron has but 40 per cent, of the conductivity of pure copper. Nickel is too brittle to work; but, alloyed with o.i per cent, magnesium or 0.3 per cent, phosphorus, it can be rolled and worked. Brittle steel is sometimes made tough and malle- able by alloying it with 0.08 per cent, manganese or magne- sium. A difference of o.ooi per cent, in the amount of phos- phorus present in the best Swedish irons can be plainly observed by a change of malleability. Art Castings in Bronze represent the most perfect work known in the department of foundry work. It has been practised from the earliest known and even pre-historic peri- ods, and the analyses of art castings found in the Egyptian tombs and in Nineveh prove that the composition then adopted was subs^ntially that of the statuary bronze, and that of the art-work of to-day. The Greeks began to make bronzes several hundred years before the Christian era, and before the commencement of that era, had attained great skill in the art. The statue of Apollo, at Rhodes, made by the pupil of Lysippus, Chares, 330 B.C., was about 100 feet (30 metres) high, and after having been shaken down by an earth- quake some 60 years later, lay over goo years prostrate, and was then carried away by a Jew who purchased it from the Saracens, making a load, as it is said, for 900 camels. The Chinese and Japanese first made use of bronze at some * Der Techniker, 1883. MANUFACTURE AND WORKING OF ALLOYS. 323 unknown but very early date. The art was long lost in Europe, but was revived in the i6th and 17th centuries, and now con- stitutes an exceedingly important industry. Art castings of large size are moulded and cast precisely as other brass-founding is done; but great precaution is taken in the selection of materials, in determining exactly the desired proportions and in all the details of foundry practice and manipulation. The usual mixtures are given elsewhere. In making statuary, the model is first formed, and is then lined off by the founder in sections in such manner that each will be likely to be easily moulded and will " draw " readily ; plaster patterns are made of these sections separately, which are used in obtaining metal copies, which latter are finally joined together. Where the piece is to be cast whole also, the mould must be often made in many parts, in order that every section of the mould may be readily removed. In some cases, an elastic mould is made within which a wax model is formed, the latter being moulded in the sand in the usual manner. For such work, a plaster cast is usually first made, which is coated with any oily or glutinous substance which will not allow moisture to be transferred, and will not permit the adherence of the cope or mould, to be formed over it. By covering the model with a thin coating of wax, an outer mould can be constructed, and the inner and outer shapes may thus be separated by a thin space which represents that to be filled by the molten bronze, and determines the thickness of the casting. This space is often filled with wax and the latter is melted out when the mould is sent into the drying room or oven. Properly made, the mould has smooth, perfect sur- faces of the exact form to be reproduced, and the bronze, when removed from it, is an exact reproduction of the model, only requiring a small amount of work to make it marketable. If the composition and the work are what is desired, the sur- face of the casting is smooth, precise in form, sharp in out- line, an4 of good color. Statues thus made acquire, with age, a color or " patina " which distinguishes all good bronzes. Statuary bronze, and bronze for art-work generally, should have, when newly cast, a fresh, yellow-red color, and a fine 324 MA TERIALS OF CONSTR UCTION—NON-FERRO US ME TALS. grain, should be easy to work with file or chisel, very fluid when melted, taking the finest impressions of the mould, and when exposed to the atmosphere in the finished casting, should take the peculiar green patina characteristic of old bronze statuary of good quality. These qualities are not usually ob- tained in so high a degree in the copper-tin or copper-zinc alloys, the common bronzes and brasses, as in alloys contain- ing the three metals. According to Guettier, the best of these alloys are : IPPER. ZINC. TIN. 92 6 2 85 II 5 65 32 3 It is very usual to add i or 2 per cent, of lead ; ancient bronzes contain as much as 6 per cent. According to Pliny, bronze was made by melting copper first, then adding 12^ per cent, of an alloy of equal parts tin and lead, known as plumbum argentarium. Stereotype Metal, of which a good quality is made of 16 parts antimony, 17 parts tin, and 67 parts lead, is worked thus : The type is brushed over with a small quantity of black- lead and oil, placed in a casting-frame, and a cast taken in plaster of Paris. This cast is dried in a hot drying-oven until absolutely free from all moisture, and is held in form, meantime, by securing it to a flat cast-iron plate. The stereo- type metal is cast upon the matrix thus produced, and the plate thus obtained is planed up to a gauge and fitted to the press, or mounted on wooden blocks of the right height to work in the press. Damaged type are cut out and replaced with perfect ones. A later process is the following : * A sheet of tissue paper covered with printing paper is placed upon a perfectly smooth metal plate, and the two sheets are pasted together. These sheete are laid over the type, beaten into their in- terstices, covered with other sheets similarly beaten down, and * Spon. MANUFACTURE AND WORKING OF ALLOYS. 325 this process is continued until the mass of paper forms a matrix of satisfactory thickness and strength. Heavier paper, as cartridge paper, is used for the last layers. This matrix is dried carefully between surfaces which hold it in shape, is then brushed over with French chalk or black lead, and laid in the casting box, where the stereotype metal is cast over it and a plate thus made. German Silver is made by English founders of small bells and similar articles of copper 57, zinc 19, nickel 19, lead 3, tin-plate 2. The copper and nickel are fused together first, the zinc added after their fusion, and the other metals last. Commercial zinc containing lead is rarely pure enough for the finer grades of this alloy which do not permit the in- troduction of lead. It is difficult to obtain this alloy in correct proportions and of good quality. Babbitt's "Anti-attrition" Metal is usually not cast directly into the " brasses " to be lined with it. It is made by melting separately 4 parts copper, 12 Banca tin, 8 regulus of antimony, and adding 12 parts tin after fusion. The anti- mony is added to the first portion of tin, and the copper is introduced after taking the melting-pot away from the fire, and before pouring into the mould. The charge is kept from oxidation by a surface coating of powdered charcoal. The " lining metal " consists of this "hardening," fused with twice its weight of tin, thus making 3.7 parts copper, 7.4 parts antimony and 88.9 parts tin. The bearing to be lined is cast with a shallow recess to receive the Babbitt metal. The portion to be tinned is washed with alcohol and powdered with sal ammoniac, and those surfaces which are not to receive the lining metal are to be covered with a clay wash-. It is then warmed sufiS- ciently to volatilize a part of the sal ammoniac, and tinned. The lining is next cast in between a former — which takes the place of the journal — and the bearing. Founders often prefer to melt the copper first in a plum- bago crucible, then to dry the zinc carefully and immerse the whole in the barely fluid copper. A report of a committee of the American Master 326 MA TERIALS OF CONSTR UCTION—NON-FERRO US ME TALS. Mechanics' Association, reporting on the use of Babbitt metal, state that thirty-five replies to their circular gave the following facts relating to the use of Babbitt metal : Four use gibs with Babbitt ; four use the solid octagon brass without Babbitt ; seven use octagon with Babbitt ; seven use half- round solid brasses without Babbitt ; four use half round brasses in three pieces with Babbitt, and one makes no re- port of the use of Babbitt. All, with one exception, report that the Babbitt metal should extend the entire length of the journal and should be put on in strips ^ to i^ inches wide, at a point between the top and the front and back points of the journal bearing ; one inserts it by drilling holes in the brass and then filling in with the metal. The Com- mittee have observed that, in engines of from thirty-two to thirty-five tons weight, the half-round brass does not give as good results as in lighter engines. Good results may be ob- tained from a hexagon-shaped brass if properly fitted. The Babbitt metal will wear until it is cut through into the cast- iron. The recess in the top of the brass is of advantage also as a reservoir for oil ; and as there is less bearing at that point, the brass wears away and the shaft beds itself into the brass, so that there is no lost motion or pounding between the shaft and the brass. The Conamittee is of opinion that the use of Babbitt metal is advisable. Solders are alloys used in joining metallic surfaces, and parts of apparatus or constructions, by fusing them upon the surfaces of contact, and allowing them to cool, obtaining a more or less firm and tenacious union. They have a wide range of composition ; the " soft solders " are made of tin and lead ; " hard solders " are usually made of brass ; and special solders are composed of various alloys of copper, zinc, lead, tin, bismuth, gold and silver. Haswell's table of solders is given later. In soldering copper, brass, or iron with soft solder, & " soldering iron " is used to melt, and to apply the solder to the work. This instrument consists of a copper head, shaped somewhat like a machinist's hammer, the large end of which is inserted longitudinally in the claw-shaped end of an iron MANUFACTURE AND WORKING OF ALLOYS. 327 holder, which is itself carried by a wooden handle ; it is heated in a small charcoal-furnace, or " brazier," which is especially constructed for the purpose. A " soldering fluid," usually a solution of zinc in hydro- chloric acid, is used to remove the oxide from the surfaces to be joined and to give them a coating of zinc, to which the solder will promptly adhere. Soldering is often successfully performed by cleaning the surfaces thoroughly, fitting them nicely together with a file, if necessary, placing a piece of tin-foil between them, binding them firmly together with " binding wire," and heating in the flame of a lamp or a Bunsen burner, or in the fire, until the tin melts and unites with both surfaces. Joints carefully made may be united, in this way, so neatly as to be invisible. When using the more fusible solders, as those containing bismuth, a fire is seldom needed. When one joint has been made with hard solder, it is not always safe to make another near it with the same composition ; a softer solder should then be used. Soft solders are not malleable, and where this quality is demanded, the solder is necessarily of the hard variety. An excellent solder for such work is made with silver and brass in equal parts. Coin silver, in thin sheets, is an excellent solder for cop- per, hard brass, and wrought iron. Cast iron cannot be readily soldered, and the attempt is rarely made. In making solders, great care is to be taken to secure uni- formity of composition ; they are often granulated by pour- ing from the crucible or ladle through a wet broom or from a considerable height into water, or they are cast in ingots and reduced to a powder by filin g or by machinery. The silver and the gold solders are usually rolled into sheets ; the soft solders are generally sold in sticks, as is also pure tin ; those which are rich in tin are distinguished by their pecuHar " tin- cry," which is destroyed by a very small admixture of other metals. In making solders, as all other such alloys, the most infusible metal is first melted, and the other constituents are added in the order of infusibility. Soft solders are very 328 MATERIALS OF CONSTRUCTION— NON-FERROUS METALS. fusible and are melted under tallow, and the hard solders are prepared under a covering of powdered charcoal to prevent oxidation. Yellow brass, containing from 65 to 80 per cent, copper, will be found useful at times in brazing wrought iron, mild steel, and all the common brasses and bronzes containing less than 10 per cent, tin or lead. Equal parts of copper and zinc make a good solder for yellow brass and is known as " broom " solder. Tin and lead are sometimes added, but probably without advantage, the one making the solder hard, the other weakening it. For brazing iron, yellow brass is excellent. In using these solders, the surfaces to be brazed should be well cleansed, sprinkled with borax, and bound tightly to- gether with fine iron wire. A clay " dam " around the joint is useful in confining the solder in place when melting. The heating should be gradual and the temperature brought slowly up to a red heat, occasionally adding borax, and, finally, the heat should be more quickly raised until the solder melts and fumes, when the piece is cooled. Silver and yellow brass make good solders for steel, melting at a moderately high heat and having considerable strength. Both solder and flux should be used sparingly to secure good work. Cast iron and alloys containing either tin or lead in considerable quantities cannot be easily soldered. The soldering fluid answers as a flux for soft solders ; borax is used with the hard varieties, as it dissolves the oxides of all metals thus treated, and leaves the clean metallic surface which is essential to perfect union. Sal ammoniac is often added to the soldering fluid when soft solders are used, and resin is a common, and in some respects the best, flux for tin- ner's work. Platinum is soldered with gold, and German silver with a solder of equal parts of silver, brass, and zinc. The essentials of a good solder are that it shall have ah affinity for the metals to be united, should melt at a consider- ably lower temperature, should be strong, tough, uniform in composition, ^nd not readily oxidized. Standard Compositions are often adopted by en- MANUFACTURE AND WORKING OF ALLOYS. 329 gineers for the various purposes to which they apply the alloys. The tables hereafter presented are full of examples. In further illustration, we have the following as the compo- sitions adopted by the Paris, Lyons, and Mediterranean Rail- way of France : STANDARD ALLOYS. PROPORTIONS. ALLOY. USES. Copper. Tin. Zinc. Lead. Ant. Gun-metal, i. 82 16 2 Bearings. " 2. 84 14 2 Valves, Screws, etc. 3- 90 8 2 Cocks, V\fiiistks, etc. Brass, i. 70 30 Tubes. " 2. 67 33 Stuffing-boxes, etc. 3- 65 35 Handles, Latches. 4- 63 37 Plates, Washers. White metal. 5 71 24 Bearings. Packing " 14 76 10 StufSng-boxes. Solder. 11 45 40 •• 55 60 For tin plate. " zinc " The useful alloys are, as already seen, exceedingly numerous, and are of extraordinary variety in appearance and physical qualities. They are applied to an almost equally wide range of uses. The following very complete lists will give an idea of their number, quality and applications.* * Chas. Haswell; Pocket-book, 1882. C. Bischoff : Das Kupfer und seine Legirungen ; Berlin, 1865. P. A. BoUey: Recherches Chimiques ; Paris, l86g. A. Herve: AUiages Metalliques, Manuel-Roret ; Paris, N.D. ' 330MA TERIALS OF CONSTRUCTION— NON-FERROUS METALS. ALLOYS AND COMPOSITIONS. — HASWELL. o u N 2 p y d § < n ii 8 S5 y At Ar Ba Br 55. 50. 3-7 84.3 75- 79-3 io" 88.8 74.3 50. 88.9 90. 10. 67. 66. 14.3 7-8 3-4 10. 10, 25. 2.9 1.6 20. 10. 7- 7- 1.4 "i.'b 10. 1 16:5 12.5 18:4 23- 20. 12.5 15-6 86'.' 80. 22. 29. Ill 28:; 21. 40. 2 S 25 20 IX 22 31 2 80 33 34 13 II 31 5 19 25 5 33 25 24 7 40 45 21 7 25 •5 .2 •4 .2 •3 is .1 .2 .5 •3 \ 4 4 • 4 .8 2.5 7-3 2-5 " hard * mathematical instruments. ' Pinchbeck ' rollsd ' very tenacious. , . ' wheels, valves. Br Br Ch Ch Ch CIc Clo Ge Go Ho La Ma Me Mu Pe\ Pri She Spe Tel Te ^h Ors 25. 25- " when fuserl add. , , mze, red 87" 86. 67.2 80. 90. 93. 93. fsX 69. 33.3 40.4 87.S 77-4 60! ■ red * " small ' statuary § nese silver ck bells 1.5 cks, musical bells rman silver 33-3 31.6 24. 2.6 2.5 use bells chinery bearings " hard tal that expands in cooling 16.7 vter, best 20. 80. M- ki 20. !■ 6616 33-4 7.4 69.8 73- culum " De and stereotype plates 69. 5l:i 15.5 " hard ide nesia 4.4 Cream of tartar 6 . 5 * For adding small quantities of copper. MANUFACTURE AND WORKING OF ALLOYS. 331 Tin. coarse, melts at 500" " ordinary, melts at 360°.. Spelter, soft " hard W Lead Steel ";.■;; Brass or copper Fine brass Pewterers' or soft Gold " hard " soft Silver, hard.. " soft.. Pewter Iron Copper SOLDERS. 67 FUSIBLE COMPOUNDS. COMPOUNDS. ZINC. TIN. LEAD. BISMUTH. CADMIUM. 33-3 25 12 2S 33-3 31 25 50 33-4 so SO Newton's fusing at less than 212° '3 Special Recipes. — The best bronze compositions for use in engineering are, according to Guettier,* the following : For pumps, bolts and similar pieces : Copper. Tin. . . . 88 I (J^opper go 12 Tin 10 The latter is the softer of the two. Often from one to four per cent, of zinc is added, as already stated. * Guide Pratique ; Paris, 1865. 332 MATERIALS OF CONSTRUCTION— NON-FERROUS METALS. For eccentric-straps and connecting-rod bearings : Copper 83 84 83 84 82 85.25 Tin 15 14 15 14 16 12.75 Zinc 2 2 1.5 1.5 2 2 Lead .. 0.5 0.5 .. 100 100 100,0 loo.o lop 100.00 The addition of lead and increase of copper gives softer alloys. Lead is often used more freely than above. Locomotive driving-axle bearings : Copper 74 80 85.25 86 8g Tin g.5 18 12.75 14 8 Zinc 9.5 2 2.00 .. 3 Lead 7 . . .... 100. o 100 100.00 100 100 The Author prefers gun-bronze to either of the above. For Locomotive Slide Valves — Copper phosphate 3.50 Copper 77.85 Tin 11.00 Zinc 7.65 100.00 Connect ing-Rod Brasses — Copper phosphate 3.5 Copper 74. 5 Tin il.o Zinc Il.o 100. o Axle-boxes — No. I. No. 2. Copper phosphate 2.5 1.5 Copper 72.5 73.5 Tin 8.0 8.0 Zinc 17.0 ig.o lOO.o loao MANUFACTURE AND WORKING OF ALLOYS. 333 Parts demanding greater strength — Copper phosphate 3.5 Copper 85.5 Tin 8.0 Zinc 3.0 100. o Zinc is here added to the bronze to aid in securing that homogeneousness which is essentially the result of the ad- dition of phosphorus. For pistons (rarely needed) : copper, 89.75 ; tin, 2.25 ; zinc, 8. For car and locomotive axle bearings : Copper 80 79 86 8g Tin 18 18 14 2. 5 Zinc 2 2.5 .. 8.5 Lead 0.5 100 lOO.O 100 lOO.O For ordinary stationary machine journal-bearings : copper, 82 ; tin, 18. For whistles of locomotives and bells : Copper -. So 81 78 79 78 71 Tin 18 17 20 23 22 26 Antimony 222 Zinc 6 .. Zinc 1.8 Iron 1.2 100 100 100 100 100 loo.o The last is the alloy of the famous " silver-bell " of Rouen. For pump-buckets, valves and cocks : Copper 88 88 86.8 Tin 10 10 12.4 Zinc 1-75 2 0-8 Lead 0.25 100.00 100 loo.o For hammers (for use on finished work) : copper, 98 ; tin, 2. This alloy will forge like copper; it may be hardened by adding more tin. 334 MA TERIALS OF CONSTRUCTION— NON-FERROUS METALS. For wagon axle bearings : Copper 78 Tin 20 Zinc 2 100 Copper 25 Cast-iron 70 Tin 5 66.5 74-5 79-5 33-0 25.0 20 0.5 0.5 0.5. The best brasses may be taken, for general purposes, as accepted by good makers, as follows : For turned work : Copper 61.6 Zinc 35.3 Tin 0.5 Lead 2.5 loo.o loo.o 100.0 loo.o The richer colors are given by the higher proportions of copper. The official recipe for work in French dock-yards is : Copper 65.80 76.0 85 Zinc 31.80 24.0 15 Tin 0.25 Lead 2.60 0.5 1 100.45 100.5 loi The hardest compositions are used for the smallest pieces. These are used in the ornamentation of engines, for brass straps, for hinges, and for pulley-sheaves. Cheap alloys for bearings have been made of the follow- ing wide range of composition : Copper 56 Tin 28 Zinc 16 100 loo.o 100 The first — Fenton's alloy — is said to wear well, not to be specially liable to heating, and to be very durable. The last — Margraff's alloy — is of similar quality. The second com- position is much cheaper and lighter, and takes the place of the white alloys used in bearings. 5-5 58 19-5 28 80.0 14 MANUFACTURE AND WORKING OF ALLOYS. 335 Kingston''s metal, formerly much used for bearings, is made by melting 9 parts copper with 24 parts tin, remelting, and adding 108 parts tin, and finally 9 parts of mercury. An alloy of 90 per cent, tin, 8 per cent, antimony, and 2 per cent, copper has been found excellent for crank and con- necting-rod bearings on the Moscow and Nishni Railroad of Russia. On the Kursk-Charcow-Asow Railroad an alloy of 78.5 per cent, tin, 11.5 antimony, and 10 copper is considered very superior for pivots of all kinds, slide valves, eccentrics, stufifing-boxes, etc. The Swiss Nordostbahn Company, in ordering locomotives recently, required the following prepa- ration as a composition for axle journals : 10 parts of anti- mony added to 10 parts of melted copper, with 80 parts of tin added, and the alloy run into bars, to be remelted for use. Bronzing is the process of staining or otherwise coloring the surface of brass, in imitation of bronze — usually imitating old bronze. The methods of bronzing and the bronzing liquids are different for different purposes and as practised in different localities and different trades. Brass is very seriously subject to oxidation, and when polished soon loses its brightness and its color. Polished surfaces are often protected by the process of lacquering (to be presently de- scribed), but the permanent preservation of the polish is rarely possible and a coloring or bronzing is very commonly resorted to. It was formerly customary to give scientific apparatus a fine polish and to cover this surface with lacquer ; it is now becoming more generally customary to bronze them or to stain them either black or brown ; these are, in fact, but modifications of one process. To obtain the golden orange color characteristic of brasses rich in copper, the piece may be polished and im- mersed in a warm bath of the neutral solution of crystallized acetate of copper for a moment, washing in clean water and rubbing dry and bright. The chloride of antimony gives a dark rich violet color, if the article is heated to nearly the boiling point of water; sulphate of copper gives a watered surface and copper nitrate a black. Larkin used the hydrochlorate of copper with a little 33^ MA TERIALS OF CONSTRUCTION— NON-FERROUS METALS. free nitric acid, largely diluted, to produce a dark bronze ; a little acetic acid added to the solution of the same salt gave a copper color, and the patina of antique bronze was imitated by adding ammonia solution in large amount, or a quantity of sal ammoniac. " Bronze " paints are used for giving to the surface of iron, or of any other material, the appearance of bronze ; they have a great variety of composition ; some are composed of filings, or powder, of brass mixed in some oil paint which does not conceal the color of the bronze. Graham's bronzing liquids, as published in i86i,*havea great range of composition and of application as follows : BRONZING LIQUIDS. To he used for Brass iy simple Immersion. No. 1 pt. d E 1 dr. S i6 B u •d 1 dr. 5 e 1 g pt. I 2 8 & oz. 1 I 30 'a ■3 1 % OZ. I xo 1 g •s a .2 p 1 1 I dr. z 3 dr. 6 ■3 I OZ. .5 *s s 1 1 pt. I a 3 t dr. 1 •3 dr. 16 16 20 u S dr. 3 4 ■1 .a •3 X OZ. I I 2 3 4 1 I 9 lO XI 12 13 14 ( Brown, and every 1 shade to black. ( Brown, and every 1 shade to red. Brownish red. Dark brown. Yellow to red. Orange. Olive-green. Slate. Blue. Steel-gray. • Blackf In the preparation of No. 5, the liquid must be brought to boil and cooled. In using No. 13, the heat of the liquid must not be under iSo*" F. No. 6 is slow in action, taking an hour to produce good results. The action of the others is, for the most part, immediate. * Brass Founders' Manual, Lond. 1870, MANUFACTURE AND WORKING OF ALLOYS. 337 To be used for Copper by simple Immersion. 1 !>. *3 u § . a. •s . t 1 1 '5 :2 .Si H 1 ■3 •a 1 ■3 ^ ^ 8 ^ g 1 No. g 1 "a a X pt. dr. oz. dr. dr. dr. OZ. dr. oz. dr. IS s Brown, and every shade to black. Dark-brown drab. J6 .■i 2 17 I 2 11 " 18 ^, 1 Bright red. Red, and every shade to black. Steel-gray, at 180" F. IQ J I 20 •' I For Zinc. Vo. pt. I I I 2 2 1 I § 1 dr. 5 •s 1 s dr. Si 1 i dr. I i dr. I I •0 1 oz. oz. 4 1 1:4 % OJ •a 1 >. •a dr. I 1 1. 1 & K diT '8 § 1 1 1 a 1 X 21 22 23 =4 25 26 27 28 29 30 Black. Dark gray. Green-gray. Red boil. Copper color. Plates so c A z. Copper color, with agitation. Purple boil. Lacquering is the process of covering a polished surface of brass or of other metal with a transparent or trans- lucent coating, which, while protecting it from oxidation and * Made to the consistency of cream. 338 MA TERIALS OF CONSTRUCTION— NON-FERROUS METALS. discoloration, does not wholly conceal it. It is a process of varnishing polished metal. It is applied also to the surfaces of bronzed objects. Lacquer is a solution, usually, of some vegetable gum or resin in alcohol or other effective colorless solvent. In its application, great care is taken to keep the piece to be lacquered warm and of uniform temperature, to apply the solution quickly, smoothly and uniformly. The usual solution is '' shellac " in alcohol, and the best can, as a rule, be made with the " stick " lac. It may be colored by any permanent transparent alchoholic solution giving the desired tint. The red coloring matters are, usually, dragon's blood, red saunders or annotto ; the yellow are gamboge, sandarac, saffron, turmeric or aloes. The following is Graham's table of lacquers : LACQUERS. SOLUTIONS. REDS. YELLOW- a •s ^ i 1 ■^ 1 No 1 ■| n 1 s 'C '5. (/3 1 a & H 1 a in 01 'a d 1 t en (A i J i cd oz. dr. dr. pt. oz. dr. oz. pt. dr. dr. gr. dr. dr. dr. dr. dr. I 4 I Strong simple. 2 I 1 I I 1 1 Simple pale. Fine pale. 4 I I I I 2 *' I 2 I I 16 4 8 *' 6 2 2 I 8 12 8 Plate gold. I 2 5 I 2 4 Pale yellow. ^' Ross's. Q I I 4 Full yellow. lO S I 2 lb 2 Gold. II 3 4 64 6 14 " 12 I I 20 2 5 " 11 3 I 4 16 Deep gold. « 14 .1 I 4 I " 15 S I 10 4° 12 10 " 16 I B ^2 Red. 17 I !i 24 27 " 18 IS so 1° 6 20 60 la Tin lacquer. Green, lor bronze. 19 I 4 X The union of red with yellow produces a fine orange color. MANUFACTURE AND WORKING OF ALLOYS. 339 The lacquers are kept in carefully stoppered bottles, and it is better that they should be of opaque material, or of glass impenetrable by actinic hght capable of altering them ; yellow glass is sometimes used. When in use, they are poured into dishes of convenient size and form and are laid on with a thin, wide flat brush.* " Clouding " is performed by pouring on the surface a mixture of fine charcoal dust in water, stirring it to obtain the pattern, and then drying. The work is finally lacquered. * See " Materials of Engineering " for lacquers and browning liquids for fire- arms, etc. CHAPTER XII. STRENGTH, ELASTICITY AND DUCTILITY OF METALS. The Strength of the Useful Metals and other mechanical properties have not attracted as much attention as the engineer would desire. Investigations have been few in number relating to soft metals, and as a rule unfruitful, in comparison with those relating to iron and steel. In recording and discussing experimental work on the various metals and their alloys, the system and nomen- clature adopted will be that employed in the study of the strength of other materials. The following summary will here suffice.* Following it, will be given a statement of the results of experiments made upon all the " useful " metals, succeeded by chapters describing investigations of the strength and elasticity of their alloys, and the conditions modifying strength. The Resistance of Metal to rupture may be brought into play by either of several methods of stress, which have been thus divided by the Author : . J. J j Tensile : resisting pulling force. ( Compression : resisting crushing force. ( Shearing : resisting cutting across. Transverse -i Bending : resisting cross breaking. ( Torsional : resisting twisting stress. • * Abridged and adapted from Part II., M. of Eng. For the theory of the elasticity and strength of materials, consult " Wood's Resistance of Materials,'' published by J. Wiley & Sons, and Burr's work on the same subject issued by the same publishers. STRENGTH OF NON-FERROUS METALS. 341 When a load is applied to any part of a structure or of a machine it causes a change of form, which may be very slight, but which always takes place, however small the load. This change of form is resisted by the internal molecular forces of the piece, i.e., by its cohesion. The change of form thus produced is called strain, and the acting force is a stress. The Ultimate Strength of a piece is the maximum resist- ance under load — the greatest stress that can exist before rupture. The Proof Strength is the load applied to deter- mine the value of the material tested when it is not intended that observable deformation shall take place. It is usually equal, or nearly so, to the maximum elastic resistance of the piece. It is sometimes said that this load, long continued, will produce fracture ; but, as will be seen hereafter, this is not necessarily, even if ever, true. The Working Load is that which the piece is proportioned to bear. It is the load carried in ordinary working, and is usually less than the proof load, and is always some fraction, determined by circumstances, of the ultimate strength. A Dead Load is applied without shock, and, once applied, remains unchanged, as, e. g., the weight of a bridge ; it pro- duces a uniform stress. A Live Load is applied suddenly, and may produce a variable stress, as, e.g., by the passage of a railroad train over a bridge. The Distortion of the, strained piece is related to the load in a manner best indicated by strain diagrams. Its value as a factor of the measure of shock-resisting power, or of re- silience, is exhibited in a later article. It also has importance as indicating the ductile qualities of the metal. The Reduction of Area of Section under a breaking load is^ similarly indicative of the ductility of the material, and is to be noted in conjunction with the distortion. E.g. A considerable reduction of section with a smaller proportional extension would indicate a lack of homogeneous- ness, and that the piece had broken at a weak part of the bar. The greater the extension in proportion to the reduc- tion of area in tension, the more uniform the character of the metal. 342 MA TERIALS OF CONSTRUCTION— NON-FERROUS METALS. Factors of Safety. — The ultimate strength, or maxi- mum capacity for resisting stress, has a ratio to the maximum stress due to the working load, which, although less in metal than in wooden or stone structures, is, nevertheless, made of considerable magnitude in many cases. It is much greater under moving than under steady " dead " loads, and varies with the character of the material used. For machinery it is usually 6 or 8 ; for structures erected by the civil engineer, from 5 to 6. The following may be taken as minimum values of this " factor of safety for the non-ferrous metals : " LOAD. SHOCK. Dead. Live. Iron and steel 3 5 4 6 8 7 8 10 + lo to 15 Copper and other soft metals Ratio of Ultimate Strength to Working Load. The brittle metals and alloys. The Proof Strength usually exceeds the working load from 50 per cent., with tough metals, to 200 or 300 per cent, where brittle materials are used. It should usually be below the elastic limit of the material. As this limit, with brittle materials, is often nearly equal to their ultimate strength, a set of factors of safety, based on the elastic limit, would differ much from those above given for ductile metals, but would be about the same for all brittle materials, thus : LOAD. SHOCK. Dead. Live. I 2 3 2 4 6 3 6 8 to 12 Ratio of Klastic Brittle metals and alloys Working Load. * The figure given for shock is to be taken as approximate, but used only when it is not practicable to calculate the energy of impact and the resilience of the piece meeting it, and thus to make an exact calculation of proportions. STRENG TH OF NON-FERRO US ME TALS. 343 The factors of safety adopted for non-ferrous metals are higher than those usually adopted for construction in iron or steel in consequence of the fact that the elastic limit and the elastic resilience, or shock-resisting power, of the latter seem to increase with strain, up to a limit; while the former gradually yield under comparatively low stresses, as will be seen hereafter. In common practice, the factor of safety covers not only risks of injury by accidental excessive stresses, but deterioration with time, uncertainty as to the character of uninspected material, and sometimes equally great uncertainty as to the absolute correctness of the for- mulas and the constants used in the calculations. As inspec- tion becomes more efficient and trustworthy; as our knowl- edge of the effect of prolonged and of intermittent stress becomes more certain and complete; as our formulas are improved and rationalized, and as their empirically deter- mined constants are more exactly obtained, the factor of safety is gradually reduced, and will finally become a mini- mum when the engineer acquires the ability to assume with confidence the conditions to be estimated upon, and to say with precision how his materials will continuously carry their loads. A characteristic distinction between the ductile non-ferrous metals and ductile iron or steel, is that the former have usually, as purchased, no true elastic limit, but yield to small stresses without recovery of form and their permanent set equals their maximum distortion. Where brittle, they are often very elastic, however, and recover fully. In such cases, the elastic limit coincides with their ultimate resistance to fracture, as is the case with glass, hard cast iron, and often with hardened steel. In the table above it is assumed that an elastic limit occurs at the point at which the elongation becomes o.cxDio of the total length of the piece stretched. In some cases it is advisable to design some minor part, or element, of a train with a lower factor of safety, to insure that when a breakdown does occur it shall be certain to take place where it will do least harm. 344 AfA TERIALS OF CONSTR UCTION— NON-FERROUS ME TALS The Measure of Resistance to strain is determined, in form, by the character of the stress. By stress is here understood the force exerted, and by strain the change of form produced by it. Tenacity is resistance to a pulling stress, and is measured by the resistance of a section, one unit in area, as in pounds or tons on the square inch, or in kilogrammes per square cen- timetre or square millimetre. Then, if T represents the tenacity and K is the section resisting rupture, the total load that can be sustained is, as a maximum, P=TK Compression is similarly measured, and if C be the maxi- mum resistance to crushing per unit of area, and K the sec- tion, the maximum load will be P=CK Shearing is resisted by forces expressed in the same way, and the maximum shearing stress borne by any section is P=SK Bending Stresses are measured by moments expressed by the product of the bending effort into its lever-arm about the section strained, and if P is the resultant load, / the lever-arm and M the moment of resistance of the section considered, P/=M Torsional Stresses are also measured by the moment of the stress exerted, and the quantity of attacking and resist- ing moments is expressed as in the last case. Elasticity is measured by the longitudinal forCe, which* acting on a unit of area of the resisting section, if elasticity were to remain unimpaired, would extend the piece to double its original length. Within the limit at which elasticity is unimpaired, the variation of length is proportional to the STRENGTH OF NON-FERROUS METALS. 345 force acting, and if E is the "Modulus of Elasticity" or " Young's Modulus," / the length, and e the extension, P being the total load, and K the section : ^ eK ~ iL ^~ EK The Coefficients entering into these several expressions for resistance of materials are often called Moduli, and the forms of the expressions in which they appear are deduced by the Theory of the Resistance of Materials, and the processes are given in detail in works on that subject. These moduli, or coefificients, as will be seen, have values which are rarely the same in any two cases ; but vary not only with the kind of material, but with every variation, in the same substance, of structure, size, form, age, chemical composition or physical character, with every change of tem- perature, and even with the rate of distortion and method of action of the distorting force. Values for each familiar ma- terial, for a wide range of conditions, will be given in the following pages. Method of Resistance to Stress. — When a piece of metal is subjected to stress exceeding its power of resistance for the moment, and gradually increasing up to the limit at which rupture takes place, it yields and becomes distorted at a rate which has a definitely variable relation to the magni- tude of the distorting force; this relation, although very similar for all metals of any one kind, differs greatly for differ- ent metals, and is subject to observable alteration by every measurable difference in chemical composition or in physical structure. Thus, in Fig. 55, let this operation be represented by the several curves, a, b, c, d, etc., the elevation of any point on the curve above the axis of abscissas, OX, being made pro- portional to the resistance to distortion of the piece, and to the equivalent distorting stress, at the instant when its dis- 34^ MA TERIALS OF CONSTRUCTION -NON-FERROUS METALS. tance from the left side of the diagram, or the axis of ordi- nates, Y, measures the coincident distortion. As drawn, the strain-diagram, a a, is such as would be made by a soft metal like tin or lead ; 6 b' represents a harder, and c c' 2. still harder and stronger metal, as zinc and rolled copper. If the smallest divisions measure the per cent, of extension horizontally, and io,ooo pounds per square inch (703 kilogrammes per square centimetre) vertically, dd' , would fairly represent a hard iron, or a puddled or a " mild " steel ; while//' and^j-' would be strain diagrams of hard, and of very hard tool steels, respect- ively. The points marked e, e', e", etc., are the so-called " elastic V ~ *" " ■— " -^ ^^ ^^ ^1^ y ^ _ U y _ --T '/ ^ ' "^ ^N ^ -^ ' ' ^ '' --- -%- J ^ IT- ^ Lt -L «^ _ r^ _ _ ^'~ - _.3 <- - ~ 1 -€ - ,^--= ::::t-T:-- =-: i' \/' ■f-- - - ^ :::£-::- ^./ m T- :*r- f - = \, :tE:l::: --|- :i? \ ■ = = ■ J4- Fig. 55. — Strain-Diagrams. limits," at which the rate of distortion more or less suddenly changes, and the citation becomes more nearly equal to the permanent change of form, and at these points the resistance to further change increases much more slowly than before. This change of rate of increase in resistance continues until a maximum is reached, and, passing that point, the piece either breaks, as at/' and g-', or yields more and more easily until distortion ceases, or until fracture takes place, and it^ becomes zero at the base line, as at JC. Such curves have been called by the Author " Strain- diagrams." Equations of Curves of Resistance or Strain-dia- grams. — These curves are, at the start, often nearly para- STRENGTH OF NON-FERROUS METALS. 247 bolic, and the strain-diagrams of cast iron, k, i, k, having their origin at o, are usually capable of being quite accurately ex- pressed by an equation of the parabolic form, as I I' e in which -j is the ratio of elongation to the length of the piece, and P the load per unit of area. The constants may be taken as: Material. British. Metric. British. Metric. Cast Iron 14,000,000 984,000 3,000,000,000 211,000,000 Forged Iron 25,000,000 1,758,000 125,000,000 8,800,000 Soft "Steels" 25,000,000 i,758jooo 125,000,000 8,800,000 Tool Steels 30,000,000 2,109,000 1,000,000,000 70,300,000 Brass and Bronze 12,000,000 844,000 50,000,000 3,500,000 Copper 10,000,000 703,000 100,000,000 7,000,000 The coefHcient A is the modulus or coefificient of elas- ticity. Such expressions as that above given, the constants being determined, for each case, by experiment, may be made to represent the method of variation of resistance with increas- ing distortion with every method of strain. The equations, therefore, the relation between ordinate and abscissa being algebraically expressed, may be made to form a means of integrating the area \y dx, and determining its magnitude. The Series of Elastic Limits — If, at any moment, the stress producing distortion is relaxed, the piece recoils and continues this reversed distortion until, all load being taken off, the recoil ceases and the piece takes its "per- manent set." This change is shown in the figure at f'f", the gradual reduction of load and coincident partial restora- tion of shape being represented by a succession of points forming the line f" f", each of which points has a position which is determined by the elastic resistance of the piece as 348 MATERIALS OF CONSTRUCTION— NON-FERROUS METALS. now altered by the strain to which it has been subjected. The distance Of" measures the permanent set, and the dis- tance/"/'" measures the recoil. The piece now has qualities which are quite different from those which distinguished it originally, and it may be re- garded as a new specimen and as quite a different metal. Its strain-diagram now has its origin at/", and the piece being once more strained, its behavior will be represented by the curve/"/" ^"/', a curve which often bears little resemblance to the original diagram 0,f,f'. The new diagram shows an elastic limit at e^, and very much higher than the original limit e". Had this experiment been performed at any other point along the line//', the same result would have followed. It thus becomes evident that the strain-diagram is a curve of elastic limits, each point being at once representative of the resistance of the piece in a certain condition of distortion, • and of its elastic limit as then strained. The ductile, non-ferrous metals, and iron and steel, and the truly elastic substances, have this in common — that the effect of strain is to produce a change in the mode of resist- ance to stress, which results, in the latter, in the production of a new and elevated elastic limit, and in the former in the introduction of such a limit where none was observable be- fore. It becomes necessary to distinguish these elastic limits in describing the behavior of strained metals, and, as will be seen subsequently, |he elastic limits here described are, under some conditions, altered by strain, and we thus have another form of elastic limit to be defined by a special term. In this work the original elastic limit of the piece in its ordinary state, as at e, /, e", etc., will be called either the Original, or the Primitive, Elastic Limit, and the elastic limit corresponding to any point in the strain-diagram produced by gradual, unintermitted strain, will be called the Normal Elastic Limit for the given strain. It is seen that the dia- gram representing this kind of strain is a Curve of Normal Elastic Limits. The elastic limit is often said to be that point at which STRENGTH OF NON-FERROUS METALS. IA,9 a permanent set takes place. As will be seen on studying actual strain-diagrams to be hereafter given, and which exhibit accurately the behavior of the metal under stress, there is no such point. The elastic limit referred to ordina- rily, when the term is used, is that point within which recoil, on removal of load, is approximately equal to the elongation attained and beyond which set becomes nearly equal to total elongation. It is seen that, within the elastic Hmit, sets and elongations are similarly proportional to the loads, that the same is true on any elastic line, and that loads and elongations are nearly proportional everywhere beyond the elastic limit, within a moderate range, although the total distortion then bears a far higher ratio to the load, while the sets become nearly equal to the total elongations. Effect of Shock or Impact ; Resilience.— The be- havior of metals, under moving, or " live," load and under shock, is not the same as when gradually and steadily strained by a slowly applied or static stress. In the latter case, the metal undergoes the changes illustrated by the strain diagrams, until a point is reached at which equilibrium occurs between the applied load and resisting forces, and the body rests indefinitely, as under a permanent load, without other change occurring than such settlement of parts as will bring the whole structural resistance into play. When a freely moving body strikes upon the resisting piece, on the other hand, it only comes to rest when all its kinetic energy is taken up by the resisting piece ; there is then an equality of vis viva expended and work done, which is expressed thus : WV^ . . , . . n^ 1 p dx —p„ Jo in which expression W is the weight of the striking body, V its velocity, p the resisting force at any instant, pm the mean resistance up to the point at which equilibrium occurs, and s is the distance through which resistance is met. 3 5° MA TERIALS OF CONSTR UCTION—ffON-FERRO US ME TALS. As has been seen, the resistance may usually be taken as varying approximately with the ordinates of a parabola, the abscissas representing extensions. The mean resistance is, therefore, nearly two-thirds the maximum, and 2g- = p dx= pmS = Y^et = ae', nearly where e is the extension, and t the maximum resistance at that extension, and a a constant. Brittle materials, like hard bronzes and brasses, have a straight line for their strain- diagrams, and the coefficient becomes ^ instead of ^, and Resilience, or Spring, is the work of resistance up to the elastic limit. This will be called Elastic Resilience. The modulus of elasticity being known, the Modulus of Elastic Resilience is obtained by dividing half the square of the maximum elastic resistance by the modulus of elasticity, E, as above, and the work done to the "primitive elastic limit " is obtained by multiplying this modulus of resihence by the volume of the bar.* The total area*of the diagram, measuring the total work done up to rupture, will be called a measure of Total or Ulti- mate Resilience. Mallett's Coefficient of Total Resilience is the half product of maximum resistance into total extension. It is correct for brittle substances and all cases in which the primitive elastic limit is found at the point of rupture. With tough materials, the coefficient is more nearly two-thirds^- and may be even greater where the metal is very ductile, as, e.g., pure copper, tin, or lead. Unity of length and of section * Rankine and some other writers take this modulus as-=r, instead of i — STRENGTH OF NON-FERROUS METALS. 35 1 being taken, this coefficient is here called the Modulus of Resilience. When the energy of a striking body exceeds the total re- silience of the material, the piece will be broken. When the energy expended is less, the piece will be strained until the work done in resistance equals that energy, when the striking body will be brought to rest. As the resistance is partly due to the inertia of the particles of the piece attacked, the strain-diagram area is always less than the real work of resistance, and, at high ve- locities, may be very considerably less, the difference being expended in the local deformation of that part of the piece at which the blow is received. In predicting the effect of a shock' it is, therefore, necessary to know not only the energy stored in the moving mass and the method of variation of the resistance, but also the striking velocity. To meet a shock successfully it is seen that resilience must be secured sufficient to take up the shock without rupture, or, if possible, without serious deformation. It is, in most cases, necessary to make the elastic resilience greater than the maximum energy of any attacking body. Moving Loads produce an effect intermediate between that due to static stress and that due to the shock of a freely mov- ing body acting by its inertia wholly ; these cases are, there- fore, met in design by the use of a high factor of safety, as above. As is seen by a glance at the strain-diagram,//", (Fig. 55), the piece once .strained has a higher elastic resilience than at first, and it is therefore safer against permanent distortion by moderate shocks, while the approach of permanent extension to a limit renders it less secure against shocks of such great intensity as to endanger the piece. When the shock is completely taken up, the piece recoils, as at e'^f'f" , until it settles at such a point on that line — as- suming the shock to have extended the piece to the point e'^ — that the static resistance just equilibrates the static load. This point is usually reached after a series of vibrations on either side of it has occurred. With perfect elasticity, this 352 MATERIALS OF CONSTRUCTION— NON-FERROUS METALS. point is at one-half the maximum resistance, or elongation, attained. Thus we have but/ varies as x within the elastic limit, which limit has now risen to some new point along the line of normal elastic limits, as ^". Taking the origin at the foot oif'f", since the variations of length along the line Ox are equal to the elongations and to the distances traversed as the load falls, and as stresses are now proportional to elongations, p=ax ;- Wh= Ws ; and W=P when the resisting force is/, the elongations x, while h and s are maximum fall and elongation, and P is the maximum resistance to the load at rest. Then p dx=a \ X ax = —s' — Ws .-. s = jf io 2 a For a static load, if s' is the elongation, W = -.P=a^ ' .-. s' = w a nee. t s ' '4 and the extension and the corresponding stress due to the sudden application of a load are double those produced by a static load. The Variation of Form of Test-Piece so consid- erably modifies the apparent tenacity of iron and steel that it is necessary to note the size and shape of the specimen tested. When a piece of metal is subjected to stress and slowly pulled asunder, it will yield at the weakest section first, and if that section is of considerably less area than adjacent parts. STRENGTH OF IRON AND STEEL. 353 Fig. 56, or if the metal is not ductile, it will often break sharply, and without stretching appreciably, as seen in Fig. 58 ; the fractured surface will have a granular appearance, and the behavior of the piece, as a whole, may be like that of a brittle casting, even although actually made of tough and ductile metal, when the piece is deeply scored ^ i ? \ Fig. 56. — Incorrect. Fig. 57. — Correct. Forms of Test Pieces for Tension. When a bar of very ductile metal, of perfectly uniform cross-section, Fig. 57, is broken, on the other hand, it will, at first, if of uniform quality, gradually stretch with a nearly uniform reduction of section from end to end. Toward the ends, where held by the machine, this reduction of area is less perceivable, and on the extreme ends where no strain can occur, except from the compressing action of the grips, the original area of section is retained, diminution taking place from that point to the most strained part by a gradual taper or by a sudden reduction of section, according to the method adopted of holding the rod. When the stress has attained so great an intensity that the weakest section is strained beyond its elastic limit, " flow " begins there, and, 23 354 MATERIALS OF CONSTRUCTION— IRON AND STEEL. while the extension of other parts continues slowly, the por- tions immediately adjacent to the overstrained (section stretch more and more rapidly as this local reduction of section continues, and finally fracture takes place. This locally reduced portion of the rod has a length which is de- pendent upon the character of the metal and the size of the piece. Hard and brittle materials exhibit very little reduction and the reduced portion is short, as in Fig. 58; ductile and tough met- als exhibit a marked reduction over a length of several diameters, and great reduction at the fractured section, as seen in Fig. 59. Of the samples shown in the figures, the first is of a good, but a badly worked, iron, and the second from the same metal after it had been more thoroughly worked. When the breaking section is determined by deeply grooving the test-piece, the results of test are higher by 5 or 10 per cent, than when the cylinders are not so cut, if the metal is hard and brittle, and by 20 to 25 per cent, with tough and ductile irons or steels. In ordinary work this difference will average at least 20 per cent, with the ductile metals. A good bridge or cable iron in pieces of i inch (2.54 centimetres) diameter cut from 2-inch (5.58 centimetres) bar, exhibited a tenacity of 50,000 pounds per square inch in long test-pieces, and 60,000 in short grooved specimens (3,515 to 4,218 kilogrammes per square centimetre). Cast irons will give practically equal results by both tests, as will hard steels and very coarse- grained hard wrought irons. Since these differences are so great that it is necessary to ascertain the form of samples tested before the results of test can be properly interpreted, it becomes advisable to use a test-piece of standard shape and size for all tests the results of which are to be compared. The figures given hereafter, when not otherwise stated, may be assumed to apply to pieces Fig. 58. STRENGTH OF IRON AND STEEL. 355 of one half square inch area (3.23 square centimetres) of sec- tion, and at least 5 diameters in length. This length is usually quite sufficient, and is taken by the Author as a minimum. For other W-0 ' * lengths, the extension is measured by a con- stant function of the total length plus a func- tion of the diameter, which varies with the quality of the metal and the shape of the test- piece. It may be expressed by the formula e = al -V f{d) The elongation often increases from 20 up to 40 per cent., as the test-piece is shortened from 5 inches(i2.7 centimetres) to ^ inch (1.27 centimetres) in length, while the contraction of section is, on the other hand, decreased from 50 down to 25 per cent., nearly. Fairbairn,* testing good round bar-iron, found that the extension for lengths varying from 10 inches (25.4 centimetres) to 10 feet (3.28 metres) could be expressed, for such iron, by the for- mula ^ = 18 -f- 25 FIG. 59. where Z is the length of bar in inches, this becomes ^ = IS + 63-5. In metric measures / = length in centimetres ; e = elongation per unit of length. This influence of form is as important in testing soft steels as in working on iron. Col. Wilmot, testing Bessemer " steel " at the Woolwich Arsenal, G. B., obtained the follow- ing figures : Useful Information, Second series, p. 301. 356 MATERIALS OP CONSTRUCTION— IRON AND STEEL. Tenacity : Form. Test-piece. Lbs. per sq. in. ; kilogs per sq. cm. Grooved, Fig. 56, Highest 162,974 n,457 Lowest 136,490 9.595 Average 153.677 10,803 Longcylinder Highest 123,165 8,658 Lowest 103,255 7.259 Average 114,460 8,047 The difference amounts to between 30 and 35 per cent., the groove giving an abnormally high figure. It is evident from the above that the elongation must be proportionably much greater in short specimens than in long pieces. This is well shown in this table of tests made by Beardslee, for the United States Board.* TESTS OP TEST-PIECES OF VARYING PROPORTIONS — TENSION. a R STRESS WHEN 1 DJAME- PIECE BEGAN BREAK ING- H TER. \ TO STRETCH STRIISS. OBSERVABLY. § % V< A S «; S "S I ^ £ D* ^ s? .J •n y •0 s-s •a IB-S ^ &B «. Bi & ■a a: % u IE \h B. fi h 5 « en In. In. In. In. Lis. Lbs. Lbs. Lis. z 5.000 6.522 10.0 .798 568 49-3 13,400 26,800 26,000 51,989 Elastic limit, 26,795 lbs. per sq. in. 2 IPlS 5.204 32.0 • 798;. 564 50.0 14,000 28^000 26,200 52,389 Elastic limit, 28,194 lbs. per sq. in. ■\ 4. SOD S-853 30.0 •797-584 46.1 14,000 28,2QO 26,190 52,495 Elastic limit, 28,062 lbs. per sq. in. 4 3-Soo 4.6=5 31.6 .791.570 480 13,000 26,450 26,070 53.052 Elastic limit, 27,268 lbs. per sq. in. .$ 3.000 3-977 33-0 •792.571 48.0 14,000 28,420 26,100 52,984 b 2.472 3.266 32.1 .799.589 45.6 14,000 27,920 26,500 52,852 7 1.989 2.644 3=-9 ■79S •.W 4.';-° 14,000 28,000 26,500 53,169 8 1.500 2.026 3S-0 ■797 •■■iflo 45.2 15,500 31,320 26,275 52,666 <) I.OOO i-.l.M 35-4 .798 .600 43- 5 16,67s 33i35o 26,590 53.169 10 0.500 0.708 41.6 .798 ■635 36.6 18,760 37i520 28,665 57.318 With such brittle materials as the cast irons, the differ- ence becomes unimportant. Beardslee found a difference of but I per cent, in certain cases. The more brittle the mate- rial the less this variation of the observed tenacity. As will be seen later, even more important variations fol- low changes of proportion of pieces in compression. No test-piece should be of very small diameter, as inaccuracy is * Report, p. 104. STRENGTH -OF IRON AND STEEL. 357 more probable with a small than with a large piece, and the errors are more likely to be increased in reduction to the stress per square inch. The length should not be less than four times the diameter in any case, and with soft, ductile metal, five or six diameters would be preferable for tension. Where much work is to be done, it is quite important that a set of standard shapes of test pieces should be selected, and that all the tests should be made upon samples worked to standard size and form. Thus, tension-pieces are often made of the shapes seen in the figure, when testing square, cylin- drical, or flat samples, or samples cut from the solid. The last is a shape called for under the U. S. inspection laws when testing boiler-plate ; but it should never be used, if choice is permitted, as it gives no chance of stretching, and is therefore nearly useless as a gauge of the quality of the metal ; it will undoubtedly be abandoned in course of time, as it invariably gives too high a figure, and does not distin- guish the hard and brittle from the better and tougher materials which are desired in construction. The dimensions adopted by the Author are one-half square inch (3.23 square centimetres) section for all metals ex- cept the tool steels (0.798 inch ; 2 centi- metres diameter, when round), and one-eighth or one- quarter square inch (0.81 to 1.61 square centimetres area; 0.398 or 0.565 inch I or 1 .4 centimetres di- ameter) for the lat- ter, at the smallest cross-section. Kent, who sketches the above, takes these ^i°- 6o.-Shapes for Test Pieces. shapes, making them, if of tool steel, fj^ inch diameter (1.75 centimetre), or | square inch (2.44 square centimetres) area ; ^ IS TOSOl ==^ 1 ■ t6"T03D" "^ ^ ^ < ■ m'rozo' fl" _ _ _ — - 1 ^ < 16'TOZO' rt -^^ ^^^^ ""'- I'""'^ ^ ^■^H - 8 TO lgl » 358 MATERIALS OF CONSTRUCTION— IRON AND STEEL. in other metals either |-inch (1.9 centimetres) diameter of 0.44 square inch (2.84 square centimetres), or as above. The edges should be true and smooth, and the fillets ^inch radius. For compression tests of metal, i inch (2.54 centimetres) long and |-inch (1.27 centimetres) diameter, ends perfectly square, is recommended. For stone and brick, a 2-inch (5.08 centimetres) cube. Transverse test-pieces should not be less than I foot, nor more than 4 feet in length, vi^hen to be handled in ordinary machines. The standard specimen will be taken as above, and good wrought iron of such shape and size should exhibit a tenacity of at least 50,000 pounds (3,515 kilogrammes per square cen- timetre) if from bars not exceeding 2 inches (5.08 centimetres) diameter, and should stretch 25 per cent, with 40 per cent, reduction of area. Such test-pieces have the advantage of giving uniform comparable and minimum figures for tenacity, and of permitting accurate determinations of elongation. Test-pieces are only satisfactory in form when turned in the lathe, as the coincidence of the central line of figure with the line of pull is thus most perfectly insured. When, as with sheet metal, this cannot be done readily, care must be taken to secure proportions of length and cross-section as nearly alike those of the standard test-piece as possible, and to secure symmetry and exactness of form and dimension ; such pieces are liable to yield by tearing when not well made and properly adjusted in the machine. The Method of Use of Testing Machines is, in general, the same for all cases, and is only modified by methods of holding the piece and of taking measurements, which may be peculiar to the machine used. The piece being carefully adjusted in the clamps, and the measuring apparatus so attached that its indications may be relied upon, and that it is not likely to be injured by any accident during the test, the load is very slowly and steadily applied. At intervals, readings of elongation and of load are taken and recorded, and the observer, noting their rate of increase, after a time detects a change in their ratio which in- dicates that the elastic limit is reached, and that extension is STRENGTH OF IRON AND STEEL. 359 taking place more rapidly with each accession of load. The fracture of the piece can usually be anticipated, and the measuring apparatus is removed before danger is incurred of its injury by the shock of breaking. With brittle materials the final break takes place suddenly, and without warning. The observer must therefore depend upon his knowledge that such material is likely to break at not far from a known load. Ductile substances usually pass a limit of maximum load, and break after stretching an appreciable amount with gradually diminishing resistance. When " sets " are to be measured, all load is removed at intervals during the test, and the piece is permitted to recoil. The difference between its length now unloaded and the origi- nal length, is the " set." This set is usually partly temporary, and the piece, if left unloaded, will often very slowly contract for a considerable time, thus perceptibly reducing the set, which then becomes permanent. It is not advisable to take measurements of set unless for a special purpose, as each re- laxation of the piece modifies its resisting power, and makes comparison with other samples less easy and satisfactory. When broken, the pieces are removed from the machine, their final length is measured, and they are carefully exam- ined to obtain such knowledge of the quality of the metal as may be secured by a study of their texture and of the charac- ter of their fracture. The Method of Record is a matter of some impor- tance in making researches relating to the strength of mate- rials. The Author has been accustomed to use printed blanks for such work. The following are the headings adopted. Of these blanks, the first is used indifferently for either ten- sion or compression, and the last for miscellaneous purposes. An examination of the records to be given in the follow- ing pages will show that the customs of the various depart- ments, as well as of individual investigators, differ greatly, not only in the extent to which the minuteness of measure- ment is carried, but also in their methods of securing results and recording them. The columns in the blanks here given are not always all filled out. 360 MA TERIALS OF CONSTRUCTION— IRON AND STEEL. I . bS \^ § i s i 'i.> fi S ■s-^^ I «OJ ^ a 1 ■0 0< 11^ % h" Ms KM 41^ ? ild ti'SS, 5S i's 35 » = p i s ■aiBui;;in ■ymii iV'V 00 ■J3)UU1!!a •lii&a ifi gz ■ll]pT33Ja % ■illSusT •HHVW ^VNIOIHO » 'HDHHOS • aMVN i n i i It 511 II yg ii" (0(3 §i ■;if. m %^s Sh* Mo •icm^v 5i S^5 •g II g S 0! ^ i M ^ S! H R riM^ s 3 ^in oi •g II a tEl s ■35, h •y •jsisureia ■? § •ipdsa S ■f 'qipeug Q ill III Q-S5 ■MHVW iVNioiao (^ z •HDHnOS 'HHVM 1 i STRENGTH OF IRON AMD STEEL. 361 I o H sS ^ \ 3. "^ M td 1 ^i^ 1 II K S-- 3 ^ ffl< 1 ^1^ fi II R^ 3S ^ K] |<* ^ % ^l5 g § II H w 0°J S I § 3 M W^ M< t?!S 2- •ajnionjjs So •mnuiixTjj^ a'^ _ ^a •JOOJJ R W J i ■z 3 _ mm a K^ ■iis Bg- «lfl 17 -j343mE!a ■I^^3^^^ ■Havw IVMIOIHO d z •aoanos • 31 ft ra o Pi 1-1 < 'A M O o o 'A < >— I o S o 362 MATERIALS OF CONSTRUCTION— IRON AND STEEL. 2Z^. Records of Tests. — The following are figures de- rived from such a test by tension, as made for the Author: TEST OF WROUGHT IRON, LENGTH 8" (19.32 CM.) ; DIAM., O.798" (2.O3 CM.). Actual. Lbs. 13,500 LOADS. MICROMETER EXTENSIONS. SETS. Actual. Per sq. inch. READINGS. Actual. Per Cent. Actual. Per Cent. 150 2,000 4,000 6,000 8,000 10,000 150 11,000 12,000 150 13,000 13.500 14,000 150 15,000 150 17,000 150 19,000 150 21,000 150 22,000 150 22,500 23,000 23,500 23.750 21,800 4,000 8,000 12,000 16,000 12,000 22,000 24,000 26,000 27,000 28,000 30,000 34.000 38,000 42,000 44,000 45.000 45,000 47,000 47.500 43,600 .6600 .662S .6637 .5546 .56o6 .5630 .6600 .6639 .6700 .6603 • 6715 .6728 .7242 ■7133 ■7535 .7417 .8474 .8326 .9720 • 9562 1.1710 1.1524 I ■ 3303 I. 3102 1-4575 1.5610 1.7646 9- 9. ■7913 ■7910 .7922 ■7930 .7946 •7948 .7914 ■7951 ■7953 ■79^ .7967 '■7959 .8424 •8351 .8712 .8632 .9618 .9518 1.0856 1.0732 I. 2811 1.2663 1.4381 I. 4212 I. 5441 1.6670 1.8693 47 54 .0013 .0023 •0035 .0050 .0058 .0064 .0070 .0080 .0087 ■0577 .0867 .1790 .3032 .5004 !6586 •7752 .8884 1.0913 I . 4700 1 . 5400 .016 .029 .044 • 063 ■073 • 030 •037 .100 .log .721 1.084 2.238 3.790 ^■255 8.233 9.690 11.105 13.841 IS. 375 19.250 .0001 .0003 .0486 •0763 !i656 .2391 •4337 .6401 .001 .004 .608 .960 2.0S3 3.613 6.043 S.ooi ELASTIC LIMIT. Kgs. 6140 Lbs. per sq. in. 27,000 Kgs. per sq. cm. 1,898 BREAKING LOAD. Original Sect. Fractured Sect. Lbs. per Kgs. per Lbs. per Kgs. per sq. m. sq. cm. sq. in. sq. cm. 47.500 3,340 69,840 4,910 Ultimate Elongation, per cent, of length = igi. Reduction of Area, per cent., ^ 31.99. Modulus of Elasticity = 24,365,000 lbs. on sq. in. Modulus of Elasticity r= 1,712,860 kilo- grammes on sq. cm. FINAL DIMENSIONS. Length = 9" -54 Diameter = o".658 STRENGTH OF IRON AND STEEL. 363 The following figures are derived from experiments upon good examples of several grades of iron plate. Sixteen experiments upon high-grade boiler plate resulted as follows : * Measures. Metric. British. Average breaking weight 3,803 54,123 Highest " " 4,007 57,012 Lowest " " 3,642 51,813 Variation \a. per centum of highest g.i Fifteen experiments made upon the best grades of flange irons gave : Measures. Metric. British. Average breaking weight 2,960 42, 144 Highest " " 3,746 53,277 Lowest " " 2,320 33,003 Variation m per centum of highest 38 Six experiments upon hard Bessemer steel gave: Measures. Metric. British, Average breaking weight 5,877 83,621 Highest " " 6,087 86,580 Lowest " " 5.237 74,509 Variation \xv per centum- of highest 14 Five experiments were made upon the best boiler plate : Measures. Metric. British. 1 Average breaking weight 4,I77 58,984 Highest " " ,-•■ 4,710 64,000 Lowest " " 3,887 55,300 Variation x^per centum of highest 14 Six experiments upon samples of tank-iron, by three makers, gave: Measures. Metric. British. Average breaking weight. Maker No. I 3,079 43,831 Highest " " " " .... 3,739 S3,i74 Lowest " " " " .... 2,533 36)1" Variation \x\.per centum of highest 32 * Journal Franklin Institute, 1872, 364 MATERIALS OF CONSTRUCTION— IRON AND STEEL. Measures. Metric. British. Average breaking weight, Malcer No. 2 2,953 42,011 Highest " " " " 3.392 48,425 Lowest " " " " 5,510 35.679 Variation va. per centum of highest 28 Average brealcing weight. Maker No. 3 2,8g6 4I>249 Highest " " " " 3,676 52,277 Lowest '• " " " 2,320 33,003 Variation \a.per centum of highest 38 In another series, of which the results were supplied to the Author by Mr. C. Huston, the following figures were obtained : TESTS OF BOILER-PLATE. TENACITY. T ELASTIC LIMIT. Lbs. per sq. inch. Kilogs. per sq. cm. Lbs. per sq. inch. Kilogs. per sq. cm. "Best boiled" "Best flange charcoal ". . . . " Double-worked boiled ". . "Best flange" 55.000 56,000 56,400 3.550 3.557 3.560 31.500 35.000 40,000 30,000 2,214 2,460 2,8i8 2,109 The figures are all higher than those usually expected by the engineer when buying iron. The last mentioned grade had a tenacity per square inch of fractured section of 87,000 pounds (5,673 kilogrammes per square centimetre), and was reduced in section 20 per cent. * Best " C. H. No. i " plate, ^th inch (0.95 centimetre) thick, tested by Kent, exhibited a tenacity of very nearly 60,000 pounds per square inch (4,218 kilogrammes per square centimetre) of original area, elongated 1 5 per cent., and its STRENGTH OF IRON AND STEEL. 365 resistance per square inch of fractured section was 76,000 pounds (5,378 kilogrammes per square centimetre.) Variations of Tenacity with Size — Bar-irons ex- hibit a wide difference of strength, due to difference of sec- tion alone. This variation may be expressed approximately with good irons, such as the Author has studied in this re- lation, by the formulas, T= 56,000 — 20,000 log d '\ T^ = 4,500 - 1,406 log d„ ) Where T and T^ measure the tenacity in British and metric measures respectively, and d and d^ the diameter of the piece, or its least dimension. Where it is desired to use an expression which is not loga- rithmic it will usually be safe to adopt in specifications the following : ^ 60,000 . ™ _ 80,000 ■^ ~ ^t/ ' ™ ^d The Edgemoor Iron Company adopt, for wrought iron in tension, the formula, _ ^ 7,000 A T= 52,000 - " „ in which A is the area, and B the periphery of the sec- tion.* The experiments made by Beardslee for the United States Board gave results for rolled iron of various qualities, rang- ing from 60,000 to 46,000, according to size {% inch to 4 in.). * Ohio Railway Report, i88i, p. 379. 366 MATERIALS OF CONSTRUCTION— IRON AND STEEL. TENACITY OF IRON WIRE. ORIGINAL DIAMETER. FINAi. TENACITY. NO. Inches. Centi- metres. Inches. Centi- metres. Lbs. per sq. in. Kilogs. per sq. cm. lO .1340 •340 ■ 1330 ■338 92,890* 6,530 II .1205 • 305 .1185 • 303 84,442 \ 5,936 12 .1040 • 255 .1010 •257 93,158 6,555 13 .0925 .225 .0920 ■234 100,297* 7,050 14 .0800 .203 .0795 .203 94,299 6,629 I'^ .0710 .178 .0680 •173 98,384!: 6,915 16 .0640 .163 .0635 .161 93,876 6,600 17 •0535 ■139 •0532 • 135 105,871 7,442 18 .0465 .118 .0400 .101 1I9,536§ 8,403 IQ .0385 .098 .0385 .098 87,617 6,159 20 •0335 .085 • 0335 .085 111,184 7,816 21 .0290 .074 .0290 .074 113,546 7,982 The tenacity of "medium soft" telegraph wire may be taken as follows, for the several sizes obtainable in the market : TENACITY OF IRON TELEGRAPH WIRE. DIAMETER. WEIGHT TENACITY. T PER YARD OR METRE. NO. B. W. G. In. Cm. Lbs. Kilogs. Lbs. Kilogs. 1 0.30 0.76 0.683 0.313 4,000 1,800 2 0.28 0.70 0-599 0.272 3,400 1,500 3 0.26 0.66 0.517 0.235 2,900 1,300 4 0.24 0.61 0.440 0.200 2,500 1,150 5 0.22 0.56 0.370 0.170 2,200 1,000 6 0.20 0.50 0.305 0.139 1,800 800 7 0.19 0.48 0.262 0.II9 1,500 650 8 0.17 0.43 0.221 O.IOO 1,200 550 9 o.i6 0.40 0.184 0.084 950 4*0 10 0.14 0-35 0.150 0.068 800 360 12 O.II 0.28 0.092 0.042 500 230 14 0.09 0.23 0.055 025 350 160 16 0.07 0.18 0.032 0.015 200 90 * Hard drawn. ■)■ Soft. :|:Soft; Extension, 0.12}. § Very hard. Comparison sample of Norway iron, very soft, broke at just one-half this figure. STRENGTH OF IRON AND STEEL. 367 Very soft and pure wire will have 20 or 25 per cent, less tenacity than is above given, while hard wire may give figures exceeding the above by an equal amount. In consequence of this variability it would be useless to express tenacities more precisely. Turning iron down has no important effect on the tenacity. The considerable variations always observable in the gen- eral rate of increase of tenacity, which, other things being equal, accompanies reduction of size of wire, are due to the hardening of the wire in the draw-plate, and occasional restoration to its softest condition by annealing. Beardslee has found the change of tenacity in forged and rolled bars, above noted, to be due to differences in amount of work done in the mill upon the iron. The extent of reduction of the pile sent to the rolls from the heating furnace is variable, its cross-sectional area being originally from 20 to 60 times that of the bar, the higher figure being that for the smallest bars. On making this reduction uniform, it is found that the tenacity of bars varies much less, in different sizes, and that the change becomes nearly uniform from end to end of the series of sizes, and becomes also very small in amount. By properly shaping the piles at the heating furnace, and by putting as much work on large as on small bars, it was found that a 2-inch (5.08 centimetres) bar could be given a strength superior by over 10 per cent., and a 4-inch (10.17 centimetres) could be made stronger by above 20 per cent, than iron of those sizes as usually made for the market. The surface of a bar is usually somewhat stronger than the interior. The Limit of Elasticity will be found at from two-fifths the ultimate strength in soft, pure irons, to three-fifths in harder irons, and from three-fifths in the steels to nearly the ultimate strength with harder steels and cast irons. Barlow found good wrought iron to elongate one ten-thousandth its length per ton per square inch up to the limit at about 10 tons. The relation between the series of elastic limits, and the maximum resistance of the iron or the steel is well shown in strain-diagrams, which exhibit graphically the varying 368 MATERIALS OF CONSTRUCTION— IRON AND STEEL. relation of the stress applied to the strain produced by it throughout the process of breaking. Experiments on Long Bars are seldom made, and but few are on record. The following data were obtained from tests made for the Phoenix Iron Company : TESTS OF LONG BARS OF WROUGHT IRON. MODULUS OF NUMBER OF SIZE. LENGTH. STRETCH. ELASTICITY. BARS. INCHES. E 23 3, X \ 35' 0" .2587 32,470,000 24 3i X li 35' 0' .2617 32,098,000 9 4 X ijif 27' 6" .2033 32,464,700 24 3i X li 35' 0" .2500 33,600,000 24 3 X J 35' 0" .2633 31,902,000 12 4 X ij 35' 0" .2692 31,203,000 24 2x1 24' 9i" .1948 30,544,000 36 z'-" 11 9" •0953. 29,380,000 48 2i" u 9 .0955 29,319,000 68 2 " II II .0998 28,056,000 48 2 " 11' 9" .1008 27,777>777 72 2 " II 9 .0940 29,787,000 120 2j" II 9 .0947 29,567,000 Repeatedly Piling and Reworking improves the quality of wrought iron up to a limit at which injury is done by overworking and burning it. Clay's experiments on good fibrous puddled iron replied, reheated, and reworked, resulted as shown in the table on next page.* The iron thus treated exhibits increasing strength until it has been reheated five or six times, and then gradually loses tenacity at a rate which seems to be an accelerating one. Forging iron is similar in effect, and improves the met^ up to a limit seldom reached in small masses. The forging of large masses usually includes too often repeated piling and welding of smaller pieces, and it is thence * Fairbaim, p. 249, STRENGTH OF IRON AND STEEL. 369 found difficult to secure soundness and strength. This is particularly the case where the forging is done with hammers of insufficient weight. The iron suffers, not only from re- heating, but from the gradual loosening and weakening of the cohesion of the metal within the mass at depths at which the beneficial effect of the hammer is not felt. EFFECT OF REHEATING. TENACITY. T NUMBER, QUALITY. Pounds per Kilogrammes per square inch. square centimetre. I Puddled Bar. 43,904 3,086 2 Piled and reheated. 52,864 3.718 3 Replied and reheated. 59,585 4,190 4 59,585 4,190 5 57,344 4,028 6 61,824 4,344 7 59.585 4,190 8 57,344 4,028 9 57,344 4,028 10 54,104 3,802 II 51,968 3,655 12 i( t( (1 43.904 3,086 The effect of prolonged heatingis sometimes seen in agran- ular, or even crystalline, structure of the iron, which indicates serious loss of tenacity. Large masses must always be made with great care, and used with caution and with a high factor of safety. Ingot iron is always to be preferred to welded masses of forged material for shafts of steamers and similar uses. The Tenacity of Ingot Irons and Steels is less sub- ject to variation by accidental modifications of structure and composition than is that of wrought iron. The steels are usually, homogeneous and well worked, and are comparatively free from objectionable elements, their variation in quality being determined principally by the amount of carbon. A singular uniformity of tenacity and of elastic limit is 24 370 MATERIALS OF CONSTRUCTION— IRON AND STEEL. observed within limited ranges of quality, with sudden changes at the limits of each range. On the whole, a gradual increase, both in tenacity and in elastic limit, is seen as the proportion of carbon is increased. The modulus of elasticity varies irregularly within a moderate range, and is evidently not affected by the proportion of carbon present. The qual- ity of the metal is usually determined principally by the pro- portion of carbon, but is also affected, to a considerable extent, by the silicon and manganese, as well as by phos- phorus. The considerable variation here exhibited is partly due to the fact that these steels were supplied by several makers, who presumably used iron from different ores and adopted different mixtures in the crucible, and partly due to the varying hardness produced by accidental variation in rate of cooling, when delivered hot from the rolls. The strength of good specimens of these metals, as they came from the mill, has been found by the Author to be, as a minimum, about T = 6o,ooo + 70,000 C \ T,^ = 4,218 4- 4,921 C) where T is the tenacity in pounds per square inch, and T^ in kilogrammes on the square centimetre ; * C is the percentage of carbon. For^ annealed samples f of good ingot iron and steel, T = 50,000 + 60,000 C ) ^m = 3,51s + 4,218 C) Thus, as illustrating these cases, the Author has found the following figures by test : * Tram. Amer. Soc, C. E., 1874. f Structures in Iron and Steel ; Weyrauch, translated by Dubois : N. Y. , 1877. STRENGTH OF IRON AND STEEL. 371 TENACITY OF STEEL. TENACITY. T By Test. By Calculation. Per cent. Lbs. per square Kilogrammes per square centimetre. Lbs. per square inch. Kilogrammes per square centimetre. 0-53 0.65 0.80 0.87 l.OI 1.09 79,062 93.404 99.538 106,979 109,209 "6,394 5,558 6,566 6,997 7,520 7,677 8,183 81,740 88,940 98,060 102,020 110,300 113.480 5.746 6,153 6,893 7.171 7.754 7.978 The Author would adopt the above formulas to deter- mine values to be inserted in specifications. Bauschinger, experimenting upon Ternitz Bessemer steel, deduced the following : Tr. = 4,350 (I + C')) i, e. T ■ = 61,870 (i + C) ) which equation expresses the results of his tests with great accuracy.* American steels are seen to be slightly stronger than the European. Weyrauch gives as a minimum set of values, such as may be used as a basis for specifications : Tr„ = 3,700 (i + Q) i.e.,. T = 52,625 (i + J which formula is probably also sufficiently exact as express- ing the strength of good, pure iron and steel containing no appreciable quantity of the hardening elements other than carbon. For C = o, T = 52,625 pounds per square inch * "Versuche ueber die Festigkeit des Bessemerstahls, " etc 372 MATERIALS OF CONSTHUCTION— IRON AND STEEL. (3,700 kilogrammes per square centimetre), which is a usual figure for good bridge, cable, and blacksmith's iron of about 2 inches (5.08 centimetres) diameter. As a general rule, the Elongations of Steel of the finest grades are diminished as the tenacity increases, and in steels tested for Trautwine* this reduction is nearly propor- tional to the increase in strength. Calling the shock-resisting power of the piece — or, more correctly, its work of resistance — equal to two-thirds the product of the ultimate resistance by the total elongation, its total resilience, R, we get R = ^ r X -£/ = 4,OQO foot-pounds nearly \ R„^ = % Trn X Elm — 2.81 kilogrammetres ) the first value being that for one square inch sectional area and one foot in length, the latter for one square centimetre area of cross-section and one centimetre in length. Then we have from the above : „,_ 6,000 £/ ^ £1 -^ ■* m for the elongation per inch or centimetre at the point of rupture. This extension varies in crucible steels containing, as in the above examples, the usual proportion of manganese, from 10 per cent, at the lower limit to ^ per cent, at the higher. When care is taken to secure freedom from those elements which produce cold-shortness, higher values of elongation and resilience may be secured. Makers of open-hearth steel have thus often been able to guarantee a tenacity of 80,000 pounds per square inch (5,024 kilogrammes per square centimetre), with an elongation of 20 per cent, and an elastic resistance of 50 per cent, of the * Civil Engineer's Pocket Book, STRENGTH OF IRON. AND STEEL. 373 ultimate. The total resilience of such metal is, therefore, about % X 80,000 X .20 = 10,667 foot-pounds per inch of section and foot of length, nearly 7.5 kilogrammetres for samples one square centimetre in section and a centimetre in length. At the elastic limit, the elastic resilience may be taken at about Re=y2T,x El, one-half the product of the elastic resistance by the elonga- tion. This elongation is usually not far from one-tenth per cent., and the elastic limit rises from two-fifths in soft irons to nearly the ultimate resistance in hardened steel ; for tool steels it may be taken at two-thirds, and for the softer grades usually at one-half. By reducing the carbon and adding manganese some ex- traordinary metals are obtained. A " steel " containing o.io per cent, carbon and 0.45 per cent, manganese, has exhibited a tenacity of 90,000 pounds per square inch (6,327 kilogrammes per square centimetre) and an extension of 25 per cent. The Elongation of Steel Bars may be reckoned at about three-fourths that of iron up to the elastic limit. Boiler and Bridge Plate Steels, made by the pneu- matic and open-hearth processes, have nearly the same strength as bars made by the same methods. The following are figures obtained by Hill* and the U. S. Board. TENACITY OF O. H. BRIDGE PLATE — FRACTURED LENGTHWISE. RESISTANCE ELASTIC. RESISTANCE ULTIMATE. T CARBON. Pounds per square inch. Kilogrammes per square centimetre. Pounds per square inch. Kilogrammes per square centimetre. ELONGATION. 0.30 0.40 0.50 49.353 63,227 65,070 3.469 4.444 4.574 93.339 86,410 83,190 6,561 6,074 5.823 .16 .14 .10 * Trans. Eng'r's. Soc. of West. Pennsylvania, 1880. 374 MATERIALS OF CON'STRUCTION— IRON AND STEEL. •AilAVHO DIdlDHJS KVHW fco r^ O lo r*-, O roco -^ to "*■ I S-co oo CO 09 CO oo t^o ro fO CO in .. , M r-. M o 00 00 CO r« CO CO , t^ c^ t^ t^ t^ b«. U K O -SEia JO sninpoM ■spunod-jooj -h\0 O*O«O>m00CO t^M^M '-' NO moo w « c^ iH ro inoo O t^oo 00 O ( «C4CIN fOfOrOWM mctMl ■spunoci ■looj n[ 'ss3J}S-jooi J assg;ss>s.?~ •uoisssjd •mo3 aiqiidaa •Z3d )SJi| ]B qsu; •B3JB T^UlSlIOJO 1I3UI sjefib's J3d 7- 10 ro K O* CJ'oJ ■^ >> O, '^^^i "- "^ °- *^ "1 "* rC ro ■* cT cTco" 4 10 «"'^=£ £2 « 2.*2. 2 ■* »o Th ■>^ ts.vo N 10 « O m - grow — 0\0^ ** O mpir^moct fO( ■O r«. N -^oo lo -H 00 H ioro-*io( ^'o m p% inco "O t^ »ooo t% >o in t^vo ( •Suiddjio ■*co c ro ^> CO "O "O mco w ov ty o- N "* f) fON ^O ■^^m>-• ^-^* moo - . . , - „ m riOO N C O VO t^ t^'O m « ■suinna ■SUIUBIJ to O- M vo m « w t--CO £ W •«• O M »0 M 10 0\0 i ■* « o in ro 1000 m m ( fOvO CO C5\ "J-CO ^o *o 10 i m 1003 "* m t^ Q. ^ O m m 1 m -^ m t^ '^ w r^ m CO lO^O f^ •Buiiunx "lOOX HaJ iilD S3H3NI anvn&s 'onujihd '3UIIIJJQ ■5UIHOIS ■ -1 r^cooo Hoo N N Noo*o^o foml ....... _ _ — o c- N r- ( 00 ro in H c O ^-00 \0 CO . T mvo lo"© ■ \0 00 M ro lovo wii V3 w:) vo vD ■*'0 worn ujM c-.^^^>o^o io« t^^o o> r^ lovo t^"© f. CO o^ ■* m 1- I- too M \o CO 00 o* »Ot»«rOf^0VOOO0'*^OO"^OCS M vocof^Ooo'-'NioOwioioinr^a w mrofOM roMOooo -*oo m 00 w 00 tovo r^ "XOCON fOCOfO•^^el tMO"4-IO 10CI'*0 0"*OvOOsONMrtt...(OV)N 0> H 00 00 loco ■* fo t^vo rooo o *0 O **• m 00 *o coo M g ■*'-' roinoicT<(oro "4-oo ^■-'*- imoic , I- O \6 r*. o « M fo»o ICONMN fOM'ejMcIctN •Suinuu t^M -.hts. tv oi b to moo o o»vo fi ' 00 00 t^ d 00 y* o\ roco 06 d e^ in d>>o >d ■3iiiujnx Wl fO»0 fl N ' rovo N H ro tn o>oo vo M I in (N. q -r f^ ' Kt^^wVd dcO d ION lOHVO roiO'^ MMWN«MWMrO«NMM-T e o •" 1-1 I- £ 3 i? STRENGTH OF IRON AND STEEL. 375 MEAN TENACITY OF ORDNANCE CAST IRON. SPECIFIC GRAVITY. TENACITY. T CARBON. Lbs. per sq. in. Kgs. per sq. cm. Graphite. Combined. I 2 3 7.204 7-154 7.087 28,805 24,767 20, 148 2,025 1,741 1,416 2.06 2.30 2.83 1.78 1.46 0.82 The best irons were generally richest in combined carbon and in manganese, and lowest in graphite, silicon, and phos- phorus. Iron supplied to the United States army must be of uni- form tenacity, should have a strength of 25,000 to 30,000 pounds per square inch (1,758 to 2,190 kilogrammes per square centimetre), and a specific gravity of about 7.245. Good gun iron is expected to range from 30,000 to 32,000 pounds per square inch (2,190 to 2,250 kilogrammes per square centimetre). Good car-wheel irons often exhibit nearly the tenacity of gun iron, and sometimes elongate three-fourths of i per cent, at fracture. Ordinary irons have a tenacity of about 20,000 pounds (1,406 kilogrammes), and often stretch less than o.i per cent. Dark irons, as No. 2, of good makes, have a tenacity equal to about two-thirds that of No. 4 of the same make, a good iron giving, in experiments by the Author,* 20,500 and 34,407 pounds per square inch (1,441 and 2,419 kilogrammes per square centimetre) respectively. These samples passed the elastic limit at 7,333 and 12,000 pounds per square inch (515 and 844 kilogrammes per square centimetre), and their moduli of elasticity were \i% and 16 millions (nearly) pounds per square inch (8,045 to 11,248 kilogrammes per square centimetre). Their densities were 7.186 and 7.259. The average of a large number of tests of iron of all grades, but usually No. 3 machinery iron, is, in * Report on Salisbury Irons, R. R, Canette, 1877. Pamphlet, 1878. 37^ MATERIALS OF CONSTRUCTION— IRON AND STEEL. tension, 18,800 pounds per square inch, as obtained by the Author, while Hodgkinson quotes, for English cast irons, about 16,000 (1,222 and 1,125 kilogrammes per square centir metre). The following may be taken as figures which should be given by the best sorts of cast irons : TENACITY OF GOOD CAST IRONS. TENACITY. T SPECIFIC Lbs. per sq. in. Kgs. per sq. cm. GRAVITY. 20,000 25,000 30,000 30,000 1,406 1,758 2,109 2, log 7.10 7.22 7.28 7.25 It will usually be found that the best single index of the strength of cast iron is its density ; and the best machinery and good gun irons should, in small castings, have a tenacity of about T = 25,ooo(Z> — 7) + 20,0001 7;.= i,758(Z?-7)+ 1,406 j between the limits of density, i? = 7 ; D = 7.28. In heavy masses the strength of cast iron may be very seriously reduced, and usually is diminished appreciably, by the internal strains due to shrinkage, and by lessened specific gravity. Even in such small variations of section as Hodg- kinson experimented upon — i, 2, and 3-inch sections — this loss of strength was very great ; the relative tenacities were as 100, 80, and "jj, in test pieces from sample-bars such as^re furnished under specification. James, repeating the experi- ment, obtained the figures 100, 66, and 60. The surface of a casting is usually, but not always, stronger than the interior of the mass. STRENGTH OF IRON AND STEEL. 377 The strength of cast iron of the usual foundry grades is generally increased by remelting, partly in consequence of the loss of carbon, and also, possibly, by the refining which occurs during the process. This change was noted by Wade when remelting No. i pig iron. Thus : TENACITY OF REMELTED CAST IRON. SPECIFIC GRAVITY. TENACITY. T Lbs. per sq. in. Kgs. per sq. cm. 7.032 7.086 7.198 7.301 14,000 22,900 30,229 35,786 984 r,6io 2,207 2,516 The same effect is produced by prolonged exposure to the flame of the reverberatory furnace, thus : TENACITY OF CAST IRON. TIME OF FUSION. TENACITY. T Lbs. per sq. in. Kgs. per sq. cm. 17,843 20,127 24,387 34,496 1,254 1,415 1,714 2,425 Bramwell increased the strength of dark grades of cast iron more than 250 per cent, by four hours' fusion. In Fair- bairn's experiments No. 3 iron was melted eighteen times, and a maximum increase of 220 per cent, was observed at the four- teenth melting. 378 MATERIALS OF CONSTRUCTION— IRON AND STEEL. Strain Diagrams of Cast Iron. — The accompanying diagram is the graphical representation of experiments made upon Salisbury cast iron referred to above. It is seen that such metal gives a parabolic strain diagram, and has no definite elastic limit. The Author has been accustomed to assume that the elastic limit may be taken at that point at which a tfC^^. ^...weA*- Sift - Si/tr Jl^cv, t^i"'" X^TTP (t^ njm Fig, 6i. — Cast Iron in Tension. STRENGTH OF IRON AND STEEL. 379 tangent to the curve' makes an angle of 45° with the axes. This is fairly accurate for the harder varieties of iron, and, although less exact for softer irons, leads to no serious error. Stays. — Where flat surfaces are secured against lateral pressure by stay-bolts, as is done in steam boilers, these bolts may yield either by breaking across, or by shear- ing the threads of the screw in the bolt or in the sheet. Such bolts should not be so proportioned that they are equally liable to break by either method, but should be given a large factor of safety (15 to 20) to allow for reduction of size by corro- sion, from which kind of deterioration they are liable to suffer seriously. Wrought iron and soft steels are used for these bolts. They are secured through the plate, and the project- ing ends are usually headed like rivets. Nuts are sometimes screwed on them instead of riveting them when they are not liable to injury by flame. " Button-set " heads are from 25 to 35 stronger than the conical hammered head, and nuts give still greater strength. Experiments made by Chief Engineers Sprague and Tower, for the U. S. Navy Department, lead to the following formula* and values of the coefficient a, p being the safe working pressure, / the thickness of plate, and d the distance from bolt to bolt : Values of a in British and Metric Measures. A. Am. For iron plates and bolts 24,000 1,693 For steel plates and iron bolts 25,000 1,758 For steel plates and steel bolts 28,000 I,g68 For iron plates and iron bolts with nuts 40,000 2,812 For copper plates and iron bolts 14,500 1,020 The" working load is given in pounds on the square inch and kilogrammes per square centimetre, the measurements being taken in inches and centimetres. The heads, where riveted, are assumed to be made of the button shape. * Report on Boiler Bracing ; Washington, 1879. 38o MATERIALS OF CONSTRUCTION— IRON AND STEEL. The diameter of stay is made about 2V^ the number of threads per inch 12, or 14 (5 or 6 per centimetre). A very high factor of safety, as above, is recommended for stays, to afford ample margin for loss by corrosion. Lloyd's Rule for stayed plates is in which p is the working pressure in pounds on the square inch, ^1 the thickness of plate in sixteenths of an inch, and /j is the distance apart of > the stays in inches. The coefificient a has the following value : a: = 90 for plate -j^ inch thick or less ; with screw stays and riveted heads ; a = 100 for plate y\ inch thick or more ; screw stays and riveted heads ; ^ = no for plate •j'^ inch thick or less ; screw stays and nuts ; « = 120 for plate ^ inch thick or more ; screw stays and nuts ; a = 140 for plate -^ inch thick or more ; screw stays with double nuts ; a — 160 for plate i\ inch thick ; with screw stays, double nuts and washers. The Board at Trade of Great Britain prescribes, ^= ' s-6 in which li is the thickness of plate as above, and s is the area of surface supported, in square inches. , a = 100 for plates not exposed to heat, and fitted with nuts and washers of 3" diameter and of ^ the thickness of the plates ; a = 90 for same case, but with nuts only ; STRENGTH OF IRON AND STEEL. 38 1 a = 60 where exposed to steam, and fitted with nuts and washers ; « = 54 for same case, with nuts only ; a = 80 where in contact with water ; with screw-stays and nuts ; a = 60 for latter case, and screw-stays riveted ; « = 36 for plates exposed to steam, screw-stays riveted. Where girder-stays are used, in which expression, d^ t^ are the depth and thickness of the girder, p^ is the pitch of bolts carrying the girder ; 4 is the distance between girders ; w is the width of crown ; / is the length of girder. Where one, two or three, and where four bolts, respectively, carry the girder, a — 500, 750, and 850. Mr. D. K. Clark,* comparing data obtained experimentally with his own formulas for stayed surfaces, gives the follow^ ing as safe values : Deflection to the elastic limit, d=^ 44 where d is the " rise " of the arch, or the deflection, and d^ the " pitch " of the stays. Maximum pressure, / = 407 ^ when t is the thickness and T the tenacity. Elastic resistance. For iron,/ = 5,000 -y a For steel, fi = 5,700 -> d * Inst. C. E., 1877-78, Vol. LIII., abstracts. 382 MATERIALS OF CONSTRUCTION— IRON AND STEEL. It is assumed that the stay-bolts are so secured that the distortion of the plate cannot break them out. The stay- bolt is equally likely to shear and to break in tension when its diameter is twice the thickness of the plate. It should, however, be made a quarter inch larger to allow for corrosion, which is more dangerous on the stay than on the sheet. Cylindrical Boiler-Shells, and other thin cylinders, have a thickness which is determined by the tenacity of the metal and the character of the riveted or other seam. If/ be the internal pressure, T the mean tenacity to be calculated upon along the weakest seam, r the semidiameter, and t the thickness, we have for axial stresses for equilibrium ; and pur" = - 27trtT, i = pr ' 2T But for transverse stresses tending to rupture longitv idinal seams, and Pr- t-- pr T With seams of equal strength in both directions, there- fore, the cylinder is at the point of rupture along the longi- tudinal seams, while capable of bearing twice the pressure on girth seams. It is evident that spheres have twice the strength of cylinders of equal diameter. Thick cylinders are considered in article 248, as they are usually made in cast iron. Flat Boiler Heads are made -both in wrought and cast iron. For these Clark's rules may be used.* • For elastic deflection, 44 * Inst. C. E., Vol. LIII., Abstracts : London, 1877-78. STRENGTH OF IRON AND STEEL. 383 For maximum pressure. or, for iron, For steel. For cast iron, / =0.215^ t P = 10,000 -r- «1 P= ii>5oo — t p= 4.000-7- whe;i t is the thickness, dx the diameter, and T the tenacity. For spherical ends, at di 4v + V where a is 108,000 for wrought iron, 125,000 for steel, 45,000 for cast iron, and v is the versed sine or rise of the head. Lloyd's Rule for cylindrical shells of boilers is abt in which a is 155 to 200 for iron, 200 to 250 for steel, b per cent, of strength of solid sheet retained at the joint, t is the thickness of the plate, and d the diameter of the shell. The value of b is thus reckoned (w = number of rows of rivets) : b = 100 -^^ ^-, for the plate ; A ^ = 100 — -, for rivets in punched holes ; Pil b= 90 — i, for rivets in drilled holes ; pil 384 MATERIALS OF CONSTRUCTION— IRON AND STEEL. where /i is the pitch of rivets ; d^ is their diameter ; a^ is the area of the rivet-section. When in double-shear, X-J^Oy is taken for a-^. The factor of safety is taken at 6, and boilers are tested by water-pressure up to 2p. The iron is expected to have a tenacity of at least 21 tons per square inch ; steel must bear 26 tons (3,307 to 4,095 kilogs. per sq. cm.). Welds are found, when well made, to carry 75 to 85 per cent, of the sheet. Steam-pipe is usually made with an enormous excess of strength, to meet accidental stresses, such as those due to motion of water within them. The Author has tested pipe broken by " water-hammer," as the engineer calls it, to 1,000 pounds per square inch (70 kilogrammes per sq. cm.) after it had been thus cracked in regular work in a long line, while the steam pressure was less than 100 pounds (7 kilogs. per sq. cm.). They had all been previously tested to about one- third this pressure. Strength of Cast-Iron Cylinders. — Cylinders for steam engines are usually given a thickness greatly in excess of that demanded to safely resist the steam pressure ; often, according to Haswell, 2500 8 for vertical cylinders, where d is the internal diameter, and t=^^\ 2000 8 for horizontal cylinders of considerable size. In metric measures, kilogrammes and centimetres, these formulas become t = — + - , nearly 200 3 ' t = —i- + - , nearly 160 3' STRENGTH OF IRON AND STEEL. 385 If ri is the external, and r^ the internal radius, T the tenacity of the metal, t its thickness, and p the intensity of the internal pressure, we have, for the thin cylinder, as an equation for equilibrium : and /rj ^T{r-,- Ti) = Tt, Tt P For the thick cylinder, however, the resistance at any internal annulus of the cylinder is less than T. Thick Cylinders, technically so called, are those which are of such thickness that the mean resistance falls considerably below the full tenacity of the metal, as exhibited in thin cylinders, in low-pressure ■ steam boiler shells, for example. Such cylinders are seen in the "hydraulic" press, and in ordnance. Barlow'^ assumes the area of section unchanged by stress, although the annulus is thinned somewhat by linear exten- sion. If this is the fact, as the tension on any elementary ring must vary as the extension of the ring within the elastic limit, the stress in such element will be proportional to the reciprocal of the square of its radius, i.e., it will be /=^^ and, taking the total resistance as /Vi, when/ is the internal strength of Materials, 1867, p. 118. 25 386 MATERIALS OF CONSTRUCTION— IRON AND STEEL. fluid pressure, since the maximum stress at the inner radius is T, that on the inner elementary annulus is Tdx, and on any other annulus • — \- dx ; while the total resistance will be, on either side the cylinder, K^^ r^ + Cri-rs) r^ + t The maximum stress is at the interior, and may be equal, as taken above, to the tenacity, T, of the metal ; then t t and the thickness ' - r-A while the ratio of the radii n_Tt . i{T-p,)_ T T-p, Lamfs Formula, which is more generally accepted, and .which is adopted by Rankine, gives smaller and more exact values than that of Barlow. In the above, no allowance is made for the compressive action of the internal expanding force upon the metal of the ring. The effect of the latter action is to make the intensity of pressure at any ring less than before by a constant quantity, / oc ^ - ^, and the tension by which the ring resists that pressure greater. STRENGTH OF IRON AND STEEL. 387 When ?• = ri, / = o ; when r = r^, p = p^; then p\=—R—b, and o = — 5 — ^, "1 '2 M — '^a and the maximum possible stress on the inner ring is n' ^Alri — 72 + ;:^ 7- = /, ^1 — ''2 ''i — '>'%' r^ + ^2' and the ratio of inner and outer radii is i/l±A Of these two formulas, the first gives the larger and conse- quently safer results, and, in the absence of certain knowledge of the distribution of pressure within the walls of the cylin- der, is perhaps best. For thick spheres. Lamp's formula becomes 388 MATERIALS OF CONSTRUCTION— IRON AND STEEL. Clark's formula* is more recent than the preceding. It is assumed that the expansion of concentric rings into which the cylinder may be conceived to be divided is inversely as their radii, and that the curve of stress will become parabolic if so laid down that the radii shall be taken as abscissas and the stresses as ordinates, the total resistance thus varying as the logarithm of the ratio of the radii. Then if the elastic limit be coincident with the ultimate strength, and 7^= the tenacity of the metal ; R — the ratio, external diameter divided by internal; / = the bursting pressure ; p — T y. hyp log R f R= er In other cases, instead of T take the value of the resist- ance at the elastic limit, and base the calculation of propor- tions upon the elastic limit and its appropriate factor of safety. . The formulas as given are considered applicable to cast iron. The strength of thick cast cylinders with heads cast in, may, however, sometimes be far in excess even of the calculated re- sistance of thin cylinders. This is shown by data obtained by test of cast iron (gun metal) cylinders made at the Watertown Arsenal, by Colonel T. T. S. Laidley, U. S. army. These cylinders were eight in number, 1 1 inches (27.9 centimetres) in diameter, and 22^ inches (57 centimetres) long, bored out and lined with a thin copper or bronze tube or an iron cjjlin- der and turned on the outside ; they were, in fact, small lined guns. All proved to be stronger than calculated as above. * Rules and Tables, p. 687. STRENGTH OF IRON AND STEEL. 389 Cast Iron has a power of resisting compression, which, as with other metals, may be taken within the elastic limit, or within the range of distortion and stress usu'al in application, as following the same law as resistance to extension. Its absolute value increases with the proportion of carbon, phos- phorus, manganese, and silicon, in combination up to some undetermined limit, and decreases as the proportion is in- creased of graphitic carbon, of silicon, and other weakening substances. Sound castings will have maximum resistance to compression at a density not far from, though a little above, that which gives maximum tenacity. In general, specifications for cast iron under pressure should be similar in form to those framed for the same iron in tension. The iron should usually be No. 3 iron for ordinary work, and should have a density of 7.26 or 7.28. The following figures are from the mean of a large num- ber of tests of iron intended for ordnance : RESISTANCE TO COMPRESSION — CAST IRON. SPECIFIC GRAVITY. RESISTANCE. C TENACITY. NUMBER. Lbs. on sq. in. Kilogs. on sq. cm. Lbs. on sq. in. Kilogs. on sq. cm. r 2 3 4 5 7.087 7.182 7.246 7.270 7-340 99.770 139,834 158,018 159,930 167,037 7,014 9,830 11,118 11.253 11.743 20,877 30,670 35,633 39.508 32,458 1,468 2,171 2,505 2,777 2,282 The tenacities are presented for comparison, and the table so completed will enable all to be compared with the chemical composition and density of irons of similar tenacity as already given. Tests of cast iron of similar grade to those reported pages 390 MATERIALS OF CONSTRUCTION— IRON AND STEEL. 37S> 378, as made by the Author in tension, gave the following results : COMPRESSION AND DUCTILITY OF CAST IRON. RESISTANCE. C TOTAL COMPRESSION. Lbs. on sq. in. Kilogs. on sq. cm. Per cent. No. 2 iron 81,488 89,127 91.674 127,323 127,323 5,699 6,265 6,445 8,951 8,951 9.48 8.72 5.86 9-95 9.50 it 4( (( It No. 4 " The formula proposed by Hodgkinson for the rather weak cast iron used in his experiments is the following : P = 170,763^ - 36,318^) P„ = 12,004^- 2,553^") in which P and P^ are the loads in British and metric measures respectively, and e the corresponding elongation up to a limit which is rarely as high as one per cent., and is usu- ally not far from the point of rupture. Strain Diagrams of Cast Iron in Compression. — The accompanying figure contains the strain diagrams illus- trating the experiments of the table. The elastic limit may be taken at one-half the ultimate resistance, although it cannot be definitely determined, since, as is best shown by the strain diagrams, the change in rate of distortion is too gradual to permit its identification. It is evident that the strain dia- grams of iron and steel under compression have equations similar to that proposed for those of metal under tension. The stronger and stiffer sample is No. 4, and the weaker and more ductile is No. 2 iron. The small circles are the obser- vations of extension and load ; the crosses indicate corre- sponding sets. STRENGTH OF IRON AND STEEL. 391 C/gtf/fiAvni^ %%Uiyi&, '^% Fig. 62.— Cast Iron in Compression. Long Bars in Compression. — The following are the results obtained by Hodgkinson, testing cast-iron bars 10 feet (3.04 metres) long and of I inch (2.54 centimetres) area of section : 392 MATERIALS OF CONSTRUCTION— IRON AND STEEL. RESISTANCE TO COMPRESSION, AND ELASTICITY OF CAST-IRON BARS. LOAD. C COMPRESSION. MOD. ELASTICITY. Lbs. per sq. in. Kgs. per sq. cm. Total, e Permanent. Lbs. per sq. in. Kgs. per sq. m. 2,064.74 6,194.24 10,323.73 14,453.22 18,582.71 24,776.95 33,030.8 145. 1 439-3 725-5 1,015.5 1,305-9 1,741-3 2,326.7 .0001561 .0004981 .00082866 .00128025 .00154218 .00208016 .0029450 .00000391 .00003331 .00007053 .00011700 .00017085 .00036810 .00050768 13,231,300 12,442,300 12,467,100 12,253,700 12,058,100 11,920,000 11,222,750 9,293 X 10' 8,744 X 10' 8,761 X 10' 8,612 X 10° 8,474 X 10' 8,377 X 10' 7,887 X 10' Wrought Iron in Compression. — The British " Steel Committee " tested iron and steel by compression in 1868-70, and found the elastic resistance of English wrought iron to lie between 10 and 14 tons per square inch, averaging about 12, or nearly 26,000 pounds per square inch (1,827.8 kilo- grammes on the square centimetre), and an extension, to the elastic limit, of 0.097 per cent, (o.ooi nearly). Experimenting on steels, this committee found the elastic limit, in compression, at 50,000 pounds per square inch, nearly (3,515 kilogrammes per square centimetre), and* at a percent- age of compression ranging from 0.000065 to 0.000080 per ton pei square inch, and a total of usually not far from 0.0018, without regard to kind of steel. Kirkaldy, experimenting on the softer and purer iron of Sweden, obtained an average of about 25,000 pounds (1,757.5 kilogrammes on the square centimetre), and an ultimate re- sistance of nearly 175,000 pounds per square inch, with a i- inch cube (12,300 kilogrammes on the square centimetre), and about one half that amount on a i^-inch (3.82 centimetres) bar 2 diameters long. Ten diameters' length reduced the fig- ures to about 1 5 per cent, of the maximum. The compres- sion was nearly 50 per cent. Tangye found the resistance of small areas of larger STRENGTH OF IRON AND STEEL. 393 masses under compression to be sensibly overcome at about 50,000 pounds per square inch, and a deep indentation to be produced by double that load (3,515 and 7,030 kilogrammes per square centimetre). Flues and Cylinders subjected to external pressure resist that pressure in proportion to their stiffness and their compressive strength if thin, and if thick sustain a pressure proportional to their thickness and maximum resistance to crushing. Fairbairn,* experimenting on flues of thin iron, 0.04 inch (0.102 centimetre), of small diameter, 4 inches (10.2 centi- metres) to 12 (31 centimetres), and from 20 inches (50.8 centi- metres) to 5 feet (1.52 metres) long, found that their resistance to collapse varied inversely as the product of their lengths and their diameters, and directly as the 2.19 power of their thickness. The following equation fairly expressed his results when/ is the external pressure in pounds per square inch, t their thickness in inches, and d their diameter. L is the length in feet : "9 / pdL ^ o , ^-^ ^ ; / = 806,000 -TF. 2.19 / 806,000'-^ ' dL or, for the length in inches, ^.19 / = 9,672,000-^ In metric measures, kilogrammes and centimetres diameter, and metres of length, ^.19 / = 68,000 -Tj , nearly * Useful Information. Second Series. 394 MATERIALS OF CONSTRVCTION-IRON AND STEEL. For elliptical flues take d = —; where a is the greater and b the lesser semi-axis. These equations probably give too small values of / for heavy flues under high pressure. Belpaire's rule, deduced from Fairbairn's experiments, is, ^081 Lloyd's rule for flues is, for working pressures, in which a is made 89,600 pounds per square inch. The British Board of Trade rule is, for cylindrical furnaces with butted joints. P (L + \)d in which a is 90,000, provided, always, p < 8,000-,. a For large joints a = 70,000 unless beveled to a true circle, when a = 80,000. If the work is not of the best quality, these values of a are reduced to 80,000, 60,000, and 70,000. The factor of safety in boiler work should not be allowed to fall below 6. " Corrugated " flues are claimed to have double the strength and much greater elasticity than " plain " flues. Resistance of Columns, Posts, or Struts. — ^The resistance of parts of structures under compression is often determined largely by their form and by the method of put- ting them together or of building them up. In construction STRENGTH OF IRON AND STEEL. 395 such parts are called pillars, posts, or struts, and are given all the various forms shown in the acconapanying figures : Fig. 63. — Cross Sections of Columns. Their ends are usually fitted with bases or " shoes " of cast or forged iron, having, in accepted practice, a minimum tliickness of I2,OOOd! where P is the total load in pounds, and d the diameter in inches of the pin sustaining the strut, as is common in Ameri- can bridge construction. The first of the forms here shown is known as the " Phoenix " column. In all such pieces the resistance to compression is less than the figures already given for short pieces yielding by actual crushing. Flexure of Columns. — It is shown, in works on the theory of the strength of materials, that the general equation for flexure of any piece subjected only to stress producing bending is, when /is the principal moment of inertia,* the second member being negative when, as in the bending of very long columns, the moment of the flexing force is negative with respect to the moment of the resisting forces, y being the ordinate of any point in the curved axis, and x the abscissa, as the curve of the beam is concave to the axis of X. From this expression is derived, by Euler and later * Discovered, and proven, by Prof. Robinson to be the principal moment for all cases. Vide Strength of Wrought Iron Bridge Members ; Van NostraruCs Science Series, No. 60, equation (5). 39^ MATERIALS OF CONSTRUCTION— IRON AND STEEL. authors, an equation for the load on a column, when boti*! ends are rounded or pinned, thus : The integral of equation 74 is. X y '= a sm -. ■ 1/^ but to make ^ = o at the extremities of the column, when x — /, we must have / P or equal some multiple of n ; thence we may put, /f. L- . /MI and, therefore, P= -j-^EI= 10 -^, nearly in which / is the length, E the modulus of elasticity, P the load, and / the moment of inertia of the transverse section. Strength of Columns of Great Length. — Since this resistance is independent of the extent of flexure, it is evi- dent that, passing the limit of elasticity, where the law of variation of resistance changes, as will be seen by studyin'g strain diagrams given later, the formula gives the breaking load, when, as in the case here taken, there is no external force aiding the column in the effort to retain its form. This expression is proposed by Navier* for columns 20 diameters or more in length. Later writers would restrict it to still more slender columns — 30 to 40 diameters. * Resum/des Ltfons : Paris, 183S, p. 204. STRENGTH OF IRON AND STEEL. 397 When the column is cylindrical, the equation becomes, and for square columns. This equation may be used for flat-ended columns 60 or 70 diameters long by multiplying the second member by 4, and to columns having one end flat, the other rounded, when 40 or 50 diameters long, by making the factor 2 ; * making the general equation, D EI 1 P= ^o-j^, nearly and „ EI , L ■ — ^w p ' "^'"V While the equations 1 become for cylinders, P = -? P = 4: ' and for square pillars. 3iErp P = P = If ^7? Hodgkinson's simple formula for the same column is given : For long columns, fixed, solid, ^8.55 Cast iron, P = 49.4 -yj^ • See Strength of Bridge Members, Robinson : New York, D. Van Nostrand, 1882, p. 107. 398 MATERIALS OF CONSTRUCTION— IRON AND STEEL. Wrought iron, P = 149.7 -jpf For hollow columns, Last iron, P = 49.6 — — j^ ■ the diameter being taken in inches, the length in feet, and the load in tons. Standard Formulas for Strength of Columns. — In all ordinary cases of yielding of columns, and in all cases of short columns, even with rounded ends, the lateral resistances must be considered. For such cases, engineers are accus- tomed to use what is generally known as Gordon's formula — more properly called Tredgold's* — or a modification of wider application proposed by Rankine.f Tredgold's formula is the following : p^ Cbd I + ^Q" for rectangular columns. The values of a and C for rupture, are given later. Working loads are usually restricted to C= 7,500, or C= 8,500, for ordinary and Phcenix sections respectively, and a — jj^y^ for pin connections to a = go^^^,,, for flat ends. This formula applies to pillars with rounded ends. Gordon obtained constants for this formula from various sources, and it has become more generally known by his name. The following are the constants obtained from hollow columns for the modified formula, for fixed ends : I + ' (-;)■ * Tredgold on Strength of Cast Iron, second edition, p, 183, \ Applied Mechanics, p. 305. STRENGTH OF IRON AND STEEL. 399 SECTION. C. MATERIAL. Lbs. per sq. in. Kilogs. per sq. cm. a. Circle. Square. Circle. Square. 80,000 80,000 36,000 36,000 5.624 5.624 2,628 2,628 0.002 Wrought iron 0.00033 C is the maximum resistance to crushing. For rounded ends, or pin-connections, multiply a by 4, and for one end fixed, by 2. In Gordon's formula, the load is in pounds the area, K, in square inches, and the length and diameter in the same units. All the values of C are lower than it is now customary to take them. Rankine's formula is of more general application than Tredgold's, although derived by a similar process. It has the following form for a strut fixed at both ends :* CK P = I + '^(i in which P is the load, C the resistance to crushing in short pieces, both in the same terms, K the sectional area, / and k the length of the column, and the least radius of gyration of its cross section in the same units. For rounded ends, a is multiplied by 4, and for one end fixed, by J/-. The follow- ing are values of C and a as given by Rankine : c. Lbs. on sq. in. Kilogs. on sq. cm. a. 80,000 36,000 5.624 2,628 * Rules and Tallies, p. 210. 400 MATERIALS OF CONSTRUCTION— IRON AND STEEL. The formula of Tredgold and its modifications may be thus derived : If the load on the head of a column be P, the intensity of the stress due that load, at any section, K, is • f~-i and this is, as a maximum in short pieces and masses, equal to the resistance, C, to. crushing given in the preceding tables. But when a long pillar or column yields, it does so by bending transversely, and follows the law already given in connection with the deduction of Euler's formula. This brings a stress, p" , due to bending solely, upon parts already strained by the stress, /', producing a maximum, p' + p" = C The value of p'' varies directly as the moment and in- versely as the breadth and the square of the thickness of a rectangular section,* or as the cube of the diameter for a cir- cular section of column, and if M is the bending moment of the load for a square section, C = p +P -^^"bd^ and since the maximum allowable deflection is proportional to the square of the length divided by the thickness, r P ^ P ^' =|[- -m * Rankine ; Applied Mechanics, p. 305. STRENGTH OF IRON AND STEEL. 4OI and the crushing load is while the maximum intensity of pressure will be P the values of which are always less than C, and decrease as the column is lengthened, finally becoming identical with that obtained with Euler's method. It is evident that the same formula, with suitable alter- ation of the constant, a, may be written, as by Rankine. The following are values of ^ for solid sections and for hollow sections with thin sides : Values of Radii of Gyration. Form of Section. k''. Solid ; rectangle -ti K'. Thin ; square \ t^. Thin ; rectangle — . ' ^ Solid ; cylinder iV '^^• Thin ; cylinder \h^. Angle iron ; equal flanges, of width b v'i ^'. Angle iron ; unequal flanges, of widths b and h. JJT^TTjI"^ • Cross of equal arms ii'l -4'. H-iron ; breadth of flanges, b ; area, A ; area of web, B — . —. = . 12 A+ B Channel iron ; depth flange + \ thickness of web , ^ ^^ . = h ; area web = B \ area flanges ■=A...h'' ^_^_-_^ + -——,y^ The value of C is 36,000 ; a = ^^f^ for wrought iron : C = 80,000 ; a = ^jVt ^°^ ^^^^ ^""O"^ • ^ — ^^^^ dimension. For octagonal and other sections approaching either of the above figures, the nearest regular figures may be takan. 26 402 MATERIALS OF CONSTRUCTION— IRON AND STEEL. Columns should always yield by alteration of form, and not by local injury. The investigations of Hodgkinson, which form the basis of the engineer's work in this direction, indicated that, in practice, the strength of long columns with fixed ends is three times as great as those with joints or rounded ends ; that the column or strut having one fixed and one rounded or loose end, is intermediate, the three cases having the relation i, 2, 3. When having flat ends, they yield at three points — in the middle and near each end — when rounded or loose, in the middle only. The increase of the diameter at the middle gives greater strength to solid pillars, but has little effect on hollow columns ; the gain, in the first case, is 10 or 12 per cent. The load carried on columns of similar form varies as the cross section. Cast-iron Columns are used in many structures, and, if sound and of good material, are reliable. They are economi- cal in cost of manufacture and of fitting, and are more dura- ble when exposed to the weather than are columns of wrought iron. They are less safe where exposed to shock, and are, for that reason, seldom used in bridges or in struct- ures liable to injury by that cause. Cast-iron pillars are more liable to defects of form of structure and of material than those of wrought iron ; they are also more subject to in- jury by shock ; they should always be designed with a higher factor of safety than wrought-iron pillars. The engineer has less confidence in cast iron, also, because of the difficulty of testing and of inspecting it satisfactorily. Cast-iron columns should not be given a thickness less than about 0.004/, nor in any case less than ^ inch (1.6 centimetres). Some engineers make this limit 0.\d. Slight inequalities of thickness do not usually impair their strength. The flanges of columns should be turned and fitted to the base, which should itself be smoothly faced to receive the column. Where the ends can be spread to form capital and base, the structure is greatly stiffened. For short pillars of large diameter, cast iron is stronger than wrought. STRENGTH OF IRON AND STEEL. 403 A moderately hard, strong, close-grained iron is best for columns, as well as for beams or other structures in which stiffness is essential. Sections Other than Rectangular are most com- mon in iron and steel beams and girders. For the general case we have, as can be shown, for moderate defiections-. where M is the moment of resistance to bending offered by the beam ; / is the Geometrical Moment of Inertia of the strained section, and_yi is the distance of the neutral axis of the beam from the adjacent surface, when either tension or compression acts alone to produce M. When the neutral axis is at the middle of the section, and the resistances are equal above and below, the total moment of resistance becomes, a d being the depth of the strained section. Prof. C. A. Smith gives a simple, handy rule for the moment of resistance of sections of " tee " and " angle " irons exposed to flexure, thus : * One-fourth the product of breadth, depth, and thickness of flange, in inches, is the moment of resistance in foot-tons ; i. e., bdt „ , — = ^nearly. 4 In metric measures, the divisor becomes 200 to give the moment in metre-tonnes. The quantity thus obtained being taken as the working load, the maximum stress is about 10,000 pounds per square inch (703 kilogrammes per square centimetre). The values of R given in the tables are not exact for * Railroad Gazette, Nov. 13, 1875. 404 MATERIALS OF CONi^TRUCTlOJSr—xKON AND STEEL. beams and girders of other than rectangular section, or for cases in which the neutral axis shifts its position under the load. If the value of R is taken as equal to the smaller of the two values T and C, any error will be on the safe side ; or the factor of safety may be somewhat increased to allow for an overestimate. The forms of section adopted will be seen in Article 270, on the working formulas for beams. It is evident that, in general, extending the extreme portions of the section where stresses become greatest, and restricting the intermediate part, or the " web," to the size needed to hold the other portions in proper relative positions, will produce forms of beam of greater strength, with a given weight of material, than can be obtained in the cases of rectangular, circular, or other simple forms of section. Where the metal has equal strength to resist tension and compression, it is further evident that the top and bottom " flanges " should be of equal size ; this constitutes the Tred- gold " I-beam " usually made in wrought iron. When the metal is stronger in compression than in tension, as is the case with cast iron, the extended side should be enlarged ; this was done by James Watt when making his " j_-beam," and by Fairbairn and Hodgkinson, who first made the " j_- beam," in which the compressed flange has an area less than that under tension in the same proportion that the resistance to compression exceeds the resistance to tension. For ordi- nary cast iron these areas are as six to one. In many cases the form of section is determined by con- venience in making or in building up. Beams and columns are often constructed of L, or " angle " iron, with plate iron, or with u , or " channel " iron, built up in various ways to form I-beams, i -beams, or various sections approaching hex- agonal or circular. For all such cases the moment of inertia can be deter- mined and inserted in the general formulas. The transverse strength of " round iron " and steel of circular section may be taken as six-tenths the strength of bars of square section circumscribing the circle. STRENGTH OF IRON AND STEEL. 405 According to Grashof a circular plate will bear a pressure if bolted along the edge, when T is the tenacity, t the thickness, and r the radius, similar units being used throughout. Shearing is produced by sets of opposed forces act- ing in the same or parallel planes, as where a punch is used or where metal is " sheared." The shearing resistance of iron is usually taken as equal to its resistance to tension, and varies with form and dimen- sions from 45,000 to 60,000 pounds per square inch (3,164 to 4,218 kilogrammes per square centimetre). The shearing resistance of steel varies from that of good wrought iron to double that value or more, according to its composition. Steel is usually, however, less capable of resisting " unfair " strains than is iron, and a good value of this form of resist- ance may be taken as 5 = 60,000 + 40,000 C \ Sm = 4,218 + 2,812 C) where C is the percentage of carbon. The shearing strength of cast iron varies irregularly from 15,000 to 40,000 pounds per square inch (1,055 to 2,812 kilo- grammes per square centimetre) of sheared section, and is most safely taken at the lower iigure. Its value is usually not far from that of the tensile resistance to which it may be taken as equal. The resistance of boiler plates to punching, of riveted wrought iron-work and of iron bridge pins to shearing has been found variable with ordinary materials be- tween the limits, usually, of 50,000 and 55,000 pounds per square inch (3,515 to 3,866 kilogrammes), and may be taken in estimates and specifications at the lower amount. A very extensive set of experiments upon the strength of bolts and nuts, conducted by the Author, gave . figures lower than the above by 20 per cent, or more. 406 MATERIALS OF CONSTRUCTION— IRON AND STEEL. In consequence of the liability, which is always to be ap- prehended, that the shearing will not take place in such a manner aS to permit the piece sheared to offer its maximum resistance, it is usual to assume a loss of from one-fourth to one-fifth, and to take S = \T, ot S = ^T. Taking the latter proportion, the ordinary working value of 5 becomes, for iron 40,000 to 45,000 pounds, and S = 48,000 + 32,000 C ) S'm = 3,374+ 2,250 CJ for steel, which value may be used in all ordinary construc- tions built of known grades of good metal. For other cases not settled by experiment, the engineer assumes the maxi- mum shearing resistance as nearly equal to the tenacity of the metal. Coupling bolts, in shaft couplings, are exposed to this action. They may be proportioned either by making this stress, as above, a safe minimum, or by direct calculation from the size of shaft, as is done by Rankine, who makes their diam- eter, d: in which d is the diameter of the bolt, d' that of the shaft, n the number of bolts, and r the radius of the circle passing through their centres. Riveted work is subject to injury by the tearing out of the rivets through the sheet, when the shearing resistance of the latter is too low, by pulling off the heads when the stress is in line with the axis of the rivet, and by the shearing of the rivet when of too small area of section. The joint has maximum value when no more likely to yield in one of these ways than in another. Loosely fitted rivets and pins have from I to ^ the shearing resistance of tightly fitted rivets ; which latter have practically the full strength due the section sheared. The diameter of the rivet should be about twice STRENGTH OF IRON AND STEEL. 407 the thickness of the plate, but the size is often determined by practical cpnsiderations. A common range of sizes is the following, although no fixed rule is settled upon : SIZE OP RIVETS. Thickness of plate, inches \s \ -h \ -h \ f f " " " centimetres 0.48 0.64 0.80 0.96 1.12 1.27 1.60 1.92 Diameter of rivet, inches \ \ \ \ I ll^li centimetres 1.27 1.60 I.g2 2.08 2.08 2.54 3.17 3.81 it it It The distance between centres, the pitch of the rivets, should be, in iron, / = 07854 y-f-^ where d is the diameter of the r'yet, t the thickness of the sheet. In steel, the same rule applies when riveted with rivets of the same quality with the sheet ; otherwise, we must have, when 5 is the shearing resistance per unit of area of the rivet-section, and S that of the Sheet, and s'^ = S'{j>-d)t, 4 = 0-7854^^7^ + d. When the sheet is of rather hard steel and the rivet of iron, the sheet is liable to cut the rivet, and the value of S should therefore be taken low. The length of the rivet is usually about 1= 2t + 2\d 468 MATERIALS OF CONSTRUCTION— IRON AND STEEL. exceeding the length of the rivet hole by 2\ times its diame- ter. For double riveting and joints held by several rows, n, of rivets, Snd^ p = 0.7854-^ + d. The lap of the joint should be sufficient to allow ample margin for chipping or planing and caulking, as well as safe against the tearing out of the rivet. Fairbairn gives the following table as exhibiting the pro- portions by him determined experimentally : PROPORTION OF RIVETS. THICKNESS OF DIAMETER OF LENGTH OF LAP or SINGLE PITCH OF PLATE. RIVETS. RIVETS. RIVETING. RIVETS. In. Cm. In. Cm. In. Cm. In. Cm. In. Cm. 1^ 0.48 5 0.95 i 2.22 :} 3-i8 li 3-18 i 0.64 i 1.27 !» 2.86 3-81 I,^ 3.81 ^ 0.79 i 1-59 I« 3-49 ij 4.76 Ji 4-13 i 0-95 i 1. 91 I^ 4-13 2 5.08 li 4 45 1.27 I 2.54 2i 5-72- 24- 5-72 2 5. OS 1 1-59 li 3-17 2| 6.99 2} 6.99 2i 6.3s 1. 91 xi 3.81 8.26 3i 8.26 3 7.62 Engineer-in-Chief W. H. Shock, U. S. N.,* finds bolts or rivets in double sRear to exceed in resistance those in single shear by the following amount : ^ inch (1.27 cm.) diameter 86.2 per cent. ^ inch (l, 59 cm.) diameter 97-0 per cent. f inch (1.9 cm.) diameter loi.i per cent. J inch (2.22 cm.) diameter 82.6 per cent. I inch (2,54 cm.) diameter 85.0 per cent. • The Torsion of Shafts may be reckoned as a case of shearing. Tlje following are safe formulas : * Treatise on Steam Boilers, W. H. Shock. N. Y. : D. Van Nostrand, 1881, STRENGTH OF IRON AND STEEL. 409 For head shafts well supported against springing: ' p_d^_dJR_ 3/125^ _ 3/2000HP 125 ~20oo' y ~E~' ^"'-y -^ • For line shafting; hangers 8 feet (2^ metres) apart : For transmission simply ; no pulleys : ^ - 90 - 1450 ' ^ - i/ ~^r ' "^"^-y—ir- P'- d^R_dJR_ ,_ 3/62.5/27' , _ &/1000HP 62.5 " 1000' '^~y ~R~ '"^"-y- R For cold-rolled iron, these formulas become : TTp_d^R_dJR »/ 7SHP _ S/ 1200HP ^^~"75~~i2oo' '^~y^R~'' '^'"~y R • Tjp-d'R-d^R_ //S5^^. ^ _ . 3/ 880//P HP = '^ = ^^^- -'- ■VSS^i'. ^ _ .V^ioHP ■■^=f-^--^~=f- 35 S50 ' |/ ie ' "* y ie ■ Here /f/" = horse-power transmitted ; d = diameter of shaft in inches ; dm, in centimetres ; R = revolutions per minute. Francis gives the following as permissible distances be tween bearings for shaftings carrying no side strain : 4IO MA TERIALS OF CONSTRUCTION— IRON AND STEEL. SPANS FOR SHAFTING. DISTANCE BETWEEN BEARINGS. Wrought Iron. Steel. Inches. Centimetres. Feet. Metres. Feet. Metres. 2 3 4 5 6 7 8 9 5.08 7.62 10.16 12.70 15-24 17.78 20.32 22.86 15-5 17-7 195 20.9 22.3 23-5 24.6 25-5 4 5 6 6 6 7 7 7 7 4 4 8 I 5 8 15.9 18.2 20.0 21.6 22. g 24.1 25.2 26.2 4 5 6 6 5 7 7 8 8 5 I 5 9 3 7 These distances may usually be safely obtained from the formulas : D= 12 ^/d; Z>„ = 3V'^ where D, I>„ are distances in British and metric measures, between bearings, d, d„ are the diameters of the shafts. In designing steam-engine shafts and other similar pieces the diameter is sometimes expressed in other terms. Thus the Author has used, in designing, the formula, for a single shaft, for long-stroked engines. in which d is the minimum diameter of the shaft in inches, d' that of the steam cylinder in inches, and L the stroke of piston in feet, and/ the steam pressure in pounds per square inch. In metric measures, kilogrammes, and centimetres, i , nearly STRENGTH OF IRON AND STEEL. 411 For screw engines the diameter is usually froni one-eighth to one-sixth greater. Shafts are generally made of wrought iron ; cold-rolled shafting is common ; steel shafts and shafting are coming into use, both of common and Whitworth steel. Cast iron is rarely used to resist this kind of stress. The Metals, and their Strain Diagrams.* — Fig. 64 exhibits a series of curves which illustrate well the gen- eral characteristics and the peculiarities of representative specimens of the principal varieties of useful metals. / In some cases two specimens have been chosen for illustration, of which one presents the average quality, while the other is the best and most characteristic of its class. Wrought iron, as usually made, has a somewhat fibrous structure, which is produced by particles of cinder, originally left in the mass by the imperfect work of the puddler while forming the ball of sponge in his furnace, and which, not having been removed by the squeezers or by hammering the puddle ball, are, by the subsequent process of rolling, drawn out into long lines of non-cohering matter, and produce an effect upon the mass of metal which makes its behavior under stress somewhat similar to that of the stronger and more thready kinds of wood. In the low steels, also, in which, in consequence of the deficiency of manganese ac- companying, almost of necessity, their low proportion of carbon, this fibrous structure is produced by cells and " bub- ble holes " in the ingot refusing to weld up in working, and drawing out into long microscopic, or less than microscopic, capillary openings. In consequence of this structure we find a depression in- terrupting the regularity of their curves, immediately after passing the limit of elasticity, precisely as the same indica- tion of the lack of homogeneousness of structure was seen in the diagrams produced by locust and hickory.f The presence of internal strain constitutes an essential peculiarity of the metals which distinguishes them from or- * From a paper by the Author ; Trans. Am. Soc. C. E., 1874. f M. of E. 412 MATERIALS OF CONSTRUCTION— IRON AND STEEL. ganic materials. The latter are built up by the action of molecular forces, and their particles assume naturally, and probably invariably, positions of equilibrium as to strain. The same is true of naturally formed inorganic substances. The metals, however, are given form by external and arti- ficially produced forces. Their molecules are compelled to assume certain relative positions, and those positions may be those of equilibrium, or they may be such as to strain the cohesive forces to the very limit of their reach. It even fre- quently happens, in large masses, that these internal strains actually result in rupture of portions of the material at vari- ous points, while in other places the particles are either strongly compressed, or are on the verge of complete separa- tion by tension. This peculiar condition must evidently be df serious importance where the metal is brittle, as is illus- trated by the behavior of cast iron, and particularly in ord- nance. Even in ductile metals it must evidently produce a reduction in the power of the material to resist external forces. Since straining the piece to the limit of elasticity brings all particles subject to this internal strain into a similar con- dition, as to strain, with adjacent particles, it is evident that indications of the existence of internal strain, and, through such indications, a knowledge of the value of the specimen, as affected by this condition, must be sought in the diagram before the sharp change of direction which usually marks the position of the lin^it of elasticity is reached. As already seen, the initial portion of the diagram, when the material is free from internal strain, is a straight line up to the limit of elas-: ticity. A careful observation of the tests of materials of various qualities, while under test, has shown that, as would, from considerations to be stated more fully hereafter in treating of the theory of rupture, be expected, this line, with strained materials, becomes convex toward the base line, and the form of the curve, as will be shown, is parabolic. The initial portion of the diagram, therefore, determines readily whether the material tested has been subjected to in- ternal strain, or whether ' it is homogeneous as to strain. STRENGTH OF IRON AND STEEL. 413 This is exhibited by the direction of this part of the line as well as by its form. The existence of internal strain causes a loss of stiffness, which is shown by the deviation of this part of the line from the vertical to a degree which becomes observable by comparing its inclination with that of the line of elastic resistance obtained by relaxing the distorting force — i.e., the difference in inclination of the initial line of the diagram and the lines of elastic resistance, e, e, e, indicates the amount of existing internal strains. Strain Diagram of Forged Iron.— In Fig. 64 the curves numbered 6, i, 23 and 100, are the diagrams produced by three characteristic grades of wrought iron. The first is a quality of English iron, well known in our market as a su- perior metal. The second is one of the finest known brands of American iron, and the third is also of American make, but it does not usually come into the market in competition with well-known irons, in consequence of the high price which is consequent upon the necessary employment of an unusual amount of labor in securing its extraordinarily high char- acter. No. 6 at first yields rapidly under moderate^ force, only about 50 foot-pounds of torsional moment being required to twist it 5°. It then rapidly becomes more rigid, as the in- ternal strains, so plainly indicated, are lost in this change of form, and at 6° of torsion the resistance becomes 60 foot- pounds, as measured at a. Here the elastic limit is reached. The next 3° produce no increase of resistance. This fact shows that this iron, which was not homogeneous as to strain, was also not homogeneous in structure. We conclude that it must be badly worked- and seamy, and that it may have been rolled too cold ; the former is the probable reason of its lack of homogeneous structure ; the latter gave it its condition of internal strain. After the first 9° of torsion, resistance steadily rises to a maximum, which is reached only when just on the point of rupture, and the piece finally commences breaking at 250°, and is entirely broken off at 285°. Its maximum elongation, whose value is proportionable to the reduction of section noted with the standard testing machines, is 0.691. 4H MATERIALS OF CONSTRUCTION— IRON AND STEEL. The terminal portion of the line, after rupture commences, is not usually accurate as a measure of the relation of the force to the distortion. The increase of resistance between the angle 9° and the angle of rupture is produced by the addi- tional effort in resistance due to the " flow," or drawing out of particles, as already indicated. Applying the scale for tension, which in the case of these curves was very exactly 24,000 pounds per square inch for each inch measured vertically on the diagram, we find that the elastic limit was passed under a stress equivalent to a tension of 19,800 pounds per square inch, and that the ulti- mate tenacity was 59,200 pounds per square inch. When nearly at the maximum the specimen was relieved from stress, the pencil descending to the base line, and the elas- ticity of the piece produced a certain amount of recoil. The angle intercepted between the foot of this nearly ver- tical line, c, and the origin at O, measures the set, which is almost entirely permanent. The distance measured from the foot of the perpendicular let fall upon the axis of ab- scissas, from the head of this line to the foot of the line e, measures the elasticity, and is inversely proportional to the modulus. A comparison of the inclination of the line made by the pencil in reascending, on the renewal of the strain with the initial line of the diagram, gives the indica- tion of the amount of internal strain originally existing in the piece. It will be noticed that the horizontal movement of the pencil is recommenced at /, under a higher resistance than was recorded before the elastic line was formed. In this case the piece had been left under strain for some time be- fore the stress was relieved, and the peculiarity noted is an example of an increase of resistance under stress,* or more properly of the elevation of the elastic limit, of which more marked examples will be shown subsequently. The exceptional stiffness and limited elastic range here shown, as compared with the other examples given, is prob- * Vide Transactions Am. Soc. C. E., Vol. II., page 290. Missing Page STRENGTH OF IRON AND STEEL. 415 ably a phenomenon accompanying and due to this increase of resistance under stress. Examining No. i in a similar manner, we find that it is far freer from internal strain than No. 6, its initial line being much more nearly straight and rising more rapidly. It is rather less homogeneous in structure, and is forced through an arc of 6°, after having passed its elastic limit, before it be- gins to offer an increasing resistance. It is evidently a bet- ter iron, but less well worked, and, as shown by the position of the elastic limit, is somewhat harder and stiffen No. i retains its higher resistance quite up to the point at which No. 6 received its incidental accession of resistance by stand- ing under strain, and the two pieces break at, practically,, the same point ; No. i having slightly the greater ductility. When the " elastic line," e, is formed just before fracture, it is seen that No. i has a greater elastic range and a lower modulus than No. 5. The elastic line formed by No. i at between 40° and 45° of torsion is seen to be very nearly paral- lel with that obtained near the terminal portion of the diagram, and illustrates the fact, here first revealed to the eye, that the elasticity of the specimen remains practically unchanged up to the point of incipient rupture; and this fact corroborates the deductions of Wertheim* and others who came to this conclusion from less satisfactory modes of research. No. 22 illustrates the characteristics of a metal which rep- resents one of the best qualities of wrought iron made, and with which every precaution has been taken to secure the greatest possible perfection, both in the raw material and in its manufacture. The line of this diagram, starting from O, rising with hardly perceptible variation from its general di- rection, turns, at the elastic limit, a, under a moment of about 80 foot-pounds, equivalent to a tension of about 24,000 pounds per square inch (1,680 kilogrammes per square cm.); and with between 2° and 3" of torsion only, and thence continues rising in a curve almost as smooth and reg- * Vide Annales de Chimie et de Physique. 4l6 MATERIALS OF CONSTRUCTION— IRON AND STEEL. ular as if it had been constructed by a skilful draughtsman. Reaching a maximum of resistance to torsion of 220 foot- pounds and an equivalent tensile resistance of over 66,000 pounds per square inch (4,620 kilogrammes per square centi- metre) at an angle of 345°, it retains this high resistance up to the point of rupture, some 358° from its starting point. The maximum elongation of its exterior fibres is 1.2, making them at rupture 2.2 times their original length. This would produce a probable breaking section in the common testing machine equal to 0.4545 of the original section. From the beginning to the end this specimen exhibits its superiority, in all respects, over the less carefully made irons, Nos. I and 6, which are themselves good brands. The homogeneousness of No. 22 is almost perfect, both in regard to strain and to structure, the former being indicated by the straightness of the first part of the diagram and its parallel- ism with the " elastic line," e, produced at 217°, and the latter being proven by the accuracy with which the curve follows the parabolic path indicated by theory as that which should be produced by a ductile homogeneous material. At similar angles of torsion. No. 22 offers invariably much higher, resistance than either Nos. i or 6, and this superiority, uniting with its much greater ductility, indicates an immensely greater resilience. It is evident that for many cases, where lightness combined with capacity to carry live loads and to resist heavy shocks are the essential requisites, this iron would be by far preferable, notwithstanding the cost of its manufac- ture, to any of the cheaper grades. Comparing their elastic- ities, as shown at 210°, 215°, it is seen that No. 22 is about equally stiff and elastic with No. i, while both have a wider elastic range and are less rigid, and hence are softer, than No. 6, whose elastic line is seen at 221". All of the character- istics here noted can be accurately gauged by measuring the diagrams. * No. 100 is the curve obtained from a piece of Swedish iron. Its characteristics are so well marked that one familiar with the metal would hardly fail to select this curve from among those of other irons. It softness and its homogeneous STRENGTH OF IRON AND STEEL. 417 structure are its peculiarities. Its curve, at first, coincides perfectly with that of No. 6. It has, however, slightly less of the condition of internal strain, and a somewhat higher limit of elasticity. The elastic limit is found at 51^° of tor- sion, and at a stress of 65 foot-pounds (9.1 kilogrammetres) of moment, equivalent to 19,500 pounds on the square inch (1,365 kilogrammes per square centimetre), in tension. Its increase of resistance, as successive layers are brought to their maximum and begin to flow, is very nearly the same as that of the specimens Nos. i and 6, and the line lies between the diagrams given by these irons up to 30°, and then falls slightly below the latter. At 220° it attains a maximum resisting power, and here the outer surface begins to rupture, after an ultimate stretch of lines formerly parallel to the axis amounting to 0.564. Had this elongation taken place in the direction of strain, as in the usual form of testing ma- chine, it would have produced a reduction of section to 0.64 the original area.* At this point the stress in tension equiv- alent to the 176 foot-pounds (24.64 kilogram metres) of tor- sional stress, is 52,800 pounds per square inch (3,696 kilo- grammes per square centimetre). From 250° the loss of resistance takes place rapidly, but the actual breaking off of the specimen did, not occur until it had been given a com- plete revolution. This part of the diagram distinguishes the metal from all others, and shows distinctly the exceptionally tough, ductile and homogeneous character which gives the Swedish irons their superiority in steel making. No. 22, even, although much more extensible, is harder than No. 100, and yields more suddenly when it finally gives way. Inspection of Fractured Test Pieces. — An exam- ination of the broken test piece gives evidence confirmatory of the record. Thus, examining the broken test pieces from the autographic machine, as shown below, and comparing them, it will be found that the specimens themselves furnish almost as valuable information, after test, as the diagrams give, and they should always be carefully inspected with a Compare Styffe, Strength of Iron and Steel, p. 133, Nos. 26-30. 27 41 8 MATERIALS OF CONSTRUCTION— IRON AND STEEL. view to securing additional or corroborative information. Fig. 65 is a sketch of specimen No. i, and shows its somewhat Fig. 65. Fig. 66. granular fracture, and the seamy structure produced by a de- fective method of working. Fig. 66, from specimen No. 16, more nearly resembles that which gave the diagram marked 6, Fig. 64. The metal is seen to be good, tough, and better in quality than No. i, but it is even more seamy, and even less thoroughly worked, as is evidenced by the cracks extend- ing around the neck, and by the irregularly distributed flaws seen on its end. Fig. 67 exhibits the appearance of No. 22 after fracture, and shows, even more perfectly than the pen- cilled record, the excellent charac- ter of the material. The surface of the neck was -originally smoothly turned and polished, and carefully fitted to gauge. Under test it has ' become curiously altered, and has assumed a rough, striated appear- ance, while the helical markings extend completely around it. The end has the peculiar appearance which will be seen to be characteristic of tough and ductile Fig. 67. STRENGTH OF IRON AND STEELi 419 metals, and the uniformly bright appearance of every par- ticle in the fractured section shows how all held together up to the instant of rupture, and that fracture finally took place by true shearing. Rupture by torsion thus brings to light every defect and reveals every excellence in the specimen. Rupture by tension rarely reveals more than the mere strength of the material. Strain-Diagrams of Low Steels. — In Fig. 64, and above the curves just described, are a set obtained during experiments on " low steels," produced by the Bessemer and Siemens-Martin processes. In general character the curves are seen to resemble those of the standard irons, as illus- trated by Nos. I and 6. The irons contain usually barely a trace of carbon. These steels contain from one-third to five- eighths of one per cent. The irons are made by a process which leaves them more or less injured by the presence of impurities, from which the utmost care can never free them. The steels are made from metal which has been molten and cast, a process which allows a far more complete separation of slag and oxides. The low steels, however, are liable to an objectionable amount of porosity, due to the liberation c5f gas while the molten mass is solidifying, whenever the spie- geleisen, employed as a conveyer of carbon, carries littk manganese. The results of these differences in constitu- tion and treatment are readily seen by inspecting the curves. They show a stiffness equal to No. 6, and about the same degree of internal strain. They contain a sufficient number of the capillary channels produced by drawing down the pores while working the ingot into bar, to cause a lack of homogeneousness in structure very similar to that produced in iron by cinder. They have a much higher elastic limit and greater strength, and the softer grades have great ductil- ity. In resilience, these softest steels excel all other metals, except the unusual example No. 22, and are evidently the best materials that are now obtainable for all uses where a tough, strong, ductile metal is needed to sustain safely heavy shocks. A comparison of the diagrams of two competing jnetals may thus be made to indicate how far a difference if 420 MATERIALS OF CONSTRUCTION— IRON AND STEEL. price should act as a bar to the use of the costlier one. For general purposes, a comparison of the resilience of the metals within the elastic limit is of supreme importance. No. 6 is seen to have more resilience within this limit than No. i, and the steels far more than either ; but No. i would take a set of considerable amount far within the true elastic limit, as indicated at a. The most valuable measure is obtained by- determining the area intercepted between the " elastic line " and the perpendicular let fall from its upper end; this meas- ures the resilience of elastic resistance, which is the really im- portant quality. No. 98 was cut from the head of an English Bessemer rail made from unmixed Cumberland ores. It contains nearly 0.4 per cent, carbon. It is quite homogeneous, has a limit of elasticity at 88 foot-pounds of torsional, or 26,400 pounds per square inch tensile stress, approaches its maximum of resist- ance rapidly, and at 210° the torsional moment becomes 225 foot-pounds, equivalent to 67,500 pounds per square inch ten- sile stress. It only breaks after a torsion of 283°, and with an ultimate elongation of 80 per cent., equivalent to a reduc- tion of cross section to 0.556. No. 76 is a Siemens-Martin steel made from mixed Lake Superior and Iron Mountain ores, and contained about the same amount of carbon as the preceding. It contains rather more phosphorus, which probably gives it its somewhat greater hardness, its higher limit of elasticity, and its some- what reduced ductility. Its elastic limit is found at 104 foot- pounds of torsion, or 31,200 pounds tensile resistance, and its ultimate strength is almost precisely that of the preceding specimen. Its elongation is 0.66 maximum. Unless more seriously affected by extreme cold than No. 98, it would be preferred for rails, and, perhaps, for most purposes. No. 67 is a somewhat "higher" steel, made by the^same process. It is less homogeneous than the two just examined, has greater strength and a higher elastic limit, but less duc- tility. Its resilience is very nearly the same as that of Nos. 98 and j6. The elasticity of all these steels seems very exactly the same. The ductility of No. 6^ is measured by STRENGTH OF IRON AND STEEL. 42 1 0.40 elongation. At d, is seen another illustration of eleva- tion of the elastic limit. The piece was left twenty-four hours under maximum stress. The torsional force was then removed entirely. On renewing it, as is seen, the resist- ance of the specimen was found increased in a marked de- gree. No. 69 is an American Bessemer steel, containing not far from 0.5 per cent, carbon. The same effect is seen here that was before noted, an increase of hardness, a higher elastic limit, and greater strength, obtained, however, by some sac- rifice of both ductility and resilience. The elastic limit is approached at 1 30 foot-pounds of torsional moment, or 39,000 pounds tensile, and the maximum is 280 foot-pounds of mo- ment and 84,000 pounds tensile resistance at 133°. Its maxi- mum angle of torsion is 150°, its elongation 0.24. No. 85 is a singular illustration of the effects of what is probably a peculiar modification of internal strain. It seems to have no characteristics in common with any other metal examined. Its diagram would seem to show a perfect homo- geneousness as to strain, and a remarkable deficiency of ho- mogeneity in structure. It begins to exhibit the indications of an elastic limit at a, under a torsional moment of 1 10 foot- pounds, or an apparent tensile stress of 33,000 pounds per square inch, and then rises at once, by a beautifully regular curve, to very nearly its maximum at 16°, and 176 foot- pounds. The maximum is finally reached at 130°, and thence the line slowly falls until fracture takes place at 195°. The maximum resistance seems* to be very exactly 60,000 pounds to the square inch. Its maximum elongation for exterior fibres is about 0.23. The resilience, taken at the elastic limit, is far higher than with common iron, and it is seen that this metal, in many respects, may compete with steel. Its elas- ticity was seen to remain constant wherever taken. This speci- men was a piece of " cold-rolled " iron. It is probably really far from homogeneous as to strain, but its artificially pro- * In exceptional cases, of which this is an example, this scale for tension givesr too high values. The tensile strength is usually rather less than above given. 422 MATERIALS OF CONSTRUCTION— IRON AND STEEL. duced strains are symmetrically distributed about its axis, and being rendered perfectly uniform throughout each of the concentric cylinders into which it may be conceived to be di- vided, the effect, so far as this test, or so far as its application as shafting, for example, is concerned, is that of perfect homo- geneousness. The homogeneousness in structure is readily explained by an examination of the pieces after fracture; they are fibrous, and have a grain as thread-like as oak ; their condition is precisely what is shown by the diagram, and the metal itself is as anomalous as its curve. Strain Diagrams of Tool Steels. — ^The " tool steels " differ chemically from the " low steels " in containing a higher percentage of carbon, and usually in being very nearly, though not absolutely, free from all injurious elements. Containing a higher proportion of carbon than the preceding class of met- als, it is comparatively easy to secure homogeneousness by the introduction of manganese, and, by the same means, to eliminate very perfectly the evil effects of any small propor- tion of sulphur that may be present. Their comparatively large admixture of carbon makes them harder and reduces their ductility, and since the reduction of ductility occurs to a greater degree than the increase of strength, the effect is also to reduce their resilience. The working of these metals is more thorough than is that of the less valuable steels or of iron. They are cast in comparatively small ingots, and are frequently drawn down under the hammer, instead of in the rolls, and are thus more completely freed from that form of irregularity in structure noticed so invariably in steels other- wise treated. The effect of increasing the proportion of car- bon is to confer upon iron the property of hardening when heated to a high temperature and suddenly cool, and the invaluable property of " taking a temper." The hardened steels are, however, comparatively brittle, the hardening be- ing secured at the expense of ductility. * Referring to the figure, a set of diagrams will be found, having their origin at 180°, which ^r& facsimiles of those au- tomatically produced during experiments upon various kinds of tool steels. STRENGTH OF IRON AND STEEL. 423 No. 58 is an English metal, known in the market as " Ger- man crucible steel." It is remarkable as having a condition of internal strain which has distorted its diagram to such an extent as to completely hide the usual indication of the elastic limit. A careful inspection shows what may be taken for this point at about 14^° of torsion, when the twist- ing moment was about 120 foot-pounds, and the tensile re- sistance 36,000 pounds per square inch (2,531 kilogrammes per square millimetre). The metal is homogeneous in struct- ure, has an ultimate resistance of 302 foot-pounds of mo- ment, or 90,600 pounds per square inch tensile resistance (6,369 kilogrammes per square millimetre). Its resilience is evidently inferior to that of the softer metals, and also less than the next higher and better grades. This metal con- tains about 0.60 to 0.65 per cent, carbon. Its elongation amounts to 0.045. No. 53 is an English " double shear steel," of evidently very excellent structure, but less strong and less resilient than the preceding. Its exterior fibres are drawn out three per cent. Nos. 41 and 61 are two specimens of one of the best Eng- lish tool steels in our market. The first was tested as cut from the bar, but the second was carefully annealed before the experiment. In this instance annealing has caused a slight loss of resilience as well as a decided loss of strength. In No. 41 the limit of elasticity can hardly be detected, but seems to be at about the same point as in No. 61, at near 130 foot-pounds moment, and 39,000 pounds tension. The ulti- mate strength is nearly 119,000 pounds per square inch. The proportion of carbon is very closely i per cent. Its section would reduce by tension 0.05. No. 70 is an American " spring steel," rather hard, but, as shown by its considerable resilience, of excellent quality, re- sembling remarkably the tool steel No. 41. It differs from the latter apparently by its much higher elastic limit. It is possible that this may have been caused by more rapidly cooling after leaving the rolls in which it was last worked. It is evident that, for exact comparison, all specimens should be 424 MATERIALS OF CONSTRUCTION— IRON AND STEEL. either equally well annealed, or should be tempered in a pre- cisely similar manner and to the same degree. Nos. 71 and 82 are American tool steels containing about 1. 1 5 per cent, of carbon. The former is notable as having an elastic limit at 69,000 pounds and a probable deficiency of manganese, producing the usual indication of heterogeneous structure. Both of these steels lack resilience, and are less well adapted for tools like cold chisels, rock drills, and others which are subjected to blows, than for machine tools. They have a maximum elongation, respectively, of but 0.013 and 0.03. Inspection of Steel Test Pieces. — Interesting and instructive as the study of these curves may be made, the in- formation obtained from them is supplemented, in a most valuable manner, by that obtained by the inspection of the fractured specimens, upon which the peculiar action of a tor- sional strain has produced an effect in revealing the structure and quality of the metal that could be obtained in no other way. Fig. 68 represents the appearance of No. 68, and Fig. 69 that of No. 58, while the peculiarities of the finest tool steels Fig. 68. Fig. 6g. are seen in No, 71 as shown in Fig. 70. The smooth exte- rior of No. 68, which is a companion specimen to that giving diagram No. 69, and its bright and characteristic fracture, STRENGTH OF IRON AND STEEL. 425 resembling that of No. 22 somewhat, together indicate its nature perfectly, the first feature proving its strength and uniformity of structure, and the second showing, even to the inexperienced eye, its toughness. This is a representative specimen of low steels. No. 58 is seen to have retained, even more than No. 68, its original smoothly polished surface. Its fracture is less waxy, and much more irregular and sharply angular. The crack running down the side of the neck shows its relationship to the shear steels, which much oftener exhibit this effect of strain, in consequence of their lamellar character. No. 58 is evidently intern;iediate in its charac- ter between the soft steels, like No. 68, and the tool steels which are rep- resented by No. 71, Fig. 70. In this test-piece the fracture is ragged and splintery, and the separated surfaces have a beautifully fine, even grain, which proves the excellence of the material, The surface, which was turned and polished in bringing the metal to size, remains as perfect as before the specimen was broken. By an inspection of the broken test pieces in this manner, the grade of the steel, and Fig. 70. Fia 71. Fig- 72. such properties also as are not revealed by an examination 426 MATERIALS OF CONSTRUCTION— IROAT AND STEEL. of the diagram of strain, are very exactly ascertained by a novice, and, to the practiced eye, the slightest possible vari- ations are readily distinguishable. Fig. 71 shows the appearance of fracture of malleableized cast iron. Its semblance to wrought iron is very noticeable. Fig. 73. Fig. 74. The lines running like the thread of a screw around the exte- rior of the neck, and the smooth, even fracture in a plane pre- cisely perpendicular to the axis, are the instructive features. Fig. 72, representing No. 33, is a specimen similar in character to No. 37. The comparative lack of ductility, its less regular structure, and its less perfect transformation, are perfectly exhibited. Fig. 73 is an excellent cut of the white iron as cast and without malleableizing. Its surface, where fractured, has the general appearance of broken tool steel. The color and texture of the metal are distinctive, however. It has none of the " steely grain." Fig. 74 represents dark-gray cast iron. Its color, its granular structure and coarse grain, are markedly^ characteristic, and no one can fail to observe in the specimen the general character which is exactly given by the auto- graphic diagrams of the testing machine. Fractured Surfaces of Tension Test pieces. — The appearance of the fracture of good iron broken by tension STRENGTH OF IRON AND STEEL. 427 varies greatly with the rate of fracture. When broken slowly it should resemble that shown in Fig. 86 following ; fract- ured rapidly it may become like that seen in Fig. 87 of the same article. Fig. 75 represents the appearance of a sur- face of slightly cold-short, but otherwise excellent iron, forged in a large mass and broken suddenly. Fig. 75. — Fracture of massive iron. The following is the record, and illustrates well the loss of tenacity due to forging in large sizes : 428 MATEKIALS OF CONSTRUCTION— IRON AND STEEL. TESTS MADE AT WATERTOWN ARSENAL, MASS., OF PIECES OF BEAM STRAP. Abt. 52 ' 52 52 " 52 4 80 4.80 4.82 4-75 4.7s 4-75 fcl 9.02 4.3 '•74,8.35 4- '■97,9-5 3-7 2.07'9.83l 8. 1.67I7.93' 8.2 1.748.26. 2.2 370,200 3.1 325,500 16.6 383,000 7. 371,500 16.8 338,1 II.l 2SX,I 41.042 38,980 40,320 ■37,790 49,620 34,020 31,170 25,75° 26,320 28,990 6,480 28,450 CLASSIFICATION OF FRACTURE, Granular. Granular from fine to coarse. Fibrous. Granulur, 90 % Fibrous, 10 i^ Fibrous. Fibrous, 80 i. Granular, 10 %. Nos. 1,^,3 were part of top of beam. Area, 5% x 6J Average strength per square inch = 40,114 lbs. (2,820 kilogrammes per square centimetre). Average elastic limit per square inch = 27,747 ibs. (1,951 kilogrammes per square centimetre). Nos. 4, 5, 6 were part of bottom member of strap. Average strength per square inch, 38,143 lbs. (2,681 kilogrammes per square centimetre). Average elastic limit per square inch, 27,973 lbs. (1,967 kilogrammes per square centimetre). Average strength per square inch of i, 2, 3, 4, 5, and 6 = 39,129 lbs. (2,751 kilogrammes per square centimetre). Average elastic limit per square inch of i, 2, 3, 4, 5, and 6 = 27,860 lbs. (1,959 kilogrammes per square ceatimetre). This iron contained about 0.25 per cent, phosphorus and less than o.i per cent, carbon, and represents what would be considered an excellent sample of " phosphorus-iron." Fig. 76. Fig. 77. — FACES of Fracture of test-piece. STRENGTH OF NON-FERROUS METALS. 429 The Tenacity of Copper varies greatly with chem- ical and physical properties. Rolled and forged metal from well fluxed castings or ingots is stronger than any castings, and becomes stronger with rolling or drawing. Major Wade * found the tenacity of Lake Superior cast copper to range from 22,000 to nearly 28,000 pounds per square inch (1,547 to 1,968 kilog. per sq. cm.), averaging above 24,000 pounds (1,705 kilogs.). Egleston gives the tenacity of both Lake Superior and Ore Knob (N. C.) copper as above, 30,000 pounds per square inch (2,109 kgs. per sq. cm.). Anderson f gives the figures for the tenacity of copper, which, in round numbers, are as below — ordinary copper is compared with that fiuxed with phosphorus : TENACITY OF COPPER. Copper, forged. " cast forged. PHOS. 0.015 0.02 0.03 0.04 TENACITY, T. Lbs. per sq. in. 34,000 ig.ooo 25,000 38,000 45,000 48,000 50,000 Kilog. per sq. cm. 2,390 1.336 1,753 2,671 3.164 3.374 3.515 , The effect of fluxing with phosphorus is here very plainly shown and amounts to an average increase of tenacity of 4,000 pounds per square inch (2,812 kilogs. per sq. cm.) for each one per cent, added up to four per cent. * Metals for Cannon, 1856, f Strength of Materials, 430 MA TERIALS OF CONSTRUCTION— NON-FERROUS METALS. SHEARING. Angle formed by shear-blades, 3 degrees. Sheet Metals. IKON. COPPEK. BRASS. STEEL, PUDDLED. Thickness. Pressure. Thickness. Pressure. Thickness. Pressure. Thickness. Pressure. In. I.O* .615 .510 .404 .283 .183 .104 ■OS7 Lbs. 144,000 53,440 39.150 25,970 IS.7'S 10,390 4,200 2.180 In. .204 .150 .05 .02 Lbs. 11,196 6,007 4,820 3.676 2,200 1,006 552 113 In. •°s .042 •03s .025 .024 Lbs. 540 423 333 220 200 In. ■ 24 •24 Lbs. Z4,020f i4,93ot IRON. COPPER. BRASS, Diameter. Pressure. Diameter. Pressure. Diameter. Pressure. Diameter. Pressure. In. 1. 142 1.040 Lbs. 35,410 30,707 24.057 19,688 In. ■% • 447 .320 Lbs. 13,979 10,593 5.543 3.093 In. ■?^ ■ill Lbs. 18,460 13,872 11.310 8,218 In. I. no •00s • 779 .648 Lbs. 29.790 22,386 17.97S 11,648 The shearing resistance of copper is usually given in office hand-books as from 22,000 to 30,000 pounds per square inch (1,420 to 2,109 kilogs. per sq. cm.). Its value may be taken as the same as in tension and as subject to the same variations. The work done in shearing copper is, according to Has- well, measured, for punched holes, by IV= 96,000 dt, in which W is the work in foot-pounds, d the diameter of the hole, and i the thickness of the sheet in inches. Resistance to Compression varies with copper, as with all ductile and malleable metals, more with variation of form of test-piece and method of application of the stresi than with the ordinary modifications of composition and of form produced in manufacture, as ingots, sheets, rods, bolts, * The cutters were parallel ; the bar 3 inches wide. f With oil. t Without oil. STRENGTH OF NON-FERROUS METALS. 43 1 etc. The application of a crushing force to a test-piece of standard size and proportions first reduces it to the barrel- form, then to that of a flat cheese-shaped mass, and finally to a sheet of which the total resistance to compression increases indefinitely as its area becomes greater by flow. The com- pression stress thus increases from about that required to pro- duce rupture by tension to that demanded to produce free flow when the intensity of the stress is a maximum ; and its total amount is limited only by the area of the sheet pro- duced. The intensity, C, of resistance to compression is usually incorrectly stated, without limitation, as about ioo,cxx) pounds per square inch (7,030 kilogs. per sq. cm.) for rolled or forged, and 120,000 pounds (8,436 kilogs.) for cast copper. The results of experiments of the Author^ presently to be given, indicate that good cast copper, in cylinders of three diameters length, will exhibit a resistance which may usually be reckoned up to a compression of one-half or more, as C = 145,000 if ^' nearly, Cm = 10,000 aY ^' nearly, where C and C„ are the resistance to compression in British and metric measures, and e is the compression in unity of length, the resistance being reckoned per unit of original sec- tion. But the volume of the piece remaining practically un- altered, the section is increased very nearly in proportion to the compression, and the resistance will thus become O — 72,000 a7 ^' nearly, C„ = S,oooj/^' nearly, when reckoned per unit of area of section actually. 432 MATERIALS OF CONSTRUCTION— NON-FERROUS METALS. The effect of impact on the tough metals having no definite limit of elasticity is modified by the velocitj' of the striking mass, and by the inertia of the piece attacked, to an extent, as yet, not fully determined. The experiments of Kick indicate a considerable increase of total work of resis- tance, when the piece is deformed in this manner, over that noted when the compression is produced slowly by steady pressure. The experiments of the Author also indicate that this work is the greater, with soft and malleable metals, as the velocity of action is increased. The real efficiency of the press, as above, is thus probably somewhat greater than the figures obtained would indicate. The above facts were well shown in experiments by the Author on the power of drop-hammers, in which the work done in crushing copper cylinders was compared with that required under the slower action of the hydraulic press. Copper, Subjected to Transverse Stress, is prob- ably always to be considered as belonging to the second class of materials treated of as above, and as more correctly rep- resented by the equation here given below, than the usually adopted equation following : M^-R^\ {' J/ dydx, and F/= 2 Tt s, [' r-dr, it) Jo Jro instead of M=^\ j r dydx, and Fl=-^\ r^ dr, the former of which, for rectangular bearers and solid shafts, would become, were T = C, M' = -E}^~\ Fl= 2.2 s,f^, 4 ' ^ instead of M^lR^-fl; F/=u6s,r,\ STRENGTH OF NON-FERROUS METALS. 433 The values of T and C are not, however, the same, and the differential expression must be integrated for the two sides of the bar separately. Cast copper, tested by transverse stress, when of fair quality should give figures equal to, or exceeding, those obtained in the record which follows: TEST OF BAR OF CAST COPPER. No. 55.— Material: Copper, cast in iron mould.— Dimensions : Length between supports, /— 22"; breadth, ^ — 0.985"; depth, .■ + 4 .267 5.635,593 Bent. 29 23 0.993 1.002 HO 3.740 8. + 3 .273 6,734,838 Bent, 130 4.150 .270 6,185,210 44° MA TERIALS OF CONSTRUCTION— NON-FERROUS METALS. Queensland tin proved very good, showing a somewhat greater strength by transverse and torsional test than Banca tin, but a less strength by tension. The transverse strength probably appears higher than it should be, both on account of different methods of test, the Banca tin being tested by dead loads and the Queensland tin by platform-scale, and on account of a perceptible flaw in the centre of the Banca bar. In the test of No. 29, as above, a load of 40 pounds pro- duced a set of 0.0095 inch, and the elastic limit appeared to be reached at about 30 pounds. At 80 pounds, a crack was observed on one of the edges on the under side of the bar, which gradually opened but did not increase in length. At no pounds the bar sank gradually, the deflection increasing more than 6 inches in ten minutes. The bar was finally broken by repeated bending, and showed that the crack above mentioned was produced by an imperfection in the casting, about one-fourth of the surface, or that portion in which the crack was observed, showing radiated lines of cooling and the remainder the close pasty appearance peculiar to tin rupt- ured by bending. The crack weakened the bar, and the final bending was resisted by but little more than three-fourths of the section. Major Wade found the tenacity of Banca tin used in mak- ing U. S. Army ordnance to be 2,122 pounds per square inch (148 kilogs. per sq. cm.) ; its density was 7,297. The Modulus of Elasticity of Tin is stated by Tredgold at 4,600,000 pounds per square inch (285,400 kilogs. per sq. cm.) for cast metal, by Molesworth at same figure nearly, and is found by the Author to vary up to nearly 7,000,000 pounds (492,000' kilogs., nearly). Some of the figures obtained are given in the records of transverse tests of cast tin already referred to. No values have been found for other forms of this metal. Tin is, however, probably less affected by the form in which' it enters the market than other common metals, and the moduli here given may be accepted for general use as sub- stantially accurate. Tin in Torsion, as tested by the Author, gives STRENGTH OF NON-FERROUS METALS. 441 figures of which the following, from the Report of the tJ. S. Board, may be taken as fairly representative : TORSIONAL TESTS OF TIN. Averages 0/ Results calculated from Autographic Sirain-DiagrrTlt. MATERIAL. -3 «t-i 1 ORDINATES OF DIAGRAM. TORSIONAL MO- MENT. ,0 i |. 1 1 a i 1 ■a i < I "0 1 S8 29 Queensland tin .. Banca tin Mean (British) . . . Metric Sg- ins. 42.78 21.26 Degrees. 691,0 556.8 Itis. 0.48 Ins. 0.22 0.13 Fi.-ibs. 13-15 12.75 Ft.-lis. 4.36 5.78 2.9029 2.197s Ft.-lbs. 208.48 105.45 3 4 32.02 20.6 623.9 623.9 0.61 1.6 0.18 0.45 12.95 1.8 S-07 0.7 2.5502 156.97 The Queensland tin showed an extraordinary ductility in the torsional tests, one of the pieces twisting through an angle of 818 degrees, or more than 2}^ turns before breaking. This represents an elongation of a line of particles parallel to the axis on the surface of the cylindrical portion of the test-piece from one inch to 4.57 inches. The average of all tests of tin is given in the following : AVERAGE RESULTS OF TESTS OF TIN. TRANSVERSE TESTS. TENSILE TESTS. TORSIONAL TESTS. i 1 Tenacity per ^ .a \ z .£P square inch £i « t a; of— •^ u ^ a ^ ■s a •3 ;« fj .§ \ 1 Rl d c u .1 Pa s • 1 bi an 1 bs e Is? u 1 •A "o 1 % I 1 1 a .a 2 a 1 a 1 1 i m ^ s I 6 1 1 ^ 1 1 130 4,150 .270 6,185,210 • 3551 3,130 .476 12.95 5-07 -392 2.5502 156.97 442 MATERIALS OF CONSTRUCTION— NON-FERROUS METALS. The Strength of Zinc has been determined by but few investigators, and, like that of all other useful metals except iron and steel, is a subject of which comparatively little is known by the engineer. Cast zinc is stated to have a tenacity of about 4,000 pounds per square inch (281.2 kilogs. per sq. cm.), and a resistance in compression of ten times that amount. Stoney states the tenacity at nearly 3,000 pounds (211 kilogs.) cast, and Traut- wine gives for sheet-zinc and zinc wire 16,000 and 22,000 pounds per square inch (1,1 24.8 and i ,546.6 kilogs. per sq. cm.), respectively. The modulus of elasticity is given by Wer- theim and by Tredgold at from 12,000,000 to nearly 14,000,000 pounds per square inch (843,600 to 984,200 kilogs. per sq. cm.), the value being higher for cast zinc. The Author has obtained much smaller figures. Pure zinc, like pure tin, is never used alone, by the engi- neer, for purposes demanding strength and toughness. The values of the several moduli are given as of interest, how- ever, and for comparison. Samples of cast zinc tested by the Author show variable tenacity, the figures ranging between 4,500 and 6,500 pounds per square inch (2,847 to 4-253 kilogs. per sq. cm.), or consid- erably above those given by earlier investigators. All the zinc thus tested by the Author was very pure, and made from New Jersey calamine. The effects of varying time and rapid- ity of strain are observable in zinc, as in tin, and are the same in kind ; they will be described later. Zinc is much less ductile than tin. The resistance of zinc to compression varies with the de- gree of reduction, and, as tested by the Author, was about 22,000 pounds per square inch (1,547 kilogs. per sq. cm.) when the compression amounted to one-tenth the original height of test-piece in pieces three diameters long, and one-half greater for a compression of one-third. Zinc is weaker under compression than any copper-zinc alloy. Zinc has no defined elastic limit, but an apparent elastic limit in compression was recorded at 5,000 pounds per.square inch (352 kilogs. per sq. cm.). STRENGTH OF NON-FERROUS METALS. 443 Records of Test of Zinc are given below, as reported to the U. S. Board. TENACITY OF CAST ZINC. Length, 5"; diameter, 0.798"- LOAD. TOTAL EXTENSION. SET. PER CENT. ELONGATION. REMARKS. 800 1,200 1,600 2,000 3,000 4,000 200 5,000 6,000 6,300 O.OOII 0.0024 . 0034 0.0051 0.0097 0.0157 0.0206 0.0240 Broke. ' o.oog6 0.02 U.02 0.07 O.IO 0.19 0.31 0.41 0.48 Diam. fractured. Section, 0.796''. Tenacity, 6,300 pounds per square inch (4,429 kilogs. per sq. cm.). COMPRESSION OF CAST ZINC. Length, 2" ; diameter, 0.625"- COMPRES- COMPRES- LOAD. SION. LOAD. - SION. Total. Per sq. in. Per cent. Total. Per sq. in. Per cent. 1,000 3,259 0.15 8,000 26,076 12.15 2,000 6,519 0.55 9,000 29,335 17.15 3,000 9,778 1.85 10,000 32,595 20.60 4,000 13,038 3-40 10,000 21.80 5,000 16,297 5.10 10,500* 34,225 24.40 6,000 19,557 7.20 Resistance fell to 7,000 22,816 10.65 10,000 32,595 33-35 * Continued one minute. 444 MA TERIALS OF CONSTRUCTION— NON-FERROUS METALS. CAST ZINC LOADED TRANSVERSELY. LOAD. DEF. SET. E. REMARKS. 20 O.OIOI 40 0.0171 6,698,725 60 0.0246 6,927.556 Modulus of rupture. 80 0.0324 6,984,644 R —1< 540 pounds per 100 0.0424 6,655,180 sq. in. (5,300 kilogs. ,120 0.0506 6,680.965 per sq. cm.). 140 0.0616 6,395,032 Most probable value 3 0.0 of .£ = 6,900,000 160 0.0753 5.973.588 Em = 428,130. 180 0.0906 5.581,549 200 0.1244 1,797,132 Broke. TESTS OF CAST ZINC BY TORSION. Length, i" ; diameter, 0.625". NO. AREA DIA- GRAM. ANGLE. MAX. ORDI- NATE. MAX. MO- MENT. EXTEN. EXTER. FIBRE. 21 A 21 C 21 D 21 B 19.63 18.81 17.24 18.13 123° ■129 151 163 2.15 2.07 1-95 2.15 N 37-83 36.55 34-42 37-83 . 2042 0.2227 0.2955 0.3380 Other Metals than those already described have been made the subject of very few experiments and the data obtainable are very unsatisfactory. The alloys of the three principal non-ferrous metals are made the subject of succeed- ing chapters. Lead\\.'as, a tenacity which is reported by Haswell as : In compression the resistance is stated to be 7,700 pounds STRENGTH OF NON-FERROUS METALS. 445 per square inch (541 kilogs. per sq. cm.) and the modulus of elasticity is given as 720,000 lbs. (49,350 kilogs.). Wertheim, however, obtains a value of 21,500,000 pounds per square inch (175,750 kilogs. per sq. cm.). Trautwine gives, for tenacity : LBS. PER SQ. IN. KILOGS. PER. SQ. CM. Lead, cast. . " pipe. " wire. " sheet 1,800102.400 1,700 to 2,240 1,600 1,925 116.5 to 168.7 119. 5 to 157.5 112. 5 155-5 as collated from various older experiments, and a resistance to compression agreeing with Haswell. The strength of lead pipe, as obtained in market, has, when tested, been found variable. The best results noted by the Author * indicate a tenacity of the metal exceeding one ton per square inch (2,240 lbs.; 157.5 kilogs. per sq. cm.). Comparing the results of a number of experiments to obtain a value of/ in Clark's formula: logR' p=TlogR; in which T is the tenacity, p the pressure, and R the ratio of external and internal radii, a mean value of T was found to be 1.4 tons per square inch (220.5 kilogs. per square cm.). The minimum value was three-fourths as great. It is prob- able that a much lower pressure, long continued, would have burst these pipes. The thickness of lead pipe is frequently determined by the rule : t — 0.0024 « ^ + 0.2, in which t is the thickness in inches, n the pressure in atmos- pheres and ^the internal diameter in inches. * Lond. Engineer ; Nov. 16, 1883, p. 378. 446 MA TERIALS OF CONSTRUCTION— NON-FERROUS METALS. Antimony has a tenacity of about i,ooo pounds per square inch (70 kilogs. per sq. cm.), and bismuth of three times that amount. Gold is a moderately strong metal, with a tenacity, cast, of 20,000 pounds per square inch, and of 30,000 in wire (1,406 to 2,109 kilogs. per sq. cm.). Silver is reported to be about equally strong (?) in the two forms, having a tenacity of 40,000 pounds per square inch (2,812 kilogs. per sq. cm.), according to Baudrimont. Platinum has a strength of from 30,000 to above 50,000 pounds (2, 109 to 3,5 1 5 kilogs.). Nickel, tested by the Author, exhibited tenacities of from 50,000 to 54,000 pounds per square inch (3,515 to 3,543 kilogs. per sq. cm.), elongating about 10 per cent. Palladium, tested by Wertheim, had a tenacity equal to that of nickel. It is ques- tionable whether any of these metals have a true elastic limit. Wertheim on Elasticity. — Wertheim gives the fol- lowing as the densities, atomic weights, and products of the two, and also the tenacities and sound-conductivity of several metals : S. G. AT. WT. < X a RESISTANCE TO RUPTURE PER MILLIMETRE. ■a 3 By extension (Guyt-Mor- veau). By compres- sion (Ren- nie). SE U ■3 11.352 7-285 19.258 10.542 6.861 21.530 8.850 7,788 12.94498 7.35294 12.43013 6.75803 4.03226 12.33499 3-95695 3.39205 0.8769 0.9907 1-5493 1-5599 1.7015 1-7454 2.2365 2.2959 0.022 0.063 0.274 0.341 0.199 0-499 0.550 1. 000 1-45 6.20 38-55 600 3.200 Tin 7-5 Gold Silver Zinc ... 9.0 9 600 Platinum Copper 12.0 20 000 17.0 • He infers a general variation of cohesion with change of intramolecular distances, and obtains his data from experi- ments upon fifty-four binary alloys and nine ternary alloys. He gives the following values of moduli of elasticity: STRENGTH OF NON-FERROUS METALS. 447 MODULI OF ELASTICITY OF METALS. LBS. PER SQ. IN. KILOGS. PER SQ. CM. Lead , 2,500,000 7,700,000 11,500,000 10,000,000 17,000,000 24,000,000 176,000 492,000 808,500 703,000 1,195,000 1,687,000 Gold Silver Palladium Bischof 's Method of Test to determine the purity and economic value of metals consists in making strips of a definite and standard size and subjecting them to repeated bending. The purer the metal, as a rule, the greater the number of changes of form required to produce fracture. Zinc, for example, was found to withstand 100, 54 or 19 bendings accordingly as it was pure zinc, best commercial spelter or the lowest quality. The ill effect of the introduc- tion of 0.0000 1 tin, or of 0.0004 cadmium is perceivable even more certainly than by analysis. Metals which do not alter by remelting, as tin or zinc, are melted in crucibles, with continual stirring and then cast in ingot moulds, 12 cm. long, 1.3 cm. square at the top and 0.3 cm. square at the bottom, 40 or 50 grammes being taken for a test, or 60 grammes for lead. The bars thus made are rolled to the desired thinness, annealed and tested. Metals, as brass, bronze or copper, which are liable to change in fusion, are rolled from the commercial form, with repeated annealing. The strips tested by Bischof were 13 cm. (4 inches) long, 0.7 cm. (2 inches) wide and of such thickness that they weigh as follows: Copper, 17; brass, 16; tin and zinc, 15; lead, 25 ; iron and steel, 12 grammes. They were tested in a " metallometer," in which they could be bent conveniently to any angle. Repeated flexure and reflexure through an angle of 67j4 degrees was found best adapted to bring out the quality of the metal. CHAPTER XIII. STRENGTH OF BRONZES AND OTHER COPPER-TIN ALLOYS. The Bronzes — under which name are included the principal alloys of copper and tin, and a few special composi- tions — vary, in strength, elasticity, ductility and hardness, with variations of composition to such an extent that they find application in an immense number of the engineer's construc- tions, their character and chemical constitution being adjusted to his needs. The most common of these alloys is " gun- bronze," which consists, usually, of 90 parts copper, 10 of tin, or 89-copper, 1 1 tin. Such bronze has a strength which will depend greatly on the soundness of the castings ^nd purity of the constituents of the alloy, but which often may exceed 50,000 pounds per square inch (3,515 kilogs. per sq. cm.) in tension. Bronze used for journal-bearings in machinery is made harder or softer, according to pressure sustained, the com- position approaching usually that of gun-bronze, and rangmg from copper, 7; tin, i; to copper, 11, tin, i; i.e., copper, 87.5; tin, 12.5, to copper, 91.67; tin, 8.33. A little zinc or lead added slightly softens it. Packing rings for steam engines are made of still softer and more ductile bronze — copper, 92, to copper, 96. These alloys have been very fully described elsewhere, and this chapter is devoted entirely to the consideration of their strength, ductility, elasticity and density. Gun-bronze, according to the "Ordnance Manual," should have a tenacity of 42,000 pounds per square inc^ (2,826 kilogs. per sq. cm.), and a specific gravity of 8.7. In Major Wade's report on " Experiments on Metals for Cannon," 1856, are given records of a number of tests of gun metal. Specimens of metal from 83 "gun-heads" (the upper part STRENGTH OF BRONZES. 449 of the casting is always deficient in strength) gave an average result of 29,655 pounds per square inch (2,085 kilogs. per sq. cm.), the highest figure being 35,484 and the lowest 23,529 pounds. This alloy was copper, 9 ; tin, i. Small bars made of gun metal gave higher figures. One set of 16 bars gave an average result of 42,754 pounds (3,006 kilogs. per sq. cm.), and another similar set an average of 41,284 pounds (2,902 kilogs. per sq. cm.), the lowest figure of the 32 specimens being 23,854 pounds and the highest 54,544 pounds. Five of the specimens gave more than 50,000 pounds (3.515 kilogs. per sq. cm.), and only three less than 30,000 pounds (2,109 kilogs. per sq. cm.). The average of 12 gun-heads was one-half that obtained from the small sample bars cast with the guns. A sample of very inferior quality fell below l8,000 pounds (1,265 kilogs. per sq. cm.). Major Wade found the quality of bronze ordnance enor- mously irregular and uncertain, and considered it very im- portant that a more reliable method of manufacture should be found. The tenacity of gun-bronze thus depends greatly upon the method of manufacture, of casting, and of cooling. By careful handling it has been given a tenacity, in ordnance, exceeding, even, 60,000 pounds per square inch (4,218 kilogs. per sq. cm.), and the Author has obtained small bars still stronger. Bronze ordnance of large size has been made here and in Europe with success ; it is, however, very liable to be irregular in composition and physical character, and the un- certainty always felt in regard to its condition is an element which enters into the question of its use for any purpose. Continual use of ordnance is thought to lead to a separation of the tin from the copper, and to final destruction. The gases of powder sometimes corrode the metal badly. The Modulus of Elasticity of gun-bronze is given by Tred- gold at 10,000,000 pounds per square inch (703,000 kilogs. per sq. cm.), and this figure is confirmed by the experiments of the Author as given later, but it is subject to great variations with the condition of the metal. 29 45° MA TERIALS OF CONSTRUCTIOlt— NON-FERROUS METALS. Manganese Bronze is another valuable alloy. That used in the construction of torpedo boats for the British navy was supplied under a contract calling for a tenacity of 26 to 31 tons per square inch (4,094 to 4,883 kilogs per sq. cm.), and an elongation of 20 per cent. This sheet bronze was from -^th to ^th inch (0.16 to 0.32 cm.) thick (No. 9 to No. 18 B. W. G.), and sustained 29 to 30 tons (4,567 to 4,725 kilogs.), stretching 25 to 35 per cent., and bending cold to a radius equal to their thickness. Manganese bronze, tested at the Royal (British) Gun Fac- tory at Woolwich,, England, by tension, gave the following figures, as reported to the Admiralty : TENACITY OF MANGANESE BRONZE. (Sheet Metal ; Rods and Bolts.) LOADS NOS. TION. Yielding. Breaking. Tons per Kgs. per Tons per Kgs. per Per sq. in. sq. cm. sq. in. sq. cm. cent. ■ 4,766 r4.o =,204 24.3 3,817 8.7 Cast in metal mould. 4,767 12.6 1,984 29.0 4,567 31.8 Ditto and forged. 4,768 14.0 2,204 22.1 3.480 5-5 Ditto. 4,769 13.2 2,079 28.8 4,535 35.3 Ditto and forged. Cast in metal mould, slight law 4,770 j6.8 2,645 23.6 3,717 3-8 in specimen. , 4,771 12.0 i,890» 30.3 4,772 25.7 Cast in metal mould and forged. V ROLLED RODS. m 6,Sl6 11. 1,732 29.0 4,567 44.6 Mild, for ships' bolts and rivets. 6,S4S 16. 6 2,615 30-7 4,83s 20.7 High, for Engineers' bolts, pump rods, etc. Medium. 6,546 14.6 2,299 30.0 4,725 26.2 6,547 34-4 5,417 39.6 6,237 11.6 Cold rolled. AREA OF SPECIMENS, O.I33 INCH. LENGTH OF BREAKING PART, 2 INCHES. f 7,364 13.8 2,173 28.57 "^^ 28.7 Pulled in direction of fibre. I, 7,365 14.06 2,205 28.46 23.2 Across fibre. 7,369 14.06 2,205 30.13 4,740 4,850 47.8 With fibre. K 7,372 .4.8 »,33i 30.78 34-1 Across fibre. I 7,374 rt,7 2,630 30. t 4i74o 28.8 With fibre. STRENGTH OF BRONZES. 45 1 The Copper-Tin Alloys, which, as has been stated, furnish a very large number of the best bronzes and engi- neers' compositions, and which are extensively used in every department of construction and the arts, had never been sys- tematically studied until the investigation was made by the U. S. Government Board upon a plan prepared, proposed, and carried out at the request of that Board, by the Author. Earlier investigations had been confined to a few familiar compositions, and it was only when appropriations made by the Congress of the United States could be applied to such a research that it became possible to determine the method of variation of strength, elasticity, and ductility, and of spe- cific gravity, and other properties, with variation of compo- sition throughout all the possible proportions of copper and tin alloys. In the research to be described the principal as- sistant employed by the Author was Mr. William Kent. Final Results. — The following table exhibits the results of the whole investigation in a compact form which permits ready comparison of data. The average results obtained by test of the copper-tin alloys, enable the engineer to reach tolerably definite conclu- sions relative to their value in construction. The results are given as obtained by the four principal methods of stress. They are very variable, and this variability is due not only to the variation of composition of the alloys, but also to their differences of physical structure, and is, therefore, to some extent, accidental. General conclusions may, nevertheless, be deduced and the principal facts revealed by test, and these conclusions are also most unmistakably exhibited by the diagrams pre- sented in this and preceding articles. The figures given by the tests have been plotted in the form of curves having for their ordinate the resistance ob- served and for their abscissas the distortion of the given test- piece. These curves exhibit the method of variation of re- sistance with progressing change of form, and constitute "strain diagrams " which exhibit to the eye every important quality of the material. 452 MATERIALS OF CONSTRUCTION— NON-FERSO US METALS. ^ •^ 'aoaajiisa^ »^ ro "1 3 ■ajqg jou3jx3 JO uoisasixg: ■uois JO S33BJU3DJ9d •TnnraixBH iBuiSuo JO saSBjuaojaa *uoi5S9Jdmo3 jo lunoinv spnnod *i{idu3j:)S Suiqsiu^ 'IBUI -Sue JO saS-EiixaDJsd 'uoij -03S pajnjorjj JO jaianiiEid ■m3a3[ iBniSuo JO sycd ui *uoi)B^ubi3 l^joj;. ■pBOi SnpTBajq JO s3J§B]ua3J3d''j(ai!i ops^i^ V I •s spao ■n0T5 -Das pajnpejjj •UOTl -D3S pUlSuQ •^jpijsup JO snppopi O ro 66 w ^ M m *0 r^ CO o 'saqDui f " JO UOTl -oagap b uiqiiM. sousiiisa-^ ■Saispsajq a J o J a q uopoagap piox ■pBOI Sui5IB3jq , JO *SLIBd UI '?[Ul!l DlJSUia ■ajmdru JO sn|npopi •AilAVaD 3IiIID35S NVaW OSS So5 (I) ^ < " o r S 2 S ■oix ■jaddoQ ■jaddo3 . iCO N s „■« Cft t«»vo h- 1^ to fO ro to c< m M H w OOOOOOOOOOOO 00 fo fT* lovo "O CO ■* fOCO fOUOO ^H-W H Q ro fo CO * en -^ o >o io« *o N foioroOVD O O O tomO dsoo 00 o mco ^co >o m oco n lO t>. ro « H \0 CO Oi^O O -^ t^'O « fo 'f d" fo d^vfT eT -^ cI vcT »o M hO^ O foBoO >OM o>co N T &C0 M ■* M Oi^O O "^ mvo I 00 ^^'O > - " N N ( ro ^ ^ O ^* m lo »o N M ro lo cS ro m" cT in" t». « >n o "^ fo -^ fo -^ CT^ a-oo lo N t- vo »o O O "^ OCO W Ci Cho ■*lolo»n^N,•,^ looo o O o roH»o -d-M loio-tt- N ro « fo -^ M c c c a OJ OJ 4J V I M U^ n H H CO M f*» (M'O^Mt-iOOO I- el d o' o" 6 d d d ro fODO ^Q ro-» fO ■■i-\5 ro M W M M t-s oi CO Ov-O M It- O>0O ifi ^00 Ov ■* w *0 « ro ro ■* ^. lO MVO VOVOIOVO^O t^t^O>»00>C3\0* cococococdcocooooooooocooooo'oococo ■ M rot^io oi «- • o ON o> ov o\ o^oo 00 00 00 tv r-. r^vo ^y^ L o P S P„ R o !2 a { I rot.*0 N rot^O H "NOIiHOdOad DIKOXV 3 3 d : d a M o w M m ■* N »ri ro'is '^^>.looo^o o» o ton totococrimM STRENGTH OF BRONZES. 453 H00000000000000000MNmMma»«mrnO^Oodmm N « ■^ m c^oo o' « 1 m M m H m ro m m^o c». m >J- m ■* m do " "~" ~~~"--^--iNVD0MHm t^^o tv o«co c?i m ■''""' 'T^p^PJ-Q t:-.co 00 t^ ts ■ - • . • . . ..".". ~. ".".".-.'.-, -.-.•;-,- -^ ^ '^ " "^ "^ " "^ O "^ .o_o_o_o_o_o o o o' o' o' o' o' d o' o' o ' o' o' o' o' o' o' o' o' o' o' o' o' H M M ci .iPhPQ.2 Q.9 o o Q Q N romo o t-^ro-*t*.oco oTn mommooonou^ ^. ^ ^ °. ^ °. T °. ° ° *^ '^^'^ '^ '^'^ vo m M lo mvS vooomt^iNO mQ mmK M O M O M O' H M M O* M H W «' W cl N CO -^ t^ m O' 4vo" H M Oo' o" N t^ mvo' M « itvG m tN M vo CO mm — ^ H H {s fo m m m 8. 8 ^ ? 8. 8. 8. 8 8 8 8 8 8 8 § § S H g S sl: R gf^- K S l?^ S S'?n.^gS;K:S£-^SS'K>^&>e,EJ,SS,a,gSS.gS?'acS'£'R??K GO \o M H CO p^ m « W M H H m « m^o ( - Oi t? : Vr, ■>K >H^ : ' ■* * m : + : : " w : + : + ■ CO ^Ti \r^ \r, '■t- \3 Tt^ooo o o Q H nooom-^t^tM ov I O 0^^0 01 I 'd-vo m •* « \o m-o p~oo vo cT i 1 m r^ o CO o o ( 1 ^-^o ooooo'oooomo I m H N cT'Oo m c\co o m tx 5 I moo m tN. m ■* o tx'o ^o ■«*- m S ■^ M vo" m>o vo vo -^ m m -^ m I t-* m w H I Q m\o o CO m o ^ - . . 1 & m CT. CT. CT>O0 M N CO lO N (S O* Tf 00 00 M 00 Ov " -•--' OOO 11 M C P- •) -tfco npjw-^ejmO-* m^o n - - , , - , I iM m o tx T^oo CO o w tx^mmo o^O— -. ]_co^ « m^o O CTi t>co m-^ts-roo o^'*-On Oi*0 • vcT cS m m" d~ M^io i^t^""-------- ■* m I ~ ~ " n m m^o 0!J_ m pT H c^ p^ d t^ p^oo '•o *o ^^ *m " P^ N M H m m P-oo M O « H ^ s E?^^ j^'s Svg^ f:?'?. js s^^g- ^ s. S- g; K^^ 3; s O OOOOOO OOOOOOHMm mco lOM ctM\o mmosmmoNcJi M M « t-» ■* p^. m-o m M M wOONi-'OogOc 0^'0 «3 « Q Q mNH oamm moo \D*^*^>■^*^'^~'*^*i^^ 1-1 MH fon hO-^h H«m -^vo CO Ti-H « (N-o O ■^mmp^ddddOfld d oooaooooooooooooHH«N-^m mco ooaJOjSajBSS o' o' o* d d d d d d d d d d d d d d d d d d d d d m M W CQ cq CQ P3 CP P3 imp^«QOOP».OOmc I mvo n mmo oi'jfmc ! tx w M \o m tv. c -.Qio mci t>t-*C ■* Oi p> p^ i-i t>co vc ^« cfmcrH'mn I « -^00 p^oo 00 Pv 0*03 m -^-o m 'J- m ■* "too o I -^co tx'o « Qco Mcovo mo Qio « p-io m tJ- h O) m H. 0_ i-i l-\0 0_n^0 iN_p-.mi>.omq;-*Px rvd" m t^co H pT m o' o" d\ d" m cS'o'oo m -^ m tx tx^D o M N m 6 « y O -^ m p>.co 00 -:^^D 'i- • o> cr< oi c^ p^^o vo m -^ - ■-« « H mMco miO"*-0 Q ^*mM Ti-txO n mO m TOoo o H m-P*»-'C0NVO • p^ N oi m m ceo H CO M N mvo in vo m ^ m p^ mM3 o ■* • « mintiiH N mem m\o P^ d> m -i^-oQ oo h mvo p- O- o\ ■ \n^mri mvo >o vo "o p». p- p^ t». ^.oo ooooooccticccon • £-i'i>S fi^ S !J, 5: !? ? ? i'S, S S ?, S"S SSS'&.'S m N d vo ir CO* tN.m i5 --fmo H H od\d m M d d j s.s'R'sa^. Maya's \o ■^yo P. S. E tv p^co 00 oo 00 o o H cm^ mm t^moo -to C C C C C Oi mo m p^ m m moo o^ >ovo mt^M vioimp^mmviH r^ m N H p^id w w p^ pv 4" M H di tC ^ ei CO r^ -^ c* h SD^o'O'O lommm■i*■■<^'*■^^■ O S w Q in in 454 MA TERIALS OF CONSTR UCTION— NON-FERROUS METALS. Strain-Diagrams, obtained from tests determining tenacity of the bronzes, are given in the accompanying figure as derived from experiments upon the first series of copper- tin alloys. No. i, pure copper, to 29, pure tin, inclusive. The curves marked A are from the upper end of the bar and B from the lower end. These curves may evidently be divided into three classes : viz., those which are very rigid and brittle, as 7 A, 7 B (cop- per, 80, bell-metal), those which are very ductile and mallea- ble but soft and weak, as Nos. 26 to 29 (tin, 95 to 100) inclusive ; and those which combine strength and ductility and possess, therefore, great resilience, as Nos. 2, 3 and 4 A (copper, 93 to 98, gun-metals). All intermediate qualities may be obtained, but these are typical and the most valuable of these compositions are evidently, for general purposes, those belonging to the last class, and of which the strain- diagrams lie between the extreme qualities, one set of which lie near the axis of abscissas, while the other set lie nearer the axis of ordinates. For some purposes, as when, for example, it is desirable to secure a high elastic limit as well as moderate toughness, alloys like ordnance bronzes, Nos. 4, 5, 6 (copper, 86 to 93), which are stiff and strong, although not very ductile, may be chosen. Cases may even arise, although certainly not often, in which the rigidity of bell-metal, No. 7 (copper, 80), may make that alloy valuable in consequence of its high elastic limit, notwithstanding its great deficiency in ductility. The Tenacities of the valuable class of these metals range not far from 30,000 pounds per square inch (2,109 kilogs. per sq. cm.), the strength increasing somewhat with the pro- portion of tin up to 18 per cent. Within that range, the expression T = 30,000 -H 1,000 t, in which T is the tenacity and t the percentage of tin, may be taken to represent a maximum which selected materials STRENGTH OF BRONZES. 45 5 should always give. Hardness is here seen to increase steadily from pure copper to copper 75, at which point that of mini- mum ductility is reached. From this point it decreases stead- ily and with tolerable uniformity to the opposite end of the series. Malleability takes an almost precisely opposite course, falling to zero at copper 60-65 and rising again to the end (pure tin). Fusibility constantly lessens, as tin is added to copper, from end to end of the whole range. The curve of ductility closely follows that of malleability in alloys rich in copper, but the lack of cohesion of tin causes a great falling off at the opposite end of the line. CHAPTER XIV. STRENGTH OF BRASSES AND OTHER COPPER-ZINC ALLOYS. The Brasses include all the copper-zinc alloys con- taining one-half copper and upwards, and a few special alloys are also given the name, as are copper-tin-zinc alloys, of which the tin forms but a small proportion. The name bronze has been applied, occasionally, to these ternary alloys, also. The terms bronze and brass are used indifferently by the older writers, but the tendency to restrict each term to a binary alloy, or to a ternary alloy in which one constituent exists in very small proportion, is decidedly observable among later writers and they will be so used in this treatise. In the cases of the brasses, as in that of the bronzes, no systematic investigation of the properties useful to the engi- neer had been made except by the U. S. Government. The U. S. Board, to which allusion has been already frequently made, authorized a determination of " the mechanical proper- ties and of the physical and chemical relations of alloys of copper, tin, and zinc," under the arrangement of committees approved by the Board,, which assigned to the Committee on Alloys the duty of " assuming charge of a series of experi- ments on the characteristics of alloys and an investigation of the laws of combination." This research was conducted in the Mechanical Labora- tory of the Department of Engineering of the Stevens Insti- tute of Technology under the direction of the Author. The facts and data thus discovered and placed on record * will bg summarized in this chapter after reference to earlier work on nearly related alloys. * Report of U. S. Board, Vol. II. ; Ex. Doc. 23 ; 46th Congress, 2nd Ses- sion. Washington : Government Printing Office, 1881. STRENGTH OF BRASSES. 457 Earlier Experiments — Mallet * found the tenacity of an alloy of copper, 90.7, zinc, 9.3, to be 27,000 pounds to the square inch (1,456 kilogs. per sq. cm.), with a specific gravity of 8.6 ; with 3 per cent, more zinc the strength was increased to very nearly 30,000 pounds (2,109 kilogs.). Copper, 85.4, zinc, 14.6, had a tenacity of about 32,000 pounds (2,249.6 kilogs,), and with copper, 83, zinc, 17, the figure became 31,000 (2,179 kilogs,). The tenacities varied little throughout the range and down to copper, 2, zinc, 1, which is a Muntz metal. Equal parts copper and zinc exhibited a tenacity of 20,000 pounds per square inch (1,406 kilogs. per sq. cm.) in Mallet's experi- ments ; the Author has obtained, in some cases, 40,000(2,812 kilogs.). Alloys rapidly become weaker, passing this maxi- mum, as the proportion of zinc is increased, as will be seen later, passing, however, a second maximum at about copper, 10, zinc, 90, which gives figures one-third as great as the first maximum. Brass cartridge metal tested with copper and steel by Lt. Metcalfe at the Bridesburg Arsenal in samples trimmed out to a contracted section of one inch (2.54 cm,), minimum breadth, and 0.03 inch (0,076 cm.) thick gave results as fol- lows : TENACITY AND ELONGATION OF CARTRIDGE METAL. LOAD. PURE tOMMERCIAL COPPER. BRASS. OPEN HEARTH STEEL. Lbs. Kilogs. COPPER. Unannealed. Annealed. I. II. Z 227 272 363 454 499 544 680 726 771 817 0.024 0.040 0.078 0.155 0.005 0.020 0.063 0.156 a.266 0.005 0.015 0.040 0.087 0.130 0.214 0.290 D.013 D.025 0.042 0.062 0.085 0,117 0.157 0.217 0.323 0.050 0.075 o.roo 0.130 0.165 0.220 0.350 0.0225 0.030 0.0425 0,060 0.0775 0.140 0.230 800 x,ooo 1,100 1,200 1,300 1,400 1,500 0.033 0.050 0.075 0.102 0.152 0.266 CO. 27 0.057 0.085 O.IZO 0.163 0.270 0.005 0.007s 0,013 0.030 0,065 1,700 * Phil. Mag., Vol. 21, 1842. 458 MATERIALS OF CONSTRUCTION— NON-FERROUS METALS. As the test-pieces were of the " grooved " form the elonga- tions serve for comparison of these specimens, but have no absolute value. Sterro-Metal, a brass which contains a little tin and iron, was tested by Baron de Rosthorn at Vienna, and gave the following results : * TENACITY OF STERRO-METAL. MATERIAL. TENACITY. Lbs. per sq. in. Kilogs. per sq. cm. 60,480 76,160 85,120 40,320 4,252 5,354 5.984 2.834 Gun-bronze j cast This alloy contained copper, 55.04; zinc, 42.36 ; tin, 0.83 ; iron, 1.77. The proportion of zinc may vary from 38 to 42 per cent, without appreciably altering the value of the alloy. The specific gravity of this metal was 8.37 to 8.40 when forged or wire drawn; it has great elasticity, stretching 0.0017 without set, and costs 30 to 40 per cent, less than gun-bronze. It has been forged into guns, cold from the casting. The strength of sterro-metal containing one per cent, and more of tin will be given in the following chapter on ternaryalloys of copp^, tin and zinc. The Moduli of Elasticity, E., of various alloys have been found, as below, to the nearest round numbers : ■ HoUey ; " Ordnance and Armor,'' p. 424. STRENGTH OF BRASSES. 459 MODULI OF ELASTICITY OF BRASSES. VALUE OF E. AUTHORITY. METAL. Lts. on sq. in. Kilogs. on sq. cm. REMARKS. Brass. 9,000,000 12,000,000 13,000,000 632,700 843,600 913,900 Tredgold. Wertheim. Bauschinger. ) II tin, 8g copper, cast ti Rolled. As will be seen, presently, the value is very variable with ordinary cast alloys of copper and zinc, but should be toler- ably uniform with rolled and drawn materials. Copper-Zinc Alloys, including the brasses, were studied by the .Author, and the investigation was, as al- ready stated, conducted in a similar manner to that described in the discussion of the alloys of copper and tin.* The specimens were in the form of bars, and were cast in an iron mould square in section, and similar in dimensions to that used in making bronzes. The experiments were made upon these bars as cast under ordinary conditions as before. The effects of different methods of casting, of slow and rapid cooling, of compression, either of the fluid metal or after soHdification, and of rolling, tempering and annealing, were to have been made the subject of a special research. Two series of these alloys were made and tested. The first series was composed of bars differing in composition by 5 per cent. The bars of the second series also differed in composition by 5 per cent., the first bar containing 2^ per cent, zinc, the last bar containing 97^ per cent. The bars were first tested by transverse stress ; the two pieces remaining after each transverse test were turned to size and tested by tension, and the four pieces thus formed * This account is mainly abridged from the Report to the Committee on Alloys of the U. S. Board. 400 MA TERIALS OF CONSTRUCTION— NON-FERROU S METALS. were tested by torsion. Some tests were made by compres- sion. The turnings from the tension test-pieces were analyzed. The specific gravities were also determined. The total weight of each casting was 4.5 kilograms (9.92 pounds). Compositions Tested. — A following table (p. 461) gives the compositions of the bars according to the original mixtures, the compositions of two portions of each bar as subsequently determined by analysis, and the specific gravi- ties. Bar No. 16 was made by melting together the upper half of bar No. 17 (21.00 copper, 77.59 zinc) and the lower half of bar No. 15 (25.98 copper, 72.90 zinc). The mould was heated each time before pouring into it the molten metal, the temperature given to it being higher the larger the amount of copper in the alloy. In melting the metal for bars. No. 7 to No. 21 (35 percent, zinc to pure zinc), inclusive, except No. 16, the copper was itielted first and covered with a layer of charcoal. The zinc was melted in a separate crucible, and poured into the crucible containing the molten copper, through the layer of charcoal. The mixture was thoroughly stirred with a dry stick. Some volatilization of the zinc took place, the amount being greater at some times than at others ; but the causes of this variation were not determined. Bars No. i to No. 6 (5 to 30 per cent, zinc) were made by first melting the copper, and then adding the zinc in the solid state. The losses of zinc vary very irregularly, and in two cases, bars Nos. 18 and 20(85 and 95 per cent, zinc), there ap- peared to have been a greater loss of copper than of zinc. The temperature of casting was then found by the formula Pc +'^' in which P is the weight of the water, P the weight of metal poured, / the temperature of the water before, and ^' after. STRENGTH OF BRASSES. 461 N If ^ g >> < g ^ s ft- U ? S 1 ^ s > >J »1 s -SI M •^lpi]5Ei3 JO sn^npoj^ I "(spanod-jooj) aonaiiisg^ 3 Q O 1U ^S •3jqg JOU31X3 JO uoisu3;xa •(S33j33p) nois -joj JO aiSuv '(mnmrxBoi JO •juaa'jod) ■(spunod-^ooj) uoissojdaioo jo junomv" o m H m'O MM M \o m m Oi i M 00 moo to t^ r- iQ^O I I 10 Q rn N -1 fnv5 -o to s m M N CO 00 0\^ 5 w '^ S £r > -4. u <^ H -r\0 ^O GO H t^O Ol ■> ro r^^ 00 ^> 0> M O 00 I M o* t>.QO ^^ -^ r^vo ■* I M o d O O O O O O Ji Q O » 3 vo m ( OOOOOOOcoa m m t^ ^> moo 00 co - I*- ^.co o t*.oo -ih . , \o N 00 « ^o o\ M I O N ^ !-• VS 01^3 O »0 « O N 'o o^ ^« m o^ O ( I ■^^ ^1- 10 m o\ moo m m^o « m I- H o in*o vovovo^ -^c^o Os ll-l>lMHHHf00 ro 5\o^ minioMO TOO ro"© m O m - -^ in o >noo M^ ej_ aaccadCBCfl m« N -^O « -^H ^s^OM T *CO O Oi'O 00 H H in\o m h«.vo moo ^00 »n ■* -^h m h m O^O^N ^r^^inmcTst^t^'O -^co t^ m o COC0COC000C0C0C0O303C00300CO03O ■ 00 m m o>oo o^ o\ rco w m f^ ci '' t^vo CJ d 00 eji N M (^\o m H CKJ. ■ Ov CT! Ol CTiCO 00 00 00 fs t>. e«. t%VD \£ u m o »n o m o > Q, N iO tN d M in !>. d « I E^ U M M b4 N » r TO mo mo 4J m o m o m O » d t^mN Q t-mis O c^m« o f^wiw CJ 0\ o^ Of Otoo 00 CO og t^ ts t>. t^« — " m « O 1 M m N ; V m m ^vo m r^>o co ts o^oo 4^2 MA TERIALS OF CONSTRUCTION— NON-FERROV S METALS. si •Xjppspp JO sntnpoi^ | •(spunod-;ooj) aDusinso^ 21 •3jqg jouaixa JO U0ISU9)X^ '(S39x33p) UOIS -jo; jo aiSuy •(ranrajireai JO 'juao 'j3d) •(spunod-iooj) ■(•inao J3d) noissajdtuoD jo lunoray •(sponod) ijDui ajBnbs jad q;Su3j)s .3ojqsiu3 •(uopoas iBuiSuo JO 'juao jad) uohoss painp^jj JO jB^amgiq •('inaD jad) iiopcSuo[a l^^ox o CO vg m oio iovg o^oo CO o m oivo oooo'4-Novor>. \ocn < O o n o^ -^co lo r- ooooooooooooooooooo'oopoo'q' I ^» rOM M f-OiOin 00000000000 CO fOlOCO ► O r>. W W « lO M fT) (r)\0 "^00 00 On»0 O fOO^^MOO O M OvVO O to tx"* o O OOOOOOOOCst* 88888888g!8 (*i o fo CO voo *£ o' d o' o' o' o' o* o' o' o' ■(pEO[ SuT3|cajq jo ■^OOMOOOOOOOOOOOOOOOOOOOOOOO «5 S I " ^3 o •(spunod) uoij -oas pajn^oHj^ •(spunod) noi; -oas j-BaiSijo »o ^VO VO Oi o o t> ^.>o 00 p^ m ro to ^ ~ .. — .- r«. f^ t^VO N lO 1^ I^ U ^l C^tCrn"flv■*d^f0^oe I«»'«J-'*QmO0OO Osl^MOO MO O »00 m tv ^'^ "5 * *i r2 ''^ cT M^"o" f o cd" h" »n 8,y AO ovo "troifotnci >tN.^"«-OM QOOroO 0"lOO lOvt^Moo no 5 loo M m mvo o o • "2 r;; "^^o^ cj o q_ ro in '*■ "^-^o o tt ■* ) M « "*VO w 00 ' t^ T^vo ■* N -a- H • (spunod- JOOJ) saqDUT \Z JO uopoag -ap H uiqjTM. aauaiiisa^ Y X>. Tj-00 VO VO « 'S. fO Oi N lO I HtO N ■* 0\i- •(saqauT) SuisiBajq ojojaq Do'poayap ibiox *i +; t; t«. Tf o\oo mmfO lo■^^^ofo■^^mc^ ■*« m dvod M ■* b\>b o w o M miotvTf^M ■ OhOOOOOOOOOOOOOOOOOOO ■(PBOI SU13(^J^ JO •^uao jad) ?iu3i|DiisBi3 t moo O^OHOOO0OO0OOOOOOOOt» t^oo • ■ajnjdtu JO sninpoj\[ ■ '*'niop^o\M H M (Nfo lovo o\ m ^^ w O "o mvD CO r* pvvo « o^o^o^tvo^>np^ tv.^ ^-^ (Ti * a^ ^ M ^*5 ^> p».co oo « moo oo m ov o t«« O i 'AXIAVaO DIdlDSdS Nvaw mrniHMvo '^inN "**o M vp »n ^v^o oo m moo n h oo o r* ^oooo>M •oo'coo tN>B w\5 r^M « m n\o tnmrooco tj mWrO«N -OOOlOs OnOO r^"0 "O^-vnmnwmhoh CO 00-00 0000 cooo ^«.p^^^^»•t^^N^sC>.^^^^^«.^.^«.ls,^-.^^^■ §E^ d-r^oo n N moo m r" 1CO«MMOP^»nPI 0^ mvO VO 0\\0 vOO\O«00-d-0tNWNHOvC -n-woovo o\m(nm mvo *£> os w t^ oico oo ■* •- ^ m 5 K TO mo "TO mo mo momommowo momo* ■s- -«• -^ m m w 3< 7°S = 2,812 kilogs. per sq. cm. For brass (copper and zinc) the tenacity may be taken as T^ = ^0,000 + 500 z ; where the zinc is not above 50 per cent.; and 7;' = 2,109 + 35-15 •^• Thus copper 70, zinc 30, should have a strength of 30,000 + (500 X 30) = 4S,000 pounds per square inch, or 2,109 + (35-15 ^ 30) ~ 3>i65 kilogrammes per square centimetre. Referring once more to Figures 80 and 81, it is seen that a line of maximum elevation crosses the field marking the crest of the mountain in Fig. 81, of which the "maximum bronze" is the peak. This line of valuable alloys may be practically covered by the formula; M = s + 3t = Constant = 55, in which s is the percentage of zinc, and i that of tin. Thus a maximum is found at about i = 0, 3= 55, while the other end of the line is ^^ = o, ^,= 18. STRENGTH OF KALCHOIDS. 47/ Along this line the strength of any alloy should be at least T„ = 40,000 + 500 z. TJ:= 2,812 + 35.15 2. Thus the alloy ^ = i, / == 18 will also contain copper = 100 — 19 = 81, and this alloy Cu, 81 ; Zn, i ; Sn, 18, should have a tenacity of at least T„ = 40,000 + (500 X i) = 40,500 lbs. per sq. in. T^„ = 2,812 + (35.15 X i) = 2,847 kilogs. per sq. cm. The alloy Cu, 60; Zn, 5 ; Sn, 16, should have at least the strength T„ = 40,000 + (500 X 5) = 42,500 lbs. per sq. in. T'„ = 2,812 + (35.15 X 5) = 2,988 kilogs. per sq. cm. while the alloy Zn, 50 ; Sn, 2 ; Cu, 48, should give, as a mini- mum per specification : T„ = 40,000 + (500 X 50) =: 65,000 lbs. per sq. in. r'„ = 2,812 + (35.15 X 50) = 4,570 kilogs. per sq. cm. These are rough working formulas that, while often de- parted from in fact, and while purely empirical, may prove of some value in framing specifications. The formula for the value of T„ fails with alloys containing less than \ per cent, tin, as the strength then rapidly falls to ^ = o. The table which follows will present, in convenient form, probably fair minimum values to be expected when good foundry work can be relied upon, and may ordinarily be used in specifications with the expectation that a good brass-founder will be able to guarantee them. 478 MA TERIALS OF CONSTRUCTION— NON-FERROUS METALS MINIMUM TENACITY OF AI.LOYS. ALLOY. TENACITY.— Probable Minimum. Cu. Zn. Sn. Lbs. per sq. in. Kgs. per sq. cm. ICO 30,000 2,109 95 5 32,500 2,285 90 10 35,000 2,460 85 15 37,500 2,636 90 10 40,000 2,8ia 95 5 35,000 2,460 97i 2i 32,500 2,28s 90 5 5 37,500 2,636 85 10 5 40,000 2,812 75 20 5 45,000 3,163 68 30 2 47,000 3.304 64 35 I 48,500 3,410 60 40 50,000 3,515 Ductility. — The ductility of these alloys is a subject of as much interest to the engineer as their strength ; and in this quality the ternary alloys are as variable as in every other. Referring again to the map, Figure 80, it is seen that a closely grouped set of slightly curved and slowly converging lines cross it from tin = 25, to zinc = SS. the mean line having an equation nearly 2.2t + ^r = 55. Along this line the alloys have immense tenacity, as exhibited by the fact that some of them, if not nearly all, are too hard to be cut by steel tools, and in shaping them only grinding tools — either the emery wheel or the grindstone — could be used, and even then with most unsatisfactory results. Yet such was the brittleness of these metals that no reliable test of their strength could be obtained. The strain-diagrams obtained were straight, and nearly vertical lines, terminating suddenly, when the piece snapped, without indication of approach to an elastic limif. They were perfectly elastic up to the point of fracture, but were so destitute of resilience that no use can probably be made of them by the engineer. Their brittleness was such that they would often break in the mould by contraction in STRENGTH OF KALCHOIDS. 479 cooling, although cast in a straight bar. In some cases they crack by the heat of the hand, and were broken at one end by the jar transmitted from a light blow struck at the other end.* The border line of this valueless territory is shown Fig. 83.— Ducthity of Copper-Tin-Zinc Alloys. ft ^^. x &>'''- ■' ■"• •' '■ ■••■ ■•' '•■ -'^ ■•■ ■■■ j/iiii-" on the map by a slightly curved dotted line to which a line having the equation 2.5/ + ^ = 55 is nearly tangent. The alloys lying along this line have nearly equal ductility, ex- tending, according to measurements obtained by the auto- graphic machine, about .03 of one per cent^ * Report to U. S. Board. 48o MA TERIALS OF CONSTRUCTION— NON-FERROUS METALS. Above this line is another having nearly the equation 4? + ^ = 50, which last line is that of equal ductiHty for alloys exhibiting extensions of 3 per cent. Still nearer the "pure copper corner" is a line fairly representing alloys containing about 3^^ + .ar = 48, and along which the extensions were 7.3 per cent., and another such line extending from standard gun- metal compositions on the one side to Muntz metal on the other — Cu 90, Sn 10, to Cu 55, Zn 45, — of which the equa- tion is nearly 4.5/ + ^^ = 45, represents alloys averaging an extension of 17 per cent. These lines are best seen on the sheet of extensions, Fig. 83. All alloys lying above the line taken here as a boundary line give figures for tenacity that ex- ceed 30,000 pounds per square inch (2,109 kilogs. per sq. cm.). The addition of tin and of zinc to cast copper thus in- creases tenacity at least up to a limit marked by the line 3^ + 2'= 55 ; but the influence of tin is nearly twice as great as that of zinc, and the limit of useful effect is not reached in the latter case until the amount added becomes very much greater than with the former class — the copper-tin alloys. Brasses can be obtained which are stronger than any bronzes, and the ductility of the working compositions of the former class generally greatly exceeds that of the latter. Ternary alloys may be made containing about i,t -^ z = 50, which ex- ceed in strength any of the binary alloys, and compositions approaching copper, 55; tin, 2; zinc, 43; may be made, of extraordinary value for purposes demanding great strength, combined with the peculiar advantages offered by brass or bronze. The addition of one-half per cent, tin to Muntz metal confers vastly increased strength. The range of useful introduction of tin is thus very much more restricted than that of zinc; alloys containing 12 to 15 per cent, tin are so hard and brittle as to but rarely find ap- plication in the arts, while brass containing 40 per cent, zinc, is the toughest and most generally useful of all the copper- zinc "mixtures." The moduli of elasticity of these alloys are remarkably uniform, more than one-half of all those here described ranging closely up to fourteen millions, or one-half that of well-made steel-wire. The moduli gradually and STRENGTH OF KALCHOIDS. 48 1 slowly increase from the beginning of the test to the elastic limit. The Fracture of these Alloys is always illustrative of their special characteristics. Those broken by torsion in the autographic testing machine were, if brittle, all more or less conoidal at the break ; ductile alloys yield by shearing in a plane at right angles to the axis of the test piece ; the for- mer resemble cast iron and the latter have the fracture of wrought iron. Every shade of gradation in this respect is exhibited by an observable modification of the surface of fracture varying from that characteristic of extreme rigidity and brittleness, through an interesting variety of intermediate and compound forms to that seen in fracture of the most ductile metals. Possibilities of Improvement.— The tenacities and ductilities given are within the best attainable figures where they relate to the most valuable working bronzes and brasses. These figures represent the result of ordinary founders' work ; and metals rich in copper, m'ade with no greater precaution against oxidation and liquation than is usual in brass foun- dries, may be vastly improved by special treatment sug- gested, by using pure ingot metals, fluxing carefully, as with phosphorus or manganese, casting in chills, rapid cool- ing, and finally rolling, or otherwise compressing, either hot or cold. Unannealed copper wire is reported by Baudrimont* as having a tenacity of about 45,000 pounds per square inch (3,163 kilogs. per sq. cm.), and Kirkaldy reports 28.2 tons per square inch (63,168 pounds per square inch, 4,440 kilogs. per square cm.), the wires having diameters of 0.0177 ^'id 0.064 inches (0.044 ^'^^ 0.165 cm.) respectively. A way should be found to secure equal purity, homo- geneousness, and density in cast copper, and such metal should then possess tenacity and toughness equal to that of rolled metal. Gun-bronze, which ordinarily has a tenacity of about 35,000 pounds per square inch (2,460 kilogs. per sq. cm.) has been made at the Washington Navy Yard. * Annales de Chimie, 1850. 31 482 MA TERIALS OF CONSTRUCTION— NON-FERROUS METALS. Alloys to be hammered or rolled will be found more diffi- cult to work as the percentage of tin is increased, and the minutest addition of tin to the brasses usually rolled is found to sensibly decrease their manageability. The "Maximum Bronzes" form a group demand- ing special consideration as including a collection of generally unfamiliar but exceptionally valuable alloys. The work planned by the Author in the investigation of this part of the subject was left incomplete by the U. S. Board, but was continued, as opportunity offered, at intervals up to the present time.* The position and characteristics of the strongest possible alloys of the three metals constituting the "Kalchoids" having been determined with a fair degree of accuracy, as al- ready described, the next step was to ascertain what modifi- cations might be produced in them by careful fluxing and the use of still more carefully prepared alloys. This later study was made in the years 1882-3, in the same manner as the earlier investigations for the U. S. Government, at the sug- gestion and under the supervision of the Author, by Mr. W. E. H. Jobbins, whose report is here abridged. f The area chosen as the field of this investigation was a small triangular portion surrounding the peak of the moun- tain. Fig. 81, marked 65,000 on Fig. 80, as this area embraces all that portion of the field in which the most valuable alloys had been proven to be located. The data obtained gave ex- ceedingly high figures, the lowest average value of tenacity being above 50,000 pounds per square inch (3,515 kilogs. per sq. cm.). As this research extended over a very limited area, it was possible to conduct the investigation with much greater exactness than before, and thus settle the composition of the " strongest of the bronzes." The metals varied with differences of but one per cent.; 23 combinations were chosen ; 2 test-pieces were made of each^ * The U. S. Board was strangled by refusal of appropriations, leaving the •work in hand unfinished. Some of the work necessary to the presentation of the reports actually made. was, however, concluded by the Author, at some expense, in the Mechanical Laboratory of the Stevens Institute of Technology. \ " Investigation Locating the Strongest of the Bronzes," J. F, I., 18S4. STRENGTH OF KALCHOIDS. 483 composition, making 46 test-pieces. Usually, the data obtained from two specimens of the same composition agreed so closely that the average value was safely taken ; but, when there was a marked difference, the data agreeing more closely with the results anticipated from analogy were adopted, and the other value rejected as being probably erroneous. The copper em- ployed was from Lake Superior, the zinc from Bergen Port. In the use of tin, phosphorus was added to give soundness in these copper-tin and copper-tin-zinc alloys, which are so liable to be made seriously defective by the absorption of oxy- gen and the formation of oxide. It has been found possible to produce, on a commercial scale, an alloy of phosphorus and tin, which, while containing a maximum percentage, does not lose phosphorus when remelted. The best proportions for practical purposes are said to be tin 95 per cent, and phos- phorus 5 per cent. After careful study, the following limits of the field were decided upon : Copper, maximum 60, minimum 50 ; Zn, 48 and 38 ; Sn, 5 and o. These limits include the best alloys for purposes demanding toughness as well as strength. The compositions are given in the following table : BEST COPPER-TIN-ZINC ALLOYS, OR KALCHOIDS. NO. cu. ZN. SN. NO. cu. ZN. SN. NO. CU. ZN. SN. I 55 43 2 9 53 43 4 • 17 58 40 2 2 54 44 2 10 55 41 4 18 54 45 I 3 54 43 3 II 57 41 2 19 53 44 3 4 55 42 3 12 57 43 20 54 42 4 5 =;6 42 2 13 55 45 21 5(5 41 3 6 56 43 I 14 52 4b 2 23 57 42 I 7 5S 44 I 15 52 43 5 23 58 41 I 8 53 45 2 16 55 40 5 The castings were made much as in all the earher investi- gations, the same precaution being taken to prevent volatili- zation of zinc, and care was taken to secure rapid cooling to prevent liquation. All the compositions thus made were 484 MA TERIALS OF CONSTRUCTION— NON-FERROUS METALS. strong and usually tough ; all could be turned and worked safely, and all were evidently of commercial value for the pur- poses of the engineer. All test-pieces were sound, and even microscopic examination revealed no defects in structure. The investigation was made by the use of the Author's auto- graphic machine as permitting most rapid work and most ex- act determinations of quality and behavior, especially as to the latter near the elastic limit. The samples were all re- duced to the standard form and size. Results of Tests. — The formula used is M= wh 4- /; where «/ = moment necessary to deflect the pencil one inch ; h = height of the curve above the base line at 0^, f = friction in foot-pounds, and M is the total torsional moment. In this case, w = 96.93 foot-pounds, andy= 4.7S, ^ being measured on the strain-diagram of each test-piece. To obtain the required values of 7" the formula T= [300— )^(?J M,* in which M is known, and 6^ is measured directly from the autographic record; T is the calculated tenacity. The values of M, T, B, and i9„ the total moment, the approximate tenacity, and the angles of torsion at the elastic limit and at rupture, have been included in the following table : STRENGTH OF BEST COPPER-TIN-ZINC ALLOYS OR KALCHOIDS. ORIGINAL MARK. STRESS IN TORSION. FOOT-POUNDS. ' M. •» APPROXIMATE STRESS IN TENSION. FOOT-POUNPS. T. ANGLES. Ultimate. Average. Ultimate. Average. 9« Sr IXI ^ OB \ Z3 \ 270.208 251.922 178.321 208 . 400 251.922 219.935 243-392 258.319 261.065 193-369 235-929 250.851 77,309 72,301 53,946 59,810 75,576 65,980 73,017 74,912 74,805 56,653 70,778 73,965 1-5° I i.i 0.7 I I 2 2 43° 40 5.05 40 13-77 10 19.8 30.3 * This relation between torsional and tensional resistances was obtained bj experiment on the machine used in this investigation. Trans. Am. Soc, C E., no. clxiii., vol. vii., 1878, STRENGTH OF KALCHOIDS. 48S ORIGINAL MARK. G6 R8 Sg L 10 Z II DB M 13 U14 V15 N16 A 17 P18 Tig QBO HBI EBB B33 STRESS IN TORSION. FOOT-POUNDS. M. Ultimate. Average A 268.881 B 263.543 A 227.689 i! 220.612 A 286.847 li 250.855 A 194.634 B 184.331 A 222.853 B 230.597 A 249.014 B 252.881 A 260.645 B 237.382 A 227.689 B 241.259 A 227.689 B 208.303 A 163.715 B 177.185 A 189.886 B 227.689 A 225.750 B 253.198 A 227.689 B 250.952 A 254.829 B 260.645 A 231.566 B 196.671 A 229.628 B 258.707 A 283.908 B 229.628 A 305.233 B 221.773 A 225.750 B 175.247 266.212 224.151 268.851 189.488 226.725 250.948 249 . 014 234.474 217.996 170.450 208.788 239.974 238.771 259.737 214. 119 244.168 266.768 263.508 200.499 APPROXIMATE STRESS IN TENSION. FOOT-POUNDS. T. Ultimate. 75,824 75,109 64,208 63,193 80,910 70,741 58,390 55,299 66,853 69,179 74,704 75,864 74,269 63,964 61,020 61,762 64,208 57,908 49,"3 53,155 56,965 68,306 67,725 75,959 63,200 73,488 72,871 71,501 69,459 59,001 68,888 77,612 81,381 68,888 85.770 60,986 63,084 45,038 Average, 75,467 63,700 75,826 56,844 68,017 75,284 6g,ii6 61,390 61,058 51,139 62,636 71,842 68,344 72,186 64,230 73,250 75,135 73,378 54,061 4.6- 2 2.05 2 2 2 2 2.6g 1.5 1.79 2.1 2.8 2.4 1.9 2.3 1.6 2 l.I 2.3 2 2.6 2 1.6 1.6 1.4 1.8 1.6 1.8 2.2 1.4 1.6 1.8 2.9 2.4 2 2.5 I 6 1.2 55 46 53.3 42.1 54 53 9 I 73 73 5 6 3 3 39 35 95 131 52 65 4 7 4 5 3.3 6,8 54 43.2 43.4 54 4.8 6.4 7.2 38 56 76 63 128 The neck subjected to distortion is in all cases, one inch (2,54 cm,) long between shoulders and ^ inch (1.5875 cm.) in diameter. Experiments thus made proved, notwithstanding the prC' 486 MA TERIALS OF CONSTRUCTION— NON-FERROUS METALS. cautions taken in making these alloys, to be a matter of some difficulty to decide satisfactorily the relative positions of the alloys studied. Nos. 7 and 22 were the best alloys made. Fig. 84. — Strongest of Bronzes. Conclusions. The Strongest Bronzes. — The results obtained from this investigation are well exhibited in the ac- panying diagram, Fig. 84. It was concluded that the alloy numbered 22 was what the Author has called the " strongest of the bronzes," and that its composition (Cu, 57; Sn, i ; Zn, 42) should locate the peak seen in the model, Fig. 81, and on the map. Fig. 80. No. 5, however (Cu, 56 ; Sn, 2 ; Zn, 42), is likely to prove a more generally useful alloy in consequence STRENGTH OF KALCHOIDS. 487 of its greater ductility and resilience ; and alloys with a little less tin may often prove even better than that. The Author has called the compositions, copper, 58 to 54; tin, ^ to 2J^; zinc, 44 to 40, which may be considered as representative of a group having peculiar value to the engineer, the "maximum- bronzes." This cluster lies immediately around the peak seen on the model. Fig. 81, including the point of maximum alti- tude. The safest alloys under shock are those containing the smallest quantities of tin. The Conclusions reached after concluding the in- vestigations which have been described in the present chap- ter are confirmed by the fact that a number of single compo- sitions have been independently discovered by other experi- menters, accidentally or incidentally to special investigations, which have peculiarly high tenacity, all of which approximate more or less closely, in their proportions, to these " maximum " bronzes and strongest " Kalchoids." Thus, Mr. Farquharson, president of the Naval (British) Commission, proposed, in 1874, an alloy composed of 62 parts of copper, 37 parts of zinc, and one part of tin. This is the reglementary naval alloy. When cast in bars it has shown on test a resistance of 70,000 pounds per square inch (5,000 kilogs. per sq. cm.). It rolls and works well, can be hammered into sheets, and is fusible only above red heat. It may be used as a lining for engine-pumps. It is but slightly oxidiz- able, and is not sensibly attacked by sea water, as shown by experiments with it extending over a period of years. A slight loss of zinc during melting must be taken into account. The British naval bronze for screw-propellers, stern bearings, bow- castings, and similar work, is composed of copper, 87.65 ; tin, 8.32 ; zinc, 4.03, and is reported to have a tenacity of 15 tons per square inch (2,362 kilogs. per sq. cm.), and to average 13)^ tons (2,126 kilogs.) in good castings. Tobin's alloy, al- ready described, is one of the " maximum " bronzes, also, containing copper, 58.22; tin, 2.30; zinc, 39.48. Sterro-metal is always a brass of nearly the same proportions of copper and zinc, i.e., a Muntz metal, containing from a fraction of i per cent, to sometimes 2 per cent, of tin, as well as some iron. 488 MA TERIALS OF CONSTRUCTION— NON-FERROUS METALS The bronze used for journal bearings in the U. S, Navy contains copper, 88 ; tin, lO ; zinc, 2. The strongest U. S. copper-tin-zinc alloy is that discovered by Mr. Tobin and described by the Author in earlier articles, and, as has been stated, had a tenacity of 66,500 pounds per square inch of origi- nal section, and 71,378 per square inch of fractured area (4,575 and 5,019 kilogrammes per sq. cm.) at one end of the bar, which was, as usual, cast on end, and 2 per cent, more at the other. This, like the " maximum alloy," was capable of being forged or rolled at a low red heat or worked cold. Rolled hot, its tenacity was 79,000 pounds (S,553 kilogs. per sq. cm.), and when cold-rolled, 104,000 (7,311 kilogs.). It could be bent double either hot or cold, and was found to make excellent bolts and nuts. These and other compositions which have been occasion- ally introduced as having extraordinary strength and excep- tional value, all contain a small amount of tin, and invariably fall within the field mapped out as described in this chapter as that containing the kalchoids of maximum possible strength. The latter, the " maximum alloys," as the Author has called them, will probably be very generally, if not exclusively, used when alloys are required of peculiar strength. It will be found that they are capable of improvement by fluxing with phosphorus, by mechanical treatment, and by small doses of iron and other metals, as are the more familiar bronzes and brasses. CHAPTER XVI. TEMPERATURE AND TIME AS AFFECTING THE IRONS ; FLOW OF METALS; FATIGUE; WOHLER'S LAW; LAUNHARDT'S FORMULA. The Effect of Heat and of variations of tempera- ture upon the mechanical properties of metals has long been a subject of debate, and one which has not yet been satisfac- torily settled by experiment.* In general, it would appear that, in a perfectly homoge- neous material, entirely free from internal strain, change of temperature would produce an alteration of strength and of ductility which would both be the reverse in direction of the variation of temperature. The forces acting to produce mechanical changes being cohesive force, on the one hand, resisting external forces tending to produce distortion or rupture, while the force pro- duced by the energy of heat-motion conspiring with external force to produce that distortion, and the molecules being at every instant in equilibrium between the force of cohesion on the one side, and the sum of the other two forces mentioned on the other, variations of form must ensue with every change in the relative magnitudes of thesa forces. A change of temperature produced by an increment of heat energy produces a reduction of cohesion by separation of particles, and the opposite change must cause an increase of cohesion by their approximation. Increase of temperature, by re^ ducing the range of action of cohesion, by separating the par- ticles and causing them to approach the limit of reach of co- hesive force, reduces ductilit)'-, and the opposite change of temperature increases extensibility. * This chapter contains some new matter and some extracted from a paper read by the Author before the Am. Society of Civil Engineers, 1874, and from a paper " On Molecular Changes in Iron," published in the Iron Age, 1873. 490 MATERIALS OF CONSTRUCTION— IRON AND STEEL. Mr. Jouraffsky has recently experimented upon rails cold (- i6° to - 21° C, + 3° to - 6° R), and warm. The total of the three hardening elements are expressed in phosphorus units thus : For rails which stood the test, 19 units; for rails which broke under the same test, 31 units. In the first, the units vary from 16 to 22 (in one case only 25 being reached), and in the second, the difference was from 22 (and that only in two cases, all the others being higher) to 45 units. Reports from Russian railways have shown that "]"] per cent, of the breakage of rails and tires occurs on those roads at a temperature below 32° Fahr. (0° Cent.). The variation of strength of iron with rising temperature has been studied by KoUmann, using metal of the following composition : WELD-IRON'. WEI.D-IRON. INGOT IRON. c O.IO 0.09 0.34 0.03 0.07 0.07 99-30 0.12 O.II 0.20 trace 0.14 0.06 99.36 0.23 Si 0.30 P 0.09 0.05 0.86 s Mn Cu 0.07 98.40 Fe These metals were rolled at temperatures varying from a maximum of 2,41 7^° Fahr. (1,325° Cent.) at the roughing rolls, to 1,112" Fahr. ($00° Cent.) as a minimum at the last pass. The results of the test of finished iron at various tempera- tures were the following, and the steel nearly the same in method of variation, although about 50 per cent, higher in tenacity at each point : (Cent 0° 200° 300° 400° 600° 1,000° Temperature, I p^j^^ ^^o ^^,0 ^^^. ^^^. j_jj2° 1,832° Lbs', per sq. in 53.30O 52,600 47,900 38,950 9,860 112 Tenacity, kgs. per sq. cm. 3,750 3,570 3,370 2,734 69° 8 The elastic limits decrease in nearly the same proportion. The extensions increase slowly up to the red heat. CONDITIONS AFFECTING STRENGTH. 491 eoo 1600 1100 C, ieS2 1332 p. Fig. 85. — Heat vs. Tenacity. The diagram above* (Fig. 85) graphically represents the results of several series of experiments, some of which have not been described. It exhibits the general accordance of all later investigations with that of the Franklin Institute of 1835. Curves Nos. i and 2 represent Kollmann's experiments on iron, and 3 on Bessemer " steel." No. i is ordinary, and 2 steely puddled iron. Curve No. 4 represents the work of the Franklin Institute on wrought iron. Curve No. 5 gives Fairbairn's results, working on English wrought irons. Nos. 6 to 1 1 are Styfle's, and represent the experiments made by him on Swedish iron. The numbers do not appear, as these results do not fall into curves ; these results are indicated by circles, each group being identified by the peculiar filling of .the circles, as one set by a line cross, * Eisen und Stahl, A. Martens ; Zeitschrift des Verdns Deutscher Itt- genieure ; Feb. 1883, p. 127. 492 MATERIALS OP CONSTRUCTION— IRON AND STEEL. ing the centre, another by one across, a third by a full circle, etc. The broken lines, 12 and 13, are British Admiralty experi- ments on blacksmiths' irons, and No. 14 on Siemens steel. The first five series, only, 'are of value as indicating any law ; and they exhibit plainly the general tendency already referred to, to a decrease of tenacity with increase of tem- perature. Fairbairn's experiments. No. 5, best exhibit the maximum, first noted by the Committee of the Franklin Institute, at a temperature between that of boiling water and the red heat. It will be observed that the measure of tenacity, at the left, is obtained by making the maximum of Kollmann unity. It will also be noted that Kollmann does not find a maximum as in curves 4 and 5, but, on the contrary, a more rapid re- duction in strength at that temperature than beyond. It would seem, therefore, that that peculiar phenomenon must be due to some accidental quality of the iron. The Author has attributed it to the existence in the iron, before test, of internal stresses which were relieved by flow as the metal was heated, disappearing at a temperature of 300° or 400° Fahr. (149° to 204° Cent.). A singular modification of set by change of temperature was noted while testing springs in the mechanical laboratory of the Stevens Institute of Technology ; * tested at 32° Fahr. (o°Cent.) the set of a coil of ^ inch (1.59 centimetres) wire, twenty inches (50.8 centimetres) long, after a compression of 3 J^ inches (8.9 centimetres) under a load of 5,000 pounds (2,268 kilogrammes) was 0.188 inch (0.48 centimetre), while at 212° Fahr. (100° Cent.) it was but 0.016 inch (0.04 centimetre). The experiments of Mr. Oliver Williams,f in determining the change produced in the character of the fracture of iron by transverse strain, at extreme temperatures, indicate loss of ductility at low temperatures. Two specimens of nut iron, from different bars, made at * Van Nostrand's Mag., 1878, p. 528. De Bonneville, t The Iron Age, New York, March 13, 1873, p. 16, CONDITIONS AFFECTING STRENGTH. 493 Fig. 86. — Fracture at Ordinary Temperature. Catasauqua, Pennsylvania, were first nicked with a cleft on one side only, and then broken under a ham- mer, at a temperature of about 20° Fahr. (— 7° Cent.). At this temper- ature, both specimens broke off short, showing a clearly defined gran- ular, or steely iron fract- ure. The pieces were then gradually heated to about 75° Fahr. (24° Cent.), and then broken as before, developing a fine, clear, fibrous grain. The two fractures were but four inches (10.16 centimetres) apart, and are entirely different. The accom- panying illustrations, from the Author's collection, exhibit this case. It has been long known that a granular fracture may be produced by a shock, in iron which appears fibrous when grad- ually torn apart. This was fully proven by Kirkaldy.* Mr. Williams was, probably, the first to make the experiment just described, and thus to make a direct comparison of the characteristics of fracture in the same iron at different temperatures. Valton has found t that some iron becomes brittle at temperatures of 572° or 752° Fahr. (300° to 400° Cent.), iind regains ductility and toughness at higher temperatures. On the whole, the fracture of iron at low tempera- tures has been found to be charac- FiG. 87.— Fracture at Low Temperature. ' * Experiments on Iron and Steel. f Bulletin Iron arid Steel Assoc, Feb. 1877 494 MATERIALS OF CONSTRUCTION— IRON AND STEEL. teristic of a brittle material, while, at higher temperatures, it exhibits the appearance peculiar to ductile and somewhat viscous substances. The metal breaks, in the first case, with slight permanent set, and a short, granular fracture, and in the latter with, frequently, a considerable set, and the form of fracture indicating great ductility. The variation in the behavior of iron, as it approaches the welding heat, illus- trates the latter condition in the most complete manner. The effect of alteration of temperature upon cast iron has been less studied than its influence on the malleable metals. The few experiments made by the Author indicate greater susceptibility to the influence of heat than is observed in either wrought iron or steel. The accompanying diagram exhibits this comparison, as made by the use of the Auto- graphic Testing Machine.* In these experiments the testing machine was placed in the open air in mid-winter, and exposed with test piece under stress to temperatures falling as low as — io° Fahr. (— 23° Cent.), and, again, taken indoors, and the tests continued at temperatures rising to te^'^r.a + 70° Fahr. (+21^ ^ 20,000 n .\ T it. J ■ Cent.). In the dia- gram, the horizontal scale at the top is a scale of temperatures, several points, as 70°, 25°, 18°, 10°, being marked ; the vertical scale is one of tenac- ities. The several numbers attached at the extremities of the curves are those of the specimens tested. It will be noted that Temperature 00,000 «),(»0 Fig. 88. — Effect of Cold. * Trans. Amur, Soc, C. E., 1874 ; Jour, Franklin Inst., 1S74, Vol, LXVII., PI. III. CONDITIONS AFFECTING STRENGTH. 495 cast iron shows greater loss of tenacity, at the higher tem- peratures than other metals, except a single piece of double shear steel. Swedish iron seems almost unaffected, and cast copper is but slightly weakened. The effect of change of temperature is invariably, so far as observed, to produce a change of tenacity in the opposite direction, rise in temperature being accompanied by a decrease of strength and vice versa. Valton found that a steel rod bent very welt at a tem- perature a little below dull red, but broke at a temperature which may be called blue, the fracture showing that color; Portions of the rod which were below this temperature mani- fested much toughness, and bent without fracture. Charcoal pig iron from Tagilsk, made in 1770, irons obtained from the Ural in rods and sheets, soft Bessemer and Martin steels from Terrenoire, soft English steel and good English mer- chant bars, all gave the same results, whether the metal tested had been hammered or rolled. Valton found that the phe- nomenon had been long known to the workmen under his direction. In working sheet iron with the hammer, they wait until the metal had cooled further when approaching the temperature which would give the blue fracture when broken. Fig. 89. — Increase of Volume, o" up to 2,776° F., (1,525" C). He concludes that wrought iron, as well as some kinds of soft steel, even when of excellent quality, are very brittle at 49^ MATERIALS OF CONSTRUCTION— IRON AND STEEL. a temperature a little below dull red heat — 577° to 752' Fahr. (between 300° and 400° Cent.). The variation of strength follows quite closely the change of density, which latter is illustrated in the preceding diagram, which exhibits increase of volume from the freezing point. The sudden fall of the line before reaching the melting point indicates the sudden increase of volume which castings exhibit while cooling, and which enables " sharp " castings to be secured. It is at the crest noted near this point that viscosity is observed. From this point back to the freezing point the variation follows a regular law. Conclusions as to Effect of Change of Tempera- ture. — It would thus seem that the general effect of increase or decrease of temperature is, with solid bodies, to decrease or increase their power of resistance to rupture, or to change of form, and their capability of sustaining " dead " loads ; and we may conclude : (i.) That the general effect of change of temperature is to produce change of ductility, and, consequently, change of resilience, or power of resisting shocks and of carrying " live loads." This change is usually opposite in direction and greater in degree at ordinary temperatures than the variation simultaneously occurring in tenacity. (2.) That marked exceptions to this general law have been noted, but that it seems invariably the fact that, wherever an exception is observed in the influence upon tenacity, an ex- ception may also ^e detected in the effect upon resilience. Causes which produce increase of strength seem also to pro- duce a simultaneous decrease of ductility, and vice versa. (3.) That experiments upon copper, so far as they have been carried, indicate that (as to tenacity) the general law holds good with that metal. (4.) That iron exhibits marked deviations from the law between ordinary temperatures and a point somewhere be- tween 500° and 600° Fahr. (260° and 316° Cent.), the strength increasing between these limits to the extent of about 15 per cent, with good iron. The variation becomes more marked and the results more irregular, as the metal is more impure. CONDITIONS AFFECTING STRENGTH. 497 Experiments upon the steel castings of which analyses have been already given illustrate well the importance of an- nealing such metals as are liable to internal stresses, thus : EFFECT OF ANNEALING STEEL CASTINGS. ULTIMATE STRENGTH. ELONGATION. TREATMENT. Lbs. per square inch. Kilogrammes per square centi- metre. Per cent. 89,289 104,362 6,177 7,337 4 8 Annealed 71,904 81,984 5,055 5,763. 4.16 Not annealed 53,782 63,616 3,781 4,572 I 13 99,496 98,560 6,995 6,929 2 12 71,944 107,744 5,25s 7,574 1.65 7.2 67,200 67,296 4,724 4,731 13-3 27.5 The Author has found, during an investigation extending from some time in 1881 to date, that annealing renders iron wire subject to a gradual yielding under permanent loads much less than those determined by ordinary test. The dif- ference in this respect between the annealed and the hard, unannealed wire from the draw-plate was remarkably great. Professor Abel concludes, after a careful investigation, that the condition of the carbon in steel determines its condi- tion as to hardness ; that in annealed steel, the carbon exists 32 498 MATERIALS OF CONSTRUCTION— IRON AND STEEL. as a definite carbide of iron, which is found in smaller pro- portion as steel is hardened. Professor Hughes, studying the magnetic condition of the steel, concludes it to be an alloy of carbon and iron, the carbide described by Abel being broken up, on hardening, to form this alloy.* Effect of Age and Exposure.t — There are many phenomena which cannot be conveniently exhibited by strain diagrams ; such are the molecular changes which occupy long periods of time. These phenomena, which consist in altera- tions of chemical constitution and molecular changes of structure, are not less important to the mechanic and the en- gineer than those already described. Requiring usually, a considerable period of time for their production, they rarely attract attention, and it is only when the metal is finally in- spected, after accidental or intentionally produced fracture, that these effects become observable. The first change to be referred to is that gradual and im- perceptible one which, occupying months and years, and under the ordinary influence of the weather going on slowly but surely, results finally in important modification of the proportions of the chemical elements present, and in a con- sequent equally considerable change of the mechanical proper- ties of the metal. Exposure to the weather, while producing oxidation, has another important effect : It sometimes produces an actual improvement in the character of the metal. Old tools, whigh have been laid aside or lost for a long time, acquire exceptional excellence of quality. Razors which have lost their keenness and their temper recover when given time and opportunity to recuperate. A spring regains its tension when allowed to rest. Farmers leave their scythes exposed to the weather, sometimes from one season to an- other, and find their quality improved by it. Boiler makejs frequently search old boilers carefully, when reopened for repairs after a long period of service, to find any tools that * Trans. Inst. Mechanical Engineers of G. B., 18S3. f Journal of the Franklin Institute, June, 1875. Scientific Anietican, Marcli 27 ; 1875.— R. H. Thurston. CONDITIONS AFFECTING STRENGTH. 499 may have been left in them when last repaired ; which, if found, are almost invariably of improved quality. The Au- thor, when a boy, amusing himself in the shop, if denied the use of their tools by the workmen, looked about the scrap- heaps and under the windows for tools purposely or carelessly dropped by the workmen ; and when one was found badly rusted by long exposure, it usually proved to be the best of steel. A most striking illustration of this improvement of the quality of wrought iron with time has come to the knowl- edge of the Author. The first wrought-iron T-rails were de- signed by Robert L. Stevens^ about the year 1830, and were soon afterward laid down on the Camden and Amboy Rail- road. These, when put down, were considered, and actually were, brittle and poor iron. Many years later, some still re- mained on sidings. Son^e of these rails v/ere taken up and re-rolled into bar iron. The long period of exposure had so greatly changed the character of the metal that the effect was unmistakable. There are probably, as the Author has concluded, two methods of improvement, each due to an independent molec- ular action. In the case of the razor and the spring, which regain their tempers when permitted to rest, a molecular re- arrangement of particles, disturbed by change of temperature in one case and by alternate flexing and relaxing in the other, probably goes on, much as the elevation of the elastic limit and the increase of resisting power, discovered by the Author and shown on the strain diagram, takes place under strain and set, The other cases may probably be due to a combination of this physical change with another purely chemical action, which is illustrated best in the manufacture of steel by the cementation process. Here the element carbon enters the solid masses of iron, and diffuses itself with greater or less uniformity throughout their volume. There seems to exist a tendency to uniform distribution which is also seen in other chemical changes. Many chemical processes are accelerated, checked, and even reversed by simple changes of relative proportions of elements. When, therefore, wrought iron containing injurious ele- SOO MATERIALS OF CONSTRUCTION— IRON AND STEEL. merits capable of oxidation, is exposed to the weather, the surface may be relieved by the combination of these elements with oxygen, and the surcharged interior, by this tendency to uniform diffusion, is relieved by the flow of a portion to the surface, there to be oxidized and removed. This process goes on until the metal, after lapse of years, becomes com- paratively pure. Meantime, the occurrence of jarring and tremor, such as rails are subjected to, may accelerate both this, and the previously described change. Crystallization. — The effect of strains frequently ap- plied, during long intervals of time, is quite different, how- ever, where they are so great as to exceed the elastic range Fig. go. — Fractured Surface of Connecting Rod. of the material. The effect of stresses which strain the metal beyond the elastic limit has already been referred to. CONDITIONS AFFECTING STRENGTH. 501 A still more marked case has come to the notice of the writer. The great testing machine at the Washington Navy Yard has a capacity of about 300 tons, and has been in use 35 years. Quite recently, Commander Beardslee subjected it to a stress of 288,000 lbs. (130,000 kilogrammes), which stress had frequently been approached before ; but it subsequently broke down under about 100 tons. The connecting bar which gave way had a diameter of five inches, and should have origi- nally had a strength of about 400 tons (406,400 kilogrammes). Examining it after rupture, the fractured section, Fig. 90, was found to exhibit strata of varying thickness, each having a characteristic form of break. Some were quite granular in appearance, but the larger proportion were distinctly crystal- line. Some of these crystals are large and well defined. The laminae, or strata, preserve their characteristic peculiarities, whether of granulation or of crystallization, lying parallel to their axis and extending from the point of original fracture to a section about a foot distant, where the bar was broken a second time (and purposely). Fig. 91, under a steam ham- FiG. 91.— Fractured Surface of Connecting Rod, men It thus differs from the granular structure which dis- tinguishes the surfaces of a fracture suddenly produced by a single shock and which is so generally confounded with real 502 MATERIALS OF CONSTRUCTION— IRON AND STEEL. crystallization. This remarkable specimen has been contrib uted by the Navy Department to the cabinet of the Author. In a discussion which took place many years ago before the British Institution of Civil Engineers, Mr. J. E. McCon- nell produced a specimen of an axle which he thought fur- nished nearly incontestable evidence of crystallization. One portion of this axle was clearly of fibrous iron, but the other end broke off as short as glass. The axle was hammered under a steam hammer, then heated again and allowed to cool, after which it was found necessary to cut it almost half through and hammer it for a long time before it could be broken. While forging a large shaft for a sea-going steamer at the Morgan Iron Works, in New York, some years ago, a " porter- bar " was employed which had been in the works and in fre- quent use for many years. This bar was about 20 feet (6.1 metres) long, 12 inches (29.5 centimetres) in diameter at the ^r'' bt= 'KM.:. ■ .'S* Fig. 92. — Fractured Surface of Link of Testing Machine. small end, and 23 inches (58.4 centimetres) at the end welded to the forging. The whole mass was slung from the crane in CONDITIONS AFFECTING STRENGTH. 503 the usual way. While the shaft was under the hammer, the jar detached the end of the porter-bar on the free side of the sling, breaking it where it was about ij inches (26.7 centime- tres) in diameter, and entirely unstrained by the load,' and detaching a piece weighing a ton or more. The load which would have been calculated as the breaking load hung upon the extreme end would have been about 14 tons (or tonnes). The fracture was partly granular, but largely crystalline. One crystal had faces a half inch (1.27 centimetres) square. A student at the Stevens' Institute of Technology, while annealing a number of steel hammer heads, left them ex- posed all night to the high temperature of the air-furnace in the brass foundry. When finishing one of them, a careless blow broke it, and the fractured surface was found to possess a distinctly crystalline character. In the illustration, Fig. 93 is a magnified representation of the surface of fracture. The two holes shown, penetrating the mass, are those drilled in the first opera- tion, preparatory to fit- ting the handle. The facets of the crystals are seen to be remark- ably perfect and well defined. Fig. 94 rep- resents the hammer on very nearly the natural scale. In this example, however, the faces were nearly all pentagonal, and were usually very perfectly formed. When imperfect crys- tals are developed, it is easy to mistake them, but the formation of pentagonal dodecahe- dra, in large numbers and in perfectly accurate forms, may Fig. 93. — Hammer-Head, Magnified. Fig. 94. — Hammer-Head, Natural Size. 504 MATERIALS OF CONSTRUCTION— IRON AND STEEL. be considered unmistakable evidence of the fact that iron may crystallize in the cubic, or a modified system. This may apparently take place, according to some authorities, either by very long continued jarring of the particles beyond their elastic limits, or under the action of high temperature, by either mechanical or physical tremor. But no evidence is given here that a single suddenly applied force, producing fracture, may cause such a systematic and complete rearrange- ment of molecules. The granular fracture produced by sud- den breaking, and the crystalline structure produced as above during long periods of time, are, apparently, as distinct in nature as they are in their causes. But simple tremor, where no sets of particles are separated so far as to exceed the elastic range, and to pass beyond the limit of elasticity, does not seem to produce this effect.* In fact, some of the most striking illustrations of the im- provement in the quality of wrought iron with time have occurred where severe jarring and tremor were common. As one example, the case of the wrought-iron T-rails, laid down on the Camden and Amboy Railroad in 1832, which have been already referred to, may be taken. Here the metal has been subjected for many years to the strains and tremor accompanying the passage' of trains with- out apparent tendency to crystallization, and with evident improvement in its quality. Such crystallization as that last described has often been observed. Wohle* found cubic crystals in cast-iron plates which had been for some time kept at nearly the tempera- ture of fusion in a furnace, and Augustine found similar crys- tals in gun-barrels; Percy found octahedra of considerable size in a bar which had been used in the melting pot of a glass furnace. Fairbairn asserts the occasional occurrence of such change due to shock, jar, and long-continued vibration. Miller found cubic crystallization plainly exhibited in Besse- mer iron, which may, however, have been due to the presence of manganese. Hill shows that heat may produce such * The Author finds such effect sometimes to follow intermitted strain. CONDITIONS AFFECTING STRENGTH, S^S changes in the process of nianufacturing large forgings, and denies the occurrence of true crystallization in cold iron. This can only be settled by further investigation. Dr. Sorby has examined the structure of wrought iron and of steel with the microscope, and found the hammered bloom to be a mixture of crystals of iron and portions of slag. The rolled bar, Fig. 97, contained crystals also, but they were fresh crystals formed on the cooling of the bar, and it was apparent that the fibre seen at the fracture was produced during rupture, and was not a characteristic of the unaltered iron. The cementing process of steel making was found to develop a network of flat crystals of a hard carbide of iron. Cast steel, Fig. 96, was found to contain larger crystals, of a different form, which were reduced in size by subsequent working. Meteoric iron, Fig. 95, was found to exhibit the characteristics of an iron crystallized by long exposure to a temperature beneath that of fusion. Dr. Sorby's paper has been published with illustrations, which are reproduced in the accompanying engravings.* The samples in question comprised armor plates, meteoric iron, cast iron and cast steel, and, as an inspection will show, ex- hibit a greatly varying structure. The specimen of cast steel is of very uniform structure, with no lines of weakness, while, an inspection of the specimen of cast iron will reveal a num- ber of plates of graphite, that naturally tend to diminish the strength of the metal. The armor plate, on the other hand, shows varying crystals and lines of welding, while the sample of meteoric iron shows a structure altogether unlike that of any artificial iron. Martens has used the microscope in the examination of various grades of iron and of spiegeleisen,t with somewhat in- conclusive results. But he finds that certain peculiarities and characteristics, due especially to the various mechanical oper- ations which the material undergoes, either during the proc- ess of manufacture, or molecular changes to the manner in * J. C. Bayles, in Iron Age and Mechanics, Mch., 1883 ; Trans. Am. Inst. M. E., 1880. f Verein zur BefSrderung des Gewerbfleisses, 1882. 5o6 MATERIALS OF CONSTRUCTION-IRON AND STEEL. Fig. 95. — Meteoric Iron. Fig. 96.— Cast Steel. FiG. 97.-ARMOR Plate. ^ig. 98.-CAST Iron. MICROSCOPIC STUDIES OF IRON AND STEEL, BY DR. SORBY. CONDITIONS AFFECTING STRENGTH. lOJ ■which it is strained in performing its functions as part of a structure, can probably thus be best and most satisfactorily in- vestigated. An examination of a specimen of gray pig iron showed that the sharp veins of graphite seldom touched the surface of the white iron. In a specimen of spiegeleisen the individual figures have a definite shape, often resembling small fir trees, which end in lines, and finally in points. Sim- ilar observations have also been made in connection with soft gray pig iron, which is distributed through the spiegeleisen. The dark portions correspond to gray pig iron, Fig. 98, while the light portions represent the spiegel. These phenomena are undoubtedly closely connected with the crystallization of iron. The crystals of graphite consist of a series of hexagonal scales, and the flakes, as a rule, occur in a developed state only in gray pig iron, being either entirely absent in spiegeleisen, or only of rare occurrence. As shown by microscopical examinations, crystals of iron are not per- fectly pure, although it has been stated that crystals having the shape which Martens calls " fir-tree crystals," actually con- sist of pure iron. Crystals of both gray and white iron occur together, especially in iron which contains a large proportion of manganese. Martens finds that the fractured surfaces of bars which are broken under repeated use exhibit distinctive features. Surfaces to be examined by the microscope should be first very carefully planed up and smoothly polished ; they should next be well washed with a dilute alkaline solution, to remove all greasy matter, and finally " etched " with dilute acid sufficiently to exhibit well the structure of the metal. The latter process is best practised v/ith very dilute nitric or hydrochloric acid, exposing the surface to its action a few minutes at a time, washing with water and repeating the operation until the surface is brought into the condition in which the microscope is found to best exhibit its character- istics. The Flow of Metals. — M. Tresca, published in La Poinqannage des M^taux, some experiments on the punching of iron, using a large diameter of punch upon small thickness So8 MATERIALS OF CONSTRUCTION— IRON AND STEEL. of metal. His experiments were made with a punch of I.l8 inches diameter (3 centimetres) on iron plates, the greatest thickness being 0.669 inch (1.8 centimetres), and the least 0.1968 inch (0.5 centimetre). He announced, as the result of his investigations, the general law that " when pressure is ex- erted upon the surface of any material, it is transmitted in the interior of the mass from particle to particle, and tends to pro- duce a flow of metal in the direction in which the resistance is least." Experiments made with nuts or bars punched cold show that an actual flow does take place in metals under pressure, which flow is governed by some law not yet enunciated.* As the punch entered, a flow took place, which was great- est in the width — the direction of least resistance — the length being but slightly increased ; the increase was greatest on the bottom face of the block. The metal at the top was there- fore compelled to spread laterally, producing increased width. The punched block was bulged in the width, producing a curved surface, concave toward the axis, greatest at the cen- tral line of its width, and decreasing gradually in either direction as it departs from that line. The length is also increased, but not as noticeably as the width. Fig. 99 represents, full size, the core punched from the block. Fig. 100, it is only ifj- inches in depth, while the hole from which it came was i ^ inches deep. At first sight, it would seem that all the metal from the hole had been squeezed into the core, and, therefore, that its density must be in- But the density of the block was 7.82. The core itself had a density of 7.78. The density of the block is slightly more than that of the core ; but this difference is probably due to the density o| the surface being increased by chipping and filing, and also to the greater soundness of the block. As the density has not increased, there must have been a flow of metal from the core into the block. * Journal of tin Franklin Inslitute, March, 1878. D. Townsend. CONDITIONS AFFECTING STRENGTH. 507 Fig. 101. These experiments were tried with other thicknesses ; the flow seemed to -decrease directly as the diameter of the hole increased, and as the thickness of the bar decreased. In order to show to the eye the flow which had thus oc- curred, several of the large nuts with their cores were planed in half. The resulting rectan- gular faces being brightly pol- ished and perfectly clean, were then etched with acids of various strengths, when they presented the appearances of Figs. 100 and loi. The curved lines mark the laminae, or plates, which were piled and rolled together to make up the bar. It will be noticed that they all curve downward, and that the greatest curvature occurs at the top, remaining nearly constant for some distance, and then decreasing toward the bottom. The flow must have occurred when the punch first entered the bar, and continued regularly, until the pressure above parted the under face, and the core was forced out. In the case represented in Figs. 102 and 103, the hole was punched with the grain, instead fVljriL OfltB''^ ZF^lMffl of across it, the result being that the superposed laminae, Fig. 100. Fig. 102. Fig. 103. instead of being curved down, ward, were wrinkled or warped, from the flow and the conse- 510 MATERIALS OF CONSTRUCTION— IRON AND STEEL. quent pressure which took place, acting against their sides or faces. Several experiments were tried by partially punching bars of the same thickness with punches that had the same diam- eter, but which varied in length according to the depth of the hole to be punched. The bars were uniformly i|f inches thick, and the punch -^ inch in diameter. In the last experiment the punch' was stopped at a depth of i^ inches, the resulting block being shown in Fig. 105. The core projects from the bottom face nearly \ inch, Fig. 104. Punched Nuts ; Full Size. Fig. 105. and measures, as before, almost \ inch in depth. The layers, in this case, are all severed, and the line of parting of the core from the block is plainly visible. The process of punching these thick bars does not depend for its successful perform- ance upon the time taken, but upon the accuracy and power of the machine, and the quality of the punch. The flow re- mains the same, whether the motion is fast or slow. , Relief of Internal Stresses by Rest. — A method of improvement under such conditions as have been elsewhere described, and one which seems likely to have an effect of especial importance in castings in hardened and in tem- pered metals, is the gradual relief, by rest, of those internal CONDITIONS AFFECTING STRENGTH. 511 stresses which are induced by working malleable iron, by casting cast irons, and by tempering steel. These stresses are apparently relieved by the process of flow investigated by Prof. Henry and Mon. Tresca. Rodman reports* the following tests of cannon tested in one case, a few days after they were cast, and in another case, more than six years later : Cast in 1851 and proved same year Cast in 1846 and proved in 1852. . 7.287 7.247 7.220 TENACITY. 37,8il) ; 2.658 s ' 29,423 1 2,068 f 22,g8g i 1,616 f ENDURANCE. 72 fires. 2,582 " 800 " Tenacities are given in pounds per square inch, and in kilogrammes per square centimetre. Rodman calls attention to this extraordinary difference, and explains the change in the manner already indicated, illustrating it by reference to the readiness with which pieces of metal under strain con- form to the new shape given them, as when hoops bent upon barrels at first lose but little of their power of restoration, but afterward take permanently the bend given them ; this process being repeated, they may finally take a bend that would at first have broken them. In the case above cited, the metal of highest tenacity proved weakest under fire. Hard and strong cast iron is most liable to internal stress. The Author tested wires from the cables of the Fair- mount Suspension Bridge, at Philadelphia, when taken down after 32 years' use, and found them fully equal in tenacity and ductility to wire of similar grade just from the wire mill. This tenacity was about 90,000 pounds per square inch (6,327 kilogrammes per square centimetre) ; they were 0.1236 inch (0.314 centimetre) in diameter. Iron and steel wire is found stronger and more ductile after having been kept long in stock, than if tested when first * Report on Metals for Cannon, p. 217. 512 MATERIALS OF CONSTRUCTION— IRON AND STEEL. made. It is, therefore, advisable to keep all strained metals out of use as long as is possible or convenient before subjecting them to stress. The Effect of Time under Stress is also often ob- servable, and is frequently even important. It is not the same for all metals, or even for different specimens of the same class. M. Vicat states that in his experiments* four wires were loaded, respectively, with \, \, \ and | their ultimate resistance, and their elongations were observed and recorded at intervals of one year. The relative extensions observed indicated a gradual lengthening of the three which were strained beyond the elas- tic limit, and that most strained finally broke, after sustaining three-fourths its original ultimate breaking weight two years and nine months, the point of rupture being finally deter- mined by the action of corrosion, which had not been entirely prevented. The several extensions were as follows : No. I, sustaining \, 33 months 0.000 per cent. No. 2, sustaining i, 33 months 0.275 per cent. No. 3, sustaining j, 33 months 0.409 per cent. No. 4, sustaining \, 33 months 0.613 P^r cent. The rate of extension was nearly proportional to the times, and the total extension to the forces. M. Vicat con- cludes that metal, thus overstrained will ultimately break, and his paper has been supposed to indicate a possibility of the ultimate failure of structures having originally an ample fac- tor of safety. Fairbairn made experiments of nearly the same nature as those of Vicat, upon cast-iron bars loaded transversely. These bars were 4^ feet (1.4 metres) between supports, and loaded with two-thirds their breaking weights. Cold-blast iron in- creased in deflection from 1.27 inches (3.23 centimetres) to 1. 31 inches (3.33 centimetres) in five years; the hot-blast bars * Annales de Chimie ct de Physique, 1S34. Tome 54, p. 35. CONDITIONS AFFECTING STRENGTH. SI3 deflected 1.46 inches (3.7 centimetres) to 1.62 inches (4.1 cen- timetres) in the same time. The deflection decreased after 13^ years, and increased again during the last two. The Author has similarly investigated the action of pro- longed stress, using wire of Swedish iron ; but one set ol samples was annealed ; the other, of two sets, was left hard, as drawn from the wire blocks. The size selected was No. 36, 0.004 inch (0.01 millimetre) diameter, and were loaded with 95, 90, 85, 80, 75, 70, 65 and 60 per cent, of the breaking load as obtained by the usual method of test. The result was, dating from 1882 : the hard was twice as strong as the soft wire. ENDURANCE OF IRON WIRE UNDER STATIC I.OAD (1884). TIME UNDER. LOAD BEFORE FRACTURE. PER CENT. MAX. OAD. STATIC 1. Hard wire (unannealed). Soft wire (annealed). 95 8 days. 3 minutes. go 35 days. 5 minutes. 85 Unbroken at end of 16 mos. I day. 80 91 days. 266 days. 75 \ Unbriiken. 17 days. 70 S- (Still whole after g years 455 days. 65 \ -1S92.) 455 days. This very remarkable difference between hard drawn and annealed iron, thus discovered by the Author, throws some light upon the discrepancy previously supposed to exist be- tween the results of Vicat's experiments and common experi- ence, as well as upon the conditions of safety of loaded iron structures. Soft irons and the " tin class " of metals and the woods are thus found to demand a higher factor of safety than hard iron. The elegant and valuable researches, also, of Mons. H. Tresca on the flow of solids,* and the illustrations of this action almost daily noticed by every engineer, seem to lend * Sur V Ecoulement des corps solides, Paris, 1869-72. 33 514 MATERIALS OF CONSTRUCTION— IRON AND STEEL. confirmation to the supposition of Vicat. The experimental researches of Prof. Joseph Henry, on the viscosity of materi- als, and which proved the possibility of the co-existence of strong cohesive forces with great fluidity,* long ago proved also the possibility of a behavior in solids, under the action of great force, analogous to that noted in more fluid sub- stances. On the other hand, the researches of the writer, indicat- ing by strain diagrams that the progress, of this flow is often accompanied by increasing resistance, and the corroboratory evidence furnished by all such carefully made experiments on tensile resistance as those of King and Rodman, Kirkaldy and Styffe, have made it appear extremely doubtful whether hard iron is ever weakened by a continuance of any stress not originally capable of producing incipient rupture. Velocity of Rupture ; Shock. — Kirkaldy concludes that the additional time occupied in testing certain specimens of which he determined the elongation " had no injurious effect in lessening the amount of breaking strain." f An exami- nation of his tables shows those bars which were longest under strain to have had highest average resistance. Wertheim supposed that greater resistance was offered to rapidly than to slowly produced rupture. The experiments of the Author prove that, as had al- ready been indicated by Kirkaldy, a lower resistance is offered by ordinary irons as the stress is more rapidly applied. This effect conspires with vis viva to produce rupture. We conclude that the rapidity of action in cases of shock, and where materials sustain live loads, is a very important element in the determination of their resisting power, not only for the reason given already, but because the more rapidly common iron is ruptured the less is its resistance to fracture. This loss of resistance is about 15 per cent.:j: in some cases, noted by the Author, of moderately rapid distortion. ^ Proc. Am. Phil. Society, 1844. \ Experiments on Wrought Iron and Steel, pp. 62, 83. % Compare Kirkaldy, p. 83, where experiments which are possibly aSected by the action of vis viva indicate a very similar effect. CONDITIONS AFFECTING STRENGTH. S15 The cause of this action bears a close relation to that Operating to produce the opposite phenomenon of the ele- vation of the elastic limit by prolonged stress, to be de- scribed, and it may probably be simply another illustration of the effect of internal strain. Metals of the "tin class" exhibit, as has been shown by the Author,* an opposite ef- fect. Rapidly broken they offer greater resistance than to a static or slowly applied load. It has also been seen that annealed iron has, in some respects, similar quali- ties. With a very slow distortion the " flow " already described occurs, and but a small amount of internal strain is produced, since, by the action noticed when left at rest, this strain re- lieves itself as rapidly as produced. A more rapid distortion produces internal stress more rapidly than reHef can take place, and the more quickly it occurs the less thoroughly can it be relieved, and the more is the total resistance of the piece reduced. Evidence confirmatory of this explanation is found in the fact that bodies most homogeneous as to strain exhibit these effects least. It does not now seem remarkable that, at extremely high velocities, the most ductile substances exhibit similar behav- ior when fractured by shock or by a suddenly applied force, to substances which are really comparatively brittle.f In the production of this effect, which has been frequently observed in the fracture of iron, although the cause has not been recog- nized, the inertia of the mass attacked and the actual depre- ciation of resisting power just observed, conspire to produce results which would seem quite inexplicable, except for the evidently great concentration of energy here referred to, which, in consequence of this conspiring of inertia and re- sistance, brings the total effort upon a comparatively limited portion of the material, producing the short fracture, with its granular surfaces, which is the well-known characteristic of * Trans. Am. Soc. C. E., 1874, et seq. f Specimens from wrought-iron targets shattered by shock of heavy ordnance, in the possession of the Author, exhibit this change in a very unmistakable manner. Sl6 MATERIALS OF CONSTRUCTION— IRON AND STEEL. sudden rupture. Any cause acting to produce increased den- sity, as reduction of temperature, evidently must intensify this action of suddenly applied stress. ' The liability of machinery and structures to injury by shock is thus greatly increased, and it is quite uncertain what is the proper factor of safety to adopt in cases in which the shocks are very suddenly produced. Meantime the precautions to be taken by the engineer are : To prevent the occurrence of shock as far as possible, and to use in endangered parts light and elastic members, composed of the most ductile materials available, giving them such forms and combinations as shall distribute the distortion as uniformly and as widely as possible. The behavior of materials subjected to sudden strain is thus seen to be so considerably modified by both internal and external conditions which are themselves variable in charac- ter, that it may still prove quite difficult to obtain mathemat- ical expressions for the laws governing them. An approxi- mation, of sufficient accuracy for some cases which frequently arise in practice, may be obtained for the safety factor by a study and comparison of experimental results. " Rate of Set " of Metals Subject to Stress for Considerable Periods of Time. — The results of experi- ments made by the Author to determine the time required to produce " set " in metals loaded more or less heavily, and to ascertain what law governs the influence of time in deter- mining the progress and the limit of change of form as the metal yields under loads, either very small or approaching the ultimate strength of the piece, were reported to the American Society of Civil Engineers, January, 1877.* Two methods of testing bars by transverse stress were adopted. By the first method, the bar was bent to a certain carefully measured deflection, and there held, and its effort to straighten itself was as carefully measured. This effort was at first equal to the load required to bend the bar to the observed deflection, but it gradually became less and less as * Trans. Am. Society of Civil Engineers, 1877, Iron Age, 1877. CONDITIONS AFFECTING STRENGTH. 517 the bar took a set, and finally either became constant, or the bar broke. In the first case, this loss of straightening power ceased when the bar had taken its set completely. By the second method, the bar was similarly mounted be- tween supports, but was then loaded with a " dead load " of a certain carefully measured amount, and the manner in which deflection took place and its amount, were very accu- rately observed. When the deflection no longer increased, and the bar remained at a constant deflection, the set was complete. In some cases the increase of deflection did not cease until the bar broke. Pounds Oec. of Load. Kilo- grammes 300 T la 270 265 am 225 210 195 ISO 165 UO ' i m 6 I-' / S» .c f / f .N Mi ^uu GUT (.'JrlL — 135 120 105 to 75 DO "^ .-a^ / 7 f4 \ \t- TEEft i:fA£. NO. Gl.c( PPEB 27.5 Irra 72.5 69 H'as 27 / — N0.5 ;lam lis I0.61 rEKU INAL N0.(j a CO ■PEE 47.50 . ZIN C.62. 22 b 'BB. ■& k^- H^BJ 7.50. ZINC 52.5( 1 ijc .612. so- li /•- 1 0^ K .643 \VRO GHT IBON l3T. TEIA :.■ ~ NO sai. r- =- Wi iToo 'FEE 27. 5i) Ti> 7:2.50 B • \ a S s t s ■ - -* Si " 31 3 ■ 31 £ t- t h S \ *"• s; 3 : ? n ? Fig. 106. — Decrease of Resistance with Time. This research was thus divided into two parts : The first on the observed decrease of resistance at a fixed distortion ; the 5l8 MATERIALS OF CONSTRUCTION— IRON AND STEEL. second on the observed increase of deflection under static loads. We here present the principal deductions. Bars were prepared of square section, i inch (2.54 centi- metres) in breadth and depth, and 22 .inches (56 centimetres) in length between bearings. They were flexed in a machine for testing the resistance of materials to transverse stress, and the load and deflection carefully measured. As the bars were retained at a constant deflection, their effort to resume their original form gradually decreased, and the amount of this effort was from time to time noted. When this effort or resist- ance had become considerably decreased, the bar was released and the set measured. This operation was repeated with each until the law of decrease of elastic resistance was de- tected. Curves (Fig. 106) were constructed, illustrating graphically this law. In all of these metals the set and the loss of effort to resume the original form were phenomena requiring time for their progress, and in all, except in the case of No. 599 — which was loaded heavily — the change gradually became less and less rapid, tending constantly toward a maximum. So far as the observation of the Author has extended, the latter is always the case under light loads. As heavier loads are added, and the maximum resistance of the material is ap- proached, the change continues to progress longer, and, as in the case of the brass above described, it may progress so far as to produce rupture, when the load becomes heavy, if the metal does not belong to the " iron class." The brass broke under a stress 25 per cent, less than it had sustained previously. Other experiments were conducted with the same object as those above described. In these experiments, however, the load, instead of the distortion, was made constant, and deflection was allowed to progress, its rate being observed, until the test piece either broke under the load or rapidly yielded, or until a permanent set was produced. The results of these experiments are in striking accordance with those con- ducted in the manner previously described ; they exhibit the fact of a gradually changing rate of set for the several cases of light or heavy loads, and illustrate the distinctions between CONDITIONS AFFECTING STRENGTH. SI9 the two classes of metal equally well. In one case, a " time- test" was made when the deflection was 0.546 inch and the load 1,233 pounds. The result noted was singular. The effort steadily decreased at a varying rate, which is indi- cated by the diagram of time and loads, and the bar finally snapped sharply, and the two halves fell upon the floor. The effort had decreased to 91 1 pounds. The deflection was pre- cisely what it had been under the load of 1,233 pounds. The beam had balanced at 911 pounds for about three minutes when the fracture took place. The bar was hard, brittle, and elastic, but must a'pparently be classed with tin in its behavior under either continued or intermitted stress. There seems to the Author to exist a distinction, illus- trated in these cases, between that " flow " which is seen in these metals, and that to which has been attributed the re- lief of internal stress and the elevation of the elastic limit by strain and with time. This last phenomenon — the exaltation of the elastic limit by strain — has been observed very strikingly, by the writer, in the deflection of iron bars by transverse stress. The plate exhibits also the strain diagrams obtained by transverse de- flection of four bars of ordinary merchant wrought iron, which were all cut from the same rod. Of these, two were tested in the Fairbanks machine, in which the deflection remains con- stant when the machine is untouched, while the load grad- ually decreases — or, more properly, while the effort of the bar to regain its original form, decreases. The other two were tested by dead loads — the load remaining constant, while the deflection may vary when the apparatus is left to itself. These two pairs of specimens were broken ; one in each set by adding weight steadily until the end of the test, so as to give as little time for elevation of elastic limits as was pos- sible, and one in each set by intermittent stress, observing sets, and the elevation of the elastic limit of metals. As seen by study of these diagrams, both classes, when strained by flexure, gradually exhibit less and less effort to restore themselves to their original form. 520 MATERIALS OF CONSTRUCTION-^IRON AND STEEL. In the case of the tin class, this loss of straightening power seems often to continue indefinitely, and, as in one example here illustrated, even until fracture occurs. With iron and the class of which that metal is typical, this reduction of effort becomes gradually less and less rapid, and finally reaches a limit after attaining which the bar is found to have become strengthened, and the elastic limit to have become elevated. In this respect the two classes are affected by time of stress in precisely opposite ways. The plate exhibits superior ultimate resistance of bars which have been intermittently strained, as well as elevation of the elastic limit. The parallelism of the " elasticity lines " obtained in taking sets, shows that the modulus of elasticity is unaffected by the causes of elevation of the elastic limit. Evidence appealing directly to the senses has been pre- sented in the course of experiment on the second class of metals, of intra-molecular flow. When a bar of tin is bent, it emits while bending the peculiar crackling sound, familiarly known as the " cry of tin." This sound has not been observed hitherto, so far as the Author is aware,' when a bar has been held flexed and perfectly still. In several cases in experiments on flexure of metals of the second class, bars held at a constant deflection have emitted such sounds hour after hour, while taking set and losing their power of restoration of shape.* The Variation of the Normal Elastic Limits with Time. — The elevation of the normal series of elastic limits, and of strength, by intermitting strain was discovered by the Author and by Captain Beardslee, of the U. S. Navy, inde- pendently, in the year 1873. It is variable in amount with different materials of the iron class,' and the rate at which this exaltation progresses is also variable. With the same material and under the same conditions of manufacture and of subsequent treatment, the rate of exaltation is quite definite, and may be expressed by a very simple formula.* The process of exaltation of the normal elastic limit due to any given degree of strain usually nearly approaches a maxi- * Trans. Am. Soc. C. E., 1876. CONDITIONS AFFECTING STRENGTH. 521 mum in the course of a few days of rest after strain, its progress being rapid at first and the rate of increase quickly diminishing with time. For some good bridge or machinery irons, the amount of the excess of the exalted limit, as shown by subsequent test, above the stress at which the load had been previously removed may be expressed approximately by the formula : E — ^ log T + 1.50 per cent. ; in which the time, T, is given in hours of rest after removal of the tensile stress which produced the noted stretch. Captain Beardslee found that, with good ductile iron, the ultimate strength is increased over 15 per cent, by being strained nearly to its limit of tenacity and then allowed to rest for at least one day. With coarse brittle iron, the increase of strength is not so great, a number of specimens of this character showing an average gain of about 6 per cent. Another set of experiments upon the action of this law was made by breaking a bar in its normal condition, and again, several days afterward, breaking one of the pieces. The second piece invariably showed a very much greater strength than the first, the gain in some cases being 20/D00 pounds per square inch, or nearly 40 per cent. This peculiar effect is well shown by a pair of bars exhib- ited in Fig. 107. Fig. 107. — Effect of Intermitted Strain. The upper specimen broke in the eye while under test and after it had begun to draw down, as seen near the letter T in the name at the right. When repaired and again tested next day, it broke in a new place nearer the middle. S22 MATERIALS OF CONSTRUCTION— IRON AND STEEL. The extent of this elevation of the normal series of elastic limits by intermitted stress is affected by quality, thus : Effect of Rest Upon Irons. Test pieces rested 1 8 hours. NUMBER AND MARKS. 62 boiler iron . . . 63 64 65 66 67 contract chain 68 69 70 71 72 iron K 73 74 ULTIMATE GAIN IN STRENGTH. STRENGTH. i u a a u £ en & Lbs. Lbs. 48,600 56,500 7,900 16.0 49,800 57,000 7,200 16.4 49,800 58,000 9, 200 18.4 48,100 54,400 6,300 13.1 48,150 55,550 7,400 15.0 50, 200 54,000 3,800 7-5 50,250 53,200 2,950 5-8 50, 700 55,300 4,600 9.0 49,600 52,900 3,300 6.6 ^1,200 52,800 1,600 3-2 58,800 64,500 5,700 9.6 59,000 65,800 6,800 II-5 56,400 60,600 4,200 7-3 Not broken Brolien Broken Broken Broken Broken Not broken . . Not broken . . Not broken . . Not broken. . Broken , . Broken ( Average, 9.4 Broken \ P^-" <=ent. Average, 15.8 per cent. Average, 6.4 per cent. These experiments indicate that a structure composed of iron of low ductility will receive comparatively slight benefit from the operation of this law, while ductile, fibrous metal, which possesses greater power to resist sudden strains, although less capable of resisting steady stress, gains in this latter power to a greater extent by the effect of strains already successfully borne. Evidence of Overstrain. — Thus a piece once over- strained, carries, permanently, unmistakable evidence of the fact, and can be made to reveal the amount of such over- strain at any later time with a fair degree of accuracy. This' evidence cannot be entirely destroyed, even by a moderate CONDITIONS AFFECTING STRENGTH. 523 degree of annealing. Often, only annealing from a high heat, or reheating and reworking can remove it absolutely. Thus, too, a structure, broken down by overstrain, retains in every piece a register of the maximum load to which that piece has been subjected ; and the strain-sheet of the structure, as strained at the instant of breaking down, can be laid down. Here may be found a means of tracing the overstrains which have resulted in the destruction or the injury of any iron or steel structure, and of ascertaining the cause and the method of its failure, in cases frequently happening, in which they are indeterminable by any of the usual methods of in- vestigation. In illustration of an application of facts known to the determination of the causes and the method of the injury or the destruction of a structure, assume a bridge to have been built with a span of 150 feet, and to have been given such proportions that, with a weight of 1,200 pounds per running foot, and a load of one ton per running foot, the maximum stress on end rods, or other members most strained, is as high as 20,000 pounds per square inch of section of metal. Suppose this bridge to have its tension members composed of a fair, but unrefined, iron, having an elastic limit at about 17,000 pounds per inch (1,195 kilogrammes per square cen- timetre), and a tenacity of 45,000 to 48,000 pounds (3,164 to 3,374 kilogrammes per square centimetre), and with an exten- sibility of about 20 per cent. Suppose this structure to break down under a load ex- ceeding that usually sustained in ordinary work and portions of the several tension members to be subsequently removed, and, a few days after the accident, to be carefully tested, with the results shown on page 524. The extensibility is found to be as little as from ten to fifteen per cent. The tension members are straight bolts without upset ends, the threads being cut, as was formerly common, in such a manner that the section at the bottom of the thread is one- third less than the sectional area of the body of the bar. The location of the tested pieces in the structure being 524 MATERIALS OF CONSTRUCTION— IRON AND STEEL. noted, it is found that' the stronger metal, having also the highest elastic limit, came from the neighborhood of the point at which the bridge gave way, and that the weakest metal, and that exhibiting the lowest elastic limit, came usu- ally from points more or less remote from the break. It is not likely that in all cases the increase in the altitude of the elastic limit, and the increase noted in the ultimate strength of the samples would exhibit a regular order coincident with the order of the rods as to position in the structure ; since the magnitude and the arrangement of the bars would, to a certain extent, determine the relative amounts of strain thrown upon them by overloading any one part of the truss. For present purposes, we may assume the order of arrange- ment to be thus coincident. Sample No. i " 2 " 3 " 4 " 5 " 6 <( » / " 8 " 9 " 10 ELASTIC LIMIT. British. 16,500 18,000 20,000 22, 500 25,000 27,500 28,000 30,000 32,000 34,000 Metric. 1,160 1,265 1,406 1,582 1.758 1.933 1, 968 2,109 2,250 2.390 TENACITY. British. 46,000 48,000 48,000 50,000 52,000 52,000 52,000 52,000 53,000 53,000 Metric. 3,243 3.374 3.374 3.515 3,656 3.656 3.656 3,656 3,726 3,726 On examination of the figures as above given, the engi- neer would conclude : First, that the original apparent elastic limit of the iron used in this case must have been not far from 17,000 pounds per square inch, and that its tenacity was between 46,000 and 48,000 pounds ; secondly, that this prim-^ itive elastic limit had been elevated, by subsequent loads ex- ceeding that amount, to the higher figures given by the bars numbered from 3 to 10 inclusive; thirdly, that the ultimate strength of the material had been, in some examples given above, increased by similarly intermitted strain. CONDITIONS AFFECTING STRENGTH. S^S It would be concluded that the ordinary loads, such as had been carried previously to the entrance upon the bridge of that which caused its destruction, never exceeded, in their straining action, 16,500 pounds per square inch of section of tension rod at the part of the truss from which No. i had been taken, and that the other rods tested had carried, probably, at the time of the accident, loads approximately equal to those required to strain them to the extent measured by their elastic limits at the time of testing them. It would be concluded that the rod from which No. 10 was cut was either that most strained by the load, and there- fore nearest the point of fracture of the truss, or that it was very near that point, and it would be made the basis of com- parison in further studying the case. As this elastic limit approaches most nearly the breaking strength of the metal, we may apply the formula for the ele- vation of the elastic limit with time after intermitted strain, which has been above given as derived from tests of a metal of very similar quality. Takingthe time of intermission as one week, the extent of the increase has a probable value not far from^' = S log. 168 + 1.5 = nearly 12^ percent. The magni- tude of the stress upon this piece at the time of the accident was therefore 34,000, less one-ninth of that value, or about 30,000 pounds per square inch of cross section of the bar. This corresponds to about 45,000 pounds per square inch at the bottom of the thread, and is within 5 per cent, of the primitive breaking strength of the iron. The bar, if broken at the screwed portion, has therefore yielded under a dead load which was at least equal to its maximum resistance, or under a smaller load acting so suddenly as to have the effect of a real " live load." Or, the slight difference here noted may be due to a flaw at the point of fracture. However that may be, it is almost certain that the body of the rod has sustained a stress of not far from 30,000 pounds per square inch. But it is found, on further investigation, that the load on the structure at the time of the accident was but sufficient to make the maximum stress on these rods — if properly distrib- 526 MATERIALS OF CONSTRUCTION— IRON AND STEEL. uted — 20,000 pounds per square inch (1,406 kilogrammes per square centimetre) at the threaded part of the piece, which piece, it has been seen, has been broken by a strain nearly double that figure. The fact is at once inferable that the load came upon these members with such suddenness as to have at least the effect of a live load, and giving a maximum stress equal to twice that produced by the same load gradu- ally applied, i. e., the case in which the load falls through a height equal to the extension of the piece strained by it, the resistances being assumed to increase directly as the exten- sion up to the point of rupture — an assumption which is ap- proximately correct for brittle materials like hard cast iron, but quite erroneous in the case of some ductile materials, which latter sometimes give a " work of ultimate resistance " amounting to three-fourths or even five-sixths of the product of maximum resistance by the extension. This accident was therefore caused by the entrance upon the bridge of a load capable of straining the metal to about one half of its ultimate strength, if slowly applied, but which, in consequence of its sudden application, doubled that stress. This sudden action may have been a conse- quence either of its coming upon the structure at a very high speed, or a result of the loosening of a nut, or of the breaking of a part of either the bridge floor or of one of the trucks of the train. The latter occurrence, permitting the load to fall even a very small distance, would be sufficient. Effect of Orthogonal Strains. — In whatever direc- tion the stress may be applied, and whatever the line of strain, the effect is the same so far as it concerns the normal elastic limit. Iron and steel wires broken by tension are found to have the transverse elastic limit abnormally elevated, and to have become very stiff and of comparatively slight ductility. This is true of wires of some other metals, and of heavier sections of metal. A large quantity of cold-rolled shafting of all sizes, of which both the longitudinal and the transverse di- mensions had been altered by rolling cold, when tested by the Author exhibited great increase of stiffness and strength, and CONDITIONS AFFECTING STRENGTH. 527 an even more considerable exaltation of the normal elastic limit. Torsion similarly stiffened wires and rods longitudi- nally, and test pieces longitudinally strained become stiffer against torsionally and transversely applied stress. Thus or- thogonal strains mutually affect orthogonal resistances of metals ; and the engineer is, by this fact, compelled to study these mutual influences in designing structures in which the stresses approach or exceed, separately or in combination, the normal primitive elastic limit of his material. The following is, in detail, an account of the behavior of a bar of "good merchant iron " under the action of intermittent and successively applied orthogonal strain (transverse suc- ceeded by tension) : * A bar of .good bridge or cable iron 2 inches (5.08 centi- metres) square and about 4 feet (1.2 metres) long was split longitudinally ; one half was cut into tension test pieces, and the other half bent on the transverse testing machine to an angle at the middle of about 120° ; the bent bar was then cut into tension test pieces like the first, and finally all these pieces were broken in tension. On examining the results thus obtained, it was found that the original elastic limit of the metal, as exhibited by the test of the unbent bar, had been exalted by transverse strain in all parts of the bar which had been so strained before being tested by tension. This elevation of the primitive normal limit had not occurred, as would have been expected, to the greatest extent at the points most strained, i. e., nearest the bend at the middle of the strained bar and less and less as the point of maximum strain was departed from, until, at the ends of the bar, this elevation became much less observable, but took place irregu- larly, and, on the average, about as much at one part as .at another. The elevation of the primitive elastic limit, in this in- stance, is 30 per cent, as an average, and in some parts of the bar about 50 per cent. The new series of the elastic limits are less uniform in value than in the original bar." * Tram. Am. Society C. E. Vol. IX., No. cxci., 1880, CHAPTER XVII. CONDITIONS AFFECTING STRENGTH OF NON-FERROUS METALS AND ALLOYS. The Conditions Affecting the Strength of the Non- Ferrous Metals, are precisely such as have been found to modify the valuable properties of iron and steel, and of other materials of construction used by the engineer. The effect of every change, whether chemical or physical, of internal or of external conditions, affecting the metal is seen in a modifi- cation of its strength, elasticity, ductility and resilience. Change of temperature, either gradual or sudden, alteration of methods of manufacture, differences, however slight, of composition and of density, and every variation of the mag- nitude, and of the number of applications, of the load has an effect, more or less marked and important, upon the value and reliability of the metal as a structural material. The effect of heat and of variation of temperature upon the non-ferrous metals and upon the alloys has been but little studied ; but some important facts have become well ascertained. The Strength of Copper is modified by tempera- ture in the same general way as iron (Part II., Mat. of Eng.). It is reduced steadily, and according to a simple law, as tem- perature rises, finally becoming zero at the point of fusion. Decrease of temperature causes increase of strength. A committee of the Franklin Institute, of the State of Pennsylvania, consisting of Professor W. R. Johnson, Benja- min Reeves, and Professor A. D. Bache, were engaged, during ,a period extending from April, 1832, to January, 1837, in experiments upon the tenacity of iron and of copper, under the varying conditions of ordinary use. CONDITIONS AFFECTING STRENGTH OF ALLOYS. 529 The effect of change of temperature upon those metals was investigated with equal intelligence and thoroughness, and most valuable results were obtained. Upward of one hundred experiments upon copper, at temperatures ranging from the freezing point up to 1,000° Fahrenheit, exhibited plainly the fact that a gradual diminu- tion of strength occurs with increase of temperature, and vice versa, and that the change is as uniform as the unavoidable irregularities in the structure of the metal would allow. The law of this variation of tenacity, .within the limits between which the experiments were made, was found to be closely represented by the formula, i. e., the squares of the diminutions of tenacity vary as the cubes of the observed temperatures measured from the freez- ing point. The following are the tenacities of copper at various temperatures, as determined by experiment, to the nearest round numbers : TENACITIES OF COPPER WITH VARYING TEMPERATURES. TEMPERATURE. TENACITY. TEMPERATURE, TENACITY. F C. Lbs. per Kilogs. per F. C. Lbs. per Kilogs. per sq. in. sq. cm. sq. in. sq. cm. 122° 50° 33,000 231 602° 316° 22,000 212 100 32,000 225 801 427 19,000 144 302 150 31,000 218 gl2 490 15,000 105 482 250 27,000 190 1. 016 546 11,000 77 545 2go 25,000 176 2,032 i,rii The Effect of Heat on Bronze and the kalchoid al- loys of copper, tin and zinc was determined by the British Admiralty at Portsmouth in the year 1877.* 34 * London Engineering, Oct. 5, 1877. 5 3° MA TERIALS OF CONSTR UCTION—NON-FERRO US ME TALS. The metal was cast in the form of rods one inch in diameter, and composed of five different alloys as follows : No. I. Copper, 87.75 ; tin, 9.75 ; zinc, 2.5 No. 2. Copper, 91 ; tin, 7 ; zinc, 2. No. 3. Copper, 85 ; tin, 5 ; zinc, 10. No. 4. Copper, 83 ; tin, 2 ; zinc, 15. No. 5. Copper, 92.5 ; tin, 5 ; zinc, 2.5. The specimens were heated in an oil bath near the test- ing machine, and the operation of fixing and breaking was rapidly and carefully performed, so as to prevent, as far as possible, loss of heat by radiation. The strength and ductility of the above test-pieces, at atmospheric temperature, were as follows: No. I, 535 pounds, 12.5 percent.; No. 2, 825 pounds, 16 per cent.; No. 3, 525 pounds, 21 per cent.; No. 4, 485 pounds, 26 per cent, and No. 5, 560 pounds, 20 per cent. As the heat increases a gradual loss in strength and ductility occurs, up to a certain temperature, at which, within a few degrees, a great change takes place, the strength falls to about one half the original, and the ductility is wholly gone. Thus in alloy No. i, at 400° F. (204° C.) the tensile strength had fallen to 245 lbs., and the ductility to 0.75 per cent. ; the precise temperature at which the change took place was ascertained to be about 370° F. (188° C). At 350° F. (177° C), the tensile strength was 450 lbs., and ductility 8.25 per cent. At temperatures above the point where this change begins and up to 500° F. (260° C.) there is little if any loss of strength. Phosphor-bronze was less affected by heat, and at 500° F. (260° C.) retained two-thirds its tenacity and one-third its ductility. Muntz metal (copper, 62; zinc, 38) was found rehable up to the limit, and iron and steel were not injured. The following table exhibits the results of these experi- ments in convenient shape. ' CONDITIONS AFFECTING STRENGTH OF ALLO VS. 53 1 o o ^ a -J K < u • •jtinn'na >. ifl w Ot o> to S' 1^ < E *S 11" w 1^ •ansnsx ^ ^ I i ^ to >& !> 1 to ^0 1 Us i§ g S.'' •xinn^na ss^o " feS5g ■ m 1, i ■ansnax S ■* s !• ~° ft 3- ■>!■ % 8 UHti ■ ?■?" •/iinpna ti k 6 s \D «; " Ih ■ 1^^ •SIISU3X 4 1-1 >§, 1 to ; U£hn > •itl!l!JDna J. VO \n vd ■fl .^ < « « u O i O CJHS ajisnax ■^ %■ ? ii-' 3 ¥ H n-: w •Xjilipna V l-t n n S < ^ < « " M Q ■ to »o 5» s ■X}!i!lona k y? "J. ? o» CO a '. a r d J s ij; s o^ u G •snsnsx s s> i tn ^ i : to o oPt; t; , , h ^o.« •jfjiiTPna k c«« ', m < g iJ to a^l •sirenax ft m % y tnino t^ i-> o R M £■ = fi « •jtinnona 1 M s H g CO a B lO »o « Uh'N ■aijsuax ^ tn a, u> •* tf, J? ^ ^ i? k ^■^ I K if. H ."o H ,"n K « "^ '^■^ 1 In tn H o\ r^ n vS- 04 « ■^ •8. N vR H « 532 MATERIALS OF CONSTRUCTION— NON-FERROUS METALS. Various Metals. — Variations of temperature, accord- ing to Baudrimont,* produce alterations of the tenacity of metals, as below. The metals were in the form of wire, nearly 0.4 millimetre (0.0158 inch) in diameter, except the iron, which was 0.175 mm. (0.0067 inch), and the copper, 0.48 mm. (0.0189 inch). The tenacity is reduced to kilo- grammes per square centimetre. All, except iron, are weak- ened by increase of temperature. TENAaTIES OF METALS AT VARYING TEMPERATURES. TENACITY IN KILOGS. PER SQ. CM. 0° c. (32° r.) 100° C. (212° F.) 200° C. (392° F.) Gold 1,840 2,263 2,510 2,832 3,648 20,540 1,522 1,928 2,187 2,327 3,248 19,173 1,288 Platinum 1,728 1,822 Copper Silver 1,858 2,708 21,027 Iron ■• . The Modulus of Elasticity of hard-drawn iron, cop- per, and brass wires was found by Loomis and Kohlrausch f to vary with temperature according to a law expressed by the equation E = E„{j. - at - bf), in which E is the modulus at the temperature t, E„ that at 0° and a and b experimentally determined co-efficients ; for the Centigrade scale, their values are Iron . . . Copper Brass . . 0.0O0483 0.000572 0.000485 0.000000 12 0.00000028 0.000 001 36 *Annales de Chimie et de Physique, 1850. f Am. Jour. Science and Arts, vol. 1., Nov,, 1870. CONDITIONS AFFECTING STRENGTH OF ALLOYS. 533 Thus, the reduction of the value of the modulus between the melting point of ice and the boiling point of water is, for iron 4.6 to 5 per cent.; for copper, 5.5 to 6 per cent.; for brass, 5.6 to 6.2 per cent., and this variation is most rapid at the highest temperature. The values of the moduli were found to be very closely proportional to the co-efificients of expansion. The following determinations were made by Wertheim : VARIATION OF MODULI OF ELASTICITY WITH TEMPERATURE. Lead Gold Silver Palladiuni . . Copper Platinum. . . Steel (wire). Steel (cast). . Iron 15° c. 59° F. 100° C. , 212° F. 200° C , 392° F- S G. Metric. British. Metric. British. Metric. British. 11.232 173 2.4 163 2-3 18.035 558 7-9 531 7.6 54S 7-9 10.304 715 10.2 727 10.4 637 9.1 11.225 979 14.0 .... 8.936 1,052 15.0 983 14.0 786 II. 2 21.083 1,552 22.2 1,418 20.3 1,296 18.5 7.622 1,728 24.9 2,129 30.4 1,928 27.5 7.919 1,956 26.5 1,901 27.1 1,792 25.6 7-757 2,079 . 26.8 2,188 31.2 1.770 25.2 The metric values are in thousands of kilogrammes per square centimetre; the British in milHons of pounds per square inch. The Stress produced by Change of Temperature is easily calculated when the modulus of elasticity and the coefficient of expansion are known, thus : Let E — the modulus of elasticity ; X = the change of length per degree and per unit of length ; Af = the difference of initial and final temperatures; / = the stress produced. S 34 MA TERIAL S OF CONSTR UCTION—NON-FERRO US ME TALS. Then: P'.EwXAfw, .-.p = XE Af For good wrought iron and steel, taking E as 28,000,000 pounds on the square inch, or 2,000,000 kilogrammes on the square centimetre, and A as O. 0000068 for Fahrenheit, and as 0.0000120 for Centigrade degrees : / = 190 At° Fahr., nearly | = 25 Af Cent., nearly \ For cast iron, taking E = 16.000,000; A = 0.0000062: p = 100 Ai° Fahr., nearly ) = 12 At° Cent., nearly ) This force must be allowed for as if a part of the tension, T, or compression, C, produced by the working load when the parts are not free to expand. Sudden Variation of Temperature has an effect, very usually, upon the non-ferrous metals, which is afterward seen in a permanent alteration of their properties. Repeated heating and cooling causes a permanent change of form, and sudden cooling from high temperatures causes a modification of the tenacity an^ ductility of the kalchoids and the metals composing such alloys, which is precisely the opposite of that produced on steel. Thus copper, brass, and bronze, suddenly cooled from a low red heat, are softened and weakened and greatly improved in malleability and ductility. This process, which is one of hardening and tempering with steels, is thus one of softening with other metals. On the other hand, veiy slow cooling softens or " anneals " steels, while it hardens the non-ferrous metals and alloys. Thus, also, casting bronze ord- nance or other castings in chills increases the value of the metal by preventing liquation and securing homogeneousness and maximum density. CONDITIONS AFFECTING STRENGTH OF ALLO YS. 535 The Effect of Chill-casting is exhibited in the fol- lowing tables of tests by tension furnished to the Author by the U. S. N. Department in the course of a series of investi- gations in 1877. The metal has the composition, copper (Lake Superior), 9 ; tin, i ; it was cast either in chills or in sand as specified, after having been melted in a reverberatory furnace, the copper first and the ,tin three hours later. The specimens tested were of the " short " pattern, and the reduction of sec- tion, rather than the elongation recorded, is the measure of relative ductility. The tables also exhibit the method of testing usual in the Ordnance Department of the U. S. Navy in 1875-6. British measures are here used. EFFECT OF CASTING BRONZE IN "CHILLS." Navy Ordnance-Bronze. TENSILE STRENGTH PER SQUARE INCH OF— J. (*- s 2 3 c 5 J s I u MARK, e g 5". In 3.2 PERMANENT TION OF PARTS OF Ml, 3-4-75 BB'iS^!';:::: 22,385 45,737 49,77= 48,000 35,820 29,818 33,630 51,459 70,000 71,600 65,6.^ 60,000 . 39,000 91,600 73.450 71,600 .100 .2603 .211 -4075 .0291 ■:$, .396 .415 40 417 470 240 20 50 134 438 376 373 8.392 8iS53 Pull of larg-e tin spots. M 2, 5-6-75 No. 3, 8-21-75 • • No. 2, 8-21-75 . . GB 2, 5-6-75 ... GB 3, 5-6-75... B 3 L, j2-7<^75. M 1 C, 3-11-76 . B2C, 3-11-76.. 830,3-11-76.. . Cast in chill mould. Cast in chill mould. Flaw in the breaking portion. Cast in chill mould. Cast in chill mould. Cast in chill mould. The guns cast in chill moulds were composed of 10 parts of copper to I part of tin ; the others were of 9 parts of cop- per to I part of tin. In the course of experiments made by Major Wade,* three ' Report on Ordnance. 53^ MA TERIALS OF CONSTRUCTION— NON-FERROUS METALS. howitzers, Nos. 27, 28, and 29, were cast from the same liquid metal. No. 27 was cast when the metal was at the highest temperature, No. 28 was cast fifteen minutes later, and No. 29 fourteen minutes after No. 28. The following results were obtained : SPECIFIC GRAVITY— TENACITY — Of gun- heads. Of entire gun. Of small bars cast in — Of gun- heads. Of small bars cast in— E Gun mould. Separate mould. Guu mould. Separate mould. 29 IS 29 Highest ... Mean Lowest 7.986 8.351 8.538 8.195 8. 531 8.752 8.686 8.823 8.816 8. 554 8.447 8.376 17,761 28,99s 23,722 50,973 52,330 56,786 31,132 28,153 28,082 In casting another howitzer, No. 30, small test-bars were cast in separate moulds, one of which was of cast iron, to ascertain the effect of sudden cooling, and the others were of clay, similar to the gun-mould. The tests of all the samples from this casting were as follows : TENACITY. Small bars cast separately in iron mould . . Small bars cast separately in clay mould . . Small bar cast in guti mould Gun-head samples Finished howitzer 37,688 25,783 53,798 35,578 The effect of the chill is evidently very beneficial, and iron moulds should, therefore, always be used where possible in the casting of bronze ; with brass they are less necessary. Effect of Tempering and Annealing. — Riche de- termined the effect of tempering and annealing upon the CONDITIONS AFFECTING STRENGTH OF ALLOYS. S37 density of the bronzes, finding that tempering increased the density of those rich in tin but not of others, as gun-bronzes ; and that annealing reduces the density of tempered bronze although it does not entirely destroy that effect. Density is increased to a considerable degree by mechanical action as well as by tempering. Successive temperings and annealings produce, on the whole, an increase of density. Tempering, according to both Darcet and Riche, softens the bronzes rich in tin, i.e., those containing about 20 per cent. tin. Thus, Riche obtained the result that such bronzes, tempered, can be moulded in the press, while they will crack if untempered or annealed. Bronze and steel exhibit opposite behavior in this respect. The same author finds that working hot does not increase the density more than working at low temperature. The metal increases in density very rapidly by working hot, and without danger of rupture ; while cold the action is ex- tremely slight and very difficult. There is evidence that the method of making gongs by the Chinese involves working hot under the hammer.* Riche, reaches the following conclusions : f "The bronzes rich in tin (18 to 22 per cent.) increase in density with tempering ; and annealing lessens the density of tempered bronze, but in a less proportion. The density is considerably increased by the alternate action of tempering and annealing, and of the press. These effects, the reverse of those in steel, coincide with the fact that tempering softens bronze while it hardens steel. " This softening, discovered by Darcet, is not sufficient to allow of this bronze being worked cold for industrial pur- poses. It was shown that this metal — extremely hard when cold and pulverizable at red heat — is forged and rolled at dark red heat with remarkable facility. This fact enabled me, in common with M. Champion, to succeed in the manu- facture of tamtams, and other sonorous instruments, by the method followed in the East. * Industries Anciennes, etc. Lacroix, Paris, 1869. f Annales de Chimie et de Physique, vol. xxx., 1873. 53S MATERIALS OF CONSTRUCTION— NON-FERROUS METALS. " Tempering produced no apparent softening in the bronzes less rich in tin (12 to 6 per cent.); and if they are tempered for industrial uses it is more especially in order to detach the oxide produced during the reheating of the matter in the course of the operations. " It was found that in the axis of a cannon, and especially toward the muzzle, there are some parts very rich in tin and in zinc. " The density of copper, subjected alternately to me- chanical action, then to tempering or annealing, displays in- verse variations according as it is exposed to the air or sheltered from it during the reheating ; while in the first case the mechanical action increases the density, in the second mechanical action diminishes the density. " Mechanical action incj;eases the density of yellow brass, and this effect is counteracted in part by tempering, and especially by annealing. It is thought that annealing is pref- erable to tempering in working with brass. " Mechanical action, tempering, and annealing, do not sensibly change the volume of similor and of the bronzes of aluminium, alloys remarkable for the facility with which they can be worked. " While repeated mechanical action increases the density of the bronzes rich in tin, especially of porous copper, of copper alloyed with iron, of brass, it evidently diminishes the density of copper exposed to the air during reheating, and it produces no noticeable alteration in the volume of similor or of aluminium bronze. Tempering produces on brass, and especially on the bronzes rich in tin previously annealed, an increase in density, contrary to what takes place in steel, cop- per and glass. " It will be perceived that tempering diminishes the density of a body, because the surface, cooled before the. centre, cannot contract freely by reason of the resistance that the interior parts dilated at this moment offer to con- traction." Copper and its alloys should not be exposed to the air •vhile heated, since it is certain to oxidize. CONDITIONS AFFECTING STRENGTH OF ALLOYS. 539 The singular avidity with which copper and the alloys absorb oxygen, causes great variation of their strengths. The Effect of Jime, and Velocity of Rupture, on the action of stress is not less important with the non- ferrous than with the ferrous metals. A very important difference is found to exist between the two classes. (See- Part II., Art. 29s, M. of E.) The rupture of the non-ferrous metals takes place under lower stresses, as the time of oper- ation is greater, and the fracture is more slowly produced. The contrary is the case with iron and steel. With non-ferrous metals, the piece strained may give way, ultimately, under static loads greatly less than those required to produce im- mediate rupture. This occurs to a less extent with soft annealed iron, and still less with harder irons and steels. Cast iron is stated by Hodgkinson to be capable of sustain- ing, indefinitely, loads closely approaching the breaking load under test. Some of the alloys will probably exhibit similar differences. With rapid distortion, the resistance is increased with nOn-ferrous metals, decreased with iron. The Author has, therefore, enunciated a principle which had been deduced from experiments on wrought iron, which is, evidently, of vital importance to the engineer, viz. : " That the time during which applied stress acts is an important ele- ment in determining its effects, not only as an element which modifies the effect of the vis viva of the attacking mass and the action of the inertia of the piece attacked, but also as modifying seriously the conditions of production and relief of internal strain by even simple stresses." * Should it be true, as suggested by the Author, that the cause of the variation of resistance, sometimes observed with increased velocity of distortion, is closely related to the cause of the variation of the elastic limit by strain,t it would seem to be a; corollary that materials so inelastic and so viscous as to be incapable of becoming internally strained during dis- tortion, should offer greater resistance to rapid than to * Trans. American Society of Civil Engineers, vol. iv., p. 334. f See Part II., Mat. of Eng.; figures 135-138. 540 MA TERIALS OF CONSTRUCTION— NON-FERROUS METALS. slowly-produced distortion, in consequence of their inability to " flow " so rapidly as to reduce resistance by such fluxion at the higher speed, or by correspondingly reducing the fractured section. This principle has been shown, by a large number of experiments, to be frequently, if not invariably, the fact. Copper, tin, and other inelastic and ductile metals and alloys, were found by the author to exhibit this behavior, and are therefore quite opposite in this respect to commercial wrought iron and worked steel. The records of the Mechanical Laboratory of the Stevens Institute of Technology frequently illustrate the proposition that metals which gradually yield under a constant load offer increased resistance with increased rapidity of rupture. The curves of deflections of a considerable number of ductile metals and alloys are very smooth when the time dur- ing which each load has been left upon them is the same ; but whenever that time has been variable the curve has been irregular. Bars of such metals broken by transverse stress give a greater resistance to rapidly increasing stress than to stress slowly intensified. Two pieces of tin, as described in pp. 542-544, were broken by tension, the one rapidly and the other slowly. The first broke under a load of 2,100 and the latter of 1,400 pounds. The example illustrates well the very great difference which is possible in such cases, and seems to the writer to indicate the possibility, in extreme cases, of obtaining results which may be fatally deceptive when the time of rijpture is not noted. The depression of the elastic limit has been observed pre- viously in materials, but less attention has been paid to it than the importance of the phenomenon would seem to demand. The strain diagram of a bronze bar is nearly hyperbolic ; but the law of Hooke, ut tensio sic vis, holds good, as usual, up to a point at which the load is about one-half the maximum^ The curve of times and loads exhibits the rate of loss of effort while the bar was finally held at a deflection of 0.5456 inch, the load being carefully and regularly reduced, as the effort diminished, from 1,233 to 9" pounds, at which latter figure the bar broke. The curve is a very smooth one. CONDITIONS AFFECTING STRENGTH OP ALLO VS. 541 EFFECT OF TIME ON BRASS. BAR NO. 599, 90 parts zinc, 10 parts copper : i x 0.992 x 22 inches. Pounds. =3 43 63 103 143 '63 Inch. 0-0033 0.007S 0.0127 0,0225 0.031 Inch. ^ 0.0347 Resistance fell in 15 h. 25 m. to 143 203. 243 283 323 •^•0347 0.0391 0.0471 0.0544 o.o6n 0.0692 Inch. 0.0781 0.08S1 Pounds. 363 403 3 403 0.0886 Resistance fell in 8 o.oJ to 333 Inch. 0.0079 h. 30 m. 0.0246 to 302 303 403 S03 603 o.q8q6 R.esistance fell in 15 h. 0.0896 0,0876 o. 1072 0,1282 0.1521 Pounds. 3 643 803 1,003 1,103 1,203 V-^3 Inch. 0.1641 0.2149 0.3178 0.3921 0.481 0.5209 Inch. 0.0336 Resistance fell in 15 m. to 1,137 3 1,137 1 1233 0.5209 0.5131 0.5456 0.2735 The bar -was left under strain at iii* 22" A.M.,aQd the effort to restore itself measured at intervals, as follows ; HOUE.— 11:37 11:50 A.M. 12:2 12:8 12:25 13:39* 12:53* 12:58* 1:20 P.M. Effort. — 1,133 1,093 ^^°T^ ^1063 1,043 i)023 ijoos 993 911 pounds. At i'' 23*" P.M. the bar broke. BAR NO. 596. 75 parts zinc, 25 parts copper; second casting;; 0.985 x 0.985 x az inches. § g a H s LOAD. .J a SET. .J n SET. LOAD. a SET-. Pounds. Inch. Inch. Pounds. Inch. Inch. Pounds. Inch. Inch. 23 0.0057 463 0.0799 503 0.0894 63 0.0142 503 0.0866 543 0.0952 103 0,0207 3 0.0014 583 0.1012 M3 0.0275 ^°i 0.0866 603 0.1042 183 0.0346 Resistance fell nsh. 623 .0.1075 223 0.0414 to 489 0.0866 643 0.1102 263 0.0485 3 0.0074 „^^3 0.1136 303 0.0549 489 0.0866 Broke s seconds after with 343 0.0610 Resistance fell in i 3 h. 30 m. ringing sound. 383 o.o66g to 473 . 0.0866 423 0.073 3 0.0092 An example of somewhat similar behavior, but exhibited by a metal of very different quality, is shown above. 542 MA TERIALS OF CONSTRUCTION— NON-FERROUS METALS. This bar was hard, brittle, and elastic, but must ap- parently be classed with tin in its behavior under either con- tinued or intermitted stress. These latter specimens were broken ; one in each set by adding weight steadily until the end of the test, so as to give as little time for elevation of elastic limit as was possible ; and one in each set by intermittent stress, observing sets, and the elevation of the elastic limit. There seems to the Author to exist a distinction, illus- trated in these cases, between that " flow " which is seen in these metals, and that to which has been attributed the relief of internal stress and the elevation of the elastic limit by strain and with time. If the long-known effects of cold-hammering, cold-rolling, and wire-drawing in stiffening, strengthening, and hardening some metals can be, as the Author is inclined to believe, at- tributed in part to this molecular change, as well as to simple condensation and closing up of cavities and pores, this vari- ation of the elastic limit by distortion under externally ap- plied force has been shown to occur in iron and in metals of that class in tension, torsion, compression, and under trans- verse strain. Effect of Prolonged Stress on Tin and Zinc. — In testing a bar of tin, in work done as described in earlier chap- ters-, the Author studied this phenonienon. An experiment on No. 29 A (a bar of pure tin) was made to determine the difference in resistance to slow and rapid rupture. This bar was a good casting, and tests of the two pieces, one from the upper and one from the lower end of the bar, should show little, if any, difference in strength. No. 29 A was tested with a load of i,7CXD pounds, which caused an elongation of o.iS inch. This load was then reduced to 1,250 pounds, and the reading again taken, showing an elongation of 0.19 inch, which increased in two minutes to 0.27 inch. The load was then increased to 1,400 pounds, and the elongation was 0.32 inch. The load was allowed to remain on the bar for ten minutes, and the elongation gradually increased to 1.7 inches, when the bar broke. It seems probable from this test that CONDITIONS AFFECTING STRENGTH OF ALLOYS. 543 the load of 1,400 pounds would have broken the piece, even if the load of 1,700 pounds had not been placed on it at the beginning of the test. vs. (- X o o H ; - Jj-r '■ ■'>■ !±:|:!:||:;;;[ CJHg:: i. i nil..... :.;.. _. ; . . i it . . : T7 ■ ■ r : - ■ - --■-5-- -- -■'-■■ - ::::J:; i:::::::i: , : : ' ■ , - T - J— - >" ■ -7 . *i._. ( ..J - _ : : :::: J:::::::: ::::: .^.,. i .__ . . 'ti g ■ - ._ ._.__.... ,_ ■*"■ '^"""'-"^ — ■""■ - - ""fHS" ^ -l.iJci-.- . 1 -_ - " - - '■-■ .z:i- - -- : - :'i £ a i" - - -' ' '' LU - -- ,_._ij^_..,_ ".r ' - '^h r - L.., : ^- -&^ . i..:iE _ ■ ---^-l :: """"S"""" "" ; _,:j:« .X.- . ._ ■ _..._} — .__ _ " "' E '" ' . -■r - ■ - - - - ::::: ::" ""::::"p:::: ^:--3-'ih--±:y: r r " T ■ . . 3-- -- - X _ -^ _ 2""" T--- . 5-- ._.__ §_.__:..:._ ._ "X , — ^ 51 f 1 • 3 5 i :::: :: ■ _ — - p -p — ---- : i ;;:::l:;:::::-|:: i:|:|±: :::::::::: - ■ |- r : ... ::;;:!;;;:[:;:;; h; ::::|::: :::::::::: - -+-^ — M Ji"" " " '- ■ . ::::i::;::;::;:±5i i -M. ._ L - ■■■"l""""i"""i::: — -^ — , - - - :::::::::::5!e:S:: — ~x'""' "" : - ' ? " -i---i±i::::]J::: _- - — L -_p . .J H 3 ; r - i^ .. . - - - J ;::i£::±:::!3^::: |::i5±5:!|::: :;:E::;:;l|:;i;;3 -■■ --ii Bar No. 29 B was tested in a different manner. The load was gradually, but rapidly, increased to 2,100 pounds, with- out stopping longer than was necessary to take the reading 544 MA TERIALS OF CONSTRUCTION— NON-FERROUS METALS. of the elongations at 975, 1,180, 1,290, 1,600, and 2,000 pounds. At 2,100 pounds, the elongation read 1.88 inches. The piece then extended very rapidly, and, at the same time, its resist- ance, as measured by the scale-beam, reduced to 1,700 pounds. The pump of the hydraulic press was worked as fast as pos- sible, but the beam could not be balanced beyond 1,700 pounds. The piece sustained this load a few seconds, then broke after an elongation of 2.58 inches. Comparing the tests, it is seen that the resistance of No. 29 A to an elongation greater than 0.19 inch was never greater than 1,400 pounds, while that of No. 29 B was 2,100 pounds, or 50 per cent, more than the former ; which 50 per cent, ap- parent increase of strength was evidently due to the greater rapidity of the test of No. 29 B. The fact that the difference in strength is only apparent is confirmed by the experiments by torsional stress on pieces from the same bar. These showed that torsion-pieces No. 29 A and No. 29 B, from the top and from the bottom of the bar, tested by moderately slow motion, each gave a resistance of 14.2 foot-pounds tor- sional moment ; piece No. 29 C, from the middle of the bar, tested in the same manner, resisted 13.2 foot-pounds, while No. 29 D, a piece taken from the middle of the bar and ad- joining No. 29 C, tested by very slow motion and left under stress for hours, resisted only 9.2 foot-pounds or some 30 per cent, less than either of the other pieces. The effect of slow and rapid test is shown by both bars in the tensile test. The average tenacity of all the pieces tested is given as 3,130 pounds per square inch, but it is probable that all the pieces would have broken at as low as 2,000 pounds if the test had been of long duration, say one hour, or as high as 4,000 pounds if each test had been made in, say, five minutes. The records of several tests follow. The effect of time is also shown in the autographic strain- diagrams (Fig. 108), and in the records calculated from th^m. CONDITIONS AFFECTING STRENGTH OF ALLOYS. 545 This peculiar property of variation of resistance with time affects a large proportion of the alloys. With zinc, experiments were made to determine the effect of rapid and of slow stress and of resting under stress. They indicated a decrease of resistance when resting under stress, a uniform resistance to very slow motion, and a rapid increase of resistance to rapid motion, except after the resist- ance has reached the maximum, when rapid motion then keeps the resistance constant. It was observed that very ductile metals, such as tin itself and alloys containing a large amount of tin, all exhibit different amounts of resistance to slow and to rapid stress, and a decrease of resistance on resting under stress. The same phenomenon is exhibited by cast zinc, which is much less ductile than the copper-tin alloys, and is less ductile than several of the alloys of copper and zinc (those contain- ing from 20 to 40 per cent, of zinc), which either did not show the phenomenon at all, or but slightly. The Effect on Bronze of long continued stress in producing continuous distortion, even when the loads are far within those required to produce the same effect on first application, is well exhibited below. EFFECT OF TIME ON BRONZE. Tests by Transverse Stress — With Dead Loads. Samples z x z x 22 inches. i MATERIAL PARTS. LOAD. DEFLEC- TION. TIME. INCREASED DEFLEC- TION. BREAKING WEIGHT, Tin. Copper. \ 9 10 zz '1-9' 7.2 zo. 90.3 zoo 98. z 92.8 90, 9-7 Pounds, 600 47S Soo 950 950 1,485 100 I2Q 140 Z40 Z40 Inc z 2 3 hes. 108 395 447 08s 140 221 357 S minutes .... 3 minutes 3 minutes S minutes 5 minutes 13 minutes 10 minutes ... . 10 minutes 10 minutes xo minutes 40 hours Inches. o.oog 0.29T 0.488 O.OSi ; 0.021 4.087 0.021 ' 0-0S5 0.098 0.038 0.920 Pounds, 650 500 1,350 1,485 35 S46 MA TERIALS OF CONSTRUCTION— NON-FERROUS METALS. h MATERIAL PARTS. LOAD, DEFLEC- TION. TIME. INCREASED DEFLEC- TION. BREAKING WEIGHT. i Tin. Copper. 12 '3 98. 8g 100 1. 11 Pounds. 160 160 j6q 160 go 120 120 120 120 80 Inches. 1.294 1.320 2.320 3.320 0.243 0.736 1. 791 2-539 0.218 10 minutes 1 day I day I day 5 minutes 15 minutes 30 minutes 45 minutes 12 hours 5 minutes Inches. 0.025 1. 000 1,000 1.000 0,063 i-oss 0.748 0.595 8.000 0.064 Pounds. 120 TIO Metals having a composition intermediate between these extremes have not been observed to exhibit flow or to in- crease deflection under a constant load. The same phenomena are exhibited by tests made in the autographic testing machine,* thus : MATERIAL. TIME UNDER STRESS. ANGLE OF TOR- SION. FALL OF PENCIL. REMARKS. g Tin. Cop- per. t2 t3 4 100 98.89 0.55 40 hours . . . 1 hour 2 hours . . . 12 minutes . 65 180 280 tsSo ■■58 0.06 inch.. . o.i inch... 0.1 inch... 50 per cent. Recovered after further distortion of 1° Recovered in 8". Recovered in 80°. Did not recover. Behaved like No. 4. Allov. 0.2 inches. * Tests by tension with similar materials exhibit similar results, and these observations and experiments thus seem to indicate that, under some conditions, the phenomena of flow, and of variation of the elastic limit by strain, may be co- existent, and that progressive distortion may occur with " viscous " metals. * A Fluctuation of Resistance with Time, illustrated in the table here given, is a singular phenomenon which has been observed by the Author, but the causes of which remain * M. of E., p. 379. f Same piece. % Taking "elasticity line," CONDITIONS AFFECTING STRENGTH OF ALLOYS, S47 FLUCTUATION OF RESISTANCE. Test by Transverse Stress. ALLOY OF COPPER AND TIN. No. 47.— Material : Alloy.—Original mixture: 17.5 Cu, 82,5 Sn.— Dimensions : Length between supports, 22"; Breadth, 0.996"; Depth, 0.983". Pounds. DEFLEC- TION. Inch. 0.0027 0.0070 0.0153 0.0256 0.0365 0.0499 MODULUS OF ELASTICITY. /V3 4 A bd^ Beam sinks slowly, 8,039,339 7,356,258 6,594)770 6,167,163 0.0617 0.0804 0.1042 0.1343 0.1666 0.1798 0.2145 0.2503 0.3021 0.3367 0.3762 0.4147 0.4S97 0.0821 40 60 80 100 5 120 140 160 180 200 5 200 220 240 260 270 280 290 300 The beam was observed to rise, and another reading of set was taken in 2 minutes. 5 I I 0.3022 I _ The beam rose again, pushed forward the poise till beam balanced at 10 pounds. Time 2 minutes. 0.3084 5,638,814 5,472,481 4^899,597 4.320,565 3,771,245 3,377.873 2,697,980 832,406 DEFLEC- TION. MODULUS OF ELASTICITY. 4 A bd^ In 2 minutes more, beam balanced at 14 Sounds. The pressure-screw was then run ack till beam balanced again at 5 pounds, and another reading of set taken. PoundsA Inches. I Inch. I „S I I 0.2998 I Beam rose again. In 2 minutes balanced at 10 pounds. In 10 minutes balanced at 16 pounds. In 39 minutes balanced at 23 pounds. Ran back pressure-screw till beam balanced again at 5 pounds. ^5 I . I 0.2902 1 In 4 mmutes beam rose agam. In 23 minutes beam balanced at 14 pounds. In I hour and ^^ minutes beam balanced at 20 pounds. Ran back pressure-screw till beam balanced again at 5 pounds. „5 I I 0.2845 I Total decrease of set in 2 hours and 20 min- utes 0.3084 — 0,2845 = 0.0239 inch. Replaced load of 280 pounds. 280 I 0.4849 j I 300 0.5332 I 1 310 I Broke on applying strain. Breaking load, 300 pounds. ■i,Pl Modulus of rupture, R = ^jr ~ 10,288. zoa-' No. 48.— Material : Alloy.—Original mixture: 12.5 Cu, between supports, 22"; Breadth, 0.985"; 87.5 Sn.— Dimensions : Length Depth, 0.990". 30 0.0025 Scale beam rose. 20 0.0050 In 2 minutes balanced at 20 pounds. 40 0,0141 In 4 minutes balanced at 29 pounds. 60 0.0230 7,249,195 In 15 minutes balanced at 34 pounds. Ran back pressure-screw till beam balanced 80 0.0352 6,330,144 Be£ m sinks slov Ay. again at 5 pounds. lOO 0.0508 5,482,803 5 1 J 0.6555 1 .-•• Beam rose again, balanced at 12 pounds in 5 O.OIgO 120 0.0760 4.397.784 5 minutes. 140 0.0969 4,024,116 5 1 1 0.6508. 1 160 0.1262 3.531,237 Total decrease of set m 20 minutes, 0.6742— 180 0.1592 0,6508 = 0.0234 inch. 2.725,307 Beam fose again, but test was continued 5 0.1238 without further waiting. 200 0.2268 260 0.8304 280 0.9018 1 1,639,194 300 1 . 0760 Beam sank rapidly. 260 D.5210 300 Repeated. Bar broke just as beam 0. 5763 ^ rose. 280 0.6458 1,207,609 Breaking load, 300 pounds. 290. 300 5 0.7185 0.8025 0.6742 1,041,220 Modul us of rupture, R ~ -7^ = 10, 254. 548 MA TERIALS OF CONSTRUCTION— NON-FERROUS METALS. to- be determined. The bars tested as shown were not per- fect in structure, and do not exhibit any considerable strength ; they consist principally of tin (82.5 and 87.5 per cent.) and are valueless for the ordinary work of the constructor, although useful " white metals." It is seen that the resistance of both bars was, at times, overcome by the load, but, on balancing the weigh-beam, the bar each time gradually re-acquired a power of raising the load which had deformed it, and straight- ened itself sufficiently to raise the beam against the upper " chock." A decrease of set took place of 0.02 inch — in the first beam in two hours and twenty minutes, and in the second in twenty minutes. In two minutes, recovery occurred to such an extent that the bar exerted an effort of 20 pounds tending to straighten itself, and in 15 minutes of 34 pounds. The phenomenon is one which will demand careful investi- gation. The Effect of Unintermitted and Heavy Stress on Resistance is well exhibited on the two sets of strain-dia- grams* here reproduced from Part II. of M. of E. The first series of tests exhibited decrease of resistance with time. No. 655 was a bar of Queensland tin, presented to the Author by the Commissioner of that country at the Centen- nial Exhibition, and which was found to be remarkably pure. A load of 100 pounds gave a deflection of 0.2109 inch, and produced a set of 0.1753 inch. The same load restored de- flected the bar 0.2415 inch, which deflection being retained, the effort to regain the original shape decreased in one min- ute from 100 to 70 pounds, in 3 minutes to 62, and in 8 min- utes to 56 pounds. The original load of 100 pounds then brought the deflection to 0.3033 inch, nearly 50 per cent, more than at first. A bar, No. 599, of copper-zinc alloy, similarly tested, deflected 0.5209 inch under 1,233 pounds, and took a set»of 0.2736 inch after being held at that deflection 15 minutes, the effort falling meantime to 1,137 pounds. Restoring the load of 1,137 pounds, the deflection became 0.5 131 inch, and the original load of 1,233 pounds brought it to 0.5456 inch. * Trans. American Society of Civil Engineers, 1877, CONDITIONS AFFECTING STRENGTH OF ALLOYS. 549 The- bar was now held at this deflection and the set gradually took place, the effort falling in 15 minutes to 1,132 pounds (4 per cent, more than at the first observation), in 22 minutes to 1,093, ii^ 4^ minutes to 1,063, in 63 minutes to 1,043, i^i 91^ minutes to 1,003, ^i^d in 118 minutes to 911 pounds, at which last strain the bar broke 3 minutes later, the deflection remaining unchanged up to the instant of fracture. This remarkable case has already been referred to in an earlier article, when treating of the effect of time in producing varia- tion of resistance and of the elastic limit. Nos. 561, copper-tin, and 612, copper-zinc, were composi- tions which behaved quite similarly to the iron bar at its first trial, the set apparently becoming nearly complete, in the first after i hour, and in the second after 3 or 4 hours. In all of these metals, the set and the loss of effort to resume the original form were phenomena requiring time for their progress, and in all, except in the case of No. 599 — which was loaded heavily — the change gradually became less and less rapid, tending constantly toward a maximum. So far as the observation of the Author has yet extended, the latter is always the case under light loads. As heavier loads are added, and the maximum resistance of the material is approached, the change continues to progress longer, and, as in the case of the brass above described, it may progress so far as to produce rupture, when the load becomes heavy, up to a limit, which closely approaches maximum tenacity in the "iron class." The brass broke under a stress 25 per cent, less than it had actually sustained previously. The records are herewith presented, and the curves repre- senting them shown in the figures which follow. 550 MA TERIALS OF CONSTRUCTION— NON-FERROUS METALS DECREASE OF RESISTANCE AND INCREASE OP SET OF METALS, WITH TIME. Bars I inch square j 22 inches between supports. No. 648 Wrought iron. First Trial, Min. 100 =75 320 320 322 322 Pounds. Pounds. Inches. 1,003 0.0995 3 1,003 O.IOOI 999 4 O.IOOI 991 12 O.lOOI 987 16 O.IOOI 987 16 O.IOOI 3 987 0,9910 i;oo3 0.1003 2,720 2.6400 Inch. 0.0049 Second Trial. I I1003 I ... I 2.2548I No. 561.- — 27.5 PARTS COPPER, 72.5 PARTS TIN. 3 2,640 4,140 160 0,0696 5 160 0.072 154 6 0.07Z 150 10 0.072 104 56 0.072 100 60 0.072 5 100 0.0763 160 0.0970 320 0.22QO 0.0145 Broke No. 599. — 10 PARTS COPPER, go PARTS ZINC. 15 28 46 63 77. s 91.5 1,233 0.5209 I.I37 3 1.137 0.5209 0.5131 '1=33 0.5456 1. 133 100 0.5456 1,093 140 0.5456 1,070 163 0.5456 1,063 170 0.5456 1,043 190 0.5456 1,023 210 0.5456 1,003 230 0.5456 0.2736 Min. 3 23 53 "33 '93 363 363 96.5 118 Pounds. 3 1,603 1,521 1,493 1,483 1,463 1,461 1,459 '.457 1,457 3 1,457 1,603 2,720 993 911 911 82 no 120 140 142 144 146 146 240 322 326 Inches. 0.287 0.287 0.287 0.287 0.287 0.287 0.287 0.287 0.287 0.2863 0.3016 2.6400 0.5456 Inch. o. 109Z 0.1481 Broke No. 6l2. — 47.5 PARTS COPPER, 52.5 PARTS ZINC 5 25 120 480 1,320 800 0.3332 3 8o5 . .. 0.3366 790 10 0.3366 778 22 0.3366 766 34 0.3366 7S6 44 0.3366 751 49 0.3366 3 7SI 0.3364 800 0.3490 I, loo 0.1478 No. 655. — Queensland tin. 100 3 xoo 70 62 S6 100 1,50 3? 38 44 0.2109 0.2415 0.2415 0.2415 0.2415 0.3033 I Bent rapidly.- The Observed Increase of Deflection Under Static Load. — In the preceding article tlie writer presented results of an investigation made to determine the tinje required to CONDITIONS AFFECTING STRENGTH OF ALLO YS. 551 produce " set " in metals belonging to the two typical classes, which exhibit, the one exaltation, and the other a depression of the elastic limit under strain. The experiments there described were made by means of Fig. log. — Decrease of Resistance with Time. Rate of set of Bars i inch square 22 inches between supports. soo T 1 / / / v •^ / i V / ■^ t _^ j.64i WKO naoii - A f Ln- ' iniut IXAX NO. Ql c( PPEE 27.5 ITIS 72.6 69 K'KS u IKEN N0.5 N0.61 2~" rERa INAL NO. 6 2 CO 'FEE 47.6t . ZIN C.53. 32 E 'BS, 0" Jn.61 COP ■EB^ 1.50. W.ll 30 15 ■""■ — '-- N .M8 WEO GET IKOf 1st. TRI> ;,. ^ _ NC 561. —^ NO.^ y 'PER 27.6 TIM 12.5 1 a 9 S 1 " 91 I s. f \ ; I 3 i j9 s ■ fe- g S r s J ? ^ a testing-machine, in which the test-piece could be securely held at a given degree of distortion, and its effort to recover its form measured at intervals, until the progressive loss of effort could no longer be detected, and until it was thus in- dicated that set had become complete. The deductions were : That in metals of all classes under light loads this de- 552 MA TERIALS OF CONSTRUCTION— NON-FERROV S METALS. crease of effort and rate of set become less and less notice- able until, after some time, no further change can be observed, and the set is permanent. That in metals of the " tin class," or those which had been found to exhibit a depression of the elastic limit with intermitted strain, under a heavy load, i. e., a load consider- ably exceeding the proof strain, the loss of effort continued, until, before the set had become complete, the test-piece yielded entirely. And that in the metals of the "iron class," or those exhibiting an elevation of elastic limit by strain, the set be- came a maximum and permanent, and the test-piece remained unbroken, no matter how near the maximum load the strain may have been. The experiments here described were conducted with the same object as those above referred to. In these experi- ments, however, the load, instead of the distortion, was made constant, and deflection was allowed to progress, its rate being observed, until the test-piece either broke under the load or rapidly yielded, or until a permanent set was pro- duced. The results of these experiments are in striking accordance with those conducted in the manner previously described. They exhibit the fact of a gradually-changing rate of set for the several cases of light or heavy loads, and illustrate the striking and important distinctions between the two classes of metals even more plainly than the preceding. The accompanying record and the strain-diagrams, which are its graphical representation, will assist the reader in compre- hending the method of research and its results. All test-pieces were of one inch square section, and loaded at the middle. The bearings were 22 inches apart. No. 65 1 was of wrought iron from the same bar with No. 648.* This specimen subsequently gave way under a load of 2,587 pounds. Its rate of set was determined at about 60 per cent, of its ultimate resistance, or at 1,600 pounds. Its de- flection, starting at 0.489 inch, increased in the first minute, 0.1047; ill tlie second minute, 0.026; in the third minute, * Trans, Am. Soc, C, E., vol. v.. page 208. CONDITIONS AFFECTING STRENGTH OF ALIO VS. S S3 0.0125; in the fourth minute, 0.0088; in the fifth minute, 0.0063; and in the sixth minute, 0.0031 inch; the total de- flections being 0.5937, 0.6197, 0.6322, 0.641,0.6473, and 0.6504, inch. In the succeeding 10 minutes the deflection only increased 0.0094 inch, or to 0.6598 inch, and remained at that point without increasing so much as o.oooi inch, although the load was allowed to remain 344 minutes untouched. The bar had evidently taken a permanent set, and it seems to the writer probable that it would have remained at that deflection indefinitely, and have been perfectly free from liability to fracture for any length of time. This bar finally yielded completely, under a load of 2,589 pounds, deflecting 4.67 inches. No. 479 was a bronze bar containing 3^ per cent, of tin. Its behavior may be taken as typical of that of the whole " tin class " of metals, as the preceding illustrates the behavior of the " iron class " under heavy loads. It was subjected to, two trials, the one under a load of 700 and the other of 1,000 pounds, and broke under the latter load, after having sus- tained it ij^ hours. The behavior of this bar will be con- sidered especially interesting, if its record and strain-diagram are compared with those of No. 599, previously given, which latter specimen broke after 121. minutes, when held at a con- stant deflection of 0.5456 inch ; its resistance gradually falling from an initial amount of 1,233 pounds, to 911 pounds at the instant before breaking. This bar. No. 479, was loaded with 700 pounds " dead weight," and at once deflected 0.441 inch. The deflection increased 0.118 inch in the first five minutes, 0.024 in the second five minutes, o.oiS in the second ten minutes, 0.17 in the fourth, 0.012 in the fifth, and 0.008 inch in the sixth ten minute period, the total set increasing from 0.441 to 0.65 inch. The record and the strain-diagram show that at the termination of this trial the deflection was regularly increas- ing. The load was then removed and the set was found to be 0.524 inch, the bar springing back 0.126 inch on removal of the weight. The bar was again loaded with 1,000 pounds. The first 554 MA TERIALS OF CONSTRUCTION— NON-FERROUS METALS. deflection which could be measured was 3.118 inches and the increase at first followed the parabolic law noted in the pre- ceding cases, but quickly became accelerated ; this sudden change of law is best seen on the strain-diagram. The new rate of increase continued until fracture actually oc- curred, at the end of I j^ hours, and at a deflection of 4. 506 inches. This bar was of very different composition from No. 599 ; it is a member of the " tin class," however, and it is seen, by- examining their records and strain-diagrams, that these specimens, tested under radically different conditions, both illustrate the peculiar characteristics of the class, by similarly exhibiting its treacherous nature. No. 504 was a bar of tin containing about 0.6 per cent, of copper — the opposite end of the scale — and exhibited pre- cisely similar behavior, taking a set of 0.323 inch under no pounds and steadily giving way and deflecting uninterruptedly until the trial ended at the end of 1,270 minutes, over 21 hours. This bar, subsequently, was, by a maximum stress of 130 pounds, rapidly broken down to a deflection of 8.1 1 inches. No. 501 presents the finest illustration of this phenomenon yet met with by the Author. The test extended over nearly 2]4 days under observation, and the bar left for the night was found next morning broken. The time of fracture is there- fore unknown, as is the ultimate deflection. The record is, however, sufficient to determine the law, and the strain- diagram is seen to be similar to that of the second test of No. 479, exhibiting the same tendency to the parabolic shape and the same change- of law and reversal of curvature preceding final rupture, and illustrates, even more strikingly, the fact that this class of metals is not safe against final rupture, even though the load may have been borne a considerable time, and have apparently been shown, by actual test, to be capable of sustaining it. A strain-diagram of each of the latter two bars is exhibited on a reduced scale to present to the eye more strikingly this important characteristic. A comparison of the records and the strain- diagrams with CONDITIONS AFFECTING STRENGTH OF ALLO YS. 555 those of the preceding article, in illustration of the behavior of the two classes of metals under constant deflection, is most Fig. no. — Increase of Deflection with Time. Rate of Set of Bars i Inch Square 22 Inches Between Supports. 6-. 1.76 *j£ jOP( DUni tBDU «J4 CALE (Jo* 1, --a 50A. SOP^ ;ll9. OTl yu.< — e ou; tf ^ — -1 .o.^ 1,*.^ a « r ! § £ s . . 8 ° s c 5 S _j 1 «» IW*> ■i^ v)l ■ "^ ?SI S'M ■'• . . — s. 2.75 a.60 -lU] h'j^ ~ \^ I'lis. Jl K s CJ; / ) TEi HO. itrs 04. i. 1.76 1.50 1.26 0.76 0.50 0.25 Oi25 ■< l-~ ISH S^ OMI — — 21 "= = t '^ 5 >ll> IIB jis. 5 MIT* I">T3S EREJ 0.55 TI> 99.4- 30; ADj U I'U ^— 1 . — — NO si ~x/ *■ yprf 'S(2 rn.Ti lAI.) 60-P juNH >■>- r' iS.c PPBI 96.2 TM ji.1« OAD 700 ODTi 10.50 .COP 'EB .7UT f^ "^Z ^. ^B= ^^ ' ~N0 ,651. /BO\J SHT ROK LOA . 1650 poiJkds » I ■ I 7t 7* ; ■: => a- ^ ; i fi (= J" 1 ^ ; s ; =" ^ ^ instructive. It will still be necessary to make many experi- ments to determine under what fraction of their ultimate resistance to rapidly applied and removed loads, the members of the " tin class " — the viscous metals — will be safe under static permanent loads. 556 MA TERIALS OF CONSTRUCTION— NON-FERROUS METALS INCREASE OF DEFLECTION WITH TIME. Bars, t inch square ; 22 Inches between supports. Load applied at the middle. § INCREASE. g INCREASE. M Q TllyiE. TIME. Difference. Total. Difference. Total. No. 651. — Wrought iron. Min. Inches. Inches. Inches. 30 0.618 0.017 0.177 Loady 1^600 pounds. 40 0.630 0.012 0.189 Min. Inches. Inches. Inches. 1° 0.642 0.012 o.2or I 0.4890 0.5937 0.1047 0.1047 60 Set 0.650 0.524 0.008 0.209 2 0.6197 0.0260 0.1307 3 0.6322 0.0125 0.1432 Second Trial.— Load, ^-floo pounds. 4 0.6410 0.0088 0. 1520 S 0.6473 0.0063 0.1583 3. 118 6 0.6504 0.0031 0.1614 5 3.540 0.422 0.422 16 0.6598 0.0094 0.1708 IS 3.660 0.120 0.542 344 0.6598 0.0000 0.1708 45 4.102 0.442 0.984 Maximum load, 2,589 pounds; maximum 75 ,. 7-634 bar under i. 3.522 4.506 deflection, 4.67 inches. Broke Doo pounds. No. 504. — O.S57 PARTS COPPER, 99.443 PARTS TIN. No. 5or. — 9.7 PARTS COPPER, 90.3 PARTS TIN. Load^ 110 pounds. Load^ 60 pounds. 0.323 1.294 5 0.406 0.0S3 0.083 10 1.319 0.025 0.025 Sl5 '•945 1-539 1.622 70 1.463 0.144 0.169 865 2.005 0.039 1. 681 130 1.530 0.067 0.236 895 2.138 0.134 - I. 815 310 1. 691 0.161 0.397 1,025 2.248 0. no 1.925 400 1.766 0.075 0.472 1,110 2,378 0.130 2.055 460 1. 811 0.04s 0.517 1,270 2.626 0.248 2.303 1,360 2.534 0.723 1.240 Maximum load. 130 pounds; maximum deflection, 8.11 inches. 1,475 2.697 0.163 1.403 1,565 2.782 0.085 1.488 1.644 1,730 2.938 0.156 1,880 3-136 0.198 1.842 No. 479.-96.27 PARTS COPI'ER, 3.73 PARTS TIN. 2,780 2,940 3.798 4.274 0.662 0.476 2.504 2.980 Load^ 700 ponnds. 3>o°° 3j?95 , Bar 1 4-349 5.097 0.07s 0.748 3.055 3.803 0.441 eft under st rain at nigh t and found 5 0.5S9 o.iiS o.iis" bro ten in the m 3ming. 10 0.583 - - 0.024 - 0.142 20 0.601 0.018 0.160 Depression of Elastic Limits. — The effects of inter- mitted stress and of interrupted strain are of peculiar interest and importance with the non-ferrous metals and the alloys. So far as they have been observed by the Author, they are often precisely the opposite of those noted in experiments on merchant iron and commercial grades of steel. They are well illustrated in Fig. iii, which is here reproduced from Part II., Mat. of Eng. CONDITIONS AFFECTING STRENGTH OF ALIO YS. 557 These strain-diagrams are obtained by transverse test, from bars of common iron, Nos. 648, 649, 650, 651, and from two specimens of bronzes, Nos. 596, 599, all of the same size. 1-1 o H < w Iz o i § i § 3 3 § fe ?■! S-0 S o ■* ^ = O S'Et isr ISt 9-tI rit 9-01 rot O'S Id 9-g,, i:» 'A rfl 9-9 0-9 e'9 1-3 s-f' 0-t 9X1 O'S 5-S o:3 I inch (2.54 cm.) square and 22 inches (55.9 cm.) between sup- ports. The first strain diagram to be studied is that of a bar of the most ductile metal (No. S99, copper, 10 ; zinc, 90). It exhibits clearly the phenomenon of flow with a depression of the elastic limit under constant load. S^SmA TERIALS of construction— NON-FERROUS METALS. This bar was left deflected under a load of 163 pounds (74 kgs.). It gradually lost its power of restoration until it only exhibited an effort of 143 pounds (65 kgs.). The curve exhibits the relation of deflection to deflecting force. The resistance gradually increased as deflection progressed until the load — 403 pounds (183 kgs.) — produced a deflection of 0.09 inch (0.23 cm.). The bar was again left, and, under a fixed deflection, again lost resisting power, and the effort to straighten itself fell to 333 pounds (151 kgs.). Finally, the bar offered its maximum resistance of 1,233 pounds (560 kgs.) under a deflection of 0.545 inches (1.3 cm.), and was then held in its flexed position. Gradually its effort to restore itself grew less and less, until, when it had fallen to gi I pounds (414 kgs.), the bar suddenly snapped and the two halves fell to the floor. No. 596 (copper 25, zinc 75) similarly exhibited a depres- sion of the elastic limit by strain, but, vastly harder, more elastic and brittle, it broke under 663 pounds (301 kgs.) and at a deflection of 0.1136 inch (0.3 cm.), before apparently pass- ing the point termed the primitive or apparent limit of elastic- ity by the Author, i. e., that point at which the sets become nearly proportional to the strains, and at which the line of the strain-diagram turns sharply away from the vertical. The strain-diagram No. 648, common iron, is that of the type of that class in which the elevation of the elastic limit has been detected by the Author. The bar was^ like the preceding, of I inch (2. 54 cm.) square section and 23 inches (55.88 cm.) in length between bearings. It reached its elastic limit at 1,450 pounds (659 kgs.) and at a deflection of 0.15 inch (0.4 cm.). Pass- ing this point, and at a deflection of 0.287 inch (0.7 cm.), the bar was held at a constant deflection, under a load of 1,600 pounds (727 kgs.). Flow occurring, the effort to regain its original shape became less and less, until in six hours it had fallen to 1,457 pounds (662 kgs.). Continuing the test, re- sistance and deflection increased as indicated by the curve, instead of following the original direction. Similar increase of resisting power under strain is seen at CONDITIONS AFFECTING STRENGTH OF ALIO VS. 5 59 other points on the curve, and whenever the process of dis- tortion was interrupted long enough to permit flow and that re-arrangement of particles which has been described. An hour or two usually gave time enough to bring out this re- markable phenomenon. This action has been discovered in iron and steel, and under every form of strain — tension, torsion, compression and cross-breaking — and it would seem that aside from accidental overstrain, producing incipient rupture or loss of strength due to such action as abrasion or corrosion, length of life of iron structures under strain was in itself, apparently, a source of increased safety. On the other hand, as is here seen, the be- havior of non-ferrous metals is precisely the opposite, and the engineer is compelled to use them with greater caution and to base his calculations upon a higher factor of safety, a conclusion fully corroborated by the work of Wohler. Recurring to Fig. iii,a resemblance is to be noted in the behavior of both classes of metals. The bars No. 649, 650 and 65 1 were tested by rapidly in- creased load up to the breaking point, allowing no time for reading of sets. The first of this set deflected 0.014 inch (0.04 cm.) under 100 pounds (45 kgs.), 0.052 under 500 pounds, 0.098 under 1,000 pounds, and 0.18 under 1,500 pounds. At 1,600 pounds the deflection was 0.2854 inch, and the bar yielded to the stress, and the deflection became 0.363 in 2^ minutes. Under 1,640 pounds the deflection increased in six minutes from 0.383 to 0.440 inch, and a maximum resistance was recorded of 2,350 pounds (1,070 kgs.), and a deflection of 5.577 inches (15 cm.). This bar was tested in a similar manner to the preceding, and in the same machine. Numbers 650 and 651 were tested by dead loads — i.e., by laying upon them heavy weights. By this method the deflection could increase to a maximum under each load, in- stead of being kept constant, as in the testing machine. No. 650 was rapidly broken without allowing time for completion of set or any considerable exaltation of the elastic limit. The plotted curves of results exhibited well the striking difference 56o MA TERIALS OF CONSTRUCTION— NON-FEKROUS METALS. of behavior between this bar and 651, which was purposely given time for set and for exaltation of the elastic limit. At 1,500 pounds (682 legs.) each had deflected nearly the same atnount, and had passed the elastic limit, as usually called. The first, however, gave way completely with 2,260.5 pounds (1,027 kgs.), while the second, after several times exhibiting an elevation of the elastic limit — as at 1,500, 1,600, 1,700, 1,900, 2,300, 2,400 and at 2,500 pounds — finally only yielded entirely at 2,589. The first only deflected 2^ inches (7 cm.); the second, 4.67 inches (11.9 cm.); although when the latter was loaded with about the weight at which the first yielded, it deflected about the same amount. The last bar was left two and a half days under its final load, and its deflection increased from 4.275 inches (10.9 cm.) to 4.67 (11.9 cm.), when the weights reached the supports of the frame and the test was ended. The other bar sank rapidly after being loaded with 1,600 pounds (726 kgs.). Both classes of metals, when flexed, were shown to exhibit less and less efiort to restore themselves to their original form. In the case of the tin class, as the Author has called it, this continues indefinitely. With the iron group this loss of effort gradually becomes less and less and reaches a limit at which the bar is found to become stronger than at first. The two classes are thus seen to be affected by time in precisely the same manner initially, but finally in exactly opposite ways. The Effect of Variable Stress in causing variation of the normal series of elastic limits observed during ordinary tests is well shown by the records of test of the copper-zinc alloys. The following are extracts from the memoranda taken during tests made for the U. S. Board to which fre- quent references are made. Similar illustration may be found among the records of tests, both of bronzes and of brasses, already given. Bar No. 8 (60.94 copper, 38.65 zinc) bent to a deflection of 3j^ inches under a load of 1,140 pounds. The apparent elastic limit was reached at about 640 pounds. At 400 pounds the bar was left under stress for eighteen hours, at the CONDITIONS AFFECTING STRENGTH OF ALLOYS. 561 end of which time no change had occurred. At 800 pounds, the beam fell and the resistance decreased 18 pounds in an hour. It finally broke under about 1,200 pounds. Bar No. 19 (10.30 Cu., 88.88 Zn.) at 1,233 pounds yielded. In an hour and a half the decrease of resistance was nearly 250 pounds, the decrease of resistance then in twenty-two minutes being 82 pounds. In three minutes after taking the last reading, when it balanced at 911 pounds, the bar sud- denly broke without warning. The deflection was unchanged during this entire " time test." The elastic limit was reached at about 900 pounds. The Effect of Repeated Strain is greater with the non-ferrous metals, and usually with the alloys, than with iron and steel. The investigations of Wohler and Spangenberg were made principally upon the latter class of materials, but were also made to cover the action of a few other metals. Wohler's law, that the rupture of a piece may be pro- duced by the repeated action of a load less than that which, once applied, would cause fracture, is true, probably, of all the non-ferrous metals, and this effect is with them much more serious than with the ferrous metals. Spangenberg found that gun bronze in tension would endure a stress of 22,000 pounds per square inch (1,547 kgs. per sq. cm.) laid on and at once removed 4,200 times before rupture ; a stress of 16,500 pounds (1,160 kgs.) 6,300 times, and 11,000 pounds per square inch (773 kgs. per sq. cm.), 5,547,600 times. It may be con- sidered safe under indefinitely repeated loads falling well under one-half its tenacity as determined by ordinary test. Phosphor bronze, forged, bore 53,900 repetitions of the small- est of the above loads, and 2,600,000 of the next load, but broke under 1,621,000 repetitions of a load of 13,750 pounds per square inch (967 kgs. per sq. cm.). The cast metal sus- tained 408,350, 2,731,161 and 2,340,000 repetitions of the same loads. This peculiar behavior is not explained by the experimenter. Further experiment in this direction is desirable. Mean- time, the engineer will probably find it advisable to allow, for intermittent loads, but one-half the stresses which would be 36 562 MA TERIALS OF CONSTRUCTION— NON-FERROUS METALS. permitted for single applications of load, and one-quarter where suddenly applied, while the factor of safety should be probably not less than one-half greater for non-ferrous material than with iron. The limits of stress sometimes pro- posed are not far from the following, which may be compared with the values already given for factors of safety and ulti- mate strength. PERMISSIBLE REPEATED STRESSES FOR NON-FERROUS METALS. FACTOR OF SAFETY. MAXIMUM STRESS. Dead Load. Live Load. Dead Load. Live Load. Lbs. per sq. in. Kgs. per sq. cm. Lbs. per sq. in. Kgs. per sq. cm. 4 4 4 4 4 4 4 4 8 8 8 8 3 8 8 8 5,ooo 15,000 16,000 10,000 5,000 10,000 12,000 1,000 352 1,055 1,125 703 352 703 845 70 2,500 7,500 8,000 5,000 2,500 5,000 6,000 500 X76 528 563 352 176 352 423 35 " forged wire Gun-bronze, cast Brass, yellow, cast . . . rolled . " " wire . . Lead, rolled When the stresses are reversed, as in connecting rods, the factor of safety should be doubled and the maximum stresses reduced at least one-half. THE MATERIALS OF CONSTRUCTION. PART II. NON-METALLIC MATERIALS. CHAPTER XVIII. STONES AND CEMENTS. Use by Constructing Engineers. — The mechanical engineer makes comparatively little use of the materials of masonry. Their principal value to him is for use in foundations, and in structures protecting his machinery. Of the stones, but few are well adapted to his purpose. Using them in foundations, he requires them to be strong and dense, hard and durable ; and the mortars and cements used to bind them should be of the best possible quality. A foundation intended to bear the shock and tremor of moving machinery must, necessarily, be more carefully built, and must be constructed of more carefully selected materials than a foundation carrying a load at rest. The civil engineer makes frequent use of all the materials of masonry. There is given here a summary of the most important characteristics of these materials. For a more detailed account of them, reference may be made to special works on civil engineering. The mason uses for foundations natural stones which are either silicious, calcareous, or argillaceous in their character, and artifical stones, including the several varieties of brick. For special construction, he sometimes employs materials not in common use, as fire brick and fire clay. These materials are very rarely used in structures, to resist other than compressive stress. 564 NON-METALLIC MATERIALS. Classification of Stones. — The system of Durocher and Bunsen classifies the stones by reference to their proportions of silica, and under a geological classification into igneous, metamorphic, sedimentary, and calcareous. The igneous orplutonic rocks were produced by original solidifica- tion from fusion or by later volcanic action. The metamorphic rocks are sedimentary, but have been altered by heat, pressure, chemical, and other agencies. The sedimentary ' rocks are the result of the abrasion of older rocks by water, and subsequent condensation and solidification under pres- sure. The calcareous rocks are composed of corals, as the " Coquina," or of the shells of marine animals which have usually, under the pressure of superincumbent rock and soil, become so compacted as to have lost their form, and to have united into dense and granular masses. Silicious Stones. — The granitic group of igneous rocks is richest in silica, and its members are known as sili- cious stones. Of this class, granite, sienite, gneiss, green- stone, and trap, and the harder varieties of sandstone, are most commonly used. Granite is a primary rock, and underlies the stratified rocks ; it is, itself, unstratified. It is of a compact crystalline structure, and is composed of quartz, feldspar, and mica ; its principal impurities are talc and hornblende. Its quartz is in the form of colorless or gray crystals; its feldspar is in opaque-white or flesh-colored crystals; and the mica appears as shining scales or grains. Its quality varies considerably according to the propor- tions and mode of aggregation of its constituents. The greater the proportion of quartz, and the less the proportion of feldspar and mica, the greater the durability and hardness of the stone. Feldspar renders the stone lighter in color and easier to cut, and more susceptible to decomposition by solution of the potash contained in it ; mica renders it fria- ble ; and hornblende heavy and tough. The best kinds of granite are the most valuable of all our building stones, and it is extensively used for important works. Under exposure to fire it is, however, less durable STONES AND CEMENTS. 565 than many, in other respects, inferior stones. It is easy to quarry, and blocks may be obtained of any size. It is diffi- cult to work, and therefore very costly to use if the stone must be dressed. Its specific gravity is about 2.66, and its weight 166 pounds per cubic foot, or nearly 2 tons per cubic yard (2659 kilogrammes per cubic metre). Granite is found throughout all the Eastern States, in Canada, in many parts of the Alleghanies and Rocky Moun- tains, and usually wherever the later rock formations have been removed and left the underlying beds exposed. It is generally classified into gray and red. Gray granite in immense quantities is found in Maine. That from Dix Island is very hard and strong. The granite of Hallowell has a greenish tint, and is very light-colored when first cut ; it is fine-grained and durable, yet easily worked. Morgan's Bay granite is very strong, easily worked, and light-colored. Round Pond granite is fine-grained and polishes well. Gray granite of good quality is also found at Mt. Desert, Clark's and Fox Islands. In New Hampshire, the "Granite State," a fine granite is found at Pelham. Much of that found in this State is coarse, micaceous, and not very durable. In Vermont, the Barre granite is perhaps the best ; it is light-colored and homogeneous. In Massachusetts, the Quincy granite has the greatest reputation. It has a dark bluish-gray color, is very durable, and takes a fine polish. In Rhode Island, at Westerly, is found granite of the very best quality; it is fine-grained, containing small particles of horn- blende and black mica. It is much used for monumental purposes. In Connecticut, a very fine-grained granite is quarried at Mystic Bridge. That obtained at Millstone Point is rather dark in color, compact, with comparatively little hornblende or mica. In New Jersey, an excellent granite is found at Newfoundland. Red granite of excellent quality is found along the Bay of Fundy. It is composed of red orthoclase, bluish quartz, and a little hornblende, with scarcely any mica. It is hard, and takes a fine polish. Similar granite is found on Forsyth's Island in the St. Lawrence. The Calais red granite comes 566 NON-METALLIC MATERIALS. from near the St. Croix River in Maine. Very fine red granites are found at Lake Superior, and at many points in the Rocky Mountains. The Scotch Peterhead red granite is used considerably, but is not greatly superior to many of our native stones. Talcose Granite, or Protogine, is one in which the mica is replaced by talc. Sienite was first quarried at Siena or Syene in Egypt, whence its name. It is a granular rock closely resembling granite, and consists of feldspar and hornblende, with fre- quently quartz and mica. It is hard, tough, rather coarse- grained, and not susceptible of taking a polish. When the feldspar is not too readily decomposed by the removal of its potash when acted upon by the weather, it is the most durable of our granitic rocks, and affords an excellent material for rough and substantial work. Before making use of this stone, if it is obtained from a new quarry, its quality should be care- fully tested. Gneiss and Mica Slate are similar to granite in com-, position, are metamorphic rocks, and are found stratified. Granite, sienite, and gneiss resemble each other so closely that they are all frequently called granite by those who are not familiar with mineralogy. Gneiss is less valuable than granite, as it cannot be obtained as readily in blocks of definite size, and as it does not as readily split in directions other than that of stratification. It is, however, a good building material, and, for the purposes of the engineer, answers frequently as well as granite. Sienitic gneiss is a stratified sienite in which hornblende takes the place of mica. A very slightly stratified gneiss is found at Concord, Nev/ Hampshire, and there are quarries of dark compact gneiss at Greenwich, Connecticut. These stones are all much affected in quality by the pres- ence of foreign elements. The oxides of iron are particularly injurious, causing discoloration and serious disintegration of the stone. Quarries of excellent granite, sienite, and gneiss STONES AND CEMENTS. 567 are found throughout the mountain ranges which extend along the coast of the United States. Greenstone, Trap, and Basalt are plutonic unstra- tified rocks, consisting of hornblende and feldspar.* In the first named, the grains are not as coarse as in granite, and the others hardly exhibit the granular structure to the eye. The first two break up into blocks, and the latter into columns of prismatic form. They are found in veins and dykes, injected among stratified rocks of all ages. In color, they vary from nearly white, in some varieties of greenstone, to as nearly black in basalt, the difference of color being determined by variations in the proportion of hornblende and feldspar, the former being dark, the latter light, in color. The green of these stones is due to the chromium which is present. The " Palisades," the bluff skirting the western shore of the Hud- son River, opposite and above New York, is composed of trap rock. Greenstone also is found in considerable quantities in the same locality, and in the Orange Mountain of New Jersey. These stones make a very durable building material, but cannot be obtained in large blocks, and are very difficult to cut. They form excellent metalling for macadamized roads, and the last two are especially useful for block pave- ment. The Porphyries contain from 60 to 80 per cent, silica, 10 to 20 per cent, alumina, and the remaining constituents are iron-oxide, lime, magnesia, soda, and potash in small propor- tions, and have some value in construction. Their specific gravity ranges from 2.5 to 2.6. They are used principally for ornamental purposes. A fine quality of this stone is obtained near Boston. Sedimentary Rocks. — Sandstone was formed in all of the later geological periods ; it is a stratified rock, consisting of grains of sand derived from the disintegration of silicious rock, and cemented together by a natural cement composed usually of silica, lime, and alumina- In the most durable kinds of sandstone, this cement is found to be almost pure '•■ School of Mines Quarterly makes Trap the generic name for these rocks. $68 NON.METALLIC MATERIALS. silica, and in the weaker stones it consists quite largely of alumina. Lime, when present in this cement, renders the stone peculiarly liable to disintegration when exposed to an impure atmosphere, or when it is used for foundations washed by water containing acids capable of attacking the lime. The presence of clay (silicate of alumina) or of protoxide of iron, is very injurious. Sandstones vary in quality between very wide limits, some being nearly as valuable as granite, and other kinds being the most friable of our building stones. A small grain, a minute proportion of cementing material, and a sharp, clear, bright fracture are the characteristics of the best stone. Such stone is usually found in thick beds, exhibiting but slight evi- dence of stratification. Water readily penetrates between the layers of this stone. In foundations, care should therefore be taken to lay it on its " natural bed," that such penetration of moisture and conse- quent disintegration by freezing, may, as far as possible, be prevented. The strength and durability of good sandstone, together with the ease with which it is cut and dressed, makes it the most commonly used of all our building stones. It is fre- quently denominated " free-stone," from the facility with which it is worked, and " brownstone " when of that color. The colors of sandstones vary greatly. The Ohio or Am- herst and Nova Scotia sandstones are of a yellowish or cream- color, or nearly white. Missouri furnishes a durable stone of a fine yellowish drab color. The stones from Portland, Con- necticut; Newark, New Jersey; and from Marquette and Bass Islands in Lake Superior, are dark brownish red. A purplish red stone is found along the Rappahannock and Acquia Creek, which is, however, not of the best quality. Avery hard and durable, highly silicious, reddish sandstone is ob- tained at Potsdam, New York. A dark brownish stone is found at Hummelstown, Pennsylvania. Soapstone, the silicate of magnesia, is a very widely distributed stone in the United States, and is very valuable STONES AND CEMENTS. 569 where a stone capable of resisting high temperatures is re- quired. Calcareous Stones consist largely of lime; lime- stones and marbles are familiar examples. Limestones — carbonates of lime — effervesce freely when attacked by acids which are capable of displacing carbonic acid from its combinations. The carbonic acid may also be expelled by high temperature. In the former case, new lime salts are formed by the union of the lime with the attacking acid ; in the latter case, caustic lime, " quick-lime," is left un- combined. The limestones vary greatly in their qualities as building materials. While some are as strong and hard as granite, others are as soft and friable as the weakest sand- stone. They are, usually, easily worked. There are two sorts : the granular and the compact ; both of which classes yield excellent stones. Those limestones which take a smooth surface and a fine polish are usually called marble ; the coarser kinds are known as common limestone. The granular varieties are generally superior in quality, for building purposes, to the compact. The impure carbonates of lime are sometimes of great value. The magnesian limestones, or dolomites, are frequently found to be exceptionally excellent. Chalk is a soft limestone in which pressure has not usu- ally wholly destroyed the organic textufe of the minute shells of which it is composed. It is generally too soft for constructive purposes, and can only be used for making lime. All the varieties of calcareous stone are found in the United States, one of the most extensive deposits being that which follows a line parallel to the Atlantic coast, and near the deposit of primary rocks already referred to ; another un- derlies the Middle States. The marbles are mostly confined to mountainous districts ; the common limestones are often found in immense strata as deposited on the bed of the ancient ocean. Marbles are usually classified into white and colored, or variegated. 570 NON-METALLIC MATERIALS. White marble is found in the Laurentian rocks of Canada. Most of the white marble used in the Northern Atlantic States comes from the Green Mountain range extending through Vermont, western Massachusetts, western Connecti- cut, and south-eastern New York. Valuable quarries exist in Vermont, at Brandon, Rutland, Danby, Dorset, and Manchester ; in Massachusetts, at Lanesborough, Lee, Stock- bridge, Great Barrington, and Sheffield ; in Connecticut, at Canaan ; and in New York at Pleasantville and Tuckahoe. Lee marble was used for the extension of the Capitol at Washington, and for the City Hall at Philadelphia. Pleasant- ville and Tuckahoe marbles were used for the St. Patrick Cathedral, New York. The Pleasantville quarries supply the " Snowflake " marble. A fine quality of statuary marble is found at Rutland, Vermont, which is almost equal to the Italian marble of Europe. The marbles become coarser, harder, and more suitable for building purposes as we proceed south from Rutland. In south-eastern New York they are of a dolomitic character. In Delaware is found a coarse dolo- mitic marble which resembles the Tuckahoe. Colored marbles. — In Vermont, on the shore of Lake Champlain, is found a brecciated marble, and a dove-colored marble with greenish streaks is quarried at Rutland. Shore- ham, Connecticut, supplies a black marble, and a fine marble of the same color is obtained at Williamsport, Pennsylvania. The black Trenton limestone is found at Glen Falls, New York. The Warwick marble from Orange Co., New York, is very beautifully colored with carmine in different shades, veined with white. The Knoxville marble is of a reddish hue, with lines of blue. The Tennessee so-called marble is mottled with chocolate and white. Of foreign marbles, the best known are — the Brodliglio of Italy, in gray, shaded with black ; the Sunna of Spain, a pal^ yellow ; the Lisbon of Portugal, a pale reddish color ; and the Belgian black of Belgium. Verde Antique is composed of bands of Serpentine and white marble. The limestones are also of special interest in furnishing, by calcination, the several varieties of lime and cements. STONES AND CEMENTS. 57 1 Common Lime is obtained from the limestones of the middle and western States, from Maine and other States of New England ; hydraulic lime is made in the State of New York ; and many other localities furnish excellent lime or cement. The limestones are indispensable to the iron-maker, also, as affording him an alkaline ba^e, which, uniting chemically with the silica and alumina, and other impurities of the ores, allow the metal to separate in a state approximating pu- rity ; the silica and lime form the slag, which is also useful as a protecting covering over the molten iron, while it re- mains in the blast furnace. For such purposes a limestone free from magnsia or other impurity is generally most val- uable. Gypsum, Alabaster, or " Plaster of Paris " is a sulphate of lime, containing some water of crystallization. It takes the latter name from the fact that large deposits of this stone, of excellent quality, underlie the City of Paris and its environs. When pure it contains, lime, 32.6; sulphuric acid, 46.5 ; water, 20.9. It is translucent and colorless. Its specific gravity is 2.6 to 2.8. When raised to a high temperature, it loses its water of crystallization, and, if then finely powdered and made into a thin paste with water, it may be readily moulded and worked into any desired form. It then quickly " sets " or hardens. It is thus beautifully adapted for use in making casts and ornaments, and has recently been largely used in special branches of pattern making. It also makes an excellent cement for many purposes. The engineer uses it in making models and patterns, and for moulds which are not to be sub- jected to great heat. This stone is found in many parts of our country. Con- siderable quantities are quarried in the State of New York. Argillaceous Stones are of little value, as a rule, for the masons' work ; they are generally weak, soft, and readily decomposed by the action of the weather. Clay slate is a sedimentary, argillaceous rock, of fine- 572 NON-METALLIC MATERIALS. grained, compact, and laminated structure, and is usually of dark shades of purple, blue, and green. The best varieties of clay slate, or roofing slate, and of grauwacke-slate are used for roofing and flagging. The former are obtained of remarkably fine quality in the State of Vermont ; the latter, on the banks of the Hudson. " Fire-Stones " are those capable of resisting the action of great heat, neither fusing, exfoliating, nor cracking. Lime and magnesia, except where existing as silicates, are injurious in fire-stones. Potash is very injurious, increasing the fusibility of the stone, and, melting, causing the formation of a fusible glass. Quartz and mica, alone or in combination, make stones of great infusibility. Mica-slate and gneiss are examples of excellent combina- tions, the latter, more particularly, when containing a con- siderable proportion of arenaceous quartz. Limestones do not usually withstand the effect of heat well. The heat of the fire calcines it deeply where exposed, and thus destroys walls built of it. In the midst of great conflagrations, as at Chicago in 1 87 1, the carbonic acid with which the lime is combined has been expelled so rapidly as to produce violent and explosive disruption. Magnesian limestones are little if any better than the pure limestones. Oleiferous limestones, containing silica and alumina, some- times, however, resist fire moderately well. Granite, gneiss, sienite, quartz, mica-slate, and other pri- mary rocks usually contain some water; when exposed to fire, they crack and even explode. Walls constructed of these stones are apt to crumble rapidly in a hot fire. Sandstones free from feldspar, somewhat porous and un- crystallized, are the most refractory of common building stones. Brick, especially when approximating to fire-brick m composition, is an excellent fire-resisting material, perhaps the best now known. Concrete and b^ton, even when well made and completely set, and artificial stones consisting principally of silicates of STONES AND CEMENTS. S73 lime and alumina, are not generally very good heat-resisting materials. Thin walls of any known building material will rapidly crumble in the midst of a large fire. The natural stones are less generally used where refractory materials are required, than are fire-brick,' the characteristics of which latter will be described hereafter. The Hardness of Stones is measured by comparison with some well-known stone of which the hardness is taken as a standard. The mineralogist's scale of hardness is the following, the softest being i, and the hardest lo, on the scale. HARDNESS OF MINERALS. Diamond lO Ruby g Cyraophane 8.5 Topaz 8 Spinell 8 Emerald 8 Garnet 7.5 Zircon 7 Quartz 7 Tourmaline 7 Opal 6 Turquois 6 Lapis lazuli 6 Feldspar 6 Amphibole 5.5 Phosphorite 5 Fluor-spar 4 Coclestine 3.5 Barytas 3.5 Carbonate of lime 3 Mica 2.5 Gypsum 2 Chlorite 1.5 Talc I The cortiparative resistance of stones to abrasion is as follows* : RESISTANCE OF STONES TO ABRASION. Statuary marble 100. Bath stone 12. Stock brick 34 • Roman cement stone 69. Old Portland cement stone. . 79. Kilkenny black marble no. Yorkshire paving stone 327. Aberdeen granite 980. The Strength of Stone to resist Crushing varies "^ Journal Franklin Institute, Oct., 1835. 574 NON-METALLIC MATERIALS. immensely with different classes, and often very considerably even in the same class. Gen. Q. A. Gillmore has shown that under a maximum varying for each specimen, the resistance varies as the cube root of the length of the side of cubes ex- perimented on. The densest stones are usually strongest.* For Berea sandstone, Gilmore gives y = a 3/~ in which y = crushing resistance per square inch of surface in pounds ; X — the length of the side of the cube in inches, and. a = from 7000 to 9500. (For metric measures : kilogrammes per square centimetre and centimetres, a = 350 to 500). STRENGTH OF STONES UNDER COMPRESSION. Granite and sienite " " " good quality Basalt and trap Limestone and marble. Best sandstone, average Conn, and New Jersey sandstone, ordinary Slate KILOGRAMMES PER SQUARE METRE. 7,655,000 to 10,936,000 9,842,000 8,749,000 7,655,000 5,468,000 3,827,500 5,468,000 3,827,500 to 2,187,000 to TONS PER SQ. FT. 700 to ICOO 900 800 700 500 500 350 to 200 to Trautwine ta^es the average weight of granite at 165 pounds per cubic foot (about 2643 kilogrammes per cubic metre), and of ordinary sandstone at 145 pounds per cubic foot (2323 kilogrammes per cubic metre), and calculates the height of column it would stand without crushing at the base as 8cxD0 feet (2440 metres), and 4000 feet (1220 metres) respec- tively, for these two kinds of stone. ^ The weight imposed ordinarily on foundations of struc- tures is rarely ten tons per square foot (109,360 kilogrammes per square metre). Stones of a soft quality usually begin to ^ Journal Franklin Insiiluk, Vol. LXV., p. 336. STONES AND CEMENTS. 575 crack at not far from one-half their crushing loads. The hardest stones yield suddenly and completely when the crush- ing load is reached. Stones crush most easily under soft materials, and offer most resistance between surfaces of hard steel. Polished stones resist better than rough. Transverse Strength. — Experiments made upon building-stones by Mr. R. G. Hatfield are recorded in the Transactions of the American Society of Civil Engineers for 1?>'J2 et seq. The formula W = S —j^ is used in expressing transverse strength. 5 = weight in pounds required to break a bar one inch square and one foot long between bearings ; L = length in feet ; b and d measure the breadth and depth of the sample in inches. The coefficient, S, varies from 250.76 to 227.19 for Hudson River blue-stone ; for Kingston it is 203.50 ; for other speci- mens of blue-stone or grauwacke the coefficient falls as Ipw as 122.31. Marble from East Chester gives about 150." Sand- stone from Belleville, New Jersey, gives coefficients varying from yS to 88.472 ; from Portland, Connecticut, 64.7 to 94 ; from Dorchester, Nova Scotia, 63 to Sy. Ohio light sand- stones give from 32 to 62. For weights in kilogrammes, lengths in metres, and Jsreadth and depth in centimetres, the values of S become eight one- thousandths those above given. The factor of safety should never be less than ten, and a far higher value is adopted in practice by many experienced engineers. At the same time, it should be remembered that the crushing strength of the mortar or cement, used as the binding material, in many cases, rather than the strength of the stone itself, determines this limit. The Durability of Stones varies as greatly as do their other qualities. It is determined by strength, hardness, and porosity. Experience and careful observation are the most reliable guides in judging of durability ; but a series of care- fully conducted tests may be made to yield valuable aid in estimating the values of newly opened quarries. In all im- portant work a careful test should be made of the stone pro- 576 NON-METALLIC MATERIALS. posed to be used. The quantity of water absorbed gives a good comparative test of the durability of stones of the same class, as does also their greater or less clearness of sound when struck with a hammer. The following are the percentages of their own weight absorbed by different stones : Granite absorbs from \ per cent, to i per cent, of water. Gneiss " " i " " " 2^ " " Sandstone " " li " " " 3 " ' Fine-grained stones absorb least water, and are usually hardest and best. M. Brard's method* of determining power of resisting the action of frost, is as follows : A cubical block, containing about 8 cubic inches (131 cubic centimetres), is sawn from a stone to be tested. This is suspended in a boiling supersatu- rated solution of sodic sulphate during a period of thirty minutes. It is then taken out, suspended above the liquid, and allowed to cool. This is repeated daily, or oftener, for at least a week. The crystallization of the salt within the pores of the stone disintegrates it, precisely as does the action of frost. The weights of the specimens tested are carefully taken before and after the test, and their comparative value is thus learned. The chemical composition of stones greatly affects their durability. Potash is apt to wash out, leaving the stone soft and friable ;' clay absorbs water and thus softens stones of which it is a constituent ; and iron, as has already been stated, injures by discoloration, and by the disintegration produced by changes from protoxide to peroxide. Hatfield gives the following as the result of his examina- tion of building-stones : , *■ * Annales de Chimie et de Physique, vol. xxxviii. STONES AND CEMENTS. 177 DURABILITY OF STONES. NO. NAME. LOCALITY. WEIGHT PER CU. FT. POUNDS. DURABILITY — YEARS REQUIRED TO DISIN- TEGRATE TO THE DEPTH OF "Ai INCH. (0.254 CM.) I 2 3 4 5 6 Sandstone Coquina Portland, Conn . . Berea, Ohio Marietta, Ohio . . . Dorchester, N. S. Amherst, Ohio . . . Florida 150.41 134.14 161.94 140.52 133-16 105 - 76 2603.3 2000.8 1794-8 811.40 306.46 6-92 Effect of Heat. — Expansion by heat takes place as fol- lows, the range of temperature being 180° Fahr. (ioo° Cent.) : Granite and sienite 0008 Sandstone 0006 to .0010. Mr. H. A. Cutting, State Geologist of Maine, experiment- ing upon the granites, finds that their specific gravity varies from 2.50 to 2.83 ; that their power of absorption of moisture decreases irregularly with increase of density from 0.00300 to 0.00125, although, the heaviest stones are not invariably least penetrable ; that they are uninjured when heated to 500° F. (260° C.) and plunged into water, but that they are all injured at temperatures below the red heat ; that at the latter tem- peratures they often crumble and become entirely worthless. The common sandstones vary in specific gravity from 2.2 to 2.6 ; they usually absorb from 0.06 to 0.25 water, although occasionally as little as 0.025, and withstand heat better than the granites, bearing temperatures 100° F. (55° C.) higher with equal safety. The limestones examined compared favorably with the sandstones. Conglomerates are not equal to either of the former sorts. " Soapstone " withstands heat much better 37 578 NON-METALLIC MATERIALS. than any other stone tested, receiving no apparent injury up to 1,200° F. (649° C). Artificial Stones comprise brick of various kinds, con- crete, b^ton, and artificial imitations of sandstone. Bricks are made by submitting clay, which has been prepared properly and moulded into shape, to a temperature which converts it into a semivitrified stone. Common brick is a most valuable substitute for stone. Its comparative cheapness, the ease with which it is transported and handled, and the facility with which it is worked into structures of any desired form, are its valuable characteristics. It is, when properly made, nearly as strong as good building- stone ; it is but slightly affected by change of temperature, or of humidity, is well cemented by mortars, and is also lighter than stone. Common Clays, of which the common brick is made, consist principally of silicate of alumina ; but they also usually contain, lime, magnesia, and oxide of iron. The latter is use- ful, improving the product by giving it hardness and strength ; hence, the red brick of the Eastern States is often of better quality than the white and yellow brick made in the West. Silicate of lime renders the clay too fusible, and causes the bricks to soften and to become distorted in the process of burning. Carbonate of lime is certain to become decomposed in burning, and the caustic lime left behind absorbs moisture and promotes disintegration. Uncombined gilica is beneficial, if not in excess, as it pre- serves the form of the brick at high temperatures. In excess, it destroys cohesion, and renders the bricks brittle and weak. Twenty or twenty-five per cent, silica makes a good propor- tion. Preparing the clay consists in clearing it carefully of peb- bles, and after mixing it with about one-half its volume of water, "tempering" it either in a "pug-mill" or by hand stirring. The clay is moulded into bricks by pressing it into forms, either by hand or by a machine, and they are then piled and burned, after having been well dried in the open air. STONES AND CEMENTS. t^-jg Burning occupies about two weeks. The bricks are first subjected to a moderate heat, until all remaining moist- ure has been expelled. The heat is then increased slowly, until, at the end of twenty-four hours, the "arch-bricks" attain a white heat ; the temperature is then lowered some- what, and a moderately high furnace heat is kept up until the burning is complete. Filially, all openings are closed, the fire is smothered, and the mass is then very slowly cooled. In the more modern processes of burning brick the prin- cipal yards have permanent kilns built of brick, either cir- cular or in the form of an ellipse, and made in compart- ments, each of which has a separate entrance and indepen- deiit connection with the chimney. A down draught is se- cured from the top, where the fuel is placed, to the chimney, which is either built within the kilns or entirely outside, but which has its draught invariably connected with the bottom of the kilns. The fuel used is generally fine coal, which falls around the bricks, and the flame and heated gases surround and pass through all portions of the mate- rials being burned. While some compartments are being burned, others are being filled, and still others are being emptied. Bricks of three kinds are taken from the kiln. Those form- ing the top and sides of the " arches " in which the fire is built are overburned and partially vitrified. They are called " arch-bricks" are hard, brittle, and weak. Brick from the interior of the pile are called " body-bricks," and sometimes hard or cherry brick ; they are of the best quality. Those brick which have formed the exterior of the mass are under- burned, and are called soft, sammel, or pale brick. They are too soft, and are of insufifieient strength for use, except for filling, in even ordinary work. Their price in the market is about twenty per cent, less than body brick, and vari- able. Good Bricks should be of regular shape, with parallel surfaces, plane faces, and sharp edges and angles. They should exhibit a fine, compact, uniform texture, should be S80 NON-METALLIC MATERIALS. quite hard, and should ring clearly when struck a sharp blow. They should absorb not more than six per cent, of their weight of water. Brick of fair quality bears a compressive force of 3,000 pounds on the square inch (211 kilogrammes per square cen- timetre) without completely crushing.* Very soft bricks will yield at as low a pressure as one-eighth this amount ; while the very best of pressed brick have been known to bear more than double 3,000 pounds.f Good brick maybe taken to aver- age about 2,000 pounds per square inch (141 kilogrammes per square centimetre.) Trautwine takes a minimum crushing strength for red brick at thirty tons per square foot (328,060 kilogrammes per square metre), and their weight at 1 1 2 pounds per cubic foot (1,794 kilogrammes per cubic metre) and thus estimates that a vertical column 600 feet (183 metres) high, would just crush at the base under its own weight. The ex- periments of Hatfield upon the transverse strength of brick L W gives a value for 5 in the formula S = -r-x", of from 19.6 to 26.36 for Perth Amboy; of from 24.35 to 42.74 for Hudson River hard brick ; and of from 32.29 to 41.85 for Philadelphia pressed brick. (In using metric measures, .S becomes eight one-thousandths of these values as before.) Masses of brickwork crush under smaller loads than single bricks ; and first quality brick, laid in first quality cement, should not be subjected to much above ten tons per square foot (i I kilogrammes per square centimetre) as a permanent load. The size and weight of bricks vary considerably. The British legal standard is 8^ X 4>^ X 2.Y^ inches. In the United States, 8>^ X 4 X 21^ is a usual size. Brickwork may be estimated at an average weight of 116 pounds per cubic foot. A good bricklayer lays from lOO to 200 bricks an hour, according to the character of the work. * G. S. Greene, Journal Franklin Institute, Vol. LXV., p. 332. f The author has tested specimens capable of resisting more than 10,000 pounds per square inch, (703 kilogrammes per sq. cm.). STONES AND CEMENTS. The following averages are given : S8i DESCRIPTION. INCHES. DESCRIPTION. INCHES. Baltimore front [si X 4i X 25 8| X 4 X 2i 8} X 35 X 2| 7l X 3| X 2| 8i X 4ix 2i 8 X 3i X 2i i 71 X 3J X 2Jr (8 V 4^ X 2j Philadelphia " Wilmington " Croton " Milwaukee North River Colabaugh Stourbridge fire-brick g^ x 4f x 2| inches. American (N. Y.) 8 J x 4^ x 2| Fire-brick is used whenever very high temperatures are to be resisted. They are made either of a very nearly pure clay — silicate of alumina — of a mixture of pure, clay with clean sand, or, in rare cases, of nearly pure silica cemented with a small proportion of clay. The presence of oxide of iron is very injurious, and it has been accepted as a rule by good engineers, that the presence of six per cent, ferric oxide in the brick justifies its rejection. It should generally be stipulated that fire-brick proposed for purchase should con- tain less than six per cent, of oxide of iron, and less than an aggregate of three per cent, of combined lime, soda, potash, and magnesia. The sulphide of iron — pyrites — is even worse in its effect on fire-brick than the substances just named. Where intended to resist extremely high heat simply, silica should be in excess ; and where exposed to the action of metallic oxides, which would tend to unite with silica, alumina should be in excess. Good fire-brick should be uniform in size, regular in shape, homogeneous in texture and composition, easily cut, strong and infusible. A good bricklayer should lay sixty per hour. The strength of fire-brick, as determined by experiments at the Royal (British) Arsenal, in 1871, is sufficient to enable it to sustain from 900 up to 2,000 pounds per square inch {6^ to 141 kilogrammes per square centimetre) before crushing. Bricks tested by the author have usually borne the maximum figure and often exceed it two or three times. 582 NON-METALLIC MATERIALS. Excellent fire-brick is made at Newark, South Amboy, and other places in New Jersey. The most infusible known fire- bricks are the Welsh Dinas bricks, which consist of 97 per cent, silica and 3 per cent, alumina and other constituents. The Mount Savage brick, of Maryland, U. S., is also noted for its infusibility. In lead smelting furnaces preference is given to fire-brick made of kaolinitic clay. Retorts for gas manufacturers, for glass-makers, and for other purposes, are made of fire-clay, in a similar manner to fire-brick, but they necessarily require more care in selecting materials, in moulding, and, particularly, in baking. A celebrated fire-clay has the following composition : SAMPLE. SILICA. ALUMINA. PROT. IRON. LIME. MAGNESIA. POTASH. Sample I 59-87 67.69 70.32 33-49 27.91 26.42 3.01 2.35 1.04 1.42 0.63 0.36 0.31 O.II 0.43 2 21 Sample 2 Sample 3 1.40 It makes excellent fire-bricks and crucibles, burns per- fectly white, and makes a fine glass-house clay. Artificial Sandstones are made by several processes. Of these b^ton and concrete will be referred to after explaining the methods of making mortars and cements. Mortars and Cements are used in masonry for the purpose of uniting the natural and artificial stones. They usually, when coitipletely hardened or " set," consist wholly or partially of carbonate of lime united with sand, or with sand and clay. Plaster of Paris — sulphate of lime — is also sometimes used. Carbonate pf lime is formed by the absorp- tion of carbonic acid from the atmosphere, which unites with the lime with which the mortars and cements were originally made up. * In the structures of the ancient Egyptians, as in the Great Pyramid, mortar was freely employed ; but it consisted almost entirely of sulphate of lime. A specimen taken from an ancient Phoenician temple, the highest stone of which was, STONES AND CEMENTS. 583 a few years ago, five feet below the level of the ground, was quite similar to that found in some of the castles in Europe, and was like a piece of solid rock. It was made of burnt lime, fine sand, coarse sand, and gravel. It was a concrete rather than a mortar ; the lime had become completely car- bonated. Ancient Greek mortars from ruins in the neighbor- hood of Athens are in very perfect condition ; they contain no gravel. Mortars from ruined buildings in Herculaneum, and from Rome and its vicinity, appear to have been made from burnt lime and puzzolana, or volcanic ash. Lime, as a building material, is of three principal kinds : common or air lime, hydraulic lime, and hydraulic or water cement. Common lime, called also pure, rich, or fat lime, is pro- duced by calcining limestone, which is nearly pure carbonate of lime, and thus expelling its carbonic acid. It " slakes " by greedily absorbing moisture, becoming converted into a dry hydrate, if water is not used to excess. Made into a paste with water, it hardens slowly in the air, but not at all under water. Hydraulic Limes are made from stones containing from 18 to 30 per cent, of silicate of alumina, of carbonate of magnesia, or of a mixture of both. They slake more slowly than air lime, and the paste hardens very slowly under water, or in wet localities. Hydraulic cements are made by calcining limestones con- taining from 30 to 60 per cent, of clay. They do not slake, and their pastes harden with rapidity under water. They are, therefore, of greatest use in building foundations. Where the proportion of silicate of alumina is greater than 60 per cent., the material is called puzzolana, and it requires the addition of fat lime to render it useful. Natural puzzo- lana is of volcanic origin. Brick-dust has a similar power of rendering fat limes hydraulic, as has also trass, terras, or blue trap-rock. The hydraulic limes and cements are sometimes obtained from stones which contain the desired proportion of lime- stone and clay; in which case they are known as natural S84 NON-METALLIC MATERIALS. limes or cements. Sometimes the lime and the clay are mixed artificially in proper proportions. Limestones of all qualities are found in New England, New York, and many other portions of the United States. The English Portland cement is made by grinding together chalk and clay. That from the gray chalk is said to be heaviest and best. This is the strongest cement known in the market, and it is by far the most expensive. Roman cement is made from nodules of limestone contain- ing clay and iron. It makes a cement which sets more quickly than the Portland, but does not become as hard. Mortars are made by mixing lime and sand with water in such proportions as will give the desired quality, thus form- ing a paste which may be used for uniting stone and brick- work. Common Mortar is made with fat lime, and clean, sharp sands, in the proportions, usually, of i to 5 by volume. It hardens promptly in the air, and becomes, finally, very hard, if of good quality, and if frost, or too great dryness, or excessive dampness does not injure it while setting. Hj/draultc moriar is ma.de with, hydraulic lime and sand. It hardens in damp situations, and is a strong binding ma- terial. Under water it often requires weeks to harden ; but hydraulic mortar of fair quality requires from three or four days to a week. Very excellent varieties harden in from one to four days. It is often tempered with clay or lime to re- tard its setting. The slower this action, usually, the firmer and harder does tlie mortar finally become. Hydraulic cement is a mortar made with the very hydrau- lic lime, also termed cement, already described. Hydraulic cements sometimes set in a few minutes after mixture, if warm; they do not shrink much in setting, and are often used without admixture of other material. Where even slight shrinkage is objectionable, an addition of three timis its own volume of sand will prevent change of volume. Hydraulic cement is generally indispensable in the construc- tion of foundations. It should be laid in thin joints, and should generally, if great strength is desired, be used un- STONES AND CEMENTS. 585 mixed with sand ; it requires about one-third its volume of water. The sand used in mixing mortars should be free from clay and perfectly clean ; it should be sharp and rather coarse. River sand is usually found to be better than sea sand, as it is free from salt, and is less liable to be found water worn. Mortars and cements are given different proportions for different kinds of work. Mortar for stone may be made by mixing 15 to 20 per cent, cement, 6 to 8 parts lime, and the remainder of the 100 parts sand ; mortar of good quality for brickwork should contain 10 per cent, less sand. Stucco is made of two parts sand to one part cement ; to this is some- times added a little sugar or molasses. Plaster for inside finish is usually of several grades. Coarse plaster is made by adding to common mortar about five per cent, of its volume of cows' hair; "fine stuff," or putty, is a paste of lime mixed without sand ; " hard-finish" contains of lime 3 or 4 volumes, of plaster of Paris i part. Concrete is made by mixing gravel or broken stone with lime and sand, using a limited amount of water. Frag- ments of brick are often also added. It is mixed in about the proportion of i part lime, by volume ; 6 parts or more of sand and other solid components, and i]/2 parts water. In using it, it should be thoroughly mixed and carefully rammed in place. It swells about three per cent, in setting. Each cubic foot ( 028 cubic metre) of gravel makes about four-fifths of a cubic foot (.022 cubic metre) of well rammed concrete. Used as a foundation for masonry, it should be laid in layers of about a foot in thickness, each being carefully ram- med before another is added. It is not well fitted for use in damp localities. Beton is the name usually applied by engineers to a con- crete in which hydraulic lime or cement is used, instead of fat lime. It should always be given the preference in wet, or even in damp, situations ; and is often used on dry work also, when strength is sought. 586 NON-METALLIC MATERIALS. Occasionally a little lime is added to retard the setting of strong hydraulic cement concrete. A commission reporting on the submarine work of the New York Dock Commission states that in order to produce good submarine masonry by depositing freshly mixed con- crete under water, certain precautions are necessary, viz.: The cementing material should possess the properties of unctuousness and adhesiveness, to enable it to retain the sand while the concrete is assuming a state of rest in the water; and it should be capable, as much as possible, of as- suming that state by spreading, rather than by breaking down and rolling off on the sides of the mass deposited. If it be deficient in the properties last named, it should then be quick-setting, in order that the washing out of the ■ sand from the cement may be arrested in a few minutes after deposition by its prompt hardening. If a box is used for depositing the concrete, the shape of the box, and the method of emptying it, should be such that the concrete will be subjected to as little wash as possible. Hence a large box is preferable to a small one, as it will expose a less area of surface in proportion to the volume deposited. B6ton has been used in the following proportions on the works named, with excellent results : Croton Aqueduct, New York — New York cement, i part, by volume ; sand, 3 parts ; broken stone, small enough to pass through a ring i^ inches (3.8 centimetres) in internal diameter, 3 parts; — Cherbourg Breakwater, France — Port- land cement, i part ; sand, 3 parts. Properly made, this concrete, or bdton, is found to be strong enough to take the place of stone ; walls, chimneys, and even bridges have been constructed of it. The addition of a small quantity of sulphuric acid, or the presence of a sulphate, is found to add very considerably to the strength of mortars. Beton-Coignet, as made by the French engineer, M. F. Coignet, arid which attracted much attention at the Inter- national Exposition at Paris, in 1867, is composed of: lime, 4 parts ; hydraulic cement, i to 2 parts ; and sand, 20 parts. STONES AND CEMENTS. 587 The ingredients are first tlioroughly intermixed dry, by hand, and again in a mill, moistening them very slightly with clean water. Moulds are then filled with the mixture, and it is compacted by ramming or hammering Four bushels of the ■ mixture, occupying, when dry, five cubic feet (141.6 litres), make three cubic feet (85 litres) of finished work, weighing 140 pounds per cubic foot (2,243 kilogrammes per cubic metre). Its peculiarities are the small quantity of water used in its manufacture, and the thorough- ness with which the mixing and ramming are done. It sets quickly, is very strong, and is the best example of mixed b6ton. It may be made into blocks to be used as cut stone, or may be built up in masses of any desired shape. The cheapness and strength of construction of Beton-Coignet are so remarkable as to have led to its use for even orna- mental work. It is used to a considerable extent in con- structing the walls of houses and public buildings. Strength of Mortars, Cements, and Concrete. — Mortar has a tenacity of from 6 to 34 pounds per square inch (0.42 to 2.39 kilogrammes per square centimetre) when six months old ; and the average, as determined by General Totten, U. S. A., was about 15 pounds per square inch (iJ3$ kilogrammes per square centimetre), or nearly a ton per square foot (10,937 kilogrammes per square metre). The increase in strength, with age, is very variable, amounting sometimes to twice or three times these figures, and, at other times, to a mere fraction. The resistance to crushing, a year and a half after setting, is given by Rondelet as from 440 to 580 pounds per square inch (31 to 41 kilogrammes per square centimetre) when sim- ply laid in place ; and from 600 to 800 pounds (42 to 56 kilo- grammes) when well rammed. These figures correspond to about 30 and 35, 40 and 50 tons, respectively, to the square foot (328,066 and 382,750, 437,450 and 546,800 kilogrammes per square metre). Its adhesion to brick or stone work is about equal to its cohesive strength, on good work, of moderate age; if very old, the adhesion is greater. 588 NON-METALLIC MATERIALS. Gypsum is used in taking casts, and in stereotyping. Gypsum is also employed for glazing porcelain, and, being an excellent non-conductor of heat, with alum for filling fire-proof safes. Made into a mortar with sand and lime, it is used for cementing floors and vaults. The best gypsum quarries that are worked on this con- tinent are those of the Bay of Fundy, Nova Scotia, and Hills- boro. New Brunswick. Over one hundred thousand tons of the finest' quality have been annually imported from these places into the United States. The Bituminous Cements are usually composed of mixtures of bituminous substances, as asphalt, with less costly materials. Bitumen or mineral tar, asphalt, and a bitumin- ous limestone are thus used. The latter sometimes contains ten or fifteen per cent, bitumen. The mixtures are made by breaking up the materials, and heating them in large iron kettles or boilers. The propor- tions in mastic are usually from one to one and a half parts bitumen to each ten of asphalt. Coal-tar, although of far inferior value, is frequently used instead of the natural bitumen, as is also pitch. Fire-clay is sometimes used in place of limestone, and the preparation so made makes excellent joints for water-pipes. Bituminous cements mixed with broken stone to form a bituminous concrete sometimes make a good road covering. Masonry is the art of making structures of stone, brick, or other earthy materials. Good masonry is built in " courses," which are usually per- pendicular to the lines of pressure bearing upon them, with discontinuous or " broken " joints in the lines of stress. The stone-mason selects the heaviest stones for his lowest courses in all foundations or structures, lays all stones on their natural beds, and secures the most perfect union between them and the cementing material. The nomenclature of stone masonry has been revised by a committee of the American Society of Civil Engineers,* and * Trans., No. CLI., 1877. STONES AND CEMENTS. 589 the specifications of the engineer are recommended to be made in accordance with their report and as below. Stones are classed thus : In practice, one class merges into the next. I. Unsquared Stones or Rubble. — This class includes stones used as they come from the quarry, without other preparation than the removal of sharp angles and projections. The term " backing," frequently applied to this class of stone, properly designates material used in certain relative positions in the wall ; while stones of this kind may be used in any position. II. Squared Stones.— 'Y.\i\s class includes stones roughly squared and dressed on beds and joints. The dressing is done with the face hammer or the axe, or, in soft stones, with the tooth hammer. On gneiss, the point is sometimes used. Where the dressing on the joints is such that the average dis- tance between the surfaces of adjoining stones is one-half inch or more, they properly belong to this class. Three subdivisions of this class may be made, depending on the character of the face of the stone. («.) Quarry-faced stones are left untouched as they come from the quarry. (p.) Pitch-faced stones have the arris clearly defined by a line beyond which the rock is cut away so as to produce edges approximately true. (a) Drafted stones have the face surrounded by a chisel draft, the space inside the draft being left rough. Ordinarily this is done only on stones in which the cutting of the joints is such as to exclude them from this class. In ordering stones the specifications should state the width of bed and end joints, and how far the surface of the face may project beyond the plane of the edge. In practice the pro- jection varies between i" and 6". It should be specified whether the faces are to be drafted. III. Cut Stones. — This class includes all squared stones with smoothly dressed beds and joints. As a rule, all the edges of cut stones are drafted, and between the drafts the stone is smoothly dressed. The face, however, is often left rough, when the constructions arc massive. 590 NON-METALLIC MATERIALS. ^\ "^ ~^-^ ^ ^ ^ - -, ^ >--^. ■^ ^ ^■^ ^ ^- X\-\^ \:^ "-^- \ ^-- -^ --N-V, ^ \ -- ,^ ^ Fig, 112. Fine-Pointed. — When The following are usual methods of dressing stones : Rough-Pointed. — When necessary to remove an Inch or more from the face of a stone, it is done by the pick until the projections vary from Y2" to i". The stone is then said to be rough pointed. This is the first operation in dressing lime- stone and granite (Fig. 112). smoother finish is demanded, rough pointing is followed by fine pointing (Fig. 113). It is used where the finish is to be final, and not as a preparation for final finish by other tools. Crandalled. — This is a Fig. 113. . , , , r - ■ , rapid method of pointmg, the effect is the same as fijie pointing, except that the marks on the stone are more regular. The variations of level are about ^", and the rows are parallel. When other rows, at right angles to the first, are introduced, the stone is said to be cross-crandalled {Fig. 1 14). Axed or Pean Hammered, and Fig. 114. Patent Hammered. — These vary only in the degree of smoothness of the surface (Fig. 115). The number of blades in a patent hammer varies from 6 to 12 to the inch, and in specifica- tions the number of cuts to the inch is stated, such as 6-cut, 8-cut, lO-cut, i2-cut. The effect of axe- ing is to cover the surface with chisel marks which are made par- allel as far as practicable. Axeing is a final finish. Fig. 115. Tooth- Axed.- -The tooth-axe is practically a number of STONES AND CEMENTS. 591 Fig. ii6. points and it leaves the surface of a stone in the same con- dition as fine pointing. It is usually a preparation for bush hammering, and the work is then done without regard to effect, provided the surface of the stone is sufficiently levelled. Bush Hammered. — The inequalities of a stone are pounded off by the bush hammer, and the stone is then said to be " bushed " (Fig. ii6). Sandstone thus treated is very apt to scale. In dressing limestone which is to have a bush-hammered finish, the usual order of operations is : 1st, rough pointing ; 2d, tooth axeing; 3d, bush hammering. Rubbed. — In dressing sandstone and marble, it is very common to give the stone its surface at once by the use of the stone saw. Any inequalities left by the saw are removed by rubbing with grit ot sandstone. These stones are used in architecture for string courses, lin- tels, door jambs, etc., and are well ^^°- '^7- adapted for use in localities where a stone surface is liable to be rubbed by vessels or other moving bodies. Diamond Panels. — The space between the margins is sunk immediately adjoining them, and thence rise the four planes forming an apex at the middle of the panel ; this makes a sunk diamond panel. When the sur- face of the stone rises gradually from the inner lines of the margins to the middle of the panel, it is called a raised diamond panel (Fig. 118). The term stone masonry includes : (i.) Rubble Masonry is composed of unsquared stones ; it may be Uncoursed Rubble (Fig. 119), laid in irregular courses. Fig. 118. 592 NON-METALLIC MATERIALS. or Coursed Rubble (Fig. 120), levelled off at specified heights, The stone may be required to be roughly shaped with the hammer, so as to fit fairly. Fig. 119. Fig. 120. (2.) Squared Stone Masonry. — This is classified as Quarry- faced (Fig. 121), or as Pitch-faced (Fig. 122). If laid in regular Fig, 121. Fig. 122. Fig. 123. courses, it is Range work (Fig. 123). If laid in courses that are not continuous throughout the length of the wall, it is Broken Range work (Fig. 124). If not laid in courses, it is Random 1 1 1 L I 1 1 1 1 1 1 1 1 1 III - 1 - 1 I — I II — L^ Fig. 124. Fig. 125. work (Fig. 125), and this is generally the method adopted. In quarry-faced and pitch-faced masonry, quoins and the sides of openings are hammer-dressed, in removing projections to secure a rough-smooth surface, with the face hammer, the STONES AND CEMENTS. 593 plain axe, or the tooth axe. This is done for doors or win- dow-frames, and improves the general effect if used where a corner is turned. (3.) Ashlar Masonry — This is " cut-stone masonry," or masonry composed of any of the kinds of cut stone. The courses are continuous (Fig. 126), but sometimes are broken II 1 1 1 1 111)11 1 1 1 1 1 1 II 1 1 1 1 II 1 1 1 1 1 1 1 1 1 1 fe Fig. 126. Fig. 127. by the introduction of smaller stones ; it is called Broken Ash- lar (Fig. 127). If the stones are less than one foot in height, the term Small Ashlar is proper. The term Rough Ashlar is sometimes given to squared stone masonry when laid as Range work ; but it is better to call such masonry " Squared Range work." Dimension stones are cut stones, whose dimensions have been fixed. Specifications for Ashlar masonry prescribe the dimensions to be used. General Rules. — Range work is usually backed up with Rubble masonry, which is specified as coursed Rubble. Every specification should contain an accurate description of the character and quality of the work desired. Samples of cutting and masonry should be prepared beforehand. The softer stones should have a depth equal to at least one- third their length to prevent crossbreaking, and a breadth of one-half their length. The hard and strong stones are allowed a double length. A rough natural surface is of advantage when strong adhesion of mortar is important. The thickness of Joints in Ashlar masonry is about }i inch, and in fine work as little as can be secured. All spaces should be completely filled. In coursed masonry one-fourth or more of all stones should be headers, /. e., should extend from front to back, and the re- 38 594 NON-METALLIC MATERIALS. mainder are stretchers, i. e., lie lengthwise in the wall. Com. mon Rubble has about the strength of mortar ; coursed and fine work has nearly the strength of the stone itself. Ashlar is usually backed with Rubble, and both should be carried up together, and, as nearly as possible, the whole length of wall should rise together. The top of the wall is protected by its cope, which is a " string-course," i. e., a projecting course ; and is made of stones long and broad enough to protect the wall from rain, and heavy enough to be displaced with difficulty ; they should be of good shape to shed rain. Adjacent stones in the coping, and in engine and in lighthouse foundations, or other places in which great strength is demanded, are secured together by iron " cramps " or " dowels " of metal or stone. The joints of masonry are finished on the surface by "pointing" with cement, plaster, or fine mortar, to give smoothness of surface and to cause them to shed rain. In the Measurement of Masonry, stone-work is measured by taking openings less than 3 feet (0.9 metres) wide as solid wall, and adding 18 inches (0.45 metres) for each jamb. Arches are usually taken as if solid from the springing line; corners are measured twice, and pillars are measured by the area of three sides multiplied by the fourth. Foundations and dimension stones are measured by cubic measure ; water- tables and base courses in lineal feet, and sills and lintels in superficial feet. Brickwork is laid like stone-work, with the line of courses perpendicular to that of pressure. Broken and soft bricks are rejected, and each brick laid should be wetted and cleaned before laying it in place ; the joints should be as thin as J^ or j\ inch (0.64 to 0.48 centimetres). About one-fifth as much mortar as brick is generally used. The " English bond," in which entire courses of stretchers and headers are laid at regular intervals as the wall rises, is considered strongest ; when laid one course of headers to each two courses of stretchers, the strength is very nearly the same lengthwise and crosswise. " Flemish bond " is laid header and stretcher alternating in each course ; it is easier to retain regularity in breaking joints STONES AND CEMENTS. 595 in this bond, but it lacks strength and is not as neat in ap- pearance as English bond. Brickwork is measured by the thousand bricks; with average sizes and good work, the following are the number of bricks laid by the superficial foot : 4-inch (10.16 centimetres) wall 7 to sq. ft., 75 to square metre. 9 " (22.36 " ) " ...14 " 150 13 " (33.02 " ) " ...21 " 216 iS " (45.72 " ) " ...28 " 300 22 '• (55.88 " ) " ...35 " 377 Corners are measured twice ; small openings are taken as solid work ; arches are measured as solid from the springing line, and pillars are measured on the face. Masonry will carry safely from 2 to 10 tons per square foot (21,875 to I09>379 kilogrammes per square metre), accord- ing to quality ; and carefully built masses of cut and dressed granite may carry four times the higher figure. Masonry in damp situations is always laid in hydraulic mortar or cement, and the lower courses of walls and founda- tions are usually carried below the frost-line. The soil should be carefully drained. Where new masonry abuts upon old work, there is always danger of cracking by the settling of the new work ; but every precaution should be taken to se- cure a good bond between the two portions and to make the joints of the new work thin, and of cementing material of such consistency as will prevent excessive shrinkage. The Cost of Masonry cannot be given except on the assumption of a fixed rate of wages. Taking the wages of a laborer at $2.00 per day, and for the mason $4.00, we may reckon as below. Where wages fluctuate, and, indeed, in all cases, if possi- ble, a careful estimate should be made after ascertaining the conditions actually affecting prices. In the estimate below stones are assumed of moderately large size. Smaller stones cost less to handle, but more for dressing. Rubble masonry should cost probably one-half these fig ures if of good quality, and may fall to one-fourth when the stones used are small. 59^ NON-METALLIC MATERIALS. COST OF MASONRY ; ASHLAR. Quarrying li cubic yards' $4 00 Dressing 16 sq. ft. face @ 40c 6 40 Dressing 48 sq. ft. joint @ 20c 9 60 Cost of stone per yard $20 00 Haulage, variable, say 2 50 Mortar i 00 Laying one cubic yard and incidentals 2 SO Cost of placing 6 00 Contractor's profit, 15 per cent 4 00 Total cost I30 00 J_jU-^ ^^SI^^^ Fig. 128. — Coursed Rubble Wall. CHAPTER XIX. TIMBER. "Timber" is that portion of the woody material of trees which is used in carpentry and joinery. Hence the term only applies to the wood of particular kinds of trees, which are therefore designated as " timber-trees." In some districts of the United States, timber cut and dressed is dis- tinctively called "lumber," the term timber being restricted to the standing wood. The timber-trees are nearly all of those classes known by the botanists as exogenous, i. e., those in which growth takes place by the formation of woody fibre on the external surface of the sap-wood, immediately beneath the bark. Endogenous trees, as those of the palm family, do not furnish timber. Their growth takes place by an internal formation of ligneous fibre, and the wood is not firm and solid enough for the purposes of carpentry. If the trunks . of timber-bearing trees are cut, they are found to be composed of concentric cylindrical layers, whose cross sections form rings, separated from each other, and evi- dently quite distinct. These layers are formed, one each year, during the period of growth of the tree. They vary in thickness, in density, and in color, according to the rapidity of growth, the length of the season, and other circumstances which may change from year to year. The outer portion of the trunk is called the " sap-wood," and is usually lighter in color, and less strong and dense than the interior portions, or heart-wood. The circulation of the sap through the sap-wood occurs during favorable weather. In winter it is supposed to cease, and this period of checked circulation causes the line of de- markation between successive annual rings. 598 NON-METALLIC MATERIALS. In midsummer also, in our climate, and in the height of the dry season in tropical climates, the sap flows less freely than either earlier or later in the season. During the montli of July, with us, it almost ceases flowing. The heart-wood is nearly, or quite, impervious to sap, its vessels being closed up, and the wood is dense and hard. It is almost pure woody fibre, is free from sap, and contains almost none of the sugar and the mucilage which are found in sap-wood. It is usually far more durable therefore than the latter. Different kinds of trees, and different individuals of the same species, have different proportions of sap-wood. The slower-growing trees usually contain least. The complete conversion of sap-wood into heart-wood occupies from one year, as with the softer woods like beech, to twenty or thirty years, and even longer, as with the oak. In the first class, slow growth, and in the second, a compara- tively rapid growth, produces the best wood. Decandolle gives the following as the maximum age of timber-trees, the figures being obtained by counting the annual rings of old trees : Elm, 335 years ; cypress, 350 ; larch, 575; cedar, 800; linden, 1,150; oak, 1,500; and the adan- sonia, 5,000. The longevity of various trees has been stated by others to be, in round numbers, as follows : Baobab tree of Senegal, 5,000 years ; dragon's-blood tree, 4,000 ; yew, 3,000; cedar of Lebanon, 3,000; olive, 2,500; oak, 1,600; orange, 1,500; ori- ental plane, 1,200; cabbage palm, 700; lime, 600; ivy, 600; ash, 400 ; cocoa-nut palm, 300 ; pear, 300 ; apple, 200 years. These estimates are disputed, however, and are by some writers thought greatly in excess of the correct figures, as it is found that several rings may be sometimes formed in a single year. The length of the life of trees seems largely dependent upon the proportion of heart-wood, and, particularly, upon its durability, decay usually originating and progressing, in grow- ing trees, only in the heart-wood. The sap-wood and bark are peculiarly subject to the attacks of worms and insects. At the period of maturity the heart-wood is of maximum density and uniformity of texture. TIMBER. S99 " Felling " Timber should always, if possible, be prac- tised at the period of maturity ; if earlier, the wood will not have acquired its greatest strength and density, and will con- tain too great a proportion of sap-wood ; if later, the wood will have become weakened by incipient decay. The oak is said to reach maturity when about lOO years of age, and it should not be felled at less than 60. Pine timber should be cut at from 70 to 100 years of age, and ash and elm at from 50 to 100. The season of the year best adapted for felling timber is either midwinter or midsummer. The months of July and August are often selected, as at those seasons the sound trees remain green, while the unsound trees are then turning yellow. Healthy trees then have tops in full foliage, and the bark is uniform in color, while unsound trees are irregularly covered with leaves of varying color, having a rougher, and often a loosened, bark, and decaying limbs. The cut should be made low, and the opposite incisions should be so made, especially with oak, as to enable the trunk to be cut clear of the stump while falling; otherwise the trunk may be split. The trunk should be immediately stripped of its bark, and, when heart-wood only is wanted, the sap-wood removed as soon as possible. The bark is often removed from trees in spring, and the felling deferred until autumn or winter. This is probably the best course to pursue, usually. Handspikes and similar " uses " should be cut from young straight trees, and near the butt. Seasoning Timber is simply driving out the sap from its pores by either natural or artificial means. This should always be done as gradually as possible, otherwise the timber is liable to crack or " check," from irregular drying. Natural or air seasoning gives the best results. The tim- ber should in all cases be squared as soon as cut, and all large logs should be halved, or even quartered. It is then piled in the seasoning yard in such a manner as to be protected as far as possible from the sun and rain. It should be placed where the air may circulate freely on all sides, not only of the pile, but of each log; bad ventilation is sure to cause rot. After 600 NON-METALLIC MATERIALS. remaining thus for some months, the logs may be cut into smaller joists, if needed in such form, or into planks and boards, and again piled for further seasoning. For heavy work, two years, and for lighter work, four years, is suiificient time for seasoning boards ; but timber is rarely overseasoned. The loss of moisture in the first year of seasoning may be taken usually as about twenty per cent. When piled for seasoning in air, the lower sticks should be placed on supports one to two feet high, to keep them from contact with the damp earth. At least an inch should separate adjacent pieces. The timber should be replied often enough to secure the de- tection and removal of unsound pieces. Water seasoning is accomplished by immersion in water for a long time. It is a slow and imperfect method, but for timber to be used in water, or in damp situations, it answers well. The sap, in this case, is removed by solution. In salt water there is usually some danger that the wood may be attacked by the ship-worm. Teredo navalis, or by the Limnoria terebrans, both of which destroy timber very rapidly. It should therefore be carefully watched. Two or three weeks water seasoning is sometimes found to be a good preparation for air seasoning, by dissolving out the more soluble salts contained in the wood. Steaming timber is resorted to where it becomes necessary to soften wood, in large pieces, for the purpose of bending it, as in ship-building. An hour to each inch of thickness is the period of time allowed. This process sometimes impairs the strength ; but it is also a seasoning process, and preserves from decay as well as from injury by warping or cracking. Hot-air seasoning is resorted to where it becomes necessary to season wood rapidly. The timber is piled in large cham- bers or ovens. The sap is expelled by a current of hot air, having a temperature of from ioo° Fahr. (38° Cent.), with large logs of hard wood, to 250° to 300° Fahr. (121° to 149° Cent.), with thin boards of the softer kinds, the wood losing, in the latter class of materials, about thirty per cent, of its weight. The time required maybe stated to be generally one week TIMBER. 60 1 for each inch (2.54 centimetres) of least thickness, to insure good work. In seasoning birch sticks, one inch or one and a quarter inch (3.2 centimetres) square, sixty hours are allowed. The fuel used amounts to about ten per cent, of the weight of seasoned wood. Seasoning by passing the smoke-laden products of com- bustion from the furnace, directly through the pile of timber, has been found not only a good method of seasoning, but also to have an important and useful preservative effect. Seasoning by boiling in oil is resorted to for some purposes, as the preparation of hickory for use in making teeth of mor- tice gears. If carelessly done, the wood may be seriously injured by the charring of its fibre in the overheated liquid ; but if the temperature is carefully kept at, or somewhat under, 250° Fahr., the result will be most satisfactory. The wood should be seasoned in blocks roughed out to nearly the finishing size, and they become not only well and uniformly seasoned, but, as shown by the experiments of Mr. G. H. Corliss, considerably strengthened. If well done, seasoning usually increases the strength of timber, but the amount of this increase is very variable. Pine gains about ten per cent., elm from ten to fifteen, oak from five to twenty-five, and ash and beech often gain forty per cent, or more. The amount of water contained in green timber varies from twenty-five or thirty per cent., in willow and ash, to thirty-five per cent, in oak, and forty per cent, in pine. Large beams are best built up of small pieces, in order to secure thorough seasoning, and to avoid risk of decay. Shrinkage always occurs to a greater or less extent, in consequence of the expulsion of moisture while seasoning; and some woods not only shrink, but warp badly, while others are seriously injured by the occurrence of " seasoning cracks." The shrinkage of timber is not usually very noticeable in the direction of its length; but transverse shrinkage often occurs to a marked degree. In soft timber, as birch, it amounts to about eight per cent. 602 NON-METALLIC MATERIALS. The tree consists of a bundle of capillary tubes, cohering laterally, the sap-wood, when green, filled with sap, and having the heart-wood moist, but choked with resinous matter. These fibrous bundles, or the vascular tissue, are bound together by a cellular tissue, the membrane which constitutes the medullary rays, which latter form, in many woods, as in oak, well-marked dividing planes and lines of weakness ; they consequently determine the surfaces along which season cracks may be developed while shrinking. The inner portion of the tree, the heart-wood, being denser and less fully saturated with moisture than the external or sap-w®od, shrinks less, and thus it happens that all planks, composed of portions of both kinds of wood, or of different qualities of the same kind, are certain to be warped or otherwise distorted while seasoning. Usually a log is cut into planks, when green, by gang or circular saws, and these planks, origin- ally of the shape seen in Fig. 129, ^'°-"9. are likely, when seasoned, to take the shapes seen in Fig. 130. The shrinkage warps the outer planks and distorts the middle one, by reducing its 'thick- ness to a greater extent at the edges than in the middle. The simple inspection of the position of the medullary rays and annual rings in a piece of green wood will enable any one to determine from what part of the trunk it has come, and to predict its change of form while seasoning. Nomenclature. — The term timber is seldom applied, in the trade, to logs cut from trees less than eight inches (20.3 centimetres) in diameter. Smaller sizes are csXled joists. Before felling, it is called standing timber ; when first cut, Fig. 130. TIMBER. 603 it is called rough timber ; and after it has been sawn, it is called converted timber ; and is also known as sided timber, joists, plank, or board, according to dimensions. Wood is either soft or hard wood. The first class includes the wood of all coniferous trees, as the pines, and of a few others, as for example white birch. The second class includes the wood of all other timber- producing trees. The soft woods generally contain turpentine and pitch, and are usually of rapid growth, straight-grained, of slight density, quite uniform in texture, and comparatively free from knots. They have but little lateral adhesion of fibre, and are easily worked. The hard woods are denser, heavier, and stronger, less easily sawn, split, or cut, and are more liable to warp and to crack than are the soft woods. They usually excel in dura- bility, and in some cases are very tough and elastic. Characteristics of Good Timber. — Good timber has the following characteristics : The heaviest is usually the strongest and most durable. That which has least sap or resin is the best. The freshly cut surfaces are firm and smooth, and the shavings are translucent, and should nowhere appear chalky or roughened, that being the first indication of decay. The annual rings should be closely packed, and the cellu- lar tissue of the medullary rays should be hard and dense. The tissues should cohere firmly, and when sawn there should be no wool-like fibre clogging the saw-teeth. In general, the darker the color, the stronger and more durable the wood. Inspection of Timber. — Timber should be inspected in dry weather, when the defects are not concealed by moisture. The color should be bright and' uniform, slowly changing from sap-wood'to heart-wood, and free from the white spots which indicate incipient decay. Dry-rot is indicated by yel- low stains. Usually sap-wood should be thrown out, except in a few cases, as in ash, lancewood, and hickory, where it is sometimes 604 NON-METALLIC MATERIALS. even better than heart-wood. The use of tjae centre heart- wood of mature trees is also usually avoided as being liable to early decay. Brash-wood, which is old and brittle in consequence of age, is rejected, as is also knotty timber, twisted wood, and the timber which has been felled after having died from natural causes, like belted timber. The preparation and inspection of small pieces is best illus- trated by the regulation system adopted in the national armories with reference to securing good material for musket and rifle stocks and butts. The wood used is the best American or Italian walnut. After seasoning about three years in the rough, or, if arti- ficially dried, after being exposed to a temperature of 60° F. (iS^S C), slowly raised to 90° F. (32° C), and held at the latter temperature six to eight weeks, the pieces are handed over to the inspector. If defective in either of the following respects, they are rejected : * 1. Under size. 6. Discolored wood. 2. Misshapen. 7. Knots or bines. 3. Galls, f 8. Crooked or cross-grain. 4. Shakes. 9. Decay. 5. Rind galls, t 10. Worm-holes. They are also examined to determine whether they have been soaked in salt water, the presence of which produces rapid corrosion of the metal of barrel, lock, and mountings. The test is made by dipping a shaving into a solution of nitrate of silver ; the presence of salt is shown by the formation of a white precipitate of silver chloride. Influence of Climate and Soil. — These greatly affect the value of timber. Generally the strongest varieties of wood come from tropical climates, but the best examples of any one variety are usually from the colder* portion of the range of country in which it abounds. * Ordnance Notes, No. 197. f Produced by insects depositing their eggs in the tree. J Due to surface injuries to the sapling. TIMBER. 605 Timber Onslow growth, in situations protected from violent winds, cut at the right time of year, and properly seasoned, is free from " cracks " and " shakes," or " checks." Cup-shakes are produced by the wrenching of the tree by winds, and are cracks separating one layer from another. Timber thus injured is sometimes called " rolled timber." Longitudinal cracks are produced by heavy winds also, and by too rapid seasoning ; in the latter case they are called seasoning cracks, in the former, wind-shakes or cracks. Frost, in cold climates, sometimes produces this kind of injury. Cup-shakes often injure oak, hard pine, mahogany, and elm, but they do not as generally affect soft pine. " Heart-shakes" which are cracks crossing the heart-wood, sometimes single and sometimes grouped, making a " star- shake" affect all kinds of timber. Decay of -Timber. — Timber decays in two quite dif- ferent ways, the causes of decay being, however, the same in both cases, namely, fermentation and putrefaction. Dryness is the best preventive of decay of timber used in general construction, and wood kept dry has been found to last several centuries. Still, it finally becomes brittle and weakened, and may ultimately give way under a light load. Water seems to act as a preservative, and some kinds of timber constantly immersed in water not in motion may endure for an indefinite period. The first effect of water is to dissolve out soluble matter, leaving the woody fibre or lignine uninjured, except perhaps very slightly by oxygen in solution in the water. This oxygen being exhausted however, no further action occurs unless a fresh supply of air-laden water displaces that originally in contact with the wood-. Alternation of moisture and dryness induces rapid decay. This takes place partly by solution and removal of a portion of the substance at each moistening, and partly by the action of oxygen dissolved in the water, a fresh supply of dissolved oxygen being furnished at each repetition of the moistening. Continued dampness in a warm atmosphere is most favor- able to fermentation, and consequently to rapid decay. This 6o6 NON-METALLIC MATERIALS. putrefaction of woody fibre is known as " rot " among those who use timber. The products of this decomposition are, as in cases of rapid combustion of wood, carbonic acid and water. The presence of water is necessary, as well as that of air, to the rapid progress of this chemical change, although the oxygen, which is essential, may sometimes be obtained from some source other than the atmosphere. Sap-wood is more perishable than heart-wood, in conse- quence of the presence of saccharine and other matters having a peculiar tendency to fermentation. It is in consequence of this fact that the complete removal of the sap by seasoning is necessary. Lime, by its tendency to abstract carbon, which, uniting with oxygen, combines with lime to form the carbonate, hastens the rotting of wood wherever it is damp. Dry lime and the carbonate do not have this effect. "Wet-Rot" and "Dry-Rot" are the two forms in which the decay of timber exhibits itself. Wet-rot occurs in any portion of the wood, if damp, and attacks the heart-wood of standing timber. Dry-rot is usually produced by the want of circulation of air, and by high temperature, where the timber has not been well seasoned. The most rapidly growing trees are most subject to decay, and those growing in sheltered localities are more liable to rot than those in. exposed situations. Of soft timbers, that containing most turpentine is least liable to rot. Woodwork embedded in damp plaster, and unseasoned timber covered with a coating of paint, are subject to dry-rot, and are apt to decay early, in consequence of the confinement of air aind moisture within their pores. Any thing which absorbs moisture and confines it in contact with wood is likely to accelerate decay. Marine Animals frequently attack timber immersed in salt water, as the bottom of vessels, piles, etc. The Teredo navalis, commonly known as the ship-worm, TIMBER. 607 converts the wood which it enters into a perfect honeycomb. It enters the wood when very small, and there increases in size, and enlarges its chambers correspondingly, until it some- times makes borings an inch (2.54 centimetres) in diameter, and several feet long. Soft woods are very rapidly destroyed by it, and the hardest woods are not safe against its attacks. The Limnoria terebrans is a smaller creature than the Teredo, shaped somewhat like a wood louse, and is rather more than an eighth of an inch (.3 centimetres) long. It is very destructive, cutting out the wood along the annual rings. There are several other marine animals which attack tim- ber, and it is usually necessary to protect it, when immersed in salt water, by sheathing with copper, as ships are protected, or otherwise covering it with a coating impenetrable by these animals. Some kinds of timber are much less liable to this kind of injury than others. The East Indian teak is said never to be attacked by either of these creatures, and live oak is compara- tively little injured by them. The Varieties of Timber used in carpentry, joinery, and pattern-making are very numerous ; and the forests of our own country yield immense quantities of some of the most useful kinds. They are divided into two great classes : Pine Woods, or the Coniferse, are distinguished by their spine or needle-like leaves and resinous turpentine-yield- ing sap. Leaf-wood comprehends all other timber-trees, and bears leaves of the ordinary broad, thin, and irregular shapes ; its sap is destitute of turpentine. The latter woods are usually best where strength, dura- bility, and hardness are demanded ; the former excel in light- ness, elasticity, and flexibility. The Leaf-woods are divided into two classes : (i) Those which have their medullary rays broad and well marked; (2) woods in which those rays are indistinct. These classes include each two sub-classes : (a) Those in which the annual rings are distinctly marked, as in the oak of 6o8 NON-METALLIC MATERIALS. the first, and in the ash of the second class ; (b) those in which the rings are obscure, as in beech of the first, and walnut and mahogany of the second class. White Pine (Pinus strobus) is a native of North America, and takes its name from the color of its wood. It grows in all kinds of soil. The best timber is found in cool, damp situations in the forests of the Northern United States and Canada, between the forty-third and forty-seventh paral- lels of north latitude. It rarely flourishes well as far south as Virginia. It grows to a great size, reaching a height of up- wards of 2CX) feet (6i metres), with a diameter of lo feet (3.05 metres) at the height of a man's shoulder from the ground. It is the tallest tree in our forests. It sometimes reaches the age of 350 years. Single logs have been cut 36 inches (91 centimetres) square and 60 feet (18.3 metres) long. Its wood is yellowish-white in color, light in weight, rather soft, free from knots, straight grained, and is very easily cut. It is du- rable only in dry air. It contains very little resin. Its leaves are very slender, and are pale green in color ; its cones are nearly cylindrical, and four or five inches (10 to 12.7 centi- metres) long. Its specific gravity is about O.70 green, and 0.50 seasoned, its weight being quoted at 44 and 30 pounds per cubic foot respectively (705 and 480 kilogrammes per cubic metre). It is used for light carpenters' and joiners' work, and is remarkably well adapted to pattern-makers' use. It has been employed' to a considerable extent in building wooden bridges. It is not a very strong wood, and swells or shrinks seri- ously when the hygrometric state of the atmosphere changes considerably. For many purposes its softness is a serious objection. The Canadian Red Pine (Pinus resinosa) is fcmnd growing on the poorer soils of the northern portion of the United States, and in Canada, reaching a height of 80 feet (24.4 metres), and attaining a diameter of 2 feet (.6 metres). It is wrongly called, in various localities, " Norway Pine " and " Yellow Pine." TIMBER. 609 The leaves are in pairs, and five or six inches long (12 to 1 5 centimetres). The wood is fine-grained and white, with a reddish tinge, somewhat soft, but quite strong and durable. It is so soft and flexible, and so readily worked, as to be a favorite timber for light work. It makes excellent planking and spars for ships. The American Yellow- Pine, " Spruce Pine" or Short- leaved Pine {Pinus mitis, Pinus variabilis), is found throughout the country, in dry sandy soils, from New England to Georgia. ^1°- i3i.-Pine. It attains a height of 60 feet (18.3 metres), and a diameter of 18 inches (45.6 centimetres). The trunk is straight and slender. Its cones are small, its leaves are in groups of threes, and from 3 to 5 inches (7^ to 12^ centimetres) long. The heart-wood is fine-grained, moderately resinous, strong and durable. The sap-wood is poor in quality, and decays rapidly. It is much used in carpentry, and for framing and floor- ing, and in ship-building; it is also used for the masts and yards of large vessels. The Southern Pine, "Long-leaved Pine" (Pinus ausiralis, Pinus palustris), is distributed along the Atlantic coast from Maryland southward, on sandy, hght soil. It is probably the most generally useful of our woods, and im- mense quantities are brought into market. Its name is very commonly confused with that of the pitch pine, and both kinds of wood are known in the Eastern States as hard pine. Both the yellow pine and the pitch pine are extensively used, by Atlantic ship-builders, for planking, beams, keelsons, etc., but seldom for any part of the frames. The yellow pine sometimes attains a height of 150 feet 39 6lO NON-METALLIC MATERIALS. (45.7 metres), and a diameter of 4 feet (12. i metres) ; but the pitch pine seldom exceeds two-thirds this size. The former is principally obtained from the States of Virginia, North Carolina, and Georgia, while the latter is abundant in all the Atlantic States south of Chesapeake Bay. The yellow pine required for navy-yard use is described as long-leaved, fine- grained. Southern yellow pine. Its leaves are rigid, and 8 to 11 inches (20.3 to 27.9 centi- metres) long ; they are dark green in color. The cones are 6 to 8 inches (l5-2 to 20.3 centimetres) long. It has but little sap-wood, and the heart-wood is of very uniform quality, its resinous matter being very regularly dis- tributed. Its grain is fine and close, and it has greater strength, durability, and hardness than any other species of pine. Though not so tough and elastic as white oak, the yellow pine, especially that from Georgia, successfully rivals it in stiffness. If a beam of each kind of timber, equal in dimen- sions, be supported at the ends, the oak beam will depart most from its " mould," but will break under about the same load. In dry situations the pine is extremely durable, and where the properties of lightness and solidity are required in com- bination, it is to be preferred to oak. Experiments upon the shrinkage of various woods, by Mr. James Jarvis, at the U. S. Navy-yard, Norfolk, Va., indi- cate that yellow ^ine should be cut in summer. The Pitch Pine {Pinus rigidd), is common throughout our country, frequenting sandy or lean rocky soils. The best qualities come from Florida. It is distinguished by peculiarly rough, dark bark, and by the abundance of its resin. Its leaves are in groups of , three, 3 to 5 inches long (7.6 to 12.7 centimetres). The wood is close-grained, heavy, free from knots, elastic, quite strong, and very durable. It is more dense than yellow pine ; which latter has the preference for all work to be cov- ered by paint. Pitch pine is very stiff, and moderately fine- grained. TIMBER. 6ll Fig. 132. — Red Fir. In using yellow and pitch pines, the best timber lot strength and durability is not necessarily that of the greatest density. The timber of greatest weight is often heavy simply because of the presence of a surplus of turpentine in its vessels. The Foreign Northern Pine, Yellow Fir, Red Fir, or Scotch Fir {Pinus sylvestris), is found in all parts of Northern Europe, including Great Britain, where the forests are largely com- posed of it. It is very much used in Europe, and is obtained in Great Britain, Norway, Sweden, and Russia, and from the Prussian ports of Memel, Dantzic, and Stettin. The logs are sometimes as large as 80 feet (24.4 metres) long and 26 inches (66 centi- metres) square. The yellow deals from Christiania are most durable, but a large waste occurs in working them, in conse quence of their large proportion of sap-wood. The durability of the better quality of this timber is con- sidered by some engineers to equal that of oak. Like the American white pine, it is excellently adapted for fram- ing, and for light carpenters', join- ers', and pattern-makers' work. In Great Britain, the American white pine is, however, consider- ably used instead of the native fir. The Cypress {Cupressus disticha, Taxodiuin distickuin), or deciduous cypress, is a tree of the pine family, having a trunk some- times 10 or even 12 feet (3.05 to 3.66 metres) in diameter, and at- taining the height of from 120 to Fig. 133.— Southern Cypress. 130 feet (36.58 to 39.63 metres). Its foliage is a delicate light green in color, the leaves linear, 6l2 NON-METALLIC MATERIALS. awl-shaped, and spreading, and borne upwards on slender branch lets. The tree is found from the Hudson to the Gulf of Mexico, and flourishes best in southern latitudes, attaining greatest size in the swamps of the South, where the soil is a deep, rich, black, and wet mud. The roots of old trees are often partly exposed and singularly contorted. The lower portion of the trunk is frequently hollow. The wood is considered excellent for many purposes. It is soft, light, straight-grained, and easily worked, and is im- perishable where covered with water. It is extensively used in those localities throughout which it is most abundantly distributed, and sometimes as a substitute for oak. The Qualities of Pine Timber are readily determined by a practised observer. Good wood has a close grain, and its slow growth should be evidenced by the thinness of the annual rings, which should not exceed a tenth of an inch (0.25 centimetre). The trunk, and consequently its rings, should be sym- metrical. The best timber is charged with resin, and this preserves it from decay, and gives it strength and elasticity ; its pres- ence is indicated by strong odor. The color of the wood should be a clear tint of yellow and red, alternating, and the texture should be very uniform, as well as the colors. The working of the timber gives a reliable indication of its quality. It should offer considerable resistance to splitting along the grain ; it should be strong and free from wooliness, and the cut of the saw and of the plane or chisel should leave smooth surfaces. The shavings and chips should be strong and elastic, and the former capable of being twisted about the fingers without breaking. The Firs are closely related to the pines, and furnish a large quantity of excellent timber to the markets of Europe and of America. The White Fir, Norway Spruce, or White Deal {Abies excelsa), grows in the mountainous portions of Northern Europe. It is tall and straight, excelling all its congeners in TIMBER. 613 these respects. It reaches the height of 100 feet (30.5 metres) and attains a diameter of 3 feet (.91 metre). Its cones are cyHndrical, 5 to 7 inches long (12.7 to 17.7 centimetres). It is used in Great Britain largely, being imported principally from Christiania and other Northern European ports. It is now frequently met with in North America. This wood adheres well to glue, and is quite durable and strong, but it is not equal to the best varieties of pine. It takes a fine polish, and is largely used for flooring and panelling, and is well adapted for spar-making. Burgundy pitch is ob- tained from this tree. The American Black Spruce Fir {Abies nigra) is so called from the dark color of its leaves. It is found in the rougher por- tions of the North Ameri- can forest-covered country, and grows to a height of 80 feet (24.3 metres). Its cones are but i or 2 inches long (2.5 or 5.2 centimetres). It is quite similar in quality to the Norway spruce fir, and excels it in toughness. It is rather less durable and is less dense ; it is also more liable to warp in seasoning. Hemlock Spruce Fir {Abies Canadensis) is found in the same range of climate as the black spruce, but it prefers a more hilly country. It forms extensive forests in Lower Canada. It attains a height of 70 feet (20.73 metres), and occasionally even 100 (30.5 metres), and reaches a diameter of 2 feet (0.61 «-<■-=>.._ Fig. 134.— "Spruce.' 6i4 NON-METALLIC MATERIALS. metre). The leaves are dark and stiff, four-sided, and needle- shaped. The cones are ^ or i inch (1.9 or 2,5 centimetres) long. The wood resembles that of the white spruce, and is generally more highly valued. Its strength, durability, lightness, and elasticity form a combination of good qualities that makes it, for some purposes, the best wood in our markets. The Red Spruce Fir {Abies rubra), or Newfoundland Red Pine, as it is also called, grows in the north-east por- tions of North America, and affords an excellent material, perhaps hardly excelled by the black spruce. Its size is about the same as that of the black variety. It is especially prized for yards and spars of ships. Fir timber has a specific gravity of from 0.6 to 0.8, weighing from 36. pounds dry to 48 pounds green per cubic foot (577 to 769 kilogrammes per cubic metre). The Larches {Larix Europcea, Larix Americana) are natives respectively of Europe and America. Their wood is hard and strong. Their leaves are very slen- der, light green in color, and short. Their cones are about one inch long. This wood has not the lightness nor the elasticity of white pine, but is tougher and more close- grained, and is far less in- flammable than are woods generally. Larch is hardly excelled by any other wood in dura- bility. The European larch was celebrated for this quality from a very early period. Even when exposed Fig. 135.— Larch. ^^ alternately wet and dry weather, it is quite durable, lasting sometimes thirty years TIMBER. 615 under most unfavorable conditions. The American variety of larch, known as Hackmatack, is highly prized by our ship- builders. It attains a height of 100 feet (30.5 metres), and a diameter of 3 feet (.91 metre). It is found from Virginia to Canada. The Linden, Basswood, Lime {Tilia Americana, T. Europma, etc.), is found ..._«isigi^m^^^ throughout a w i d e , fj. g^"^.»-^'y, range of climate in r 'T^^^' t *-*<*'" both the United States .'^-Iw fiV^'t'^'^t -^ and Europe, and has ■*"■ V '•«' 'UVi'A''''" '. '<■ many varieties. The T'^ ''-V§ -i* V- useful varieties are '*3k?'.l£j4i^-*^ trees of moderate size. bearing large, smooth, heart-shaped leaves a,- J^S-'f ^ ,, jg'*^' alternating on the ' '.i'''*''^.' * ' i-" s« stem, and having fra- . >'?^-^iS^^ME ^i^^V Fig. 136. — Basswood. grant flowers which are '^^^f^' ' ?^ favorites with the bees. The foliage is dense, and the tree is an ex- cellent shade tree, but very subject to the attacks of insects. The wood is yellowish-white, soft, and light, but moder- ately close-grained and tough. It is used largely for furniture, coarse carvers' work, and to some extent in carpentry. The inner bark, or "bast," is used for making coarse matting, baskets, etc. The Cedars and Junipers are woods of less general application than the pines ; but have, nevertheless, great value in construction. The White Cedar (Cupressus thy aides) is found on the Atlantic coast of the United States from New York to Georgia, wherever the soil is wet. It is the principal inhabi- tant of the interior swamps of New Jersey and of Virginia, and trunks are often found of large size, sound and merchant- able, lying far below the surface, embedded in mud and peat. 6i6 NON-METALLIC MATERIALS. It grows to a height of 80 feet (24.4 metres), and to 3 feet (.91 metre) in diameter, with a straight stem and branches up to within 30 feet (9.1 metres) of the top. Its resin is yellow, slightly odorous, and small in quantity. The cones are small, greenish in color, becoming bluish at the end of the season. The wood is odorous, soft, fine-grained, light, and easy- working, taking a red tint, and often a decided color, when seasoned. It resists the weather remarkably well, and is, therefore, used very extensively for shingles. The wood makes the best of railroad ties for light traffic, but is too soft for general use ; it makes excellent fencing and telegraph poles, and domestic utensils are often made of this wood. It is cut at all seasons, but best when the sap flows most slowly. The Virginia " Red Cedar " {Juniperus Virginiand) is found in nearly all parts of the United States and Canada. It is, when fully grown, from 30 to 50 feet (9.1 to 1 5.2 metres) high, and sometimes 12 inches (.3 metre) in diameter. It is found on dry, sterile, rough country. The wood is light in weight, weighing 32 pounds green, and 28 pounds seasoned, per cubic foot (512 and 448 kilogrammes per cubic metre). The color of the heart-wood is red, while that of the sap-wood is white. It is brittle, compact, and dura- ble, and has a strong charac- teristic odor and a bitter taste, which preserves it from the at- tack of insects. It is especially valuable for drawers, chests, boxes, and some kinds of furniture. When well-seasoned it makes excel- It is extensively used for covering Fig. 137.— Cedar. lent rulers and T-squares. lead-pencils, and is sometimes called Pencil Cedar. TIMBER. 617 The Bermuda Juniper {Juniperus Bermudiana), or Bermuda Cedar, is a native of the West Indies. It is harder and heavier than the pencil cedar, and has a similar odor and appearance. It is very durable when well seasoned and free from sap-wood, and has been considerably used by ship-build- ers for planking. These cedars, or more properly junipers, are largely used for drawers, wardrobes, and church furniture. The California " Cedars " grow to enormous size. Tar, Pitch, and Turpentine are obtained from the more resinous trees of the pine family. Tar is obtained by a rude distillation of the heart-wood of pine. It is viscous and semifluid at ordinary temperature, solid when cold, and quite liquid when heated. It is brown- ish-red in color, becoming black with age or by the presence of impurities, or by overheating when made. It iS' used for preserving cordage, and the oakum which is Fig. 138. — Tapping the Pine. used in calking the seams of vessels, and as a binding mate- rial in artificial fuels, and in some kinds of cement. 6i8 NON-METALLIC MATERIALS. Pitch is made by boiling tar until its consistency is consid- erably increased. It is hard at ordinary or low temperatures, but is softened by the heat of the hand. It is used as a cementing and preservative material. Rosin, or colophony, is a pitch obtained by distilling turpentine. The best is lightest in color. Turpentine is the sap of the pine. The tree is tapped an- nually when the sap is flowing most freely. White, or " vir- gin " turpentine, is obtained from the tree the first season ; during succeeding seasons the product becomes gradually darker, and is known as " yellow-dip." Trees are tapped twelve or fifteen years in succession. A large part of the tur- pentine in the market comes from North Carolina. The following description of the process of distillation may explain further : * A fifteen-barrel copper still, the barrel weighing 220 lbs. (100 kilogrammes), is charged early in the morning. Heat is applied until the mass attains a uniform temperature of from 212° to 316° Fahr. (100° to 158° Cent). This is contin- ued until the water contained in the crude turpentine as it comes from the forest has been driven off. The first product distilled over contains pyroligneous acid, formic acid, ether, and methylic alcohol, with water. This is known as low-wine. All the water having been distilled off, a small stream of cold water is now let in, so that the heat is kept at or below 316° Fahr. (158° Cent.), the boiling point of oil of turpen- tine. The oil of turpentine and water now come over, and the mixture is caught in a wooden tub. This tub is con- structed as follows : * The distillate is caught at A from the still, and separates into water and oil. At B Fig. 139. — Separator. * Scientific American, TIMBER. 619 there is an overflow spout, which discharges into the tub D. The water is kept low enough in the lower part of the tub to prevent its overflowing through the cock B into the recep- tacle D. From this receptacle it is put into oak casks, well secured with iron hoops, and thoroughly glued inside. The distiller tests the quality of the flow from time to time in a proof glass. The distillation is continued until the pro portion of fluid coming over is nine of water to one of oil of turpentine. At this stage the heat is withdrawn, the still-cap is taken off, and the hot rosin, which remains in a fluid state in the still, is drawn off by a valve or cock at the side of the still near the bottom. The yield of oil of turpentine from " virgin dip " is about 6 gallons (27 litres) to the barrel. The yield of oil of turpentine from " yellow dip " is about 4 gallons (18 litres) to the barrel. W. i- 111?.-'-! !. Fig. 140. —Turpentine Still. Venice turpentine is that obtained from the larch. It is sometimes imitated by mixing rosin and spirits of turpentine. 620 NON-METALLIC MATERIALS. Spirits of turpentine is the essential oil of turpentine, and is obtained by distillation. The Oaks form a most valuable class of timber-trees, and a large number of species are known and used. Of more than sixty species known to botanists, over forty are natives of North America, and several produce very excellent timber. The best kinds of oaks are, if properly prepared for use, the hardest and most durable of woods. Kept either under water or perfectly dry, oak has been known to last several centuries. It is strong, tough, and moderately stiff. There are, however, varieties of oak which yield inferior timber, and trees of the same species may yield either superior or inferior timber, according to the nature of the soil and the climate in Fig. 141.— Oaks. which they have grown. The texture is alternately dense and porous. The wood has a peculiar odor and taste, the latter being due to the presence of gallic acid, which, by contact with TIMBER. 621 iron, produces an ink which blackens the wood and corrodes the metal. The oaks grow on a great variety of soil, preferring a clayey subsoil overlaid with rich loam. The Live Oak {Quercus virens) is one of the best known ship-timber trees. It is evergreen, and grows on the sea-coast from Maryland to the Gulf of Mexico and the Mis- sissippi, and is now so scarce and so valuable that the govern- ment has reserved all of the Florida live-oak forests for naval purposes. The tree grows to a height of 60 feet (18.3 meti-es), and to a diameter of 4 feet (1.22 metres), but is usually 40 or 45 feet (12 to 13.7 metres) high, and 12 to t 8 inches (30.5 to 46 centi- metres) in diameter. The sap-wood is whitish in color. It is free from the glutinous matter which fills the capillary vessels of the denser heart-wood. Unlike other varieties of oak in our country, it is free from acid. This timber is used almost exclusively for the purposes of ship-building, and is the most costly ship timber in the market. It is heavy, compact, fine-grained, yellowish in color, and is the strongest and most durable of all American woods.* It is not well adapted to the reception of spike fastenings, as the grain refuses to receive the point in the cutting direc- tion, and permits splitting of the wood. There is no diffi- culty, however, in fastening with bolts and treenails. Live oak, if exposed long in the open air, in the rays of the sun, or to winter winds, will check badly. It does not re- quire many months of air seasoning, however, to fit it for its ordinary uses. The White Oak {Quercus alba) is a more common and a very valuable variety of oak. It is especially valuable for ship-building, for which its trunk furnishes the heavier beams, and its large roots and branches yield the compass timber. * The Author possesses a. live-oak cane, taken in 1865 from the /f«/ of the frigate United States, a naval vessel built very early in the present century. It is as perfectly sound, apparently, as when first cut. 622 NON-METALLIC MATERIALS. It is used for water-wheel shafts and steps, and other millwrights' works, and for artillery carriages. The wood from the roots makes beautiful furniture. The cost and the difficulties of working it preclude its extensive use. The bark is rich in tannin, and is of great value for tanning leather. This tree is found from Canada to the Carolinas, and is most abundant in the Mid- dle States, forming large forests west of the Alleghany range of mountains. It reaches a height of 8o feet (24.3 metres) and more, and its trunk is sometimes 6 or 7 feet (1.8 or 2.1 metres) in diameter. It is one of the few trees which retain any of their leaves throughout the winter. The leaf is deeply indented, long and narrow. Its bark is of a light grayish-white color, giving it its name ; the wood is light straw-colored, with a tinge of red, and is very tough, strong, d,nrable, elastic, and pliable, with strong lateral cohesion. It is very liable to shrink, warp, and crack in seasoning, and is therefore of little value for boards. The shrinkage amounts to about one thirty-second. The wood of trees 60 to 1 00 years of age is much tougher, particularly on high lands, than that of older trees. No cer- tain data exist for comparing the properties of white oak grown in various districts, but it is generally supposed that the best timber for durability is that grown near the Atlantic seaboard, or along the borders of the great lakes. Generally the strongest timber is grown on wet lands. The experi- ments of Jarvis prove, first, that there is ten per cent, in one year, and five per cent, in four years, more shrinkage in Fig. 142. — White Oak. TIMBER. 62 3 weight of the squared timber which is cut in the warm sea- son, than in that cut during the cold season ; secondly, that in the case of round logs, in bark, there is eight per cent, in one year, and seventeen per cent, in four years, more loss by evaporation if cut in the summer season. It has a specific gravity of from .7 to i.i, weighing from 44 pounds, dry, to 70 pounds, green, per cubic foot (705 to 1,121 kilogrammes per cubic metre). The Post Oak (^Querctts obtusilobd), or Iron Oak, is common in Maryland, and eastof theAUeghanies in Virginia, where it is also called the Box White Oak. It is occasionally found as far north as New York and New England. It produces excellent timber, but seldom exceeds a foot or 15 inches in diameter, and a height of 50 feet (15.24 metres). The wood is of a yellowish hue, close-grained, and is often superior to the white oak in durability and strength. It is also finer grained. It is a most excellent wood for construc- tive purposes where of sufficient size, and is used for knees in ship-building, and for staves. V .?;,! "^ Fig. 143. — Swamp Oak. The Swamp Post Oak is found in the Carolinas and in Georgia, in swampy and often inaccessible districts. It is 624 NON-METALLIC MATERIALS. larger than the preceding species, and is an excellent timber- tree. The Red Oak {Quercus rubra) is a Canadian tree, which grows with considerably greater rapidity than either of the preceding. It is usually smaller, but attains a height of loo feet (30.5 metres). Its leaves change to a red color before falling, in autumn, and this fact gives the tree its name. The wood is easy to work, light and spongy, and lacks the durabihty of the better kinds. It is coarse-grained, and is only used to any considerable extent for staves. The Rock Chestnut Oak {Quercus primus monti- cold) grows in the Middle States, and as far north as New England. It is most plentiful among the Alleghanies, and is more durable, and is, in other respects, nearly as valuable as the white oak, but its scarcity prevents its equally extensive use. The Chestnut White Oak {Quercus primus) is found in the Southern Atlantic States. It produces a strong and durable wood, although not equal to the white or the post oak. It is used to some extent in wheelwrights' work, and is considered nearly equal to white oak for ships' frames. The British Oak {Quercus pedunculatd) is found all over Europe, and is most common in England and France. It grows to a height of from 70 to 100 feet (20.7 to 30.5 me- tres), and attains a diameter of 6 feet (1.8 metres). The wood has a light brown or reddish tinge, with numer- ous large medullary rays. It is tough and strong, quite hard, straight-grained, free from knots, splitting freely, and is said to be one of the best kinds of oak for joists, or where a stiff timber is desired. It bears changes from wet to dry, and the reverse, well, and is almost unalterable when protected from the actiqp of oxidizing agents, when in air or under water. The Sessile Fruited Oak {Quercus scssiliflora) is another very valuable European timber-tree, which is most common in the German forests. The wood is rather dark, of uniform color and grain. TIMBER. 625 heavy, hard, and quite elastic, resembling chestnut slightly in appearance. Like other varieties of oak, it is liable to warp and crack in seasoning. Its durability is equal to that of the preceding sort. It is somewhat more difficult to work. The Beech {Fagus sylvaticd) is a native of Great Britain and of Northern Europe. Its closeness and uniformity of texture make it valuable for tool-makers and furniture manufacturers, a large propor- tion of ordinary English furniture being made of it. It is used in dry situations by millwrights for the cogs of mortice gears. The lighter-colored wood is best. The American Beeches (Fagus sylvestra and Fagus ferrii- ginea), the white and the red, are of somewhat less value, although similar in general characteristics. It generally congregates in great quantities wherever the soil is most favorable ; hundreds of acres are sometimes cov- ered with this alone. Such tracts are familiarly called beech- woods. Beech is used for furniture, gearing, submerged water- wheel bearings, tool handles, plane stocks, saddle-trees, wallets, chair-making, etc. The Chestnut [Castanea vescd) is a native both of Europe and America. It attains a height of 70 feet (20.7 metres), and a diameter in our Middle States of 6 feet (1.8 metres); its average size is about 45 feet (13.7 metres) high, and 2 feet (.6 metre) diameter. It is a very long-lived tree, and has been known to attain the age of lOOO years. When of great age, it is invariably hollow, and valueless for timber. It is very similar in color to white oak, although exhibit- ing a stronger contrast between sap and heart-wood than the latter. It is distinguished from oak very readily by the lack of marked medullary rays, and by its lightness. The wood is of great value. It is extremely durable, last- ing under water even longer than oak or elm. It is hard and compact, and, when young, tough and flexible ; but it acquires brittleness with age. Breaking transversely, it first bends considerably, and then fractures suddenly. 40 636 NON-METALLIC MATERIALS. Fig. 144. — The Great Chestnut of Mount Etna, TIMBER. 627 The Ash {Fraxinus excelsior) of Europe, and the White Ash (J^raxinus Americana) of America, are very valu- able timber-trees. They grow to a height of 60 feet (18.5 metres), and acquire a diameter of 20 inches (50.8 centi- metres) in rich, moist, loamy soils. They have no observa- ble sap-wood. Their woods have many useful applications. Ash is quite similar in color to oak, and in texture to chestnut. It is straight-grained, remarkably tough and elastic, excelling in these qualities all other common woods, and answers admi- rably for handspikes, heavy oars, ship blocks, tool handles, the wooden portion and framing of machinery, and for wheel carriages and agricultural implements. It is durable under cover, but decays rapidly if exposed to the weather. The Common Elm {Ulmus Americana) is a native of New England, where it attains a height of 100 feet (30.5 me- tres), and a diameter of 6 feet (1.83 metres) or more. It grows along river-banks and in rich soil, and is a noble, ornamental tree. The heart-wood is brown, and the sap-wood is nearly white. The wood is porous and cross-grained, and does not split when nails are driven into it. It is most valued for its great durability in situations where it is constantly wet. It is used for piles under wet foundations, framing and sheath- ing around wheel pits in mills, for pumps, water-ways, the keels of ships, planking, and for flumes and water conduits. It is used also by wheelwrights. The European Elm {Ulmus campestris and several other species) is said to be even harder and more durable than the American, and is applied to similar uses. It is hard, flexible, and tough, but difficult to work. The wood is used for wheel naves and rims, and for wh'eelwrights' use generally. Wych Elm is the best variety. The Canada Elm, or Mountain Elm {Ulmus race, mosa), is a less valuable tree. Its wood is close and iine-grained, flexible and tough, but it shrinks, twists, and cracks in seasoning. 628 NON-METALLIC MATERIALS. The Locust, or Common Acacia {JRobinia pseuda- acacid), is a flowering tree found in the mountainous and hilly portions of the country from Canada to the Southern States. It grows rapidly, and reaches a height of 70 feet (20.7 me- tres), with a diameter of 4 feet (1.21 metres) ; but it is usually considered full-grown if 40 feet(i2.i metres) high, and a foot (.305 metre) in diameter. It is a fine ornamental tree. The wood has a peculiar greenish-yellow color, slightly resembling boxwood. The structure is alternately very com- pact and quite porous ; its annual rings are thus very distinctly marked. It exhibits no medullary rays, and the wood has neither taste nor odor. It is a very valuable timber, especially, for fence-posts and rails, or boarding, but is seldom found of suf- ficient size and quantity to be used in the latter form. It turns well in the lathe, but otherwise is difificult to cut and work. It has great torsional strength and resilience, excelling all other common woods in this quality. The Hickory, or White Walnut {Gary a alba), is a tall, handsome, American timber-tree, having great value for many purposes. It is common throughout the northern and eastern portions of the United States. It grows to the height of 50 or 60 feet (15.2 or 18.3 metres), and reaches a diameter of 3 feet (.9 metre). The wood is alternately very dense and somewhat porous, and it is one of the heaviest of our woods. It is very strong and stiiT, yet el^tic and tough. The wood, when freshly cut, has a slightly bitter taste and a mild odor ; it is then almost white in color, but by ex- posure becomes gradually darker. Its heart-wood contains brownish-colored pores. It malces excellent cogs for mortice gears, and is well adapted for handspikes, although rather heavy, and for axles, shafts, spokes, and other wheelwrights' work. The Black Walnut (Juglans nigra) is found throughout our Middle and Western States, and as far south as the Gulf of Mexico. The tree presents a fine appearance, attains considerable size, and yields a much-prized timber. TIMBER. 629 The wood varies considerably in quality. It is of a brown color, approaching red in some specimens, and of a dark chocolate color in others; The sap-wood is frequently quite light in color. The best wood has a fine grain and a dense structure, although usually ex- celled in both particu- lars by good mahogany. It isnearlj- as strong as mahogany, and is tougher. It is durable, and easily worked. It is more generally used in the United States for furniture and for ornamental purposes Fig. 145. — Black Walnut. than any other wood, and immense quantities of it are annu, ally worked up. The Cherry and Plum (Prunus) are found both in Europe and America. The wood is excellent, quite hard, of a pale pinkish brown, or yellow color, and of close grain. It makes very neat furniture, and is used for handles of tools. As its price is about that of panel pine, it is very extensively used for hard patterns. The Holly (Ilex opacd) is an American wood, found from Maine to Pennsylvania. The tree attains a height of 30 or 40 feet (9.1 to 12.1 metres), is distinguished by the bright red of its berries, and by its glossy leaves. Fig. 146. — Cherry. 630 NON-METALLIC MATERIALS. The wood is white in colqr, close in texture, with a beauti- fully fine grain. It requires to be carefully and thoroughly seasoned, and then is found most excellent for "T-squares," painted wooden wares, cabinet work, blocks for calico-printers, and for turned work. The Maple {Acer) is another American wood. It is remarkable for the beautiful variety of its grain. It is very largely used in joiners' Fig. 147.-H0LLY. ^Q^-^^ and in cabinet- making. It has been used, with good results, as a packing in pump-buckets. The Sugar Maple, or Bird's-Eye Maple (Acer saccharinum), produces a sap charged with sugar, and the tree is there- fore called the Sugar Maple. This wood is full of small knots which give it its name, and which make it the most beautiful of our light -colored woods. The Dogwood (Cornus Florida) is a small deciduous tree attaining a height of 30 feet (9.1 metres), and bearing beautiful large white flowers, Massachusetts to Florida, in moist, rich soil Fig. 148 — Maple. It grows from TIMBER. 631 The wood is hard, fine, and close-grained, rather difficult to work, and can be given a fine polish ; it is used in making tool handles, mallets, drifts, toys, harrow-teeth, hames for harnesses, and for small articles of turned work. Mahogany (Swietenia mahogani) is a West Indian and Central American tree, growing in greatest size and perfection in the fertile regions of Honduras, and in the valleys of Cuba. The tree is remarkable for its beauty of form and rapidity of growth, as well as for its noble size. Specimens measuring 20 feet (6.1 metres) in circumference are often found. Mahogany is of various shades of brownish red, quite uni- form in its tints in the same piece, but varying greatly in dif- ferent specimens. The texture is very uniform, and its medullary rays and annual rings are not usually very well marked. The pores are quite noticeable, and, in mahogany from the West Indian islands, are filled with a white sub- stance which distinguishes this variety, called also Spanish Mahog- any, from the Honduras wood. It has no perceptible taste, and but slight odor. In seasoning, it is less subject to cracking or dis- tortion than almost any known wood, which fact, and its excep- tional beauty, make it a much sought and highly prized wood for _ ° . , f Fig- 149.— Mahogany. fine furniture, and for many special uses, among which those of the pattern-maker are not the least important. The Honduras wood, often called Baywood, holds glue re- markably well. The Spanish mahogany is imported in logs measuring, often, 2 feet (.61 metre) square, and 10 feet (3.05 metres) long. The Honduras mahogany comes in logs 14 or 15 feet (4.6 metres) long, and from 2 to 4 feet (.6 to 1.2 metres) square. The former is harder, of closer grain, and darker in color than the latter, which is comparatively porous, of irreg> ular color, and is rather a weaker wood. 632 NON-METALLIC MATERIALS. Mahogany is also found in the East Indies and in Africa. It is of excellent quality, but less beautiful than the American woods. Its specific gravity is .8. Lignum- Vitse {Guaiacum officinale) is obtained from the West Indies in logs of small size, and 3 to 12 feet (.9 to 3.6 metres) long. It is the hardest and heaviest wood gener- ally used in the arts, its specific gravity being about i.5. The wood is dark brown in the heart, and light yellow in the sap- wood. Its immense strength and hardness make it very valuable for sheaves of pulleys, and ships' blocks, or wher- ever great weight and friction are to be sustained. In making sheaves, care is usually taken to turn them so as to leave a ring of sap-wood on the outside, and the heart-wood within. The sheave is thus rendered less liable to crack. Lignum-Vitce is used for steps of water-wheels, for the stern or outboard bearings of the screw shafts of steam- vessels, and, occasionally, for other kinds of machinery bear- ings. Thus used, it bears an immense pressure under water, without wear or heating, and is better in such positions than any metal. It is necessary to secure efficient lubrication with water, as, although the friction is greater than if lubricated with oil, the latter lubricant does not effectively carry away the heat developed. The " end grain " should take the wear, if possible. The Spanish Cedar {Cedrela odoratd) is a West Indian wood, red in color, soft, light in weight, brittle, and odorous. It is best known as the material of which cigar-boxes are made. The Teak ( Tectonia grandis) is an extremely valuable East Indian wood. It is also called Indian oak. Although comparatively little known in this country, it is very exten- sively used in Great Britain by ship-builders. The finest qualities come from the forests of Burmah, Ceylon, Malabar, and Java, where it grows to the height of 150 feet (45.7 metres), with wide-spreading branches, and a straight, grace- ful trunk which is sometimes 9 feet (2.75 metres) in diameter. The wood is said by British ship-builders to be the best in TIMBER. 633 the world for their purposes as well as for general ship-car- pentry. It has some resemblance to oak in its color, but it is rather lighter, and is more uniform in density and in com- pactness of grain. Its specific gravity, seasoned, is about .6. It is light, strong, and durable, and is easily worked. It seasons quickly, requiring comparatively little drying. It is somewhat liable to check. It is less frequently attacked by insects than other woods, its peculiar oily, odorous, and perhaps poisonous, sap generally preserving it from even the white ant and from the teredo. The acidity of the sap of the common oak forbids the use of iron fastenings ; but the teak, to the other good qualities of oak, adds that of preserving iron embedded in it, by its oily sap. • Camphor Wood {Guttiferd) is also a valuable East Indian wood. It grows to a large size. The wood is very strong, durable, and easily worked. It weighs about 70 pounds per cubic foot (1,121 kilo- grammes per cubic metre). It has a powerful odor which preserves it from the attacks of insects and of marine animals. Boxwood {Buxus Balearicus) is usually of South European and Asiatic growth, but it is found also in America. The tree is low, and the imported logs are seldom over a foot (.305 metre) in diameter. The wood is yellow, brighter in color than our locust, with thin bark and numerous small knots, and is often twisted and somewhat unsound. It is extraordinarily smooth and com- pact in texture. It is used principally for small work. The engraver uses it almost exclusively, and it is largely used for rulers and scales, and for small turned work. Ebony (Diospyros) is found in nearly all tropical countries. The best {D. ebenus) comes from Mauritius. It is black (sometimes jet black), extremely hard and heavy, with a fine, close grain. It is chiefly applied to ornamental purposes, and is used by the engineer for some kinds of model work. A green ebony, so called (Americamis ebenus), is found in the West Indies. 634 NON-METALLIC MATERIALS. Lancewood {Uvaria lanceolata) is brought from the West Indies. It is lighter in color than boxwood, splits easily, but is very tough, strong, and elastic. It is, therefore, well adapted for pole-springs, and is useful wherever an elastic and strong wood may be needed. Greenheart [Nectandra RodicBi) is brought from the West Indies and the north-east coast of South America in logs from 30 to 50 feet (9.1 to 15.2 metres) long, and from I to 2 feet (.61 metre) square in section. The wood is dark green varying to dark chestnut in color, sound, straight-grained, strong, elastic, and tough. It is very heavy, having a specific gravity of about 1.15. When broken, it yields suddenly and completely. It is also very durable, resisting both weather-wear and the attacks of insects re- markably well. It is used for ship-work, engine-keelsons, beams, and piles. Rosewood {Amyris balsamifera) is a native of tropi- cal America, the best wood coming from Brazil. It is the most beautiful and highly prized of the dark or- namental woods. Its color is a very dark brown, or nearly black, shading off in spots into a deep, rich, brownish red, and presenting a beautiful variety of color and of patterns in its grain. It is hard and heavy, rather difficult to work, and takes a beautiful polish. It is largely used in the form of veneers. Timber is measured, when bought in market, either by the cubic foot or by board measure. The unit of the latter is the square foot of one inch thickness, and is denoted by the abbreviation B.M. Sawed or hewed timber is often measured by the cubic foot. Round timber is measured by multiplying the length by the square of one-fourth its mean girth to obtain the cubic contents. Oak timber should measure in the shortest logs one foot or more in length for each inch in diameter. Timber sup- plied for general purposes is usually cut to a standard length for convenience of measurement. CHAPTER XX. STRENGTH OF TIMBER; Its Special Adaptations and its Preservation, The woods vary immensely in strength, and even in the same kind there may be a great variation among several specimens, arising from differences of age, and of climate, soil, exposure, seasoning, any circumstances, in fact, which may differently affect each individual tree. Wherever a definite statement of strength is hereafter given, it will be understood that it applies to well-preserved and well-seasoned mature specimens of the kind referred to. As a general rule, the heart-wood of the tree is strongest and most uniform in character. If the tree has begun to decay while standing, however, the heart-wood is first affected. A tree, sound when felled, decays externally first, the sap- wood usually rotting away much sooner than the heart. The pines are rich in resin, which is an excellent preserva- tive, and as it abounds principally in the heart-wood, knotty portions of these trees are almost indestructible by exposure to the atmosphere. It is evident that experience and excel- lent Judgment are required to determine when the tree has arrived at just the proper age to yield the best and strongest timber. After the tree has been felled, the strength of its wood is largely influenced by the method of seasoning. If this be done gradually and thoroughly, the seasoned wood is far stronger than the green ; sometimes it is of double strength. If seasoned in oil, as described on page 6oi, on season- ing, the strength of hickory has been found by Mr, Geo. H. Corliss, who first made the most successful experiments, to be upwards of fifteen per cent, greater than good speci- mens seasoned in the usual manner. This is confirmed by Him, who found a gain due to this process of from ten to twenty per cent, with various woods. 636 NON-METALLIC MATERIALS. In fir, the thinner the annual layers, the greater the coef- ficient of elasticity. In other woods, no difference was de^ tected, arising from this cause. Timber has no defined limit of elasticity. One is taken by some writers, assuming as a limit in extension that point at which the set becomes g„^',„„ ofthe original length (.00,005 /), It may be taken, for purposes of estimation, at one-third 01 one-fourth of the breaking weight. Coefficients of Elasticity. — The following values of E are given by various experimenters : COEFFICIENTS OF ELASTICITY. BRITISH. METRIC. Lbs. on Sq. In. Kg. on Sq. Centim 1,600,000 112,480 1,800,000 126,540 1,250,000 91,250 1,500,000 105,450 1,800,000 126,540 1,200,000 84,360 1,400,000 98,420 1,000,000 70,300 1,400,000 98,420 1,700,000 119,510 1,900,000 133,570 1,800,000 126,540 1,600,000 112,400 1,000,000 70,380 2,100,000 147,030 1,400,000 98,420 Ash Box Chestnut, dry . . . . , Elm Fir, Baltic Fir, New England, Larch Lignum-Vitse Mahogany Oak, English . . . . , Pine, Pitch " Red " Yellow " White Teak, Indian Willow The Factors of Safety used with the woods are gen- erally large, especially where the attempt is made to use it in tension, or when beams are fished or scarfed. They may be taken, for ordinary work, at 5 for " dead " loads, 10 for a moving load, and 10 to 20 under shock. In the latter case, however, they should be carefully determined after calcula- tion of the resilience of the parts attacked. The Tenacity of timber is very variable. The fol- lowing are values of T for good samples. STRENGTH OF TIMBER, 637 CO-EFFICIENTS OF TENSILE RESISTANCE. Ash Birch, Black Beech , Box , California Spruce Cedar, Bermuda " Guadaloupe , Chestnut , " Horse Cypress Elm Fir (New England Spruce) . " Riga Greenheart Holly Hickory, American Lancewood Larch Lignum-Vitss Locust Mahogany, Honduras " best Spanish. . . Maple Oak, American Live " " White " English " best English Oregon Pine Pear Pine, Pitch " Red " White " Yellow Plum Poplar Spruce ■pgak Walnut, Black Willow Lbs. per Sq. In 10,000 to 15,000 7,000 8,000 10,000 12,000 4, COO 5,000 7,000 8,000 4,000 8,000 5,000 5,000 6,000 10,000 10,000 8,000 6,000 10,000 10,000 5,000 8,000 8,000 10,000 10,000 g,ooo 12,000 9,000 7,000 8,000 5,000 3,000 5,000 7,000 7,000 5,000 10,000 8,000 10,000 10,000 12,000 15,000 14,000 7,500 9,500 10, 500 12,000 6,000 13,000 10,000 12,500 9,000 15,000 14,000 15,000 10,000 12,000 15,000 8,000 15,000 10,000 14,000 10,000 10,000 8,000 7,500 12,000 10,000 10,000 15,000 Kg. per. Sq. Cm. 703 to 492 562 703 844 281 352 492 562 281 562 352 352 422 703 703 562 4^2 703 703 350 562 562 703 703 633 844 633 492 562 352 362 352 492 492 352 703 562 703 1,05s 703 844 1,055 984 527 668 738 844 422 914 703 879 633 1,055 984 10,55 703 844 1,655 560 1,055 703 984 703 703 562 844 703 703 1,055 Across the grain the tenacity is much less, being for the pines and spruce woods from one-tenth to one-twentieth ; and in harder woods from one-sixth to one-fourth the figures just given. In oak it is one-fourth, in pine hardly one-tenth. %i NON-METALLIC MATERIALS. The Crushing Resistance of timber is as variable as its tenacity. Mean values for good quality only can be given. The following Moduli of Crushing Strength are deduced from experiments upon pieces one inch (2.54 centimetres) in diameter, and two inches (5.08 centimetres) long. Hodgkinson found the compressive strength of wet wood to be frequently less than half that of dry. COEFFICIENTS OF RESISTANCE TO CRUSHING. [In direction, parallel with fibres.] Alder Ash Beech Birch , ' ' English Box Cedar Cherry Chestnut Elm Greenheart Hickory Larch Locust Lignum-Vitse Maple Mahogany, Spanish Oak, English " Live " White Pear Pine, Red " White " Yellow Spruce Teak Walnut, Black White Waiow BRITISH. Lbs. per Sq. In. 6,000 to 7,000 4,600 " 8,000 8,000 " 9,000 6,000 " 10,000 5,000 " 6,500 8,000 " 10,000 4,000 " 6,500 5,000 " 6,500 4,000 " 4,800 8,000 " 10,000 10,000 " 14,000 8,000 " 3,000 " 7,500 " 8,000 " 5,000 " 7,000 •■ 6,500 8,000 •■ 10,000 5,500 ■' 8,000 7,500 7.500 6,000 10,000 6,000 10,000 7,000 9,000 6,000 9,800 5,500 9.500 9,600 6,000 8,000 10,000 6,000 " 3,000 " 6,500 " 4,500 " 6,000 " 5.600 " 7,500 " 3,000 " Kg. per Sq. Cm. 422 to 492 323 " 562 562 422 352 562 281 352 281 562 703 562 211 527 562 352 492 457 562 587 537 422 211 457 316 422 394 527 211 633 703 457 703 457 457 337 703 984 689 387 668 675 422 562 703 703 562 " 527 " 422 " 703 " 422 " 703 "492 "• 633 " 422 In many cases it will be noticed that the tensile strength of wood is double its resistance to crushing, even in short pieces. STRENGTH OF TIMBER. 039 In tests hereinafter referred to the author has found the following coefificients of compression, material tested dry : * COEFFICIENTS OF RESISTANCE TO CRUSHING. California Spruce. Oregon Pine 9, 200 to 12,800 9,200 " 11,500 647 to 900 647 " 808 Across the grain, the resistance to crushing is from 1,000 lbs. per square inch (703 kilogrammes per square centimetre) upward, with various ordinary woods, but very few experi- ments have been made to determine it. A pressure of 1,000 pounds per square inch (703 kilo- grammes per square centimetre) indents white pine .1 inch (.25 centimetre) ; yellow pine, .004 (.01 centimetre) ; and the hard woods to an extent which is too slight to be de- tected. Long Pillars yield by bending. A long series of experiments were made by Hodgkinson, and his principal deductions were the following: Flat-ended pillars, of considerable length in proportion to their diameter, offer about three times the resistance of similar pillars with rounded ends. One end being rounded and the other flat, the pillar has a strength which is the arithmetical mean between the pre- vious two cases. Both ends being fixed in one case, and both rounded in another, the cross-section being equal, a pillar of a given length in the second case has no more strength than one of double that length and of the first form. The strength of a pillar may be increased one-seventh by enlarging it in the middle. Hodgkinson's Formulas ; Gordon's. — Hodgkinson * Some of the best experimental work on the strength of American woods has been done by Mr. R. G. Hatfield, and the results are published in the Ameiican House Carpenter, John Wiley & Sons, publishers. ,640 NON-METALLIC MATERIALS. deduced from experiment, for the formula of Euler, for square, flat-ended, oak timber : and for red pine, in which P = crushing weight in gross tons, d = thickness of the pillar in inches, L = length of pillar in feet. Where the pillar is less than thirty, and more than four or five diameters in length, W= ^^^ P+}iCK where W = strength of the column in gross tons. F = the strength given by the preceding formulas (7 org). C = the modulus of crushing resistance given in the table. K = the area of cross-section in square inches. A more usual formula is, in form, that of Gordon, some- times called Rankine's. Rankine's modification of the latter is the following ; the crushing weight in pounds : I + in which S is the sectional area in square inches, a and / con- stants, and / and d the length and diameter in inches. He gives for the value of a and /, for timber, 188' and 7,200 respectively. The experiments of C. S. Smith give, for well- seasoned yellow pine,/= S.ooo> ^ = 250. Morin adopts Euler's rule : in which P' is the load in kilogrammes, d the diameter in cen- STRENGTH OF TIMBER. 64I timetres, and / the length in decimetres. A is taken for pine timber at 160 for a safe load. This value of A varies with the modulus of resistance to compression. It is good practice invariably to limit the load on col- umns and other struts, to that which fails to cause perceptible flexure, and never to exceed that which causes deflection to a degree beyond which a great increase may be expected to occur with comparatively Httle additional load. This latter point is reached with from one-third to one-half the breaking load. In making struts of timber, Laslett states that his experi- ments indicate that the ratio of area of cross-section in square inches to length of inches, should not be less than from about 0.8 to i.o (using metric measures, cross-section in square centimetres = 2 length in metres), and that a resistance to crushing may then be anticipated of nearly the maximum obtained with cubic specimen, which conclusion is also reached by later experimenters. The relative values of timber and iron for columns are not far from the ratio of i to 10. It is sometimes necessary, in very long columns, to secure stiffness, as well as strength. The following formulas are given in Tredgold's Carpentry, for pillars above thirty diam- eters long : ^* W = A -y^ for square pillars W = A — U— for rectangular, and L 1.7 D' W = A ^j^ for cylindrical pillars. where W = safe load in pounds, b, t, and d = the breadth, thickness, or diameter in inches, L = the length in feet. The value of the coefficient A is about 1,500 for beech, chestnut, elm, and white pine; 2,000 for ash and mahogany; 2,500 for teak and Dantzic oak, and 2,200 for red pine. 41 642 NON-METALLIC MATERIALS. Columns for Mills. — During the year 1881, Prof. Lanza,* of the Massachusetts Institute of Technology, con- ducted a series of experiments- on full-sized wooden columns, for the purpose of determining what shape and proportions were best adapted for the support of mill flooring. Two series of tests were made, also, to determine the actual crushing strength of the wood used, with the following results : RESISTANCE TO CRUSHITfG, Average crashing strength of Yellow Pine. " " White Oak.. Whitewood . KG. PR. SQ. CM. 307-5 232.6 210.6 These iigures are deduced from the tests of unselected material, and therefore fall considerably below those ordin- arily given. For comparison with the above the following tables, of re- sults obtained at the Watertown Arsenal, will be found in- teresting : CRUSHING STRENGTH OF YELLOW PINE, Very Straight Grained, Twenty Years' Seasoning. LENGTH. DIMENSION OF SECTION. CRUSHING STRENGTH. ARSENAL • FORM OF INCHES. METRES. INCHES. CENTIMETRES LBS. PER SQ. IN. KILOS. PER SQ. CM. 573 20.4 SI. 82 Circular. 10.2 diam. 23.91 diam. 6,676 467.32 5?9 119.95 119. 9 304.67 304.07 Rectangular.. ro.97 X II 10.96 X 10. 96 27.86 X 27.93 27.8s X 27.83 6,230 6-SS2 436.1 458.64 582 20 50.8 9x9 8.02 X 8.02 22.86 X 22.86 8,322 582.54 S83 16 40.64 (i 20.37 X 20.37 8,165 571.55 .3^ 584 li 4x4 10. t6 X 10. r6 7,394 517.58 58s 2 7.62 13.24 u l.S X 1-5 3.81 X 3.81 SiS93 387.31 n.a 586 6 u 3X3 7.63 J 7.62 7.62 ; 7.62 8,644 60s .•8 1*: 587 6 IS.'4 3x3 8,133 569.31 588 3 7.62 1 *' i.S X 1.5 3.81 X 3.8i 8,3=9 583.03 *b' 589 3 7.62 '* i.S ** I.S 3.81 X 3.81 8,302 581.14 *5 59° 3 7.62 ** i.S>< 1-5 3.81 X 3.81 6,35S 444. 85 J < Avera 7,386 517.02 * Boston Journal of Commerce, January 28, 18 STRENGTH OF TIMBER. 643 CRUSHING STRENGTH OF YELLOW PINE. Very slow growth . LENGTH. FOKM OF SECTION. DIMENSION 3F SECTION, CRUSHING STRENGTIT. NUMBER. INCHES. CENTI- METRES. INCHES. CENTIMETRES LBS. PER SQ. IN. KILOS. PER SQ. CM. 591 592 593 14 17.2 19. 1 35-56 Rectangular 4.6 X 4.6 5'3 X 5.3 11.68 X 11.68 13.46 X 13.46 9i947 10,250 7,820 696.29 717-S 547-4 Avera 9.339 653.73 CRUSHING STRENGTH OF YELLOW PINE. Very green and wet. ARSENAL LENGTH. FORM OF SECTION. DIMENSION OF SECTION. CRUSHING STRENGTH, NUMBER. INCHES. CENTI- METRES. INCHES. CENTIMETRES LBS. PERSQ. IN. KILOS. PER SQ. CM. 691 692 714 180 180 ' 180 459 459 459 Open rect. 16 X 13.63 16. 2 X 7 17 X 8.7s 40.S4 X 34.67 41.15 X 17.78 44,18 X 22.22 3,070 2,795 3,t8o 212. I 195-65 222.6 Avera 3,015 2H.05 CRUSHING STRENGTH OF SPRUCE. LENGTH. DIMENSION OF SECTION. CRUSHING STRENGTH. FORM OF SECTION, NUMBER. CENTI- METRES. INCHES. INCHES. CENTIMETRES LBS. PER SQ. KILOS. PER 56s 24 60. r6 Rectangular . ->% X 1% 13.65 X 13.65 4.946 346.23 566 567 24 3S 60.96 91-44 " U u u ;; 4,811 4,874 336.77 340-98 568 91.44 4,500 315-00 56, 60 152-4 U (> 4,451 3"-57 570 60 '52*L 4,943 346.01 571 120 304-68 3,967 277-67 572 J20 304-6B 4,908 343-56 60 XX2.4 5,275 369-25 30 7.62 " u u I (( 5<372 376.04 IS 38.08 *' 5,754 402.78 977 121.2 307-85 Circular. 12.4 diam. 31*5 4,681 327.67 644 NON-METALLIC MATERIALS. We have the following average values for crushing strength of yellow pine : nESISTANCE TO CRUSHING. LBS. PER SQ. IN. KLS. PR. SQ. CM. Pine, straight grained, well seasoned, Arsenal test " slow growth, " " " " " very green and wet, " " " as used in Lanza's tests " C. Shaler Smith's tests ■7,386 9.339 3.015 4,400 5,000 517-02 653-73 211.05 308.CO 350.00 This shows a great variation between the figures of care-, fully made and authentic tests. These differences are evi- dently due both to the selection of timber and to the season- ing of material. Lanza recommends that columns should be bored from one end only, and this boring should extend throughout the length of the column. When columns are bored from both ends so as to meet in the middle, the two borings are apt to be eccentric, thereby weakening the piece. The object of the boring is to allow free access of the air to all parts of the wood. Resistance to Shearing is offered when it is at- tempted to divide the piece by a pair of forces acting along the same line in opposite directions, and parallel to the plane of separation. The shearing may take place in the case of timber, either along the grain on a plane parallel to the direction of the fibre, or across the grain in the same plane, or it may take place in a plane to which all the fibres are perpendicular. In each of these three cases, the modulus of shearing resistance has a different value. In each case the resistance is proportional to the area of the section ruptured, and is generally independent of its form. Where, however, the form is such that all parts of the section strained cannot act together in resisting shearing, the modu- STRENGTH OF TIMBER. 64s lus may be greatly reduced. Where, for example, the sec- tion is long and narrow, it will yield far more readily when attacked at the narrow, than when the shearing begins on the wider side. The following values of the Modulus of Shearing, are given by R. G. Hatfield for cases where the force acts along the grain, and parallel with the fibres : COEFFICIENT OF DETRUSIVE SHEARING. Chestnut. Hemlock. Locust. . . Oak Lbs. per Sq. Inch. 690 540 1,180 780 METRIC. Kg. per Sq. Cm. 48 38 83 55 Pine, Ohio. . . '" Spr'ce (Fir) " White.... " Yellow. . . Lbs. per Sq. Inch. 388 470 490 510 Kg.perSq. Cm. 27 33 34 3S A knowledge of this modulus is necessary in properly pro- portioning the joints in tie-beams, and the depth of notches at the foot of rafters. Following are values of the modulus of detrusive shearing in cases where the force acts perpendicular to the fibres : COEFFICIENTS OF DETRUSIVE SHEARING ACROSS THE GRAIN. BRITISH. METRIC. BRITISH. METRIC. Larch( Hackmatack) Oak 1,000 4,000 70 280 Red Pine 800 600 56 42 Spruce Pine Trautwine obtains by experiment the following values of the shearing resistance of American woods, where rupture is produced across the axis of the piece. 646 NON-METALLIC MATERIALS. RESISTANCE TO TRANSVERSE SHEARING. Ash Beech Birch Cedar, White. " C. Am. Cherry Chestnut Dogwood Ebony Gum Hemlock LBS PER 5Q. INCH. KG3. PER SQ. CM. 6,280 5.223 5,595 1,372 to 1,519 3.410 2.945 T.535 5,510 7,750 5,890 2.750 ,440 !366 '392 96 to 107 239 206 108 456 543 413 193 Hickory Locust Maple Oak, White. . . " Live.... Pine, White . . " Yellow. Poplar Spruce Walnut, Black Walnut, White LBS. PER SQ. INCH. 6,045 to 7,285 7.176 6,355 4,425 8,480 2,480 4,340 to 5,735 4,418 3,255 4,725 2,830 KGS, PER SQ. CM. 23 to 511 503 ,445 310 595 174 304 to 402 310 228 331 199 Fairbairn found that the resistance offered to forcing a ball three inches (7.62 centimetres) in diameter, through three- inch (7.62 centimetres) oak plank, was about the same as with quarter-inch (.63 centimetre) boiler plate, 17,000 pounds (7,727 kilogrammes). Rupture by Cross-breaking more frequently occurs with timber than any other kind of rupture, owing to the fact that it is more usually subjected to cross strains in situ- ations where it is generally applied. The relation between the stress and the character of the molecular change which it produces, has been made a subject of frequent mathematical investigation from the time of Ga- lileo, who seems to have been the first to attack the problem analytically. Such discussions have properly no place here, as the engineer will learn the theory of the subject from special treatises on strength of materials.* Where any beam is fixed at both ends, it is found in all actual cases that the formula gives it credit for more strength than it really has, and that it is more liable to break in the middle than at either end, although the analysis which deter- mines the formula indicates that this liability is the same at each of the three points. Barlow has therefore recommenHed for the special case that the formula read^ Rbd^ * 1 » * See Wood's Resistance of Materials ^ N. Y., J. Wiley & Sons, 1877, STRENGTH OF TIMBER. 647 as more nearly approaching the conditions of practice. The discrepancy probably arises from the fact, that in practice the beam is not perfectly " fixed " in the sense in which that word is used above. The following values of R, the modulus of rupture, in timber, have been determined by various authorities, and are given as close approximations for timber in good condi- tion. The units are pounds and inches, kilogrammes and centimetres. MODULI OF RUPTURE OF WOODS. BRITISH. METRIC Ash Beech , Birch, American Box Cedar, West Indian . . . Cherry Chestnut Ebony, West Indian. . . Elm Fir, New England " Riga " Norway ' ' American Spruce. . Hemlock Lancewood Larch, European " American 12,000 g.ooo 9,500 8,500 8,000 8,000 7,000 15,000 8,000 7,000 7,000 7,000 7,000 7,000 15,000 8,000 10,000 844 633 668 598 562 562 492 1,055 562 492 492 492 492 492 1,055 762 703 Lignum-Vitss Locust Mahogany, Spanish. . . " Honduras.. Maple Oak, Canadian ' ' English " European " Live " White Pine, Pitch " Red " Yellow Teak Walnut Willow Whitewood (Basswood) BRITISH. METRIC 12,000 12, COO 8,000 10,000 8,000 10,000 10,000 10,000 12,000 11,000 8,000 8,000 10,000 15,000 I2,OCO 7,000 8,000 844 844 562 703 562 703 700 703 844 773 562 562 703 1.055 844 492 562 From the records of about forty tests of California spruce and Oregon pine, made by the Author' at the Stevens Institute of Technology for the U. S. Geological Survey during the year 1880, the following results are taken : MODULUS OF RUPTURE (meAN). California Spruce. Oregon Pine .... BRITISH. 12,228 11,071 845 775 648 NON-METALLIC MATERIALS. Beams of the same material vary greatly in strength, and they sometimes break under one-fourth the load correspond- ing to their coefificients as above given, even when apparently sound. A large factor of safety is hence advisable. A solid cylinder varies in strength as the cube of its diameter. The formula for this case becomes, where fixed at one end and loaded at the other, 1.7 X 6 X / and if uniformly loaded this value P is doubled. Supported at the ends and loaded in the middle, /"becomes quadrupled ; supported at both ends and uniformly loaded, it is eight times as great. A beam supported at one end, fixed at the other, and loaded uniformly, has the same strength as the last case, as has also a beam fixed at both ends, and loaded in the middle. When fixed at both ends and uniformly loaded, the value of P is twelve times as great as in the first of the preceding cases. The latter statement of the relative strength of beams differently placed is correct for all solid beams. A wooden beam of triangular section, supported at both ends, is about one-sixth stronger with its base upward than with its base downward. The strongest beam of rectangular section that can be cut from a round log, has a breadth proportioned to its depth, as I is to V^, or nearly as 5 to 7. Such a beam is 10 per cent, stronger than the beam of square section that might be cut from the same log. The most resilient beam has its breadth and depth equal. Placing the beam with its annual layers in the plane in which the load acts increases its resist- ance in the proportion of 8 to 7 nearly. The following, in British measures, are the dimensions and safe distributed loads of sound pine beams, for' each inch of thickness, as used in ordinary work. In metric measures the loads are approximately metric ^^ tonnes" of 1,000 kilogrammes, for depths in centimetres as given, and per y^, decimetre width. STRENGTH OF TIMBER. 649 LOADS ON YELLOW PINE BEAMS. SAFE UNIFORMLY DISTRIBUTED LOADS IN TONS OF 2,000 LBS. FOR RECTANGU- LAR BEAMS ONE INCH IN THICKNESS. S 1 in U-) w 00 \o 10 ti- N e<^o « N M « r>. OM33 M M 'S'O o^miow 0^00 l-« NMOoMOf*^incn-ONOO^'^r-*MONoo-TMrtu-,«c r-. in en w vOoo cj C>Mvo« OM^iri^e^w NH do^ O^S M J^ ?; t^vO o O o in Iri iri S^j inr^inc+^coMwwMMiHMMMtHcfddodoodddddodd M c^O WO Ti-o wo c<^e. f^O O O tn in irj in in --^ ^ ■'t wO^l-e^^ciNMWMMWMHOOOOOOOOOOOodoodo in r^co N "ir r^o r>. r^ ^ Tt r-. CO e<>co w -^mw ^^^>.o^cnO ONa>M mo^nM cno M m ■^- in r>.o O r^O t^ O c^co e^^omMoo mmMcoo inc<^M o O^ r-^oo c> 0^ CO 00 ^ to t-i 0^ O^Q0 r-. r^o oo mminin-^-sj-^Tj-Tt" ^'cO M\neo«wMMWMMMOooooood ddddddddddd =1 coo r^C^O M Q O^coo^-^r^ inoo oo OO ininr--0 mo l^mco OQ'^^OOONinMOO too MO «oo in« O r^mcow o cor^mT+co OcomovO -^W M a^oo 1^ t^o vO uimminTfTi-'rr'^l-Ti-cocotOcoco owicoNNMMMMMOooooooodddd dddddddd S^SSS""* 2° O-^O -tOogom-^r^NO QwinOOcoM o OO OOOOcoOOmco -^O "^ \0 W O^O '^N Oooo mcoci m O CT'CO -3-MCOMO'^NO CMXl l>.I^OO »niOTr'f1-''i-'^'^COCOeOCOCOCOCOM W co-^wNMHMMOooooooooooddoddo'ddddd n- N moo o^r^«co M TTM o^ rfO co-^ooo mr»M m n a»QO l~^ r^oo 0^ M rr r* M coco m o^<:) r^ o to i^ co OO toQooo ^cow oco r^o m ■^ to to o^ ^ to r^ CO M cr-oo i^o -O ^nmTf-^^Tt-cocococoeococi N M « m m ci otoNMMMOoooooooooodoodddddddddd m M mo mco -^comc^ h cT'COCI mw m mo mcoo m»^ \rf^ oo m o ■^N M MOO Mmm-^Ti-'i-ONr>.ooo Hr^rf-Nw wco moo m m on ■^j-Mco MCO Tf-como^-^O r>.TtM o>t^o ■^ CO (N M Oco t-<.r^o m • • ■^ « -Tj- M 00 r^O U-l-^*^rtCOCOC^««WW««««MMMHMM • - ^NMMOooooooooooooddddd dddddd H COM -^-M M r*.o inco ON'^-M cor*.coO Ono^O w moo wo mo OOcomooooowr^'.mw O o^r^o m-rcocoMMMOOOOv* • • • inWOOOmT^eOCOClNMMMIHMMMMMMMMHIMWO • • • ■ ^^^666666666666666666666666 ion' 00 0^ -^ t^ o^oo r^ CO --d-co m ■^- ^to Onno Mf>iCOO*mM On coo r^ CO ^00 -^M ONr>.mTftOW m o O o^ Oioo oo f^ r- l^o r^co m'^cow««wMMwwMMMMOOOOOOOO Hoooooooooooooodddddddddd "*2 M mr»r-*NQO meow m o O^co f>. r>.o OvOmmmm-^'^ MineONPlMMMMWMOOOOOOOOOOOOO wooooooooooddddddddddddd IT) m ci CO in t:J- onoo O^z^ r^wco mw O^l-^mcOM oco r->. ci M m « CO r>-o omm'^h^TcototocococociN Oe0«MMM00000000000000000 66666666666666666666666 fX) O^xTi 0^\0 u-jMoo meOM o 0»r*.O ininrftoto r^ CO 0^\0 in-^t-^COtOMPICIPlNHMMMMMtHM O* in CO r*> ^ N a>oo i~^o ommm-^-^Tj-^co 00000000000000000000 66666666666666666666 1/ S'S } ^ M W CO ^ »no t^oo 0> M M co^mo r>.oo Q« O m « corhmO t-^cO CT* o MMMMMMHMMMWMC*NC<«OIClC^D)CO These loads are about one-eighth the breaking load. Beams supported. Rule. — To find the safe uniformly distributed load for yellow pine beams, multiply the number given in the table by the thickness of the beam in inches, or take 0.4 the given number for load per centimetre of thickness. For beams of other wood, multiply by the following numbers : White Oak. White Pine. Hemlock. White Cedar. Spruce. i.« -95 .95 .65 .85 650 NON-METALLIC MATERIALS. The Stiffness of Beams varies as their breadths, and as the cube of their depths. As the strength only varies as the square of the depth, it follows that large beams will be found to bend less, before breaking, than will small beams. From this fact, also, it happens that it becomes necessary, with flexible wood of comparatively small scantling, to proportion them to bear a given load with a certain limited deflection, rather than with reference to their absolute strength. In using any material the necessity frequently arises for employing formulas, ex- pressing stiffness rather than strength, in order to secure the requisite rigidity of parts. The stiffest beam which can be cut from a round log has its breadth and depth proportioned as i is to a/^, or nearly as I to 1.732 = -577+ to i. Formula for Flexure. — The following formula rep- resents the flexure of beams of rectangular section, lying on two supports, and loaded in the middle : Let D = the deflection, in inches or centimetres, L = the length between bearings, in feet or metres, P = the weight, in pounds or kilogrammes, d and d = the breadth and depth, in inches or centimetres : 2? = .^^^' C is a constant determined by experiment for each mate- rial. On page 99 are its values as given by the best authori- ties. In metric measure C^ = C X 8000, nearly. It is generally assumed that timber should not be loaded to a deflection greater than y^u* its length. In such cases, Soi? = Z, and, substituting this value, we get from (23), British measures : C=J^^, and P= ''' Where beams are fixed at one end and loaded at the other, they deflect 16 times as much as when supported at STRENGTH OF TIMBER. 651 both ends, and loaded in the middle.* Hence, for this case, the values of C above given must be increased in this pro- portion ; the formula then becoming Z? = \6PDC COEFFICIENTS OF DEFLECTION.' Ash Beech Birch Cedar Cherry Chestnut ' ' Spanish Elm Fir, Am. Spruce . " Hemlock . .. Larch BRITISH. METRIC 0.00030 2-5 . 00030 2.5 0.00030 2.5 0.00030 2.5 0.00040 3-0 0.00025 2.0 0.00050 3.5 0.00030 2.5 0.00025 2.0 0.00025 2.0 . 00030 2.5 Maple Mahogany, Spanish " Honduras Oak, minimum.. . . " maximum. . . " mean value. . Pine, White " Pitch Teak Walnut Willow..' 0.00040 0.00030 0.00025 0.00025 0.00050 0.00040 0.00025 0.00030 0.00030 0.00025 0.00060 3-0 2.5 2.0 2.0 4.0 30 2.0 3.5 2.5 2.0 5-0 If, in this latter case, the load should be uniformly dis- tributed, the formula becomes : D = 0.62s CPP (26); D = \2PPC bd^ The formulas just given for the deflection of beams are those most generally used. A less simple, but possibly more accurate formula has been proposed by Prof. W. A. Norton, and is well supported by the experiments from which he de- duces it. A = the deflection of a piece supported at the ends, loaded at the middle (/, b, d are in inches): P/' PI A^- + C—. 4Ebd^ bd C is given at 0.0,000,094; E = 1,427,965 pounds for pine. Trautwine gives 24 for beams as fixed in practice. 6s 2 NON-METALLIC MATERIALS. Beams should be made as deep as possible, pro- vided they are not made of such depth as to be liable to over- turn and break sideways. A formula to determine the proper proportions of section is the following, which is given for use in general practice : b = 0.6 — — The stiffest rectangular beam that can be cut from any cylindrical log has its thickness equal to one-half the diam- eter of the log. Beams of square section are equally stiff in whatever direction they may be bent. A beam fixed at both ends has twice as great stiffness as one merely supported. In framing, therefore, the joists should be made of as great length as possible, in order that they may extend over the greatest number of supports ; and they should invariably be notched over the latter, where possible. Working Loads for Floor-beams. — C. J. H. Wood- bury, of Boston, Mass.,* gives the following formulas, deduced from experiments on beams used in mill floors. The meas- ures, as will be seen, are all British : Let h = depth of beam, inches. b = breadth of beam, inches. d = deflection, inches. / = span, feet. J = width of load, feet. w = distributed load per square foot of floor, includ- ing its own weight, lbs. u = weight of floor per square foot, lbs. w' = distributed load upon square foot of floor, not including weight of floor, in lbs. W = concentrated load on floor, lbs. , R = modulus of rupture, lbs. per square inch. • See a paper read before the American Society of Mechanical Engineers (1S81), a-ai Fire Protection 0/ Mills, by C, J. H. Woodbury; N.Y., J. Wiley & Sons, 18S2. STRENGTH OF TIMBER. 653 E — modulus of elasticity, lbs. per square inch. / = factor of safety, in units. Assuming the following data : SOUTHERN PINE. SPRUCE. E = 2,000,000, 1,200,000; R = 12,960, 10,080. That in storehouse floors, / = 6, for fixed loads ; 2/ = 12, for live loads. The limit of d in mill floors, .075 inch per 8 feet, say T^T77 span. For 25 feet beams, same curvature = about .75 inch = -^\^ span. We find : In a beam loaded at centre and supported at ends, „_ i8PF7+9w/^ _ 432 WP ~ bh^ ' bh^d Strength of beams. Load uniformly distributed ; «/ = ze = gfsP' V gwfs ' V -Kb Strength of floor plank Load in bulk (as grain) : Strength of floor plank. Load in case or bale : 12// ' V3/(4«'' + «)' h= ^ l lfl\A->J^ + u) aR 654 NON-METALLIC MATERIALS. Then in a storehouse, with floors of spruce plank ; beams of Southern yellow pine, 8 feet between centres, and height of beam = twice breadth. ' The strength of beams : Strength of floor plank : w 4 ' V 4k/ + « ' V 35 Deflection of beams : Deflection of floor plank one bay in length — a form of con- struction not advised : 45w/* 2Edk^ , ^[a^wI^ '^^w' -^^-liT^' ^=y^Ed- Deflection of floor plank Hvo bays in length : 28w/* _ 2,Edk^ _ VaSw/^ 3^/««' "'" 28/^ '■ ^-y 3£^ Torsional Strains rarely occur with timber, and but little has been definitely known, until recently, of the value of the different woods to resist this kind of stress. STRENGTH OF TIMBER. 655 The following values were determined in British measure by the Author, who used a machine designed for the purpose, which recorded its own action by pencilling a curve whose abscissas represented twisting moments, and whose ordinates represented the corresponding values of the angle of torsion.* COEFFICIENTS OF TORSION. Ash Cedar, Red. . . Chestnut Hickory Locust Mahogany Oak Pine, Spruce. . , " Yellow.. " White.. . Walnut, Black, r- 410,000 0.001,055 890,000 0.000,701 355,000 0.001,783 910,000 0.000,695 1,225,000 0.000,517 660,000 0.000,960 570,000 0.001,111 211,000 0.003,000 495,000 0.001,280 220,000 0.002,880 582,000 0.001,090 When Rupture by Torsion occurs, the outer layers of fibres will be broken first. Up to the limit of elasticity of these fibres, the strain upon any one fibre will vary approxi- mately as its distance from the axis of torsion. Where C = coefficient of rupture, d = the diameter, in inches or centimetres, P = the twisting force, in pounds or kilogrammes, / = the lever arm of P. We shall have for cylindrical pieces, Cnr' '/2Pi P= 2/ r — Ctt ^ youmal Franklin Institute for 1873, p. 254. 6s6 NON-METALLIC MATERIALS. The following values for A have been determined by the Author by experiments with recording apparatus, and for the simplified equations d zl PI and P = Ad^ COEFFICIENTS OF TORSION, Ash Cedar, Red Chestnut Hickory Locust Oak Mahogany, Spanish. 328 244 296 644 ,648 ,424 •524 41.0 30-5 37.0 80.5 81.0 53-0 65-5 Black Spruce. . . Heart Sap Pine, Spruce. . . " White... " Yellow... Walnut, Black. .216 .264 .316 ,185 .412 27.0 33-0 39-5 23.1 51-5 Cauchy makes C about four-fifths the value of the coeffi- cient of transverse rupture ; but this relation must probably be variable. RELATIVE TORSIONAL RESILIENCE. NAME. VALUE. NAME. VALUE. White Pine 1.00 1.50 1. 61 1.65 2.25 2.40 Yellow Pine 387 3-95 5.80 6 60 Spruce Black Walnut Red Cedar Spanish Mahogany Ash Oak . . 6.90 Chestnut STRENGTH OF TIMBER. 657 Extended and delicate researches upon the laws of resist- ance to torsion were made by M. G. Wertheim* His most important conclusions were the following: (l.) The torsion angle consists of two parts, one of which is temporary, the other permanent. The latter increases con- tinually, but not regularly. (2.) The temporary part increases more rapidly than the applied moment, up to the limit of elastic resistance, and, in some cases, beyond. (3.) The temporary part does not precisely vary with the length twisted. The shorter the piece, the greater this dis- proportionality. (4.) Torsion causes a diminution of volume in homoge- neous substances, the density increasing from the centre to the circumference. The diminution is proportional to the product of the length of the piece, and the square of the angle of torsion. These conclusions are deduced from experii;nents upon small angles of torsion. Effect of Prolonged Stress upon the Strength and Elasticity of Pine Timber. f — Experiments made by Mr. Herman Haupt showed that timber may be injured by a prolonged stress far within that which leaves the material un- injured when the test is made in the usual way and occupies a few minutes only.ij: An extended series of experiments made intermittently in the Mechanical Laboratory of the Stevens Institute of Technology, Department of Engineer- ing, included an examination of this subject, and the result has confirmed Haupt's earlier work, and has given a tolerably good idea of the effect of prolonged stress in modifying the primitive relation of stress and strain where the wood is good Southern yellow pine. A selected yellow pine plank was obtained for test, the * Annales de Chemie et de Physique, vol. xxiii., 1st series, vol. 1., 3d series. f From the Proceedings of the American Association for the Advancement of Science, vol. xxx., Cincinnati Meeting, August, i83i. R. H. Thurston. % Bridge Construction, N, Y., 1871, p. 61. 42 6s 8 NON-METALLIC MATERIALS. origin and history of which could be traced and the extent of seasoning known beyond all doubt. The Author, from an examination of the results, concluded that a load of 60 per cent, the maximum given by the usual form of test, is for such pieces unsafe, although it would seem that a slightly smaller load might have been carried indefin- itely, or until decay should weaken the timber. A factor of safety of two would possibly have permitted indefinite endur- ance under static load. Taking the probable breaking load under unintermitted stress as 50 per cent, that sustained as a maximum under usual tests, and then applying a factor of safety of two, we obtain a safe factor, based on the ordinary test, 4. Conclusions. — In brief, the conclusions to be drawn from the research referred to, are evidently that small sec- tions of yellow pine timber yield steadily over long periods of time under loads exceeding 60 per cent, the maximum ob- tained by ordinary tests of their transverse strength, and finally break after a period, which with the lighter loads may exceed a year ; that deflections half the maximum reached under test may be unsafe for long periods of time, and that a factor of safety of at least 4 should be used for permanent static loads when the character of the material is known. The author would, in the light of what is now known, always use a factor of safety of at least 5 under absolutely static loads, and when the uncertainties of ordinary practice as to the exact character of material, and especially where shake and the impact of live loads were to be considered, would make the factor not less than 8, and for much of our ordinary work 10. Conclusions relative to the application of Wood in Engineering Construction. — From what has been already learned, and by comparison with that which is hereafter stated concerning other materials used in engineering, some conclusions may now be deduced, relative to the value of wood to resist the various kinds of stress which the engineer is compelled to meet in his constructions, and for specicd applications. STRENGTH OF TIMBER. 659 For pattern-making a light wood is generally desired, ca- pable of seasoning without checking, and of being easily worked. For large patterns, white pine or cherry is generally used ; and for small patterns, where weight is less objection- able, and where strength, smoothness of grain, and firmness of texture are more essential, mahogany is taken. For ex- tremely small patterns, boxwood and ebony are much used. For turned work, alder, beech, birch, and white pine are used when an easily worked wood is desired ; for a tough and fine-grained, clean and smooth-working material, holly is unexcelled ; it requires, however, great care in seasoning. Apple, maple, pear, locust, boxwood, ebony, oak, and elm are all valuable for lathe work. Black walnut, mahogany, and rosewood are used for orna- mental purposes, and work well in the lathe as well as at the bench. For ordinary Joiner's work, the pines are principally used, and for finer work, maple, black walnut, and mahogany are in request. Rosewood and some other tropical woods are generally used only for expensive work, such as is never necessary for the engineer to construct. Where extreme lightness is desired, white pine is gener- ally used; for purposes requiring a wood both light and strong, yellow pine is most called for. Woods of Commercial Value in connection with properties usual or peculiar as named at their heads respect- ively, are as follows : Elasticity. — Ash, hickory, hazel, lancewood, chestnut (small), yew, snakewood. Elasticity and Toughness, — Oak, beech, elm, lignum-vitse, walnut, hornbeam. Even Grain (for carving and engraving). — Pear, pine, box, lime-tree. Durability (in dry works). — Cedar, oak, poplar, yellow pine, chestnut. Wet Construction (as piles, foundations, flumes, etc.). — Elm, alder, beech, oak, plane tree, white cedar. 66o NON-METALLIC MATERIALS. Ship-building. — Cedar, pines (deals), firs, larches, elms^ oaks, locust, teak. House-building. — Pines, oak, white wood, chestnut, ash, spruce, sycamore. Furniture. — Common : beech, birch, cedars, cherry, pines, white wood. Best furniture : amboyna, black ebony, mahog- any, cherry, maple, walnut, oak, rosewood, satinwood, sandal- wood, chestnut, cedar, tulipwood, zebra wood, ebony. Machinery and Mill-work. — Frames : ash, beech, birch, pine, elm, oak. Rollers, etc. : box, lignum-vitae, mahogany. Teeth of wheels : crab-tree, hornbeam, locust, hickory, and maple.. Foundry patterns : alder, pine, mahogany, cherry. Of the above-named varieties, those that chiefly enter into commerce in this country are oak, hickory, ash, elm, pines, black walnut, maple, cherry, butternut, white wood, etc. No approximate figures even can be given of the amount annually used in this country. In parts requiring great strength and toughness, white oak, hickory, and locust are used. The first named is used for water-wheel shafts, for places where lignum-vitse cannot be used, for subaqueous bearings — as for steps for turbine wheels — and for any position in which it will be kept constantly wet. Hickory and white oak are particularly well adapted fol teeth of mortice-gear wheels, as are also maple and beech • and the former for any dry situations in which their great strength and toughness are likely to be found requisite. Locust is selected where strength and toughness are desired, and where large pieces are not necessary. The last five woods, and maple and the pines, are those most frequently used by the mechanical engineer. In the drawing office, boxwood, holly, and red cedar are used for the blades of T'-squares, and for rulers, and scales. Some of the ornamental woods are used for the heads of /"-squares. Pearwood is found to be well adapted for model work, and maple for general light work requiring a good surface. The latter makes good teeth for mortice wheels which are STRENGTH OF TIMBER. 66l not subjected to very heavy stress. Sour applewood is even better for the latter purpose, and is much sought by wheel- wrights for gearing used in dry situations. In presence of moisture, white oak is the best of all woods for this work. When wood is required in carpentry, for iloor-joists and raft- ers, the stiff woods are selected ; for carriage-shafts and poles, builders select the toughest woods, while for tie-beanis, those woods having greatest lateral cohesion and tensile strength are taken. In building railroad cars, where lightness and strength should be well combined, pine is preferred above all other woods. ' Tough and cross-grained woods are most difficult, and therefore most expensive to work ; the most brittle woods are usually easily worked, the fine-grained woods take the smooth- est polish, and the surface is best preserved by the harder varieties. The Figure of the Markings of Wood depends more upon the particular directions of the fibres than upon any dif- ference of color. If a tree were formed of cylindrical layers, the horizontal section would exhibit concentric circles, the vertical section giving parallel straight lines ; and the oblique section, ellipses. But few trees are to be found exactly straight, and, therefore, although the three sections have a general tendency to exhibit the figures described, every bend and twist in the tree disturbs the regularity of its fibre, and adds to the variety of grain and ornamentation of the wood. A perpendicular cut through the heart of the tree exhibits the most diversified surface, because in it occurs the most profuse mixture of the fibre, the oldest and newest being presented in the same plank. Curls are formed by the confused filling in of the space between the forks of the branches. The figures thus produced cause a log to be valuable in proportion to the number of curls it contains. Figures are also produced in the following manner. The germs of the primary branches are set at an early period of the growth of the parent stem, and thus give rise to knots. But many fail to penetrate to the exterior, and are covered 662 NON-METALLIC MATERIALS. over by later annual rings. When the germ forces its way to the surface, the fibres of the trunk bend aside when they en- counter the knot, and in the soft woods do not unite with it. The hardness of knots is due to the close grouping of the fibres, and to their compression by the surrounding wood, which itself is allowed to expand by the yielding of the bark. The same operation goes on in the roots of trees, and fur- niture veneers are often obtained from them. The bird's-eye maple has points or spines on the inside of the bark, which penetrate the wood and make irregular indentations. These cause that peculiar appearance from which the wood takes its name. In woods, the figure of which resembles the ripple-marks of the sea on fine sand, such as satinwood, sycamore, mahog- any, and ash, the figure is produced by the serpentine form of the grain. The fibres of all such pieces are wavy in planes at right angles to that on which the ripple is observed, if not on both, those parts of the wood which receive the light being brightest. Woods having silver grain, or marked medullary rays, exhibit a dappled appearance similar to that produced on silk by threads crossing one another. English oak, Riga and Dutch wainscot logs, Austrian wainscot, etc., have this pecu- liarity. In the oak plank the principal lines are the edges of the annual rings, which show parallel lines. Damask pencillings, or broad, curly veins and stripes, are caused by groups of the medullary rays which undulate from the surface to the centre of the tree, and creep in betwixt the longitudinal fibres. Were the fibres of trees arranged with the uniformity and exactitude of a piece of plain cloth, they would show an even, uninterrupted color ; but being arranged in irregular, curved lines, every section partly re- moves some and exposes others, thus producing a grgat variety of figure. Coloring^ of Woods. — Some woods are nearly uni- form in color, and some have several shades of the same hue or of several colors. In the transverse section of such woods the tree seems to have clothed itself with different , coats of STRENGTH OF TIMBER. 663 various colors. Tulipwood, kingwood, zebrawood, and rose- wood illustrate this case. In ordinary planks these markings are drawn out into stripes, bands, and patches, or wavy fig- ures of beautiful or grotesque form. Woods variegated both in grain and color are generally employed for objects with smooth surfaces, as in cabinet- work. Such are Amboyna, kingwood, mahogany, maple, partridge, rose, satin, snake, tulip, and zebrawood. Speci- mens of marquetry often beautifully illustrate the use of such wood for the purpose of ornamentation. The same style of work in mouldings has an inferior effect. The colors of " fancy woods " are not usually liable to fade by exposure to light, tulipwood being one exception ; but age darkens them and mellows the general effect. Only the whitest of varnishes should be laid over them, for the natural tint v/ill easily be spoiled. The rich greenish brown of walnut is esteemed for piano-forte cases, for which work, however, rosewood has hftherto been more generally used. The rich, deep orange of Spanish mahogany makes beautiful tables and counter-tops, and the size of the timber adapts it to either use. Honduras mahogany, of a brownish tint, is used for all kinds of superior cabinet-work, while oak is prin- cipally employed where durability is a necessity. Pitch-pine is pleasing in color and figure. Rosewood has very rich tints, and is much used. Carpentry is the art of construction in wood, and properly includes several divisions as joinery, cabinet-making, pattern-making, and ship-construction. The engineer will find special treatises on each subject which will give full in- formation relating to trade methods. In this place only the simplest principles involved in all wood-working can be given. The fashioning of wood is often done by machinery, and hand labor is only employed in fitting and in forming special shapes or in making constructions which are not called for in such quantities as to justify the building of special machinery for their manufacture. In constructions of wood the parts are usually straight and simply formed pieces, and stresses are almost invariably 664 NON-METALLIC MATERIALS. taken either as transverse loads or by compression ; wood is unfitted for sustaining tensile forces, as it is extremely diffi- cult to obtain such a secure hold upon the material as to permit the tenacity of the piece to be fully brought into play before fracture occurs by detrusion. In uniting timber it is advisable to be exceedingly careful to reduce the loss of section by cutting for the joints and fastening to a minimum ; to take advantage of peculiarities in the " lay " of the grain wherever possible ; to make sur- faces exposed to pressure of such shape, and to place them in such a position that the lines of pressure shall be normal to them ; to give ample area of bearing surface to insure safety against injury by the maximum stress anticipated ; to fit abutting parts perfectly and unite them securely, and to insure, wherever possible, equal strength in the pieces and their fastenings, except in places where it is found advisable to make some one point somewhat weaker than the others in order that, in case of accidental rupture, the most costly piece shall be saved at the expense of one that can be better spared. Precaution is necessary to prevent the use of such fastenings, or so locating them that they shall either cut through the wood or crush their bearings. The joints should be simple in form and carefully designed for each case. Joints receiving compressive stresses are usually made by cutting squarely across the line of pressure ; but those made to resist either transverse or tensile forces are less simple. Scarfing is th*e most usual method, and practiced in several ways, as is seen in the accompanying sketches : * Fig. 150. Fig- i5i- In the examples, Figs. 150 and 151, the pieces to be joined * See Appletons' Cyclopiedia of Applied Mechanics. STRENGTH OF TIMBER. 665 are cut diagonally at the abutting ends and a stepped surface fornied on each side in such a manner that, being fitted and bolted together as shown, the joint becomes nearly as strong as the soUd wood ; to obtain still greater strength of joint the upper and lower surfaces of the scarf are sometimes covered by a pair of " fish-plates " of boiler-iron extending some distance each way beyond the joint and bolted on by through-bolts having their heads bearing on one plate and their nuts on the other. Fig. 152. Fig. 153. In other cases joints are made as in Figs. 152, 153, above, and the lap observed in the second sketch is brought to a bearing by a key of hard wood driven into place after the parts are fitted together, as is also seen in Fig. 150. Pieces brought together at right angles are often dovetailed as in Fig. rS4. For other cases, tenons, or the end of one Fig. 154. piece fitting into mortices, cut through the bearing surface as in the sketches below, are adopted, and when the tenon enters at such an inclination as 3 i Fig. 155- Fig. 156. Fig. 157. to give rise to danger of slipping under the load or splitting out, the use of a bolt or a strap, as shown, will give security. King-posts are united with diagonal, as with rafters 666 NON-METALLIC MATERIALS. (Fig. 158), with braces and tie-beams (Fig. 159), as here shown, and straps or bolts are often added for greater safety. The vertical and one diagonal may be united as in Fig. 160. Fig. 158. Fig. 159. Fig. 160. The bridge-work timber verticals and diagonals often rest on cast-iron shoes which have a broad bearing on the chords, and thus give security against crushing ; this is also practised in the construction of heavy wooden roof-frames. Timbers crossing each other are often halved together or are lapped when simply abutting. The length of scarf or other Joints must be such that the total resistance to detrusion shall be at least equal to the resistance to transverse rupture ; the hard woods, as oak, hickory, ash, elm, locust, are usually scarfed to a length equal to six times their depth ; pine and other soft woods are given twice as long a scarf. Wooden girders should have good bearings at the ends, and should rest on the solid wall and have a bearing on stone or iron, with ample space for circulation of air. Pins used in securing parts of timber-work are often made of wood and are called " treenails "; their diameter is usu- ally about one-tfiird the thickness of the planks which they unite ; they are best made of oak or locust, and may be taken as having a strength of 3,000 to 4,000 pounds per square inch (210 to 280 kilogrammes per square centimetre) of cross- section. ■ Nails are used in small and spikes in large sections of wood-work, and are given a length of from two to three times the thickness of the thinner of the two parts united by them. The following table from Sevan's experiments shows the rela- tive value of several standard sizes :* * Tredgold : Carpentry. STRENGTH OF TIMBER. btj HOLDING POWER OF NAILS IN PINE. KIND. LENGTH. NO. PER DEPTH DRIVEN. RESIST. TO DRAWING. In. Cm. Kilogs. Lbs. In. Cm. Kilogs. Lbs. Brads 0-44 1. 12 4,560 10,000 0.4 1. 00 22 48.4 Brads 0.53 1-35 3,200 7,040 44 1. 12 37 81.4 Brads 1.25 3.18 618 1,360 0.50 1.27 58 127.6 Fivepenny. 2.00 5-08 139 306 1.50 4.81 320 740.0 Sixpenny. . 2.50 b-35 73 160 1. 00 4-54 187 411. 4 Sixpenny . . 2.50 6-35 73 160 1.50 4.81 327 719.4 Sixpenny.. 2.50 6-35 73 160 2.00 5.08 530 1,166.0 The weight of nails is often roughly taken as about W= 0.4/^ when Wis weight in pounds per 1,000, and / their length in inches; in metric measure W^=o.o6/' for kilo- grammes and centimetres. The resistance to a drawing force varies as roughly as dl, when d is the depth to which the nail is driven. Bevan found the following to be the resistance of a sixpenny nail driven i inch (2.54 centimetres) into different woods : Pine, across grain Oak Elm 187 507 327 411. 4 1,115-4 829.4 Pine, with grain . Elm, " " . 87 257 191-4 565.4 The resistance to driving by steady pressure is, in soft woods, 20 per cent, greater than resistance to extraction. A sixpenny nail forced into Christiania " deal " offered resist- ance as below: 0.25 0.64 0.5 1-77 I. 2-54 1-5 3-81 Depth, -j Cejjtin,etres 5.08 (Lbs 24 11 76 34-5 235 107 400 171 610 PRESSURE, -j j^il^g^ 276 668 NON-METALLIC MATERIALS. Wellington* found the resistance of railroad spikes driven into various woods to be : WOOD, Beech Ash Elm Maple DRIVING IN. PULLING OUT. Lbs. KgS. I.bs. KgS. '',743 3,074 5,978 2,717 S,953 2,750 4,560 ^;?i^ 4,606 2,094 3,690 3,843 1,746 3,111 1,323 Oak, green " seasoned. . Chestnut Hemlock DRIVING IN. Lbs. 5,820 6vt33 3.^91 2,zo5 KgS. 452,6 2,924 1,586 1,323 PULLING OUT. Lbs. 6,523 4,281 3,260 1,996 KgS. 2,965 1,946 1,936 907 It was found that elm and ash will hold a spike about two- thirds as well as oak or beech, and a third better than chest- nut ; soft maple and sycamore are four-fifths as effective as chestnut, two-fifths as good as oak and beech, and a half bet- ter than hemlock. Wood-screws are used wherever the parts are to be again separated ; where the stresses are likely to be greater than can be safely resisted by nails ; where the pieces joined are liable to be split or otherwise injured by the use of nails, or where nicety of fitting is important. Their resistance is nearly as the square of their diameters, if made of such length that their full strength may be utilized ; shorter screws, or any screws that pull out without breaking, resist merely as the area fractured, i. e., about as the product of the length of screw holding in the wood and diameter outside of thread. Bevan pulled screws 0.22 inch (0.5 centimetre) in diameter over the thread, having twelve threads per inch (5 to the centimetre), from a depth of a half inch (1.27 centimetre), thus : Beech USotoggolbs. / 210 ' 450 kgs. Ash \ 7f|^^- j 360 kgs. Oak \ 76olbs. ( 345 kgs. Mahogany Elm Sycamore . 770 lbs. 350 kgs. 665 lbs. 300 kgs. 830 lbs. 372 kgs. * Railroad Cazelte, No. 51, p. 668. STRENGTH OF TIMBER. 669 HOLDING POWER OF SCREWS. LENGTH. MUMBEE 2 a 1 2 a ia .a •a f 8 11 a* H^ 1 0^ "3 1, .a H s B u H a 1 i'°3 Sun lis U H H Q i 2886 2424 Fine grain dry white ash. Coarse grain dry white ash. 3.06 2.57 3 iK 20 3 3-2 6.4 ZZ02 2V 2879 Blacic walnut. 3-°5 aM iX i5 9 3.6 5.t 1217 2 i8s7 Coarse grain dry white ash. 1-97 =>4 i;!i H 10 4.0 S.t 844 2 1811 " H ii 1.92 2 i3i 20 3 3.2 3.B 824 >« 11586 i.b. 2 i« 16 9 3.6 1.8 721 I« It'll (( (t Ik 1.73 2 i^ 14 10 4.0 3-a 733 I>i 1580 1.67 2 i« 12 12 4.8 3.8 709 ■^■A IS49 (i (I 11 1.64 IX % 16 9 3.6 2-5 519 I 1142 ** '* *^ 1.2E 1^ % 14 10 4.0 2.5 '538 X ii8s 11 li .1 1.25 I>4 y. 10 13 5.1 2.5 438 I 943 I. 00 When bolts and nuts are used, the wood should be pro- tected by giving their heads and nuts a broad bearing on washers, i^ to 2 times the diameter of the head for hard and soft woods respectively. Iron fastenings should be avoided where the wood has an acid sap, as do some oaks. The sap of teak and of the pines protect iron fastenings. Where used they should be protected by paint, oil, or coal-tar, or by galvanizing. Glues. — Common Glue. — The absolute strength of a well-glued joint is given as : HOLDING POWER OF GLUE. POUNDS PER SQ. IN. KILOS. PER SQ. CM. Beech Across the grain, end to end, 2,133 1,436 1,735 1,493 1,42a With the grain. 1,095 1,124 568 341 896 Across the grain, I49-3t 100.52 121.45 104.54 99.51 With the grain. 76.65 Elm 78.68 Oak 39.76 White wood. .. . 23.87 62.72 fi/Q NON-METALLIC MATERIALS. It is customary to use from one-sixth to one-tenth of the above values to calculate the resistance which surfaces joined with glue can permanently sustain with safety. A little pow- dered chalk strengthens glue. Marine Glue. — India-rubber, i part ; coal-tar naphtha, 8 to 12 parts; shellac, 15 to 20 parts ; melted together. Use hot. Glue dissolved in skimmed milk will resist the action of moisture ; also glue softened with boiled oil or resin, and one- fourth its weight of iron oxide added. Water-proof Glue. — Boil eight parts of common glue with about thirty parts of water, until a strong solution is obtained ; add four and a half parts of boiled linseed-oil, and let the mixture boil two or three minutes, stirring it constantly. Preservation of Timber. — The causes of decay in timber have already been stated (p. 605), and the process of decay has been described. As has been seen, timber should be protected against the deleterious effects of moisture and oxidation, and the attacks of insects. Timber lasts longest either in perfectly dry and well-ventilated places, or where it is kept constantly im- mersed in water. The problem of preserving timber from decay is fully stated when it is said that the object to be attained is the prevention of oxidation. Timber which has been thoroughly seasoned by the methods already described, and which is perfectly dry, may be preserved by external applications. Under other circum- stances, internal^application of various solutions must be re- sorted to. Paints and Varnishes are used for the protection and preservation of timber by external treatment. They form a coating upon the surface, which resists the wearing action of the weather, and prevents the entrance into the pores of the wood of either moisture or corroding gases. , Should the wood not have been previously well seasoned, however, paint only hastens decay by confining the moisture and hastening the fermentation of the putrescible matter re- maining in the wood. The following are among the best of this class of preserva- STRENGTH OF TIMBER. 67 1 tive compositions ; many of them are recommended in the U. S. Army Ordnance Manual. The proportions are given for 100 parts by weight of pre- pared colors, when not otherwise designated. One gallon (3.79 litres) of linseed oil weighs. . 7.5 pounds 3.41 kgms. " " " spirits of turpentine . . 7.25 " 3.3 " " " " Japan varnish 7. " 3.18 " " " " sperm oil 7.12 " 3.23 " " " " neat's-foot oil 7.63 " 345 " BOILED OIL. Raw linseed oil 103. Copperas 3.1S Litharge 6.3 Suspend the copperas and litharge in a cloth bag in the middle of the kettle of oil. Boil 4^ hours with a slow steady fire. DRYER OR DRYING. Copperas and litharge from the boiled oil 60 Spirits of turpentine 56 Boiled oil 2 The mixture from the boiled oil to be ground and thoroughly mixed with the turpentine and oil. PUTTY. (for filling CRACKS.) Spanish whiting, ground 81.6 Boiled oil 20.4 Make into a stiff paste. If not intended for immediate use, raw oil should be used ; putty made with boiled oil hardens quickly. Also, mix finely sifted oak or other sawdust with linseed oil which has been boiled until it has become glutinous. WHITE PAINT. For inside work. Fop outside work. White lead in oil 80.0 80. Boiledoil 14.5 9- Raw oil 0.0 9. Spirits turpentine 8.0 4. Grind the lead in the oil, then add the spirits of turpentine. 6/2 NON-METALLIC MATERIALS. For woodwork use i pound to the square yard for three coats (2.75 kilogrammes per square metre). LEAD COLOR. White lead in oil 75,0 Lampblack I.o Boiled linseed oil '. 23.0 Litharge 0.5 Japan varnish 0.5 Spirits turpentine 2.5 Grind the lampblack and litharge separately in oil, then stir into the white lead and oil. Turpentine and varnish are added as the paint is required for use, or when packed in kegs for transportation. BLACK PAINT. Boiled linseed oil 73, Lampblack 28. Litharge r. Japan varnish I. Spirits turpentine i. Grind the lampblack in oil ; mix it with the other oil, then grind the litharge in oil and add it, stirring well. The varnish and turpentine are added last. This paint can be used for iron work. GRAY OR STONE COLOR, FOR BUILDINGS. f ist Coat. 2d Coat. White lead in oil 78.0 100. Boiled oil g.5 20. Raw oil 9.5 20. Spirits turpentine 3.0 o. Turkey umber o. 5 O. Lampblack 0.25 0.25 Yellow ochre 0.00 3. Mix like the lead color. A square yard of new brickwork requires, for two coats, ij^lbs. ; for three coats ij^ lbs. In metric measures, one square metre requires for two coats .61 kilogrammes ; for three coats .81 kilogrammes. STRENGTH OF TIMBER. 673 CREAM COLOR, FOR BUILDINGS. ist Coat. 2d Coat. White lead, in oil. 66.66 70.00 French yellow 3.33 3.33 Japan varnish 1.33 1.33 Raw oil 28. 00 24. 5 Spirits turpentine 2.25 2.25 A square yard of new brickwork requires for first coat ^ ; for second ^ pounds ; in metric measures, one square metre requires for first coat .4 ; for second .2 kilogrammes. BLACK STAIN, FOR WOOD. Copperas I lb. .67 kgm. Nutgalls I " .67 " Sal ammoniac 25 " .17 " Vinegar 1 gall. 3.79 litres. Stir occasionally, and it becomes ready for use in a few hours. Clean and smooth the surface, filling the cracks with black putty, allowing it to harden. Apply the stain two or three times, then leave it a day or two to dry ; finally rub with boiled oil until polished. JAPAN VARNISH. Litharge 4 lbs. i. 8 kgms. Boiled oil 88 " 40.0 " Spirits turpentine 2 " .9 " Redlead 6 " 2.7 " Umber I " .45 " Gum shellac 8 " 3.6 " Sugar of lead 2 " .9 " White vitriol i " -45 " Boil over a slow charcoal fire five hours, mixing all the ingredients except the turpentine and a small portion of the oil ; the latter is added as required to check ebullition. The mixture must be continually stirred with a wooden spatula, and great care is required to prevent it taking fire. The turpentine is added when the varnish is nearly cool, and should be well stirred in. The varnish must be put in close cans and kept tightly corked. 43 6/4 NON-METALLIC MATERIALS. Japan varnish may be purchased. Paraffin Paint. — Mix together good asphalt and paraffin in equal parts, melt, and stir well together. Add a small quantity of finely ground caustic lime, constantly stirring. Apply with a large brush. When this first coat has cooled, put on another coating of pure melted paraffin applied quickly and evenly. Brown mineral (iron oxide) paints, as sold ready for mix- ing with oil, a paint of " red lead " and oil, or of " zinc v/hite," are all used extensively on iron and for outside work. Wood work exposed to the weather is repainted, in our climate, at intervals of four or five years. The woodwork supporting the floors of bridges, and timber in damp situations as in wheel-pits, is sometimes coated with coal-tar prepared for use by boiling, and by the addition of a small quantity of chalk to give it body. This is also an excellent application for water-pipes, for smoke- stacks of iron, and other out-of-door ironwork. As a pre- servative against decay it is also excellent on woodwork, but is often seriously objectionable because of its inflammability. Boiling linseed oil, pitch and vegetable tar, applied hot, are not unfrequently used as external applications, and ai^e found to be very effective preservatives. The soot from bituminous coal, mixed with oil or with coal tar, is a very durable and excellent preservative, shedding water well, and protecting efficiently against oxidation. Fish-oil may also be used in some cases for % similar purpose. Sulphate of iron, in oil, has also been found to make a useful paint. The materials which enter into the composition of paint frequently exert a decisive effect upon its preservative qual- ities. Adulterated and impure paints may not only lack preservative qualities, but may, by their adulterations, actu- ally hasten decay. The most important constituent of paint is White Lead, which should be of good quality, and unmixed with any substances which impair its brightness. Its usual adulterations are chalk and the sulphates of lead and baryta ; the latter is the least objectionable. Zinc White is more expensive, and forms a better basis STRENGTH OF TIMBER. 6/5 than white lead ; it, however, works dry under the brush and takes longer in drying ; it does not have the covering proper- ties of white lead, but forms, however, a more dense coating, which resists the action of the weather and retains its color better than lead paint. It is frequently adulterated with sul^ phate of baryta. Red Lead makes a very durable paint and dries well. Linseed Oil is one of the most important constituents of paint ; it improves greatly by age, and ought to be kept at least six months before using. It can be made a "dryer" by boiling, or by the addition of foreign substances. It dries better than any other oil, has a heavy body, and it is owing to this fact that it is capable of resisting the action of the .weather. Pure linseed oil is of a pale, transparent, amber color, very limpid, and has little odor, is comparatively sweet to the taste, and when exposed to the light and air grows lighter in color. In adulterated linseed oil the opposite effect takes place. Nut Oil and Poppy Oil are inferior in quality to linseed oil, and are used to adulterate the latter. Of the colors, yellow ochre is used as a body color more extensively than any other — the best is very durable in color. Amber, Vandyke, and metallic browns are derived from the iron salts, and are also very durable ; they adhere to iron better, and are less affected by the air than the red lead. The real value of any paint depends upon the quality of materials used in composition, and upon the care used in its mixing and preparation. Charring the surface of well-seasoned timber is found to considerably increase its durability, and this is the method most frequently adopted for the preservation of those por- tions of fence posts which are buried in the ground. An external application of silicate of sodium, has been advised by Abel, for seasoned timber. It is said to form a hard and very durable coating upon the surface, and to act effectively as a preservative against fire, as well as against decay. The solution is laid on with alternate coats of lime wash. Two or three applications of the silicate of sodium ^7^ NON-METALLIC MATERIALS. are required to form each coat. Sulphates of iron and of copper, the chloride of mercury, common salt, and other solutions, are occasionally used for external washes. The common oil paints are, by far, the most usually ap- plied. Their durability is increased by sprinkling liberally with sand, where circumstances permit. In timber protected by external treatment, special care is required to fill cracks. The Saturation of Timber, either seasoned or un- seasoned, with antiseptic materials has become a matter of such great importance as to have attracted much attention. Many processes have been tried and recommended, but none are generally used in this country, and very few are practiced at all. A few seem to be effective, but costly ; many are of tem- porary benefit ; others, while seeming to be useful at first, are actually injurious, ultimately destroying the timber which they are intended to preserve. The external applications above described are of no value in defending the timber against the attacks of wood-boring insects. Sheathing the timber in metal, and one or two meth- ods of saturation, are apparently the only reliable expedients. Of the processes of preservation of wood by saturation, JCyan's consists in the injection of the bichloride of mercury (corrosive sublimate); Burnett used the chloride of zinc; Boucherie employed the pyrolignite of iro7i ; Margery used the sulphate of eftipper ; Bethell saturated his timber with creo- sote, or " dead oil," from gas works ; Beer used a solution of borax. The metallic salts owe their antiseptic property to the fact that they produce coagulation of the albumen, which is the fermentable and perishable part of timber. The use of metallic salts was proposed nearly a hundred years ago, but the first practical applications were made about a half century ago. " Kyanizing " was suggested by Sir Humphry Davy, some ten years before the process was patented by Sir R. H, Kyan, in England, in 1832. STRENGTH OF TIMBER. 677 The solution used consisted of one pound of the bichlo- ride of naercury in four gallons (i kilogramme to 33^ litres) of water. Timber thoroughly impregnated with the salt has great durability, but the general adoption of this process is precluded by the cost of materials. A hundred parts of timber absorbed one and a half parts of corrosive sublimate. Where it is brought in contact with iron it produces corro- sion, and its application is thus rendered still less frequently permissible. Kyanized timber was used to some extent in Great Britain and the United States when first proposed. Among other constructions of timber thus prepared may be mentioned the aqueduct of the Alexandria Canal, crossing the Potomac River at Georgetown. " Burnettizing " was proposed by Sir Wm. Burnett, in 1838, and has been quite largely practiced for special pur- poses. The chloride of zinc, in the proportion of one part dis- solved in ten parts of water, is forced into the pores of the wood under a pressure of from one hundred to one hun- dred and twenty-five pounds to the square inch (7 to 8.75 kilogrammes per square centimetre). Burnett's method was, originally, simple immersion in the solution two or three weeks. An establishment was organized at Lowell, Mass., in 1856, in which burnettizing under pressure was practiced. Subse- quently several railroad companies adopted this method and process, and erected burnettizing works. The cost of preserving timber by this method, including interest on capital and all other expenses, ranged from five to seven dollars per thousand feet, board measure. The process is not, however, believed to afford as perfect protection as the more expensive method of kyanizing. The Bethell Process was also patented in England in 1838, and its cheapness and effectiveness have given it a considerable commercial success, both in Europe and the United States. It consists in the saturation of the wood with bituminous substances obtained by the distillation of coal tar. 6/8 NON-METALLIC MATERIALS. Like the metallic salts, these substances produce coagulation of the albumen, and thus destroy the tendency to fermen- tation. Timber thus prepared is rendered very durable, and the process is comparatively inexpensive. Its use has, however, been given up, in some instances after extended trials, on the ground that the increase in durability was not sufficient to compensate for the expense. Each cubic foot of timber, under a pressure of 150 pounds per square inch (10.5 kilogrammes per square centimetre), absorbs, in twelve hours, from eight to twelve pounds (1.55 to 0.23 kilogrammes to the cubic decimetre) of the creosote or dead oil. The smaller amount is the allowance advised for railroad cross-ties. Hard woods absorb least. The strength of tim- ber preserved by this method is unimpaired, and it requires no painting, although, with dry timber, a superficial coating of coal tar is sometimes added. This process has special advantages where the timber is exposed to alternations of dryness and moisture, and is therefore liable to wet rot. The dead oil fills the pores completely, coagulates all albumen, and absorbs all oxygen that may exist free in the wood, and, by its poisonous qualities, it acts as a protection against the attacks of insects. It does not, however, afford perfect pro- tection against the ravages of the white ant of tropical coun- tries. Even marine insects usually avoid creosoted timber, and wood so prepared is therefore used to a considerable extent in submarine work. The antiseptic element of dead oil is supposed to be the carbolic acid, which is estimated by Prof. Letherby at from one and one half to six per cent, of the whole. The cost of creosoting 1,000 square feet (3.15 square metres), board measure, of oak or spruce fir, has been given at from five to eight dollars. The Seely, the Robbins, the Leuchs, and the Hayford Processes are American modifications of the Bethell process. The Seely Process consists in subjecting the wood to a temperature between 212° and 250° Fahr. (ioo° and 121° Centi- STRENGTH OF TIMBER. 679 grade) in a bath of creosote oil, for a sufficient length of time to expel all moisture. When all water is thus expelled the pores contain only steam. The hot oil is then quickly replaced by a bath of cold dead oil. The steam in the pores of the wood is thus condensed, and a vacuum is formed, into which the oil is forced by atmospheric pressure and by capil- lary attraction. From six to twelve pounds of creosote oil to the cubic foot (0.7 to 1.4 kilogrammes per cubic decimetre) of wood is expended in this process. The amount is dependent upon the use to which the wood is to be put. For piles or other timber exposed to the depredations of worms, twelve pounds is used. An impregnation of ten pounds to the cubic foot costs twenty-five cents. For work in wheel-pits and under foundations, at least ten pounds per cubic foot (1.118 kilo- grammes per cubic decimetre) should be used. For piles the usual charge is thirty cents per cubic foot. The RoBBlNS Process consists in treating wood with coal tar or oleaginous substances in the form of vapor. The wood is placed in an air-tight iron chamber, connected with which is a still or retort, heated by a furnace. When heat is applied, the vapor of naphtha is generated at a tem- perature of 250° to 300° Fahr. (121" to 149° Centigrade), the creosote oil vapor at 360° to 400° (182° to 204° Centigrade), and the heavier tar oils at 500° to 600° Fahr. (260° to 316° Centigrade). The wood is thus exposed from six to twelve hours. By this process it would seem impossible to charge the wood with more than a fraction of the amount of carbolic acid and of other component parts of coal tar expended in the Seely process. The latter process is decidedly an im- provement on the process of Bethell. The cost of creosoting 1,000 square feet, board measure, of oak or of spruce fir, has been given as from five to eight dollars. The Leuchs Process, as perfected by Hock, is applied to cross-ties in the following manner : The ties are introduced into an iron cylinder or reservoir 68o NON-METALLIC MATERIALS. heated on the outside by a steam jacket. The wood, already as dry as possible, is raised to the highest degree of desicca- tion by the introduction of steam into the jacket, and when no more vapor escapes from it, a solution of paraffin is forced into the cylinder through a tube, by compressed air. This cylinder has a refrigerating coil which discharges into a closed receiver. Then steam is let into the jacket again. The water boiling and the vapor of petroleum not being able to escape, the pressure inside the cylinder rises to 75 or 100 lbs. per square inch (5.25 or 7 kilogrammes per square centime- tre), at which pressure the wood is completely impregnated v/ith the liquid. When this operation has been prolonged sufficiently, heat- ing is stopped, and the operator waits until the pressure has fallen, and the excess of paraffin can be drawn off into the reservoir. The wood is again heated. When all remaining vapor of petroleum has been absorbed, air is blown into the cylinder to drive out gas, which might incommode workmen while removing the wood. The paraffin remains distributed among the ligneous fibres, enveloping them with a thin coating, at the same time filling the pores and cellular spaces. The wood is thus well protected against moisture. Nails driven into wood so treated, do not rust as in wood impregnated with metallic salts, and the preserved wood retains its value as fuel. The Hayford Process is one in which the wood is placed in a cylyidrical boiler, into which steam is admitted and atmospheric air forced, until there is attained a pressure of 30 to 40 pounds per square inch (2.1 to 2.8 kilogrammes per square centimetre), and a temperature of 250" to 270° Fahr. (121° to 132° Centigrade), which pressure and temperature suffice to evaporate the sap of boards and two-inch (5.08 centimetre) plank in four or five hours — ten or twelve hours being required for heavy timber. After this, the vaporized Sap and steam condensations are drawn out by air pumps. A vacuum is then produced which completely withdraws the vaporized sap from the very heart of the wood. The oil is admitted to the wood through perforated pipes arranged STRENGTH OF TIMBER. 68 i around the interior of the cylinder, under pressure such as, with the partial vacuum within the cylinder, is equivalent to. about 75 pounds to the square inch (^.25 kilogrammes per square centimetre). This is sufficient to cause the oil to penetrate the more porous woods. For those which are more dense, a further pressure of 60 to 150 pounds (4.2 to 10.5 682 NON-METALLIC MATERIALS. kilogrammes per square centimetre) is required for a certain length of time. The illustration on the preceding page exhibits the con- dition of timber thus creosoted, and uncreosoted, exposed in sea water at Wilmington, N. C. Creosoting is at present considered the most valuable process of protecting wood against the teredo. M. Paulet considers that the petroleum products con- taining phenic acid are preferable to metallic salts for treat- ing wood exposed to the action of sea water, because naphthaline, and especially phenic acid, have an antiseptic power, coagulate the albumen, and thus check the circulation of the sap, and also of the blood of parasites ; that the vola- tility and the solubility of these preservative agents would render their antiseptic action temporary only, if the more fixed and thicker oils which accompany them did not enclose and retain those substances, at the same time obstructing all the pores of the wood, and rendering difficult the access of dissolving liquids and destructive gases ; but that grave objections have been raised, based upon restricted produc- tion of these oils, and on the fact that the wood thus im- pregnated is inflammable, while on the contrary, all the metal- lic salts render wood uninflammable. Boucherie's Process, patented in 1839, is an ingenious and inexpensive method of saturation. This process attracted much attention, and was practiced with considerable success. The timber, freshly cut, and with its terminal foliage still remaining, was set either vertically or horizontally, with the foot immersed in a vat containing the antiseptic solution. The circulation continuing in the trunk of the tree, the sap becomes ejected, and its place is taken by the preservative solution, which is thus thoroughly distributed throughout the fibre. » Growing trees were also treated by the injection of the liquid into their trunks. Where logs, deprived of their foliage and branches, were to be saturated, they were placed on end, and a waterproof bag, or a tank, containing the solution used, was mounted above it, the liquid being thus forced down- STRENGTH OF TIMBER. 683 ward through the stick by hydrostatic pressure, driving the sap before it and out of the lower end. The antiseptic proposed by Dr. Boucherie was crude py- rolignite of iron. His process of saturation was largely used with other preservatives also, and his invention of the satu- rating process seems to have been more generally appreciated than his introduction of a cheap antiseptic. Where it can be conveniently applied, it is exceedingly efficient. Numerous and elaborate experiments were tried by Dr. Boucherie, in which the action of pyrolignite of iron was carefully noted. He found that one-fiftieth of the weight of the green wood was a sufficiently large proportion of the antiseptic to insure preservation. The hardness of the wood was stated to be doubled by the use of the pyrolignite. Solutions of deliquescent salts were applied by Dr. Boucherie in the way described, and were found, in the case of chloride of lime and some others, to increase the flexibility of timber. He therefore proposed the use of such solutions, with the addition of one-fifth their quantity of pyrolignite of iron, when it was desired that the wood should retain its moisture, and its flexibility and elasticity. The same inventor proposed, as a cheap substitute for these solutions, the stag- nant water of salt marshes. Such preparation, it was claimed also, prevented the warping and splitting of wood, which is a frequent consequence of rapid drying, and yet seasoning was said to be expedited by its use. In this case, the solutions were weak, and the wood could be afterward painted over without difficulty. The process was applied by the inventor to the saturation of timber with earthy chlorides, as a protec- tion against fire. These salts, fusing upon the surface of the wood on the application of heat, rendered it quite incom- bustible. ■^ood was dyed with both mineral and vegetable colors by Dr. Boucherie, and the application of the usual method of producing " fast " colors, by the introduction of dye and mordant successively, was thus made practicable. Wood was treated with odorous solutions to give it fragrance, and with resinous matters to make it water-proof. 684 NON-METALLIC MATERIALS. The French Government, after receiving favorable reports from the Commission of Engineers appointed to examine into the merits of the process, finally conferred upon the inventor the great gold medal of honor. Stabsequently a money award was made him, and he surrendered his patent, which thus became public property. This is still considered one of the best processes yet devised. " Beerizing " consists in the saturation of the timber by any convenient process, with a solution of borax. This is claimed to dissolve the albumen, and the solution may be allowed to remain, the borax having antiseptic properties, or it may be washed out, and the wood then dried is stated to become more thoroughly seasoned and durable than it can be made by the ordinary process of seasoning. Folacci's Process of securing incombustibility and impermeability of woods consists in their impregnation with a composition consisting of : Sulphate of zinc 25 parts. Potassa 10 parts. Alum 20 parts. Oxide of manganese 10 parts. Sulphuric acid at 60° B. 10 parts. Water 25 parts. 100 parts. These chemicals are mixed and heated without the sul- phuric acid to a temperature of 45° Centigrade (113° Fahr.), and the acid is then added gradually until solution is com- pletely effected. It may be applied by Boucherie's or by any other con- venient method. Margery's Process of saturation with sulphate of copper has been found very effective in some instances. It was applied by the Boucherie method to telegraph poles and to railroad cross-ties many years ago in France, with perfect success as a preservative against decay. When thus prepared telegraph posts last from fifteen to twenty years. They are subjected to the process in the forest di- STRENGTH OE TIMBER. 685 rectly after cutting, and while yet full of sap ; the expense of thus treating them is usually from one to one and a half dollars per post. The wood is more rapidly and perfectly protected in proportion as it is porous and rich in sap. After under- going the preservative process the timber is seasoned and becomes very light and portable. Nearly all the posts of the French, German, and Belgian telegraph service are now treated by this process. The salt used is poisonous to vegetable and animal para- sites which appear at the beginning of all organic decompo- sition. The quantity of the salts of copper should be increased when the wood is intended to be immersed in water or buried in a moist soil, as water dissolves this salt slowly. There is in wood impregnated with the salts of copper a portion of the sulphate closely united with the ligneous tissue, and another portion in excess remaining free ; this latter por- tion dissolves first, and, carried off by the exterior fluids, only retards the loss of the metallic salt combined with the wood ; but this combination itself, although more stable, does not escape removal, which is accelerated or retarded according to the rapidity and ease with which the dissolving liquid is re- newed. The quantity of metallic salt should be small in wood intended for constructions in the open air, in order to pre- vent mechanical injury due to crystallization. Major Sankey found this process equally efficient in India, more recently, as a protection against the attacks of the white ant and other insects. He used a solution of one pound of the salt in four gallons (i kilograriime to 33^ litres) of water. The timber was steeped in the solution two or two and a half days for each inch (2.54 centimetres) in thickness. A simple coating of boiled linseed oil thickened with powdered charcoal, has in some cases been found a very eco- nomical and efficient preservative of timber. Statistics of railroad constru ction have given the fol- lowing data : Of unprotected oak cross-ties on European roads, 25 per cent, were renewed in 12 years, and 50 per cent, in 17 years. When impregnated with chloride of zinc, 3^ per cent, were 686 NON-METALLIC MATERIALS. renewed in 7 years, and 20 per cent, in 17 years ; when pro- tected by "dead-oil," o.i per cent, were replaced in 6 years. Of pine cross-ties impregnated with chloride of zinc, 45^ percent, were renewed in 7 years, and 31 per cent, in 2t years. Of ties of beech, protected by creosote, 45 per cent, were replaced in 22 years. On the railroad from Hanover and Cologne to Minden, of pine ties injected with zinc, 21 per cent, were renewed after 21 years; of beech ties injected with creosote, 46 per cent, after 21 years; oak not injected, 49 per cent, after 17 years; oak ties injected with chloride of zinc, 20.7 per cent, after 17 years. The ties not renewed appeared perfectly sound. In all these cases the ties were laid in favorable situa- tions. It was reported to the German Railway Union, in 1881, that, on the railways of Europe, chloride of zinc was most generally used for preserving timber, and creosoting next, while the use of sulphate of copper was declining. Preserva- tion with chloride of zinc cost less than one half as much as creosoting. Seasoning was considered desirable before creo- soting, but not when using the salts of zinc. The importance of the preservation of timber is daily increasing, not only as a matter of ordinary economi- cal policy, but because the rapid destruction of forests is continually rendering timber more scarce and more costly. It will become a matter of such vital necessity to preserve our forest trees, that legislation will soon inevitably aid in increasing the market value of timber by forbidding its whole- sale destruction. The substitution of iron for wood, in construction, is pro- ceeding so rapidly, that it will afford some relief ; but ^he preservation of timber will nevertheless remain a matter of exceptional importance. CHAPTER XXI. MISCELLANEOUS MATERIALS : Leather; Belting; Paper; India Rubber; Gutta .Percha : Cordage. Leather ; Belts. — One of the principal uses of leather in engineering is in its application in the form of belting, for driving machinery. The best quality is well tanned ox-hide, cut from the back of the animal, and very exactly trimmed, to form perfectly straight strips of uniform thickness. These strips, which are from 4 to 6 feet (1.2 to 1.8 m.) long, and usually about three-sixteenths of an inch (0.48 cm.) in thick- ness, are scarfed, spliced, and cemented end to end, to make any desired length of belt. " Single " belts are those made of a single thickness of leather. Extra strong belts are made by cementing or rivet- ing together two thicknesses of leather to form a " double belt." Under light loads the single belt has the greatest adhesion ; but under heavy loads the double belt is fully as efificient. Double belts are sometimes made 6 feet (1.2 metres) wide, and loo to 150 feet (30 to 45 metres) long. The inside of the hide is called the " flesh side ; " the out- side the "grain side." The belt wears best when placed with the flesh side next the pulley. Some engineers, how- ever, advise the reverse position, as the belt is less liable to slip. The weight of hard, well-tanned belt leather is about that of water, 62^ pounds per cubic foot (1,000 kilogrammes per cubic metre), and may be taken, at an average, as about 0.85 pounds per square foot (4 kilogrammes per square metre). The tenacity of belt leather of good quality is about 650 pounds per inch in width (115 kilogrammes per cm. wide), one half that amount when spliced and riveted, and one 688 NON-METALLIC MATERIALS. third when laced. The safe working tension is given by some engineers as 50 pounds per inch of breadth (9 kilo- grammes per centimetre). Belt lacings are strips of sheepskin, about a half inch (1.27 cm.) wide, and a yard (0.9 metre) or more in length. In joining belt ends, the belt is pulled taut, and, if heavy, stretched and held by " belt-clamps," cut so that the ends just meet, and a single, or double, or even a triple row of holes, according to its size and tightness, punched to re- ceive the lacing. The holes should be exactly in line. The lacing is then passed through these holes, backward and for- ward, joining the two ends evenly and strongly. Small, strong tempered steel hooks are often used instead of lacing. Calfskin, well tanned, stretched wet, makes good lacings. The firmest and best method of uniting leather belts is to scarf the ends so as to lap a distance of not less than ten, nor usually more than twenty, times, their thickness, adjust- ing the length of the belt carefully by setting up well with " belt-clamps," and then cementing the parts well, finally securing the lap by copper rivets. The next best method is probably that of connecting the ends by steel hooks, and the least effective, but probably most usual, way is by lacing. The stress allowed on a single belt may be taken at 300 pounds per inch of width (55 kilogrammes per cm.) Morin allows 284 pounds per square inch (20 kilogrammes per square cm.), and Claudel 355 pounds (25 kilogrammes). Rankine takes 285 pounds per square inch of section (20 kilo- grammes per square cm.), and assumes the usual thickness as 0.16 inch (0.4 cm.). In very fast-running belts, the tension given in setting up is sometimes increased by centrifugal action to such an ex^ tent that it should be allowed for in calculating the width required. This " centrifugal" tension " is thus measure^ : A belt of the section s, density d, and running at the ve- locity V, in passing from one side to the other of the pulley exhibits energy, in one direction, measured each second by E = sdv X — ; MISCELLANEOUS MATERIALS. 689 this energy is first destroyed and then is reproduced in the opposite direction on the other side of the pulley. . The total energy is thus : 2E = sd ■ g 2E The effort demanded to produce this reversal is , V or. ■p _ 2^ _ sd")? T The required section of belt now, instead of j = -— , in. which T'is the total tension, and t is the safe working tension per square unit of section of the material as above, becomes, - ^ t g The Friction of leather belts is relied upon to prevent slipping. This friction is at each point of contact propor- tional to the pressure there existing, and the total resistance to slipping is obtained by summing the resistances through- out the arc of contact. The pressure at each point is equal to the tension at that point, as is seen when it is considered that it is the same as would exist were that tension uniform, and were there no friction throughout the arc of contact. But this tension is actually variable in consequence of the existence of friction, and we have, as shown by Rankine, dR = fTd6, = dT, when R = Ti — T^, the working stress, i. e., the difference of tension at the extremities of the arc,/ = the coefficient of friction, and d = the arc of contact ; and 44 6gO NON-METALLIC MATERIALS. since T, = T^y' ; T^ = T.e-/' ; T,+ T^=T^ {e^ + i). The mean tension is -^-^ — ^, and 2 • 2{/'- 1) Since ^ = 2nn, when ;? = arc of contact in " turns," /»_ jQ2.7288/«^ and /= 0.42. When this arc is. expressed in degrees, R = Ti{i - io-''-°°3=«). The following are values * of the factors : When 6 = 7t and « = J^, as where the pulleys are of equal size : TENSION AND FRICTION OF BELTS. /= 0.15 0.25 0.42 2.7288/= 0.41 0.68 1. 15 y, + n = l.6o 2,20 = 2.66 1.84 = 2.16 1-34 3.76 1.36 0.86 In common practice, it is now usual to take / = 0.22> or/= 0.25; n^ % 71, = R: =^^^' = i-S; R = }^ {Ti + Ti) = 150 lbs. per inch (28 kilogrammes per centimetre) of width of belt ; but one-half this stress is ofCen observed. An old millwright's rule allows one horse-power for each inch in width running 1,100 feet per minute, i. e., * Rankine, Machinery and Millwork, | 310. MISCELLANEOUS MATERIALS. 69 1 hv HP = ; (2.54 centimetres width running 335 metres per minute. Belts are often driven without slip, to nearly- double this power. Nagle gives the following : For laced belts — HP= ctvw (0.55 — 0.000021 5 7^^'') For riveted belts — HP= ctvw (i — 0.00002157^2/*) when ^ _ I _ jQ-o.ooogsB/B . 6 = arc of contact, degrees ; w = width of belt, inches, i = thickness ; V — velocity, feet per second. Morin gives the following as maximum coefficients of fric- tion of belts : FRICTION OF BELTS. /- Common cases, iron pulleys 0.28 Wetbelts, " " 0.38 Common belts, wooden pulleys o. 47 New " " " 0.50 Where the arc of contact varies, the value of -^ may be altered thus (/ = 0.28) : » = 0.2 0.3 0.4 0.5 0.6 0.8 i.o —1=1.4 1-7 2-0 2.4 2.9 4.0 5.8 Raw Hide, or untanned leather, when perfectly sound, is much stronger than tanned leather, and is much used for some parts of textile machinery connections, in looms, for ship's tiller-ropes, etc. It is cut from the raw skin and dried in the sun. Its strength may be taken as one-half greater than that of leather ; its resistance to violent impact is very great. 692 NON-METALLIC MATERIALS. The Cement used for belts may be made by melting to- gether :* I part shellac ; 2 parts pitch ; 2 parts linseed oil ; 4 parts India-rubber ; 16 parts gutta percha, until thoroughly incorporated. It is applied warm, in a thin coating, very quickly, and the two parts of the belt are promptly and firmly clamped together and left until completely set. Leather is used for the packing of pumps, and often for their valves. Paper is principally used by the engineer in the draw- ing-room ; but it is occasionally applied in the making of belts, of packing and of other needed articles. Drawing paper for nice work is made of linen rags re- duced to a pulp, and formed into thin sheets having a smooth, peculiarly varied surface, which takes ink and colors well, bears erasures, and is strong and durable. The finished drawings are sometimes coated with a solution of shellac in alcohol ^ (i part shellac to from 4 to 8 of alcohol), which dis- colors the sheet, but which enables the draughtsman to wash it when soiled, and prevents that rapid soiling which always occurs when working drawings are handled. Rough drawings are made on cotton paper of a cheap grade, which comes in long rolls of considerable width. Shop dra-\Vings are usually copies on tracing cloth, or photographically prepared " blue prints," made by using the tracing as a negative, and are mounted on a board to avoid injury by rolling or bending. The principal sizes of drawing paper are : Medium 18 x22 inches. Royal 19x24 " Imperial 21^x29 " Elephant 224x27^ " Columbia 23x33! inches. Atlas ■ 26x33 " Theorem 28x34 " Double Elephant 26x40 " Of tracing paper we have : Double Crown 20x30 inches. Double D Crown 30x40 " Double D D Crown 40x60 " Grand Royal 18x24 mches. Grand Aigle . ■ • t 27x40 " * Molcsworth. MISCELLANEOUS MATERIALS. 693 Tracing cloth comes in rolls of various widths, as does vellum writing paper. Blotting paper is a thick cotton paper, perfectly free from size or greasy matters. Lithographic paper is made by coating printing paper with a composition of one part alum, two of gum arabic, and six of starch, dissolved in warm water, and laid on hot with a brush. Tracing paper is made by washing printing paper with a mix- ture of either Canada balsam and oil of turpentine, or nut oil and turpentine, and thoroughly drying before using it. Copying (" manifold ") paper is writing paper coated with lard and blacklead. Paper of considerable thickness is known as pasteboard, and is extensively used for making boxes to contain light materials. Paper Belts are sometimes used. They consist of a very hard pressed paper, have great strength and considerable durability, but a low coefficient of friction, and are very stiff and unmanageable. The same material makes the ex- ceedingly light and strong boats used for racing purposes. They are formed and compressed in moulds, and are there- fore properly "papier mach^." Calender Rolls are made of paper formed and compacted by the hydraulic press. This compacted paper becomes as close in texture as hard wood, very strong, with a very fine, smooth surface, and works like a soft metal. India Rubber and Gutta Percha are used in special forms by the engineer ; in bands for belting, in sheets for packing, and, to a limited extent, for various minor pur- poses. India-rubber, or caoutchouc, is the dried juice or sap of several tropical trees or shrubs. The best, the " Para gum," IS obtained from the Hevea Braziliensis, one of the Euphorbi- acese, Ceara rubber from the Manihot glaziovii, and Pernam- buco gum from the Hancornia speciosa. This gum is also obtained from the East Indies, Africa, and Central Amer- ica. Pui;e rubber, if of good quality, is dry, tough, strong, and 694 NON-METALLIC MATERIALS. enormously elastic. It dissolves freely in benzole, chloro- form, carbon disulphide, and the essential oils ; contact with oil or grease rapidly destroys it. All the rubber used by the engineer is " vulcanized " by heating it and incorporating with it 20 to 30 per cent, of sul- phur ; it then becomes less readily softened by heat or hard* ened by cold, and makes very durable water-proof articles of many kinds, all of which are made of woven fabrics smeared with the vulcanized rubber. When the proportion of sulphur reaches 30 or 40 per cent., various grades of "ebonite" are produced — a hard, jet black, moderately elastic substance, used by the engineer for making rulers, scales, " triangles," and curves, and in the arts generally, for a great variety of purposes. India-rubber Belts are made by weaving cotton canvas of the required length and width, and coating it with vulcanized rubber. These belts are made two, three, or four-ply, as they are required to do work demanding the strength of two, three, or four thicknesses. These belts are considerably stronger than leather belts, are usually truer and run more smoothly, and are perfectly impervious to water ; they have a higher coefficient of friction, but, if overloaded, are apt to be rapidly and seriously injured by slipping. India-rubber Valves are largely employed in hydraulic machines. They should be made of well vulcanized rubber, uniform in thickness, cut precisely or moulded exactly to size and shape, and ^ould have just sufficient thickness to safely sustain the pressure thrown on them. In many cases they outlast metal valves. Gutta Percha is the dried and hardened sap from the bark of trees of the order Sapotacem, found plentifully in the Malay Peninsula. When pure it is grayish white, becoming brown and yellow with exposure or from the presence of impurities. It is as hard as the softer woods, and can be easily moulded or rolled into sheets having considerable toughness and without elasticity, resembling in density and texture hard leather, having a specific gravity of 0.98 to i.oo. In solubility it resembles/rubber, as well as in nearly all other MISCELLANEOUS MATERIALS. 695 properties except elasticity. It is an excellent insulator, and is extensively used in telegraphic engineering. It is also used for belting, the same proportions being adopted as with leather. Cordage is usually made of hemp, flax, and cotton, and sometimes of leather, rawhide, and often of wire. Small cordage is known as rope ; it is usually composed of three or four strands of "yarns," laid up with a right-hand twist, the strands being laid up with a left-hand twist. Haw- sers are made up with three right-handed strands, and cables with three hawsers laid up left-handed. Shrouds are- made up with a central core, surrounded by four strands. Tarred Ropes are less subject to injury by the weather than white cordage, but have one fourth less strength. Cordage takes up from 20 to 30 per cent, of its weight in tar. The larger the cordage the less its strength per unit of sec- tion ; this loss amounts to nearly 50 per cent, in large cables. Three-strand cordage is ten or fifteen per cent, stronger than four-strand, if rope laid ; ten per cent, weaker in hawsers and cables. The working or maximum proof strength of cordage may be calculated by multiplying the square of the girth in inches by 200 to 300 pounds, or in centimetres by 14 or 22 kilo- grammes. White 2-inch (5 centimetres) hemp rope should carry about 5,000 pounds (2,27s kilogrammes), or nearly one ton weight per pound weight per fathom (say 1,200 kilogrammes per kilogramme weight per metre). The U. S. navy test allows 4,200 pounds on a i^-inch white hemp rope, or 1,700 pounds per square inch (1,195 kilogrammes per square centimetre). Manilla rope has about two-thirds the strength of good Rus- sian (Riga) hemp. The method of connection of ropes is usually by making knots ; although permanent union is effected by a " splice," in' which the two ends are overlaid for a considerable distance, and their strands mutually interwoven, making a connection as strong as the body of the rope. Knots cannot well be verbally described ; but the following 696 NON-METALLIC MATERIALS. engravings, selected from Molesworth, represent the principal knots used by the engineer and the seaman. In these illus- trations, W indicates the direction of the weight, P that of the pull. Fig. 162 is the "half hitch," used to secure the end of a line to any object during a steady pull ; Fig. 163 is a "timber hitch," Fig. 162. Fig. 163. Fig 164.-T1MBER Hitch^ m p 1 o y e d for the Half Hitch. Timber and Half Hitch. same purpose, when Hitch greater security or a more permanent hold is desired. Fig. 164 exhibits the two hitches used together, the timber hitch backing the half hitch ; this arrangement is used at sea in towing spars. With the "clove hitch," Fig. 165, the stick is held in position by a pull on each side ; the " rolling hitch," Fig. 166, is a still tighter knot than the timber hitch, as it rolls the lines over, and binds itself as soon as the pull is given. Fig. 165. — Clove Hitch. F1G.166. — Rolling Hitch. Fig. 167. — Reef Knot. The " square knot," or " reef knot," Fig. 167, is the sim- plest and best knot for uniting rope ends ; the " sheet bend," Fig. 168. — Sheet Bend. Fig. i6g. — Sheet Bend WITH Toggle. Fig. 170. — Bowline. Fig. 168, is somewhat similar ; the " sheet bend and toggle," Fig. 169, is easily unloosed, and the "bowline," Fig. ^70, forms a loop which can be thrown, lasso-like, over a post, to which it is proposed to make fast. Fig. 171 shows the method of putting a "stopper" on a rope or cable, for the purpose of holding it in place while shifting the end, or while fleeting it at a winch. The " black- MISCELLANEOUS MATERIALS. 697 Fig. 171. — Stopper on a Rope. wall hitch," the " fisherman's bend," the " round turn and half hitch," Figs. 172, 173 and 174, show how a rope may be made p Fig. 172. — Blackwall Hitch. Fig. 173. — Fisherman's Bend. Fig. 174. — Round Turn and Half Hitch. fast to a hook, or to a link or deadeye ; the lashing shown at A on the last two makes the line secure. A "sling," or "strop," Fig. 175, which may be either a rope or a chain, has many uses ; it is seen in Fig. 176, as used in raising a stone or other heavy mass; in Fig. 177, as used to give a hold for the hook of the tackle; and, in Fig. 178, is illustrated the attachment of stop and " guy ropes " to the head of a derrick. c 2) Fig. 175. — A Strop or Sling. Fig. 178.— Head of a Derrick. Fig. 176. — Slinging a Case. Fig. 177. INDEX. Abel, condition of carbon in steel, 497, 498. phosphorus-copper alloy, strength,227, 233- silicate of sodium as a preservative of wood, 675. Abrasive resistance of brick, cement, stones, 573. Acacia tree, 628. Acid absorption of iron and steel, 212. oxides formed in roasting ores, 54. Accumulator, hydraulic, 74. Adamsonia tree, longevity, 598. Admiralty tests, blacksmiths' irons, 492. eflect of heat on bronzes and kal- choids, sag-SSI- African iron furnaces, 44. ores, 5c. metal working, 4. Age, effect upon metals, 498-500. of trees. See Longevity of trees. Agne, brass, 320. Air-hoist, 80. Air lime. See Lime. Aitken, classification of brass work, 305. Alabama tin, 239. Alabaster. See Gypsum. Alder wood, crushing resistance, 638. durability in wet construction, 659. uses, 6sg, 660. Alkalies in fluxes, 8. Alkaline bases, copper ore reduction, 219. metals, 37. Alleghany iron ores, 50. Alligator squeezer, 109. Alloys, 37, 38, 270. aluminium, 256, 2S7. bronze, 306, 309. silver, 309. tin, 18. amalgams, 238, 2S8, 259. anti-attrition metal. Babbitt's, 325, 326, 330. antimony, 253, 312, 332-335. bismuth-lead, 312. -tin, 312. brass, 333. tin-lead, 312. argentan, 530. argentiferous, 330. Babbitt's metal, 325, 326, 330. Bath metal, 298. bell metal, 276, 278, 284, 330, 333. Berthier's antimony-tin-lead, 312. copper nickel, 310. bismuth antimony-lead, 312. -tin, 312. See Brass. Bristol metal, 299, 300. Britannia, 239, 312, 313, 330. See Bronze. button metal, 2gg, 300. cartridge metal, 454. casting, 317, 321, 463. chemical combinations, 270, 271. Chinese gong, 330. silver, 330. white copper, 330. chrisocalle, 300. 700 INDEX. Alloys, amalgams, coin, aluminium bronze, 308. bronzes, 277. copper with gold and silver, 38. copper-nickel, 309, 310. nickel, 255. See Copper. crystalline form, 26. crystallization by slow cooling, 321. electric resistances, 18, 19. expansion in cooling, compositions, 330. Farquharson, 487. Fenton's, 334. ferrous. See Iron. formed by impurities in metals, 16. fusible, 331. gold-leaf, 297. gold-palladium, 267. gong. See Bell-metal, hardness, 16. Haswell's table, 330, 331. heat conductivity, 10. influence of contained metals, 321, 322. iridium, 33, 264, 265, 313. iridium, platinum, 33, 313. iron, 177. iron, brass, 177, 302. iron, bronze, 177, 302, 304, 334. See Kalchoids. solder, 331. Kingston's, 335. lead, antimony, bismuth, 312. lead, antimony, bismuth, tin, 312. lead, antimony, tin, 312. lead, bronze, 321, 324. ' lead, tin organ pipes, 311. manganese, 265, 266, 313' bronze, 304, 306. Mackensie's, 300. Margroff's kalchoid, 334. medal bronze, 283, 301. mechanical mixtures, 270, 271. melting. See Brass, mock gold-leaf, 297. mosaic gold, 298. Muntz metal, 290, 29S, 530, 531. Alloys, nickel-brass, 309, 310, nickel iron, steel, 77. nickel copper, 310. See Ordnance, oreide, 296, 297. palladium-gold, 26^. Paris-Lyons and Mediterranean R. R standard, 329. Parsons' magnesium-bronze, 304, 306. pewter, 23, 33, 239, 312, 313, 330, SSI- phosphor-bronze, 274, 281, 381, 530i 531, 562. physical properties, 38. pinchbeck, 296, 330. platinum-iridium, 33, 313. platinum-iron, steel, 177. queen's metal, 312. safety factors, 562. silver, 138. silver-aluminium, 309. See Solder, speculum metal, 276, 277, 279, 285, 300, 330. Spence's metal, 314. stereotype metal, 312, 324, 323, 330. sterro-metal, 302, 458, 465. ternary alloys, 310, 312. thermal conductivity, 10. tin amalgams, 238, 239. tin-antimony-lead, 312. tin-antimony-zinc, 312. tin-lead, 239. tin-zinc, 311, 312. Tobin's kalchoid, 288. tombac, 289, 296, 297. tungsten,' 151, 267. tutenag, 330. type-metal, 38, 330. Vienna gold-leaf, 297. white. See Mangfanese-bronze. white copper, Chinese, 330. working. See Brass. See Zinc. Alluvial ore, tin, 235. Alum ; aluminium sulphate, 236, 257. Alumina ; aluminium oxide, 257. INDEX. 701 Alumina in porphyry, 256, 257. slag from puddling furnace, 186. silicate. See Clay. in fire brick, 581. hydraulic limes, 583. slag, g. Aluminium, 25-29, 256-258, bronze, 306, 309. electrical resistance, 19. electric conductivity, 18, 19, 26, 256. heat conductivity, 256. expansion, 33. in iron, 169, 180, 190. price, 269. silver alloy, 309. steel, 309. Aluminum. See Aluminium. Amalgamation in metal reduction, 7. Amalgams, 258, 259. tin, 238. Amboyna wood, 660. American antiquities, metal, 41, 273, 274. Amorphous metals, 27. Amphibole, hardness, 573. Ancient metals, 3, 4, 10, 12, 41, 45. Anderson , tenacity of copper, 429. sterro-metal, 465. Angle iron, 117. as columns, 401. Annealed bronze, 275. compositions; specific gravities, 282, 284. cast iron, 127, 136, 179, igo. copper, 227, 229. ingot irons and steels, tenacities, 370. steel, temper of tools to cut, 208. Annealing bronzes, 279, 536-538. See Furnaces. steels by boiling water, 210. castings, importance, 497. tempered, 205, 209. tin-plate, 129. white iron, 179. wire, effect on tenacity, 497. zinc, 243. Anodes, nickel, 255, 256. Anthracite coal, 67, in blast furnaces, 47, 57, 59, 62, 63, 75. Anti-attrition metal. Babbitt's, 325, 326. Antimony, 22, 26, 27, 29, 252, 253. alloys, 253, 312, 322, 335. atomic weight, 29. crystalline form, 26, 27. distillation, 253. electric conductivity, 26, 253. fusibility, 36. gray, 253. hardness, 17. heat conductivity, 26, 253. expansion, 253. in iron, 177. melting point, 253. ores, 253. properties, 252. salts, 253. solder, 35. specific gravity, 22, 26. specific heat, 26, 27. symbol, 247. tenacity, 252, 446. tin-zinc alloys, 312. uses, 253. weight, 22. Apple tree, longevity, 598. wood, uses, 659, 66r. Aqueous chemicals in metal reduction, 7. Arch brick, 379. « Area reduction. See Tensile tests. Argentan, 330. Argentiferous alloy, 330. Argillaceous ores, loss by roasting, 54, 55 stones. See Clay, Slate. Arsenic, 263, 267. atomic weight, 29, 263. cost, 269. in iron ore, 54, 18. melting point, 36. ore reduction, 54, 181, 263, 264. properties, 264. specific gravity, 22, 264. specific heat, 2g, sulphite, 263. Swedish iron, 187. 702 INDEX. Arsenic, symbol, 263. volatilization, 264. Arsenide of nickel, 254. Art castings, 11, 12, 322-324. Artificial stones, 582. See Beton, Brick, Concrete, effect of heat upon, 572. Art-work, 11, 12. Ashlar masonry, 593, 596. Ash tree, 627. longevity, 598. wood, columns, 641. crushing resistance, 638, 641. deflection, 651. elasticity, 636, 651. hold upon spikes and screws, 668, 669. markings, grain, 662. shearing resistance, 646. tenacity, 637. torsion coefficients, 655, 656. transverse strength, 636, 647, 651. uses, 660. Asiatic art work, 11, 12. Asphalt cement, 588. Atomic weight, aluminium, 256. arsenic, 263. bismuth, 253. copper, 29, 446. gold, 446. iron, 29, 446. lead, 446. magnesium, 262, 446. manganese, 265. platinum, 260, 446. silver, 446. tin, 236, 446. zinc, 243, 446. Austrian wainscot logs, markings, 662. Axed stones, 590. Axle bearings, alloys, 284, 332, 334. Azurite, 217. B. Babbitt's anti-attrition metal, 325, 326 330. Bailey's bronze, heat expansion, 33. Ball, puddle, 88, 89, 104-112. Banca tin, 234, 239. analysis, 237. tests, 439-444- Baobab tree of Senegal, longevity, 598. Bar. See Transverse tests. Barium, 267. cost, 268. Barlow, extension of iron by stress, 31. formula, thickness of water pipes and cylinders, 79, 384-388. strength of wooden beams, 646. Barytes, hardness, 573. sulphate, adulterant of white lead, 675. Basalt, 567. Basic dephosphorizing process, 165, 167. linings in Bessemer converters, 165. oxides formed in roasting ores, 54. Basswood, 615. Bast, 615. Bath metal, composition, 298. Bath stone, abrasive resistance, 573. Baudrimont, tenacity affected by temper- ature, 532. silver, 446. Bauschinger, formula for tenacity of Bessemer steel, 371. Baywood, Honduras mahogany, 631. Beams. See Girders. iron shapes, 118. strength, formulae, 403-405, 650-654. See Transverse tests. Beam-and-rail mill, 115. Beardslee, elastic limit, variations with time, 520, 522. tenacities modified by forms of test- pieces, 367. size of test-pieces, 365, Bearings, alloys, 329. , hardness, 16. bronzes, 283, 284. bronze-antimony, 335. kalchoids, 334. wood, 660. " Bears " in blast-furnaces, 58. INDEX. 703 Becquerel, electric conductivities of metals, 18. Bedson, continuous rolling mill, 115. Beech trees, 625. conversion of sap-wood into heart- wood, 598. wood, columns, 641. crushing resistance, 638. durability in wet construction, 659. elasticity, 659. glued joints, strength, 669. hold upon spikes and screws, 668, 669. railroad ties, 686. shearing resistance, 646. stiffness, 651. tenacity, 637. toughness, 659. transverse strength, 647. uses, 659, 660. Beer's wood preservative process, 676, 68g. Bell, analyses of furnace gases, 65. temperatures, blast furnace, 69, 70. Bell metal, bronze, 276, 278, 284, 330, 333- Belgian process of zinc manufacture, 330, 333- Belpaire, formula for strength of flues, 394- Belting, 687-695. Bending stresses, 344. See Transverse tests. test for metals, Bischof's, 447. Bends, rope, 696, 697. Benoit electrical resistance, 19, 20. Berard iron 'furnace, 165. Berea sandstone, crushing resistance, 574- Bermuda cedar wood, tenacity, 637. juniper, 617. Berthier, analysis of Swiss copper, 228. antimony-tin-lead alloy, 312. copper-nickel alloy, 310. Bessemer irons, classification, 136. basic dephosphorizing process, 165, 167. Bessemer irons, pneumatic process, 49, 149, 151-169. wire, 125. steels, analyses, 194-196. classification, 134-136. tenacity, 356, 363, 371. variation with temperature, 491. torsional resistances, 415, 421, 424, 425. Best bar iron, 105. Bethell, wood preservative process, 676- 678. Beton, 585, 586. effect of heat upon, 572. See Cement, Concrete, Mortar, 573^ 582-588. Beton-Coignet, 586. Bevan, hold of woods upon nails and screws, 666-669. Bible, references to metals, 4, 41, 42. Bichloride of mercury for wood preser- vation, 676, 677. Birch wood, English, crushing resistance, 638. shearing resistance, 646. shrinkage, 601. stiffness, 651. tenacity, 637. transverse resistance, 647. uses, 659, 660. Bird's-eye maple tree, 630. wood, grain, markings, 662, Birmingham wire gauge, 124. Bischof, bending test, 447. Bismuth, 252-254. antimony-lead alloy, 312. atomic weight, 29, 253. cost, 254, 269. crystallization, 26, 27, 253. electric conductivity, 18, 26. hardness, 17. heat conductivity, 18, 26. latent heat, 37. melting point, 26, 35, 36. specific gravity, 22, 26, 253. specific heat, 26, 27, 29, 253. tenacity, 446. 704 INDEX. Bismuth, weight, 22. Bitts, temper, 208. Bituminous cement and concrete, 588. coal for iron furnaces, 57, 181, 182. specific gravity, 67. specific heat, 67. weight, 67. Black birch wood, 637. Black copper, 223. Black oxide of copper, 217. Black spruce fir, 613, 614. torsional coe6Scient, 656. Blackwall hitch, 697. Black walnut. See Walnut, black, wood, crushing resistance, 638. shearing resistance, 647. tenacity, 637. torsional coefficients, 655-657. uses, 659. Blair, analyses of chain irons, 188. steels for U. S. Board, 193. Blanks for records of tests, 360, 361. Blastfurnace, anthracite, 47, 59. bloomary, 45-47, 83-85. blowing machinery, 76. Catalan, 83-85. charge, 57, 65-67. charcoal, 59. coke, 59, 70. construction, 61, 62. Deccan, 45. forms, 58-60. fuel consumptions, 70. height, 69. * historical, 46. putting in blast, 63. reactions, 7, 64-67. Rachette, 60. size, 46, 58-60, 68, 6g. Saxony, tin ore reduction, 236. Stuckofen, 45. temperature, 6g, 70. tin ore reduction, 235, 236. water supply, 81. Blazing off, tempering, 206, 207. Blende, oxidized by roasting ores, 54. zinc, 240, 246. Blistered copper, 219-222. Blister steel, 137-140. classification, 134, 136. Block tin, 235, 236, 238. Bloomary furnace, 45-47, 83-85. Blooming rolls, iii, 112. Bloom iron, classification, 134, 136. Blotting paper, 693. Blowing machinery, 76. Blue carbonate of copper, 217. Blue metal, copper, 222. Blue prints, 692. Blue stones, transverse strength, 575. Blue trap-rock in hydraulic cements, 583 Boats, paper, 693. Body bricks, 579. Bog iron ore, 51. Bohemian tin, 237. Boiler plate. See Plate iron. Boiler. See Steam boiler. Boiling process for iron, 90-105. tin refining, 236. BoUey, compositions of brasses, 296-299, bronzes, 282-287. Bolton, prices of metals, 268. Bolts and nuts, strength, 379-408, 406. shearing resistance, 405—408, 430. Bond's brickwork, 594, 595. Borax, as a flux, 8. for wood preservation, 676, 684. Bornite, 217. Bottoms, copper, 222. Boucherie, wood preservative process, 682-684. Bowline, 6g6. Bowling iron, analysis, 188. Boxwood, 633. crushing resistance, 638. tenacity, 637. uses, 659, 660. Bramwell, effect of fusion on strength of cast iron, 377. Brads, holding power, size, weights, 667. Brard, test of resistance of stones to frost, 576. Brash wood, 603. Brazing copper, 230. INDEX. ;o5 Brass, 38, 270, 288-299. agne, 320. ancient, 4, 41. -antimony alloy, 333. area reduction under tensile stress, 461, 462. Bath metal, composition, 298. Bearing's composition, 296. Bristol metal, composition, 297, 298. See Bronze, bronzing liquids, 336. burnishing, 294. button metal, 299, 300, cartridge metal, 457. casting, 316-323. fluxing, 463. castings, 322, 323. chrysorin, composition, 297. classification, .289. color, 288, 289, 291, 292, 296-299. colored by hydrochloric acid, 292. compositions, 288-299, 329, 33°- compressed, composition, specific grav- ity, 282, 283. compression hardening, 292. compressive resistance, 461, 462. See Copper. deflection, 461-4O3, 560, 561. deposition by electrolysis, 291. deterioration with age, 292. drawing, 294. ductility, 289, 296-299, 461, 462. Dutch, 297. early experiments, 457-459. elasticity moduli, 458, 459, 461-463. elastic limits depression by strain, 557, 558. tensile, 461, 462. torsional, 461, 462. transverse, 461, 462, 560, 561. variable, 557, 558, 560, 561. electric conductivity, 18, 296-299. electrolytic deposition, 291. expansion by heat, 30, 33, 34, 291. flow under stress, 292. forgings, compositions, 297, 298. foundry, 317, 323. Brass, furnaces, 318-320. fusibility, 289, 296-299. See Melting point. German, properties, 298. "gold" leaf, composition, 297. hardness, 289, 296-299. heat conductivity, 18, 20, 296-299. heat expansion, 30, 33, 34, 291. iron alloy, 177, 302. See Kalchoids. malleability, 289, 296-299. mallet classification, 289. melting, 317-321. See Fusibility. point, 35, 291. Muntz metal, 290, 298. strength affected by heat, 530, 531. odor, 24. oreide, composition, 296, 297. oxidation, effect upon strength, 539. increased by heat, 538. pinchbeck, composition, 296, 330. plate, shearing resistance, 463. powder, composition, specific gravity, 297-298. properties, 288-299. resilience, 461, 462. resistance, fluctuation under stress, 547, 548. interrupted strain, 557, 558. maximum, 562. oxidation, effect, 539. variable with time, 540, 548, 551. rolled, 292, 293, 297, 299, 300, 316, 319. 43°- Roman, 271. liomilly, 290. safety loads, 562. shearing, 430. plate. See Rolled, sheathing, composition, 297, 298, 330. shrinkage, 2gi, 317. solder, 298, 299, 331. solubility, 292. specific gravity, 22, 282, 283, 289, 296-299, 461. heat, 28. 7o6 INDEX. Brass, spinning, 294. squirting, 292. stamping, 316. sterro metal, 302, 458, 465. stiffness, 461, 462. structure, 289, 296-299. taste, 24. tempering, 24. temperature, effect, 291, 534. tenacity, 289, 296-299, 457, 461, 462. formulse, 476. test pieces, U. S. Board, 459, 460. Thurston's tests, 459-463, 560, 561. tombac, 289, 296, 297. tools to cut, 208. torsional tests, 461, 462. transverse tests, 403, 461-463, 560, 561. tubes manufacture, 294, 317. tutenag, composition, 330. uses, 290, 293. Vienna gold leaf composition, 297. watchmakers' properties, 299. weight, 21, 22. See Specific gravity, wire, 293, 298, 319, 330. drawing, 316. formula for elasticity modulus, 532, 533- work, Aitken's classification, 315. working, 293, 294, 315-322. See Bronze working, white, 463. button metal, 299. See Manganese bronze, whitened by ammonia, 292. yellow for brazing, 328. See Zinc. Braziers' sheets, copper, 231. Brazing, 328. Breaking down rolls, in, 112. Bricks, 578-582. compressive resistance, 573, 580. fire, resistance, 572, 581, 582. Brick-work, 594, 595. Bridge-plate. See Plate iron. Bright iron, 82, 179. Brine, as a tempering bath, 200, 206. Bristol metal (brass) composition, 297, 298. Britannia metal, 239, 312, 313, 330. British Board of Trade, rule for strength of cylindrical furnaces, 394. early metal working, 42, 44. oak tree, 624. tin, 237, 239. Brittleness. See Ductility, Malleability, and Resilience. Bronze, 38, 233, 244, 270-287, 451-455. age, 4, 5- aluminium, 306-309. electrical conductivity, 19. analyses, 452, 453. ancient, 41, 272-274, 277. annealed, 275, 282, 284. annealing, 279, 536-538. antimony alloy, 333-335. area reduction under tensile stress, 452. chilled bronze, 535, 536. art castings, 322-324. See Statuary. Asiatic, ii. bearings, 279, 283, 284, 287, 330, 332, 335, 44S. bell-metal, 276, 278, 284. See Brass. brittle, 285. casting, 322-324, 463. Chinese gongs, composition, 330. coin, 277, 283, 330. colors, 2S2-2S7. composition, 282-287, 330. compressive tests, 452, 453. See Copper. cast, a cire perdue, 11. cymbals, 330. deflection, 452, 453, 545, 546,»553- 556. dense properties, 203, 285. ductility, 282-287, 455- torsional, 452-454, 545, 546. early, 4, 5, 10-12. electric conductivity, 282-287. INDEX. 707 Bronze, elasticity moduli, 449, 463. effect of temperature, 532, 533. tratisverse, 452. elastic limits, 452-453, 545. elongations, 452-453. expansion by heat, 30, 33. Farquharson's (kalchoid), 487. fluctuation of resistances under stress, 547, 548. ' fluxing, 463. by zinc, 321. forged compositions, 282. See Malleability. French, 11. fusibility, 276, 282-287. gun metal. See Ordnance. Greek, 11. gongs, Chinese, 278. hammered, 278, 279. hammers for finished work, 333. hard, 285. hardness, 276, 282-287, 330. Haswell's table of compositions, 330. heat. See Fusibility. conductivity, 282-287. elasticity affected by, 532, 553. expansion, 30, 33. strength affected by, 494, 529-531, 534- hydraulic work, 331. influence of presence of other metals, 321, 322. iron alloy, 177, 302, 334. jewellers' punches, composition, 284. See Kalchoids. lead, presence of, 321. liquation, 278, 283, 285. malleability, 276, 278, 279, 282-287. malleable, 282, 455. machine, 279, 282, 2S3, 331-335- manganese, 281, 303-306, 450. maximum, 284, 482-488. medal, 283. melting points. See Fusibility, ordnance, 277, 278, 280, 281. casting, 281, 535, 536. compositions, 283, 330. Bronze, ordnance, elasticity, 458. maximum stress, 562. oxidation, 277, 278. effect upon strength, 539. repeated strain, effect, 561. safety factors, 562. ' specific gravity, 22, 448. tensile tests, 448, 449, 458, 481. paints, 336. patina, 278. See Pewter, phosphor-bronze, 274, 281, 381, 531.. 561. powder (brass), 297, 298. properties, 282-287. red, 330. resilience, 452, 453. resistances, 282-287, 448-455- fluctuation under stress, 547, 548. heat effect of, 529-531. intermittent stress, effect of, 557. oxidation, effects, 539. temperature, effect, 532-536. Roman, 11, 273. sheathing, 276. silicon, 20. solders, 291, 331. solubility, 282. specific gravity, 22, 276, 280, 282-287, 452. bearings, 283, 284. bell metal, 284. malleable bronze, 284-286. ordnance, 448. powder (brass), 297, 298. tempered bronze, 282. speculum metal, 276, 277, 279, 285, 330. statuary, 278, 322-324, 330. strength. See Resistance, strongest, 284, 482-488. structure, 282-287. temperature, effect upon elasticity, 532, 533- strength, 534-536- tenacity, 494. temper of tools to cut bronze, 208. 7o8 INDEX. Bronze, tempered, specific gravity, 282. tempering, 24, 275, 279, 536-538- tensile tests, 276, 282-287, 451-455- area reduction, 452. elongations, 452, 453. formulae, 474, 476. ordnance bronze, 448, 449, 458, 481. strain diagrams, 454. temperature, effect of, 494- See Thurston's tests, 451-453- See Tin. torsional tests, 451-454, 545, 546. transverse tests, 451-453, 545, 546, 553-556- working, effect of, 537, 538. See Brass working, wire (phosphor-bronze)^ 281. zinc, as a flux, 321. Bronzes, art castings, 278, 322-324, 330. Siris, 12. Bronzing liquids ; process, 335, 336. Brown hematite, 51, 52, 67. mineral paints, 674. Burgundy pitch, 613. Bush-hammered stones, 591. Button brass (kalchoid), 300- Bulb iron, 118- Bunsen. See Durocher and Bunsen. ■' Burden " of blast furnace, 63- Burden's rotary squeezer, no. Burnishing brass, 294. Burnett wood preservative process, 676, 677- , Button sheet brass, 299. C. Cabbage palm tree, longevity, 598. Cabinet-work, woods, 659, 660, 652, 663. Cables, 695. Cable wire, effect of time and use, 511. Cadmium, 266. atomic weight, 29. cost, 269. elasticity modulus, 447. electric conductivity, 19, 20. hardness, 17. Cadmium, latent heat, 37. melting point, 36. specific gravity, 22. heat, 29. sulphide, 266. Cairns, analysis of meteoric iron, 52, 19a Calamine, 240, 241, 246. Calcareous rocks, 564, 569. Calcination, 54, 57. copper ores, 220. iron ores, 535. zinc ores, 241. See Roasting. Calcium, 266. cost, 268. fluoride, 8, 254, 373. in meteoric iron, igo. Calender rolls of paper, 693. California spruce wood, crushing resist- ance, 639. tenacity, 637. transverse strength, 647. Calomel, 259. Calvert and Johnson, brasses, investi- gations, 296-299. bronzes, investigations, 282-287. oxidation of iron, 211, 212. platinum, heat expansion, 261. Camphor wood, 633. Cam squeezer, 109. Canada elm tree, 627. Canadian oak wood, transverse strength, 647- red pine, 608, 609. Cannon. See Ordnance. Caoutchouc, 693, 694. belts, 693. vulcanized, 694. Capacity for heat. See Specific heat. Carbon, absorption in copper refining, 223. • in annealed and unannealed steels 497, 498. anthracite iron, 181, 182. armor plate, Lowmoor, 187. arsenic ore reduction, 264. Barrow ingot iron, 195. INDEX. 709 Carbon, Bessemer, 194-196. in boiler plate, Neuberg-Bessemer, 195. in Bowling iron, i88. bright iron, 179. cast iron, 168, 196. Barrow ingot, 195, foundry, 178, 180-182. malleableized, 190. ordnance, 375. S. Oural mountains, 178. tenacity, effect of carbon, 369, 371, 373- cementation, change in proportions by, 191. chain iron, 188, 189. charcoal iron, 181, 182. chilling iron, 180. chisel temper steel, 194. coke iron, iSi, 182. cold blast irons, 181. compression, change in proportion by; 210. copper refining, absorption in, 223, condition in different steels, 497, 498. crucible steels, 191, 192. die-temper steels, 194. ductility, affected by, 197. electric conductivity, affected by, 197. foundry irons, 178, 180-182. German Bessemer steel rails, 195. heat effects in steel, 204. hot blast irons, 181. irons. See Cast iron, iron, l58. Lowmoor armor-plate, 187. malleable iron, 187. malleableized iron, 190. meteoric iron, 190. mottled, 179. Neuberg Bessemer steel boiler plate, 195. ordnance iron, 375. percentages in irons and steel, 133, 135. 136, 191, 210. See all entries under Carbon, plate iron, Lowmoor armor, 187. Carbon, Neuberg, Bessemer, 195. puddled iron at different stages, 182- 186. rail steel, 195, 196. South Oural iron, 187. Swedish iron, 187. steels, 168, 172, 193. annealed and unannealed, 497, 498. Bessemer, 194-196. cementation, 137-141, 191. chisel-temper, 194. crucible, 191-193. die-temper, 194. ductility, affected by, 197. hardening effects, 497, 498. heat effects, 204. nijalleability affected by, 197. percentages, 133, 135, 136, 191-193, 195, 196, 497, 498- tempered, 193-195. tenacity affected by, 369-371, 373. unannealed, 497, 498. weld iron, 187. white iron, 179. Carbonate of copper 217, 219. iron, 51, 52. lead, 248, 252. lime, flux for silica, 9. hardness, 573. reduction processes, 7. soda, reduction of nickel ores, 254. zinc, 5, 240. Carbonic acid driven from ores by wast- ing, 54- with oxygen, cause of rust, 211. in sidenite, 52. Carboniferous limestones, iron ores, 50. Carburetted cobalt, specific heat, 27. nickel, specific heat, 27. Carondelet zinc ores, 246, 659. Carpentry, woods for, 659. Cartridge metal, brass, 457. Carving, wood for, 659. Case-hardening, 137-139, Cassiterite, 234. Cast copper, 227, 232, 233. Casting, a cire perdue, 11. 7ro INDEX. Casting, brass, 316-323. chill, ordnance brass, 535, 536. house, 71. rediscovered in England, 47. Castings, 126, 127. aluminium bronze, 308, 309. classified, 311. compression of, 168. decarbonizing, 127, 136, igo. French, 11. Greek, 11. Italian, 11. malleableized, 127, 136, igo. manganese bronze, 303-306. Roman, 11. shrinkage, igS. Spence's metal, 314. tin-zinc alloys, 311. zinc, 244, 245. Cast iron, 39, 82. analyses. See Composition. appearance under microscope, 505-507. See Blast furnace. boiling process, go-105. bright iron, 82, 179. calcination ores, 54, 57, 535. See Carbon. car-wheel, tensile tests, 375. .Si?^ castings. chilled castings, 180. classification, 132-136. coke iron, 181, 182. cold blast, 181, 182. columns, strength, 398-402. composition, 168, 172-182, igo, ig6. See Carbon. South Oural Mountains, 178. compressive resistances, 374, 375, Hodgkinson, 389-3g2, 438. corrosion, 171, 210-214. crystal, 504-505. cylinders, strength, 79, 384-388. decarbonizing, irons used, 88. density. See Specific gravity, relative to tenacity, 38g. ductility, 390, 391. elasticity, 374, 392. Cast iron, elasticity moduli, 347, 374. elastic limits, 367, 375, 378. elongation. See Tensile resistance, expansion, 30, 34, ig8, igg. flux. See Blast furnace. See Forge irons, foundry iron, analyses, 180—182. carbon in, 180-182. expansion coefficients, 199. grades, 82. properties, 178, I7g. specific gravities, igg. fuel, 57. furnace. See Blast furnace, fusion, effect on density and strength 377- gray irons, 179, 182. expansion of coefficients, 199. gray irons, fuel, effect of, 58. fusibility, 198. grades, characteristics, 82. specific gravity, igg. heat, ig8. yield in roasting, 54. girders, 82, 83. hardness, 16, igg. heat, effect upon strength, 4g4, 4gs. expansion, 30, 34, igg. Hodgkinson, compression of long bars, 391. 392. hot blast, 47, 4g, 71-76, 181. See Ingot irons. See Iron, malleableized, 88, 127, 136, 190. torsional tests, opp. 415, 425, 426. manufacture, 46, 53-86. melting, effect on density and strength, 377- melting point, 35, 36, 198. mottled iron, 179. nomenclature, 132-136. « non-malleable, 375. ordnance, 171, 210-214. ores, 9, 51-55, 57, 64-67. reduction, 53-70, 82-86. oxidation, 171, 210-214. pipes, strength, 374. INDEX, 711 Cast iron. See Pipes, production, 53-86. proof stress, 374. puddled. See Carbon. See Puddling. remelting, strength increased, 377. resistances, continuous loads, 512, 513. size, effect, 376. roasting ores, 54-57. 535- Salisbury iron tests, 375, 378. sets, 375, 378, 390, 391. shearing resistance, 405, 406. shrinkage, 198, 291, 317. in cooling, 198. silicon in, 180-182, 188, 189, 196. size, affecting strength, 376. South Oural Mountains, 178. specific gravity, 23, 199, 374-377, 389. affected by fusion, 377. ordnance, 375. relation to strength, 376. specific heat, 27, ig8, Igg. See Steel. strain diagrams, opp. 415, 494, 495. Salisbury iron, 378. strength. See Resistance. affected by heat, 377, 494-496. temperature, effect upon strength, 494- 495- tensile resistances, 15, 374-376, 389. carbon effects, 369, 371, 373. car-wheel iron, 375. ordnance, 375- related to density, 389. tools to cut, temper, 208. torsional tests, 374, opp. 415, 494, 495. transverse resistance, 402-405. weight. See Specific gravity, white, 90. characteristics, 82. classification, 133. fuel, influence of, 58. fusibility, ig8. properties, 179. specific heat, 198, igg. Cast steel. See Crucible steel, appearance under microscope, 506. Cast steel. See Temper. Cast zinc, 45, 46, 53, 83, 244, 245. Catalan forge, 45, 46, 53, 83. See Bloomary furnace. See Charcoal blast furnace. See Furnace. furnace iron, classified, 134. Caulking tools, 208. Cavalo zinc ore, 242. Cedar trees, 615-617. Lebanon, longevity, 59S. wood, American, shearing resistance, 646. Bermuda, tenacity, 637. crushing resistance, 638. deflection of beams, 651. durability, wet and dry, 659. Guadaloupe, tenacity, 637. red, torsional resistance, 655, 656, uses, 660. West Indian, transverse tests, 647. white, shearing resistance, 646. Cement, 572, 573, 582-588. belt cement, 691. bituminous, 588. hydraulic, 572, 573, 583-585. Portland, 573, 584. puzzolana, 573, 583. Roman, 573. trap-rock, 583. trass, 583. lime, 571, 583. Cementation steel, changes in carbon proportions, 191. See Crucible steel. Cerium, 267. cost, 268. fusibility, 36. Chain iron, analyses, 188. Chalcopyrite, 217. Chalk, 569. flux in nickel ore reduction, 254. Chaney, table coefficients of expansion, 33- Channel iron, 117. columns, 401. Chantrey's alloy (kalchoid), 302, 712 INDEX. Charcoal, blast furnace, 57, 58, 62. See Furnace, irons, 181, 182. boiler plate, 119. specific heat, weight, 67. Charring for wood preservation, 675. Chaudet. medal bronze, 277. See Carbon. Chemical characteristics, 277. combinations of metal ores, 178. processes in ore reduction, 7, 64-67. properties of irons and steels, 170-214. puddling changes, 182-186, 187. reactions in Bessemer converter, 165. structure of iron ores, 52. See Microscopical appearance. Chenhall, analyses of zinc ores, 242. Cherry bricks, 5gg. tree, 629. wood, crushing resistance, 638. deflection, 651. shearing resistance, 646. transverse resistance, 647. uses, 659, 660. Chestnut trees, 625. white oak, 624. wood, columns, 641. crushing resistance, 638. deflection, 651. durability, dry construction, 659. elasticity, 636, 651, 659. elastic uses, 659. for furniture, 660, hold upon spikes, 668. house building, 660. shearing, 645, 646. spikes, hold upon, 668. tenacity, resistance, 637. torsional coefiicients, 655-656. transverse modulus of rupture, 647. uses, 659, 660. Chilled castings, 180. ordnance bronze, 535, 536. cast iron, temper tools to cut, 200. rolls, 112. Chinese gongs, 278, 284. silver, composition, 330 Chinese gongs, white copper, composi« tion, 330. Chipping tests, tool steels, 374. Chisel temper, 194, 208. Chloride of manganese, 265. mercury, 259. ore reduction processes, 7. produced by use of salt, 54. of zinc, wood preservative, 676, 677, 685, 686. Chlorite, hardness, 573. Chrisocalle, kalchoid, 300. Chrome steel, forging, 207. hardening, 208. Chromium, cost, 269. fusibility, 36. hardness, 17. in meteoric iron, igo. Chrysorin, brass, composition, 297. Cinder of puddling furnace, 95. Cinnabar, mercury, 259. Circular plate bolted at edge, strength, 405. boiler heads, strength of, 382, 383. Clark, cylinder, strength of, 388. lead pipe, strength of, 445. stayed plates, strength of, 381, 382. Classification of irons, 132-136. ores, 51. metals, 6. steels, 132-136. Clay. See Alumina silicate. for bricks, 578, 579. as a flux, 9. slate, 571. direct production of wrought iron, 47. effect of heating and re-working iron, 368, 369. Clemendot, tempering by compression, 2:0. Clove hitch, 696. , Clouding, 33g. See Bronzing, Lacquering. Coal for blast furnace, anthracite, 47, 57, 59, 62, 63, 67, 75. bituminous, 57, 67, 8i, 182, charcoal, 57, 58, 62. INDEX. 713 Coal for blast furnace, coke, 49, 59, 62, 63, 67, 70. Coal tar in bituminous cements, 588. as paint, 674. Coarse copper, 219. metal, copper, 222. Coating, zinc, 245. Cobalt, alloyed with iron and steel, 177. amorphic, nearly, 27. carburetted, specific heat, 27. in chain irons, 188. cost, 269. fusibility, 36. hardness, 17. in iron, 189. cast from So. Oural Mts., 178. Lowmoor armor plate, 187. meteorio, 190. specific gravity, 72. heat, 27. in steels, 193. Cocoanut palm tree, longevity, 598. Cogs, woods for, 660, 661. Cohesion. See Tensile. Coin, aluminium bronze, 308. bronzes, 277. copper, with gold and silver, 38. -nickel, 309, 310. nickel, 255. Coignet, Beton-Coignet, 586, 587. Coke in blast furnaces, 49, 59, 62, 63, 70. specific gravity, specific heat, weight, 67. irons, 181, 1S2. Cold, effect upon tin, 237. blast irons, 181, 182. chisels, temper, 208. See Heat. pot, tin plate manufacture, 130. rolled iron, torsional tests, opp. 415, 421. rolling tin plates, 129, 130. -sheets, tin casting, 437. See Temperature. Collared rolls, 115. Color bronzes, 282-287. Colophony, 618. Colors for iron ores, 51. for paints, 675. of tempered steels, 205-209. of woods, 662, 663. Columbrium or niobium, 267. Columns, 640, 641. ash, 641. beech, 641. cast iron long bar, 391, 392. strength, 398-402. channel iron, 401. chestnut, 641. See Compressive. cylindrical strength, 397. elm, 641. flexure, 395-405. formulse, 396, 398, 400, 401, 641. Gordon's formula, 398, 399. Hodgkinson's formula, 397, 398. hollow, strength, 397-399. iron and wood, 641. mahogany, 641. oak, 641. Phoenix, strength, 397, 398. pine, 639, 641. posts, 395-405. Rankine's formulae, 398, 399. rectangular, strength, 398. square, strength, 397. strength, 394-405. struts, 394-405, 644. teak, 641. Tredgold's formulae, 398, 400. wood, bored for air access, 644. and iron, 641. wrought iron, 397, 398, 402. Combining number. See Atomic weight. Combustibility of maijnesium, 262. tin, 237. Compound steels, 135, 144. Compressed brass, 292. bronze, 282, 283. ingot irons and steels, 168, 169. Compression, tempering, 210. Compressive resistance aluminium- bronze, 307-309. 714 INDEX. Compressive resistance ash wood, 638. brasses, 461, 462. brick, 573, 580. bronzes, 452, 453. carbon, proportion in steel effects of, 210. cast iron, 374, 375, 389-392, 438. See Columns. copper, 430, 431, 446, 452, 461. cylinders, 393, 394. fluxes, 382-388, 393, 394. See Hardness. lead, 446. shearing, 374, 392. spheres, 387, 388. See Steels. stones, 573, 574, 585. struts, 394-405. tin, 438, 446, 452. velocity effects, 432. woods, 638-646. detrusion by nails and screws, 667- 669. wrought iron, 392, 393. zinc, 442, 443, 462. Concrete, beton, 572, 585, 586. bituminous, 588. cement and mortar, 572, 573, 582- 588. effect of heat upon, 572. strength, 587. Condenser, copper ore reduction, 221. Conductivity. See El|ctric conductivity. See Heat conductivity. Conglomerates, resistance to heat, 577. Coniferas, trees, 607. Connecting-rod bearings, bronze-anti- mony, 335. kalchoid, 332. Constantino zinc ore, analysis, 242. Continuous mill, 115. Contraction. See Expansion. Converters, Bessemer, 156-160. Converting furnace, cementation steel, 137-139- Conversions, reduction processes, 7. Cooling of iron in rolling, 116. Coping, masonry, 594. Copper, 216, 226, 227, 228, 461. absorption of carbon in refining, 223. alloys, 30, 233, 270-310. ancient, 271, 272-274. crystalline form, 26. Prick's, 310. iron, 177. heat increasing oxidation, 538. iron and tin, 302, 303, 334. nickel, 309, 310. nickel-zinc, 310. tutenag, 310. white copper, 310. packfong, 310. phosphorus, 227. alloy and flux, 233. steel tutenag, 310. tin. See Bronze. tin-zinc. See Kalchoids. white copper, 310. zinc. See Brass, amorphous, nearly, 27. analyses, 226, 228, 431-436. ancient alloys, 10, 271, 272-279. annealed, 227, 229. area reduction under tensile stress, 452, 461. atomic weight, 29, 446. bar, 232, 233. bearing alloys, 334. See Alloys. Bessemer steel rails containing copper, 195- black, 223. oxide, 217. blister, 2ig, 222. blue carbonate, 217. metal, 222. bolts, shearing resistance, 430. bornite, 217. ► bottoms, 222. See Brass. braziers' sheets, 231. brazing, 230. See Bronze, bronzing liquids, 335-337. INDEX. ;is Copper, calcining, 220. carbon, absorption in refining, 223. carbonate blue, red, 217. green, 217. reduction, 219. cast, 227, 232, 233. strain diagrams, opp. p. 415. castings, tenacity, 232. cast iron containing copper, 178. chain iron containing copper, 188, 189. chalcopyrite, 217. Chinese white, 330. coarse, 219. commercial, 222. compressive resistance, 430, 431, 446, 452, 461. condenser, ore reduction, 221. consumption, 233. contraction. See Expansion, cost, 269. crystallization, 26. cuprite, 217. deflection. See Transverse tests, density. See Specific gravity, deposition, electrolytic, 227, 231. deterioration virith age, 229, 230. Douglas (Hunt and) method of ex- traction, 224. ductility, 23, 24, 226-229, ^82, 296. torsional, 452, 461. early vforking, 5. elasticity moduli, 433, 434, 436. heat effects, 533. torsional, 436, 452, 461. transverse, 452, 461. variations with temperature, 532, 533- elastic limit under tensile stress, 436, 452, 461. torsional stress, 436, 452, 461. transverse stress, 452, 461. electro conductivity, 18-20, 26, 227, 231, 282, 296. presence of iron, 322. electric plating, 227, 231. elongation. See Tensile tests, torsional tests. Copper, European, 227. expansion by heat, 30, 33, 34. extraction, Hunt and Douglas pro- cess, 224. wet process, 223. forged. See Hammered, fine metal, 220. fire box, tube sheets, 232. fluxed and alloyed with phosphorus, 233, 429. Frick's alloy, 310. fuel and ore reduction, 220. furnaces for ore reduction, 220-225. green carbonate, 217. hammer, 232, 233. hammered, 226, 229, 230, 232, 233. hardened, 232, 233. hardened by hammer, 232, 233. hardness, 16, 17, 215, 232, 233, 282, 296. heat effects, 494, 495, 532, 534. expansion, 30, 33, 34. conductivity, 17, 26, 232, 282, 286. strength affected by, 494, 495, 528, 529. 532, 534- See Fusibility, melting point. oxidation increased by heat, 538. strength affected by heat, 494, 495, 528, 529, 532. in hematite, red, 181. Hunt and Douglas process, 224. impact effect, 432. in iron chain, 188, 189. from red hematite, 181. Swedish, 187. iron alloy, 177. electric conductivity, 322. iron zinc alloy, 302. Japanese, 227. See Kalchoids. lead in sheet copper, 223. whit^, injured by, 322. Lake Superior, 227. malachite, 217. malleability, 24, 25, 226-230, 282, 296, See Duclility. matt ore reduction, 221, 222. 7i6 INDEX. Copper, maximum stresses, 562. melaconite, 217. melting point, 26, 35, 36, 215, 282, 296. metal, coarse, fine, 222. metallic, 216, 226, 227. in middle ages, 10. mixing in ore reduction, 220, 221. native. 216. nickel, alloy and coins, 309, 310. nickel and zinc alloys, 310. ores, 6, 216-219. oxides, black, red, 217, 219. reduction, 219-225. Parnell process, 242. sulphides, 2ig. oxidation, 225, 226. effect upon strength, 539. increased by heat, 538. oxides, black, red, 217. oxygen absorption, 225, 226. packfong, 310. phosphorus alloy and flux, 227, 233. tenacity, 429. pig, 223. pipes, 229, 233. plating, 123, 227, 231. processes, 219-225. production, 28. properties, 215. purification, 219. purple ore, 217. pyrites, 217. , reduction. See Ore reduction, refining, 212, 222, 223. regulus, 219, 221, 222. resilience, 452, 461. resistance deterioration with time, 229, 230. maximum, 562. oxidation effects, 538, 539. shearing, 430. time deterioration, 229, 230. rolled, 223, 229-233, 296. rust. See Oxidation, safety factors, 342, 562. shearing resistance, 430. Copper, sheathing, 231, 232. sheet, 223, 229-232, 296. slag, 222, 223. smelting, 218—221. solder, 230, 231. sound conduction, 446. South American, 227. specific gravity, 22, 26, 215, 216, 227, 274, 282, 292, 446, 452, 461. specific heat, 26, 29. at different temperatures, 27. in steel, 177, 193. structure, 282, 296. sulphates for wood preservation, 676, 680, 684-686. sulphide, 219. supply, 218. in Swedish iron, ]87. Swiss, 228. symbol, 215. temperature. See Heat, tempering, 24, 227. temper of tools to cut, 208. tensile resistances, 15, 226, 227, 232, 282, 296, 429, 436, 446, 453, 461, 528, 529. area reduction, 452, 461. elastic limit, 436, 452, 461. elongation, 436, 452, 461. variations with temperature, 529. wire annealed and unannealed, 481. tile, 222. See Copper-tin alloys. See Copper-tin-zinc alloys, tin, effect of presence, 16. tinned, 227. tools to cut copper, 208. torsional tests, opp. 415, 435, 436, 452, 461, 540. toughening, tough cake, 222. transverse tests, 432, 433, 436, 452, 461, 463, 540, 545, 546. tubes, 229, 233. tutenag, 310. volatilization, 35, 215. weight. See Specific gravity, western, 227. IMDEX. 717 Copper, white, 309, 310. Chinese, 330. white lead, injured by, 322. , wire, 226, 232, 233. working, 5, 226, 229-233, 537, 538. in wrought iron, 177. yield, 217. zinc-iron alloy, 302. Copperas, as a dryer for paints, 671. Coppered wire, 123. Copper-tin alloys. See Bronze, -zinc alloys. See Kalchoids. -zinc alloys. See Brass, -iron alloy, 302. Cordage, 695-697. Core, 10, II, 126. box, 126. See Casting. print, 126. Corliss, oil-seasoning wood, 601, 635. Cornwall tin, 12, 239. Corrosion of iron and steel, 210-214. See Oxidation. Corrugated flues, 394. Cort, iron processes, 48, 92, ill. Cost of metals, 268, 269. Coursed rubble, 592. Cowles' process, aluminium production, 257, 269. Cowper's hot blast stove, 74. Crab-tree wood for cogs, 660. Cracks in timber, 605. Crandalled stones, 590. Crane used anthracite in' blast furnace, 47. Crank bearings, bronze-antimony alloy, 335. Creosote for wood preservation, 676-679, 686. Crompton rotating puddler, 100. Cronstedt discovered nickel, 15. Croockewit, specific gravities of brasses, 297, 298. bronzes, 285-287. Cross-section ; iron columns, 401. Crushing resistance. See Compressive resistance, stones, 573, 574- Crushing resistance woods, 638-646. Crucible, for steels, 141, 142. Crucible steel, 141-151. analyses, 191-193. blister steel, 137-140. castings, 145. cast steel, 140-151, 494, 506. cementation steel, 137-141, 19I. classification, 134, 136. compounds, 144. elongation, 372. furnace, 142, 143. manufacture, 136. microscopic appearance, 506. open hearth steel, 145, 149, opp. 415, 520. properties, 144, 145. temper, 144, 193. temperature effects, 494. tensile resistance, 372, 373. See Tool steels. torsional tests, opp. 415, 423-425. uses, 144, 145. Wootz or Indian, 145. Crushing. See Compressive. Cryolite a source of aluminium, 257. Crystallization alloys, 26. bismuth, 253. copper ores, 217. electric deposition, 26. by slow cooling, 321. See Fracture. iron, 197, 504. ore, 51. metals, 26, 27, 500-507. See Molecular changes. Cupola furnaces for Bessemer plant, 154. for copper ore, 2l8, 224, 225. Cuprite, 217. Cup-shakes in timber, 605. Curves of resistance. See Strain-dia- grams. Cut stones, 589, 591. Cutting, H. A., heat resistances, moisture, absorptions, and specific gravities of stones, 577. Cutting tools, tempers, 208. 7i8 INDEX. Cylinders, cast iron, strength, 79, 384- 388, 394- compressive resistance, 393, 394. ■ wrought iron, 382-384, 393, 394. Cylindrical boiler shells, strength, 382- 384- furnaces, strength, 394. Cymbal bronze composition, 330. Cymophane, hardness, 573. Cypress tree, 611, 612. longevity, 598. wood, 612, 637. D. Dakota tin, 239. Dank's puddling furnace, 98-100. Dantzic oak columns, strength, 641. Darby, early iron manufacture, 46, 47. D'Arcet, bronze investigation, 275, 302, 537. Dean, bronze investigation, 282-287. Dead load, definition, 34. Dead oil for wood preservation, 676-679, 686. De Bonneville, hardening steel springs, 205. Debray. See Deville and Debray, plati- num manufacture, 261. DecandoUe, longevity of trees, 598. Decarbonization. See Pneumatic process. Decay of timber. See Timber decay. See Wood preservation. Decarbonizing castings, 127, 128. process, wrought iron manufacture, 87. Deccan iron furnace, 45. Decomposition, reduction process, 7- Deflection. See Transverse tests. Delhi, pillar of iron, 43. D' Elhuj art Bros. , preparation of tungsten . I5-' Deliquescent salts, 683. Density of cast iron affected by fusion and remelting, 377. relation to compressive resistance, 389. tensile resistance, 377. steel modified by heat, 204. Dephosphorizing processes, 165. See Basic-dephosphorizing. See Phosphorus. Depretz's chrisocalle, Kalchoid, 300. iron bronze ordnance metal, 302. DeRosthorn, tests of sterro-metal, 458. Designing machinery, 126. Despretz, heat conductivities of metals, 17- Detrusion. See Compressive resistance. See Shearing resistance. woods, 644-646. Deville, aluminium manufacture, 257. and Debray, platinum manufacture, 261. Diamagnetic, bismuth, 253. Diamond, hardness, 573. paneled stones, 591. Diaz, bronze weapons of Aztecs, 273. Didymium, 267. Die-plates wire drawing, 120. Die-temper, 194, 208. Direct process, iron manufacture, 45-47 ; 83-85. See Bloomary. Distillation. See Volatilization. Dogwood, shearing resistance, 646. Dolomite, 569-570. Double refined iron, 120. Douglas, copper furnaces, 225. copper smelting methods, 218. Dovetailed joints, 665. Dragon's blood tree, longevity, 598. Drawing paper, 692. temper of steels, 205, 207. Drawn brass, 294. Drawn plates, wire drawing, 120. Drilling tests of tool steels, 374. Drills, temper, 208. Dryers for paints, 671, 675. Dry puddling, 7, 90, 92. » See Puddling. Dry rot, 603, 606. Dyeing wood Boucherie process, 683. Ductile brass, composition, 298. Ductility, aluminium-bronze, 256, 258, 307-309. INDEX. m Ductility, brass, 289, 296-299, 461, 462. See Area reduction, bronzes, 282-287, 452, 455, 529-531, 535. 536. torsional, 452-454, 545, 546. phosphor-bronze, 281. modified by temperature, 530, 531. cable wire, effect of time and use, 511. carbon, effect upon iron and steels, 198. cast irons, 390, 391. chill-cast bronze, 535, 536. copper, 23, 24, 226-229, 282. 296. torsional, 452, 461. -tin-zinc. See Kalchoid. See Elongation, gold, 23, 25, 322. heat effects, 198, 530-536. on-kalchoids, 530, 531. impact, change produced by, 492, 493, 496- irons, carbon and manganese in, 197, 198. cast, 390, 391. . phosphorus, effect of presence, 322. See Wrought iron. Kalchoids, 471-478, 487. heat effects, 530, 531. lead, 25. magnesium, 262. See Malleability, manganese bronze, 303-306. metals, 23-25. Muntz's metal, varying with tempera- ture, 530, 531. nickel, 254. effect of phosphorus, 322. phosphor-bronze, 281. variations with temperature, 53a, 531- See Resilience, shock, 492, 493, 496. silver, 25. steel, 199. temperature. See Heat effects, time effects, 511, 512. See Tensile tests. Tin, 236, 237, 287, 452. Ductility. See Torsion tests. zinc, 25, 240, 243, 299, 462. Dudley, early iron working patents, 46. Dudley, Dr., composition of steel rails, 1 96. Dulong, expansion coefficient cast iron, 199. and Petit, specific heats of metals, 27, 28. variations with temperature, 30. Dumas coin bronzes, 277. density of vapor of mercury, 258. hardness of metals, 17. Durabihty of stones, 573-577. Durable woods, 659. Durocher and Bunsen, classification of stones, 564. Dussaussoy, ferrous bronze, 302. Dutch brass, 297. wainscot logs, markings, grain, 662. E. Early uses of metals, 1-5, 40-43. Earthy salts for fire-proof preparations, 683. Ebonite, vulcanized caoutchouc, 694. Ebony, green and black, 633. shearing resistance, 646. transverse modulus of rupture, 647. uses, 659. Eccentric straps, kalchoids for, 332. Edgemoor Iron Co. formula for tenacity of iron, 365. Eggleston, copper analyses, 228. deterioration with age, 230. electric conductivity, 228. tenacity, 429. Ehrenwerth rotating puddler, lOI. Egyptians, iron used by, 42, 43. Elasticity, 344, 345. aluminium bronze, 309. See Brass, Bronze. bridge plates, 373. cast iron, 374, 392. flexure, 516-520. See Resilience. 720 INDEX. Elasticity. See Torsional tests, Trans- verse tests. wood, effects of prolonged stress, 657, 658. moduli, 345, 347, 362, 368, 446, 463, 532, 533- aluminium, 309. ash, 636, 651. beech, 659. brass, 45S, 459, 461, 462. wire formula, 532, 533. bronze, 449, 452, 463. variations with temperature, 532, 533- ordnance, 458. cadmium, 447. cast iron, 347, 374, 375, 392. chestnut, 636. copper, 433, 434, 436, 4SZ, 461. variations with temperature, 532, 533- wire formula, 532, 533. See Elastic limits, elm, 647. See Elongation, gold, 447. variations with temperature, 533. heat effect, 532-534. iron, 446. formula, 532-533. wrought, 362. kalchoids, 480. lead, 446, 447. variations with'temperature, 533. lignumvitEe, 636, 659. ordnance bronze, 449. palladium, 447. variations with temperature, 533. pine, southern, 653. plate iron, 373. platinum, 447. spruce wood, 653. silver, 447. variations with temperature, 533. steel, variations with temperature, 533- teak, 636. Elasticity moduli, temperature effects, 532-534- tin, 439, 440, 441, 446, 452. torsional, 655, 656. willow, 636. wood, 653. wrought iron, 362. zinc, 446. Elastic limit, 346-349, 353-358, 442, 462. boiler plate, 364. brasses, 461, 462. depression, 557, 558. bronzes, 452, 453, 545. variations under test, 556, 561. cast iron, 347, 367, 374-376. Salisbury, 378. copper, 436, 452, 461. See Deflection. See Variation, depression, 556-558. tin class, 552. See Elasticity, elasticity moduli, elevation, 499, 557-560. See Elongation, ingot iron, 370. intermittent stress, effect, 552. See Variations, iron, 557-560. cast, 367, 375, 378. ingot, 370. wrought, 367. plate iron, 364. resilience formulae, 349-352. See Proof stress. See Set. steels, 367, 370. British, 392. tools, 374. See Tensile, Torsional, and Transverse tests, time variations. See Variations, tin, 438, 441. tensile and transverse tests, 452. variations with strain, 539, 558-561. time, 520-527, 556-561. tin class, 552. woods, 636. INDEX. •J 21 Elastic variations wrought iron, 367. zinc, 442, 462. woods, 659. Electric conductivity, 18, 19, 26, 256. alloys, 18, 19. aluminium, 18, 19, 26, 256. antimony, 26, 253. bismuth, 18, 26. brasses, 18, 296-299. bronze, 282-287. cadmium, Ig, 20. carbon presence, 197. copper, 18, 20, 26, 227, 231, 282, 296. injured by presence of iron, 322. heat effects, 197. iron, 19, 20, 26, 197. injury in copper, 322. wrought, 97. lead, 18-20, 26, 247. magnesium, Ig, 26. manganese, 26. metal, 6, 17, 18, 19, 26. pines, 636. platinum, 18, 20. silver, 18, 20, 26. steels, 199. thallium, 19, 20. tin, 18, 19, 287. wrought iron, 197. zinc, 29g. current. See Voltaic current, furnace, 257. Electro-negative, metalloids, metals, 6. plating with copper, 227, 231. Elevators. See Hoisting machinery. EUershausen puddling process, 98. Elliptical flues, strength, 394. Elm trees, 627. longevity, 598. wood, columns, strength, 641. crushing resistance, 638. deflection, 651. durability, 659. elasticity, 659. modulus, 647. elastic limit, 636. glued joints, strength, 66g. Elm wood, hold upon nails, screws and spikes, 668. tenacity, 637. toughness, 65g. uses, 65g, 660. Elongation, 343-352. See Ductility, Tensile tests. Eisner, volatilization of iron, 197. Emerald, hardness, 573. Enamel, tin, 237. Endogenous trees, 597. Engineer, knowledge required, 15. English bond, masonry, 594. Engraved plate alloy, 312. Engraving, wood for, 659. Epochs in history of iron trade, 49. Equations of resistance curves, 346, 347. Erbium, 266, 268. Euler, formula for columns, 640. Europe, early metal working, 4, 5. Eustis, iron in nickel ores, 255. Exaltation <3f elastic limits. See Elastic limit. Exogenous trees, 597. Expansion by heat, aluminium, 33. antimony, 253. bismuth, 253. brass, 30, 33, 34, 291. , bronze, 30, 33. cast iron, 30, 34, 198, 199. copper, 30, 33, 34. glass, 30. gold, 30, 33. iron, 30, 34. lead, 247. mercury, 258. nickel, 33. platinum, 30. -iridium alloy, 313. silver, 30, 33. steel, 30, 34, 199. tin-zinc alloy, 33. zinc, 30, 34. Exposure to weather, effect on metals 498-500. See Wood preservation. Extension. See Elongation. 722 INDEX. F. Factors of safety. See Safety factors. Faggoting, 48, 107. Fairbairn, cast iron strength, 377, 512, 513- continuous stress on cast iron, 512, 513- crystallization by shock, 504. cylinders and flues, strength, 393, 394, detrusion of oak, 646. I-beam, 404. temperature affecting strength of iron, 491, 492- time under stress, effect oa strength, 512, 513. Farquharson bronze (kakhoid), 487. electrolytic deposition of brass, 291. copper, 227. Feathered iron, 118. Feldspar in granite, 569. basalt, greenstone, and trap, 567. hardness, 573. Fenton's alloy, 334. Ferric oxide. See Hematite, Oxide of iron. Ferro-manganese in magnesium bronze, 303-306. See Manganese, Spiegeleisen. Ferrous bronze, 303. carbonate. See Spathic ore. See Iron. Fettling of puddling furnace, 91, Feuohtwanger, nickel coinage, 310. Fine metal copper, 222. Finery furnace. See Refinery. iron classified, 134. Finishing rolls, iii, 112. Fir trees, 611-614. woods, American spruce, deflection, 651. transverse strength, 647. Baltic, elastic limit, 636. hemlock, deflection, 651. New England, elastic limit, 636. transverse strength, 647. spruce tenacity, 637. Fir woods, Norway, transverse strength, 647. Riga, tenacity, 637. transverse strength, 647. spruce pine shearing resistance, 645. uses, 611, 660. Fire-box, copper, 232. iron, 119. Fire brick, 581, 582. Fire clay for cement, 588. Fire-proof preparations, 683. Fire stones, 572. Fire for tempering steels, 205, 206. Fisherman's bend (knot), 697. Fish oil, preservative from oxidation, 674. Fizeau, expansion coefficients, 33. Flange iron, boiler plate, 119. tenacity, 363. See Wrought iron. Fleitmann, manganese added to nickel, 255- nickel welded to iron and steel, 255. Flemish bond, 594. Flexure, 516-520. columns, 395-405. See Transverse tests, Torsional tests. Flogging chisels, 208. Floor Beams. See Beams, Columns, Flexure- Floor planking, 653, 654. Floor of brass, 292. metals, 507-520. under stress, 353-359. See Time. Fluctuation. See Variation. Flues, strength, 393, 394. Fluoride of aluminium, 257. calcium. See Fluor spar. manganese, 265. reduction process, 7. , Fluor spar as a flux, 8. in nickel ore reduction, 254. hardness, 373. Flux, definition, kinds, 8, 9. alkalies, 8. borax for welding steel, 204. INDEX. 723 Flux, brass castings, 463. bronze castings, 463. carbonizing in tempering steels, 205. limestone, 571. See Lime. See Litharge. nickel ores, 254. See Phosphorus. solder, 327, 328. steel, 151, 205. Fodder, pig lead, 251. Foil, tin, 237, 238. Folacci, wood preservative process, 684. Forbes, heat conductivity of metals, 17, 19. Forge. See Furnace. See Catalan forge. Forged brass, 297, 298. bronze, 282. copper, 22g, 230. nickel, 254. Forge irons, conversion into Vfrought iron, 179. grades, characteristics, 82. properties, 125, 179. tensile tests, 427, 428. See Wrought iron. Forge metal. See Forged metal. platinum-iridiura alloy, 313. Forge process. See Forge irons, Refinery. Forgible kalchoids, Farquharson's, 487. Tobin's, 488. Forging, 125. iron, proper conditions, 125. manganese-bronze, 305, 306. nickel, 255. steels, 199, 204, 205. Fossil iron ore, 17. Foundation stones, 563. Foundry brass, 317-323. irons, analyses, 180-182. expansion coefficients, 199. grades, 82. properties, 178, 179. specific gravities, 199. Fowler's aluminium-vulcanite compound, 257- Fowne's table specific gravities of pure metals, 22. Fracture, cast irons, 425, 426. See Crystallization. kalchoids, 481. malleableized cast iron, 425. See Molecular change. rapidity of, effects, 515. See Rupture. shock, 493. steels, 424, 425. temperature effects, 492-496. white iron, 426. wrought iron, 427, 428. Francis, distances between shaft-bear- ings, 409, 410. Franklin Institute, copper strength, 528, 529- iron, strength modified by tempera- ture, 491, 492, 528, 529. Franz and Wiedermann, heat conductiv- ities, 18. Freestone. See Sandstone. French art castings, 11. oxide, composition, 296, 297. Fricks' German-silver alloy, 310. Friction of belts, 689-691. Fritz, three-high rolling mill, 164. Frost, resistance of stones, 576. Fuel, blast-furnace consumption, 55, 7a copper ore reduction, 220. in engineering and metallurgy, 9. in zinc manufacture, 246. Furnace, Berard's, 165. See Blast furnace. Bloomary, 45-47, 83-85. brass-founder's, 318-320. Catalan forge, 45, 46, 53, 83. cementation, 137-139. charcoal, 57, 58, 62. converting, 137-139. copper ore, 223. crucible, 142, 143. cylindrical, strength, 394. decarbonizing, 127. Dank's rotating, 98-100. Deccan, 45. direct process, 45, 46. Ehrenwerth's rotating, loi, 102. 724 INDEX. Furnace, electric, 257. forge, 44, 46, 53, 83, 88, 89. French, 46. gas, 48. Indian, 45. malleableizing, 127. Mansley's rotating, loi. 102. Pernot's rotating, 102, 103. Ponsard's, 165. puddling, 90-103. refining, 89, 90, regenerative gas, 48. reverberatory, 90-103. See Reverberatory. rotating, 90-103. Sellers' rotating, 100, loi. Siemens' regenerative gas, 48. StUckofen, 45. wind, 142. zinc smelting, 241, 242. Furniture woods, 660. Fusible compounds, 331. brasses, 289, 296-299. bronzes, 282-287, 455- copper, 282, 296. See Melting point. metals, 34-36. nickel, 254. platinum, 261, 262. tin, 287. zinc, 299. Fusion point. See Melting point. Fusion, cast iron, effect on strength, 377. See Latent heat. metal reduction, 7. Galena, 6, 248. See Lead. limestone, 248. oxidized by ore-roasting, 54. smelting, 249. Galvanized iron, wire, 246. Gard, nickel anodes for nickel plating, 255, 256. Garnet, hardness, 573. Gamier. See Woestyn and Gamier. Gas furnace, 48. Gaseous products of puddling furnace, 49. Gaudaloupe cedar wood. See Guadaloupe. Geological classification of iron ores, 51. deposits of iron ores, 50, 51. German Bessemer rails, irons, analyses, 195. brass, properties, 298. steel classified, 134. silver, 19, 20, 309, 310, 325, 330. solder, 328. Gilchrist, basic dephosphorizing process, 165. Gillmore, crushing resistance of stones, 574- Girders. See Transverse tests. iron, 118. stays, strength formulae, 403-405, 650- 654. Glucinum, 266. Glass, heat expansion, 30, 34. Glue, 676. Glued joints, strength, 669. Gneiss in fire stones, 572. See Granite. iron oxides in, 566. sienitic, 560. water absorption, 576. Goguel, ferrous bronze, ordnance metal, 303- Gold, 266. alloyed with brass, 38, 298. copper, 38. iron and steel, 177. palladium, 267. ancients used, 40. atomic weight, 29, 446. cost, 269. crystalline form, 26. ductility, 23, 25. diminished by lead, 322. > early knowledge, 4. elasticity modulus, 447. variations with temperature, 533. electric conductivity, 17-20, 26. expansion by heat, 30, 33. hardness, 17. INDEX. 725 Gold, heat expansion, 30, 33. tenacity affected by, 532. imitation. See Brass, Bronze, kalchoids, 301, leaf, 297. Mosaic, 2g8. iron alloy, 177. leaf, imitation, 297. malleability, 24, 25. melting point, 26, 36. mock. See Imitation. Mosaic, 298. palladium alloy, 267. solders, 328, 331. specific gravity, 22, 26, 446. heat, 25, 27, 29. steel alloy, 177. tenacity, 446. affected by heat, 532. vaporization, 35. Gongs. See Bell-metal. Gordon's formula for strength of col- umns, 398, 399, 640. Gothite. See Brown hematite. Grading iron ores, 53. Graham bronzing liquids, table, 336, 337- lacquers, table, 338. Grain of woods, 659, 661-663. tin, 238. Granite, 564, 566. Aberdeen, hardness, 573. crushing resistance, 574. fire and heat resistance, 572, 577. See Gneiss, gray hardness Aberdeen, 573. heat and fire resistance, 572, 577. protogene, 566. red, 565. See Sienite. specific gravity, 577. talcose, 566. water absorption, 576, 577. weight, 574. Granular fracture by shock, 493. Granulated iron, carbon percentages, 135. tin, 238. Graphite carbon, 172. Grashof, strength circular plates, 405 Grauwacke, transverse strength, 575. Gray antimony, 253. granite, 565. irons. See Cast iron, expansion coeflScients, 199. fuel, effect of, 58. fusibility, 198. grades, characteristics, 82. specific gravity, 199. heat, 198. yiel^ in roasting, 54. Grease pot, tin plate manufacture, 130. Great Britain, early metal working, 42, 44. See British, i. Greek art castings, II. iron and steel used by, 41, 42. Green ebony, 633. carbonate of copper, 27. Greene, compressive resistance of bricks, 580. Greenheart, 634. crushing resistance, 638. tenacity, 637. Greenstone, 564, 567. Gruner, blast furnace gases, 70. Guadaloupe cedar, 637. Guettier, bronzes for engineering pur- poses, 331-334. cast iron specific gravities, 199. Guide mills, 114. Gum wood, shearing resistance, 646. Gun metal. See Ordnance. Gutta percha, 693-695. Gypsum, 571, 588. hardness, 573. mortar, 582. H. Hackmatack, 615. shearing resistance, 645. Half hitch, 696. Half-round iron, 118. Hammer, blacksmith's, 125. bronze, for finished work, 333. 726 INDEX. Hammer, copper, 232, 233. helve, 108. steam, lOg. temper, 208. Hammered bronze, 278, 279. copper, 226, 229, 230, 232, 233. metals, 16. stones, 590, 5gl. work, repousse, Grecian, 11. Hammering, 107, 109. Hard bricks, 579. pine, 6og. solder, 331. steel, 135, 136. Hardness, aluminium, 308. bronze, 308. brasses, 289, 296-299. bronzes, 276, 282-287, 33°. 455- cast irons, 16, 199. ' See Compressive resistance. copper, 16, 17, 215, 232, 233, 282, 296. iron, cast, 197, 199. ores, 51. lead, 17, 247. metals, 16. ■ Dumas' scale, 1 7. Golner's scale, 16. nickel, 7, 254. platinum, 17, 261. steels, 199. stones, 573. tin, 17, 287. wrought iron, 573. zinc, 17, 299, 573. Hardening, heat effects, 204. metals, 24. repeated effect, 204. steels, 24, 204. See Tempering. tool steels, 200-210. Hartmann puddling, chemical changes, 182-186. Haswell, alloy table, 330, 331. cylinders, cast iron, strength, 384. fusible compounds, table, 331. Hatfield, brick, transverse strength, 580. stones, transverse strength, 575, Hatfield, woods, shearing resistance, 645. Haupt, wood, effect of prolonged stress, 657, 658. Hawsers, 695. Hayford, wood preservative process, 680-682. Hazel wood for elastic uses, 659. Hebrews, knowledge of metals, 4. Heap roasting ores, 54, 55. Heartshake's, 605. wood, 597, 598, 603, 606, 63s. torsion coefficients, 656. Heat. See each material; also, Anneal- ing, Carbon, Cold, Corrosion af- fected by, 542, 543 ; Ductility in- creased by, 198, 534 ; Elasticity moduli variations, 532, 533; Ex- pansion, Fractures of iron at dif- ferent temperatures, ig8, 492-496; Latent heat, Malleability increased. Melting points. Specific gravity. Specific heat. Stresses produced by variations in temperature, 533-536 ; Tfemperature, Volatilization. Heat conductivity, 6, 17, 18, 30-34. aluminium, 256. antimony, 253. bismuth, 18, 26. brass, 20, 296-299. bronze, 282-287. copper, 17, 26, 36, 232, 282, 296. gold, 26. iron, 17, 18, 20, 26. lead, 26, 247. magnesium, 26. manganese, 26. mercury, 26, 250. metals, tables, 17, 18, 26. palladium, 26. platinum, 18. silver, 26. , tin, 26, 237, 2S7. wrought iron, 197. zinc, 17, 26, 299. Heating, ore reduction, 7. surface steam boilers, 81. Heath, use of manganese, 142. INDEX. T27 Heaton, dephosphorizing process, 165. Hedges, table electric resistances, 18. Hematite, 51, 52, 67, 181. Hemlock, 613, 614. wood, deflection, 651, hold upon spikes, 668. shearing resistance, 645, 646. transverse resistance, 647. Henderson, dephosphorizing process, 165. puddling process, 98. steel process, 157. Herve, ferrous brasses and ferrous bronzes, 302, 303. Hesiod, references to iron, 42. Hickory, 628. wood, crushing resistance, 638. shearing resistance, 646. strength, 660. tenacity, 637. torsional coefficients, 655, 656. toughness, 660. uses, 659, 660. Hill, crystals in large forgings, 504, 505. tensile resistances bridge plates, 373. Hindostan, early knowledge of metals, 4. Him, wood seasoning with boiling oil, 35. H-iron columns, 401. History of iron trade ; the six epochs, 49. Hatche's rope, 695-697. Hobson and Sylvester, ductility and mal- leability of zinc, 240. Hock-Leuch's wood preservative process, 679, 680. Hodgkinson, cast iron compressive tests, 391-393- formula for strength, 390. I-beam, 404. strength dependent upon size, 376. tenacity, 376. wood columns, strength, 397, 398, 402, 639, 640, 670. Spence's metal, solubility, 314. Hoisting machinery, 77-81. Holding power of nails, screws, spikes, 668, 669. Holland, iridium ore reduction, 264, 265. Holley, Bessemer plant, 152-164, 167. Holly tree, 629. wood, tenacity, 637. uses, 659. Holman, specific heats, iron and pla- tinum, 29. Homer, reference to ivory, 41. Homogeneous metal, carbon percent- ages, 135. Homogeneousness of castings secured, 169. Honduras mahogany tree, 631. wood, deflection, 651. tenacity, 637. transverse resistance, 647, 651. Hornbeam wood, 659, 660. Hornblende in basalt, greenstone, and trap, 567. granite, 564. sienitic gneiss, 566. Horse-chestnut wood, tenacity, 637. Hot-blast irons, 181. process, 47, 49. stoves, 71-76. House building woods, 660. Housings, rolling machinery, 113. Hunt, classification of copper, extraction processes, 223. Hunt and Douglas, copper extraction process, 224. Huntsman, crucible cast steel working, 141. Huronian deposits of iron ore, 50. Huston, tenacities of boiler plate, 364. Hydraulic cement, 572, 573, 582-588, hoist, 77-80. work, alloys for, 331-333. Hydrogen, use in metal reduction, 7. I-iron, 117. Illinois iron ores, 51. Impact, 349-352, 432. See Resilience, Shock. Impermeable woods. See Wood preser- vation. 728 INDEX. Importation of zinc, 246. Impurities in metals and ores, 5. Incombustible woods, 683, 684. Indian, East, knowledge of metals, 4, 10. steel, Wootz, 145. teak wood, elasticity limit, 636. India-rubber, 693, 694. Indiana iron ores, 51. Indian iron furnace, 45. Indians, iron used by, 42, 43. Indium, 267, 268. Ingot iron, 133. See Cast-iron, classification, 136. compression of castings, 168, phosphorus in, 168. temperature affecting strength, 490. steel, 133. classification, 136. compression, 168. See Steel. Insects and worms, wood boring, 606, 607, 676. Insulator, gutta percha, 695. Internal strain. See Strain, internal. Iridium, 264, 265. alloys, 33, 264, 313. cast, 269. discovery, 26. electric conductivity, 26. fusibility, 36. heat conductivity, 26. osmiridium, 264. ^ platiniridium, 264. specific gravity, 22, 264. heat, 26, 27. Iridium-platinum alloys, 33, 313. Iridosmine, 265. Iron, African manufacture, 44, 45, 50. annealing, effect on tenacity, 496, 497. alloys, 177. with brass, 302. bronze, 302-304, kalchoid, bell metal, 333. manganese, 313. American manufacture, 149. tribes, 41. Iron, amorphous, nearly, 27. ancient analysis, 40. angle iron, 117. as columns, 401. antiquity, 39-44. arsenic in ore, 54, 181. Swedish iron, 187. art work, 12. atomic weight, 29, 446. bar, 105, 106, 117. Barrow, ingot antiquity, 195. See Iron puddled, beam, classified, 134. See Beams strength formulae, 403-403 650-654. bell-metal kalchoid-iron, 333. Bessemer, classification, 136. manufacture, 49, 149, 151-169. for wire, 125. best bar, 105 . bloom, classified, 134, 136. See Coin, Catalan, iron furnace, wrought. See Blast furnace, bog ore, 51. Borneo manufacture, 45. Bowling, analysis, 188. brass alloys, 137, 302. brazing, 328. bright, 82, 179. British manufacture, 42, 44. brown hematite, 51. bronze alloys, 302-304, 334. bulb, 118. See Carbon, carbonate, 51, 52. See Cast iron, chain iron, analysis, 188. channel, 117. columns, 401. , charcoal, 181, 182. boiler plate, 119. chemical structure, 52. properties, 120, 214. coal measure ores, 51. cobalt, alloyed with, 177. INDEX. 729 Iron, classification, 132-136. ores, 51. colors ores, 51. cost, 269. cross-iron columns, 401. crystalline forms, 26, 51. crystallization by shock or jar, 504. deposits, 50, 51. See Direct process. See Indirect process, ductility, 23, 25. malleability, 25. See Elasticity moduli, Elastic limit, early working, 43, 44. uses, 4. Edgemoor Iron Co. formula for tenac- ity, 365- elasticity moduli formulas, 532, 533. temperature effects, 532, 533. elastic limits, 557-560. electric conductivity, 19, 20, 26. as affected by presence of carbon, 197. copper injured by iron, 322. , elongation by stress, 51. expansion. See Heat expansion, 34. Esquimaux, use, 41. factors of safety, 342. feathered, 118. finery, classified, 134. flange, tenacity, 363. forge, conversion into wrought iron, 179. grades, characteristics, 82. properties, 123, 179. tensile tests, 427, 428. forgeable, classified, 133. forging, proper conditions, 125. fossil ores, 51. fractures at different temperatures, 492-496. See Furnace, galvanized, 246. geological classifications, 51. See Girders. gray pig, classified, 133. Great Britain, iron- working, 44. Iron, hardness, 17. See Heat, heat conductivity, 17, 1,8, 20. effects, 493-496, 530, 531, 534-536. elasticity affected by, 532-534. expansion, 30, 34. fractures, characteristics, 492-496. tenacity affected by, 494, 495, 528, 529, 532- hematite ores, 50, 52, 67, iSi, hot-blast process, 47. I-beam, 404. impurities, 16. India working, 45. ingot, 133. classification, 132. kalchoids and iron bell metal, 333. magnetic ore, 51. malleability, 25. See Ductility, Elongation. effect of increase of phosphorus, 174, 322. sulphur, 175. manganese in, 173, 187-189, 195, 266, 313- melting point, 26. meteoric, 52, 190. microscopic appearance, 506. mineralogical classification, 51. native, 52. nickel welded to, 255. nomenclature, 51, 132-136. See Ores. arsenic in, 54, i8l. reduction, 53-86. oxide. See Oxide in fire brick, 581. paint, 674. peroxides, 6, 54. pig. See Cast iron classification, 133, 134- See Phosphorus. See Plate iron, properties physical, no, 191, 194-19(4, 210-214. See Puddling, pure, 170. pyrites, 6. 730 INDEX. Iron, pyrolignite for wood preservation, 676, 682, 683. sulphate as a paint, 674. rail, 106, 118. remelting pig, classification, 133. rolled, classified, 134. See Puddled, safety factors, 342. sets, 557-560. shearing resistance, 430. Silurian copper ores, 51. solder, 331. sound conduction, 446. specific gravity, 22, 26, 51, I97, 446. specific heat, 26, 29. specular ores, 51. See Steel. sulphate as a paint, 674. sulphides, pyrites, 6, 181. See Sulphur, temperature. See Heat. modification, 493-496, 530, 531, 534-536- tenacity, Edgemoor Iron Co. formula, 365. tensile, 366, 369-371, 428. transverse tests, 402-405. tinned plate, 239. weld. See Wrought. welding, 170. wire, 123, 125. See Wire. wrought, weight, 21, ^^ Italian art work, 11. Ivy, longevity, 598. J- James, strength castings dependent upon size, 376. Jamin, reflective power of metals, 20. Japan varnish, 673. Japanese art work, 12. copper, 227. Jobbins, location of strongest bronzes, 482-488. Johnson. See Calvert and Johnson. Joiners' work, wood for, 659. Joints, glued, 669. wood, 664-666, 669. Jones, metal compression by steam, 168. Jouraffsky, strength of rails at different temperatures, 490. Journal bearings. See Brass, Bronze. Junipers, 616, 617. K. Karsten, heat expansion of brass, 33. brass-iron alloy, 302. brittle bronze rendered ductile, 276. copper alloys with tin, 276. Kalchoids, 300, 301, 329, 330, 464-488. amalgams, Kingston's metal, 335. bearings, 301, 332-334. bell metal, 333. See Brass. brasses. See Bearings. See Bronze, statuary, 300. button brass, 300. chrisocalle, 300. connecting rod bearings, 332. See Copper. Depretz's chrisocalle, 300. ductile, 471, 478. ductility affected by heat, 530, 531. elasticity moduli, 480. Farquharson's alloy, 487. Fenton's alloy, 334. fractures, 481. gold, imitation, 301. Haswell's table of compositions, 33a heat, effect of sudden changes, 534. increasing oxidation, 538. tenacities affected by, 530, 531. hydraulic work, 333. Kingston's metal, 335. , lathe work, 334. machine, 333. Mackenzie's alloys, 300. Margraff's alloy, 334. table of properties, 301. maximum, 472, 473, 482-488. INDEX. 731 Kalchoids, medal, 283, 301. model, 470-472. oxidation affecting strength, 539. increased by heat, 538. pistons, 333. properties, 301. slide valves, 332. speculum metal, 300. statuary bronze, 300. strain diagrams, 473-475. strongest, 472, 473, 482-488. tenacities, 476-478, 484, 485, 487, 488. Thurston's investigations, 466-482. See Tin. Tobin's alloy, 487, 488. torsion tests, 473-475, 484, 485. tough alloys, 473. turned work, 334. See Zinc. Kelley, pneumatic process, steel making, 49. 152- Kent, blast furnace charge, materials, 66. rusting accelerated by gases of com- bustion, 212. test piece forms and dimensions, 357, 358. Kerl, analyses of tins, 237. Kerpeley, analyses of Bessemer steel rails, 195. Kilns, brick, 579. roasting ores, 54-57. King posts, 665, 666. Kingston's metal, 335. Kingwood for ornamental work, 663. Kirkaldy, compressive tests, wrought iron, 392. granular appearance of fractures by shock, 493. phosphor-bronze, ductility and tenac- ity, 281. velocity of rupture, effects, 514. Knots, rope, 695-697. Koechl, melting point, tin-zinc alloys, 311- Kohlrausch. See Loomis and Kohl- rausch. Kollman, wrought irons at different temperatures, 490-492. Kohn, manufacture of Siemens-Martin steel, 147. Kopp, relations between atomic weights and specific heats, 490-492. Krieger, hollow zinc castings, 244. Kupfernickel, 254. Kyan, wood preservative process, 676, 677. L. Lacing for belts, 337-339. Lacquer. See Bronzing. Lafond, investigation of bronzes, 282- 287. Laidley, strength thick cast iron cylin- ders, with heads, 388. Lake Champlain iron ore deposits, 50. Superior copper. See Copper, iron ores, 50. Lame, formula strength of cylinders and spheres, 386-388. Lamps, magnesium, Larkin's, 263. Thurston's, 262, 263. Lan, tempering affecting proportion of carbon, 210. Lancewood, 634. tenacity, 637. transverse strength, 647. uses, 659. Landore, experiments in elimination of sulphur from steel, 150. sulphur not eliminated from steel bath, 150. Langley, analyses of crucible steels, 191, 192. effect of hardening upon specific grav- ity of steels, 203, 204. Lanthanium, 267. Lanza, wood, crushing resistance, 642. Laplace and Lavoisier. See Lavoisiel and Laplace. Lafond, composition of bronze ramrods, 282. Lapis lazuli, hardness, 573. Larch tree, 614, 615. 732 INDEX. Larch tree, longevity, 598. wood, crushing resistance, 638. deflection, 651. elastic limit, 636. hackmatack, 615. shearing resistance, 645. tenacity, 637. transverse resistance, O361 647, 651- turpentine, 6ig. uses, 614, 615, 660. Larkin, bronzing liquids, 335, 336. magnesium lamps, 263. Laslett, proportions struts, 641. Latent heat, table, 36, 37. lead, 247. mercury, 258. tin, 237. Lathe work, woods, 659. Lavoisier, introduction of use of balance in researches, 5. and Laplace, expansion of coefficients, 30. Laurentian deposits, iron ore, 50. Lead, 247-252. America, early used, 41. amorphous, nearly, 217. antimony alloys, 322. bismuth alloys, 312. tin alloys, 312. atomic weight, 29, 446. bismuth-antimony alloys, 312. in bronze, 321, 324. compressive resistance, ^6. copper sheets, 223. copper injure white lead, 322. cost, 269. crystalline precipitations, 26. ductility, 25. in early times, 12. elasticity moduli, 446, 447. variations with temperature, 533. electricity resistance, 18-20, 26. expansion by heat, 30, 33, 34, 247, galena ore, 6, 248. hardness, 17, 247. heat conduction, 26, 247, expansion, 30, 33, 34, 247. Lead, latent heat, 37, 247. litharge, 252. malleability, 25, 247. mediaeval uses, 12. melting point, 26, 35, 36, 247. paints, 671, 674, 675. Pattinson process, 249, 250. pipe, 247, 251. lead-tin alloy, 311. rolled, weight, 21. roof covering, mediaeval, I2. solder, 35, 331. specific gravity, 23, 26, 247, 446. heat, 27, 2g, 247. in statuary bronze, 324. strength, 445. sulphate, adulterant of white lead, 675. tenacity, 15, 247, 444, 445. tin-alloys, organ pipes, 311. tin-antimony alloys, 312. volatilization, 247, 249. white copper injures, 322. paint, 671-675. sulphate of lead, an adulterant, 675. Leather : belting, packing, valves, 687- 692. cement, 692. Lehigh zinc ore, 246. Letherby, carbolic acid in dead oil, 678. Leuch's wood preservative process, 679, 680. Lignumvitae wood, 632. crushing resistance, 638. elasticity, 636, 659. tenacity, 637. toughness, 659. transverse strength, 647. uses, 659, 660. Lime air, 571, 583. alabaster, 571. brick, silicate of lime, 570, fire brick, 581. See Brick, brick-work, 5. carbonate, hardness, 573. See Carbonate of lime. INDEX. ;33 Lime carbonates, flux for silica, g. hardness, 513. common, 571, 583. copper ore reduction, 2 1 9. fire brick, 581. stones, 572-577. flux for silica, 8. in tin ore reduction, 235. gypsum, 571. hydraulic, 573, 583-585. ore reduction copper, 219. plaster of Paris, 571. preservation of timber, 606. sandstone, 568. silicate in bricks, 570. slag, g. specific gravity, 67. heat, 67. sulphate, 571. weight, 67. Vfood preservative, 606. Limestones, 569. alabaster, 571. brick, silicate of lime in, 578. fire, lime in, 581. carbonate of lime, flux for silica, g. hardness, 573- carboniferous, iron ore in, 50. crushing resistance, 574. fire stones, 572-577. flux, 9, 571. gypsum, 571. heat resistance, 572, 577, 581. magnesium, 569. plaster of Paris, 571. specific gravity, 67. heat, 67. weight, 67. Lime tree, 615. longevity, 598. uses ; carving, engraving, 659. Limonite, 51, 52. Limnoria terebrans, 607. Linden tree, 615. longevity, 598. Linseed oil, 675. Liquation, 318, 321. Liquation in bronze, 278, 283, 285. Litharge, 252. in dryers and paints, 671, 672. Litliium, 266. cost, 268. melting point, 36. specific gravity, 22. Lithographic paper, 693. Live load. See Load, live. Live oak tree, 621. wood, crushing resistance, 638, shearing resistance, 646. tenacity, 637. Loads, static, 550. See Stress. Lloyds' rule, strength cylindrical boilers, 383. flues, 3g4. stayed plates, 380. Load, dead, definition, 341. copper transverse tests, 546. iron tests, 559, 560. live, 341, 342. maximum, non-ferrous metals, 562. See Safety factors. See Static, Strain, Stress, 550. working, 341, 342. Locust tree, 628. wood, crushing resistance, 638. shearing resistance, 645, 646. tenacity, 637. toughness, 660. torsional coefficients, 655, 656. resilience, 656. transverse resistance, 647. tree-nails, strength, 666. uses, 660. Loomis and Kohlrausch, elasticity mo- duli, 532. effect of variable temperature, 532, ■ 533- Longevity of trees, 598. Long-eared pine, 6og. Loup, 84. See Hammering, Rolling, Squeezing. Lowmoor iron, analysis, 187. Low steels. See Mild steels. 734 INDEX. Low wine, turpentine product, 6i8. Lustre of metals, 20, 21. M. Machinery alloys. See Alloys. See Brass, Bronze, 279, 282, 283. design, 126. See Iron. safety factors, 342. See Steels. wood for frames, 660. Mackensie, antimony-bismuth-lead al- loy, 312. antimony-tin-lead alloy, 312. Kalchoids, 300. Magnesia in fire bricks, 581. in fire stones, 572. in slag, 9. carbonate in hydraulic limes, 583. Magnesian limestones, 569. resistance to heat, 572. Magnesium, 262, 263. atomic weight, 29, 262. cost, 269. discovery, 26. electric conductivity, 19^ 26. heat conductivity, 26. in iron, meteoric, igo. lamps, Larkin's, 263. Thurston's, 262, 263. melting point, 26. in nickel, 322. signal apparatus, Thurston, 262, 263. specific gravity, 22, 26. specific heat, 26, 29. in steel, 322. Magnetic Manganese, 265. nickel, 254. ores, 57, 180, 181. oxide. See Magnetite. specific gravity and weight, 67. specific heat, 67. steel. See Magnetism. Magnetism in iron, 171. See ores. steel, 171, 199, 498. Magnetite, 51. chemical composition, 52. roasting, 54, 55, 57. Mahogany trees, 631. wood, columns, strength, 641. crushing resistance, Spanish, 631 deflection, 651. elasticity limit, 636. grain ; markings, 662. hold upon screws, 668. tenacity, 637. torsion coefiScients, 655, 656. transverse resistance, 647. uses, 659, 660. Malachite, 217. Malleable bronze, 282, 285, 286. iron. See Wrought iron. Kalchoids, Farquharson's, 487. Tobins', 487, 488. Malleableized cast iron, 88, 127, 136, igo. Malleability, 24, 240, 243-245, 266, 289, 299- aluminium, 256, 258. -bronze, 307. brass, 289, 2g6-2gg. bronze, 276-27g, 282-2S7, 455. manganese bronze, 303-306. carbon, effect upon iron and steel, 198. copper, 226-230, 282, 287, 296. See Ductility, gold, 24, 25. See Hammering, iron, 25. lead, 25, 247. magnesium, 262. manganese, effect upon iron and steel, ig8, 322. metals, 23-25. metals, foreign, effect upon, 197. nickel, 25, 254. ^ See Phosphorus, platinum, 25, 261. silver, 25. steel, ig7, 199. See Sulphur. tin, 25, 236, 287. INDEX. ;3S Malleability, zinc, 240, 243-245, 299. See Annealing. Torsional tests, opp. 415, 425, 426. Mallet brass, properties, 289, 296-299, 457- bronze, properties, 276. oxidation of iron, 212, 213. resilience coefficients, 350. sodium-tin alloy in ordnance bronze, 229. tin, tenacity, 437. Manganese, 265, 266. alloys, 265, 313. atomic weight, 29, 265. Bessemer steels, 195, 196. bronzes, 281, 303-306. Admiralty tensile tests, 450. in cast irons, 178, 180-182, ig6. cementation changes, proportions, I91. cost, 269. crystallization effects, 27. indecarbonization, 88. discovery, 26. electric conductivity, 26. ferro-manganese, 266. fluoride, 265. fusibility, 36. hardness, 17. heat conductivity, 26. in iron, 173, 187-189, 195, 266, 313. iron alloyed with, 313. malleability effects, 198, 322. melting point, 26, in nickel castings, 255. in open-hearth process, 150. specific gravity, 22, 26, 265. specific heat, 26, 27, 29. in spiegeleisen, 265, 266. in steel, 142, 143; 167, 191, 193, 373. Manifold paper, 693. Manilla rope, 695. Mansley's rotating puddler, loi. Maple trees, 630. wood, crushing resistance, 638. deflection, 651. glued joints, 669. grain, 662. Maple wood, hold on spikes, 668. shearing resistance, 646. tenacity, 637. transverse resistance, 647. uses, 659, 660. Marbles, 569, 570, 574. abrasive resistance, 573. calcareous stones, 569. transverse strength, 575. Margery, wood preservative process, 684, 685. Margraff, Kalchoids, 301, 334. Marine animals which attack timber, 606, 607. glue, 670. Markings of woods, 661-663. Marten, microscopic investigations of iron and spiegeleisen, 505, 506. Masonry, 588-596. Matte, copper, 221-223. Matthey, platinum-iridium alloy, 313. Matthiessen, brass investigation, 296. bronze investigation, 282-287. zinc investigation, 299. Maximum bronzes, Thurston, 482-488. loads, non-ferrous metals, 562. Mechanical metallurgical processes, 9, puddling, 97. Medal bronze, composition, 283. Melaconite, 217. Melting brass, 317, 318. department, Bessemer works, 155. furnace. See Furnace, for zinc, 243, 245. point aluminium, 258. antimony, 26, 36, 253. brass, 35, 291. cast iron, 35, 36, 198. copper, 26, 35, 36, 215. effect of different elements, 197. lead, 26, 35, 36, 247. metals, tables, 26, 35, 36. Spence's metal, 314. tin, 35, 36. tin-zinc alloys, 311. wrought iroii, 35, 197. zinc, 26, 35, 36, 245. 736 INDEX. Melting zinc-tin alloy, 311. Merchant irons, 105. Mercury, 258-260. amalgams, 258, 259. atomic weight, 29. bichloride for wood preservation, 676, 677. cost, 269. electric conductivity, 19, 26. expansion by heat, 33, 258. heat conductivity, 26. latent heat, 258. melting point, 26, 35, 36. metal reduction, 7. specific gravity, 22, 23, 26, 258. heat, 26, 27, 29, 258. tempering bath, 200. Metals. See Alloys. age and exposure, effects, 498, 500. alkaline, 37. amorphous, 27. characteristics, 6, 13, 26. chemical, 37. chemical combinations, 6. cooling expansion, 330. cost, 268. crystallization, 26, 27, 500-507. distinction from metalloid, 6. ductility, 23-25. early knowledge, 3, 4, 10-12. electric conductivity, 6, 18, 19. electro-positive, 6. expansion by heat, 30-34. flow, 507-520. fusibility, 34-36. » hammered, 16. hardening, 16. heat conduction, 6, 17, 18, 30-34. effects upon strength, 16. historical, 3, 4, 10-12. impurities, effects, 5, 16, 268. lustre, 20, 21. malleability, 23-25. melting-points, 35, 36. iron, ferrous, 15. maximum loads, 562. oxidation, 498, 500. Metals, oxides, 9, 37. in slag, 9. properties, 5, 6. rare, 266. reduction processes, 7. resistance to stress, 353-359- rolled, 16. rust. See oxidation. specific gravities, 22. sulphides, 37. temperature influence, 16. tenacities, 15. See Testing. tools to cut, 205. useful, 12. volatilization, 34, 35. weather effect, 498-500. Metallic salts for wood preservation, 676. Metalloids, characteristics, 6. Metallurgy, early, 2, 3. fuels, 9. mechanical processes, 9. Metamorphic rocks, 50, 564. Metcalf, effects of hardening, 200-203. Metcalfe, tests of cartridge metal, 457, 458. Meteoric iron, 52, 190. microscopic appearance, 506. Mica in fire stones, 572. granite, 564. hardness, 573. slate, fire resistance to, 572. See Granite. Michaelis, deterioration sheet copper with age, 230. Michigan iron ore, deposits, 50. See Lake Superior. Microscopic investigations of iron and steel, 505-507. Mild steel. See Steel. Miller, analyses of irons ; same mallea- bleized, 190. cubic crystallization in Bessemer steels, 504. Millit, investigation of bronzes, 2S2-287. Mills, rolling, 115. Mill-work, 105. INDEX. 717 Mineralogical classification of iron ores, 51- Mineral paints, 674. Mine tin ore, 235. Minium, red lead, 247. Mirrors, metallic, 12. speculum metal, 276, 277, 279, 285, 300, 330. Missouri iron ores, 50. Mixing copper ore, 220, 221. iron ore, 50. Model copper-tin-zinc alloys, 470-471. Molding, 126, 127. Molecular changes. See Crystallization, elastic limit variations, Flow. Molybdenum, 266. cost, 269. fusibility, 36. specific gravity, 22. Morin, aluminium bronze, greenish, 308. belts, friction, 691. strength, 688. pillars, strength, 640, 641. Morrison, steam hammer, 49. Mortar, 572, 573, 582-588. Mosaic gold (brass), composition, 298. Moses, references to brass and iron, 41. Mottled iron, 82, 179. Mountain elm tree, 627. Muck-bar, 105. Mudge, speculum bronze, 279. Muntz sheathing metal, 290. composition, 330, 480. strength affected by heat, 530, 531. tenacity, 457. Muschenbroek, properties of bronzes, 282-287. Mushet, alloys of manganese and iron, 313- manganese in steel, 173. sheathing bronze, 276. N. Nagle, horse-power of leather belts, 691. Nails, resistance to penetration and withdrawal, 667. 47 Nasmyth steam hammer, 48. Native metals, reduction processes, 7. Navier, strength of columns, 396, 397. Neilson, hot-blast iron manufacture, 47. Neuberg, Bessemer boiler-plate analyses, 195. Newton fusible compounds, 331. New York iron ore deposits, 50. Nickel, 254, 256. alloyed with copper, 310. and zinc, 309, 310. iron and steel, 177. magnesium and phosphorus, 322, amorphous, nearly, 27. anodes, 255. atomic weight, 29, 254. carburetted, specific heat, 27. cost, 269. discovery, 26. ductile, 322. electric conductivity, 26. elongation, 446. expansion by heat, 33. fusibility, 36, 254. hardness, 7, 254. heat conductivity, 26. expansion, 33. in irons, 187-189. magnesium, effect, 322. malleability, 25, 254. manganese effect, 255. melting point, 26. in meteoric iron, 190. phosphorus, effect, 322. plating, 254, 255. specific gravity, 22, 23, 26, 254. specific heat, 26, 27, 29. in steel, 193. tenacity, 446. weight, 23. welding to iron and steel, 255. Niobium, 267. Nipping blocks, wire drawing, 121. Nitrogen in iron and steel, 176. Nomenclature of iron and steel, 132-136 Non-ferrous metals, 15. maximum loads, 562. 738 INDEX. Non-ferrous metals, safety factors, 342, 343- Norton formula deflection ; coefScients, 651. Norway fir, 613. transverse strength, 647. pine, 608. Norwegian iron ores, 50. Nut oil, 50. O. Oak trees, 620-625. conversion of sap wood into heart wood, 598. longevity, 598. wood, columns, strength, 641. crushing resistance, 638, 642. deflection, 651. durability, 659. elastic limit, 636. glued joints, 669. grain, markings, 662. hold upon nails, screws and spikes, 667, 668. resilience, torsional, 656. shearing resistance, 645, 646. tenacity, 637. torsional resistance, 655, 656. transverse resistance, 647. treenails, strength, 666. uses, 659-661. Odors of metals, 24. Ohio iron ores, 51. Ohio pine. Oil, dead, for wood preservation, 676- 679, 686. for paints and varnishes, 670-675. tempering bath, 200, 207, Oil of turpentine, 6l8, 619. Oleiferous limestone, resistance to heat, 572. Oolitic iron ores, 50. Open hearth process for irons and steels, 49, 146-151. steels, classification, 134. See Siemens- Martin, strength, 372, 373. Open hearth process for steels, tensile tests, 373. Ordnance bronze, 276-278, 280, 281. casting, 281. in chills, 535, 536. compositions, 283, 330. elasticity, 458. kalchoids, 329. maximum stress, 562. oxidation, 277, 278. affecting strength, 539. repeated strains, effect, 561. safety factors, 562. specific gravity, 22. temper of tools to cut, 208. tensile tests, 448, 449, 458, 481. cast iron, 375. copper-tin-iron alloy, 302, 303. endurance under fire, 511. Orange tree, longevity, 598. Ore, alumina, 25-29, 256-258. aluminium. See alumina, argillaceous, 54, 55. arsenic, 54, 263, 264. antimony, 252, 253. bismuth, 53, 254. black band, 54, 55. calcination, 53. character for different purposes, 57. chemistry of reduction, 64-66. copper, 216-218, 223-225. reduction, 219, 221, 222. crushing, 9. distribution ; laws governing, 13. grading, 53. impurities, 5. iridium, 264, 265. iron, 50-52, 67, i8r. black band, 54, 55. bog ore, 51. hematite, 51, 52, 67. magnetic, 51, 57, 180, 181. peroxides, 6, 54. pyrites, 6. lead, 6, 247-250, 252. magnesuim, 262. magnetic, 51, 57, i8o, 181. INDEX. 739 Ore, manganese, 266, mercury, 259, 260. See Metal, mixing, 53. nickel, 254, 255. platinum, 260-262. , arsenic, 263. reduction, chemistry of, 64-66. copper, Z18-230. See Furnace, iridium, 264, 265. iron, 44-86. lead, 248-251. magnesium, 262. manganese, 265. mercury, 259, 260. platinum, 260. processes, Wright's table, 7. tin, 234-236. zinc, 240-243. Silurian, upper, 53. sorting, 53. specular, 54, sulphur, 54. tin, 234-237, 239. wasting, 9, 10, 53. zinc, 6, 240, 242, 246. Oregon pine, crushing resistance, 639. tensile resistance, 637. transverse resistance, 647. Oreide composition, 296, 297, 330. Organ pipes, tin-lead alloy, 331. Oriental plane tree, longevity, 598. Ornamental woods, 661-663, 639. Orthogonal strains, 526, 527. Osmiridium, 264. Osmium, 264, 266. cost, 268. fusibility, 36. heat expansion, 33. Over-poling, copper ore reduction, 219, 223. Overstrain, evidence from subsequent tests, 522-527. Oxidation, 211. of aluminium, 256-258. aluminium bronze, 307-309- Oxidation ot bronzes, 277, 278. aluminium, 307-309. manganese, 306. bronzing liquids ; process, 335, 336. combustibles in roasting ores, 54. copper, 225, 226, 539. alloys, increased by heat, 538. exposure of metals to different media, 211, 212. iron, 170, 171, 210-214. lacquering, 337-339. lead, 247. manganese, 265. bronze, 306. nickel, 254. ordnance bronze, 539. paints, 670-675. platinum, 261, 262. steel, 210-214. uranium, 267. varnishes, 673, Oxide, aluminium, 257. bismuth, 253, 254. copper, block, red, 217. iron, 50. in fire-bricks, 581. hematites, 51, 52, 67, 181. in gneiss and mica slate, 566. ores, 6. lead, 247. magnetite, 51, 52, 54, 55, 57. manganese in steel production, 47. metals, 37. in slag, 9, 186. in ores, 6. reduction processes, 7. in puddling, 95, 186. tin, 237. zinc, 6, 240, 245. Oxygen, cause of rust, 211. Packing metal, 329. Paints, 670-675. bronze, 336. Palladium, 26, 267. 740 INDEX Palladium, atomic weight, 29. cost, 268. elasticity moduli, 447. variations with temperature, 533. electric conductivity, 19, 20. expansion by heat, 33. fusibility, 36. hardness, 17. specific gravity, 22. specific heat, 29. tenacity, 446. affected by heat, 532. -gold alloy, 237. Paper, kinds and uses, 692, 693. blue prints, 693. wash for, 692. Papier mache, 693. Paraffin paint, 674. wood preservative, 679, 680. Parnell process, copper ore reduction, 242. zinc manufacture, 243. Parsons, manganese-bronze, 304-306. Partridge, wood for ornamental work, 663. Passaic zinc ore, 246. Pasteboard, 693. Patina, 278, 321. Patterns for molding, iz6. woods for, 659, 660. Pattinson process, lead manufacture, 249, 250. Pear tree, longevity, 598. wood, crushing resistance, 53S. tenacity, 637. uses, 659,, 65o. Peat for iron furnaces, 47. Pennsylvania iron ores, 50. Percy, analyses of ancient alloys, 272. Delhi iron pillar, 43. Pemot rotary puddler, 101-103. Petit, see Dulong and Petit. Petroleum products for wood preserva- tion, 682. Pewter, 239, 312, 313, 330. See Bronze. expansion by heat, 33. Pewter solder, 331. specific gravity and weight, 23. Phillips, analyses of ancient alloys, 4, 271, 272. list of copper ores, 216, 217. petroleum products for wood preserva' tion, 682. Phoenix Iron Co., tests of long bars, 368. Phosphide of iridium, 265. Phosphor-bronzes, 274, 281. strength affected by heat, 530, 531. repeated strain, 561. wire, 381. Phosphoric acid in slag, 186. Phosphorite, hardness, 573. Pickling sheet iron, 129. Pig-boiling, 91. Pig lead, 251. Piles, 106, 107. Pillars. See Columns. Pinchbeck, composition, 296, 330. Pine trees, 607, 6u. Newfoundland red, 614. woods, columns, strength, 639, 641. crushing resistance, 630, 638, 639 641, 644. deflection, 651. durability, 659. elastic limits, 636. hold upon nails, 667. Ohio, shearing resistance, 645. Oregon, crushing resistance, 639. tenacity, 637. transverse strength, 647. pitch, deflection, 651. elastic limit, 636. tenacity, 637. transverse strength, 647. qualities, 612. red, columns, strength, 641. * crushing resistance, 638. elastic limit, 636. shearing resistance, 645. tenacity, 637. transverse strength, 647. railroad ties, 687. INDEX. 741 Pine wood, red, resilience, torsional, 656. shearing resistance, 645, 646. Southern, 609, elasticity modulus; transverse strength, 653. See Yellow. spruce, shearing resistance, 645. torsional coeflficients, 655, 656. strength, 659. stress prolonged, effect, 657, 658. tenacity, 637. torsional resistance, 655, 656. transverse resistance, 647, 653. white, crushing resistance, 638. deflection, 651. elastic limit, 636. shearing resistance, 645, 646. tenacity, 637. torsional coefficients, 655, 656. yellow, crushing resistance, 638, 642-646. durability, 659. elastic limit, 636. safe loads, 649. shearing resistance, 646. strength, 659. tenacity, 637. torsion coefficients, 655, 656. resilience, 656. transverse strength, 647. uses, 659, 660. Pins, wooden, strength, 666. Pipes, cast iron, strength, 384. copper, 229, 233. flues, 393, 394. lead, 247, 251, 445. organ, tin-lead alloy, 311. strength, 79, 384, 388, 394. tin, 238. See Tubes. Pitch, 618. in bituminous cements, 588. Burgundy, 613. Plane-tree wood, durability, 659. Planing tests. 374. tools, temper, 208. Planking, floor, strength, 653, 654. Plaster of Paris, 571, 588. hardness, 573. mortar, 5S2. Plate-engraving alloy, antimony-tin-lead, 3^2- Plate iron, 107, 1 18-120. See Bessemer iron and steel. boiler heads, strength, 382, 383. shells, strength, 119, 382-384. tensile tests, 364. bridge, strength, elongation, 373. classification, 134. elastic limit, 364, 373. elongation, 364, 373. stayed plates, strength, 381, 382. tank iron, tenacity, 363, 364. tensile tests, 119, 363, 364, 373. See Wrought iron. Plate mill rolls, 114. Plating, nickel, 254, 255. Platiniridium, 264. Platinum, 26, 260-262. alloyed with iron and steel, 177. atomic weight, 446. cost, 260, 262, 269. ductility, 25, 261. elasticity moduli, 447. variations with temperature, 533. electric conductivity, 18-20. expansion by heat, 30, 33, 34, 261. hardness, 17. heat conductivity, 18. effect upon tenacity, 532. iron alloy, 177. malleability, 25. phosphorus reaction, 265. specific gravity, 22, 23, 261, 446. heat, 27, 261. solder for gold, 328. sheet alloy, 177. tenacity, 446. Platinum-iridium alloy, 33, 313. Plum wood, tenacity, 637. Plunger hoist, 77-80. Plutonic rocks, 364. Pneumatic hoist, 80. 742 INDEX. Pneumatic hoist process, i86. See Bessetner. Kelly, 49. steels, classified, 134. Poling, copper ore reduction, 219, 222, 223. Ponsard furnace, 165. Poplar wood, durability, 659. shearing resistance, 646. tenacity, 637. Poppy oil, 675. Porphyry, 567. Portland cement, 584. abrasive resistance, 573. Posts. See Columns. Potash, flux for silica, 8. in fire bricks, 581. stones, 572. Potassium, atomic weight, 29. cost, 269. hardness, 17. melting point, 36. in meteoric iron, igo. specific gravity, 22. specific heat, 29, Pots or crucibles for steels, 141-143. Prechtl, malleability and ductility of metals, 25. Preservation of wood, 670-686. Primary rocks, 51. Prices of metals. Table \^See under Cost in detailed entries.] 268, 269. Proof stress, definition, 341. cast iron, 374. See Elastic limit. tool steels, 374. Protogine (granite), 566. Protoxide changed to peroxide in roast- ing, 54- of iron and of manganese in slag, 186. Puddled irons, 105-107. classification, 136. steel, 104, carbon percentages, 135. castings, large, 145. Puddled steel, classification, 136. Puddling, 7, 90-105. chemical changes, 182-187. Corts' patents, 48. dry-puddling, 103, 104. Ellershausen process, 98. furnaces, 98-103. Siemens, 48. utilization of gases, 49, 97. Henderson process, 98. mechanical, 98-103. Siemens process, 97. Puddle train, 105, m, 112. Punches, bronze composition, 284. Punching boiler plates, 405. flow of metals, 507-510. Purification, copper, 219. mercury, 260. platinum, 260, 261. See Refining. Putty, 671. Puymaurin, alloy, 275. coin bronze, 277. Puzzolana, hydraulic cement, 583. Pyrites, copper, 217. in iron ores, 6. Pyrolignite of iron for wood preserva- tion, 676, 682, 683. Q- Quartz as a flux, 8. in fire stones, 572. granite, 564. Queensland tin, analyses, 237. tests, 439-441. Queen's metal, 312. R. Rachette blast furnace, 60. use in lead smelting, 248. • Rail iron, 106, 118. mill, 115. steel, analyses, 195. variations in tenacity with tempera- ture, 490. Railroad ties, preservation, 6S5, 686. INDEX. 743 Rangework masonry, 592. Rankine, belts, strength and tension, 688-690. formula for strength of columns, 398, 399- modulus of elastic resilience, 350. Raw hide for belts, 691. Recarburization, open-hearth process, 147- Records of tests, blank forms, 360, 361. Reduction of area, under stress. See Tensile tests. Reef knot, 6g6. Reese, basic-dephosphorizing process, 165. Refining copper, 219-225. See Flux. forge process, 88, 89. iron, 88-104. lead, 249, 250. mercury, 259, 260. platinum, 260, 261. purification of mercury and platinum, 260. steel, 148, 151, 181-186. tin, 236. zinc, 241-246. Regenerative gas furnace, Siemens, 48. Regnault, ductility and malleability of metals, 25. expansion coefficients of mercury, 33, 258. specific heat of cast iron, 198, 199. metals, 27-29. platinum, 261. steel, 199. wrought iron, 197. Regulus, copper ore reduction, 219, 221, 222. Reheating of wrought iron, 105-107. Remelting of cast iron, effects, 377. Repousse work, 10, 11. Resilience, 349-352. brass, torsional, 461, 462. bronze, torsional and transverse, 452, 453- Resilience, coefficients, 350. copper, torsional and transverse, 452, 461. impact, 349-352, 432. Mallett's coefficients, 350. See Resistance, Shock, Tensile, Tor- sional and Transverse tests, steels, 372. tin, 441, 453. woods, 656. zinc, 462. Resistance. See each material and the following : Abrasion, Annealing, Carbon, Circular plate bolted at edge, 405 ; Columns, Compress- ive, Cordage, 695 ; Crushing, Cylinders, 79, 382-388, 393, 394 ; Detrusion of woods, 644-656 ; Ductility, Elasticity, Elastic Lim- its, Electric Conductivity, Elonga- tion, Fractures, Frost resistance of stones, 576 ; Hardness, Heat, Malleability, Metals, Oxidation, Safety factors,' Sets, Shearing, Specific gravity. Stiffness, Strain, Strain diagrams, Stress, Tempera* ture, Tensile tests, Thurston, Time, Torsional tests. Transverse tests. Rest from strain. See Strain. Reverberating furnace, 90-103. for Bessemer steel, 155, 156. direct-process iron, 41. copper ores, 220, 224, 225. Corts', 48. lead, 249, 250. puddling, 91. reheating, 107. Siemens, 48. tin ores, 234, 235. zinc ores, 241. Re-working of wrought iron, 105-107. Reynolds, use of oxide of manganese in steel, 47 Rhodium, 267. cost, 268. fusibility, 36. 744 INDEX. Rhodium, hardness, i6. Riche, aluminium bronze, specific grav- ity, 307, 308. brasses, specific gravities, 296-299. bronzes, annealing and tempering, ef- fects, 275, 536-538. investigations, 282-287. copper, annealing and tempering, ef- fects, 237. Riga fir, grain, 662. tenacity, 637. transverse strength, 647. hemp rope, 695. Rimmers, temper, 208. Riveted work, strength, 405-408. Roasting ores, 54-57. See Calcining, copper ores, 219—222. Robbins, wood preservative process, 678, 679. Rocks, 564. crusher, 9. Rodman, endurance of cannon, 511. Rolled brass, 292, 293, 297, 316, 319. copper, 223, 229-233, 296. iron, sheet, 117, 118. See Plate-iron. lead, 247, 251. manganese bronze, 305. tin, 238. zinc, 243-245. Rolling, 105, 107, 111-116. cooling during the operation, 116. Corts', patents, 48. machinery, 111-117, 164. mills, 164. products, shapes, 1 16-120. zinc, 243. hitch, 696. Roman art-castings, 11. brass, 271. bronze, 11, 273. cement, 584. abrasive resistance, 573. iron working, 41, 42, 44. Romilly brass, 2go. Ropes, 695-697. Rose, fusible compound, 331. Rosewood, uses, 634, 660, 663. Rosin, 618. Ross, speculum bronze, 279. Rot timber, 606. See Wood, preservative processes. Rotary squeezer, 109. Rotating puddler, 98-103. Roughing blocks, wire drawing, 121. rolls. III, 112, 114. Round turn, 697. Rubble masonry, 589, 591, 592, 594, 596 Rubidium, cost, 268. melting point, 36. Ruby, hardness, 573. Running-out fire. See Refinery. Russian hemp rope, 695. sheet iron, 131. Rust. See Oxidation. Ruthenium, 267, 268. Safety factors, 342, 343. boilers, 394. brass, 562. copper, 342, 562. ordnance bronze, 562. lead, 562. shock, 342. steel, 342. stones, 575. woods, 636, 653. Salisbury cast irons, tests, 375, 378. Salt, common, in roasting ores, 54. in tin ore reduction, 235. Salts of aluminium, 256. antimony, 253. bismuth, 253. deliquescent, for wood preservation, 683. earthy, for fire-proof preparations, 683. lead, 247, 252. metallic, for wood preservation, 676. in ore roasting, 54. zinc, 245. Sand as a flux, 8. INDEX. 745 Sand as a flux, in malleableized iron, 190. Sandal wood, for furniture, 660. Sandstone, 564, 574-577. artificial, 582. See Beton, Concrete. Sap circulation, 597, 598. wood, 597, 598, 603, 606. torsion coefficients, 656. Sardini, 660, 662, 663. Satin wood, 660, 662, 663. Saxon tin, analysis, 237. blast furnace, 236. Scarfing, 664, 665. Schinz, analysis of gases in ore reduction, 65. Schuchardt, prices of metals, 269. Scrap for merchant iron, 107. Screws, hold upon woods, 669. Seasoning. See Timber seasoning. Sectional reduction. See Tensile tests. Seebohm, temper for steel, 193, 194. .Seely, wood preservative process, 678,679. Sellers, regenerator, 74. rotating puddler, 100, loi. steam hammer, 49. Sets, 346-349. cast irons, 390, 391. Salisbury, 375, 378. continuous stress, 548, 551. copper, transverse tests, 433. decrease under test, 547, 548. See Elastic limit. springs, 492. See Strain diagrams. static loads, 552. steel, 559. time variations, 516-520. tin, transverse tests, 439. wrought iron, 362. zinc, 443. Shafting, distances between bearings, 409-411. Shakes in timber, 605. Shapes from rolling mill, 117-120. Shearing, 344. resistance, brass bolts and plates, 430. Shearing resistance plate, 463. bolts, brass, copper, iron, 430. copper bolts and plates, 430. iron, 405, 406. bolts, plates, sheets, 430. steels, 405, 406, 430. woods, 644-646. Shear steel, 140. Sheathing alloy of copper, 23I, 232. brass, composition, 297, 298, 330, bronze, 276. copper, 231, 232. zinc, 245. Sheppard, iridium ore reduction, 264. Sheet brass, 292, 293, 297, 316, 319. copper, 223, 229-232, 296. iron, 117, 118. See Plate iron, lead, 247, 251. metal, properties, 16. nickel, 254, 255. tin, 238. zinc, 243-245. Sheet bend and toggle, 696. Shingling, 105. Corts' early patent, 48. Ship-building woods, 660. sheathing brass, 297, 298. Muntz metal, 290, 330,480, 530, 531. worms, 606, 607. Shock, W. H., strength of riveted work, 408. Shock, 341, 349-352, 514-516. effect upon ductility, 492, 493, 496. crystallization by, 504. fracture characteristics, 493. impact, 349-352, 432. See Resilience, safety factors, 342. Shrinkage of castings, 198. brass, 291, 317. See Timber. Siberian iron and steels, 177. Siderite, 51, 52. Siemens, direct-process ore reduction, 47, 85-87. in reverberating furnace, 47, 97. 746 INDEX. Siemens, direct-process ore reduction in regenerative furnace, 48. furnace for crucible and open-hearth steels, 145, 146. zinc smelting, 246. rotating puddler, 100. steel, classified, 134. heat affecting tenacity, 492. Siemens-Martin, direct ore reduction, 87. iron, classified, 136. for wire, 125. steel, 146, 147-150. classified, 134, 136. Sienite, 564, 566. crushing resistance, 574. effect of exposure to fire, 572. See Granite. Sienitic gneiss, 566. Signals, magnesium, 262. Thurston's apparatus, 262, 1263. Silesian process, zinc manufacture, 241, 242. Silica in fire-brick, 581. as a flux, 8. for copper ores, 2lg. Silicate of alumina in slag, 9. lime in bricks, 578. magnesia, soapstone, 568. nickel, 254. sodium for wood preservation, 675. zinc ores, 6, 240, 246. Silicious calamire, 240. stones, 564. Silicic acid in slag, 186. Silicon bronze, comparative conductivity and strength, 20. cementation changes proportion, igi. in iron, 176. armor-plate, Lowmoor, 187. Bowling, i83. cast iron, 180-182, 188, 189, 196. ingot, Barron, 195. malleable or weld, 187. malleableized, 190. meteoric, igo. puddled, at different stages, 1S2-186. Swedish, 187. Silicon bronze in wrought iron, 187 i8g. steel, 176, 177, 193. Bessemer, 195, 196. crucible, 191, 192. manufacture, 170, 172. Silver, 266. alloys, 138. atomic weight, 29, 446. Chinese, 330. cost, 269. crystalline form, 26. ductility, 25. elasticity modulus, 447. variations with temperature, 533. electric conductivity, 18-20, 26. expansion by heat, 30, 33. hardness, 17. heat conductivity, 18, 26. historical, 4, 41. latent heat, 37. lustre, 20. malleability, 25. melting point, 26, 35, 36. sound conduction, 446. specific gravity, 22, 23, 26, 446. specific heat, 26, 27, 29. solder, 328, 331. tenacity, 446. affected by heat, 532. weight, 23. -aluminium alloy, 309. Siris bronzes, 12. Slag in Bessemer steel rails, 195. composition, 9, 186. in copper, 221-223. hammering to remove, 48. iron, 187-189. puddling furnace, g6, 186. refineries, 90, Slate, 571, 572. clay, 571. crushing resistance, 574. mica, 566. in fire stones, 572. See Granite. Slime of tin ore, 235. INDEX. 7A7 Sling or strop, 697. Slitting rolls, 115. tests, tool steels, 374. Smelting, copper ores, 218-221. furnace, Corts', 48. galena, 249. lead ores, 248, 249. tin ores, 234, 235. zinc ores, 240. Smith, strength of T and angle irons, 403. wood columns, 640. Snakewood, uses, 659, 663. Snelus, basic-dephosphorizing process, 165. Soaking pit, 169. steel while tempering, 204. Soapstone, 568. resistance to heat, 577, 578. Soda, carbonate, 7. in reduction of nickel ores, 254. flux for silica, 8. Sodium in aluminium manufacture, 257- atomic weight, 29. cost, 269. hardness, 1 7. manganese fluoride reduction, 265. melting point, 36. silicate, wood preservative, 675. specific gravity, 22. specific heat, 29. Solders, 326-329. brass, 291, 298, 299. broom, 328. copper, 230, 231. hard, 326-328. Haswell's table,, 331. metals used, 35. soft, 326-328, 331. Spence's metal, 314. Soldering fluid, 327, 328. iron, 326. Solubility, aluminium, 256, 258. bismuth, 254. brass, 292. bronze, 282. iridium, 264. iron, 331. Solubility, platinum, 261. Spence's metal, 314. Solution of metals in aqueous chemicals, and in mercuiy, 7. Sonorous alloys. See Brass, Bronze, Cop- per-nickel, Kalchoids. Sorby, microscopic investigations of iroff and steels, 505. Sound conduction, 446. Sparry ores, roasting, 57. Spathic ores, 51, 67. Specific gravity, 632. aluminium, 256, 257. aluminium bronze, 307, 308. annealing, effect on bronze, ssfr 538. arsenic, 264. azurite, 217. bismuth, 253. blast furnace materials, 67. bornite, 217. brass, 289, 296-299, 461. bronze, 274, 276, 280, 282-287, 452, effect of annealing and tempering. 536-538. ordnance, 448. chill casting, 535, 536. carbonate of copper, blue, green, 217.' iron, 51. chalcopyrite, copper pyrites, 217. charcoals, 67. chill casting, effect upon bronze, 535, 536. cinnabar, 259. coal, 67. coke, 67. copper, 22, 26, 215, 216, 227, 274. 282, 296, 446, 452, 461. native, 217. ores, 217. cuprite, 217. • See Density, ferrous carbonate, 51. fir wood, 614. Gmhite, 51. German silver, 310. gold, 446. 748 INDEX. Specific gravity, gutta percha, 694. heat, effects upon bronze, 536-538. hematite, 51. brown, 67. red, 51, 67. iron, 446. armor plate, Lowmoor, 187. carbonate, 51. cast, 23, 24, 198, 199, 374-376, 389- gray, 198. ordnance, 375. ores, 51. pure, 197. tenacity, relations between, 389. wrought, 197. iridium, 264. lead, 247, 446. lime ; limestone, 67. limonite, 51. Lowmoor armor plate, 187. mahogany, 631. magnetic ore, 67. magnetic oxide, 51. magnetite, 51, manganese, 265. mercury, 258. metals, tables, 22, 23, 26, 446. nickel, 254. oak, white, 623. pine, white, 608. platinum, 261, 446. platinum-iridium alloy, 313. . pyrites, copper, 27. rhodium, 267, siderite, 51. silver, 446. spotthic ores, 51. steels, 199. tool, 374. tempering, effect, 536-538. thorium oxide, 267. tin, 236, 287, 440, 446. turgite, 51. xanthosiderite, 51. See Weight, zinc, 245, 299, 446, 462. Specific heat, aluminium, 26, 2g. antimony, 26, 28, 29. arsenic. 29. bismuth, 26, 28, 29, 253. blast furnace materials, 67. brass, 28. cadmium, 29. charcoal, 67. coal, 67. cobalt, 27. carburetted, 27. coke, 67. copper, 26, 27, 29. gold, 26, 28, 29. hematites, 67. iridium, 26, 29. irons, 26, 29. cast, 27, 198, 199. wrought, 27, 197. lead, 26, 28, 29, 24.7. lime ; limestone, 07. magnesium, 26, 29. magnetic ore, 67. maganese, 26, 28. mercury, 26, 28, 29, 258. metals, tables, 26-30. nickel, 26, 27, 29. carburetted, 27. palladium, 26, 29. platinum, 26, 28, 29. potassium, 29. silver, 26, 28, 29. sodium, 29. spathic ores, 67. specular ore, 51, 54. steel, 27. soft, 27. tin, 26, 27, 29. tungsten, 28. zinc, 26, 27, 29. Speculum metal, 330. , bronzes, 38, 276, 277, 279, 285, 330. Kalchoids, 300, 330. lustre, 20. Spelter, 331. Spence's metal, 314. Spencer, rotating puddler, 100. INDEX. 749 Spiegeleisen, 265, 266. See Manganese, in manganese bronze, 303-306. microscopic investigations, 505, 507. steels, 149, 313. Spikes, resistance to penetration and ex- traction, 668. Spinell-hardness, 573. Spinning sheet metal, 294. Spirits of turpentine, 620. Splicing ropes, 695. Spongy platinum, 260. Sprague and Tower, strength of stayed plates, 379. Springs, set, 492. tempering ; temper, 205, 208. Spruce, American black, 613. fir, transverse s'.rength, 647. black, 613, 614. torsion coefficients, 656. California, crushing resistance, 639. tenacity, 637. transverse strength, 647. crushing resistance, 638, 639, 643. deflection, American fir, 651. elasticity modulus, 653. fir, American, deflection, 651. transverse strength, 647. hemlock, 613, 614. Norway, 613. red, 614. shearing resistance, 645. hemlock, 613, 614. New England, tenacity, 637. Norway fir, 613. pine, 609. shearing resistance, 645. torsion coefficients, 655, 656. red fir, 614. resilience, torsional, 656. shearing resistance, 645, 646. tenacity, 637. torsional coefficients, 655, 656. resilience, 656. transverse strength, 647, 653. uses, 660. Squeezing; squeezers, 105, 109, 11 1. Squirting, brass, 292. lead pipe manufacture, 251. Stain , black, for wood, 673. Stamping brass, 316. tin ore, 235. Stannic oxide ; stannite, 234. Stannite, 23. Star shakes in timber, 605. Statuary bronze, 278, 321-324, 330. Mackenzie's Kalchoid, 300. Stay bolts, stayed surfaces, 379-382. Stays, girder, 381. Steam boilers, copper, 232. flues, strength, 393, 394. heads, 382, 383. heating surface, 81. shells, strength, 382-384. stayed surfaces, strength, 379-382. See Plate iron, types, 81. engine, influence upon iron, manufac- ture, 48, 49. hammer, 48, 49, 108, 109. hoist, 81. pipe, strength, 384. Steel, 39, 132-169, 171-178, 190-196, 199-214. alloyed with aluminium, 309. other metals, 177. analyses. See composition, annealed, 177, 205, 209. condition of carbon, 497, 498. antimony in, 177. basic dephosphorizing process, 165, 167, i6g. Bessemer, analyses, 194-196, carbon in, 191, 194-196. copper in, 191. classification, 134, 136. manufacture, 49, 149, 151-169. tenacity, 356, 363, 371. variations with temperature, 491. torsion teats, 415, 420, 421, 424, 425. blister, 137-140. brazing, 328. carbon in, 168, 172, 193. 750 INDEX. Steel, carbon in annealed and unannealed steels, 497, 498. Bessemer steels, 194-196. cementation steels, 137-141, 191. chisel temper ; die temper steels, 194. ductility affected by, 197. hardening effects, 497, 498. heat effects, 204. malleability affected by, 197. percentages, 133, 135, 136, 191-193, 195. 196, 497, 498- tempered steels, 193-195. tenacity affected by, 369-371, 373. unannealed steels, 497, 498. classification, 132-136. cast (j« crucible), 140-151, 494, 506. castings, importance of annealing, 127, cementation, 137-141. [169, 497. See cast steel, crucible steel. carbon proportion changes, igl. characteristics, 151, 167. chemical properties, 137-139. chipping tests, 274. chrome, 207, 209. classification, 132-136. cobalt in, 177-193. compositions, 133, 135, 136, I91-193, 195, 196, 497, 498. compound, 135-144. compression while cooling, 210. compressive resistance, 374, .392. copper in, 177-193. corrosion, 211, 214. crucible, 141, 142. analyses, 191-193. See Cast. castings, 145. cementation, 137-141, 19I. classified, 134, 136. compounds, 144. elongations, 372. furnace, 142, 143. manufacture, 136. See open hearth. properties, 144, 145. temper, 144. tensile resistance, 372. Steel, crucible, torsional tests, opp. 415. 423-425. uses, 144, 145. Woots, or Indian, 145. crystallization, 503, 505, 506. direct process, 149-161. See Bessemer, drilling tests, 374. ductility affected by time, 511, 512. elasticity moduli, 347. variations with temperature, 533. elastic limits, 367, 559. elastic resilience, 373. electrical resistance, 19. elements in, 135. elongation, 135, 372, 373. 573- expansion by heat, 30, 34, 199. fracture, appearance, 424-426. German, temperature and torsion tests, 494- gold in, 177. grading by eye, 134, 135. hard, 24, 135, 136. hardness, 16, 17. modifications by heat, 204. hardening. See tempering, heat. See annealed. carbon proportions affected by, 204. density modified by, 200-210. elasticity moduli variations, 533. expansion, 30, 34, 199. hardness variations, 204. specific heat variations, 27. strength affected by, 493. temperature modifications, 534. tempering, 205, 209. tenacity variations, 491, 492,494,532, Heaton dephosphorizing process, 165. Henderson processes, 151, 165. ingot, classified, 133, 136. tensile resistance, 369-371. * See Iron, Kelley pneumatic process, 152. See Bessemer, lime, presence affecting strength, 511, 512. lustre, 20. INDEX. 751 Steel, magnetic condition, 199, 498. malleability affected by carbon, 197. manganese, effect of, 142, 143, 167, 173. 174, 191. 193, 195. 373- manufacture, 132-169, 171-178, 190- ig6, 199-214. medium, temperature and torsion tests, 494. melioration by exposure, 498, 499. microscopic appearance, 505, 506. mild, analyses, 193. carbon percentages, 135. elasticity moduli, 347. elongation, 135. tenacity, 135. torsion tests, opp. 415. uses, 135. weight, 21. nickel in, 177, 193. nomenclature, 132—136. open hearth, classified, 134. manufacture, 145, 149. See Siemens-Martin. tensile tests, 373. oxidation, 210-214. phosphorus in, 165, 166, 174, 17s, 191. 192, 195. pneumatic processes. See Bessemer, properties, 199. puddled, 104. carbon percentages, 135. classification, 134, 136. shearing resistance, 430. safety factors, 342. sets under strain, 559. shear, 134, 136, 140. torsion tests, opp. 415, 423- shearing resistance, 405, 406. Siemens-Martin, classified, 134-136. manufacture, 145, 149. See Open hearth, torsion tests, opp. 415, 420. silicon in, 176, 177, 191. ^93, 195- slotting tests, 374- soft. See Mild, solder, 328, 331. spring tempering, 206. Steel, strain diagrams, autographic, opp. 415, 419-426. structure, 411. specific gravity, 23, 27, 199, 203, 204, 374- specific heat, 27. steam boiler heads and shells, strength, 383. stayed plates, strength, 379, 381, 383. sulphur in, 175, 176, 191, 193, 195. tempered, 193, 194, 203, 204. tempering, 199, 208-210. . tenacity, 15, 370, 372-374.' formula:, 371. variations with temperature, 491, 492. test pieces, forms and dimensions, 352- 358. tilted, 140. time, effect upon ductility, 511, 512. tin in, 177. titanium in, 177. tool, analyses, 191, 192. carbon in, 135, 193, 194. chipping tests, 374. compressive tests, 374. drilling tests, 374. elasticity moduli, 347. magnetism, 199. melting point, 199. planing tests, 374. slotting tests, 374. strain diagrams, 422-424. specific gravities, 374. tempers for, 208. tensile tests, 374. torsion tests, 374. turning tests, 374. torsion tests, 374, opp. 415, 419-426. tungsten in, 267. turning tests, 374. weight, 23. weld, 133. 134. 136. welded to nickel, 255. welding point, 199. Wolfram, tempering, 208, 209. Stereotype metal, 312, 324, 325, 330. Sterro metal, 302, 458, 465. 752 INDEX. Stiffness, 650-654. aluminium bronze, 307. beech ; birch, 651. brass, 461, 462. columns, 395-405. flexure, 516-520. See Transverse, Torsional tests. Stones, abrasive resistance, 573. absorption of water, 576. artificial, 578. ' See Beton, Brick, Concrete, heat resistance, 572. sandstone, 582. classification, 564. crushing resistance, 573, 574. durability, 575-577- fire, effect of exposure, 572, 577. foundations, 563. frost resistance ; test, 576. hardness, 573. masonry, 588-596. siliceous, 564. transverse strength, 575. water absorption, 576. weight, 577. Stone breaker, 9. Stone work, 588-596. Stopper, rope, 697. Strain diagrams, 346-352, 411-424. See Torsional tests. Strain, evidenced by subsequent tests, 522-527. intermittent, ellect, 557, 558, 561-563. internal, 412, 413. See Load. orthogonal, 526, 527. rest, effect upon resistance, 510, 511, 520-522. See Stress. Straits tin, 234. Stream tin, 234, 235. Strength. See Resistance. Stress, 344. continuous, 512, 514, 548-551. elastic limit, tin class, 552. variable, 560, 561. intermittent, bronze, 557. Stress, intermittent, tin class, 552. internal, relieved by rest, 5 10, 511. See Load. maximum for non-ferrous metals, 562, prolonged, bronze, tin, zinc, 542-546. pine, 657, 658. resistance to, method of, 345. See Resistance, static, iron wire, 512. See Strain, Strain diagrams. Stretch, temperature effects, 533, 534. unintermittent andheavy, brass, bronze, tin, 548, 550. variable, bronze, 547, 548. wood, prolonged stress, 657, 658. Stretch. See Elongation. Strontium, 267. cost, 268. Strop or sling, 697. Struts. See Columns. StUckofen, iron furnace, 45. Styffe, effect of temperature upon strength of iron, 491. Sugar-maple tree, 630. Sulphate of baryta, an adulterant of white lead, 675. copper, wood preservative, 676, 680, 684-686. iron, as a paint, 674. lead, an adulterant of white lead, 675. lime. See Gypsum, nickel, 254. zinc, a wood preservative, 648. Sulphides, 371. of bismuth, 252, 254. cadmium, 266. copper, 6. lead, 248. See Galena, Lead, iron, 6. in slag, 186. mercury, 259. ores, 6. reduction processes, 7. metals, 37. nickel, 254. zinc, 6. INDEX. 753 Sulphite of arsenic, 263. Sulphur in bituminous coal, 57. cementation changes of proportions, 191. in iron, Barrow ingot, 195. Bowling, 188. cast, 180-182, ig6. castings, 175. chain, 188, 189. Lowmoor' armor plate, 187. malleable, 187. malleableized, 190. meteoric, 190. puddled, at different stages, 185. Swedish, 187. ores, 54. steels, 193. Bessemer, 195. crucible, I91, 192. vulcanized caoutchouc, 694. Sulphuret of antimony, 253. Sulphuretted copper ores, 219. Swedish Bessemer steels, analyses, 195. iron, 187. ores, 50, 55. strength at different temperatures, 491. Swiss copper, 228. Sycamore wood, grain, 662. hold upon screws and spikes, 668. uses, 660. Symbol, aluminium, 256. antimony, 247. arsenic, 263. bismuth, 253. copper, 215. iridium, 264. lead, 247. magnesium, 262. manganese, 265. mercury, 258. nickel, 254. platinum, 260. tin, 234. T. Talc in granite, 564. hardness, 573. 48 Tank iron. See Plate iron. tenacity, 363, 364. Tangye, compressive tests, wrought iroo, 392. 393- Tantalum, 267. cost, 268. fusibility, 36. Taps, temper, 208. Tar, 617, 618. as a paint, 674. for wood preservation, 676-682. Tarred ropes, 695. Taste of metals, 24. Teague, utilization of waste products of puddling furnace, 49. Teak tree, 632, 633. wood, 607. columns, strength, 641. crushing resistance, 638. deflection, 651. elastic limit, 636. tenacity, 637. transverse strength, 647. use, 660. Teeth of wheels, woods for, 660, 661. Telluric iron, 52. Tellurium, cost, 269. hardness, 17. melting point, 36. specific gravity, 20. Temperature in blast furnace, 69, 70. See Cold. corrosion affected by, 542, 543. elasticity moduli affected by, 532, 533. fracture, characteristics, 493-496. See Heat. irons, effect upon, 493-496. metals, effect upon, 16, 489-498, 534- 536- specific heats, 27-30. stresses produced by changes in, 533, 534- for tempered steels, 206. tenacities modified by, 493-496. Tempering baths, 200, 206, 207. brass, 24. bronze, 24, 275, 278, 279, 282, 536-538. 754 INDEX. Tempering, by compression, 210. copper, 24, 227. See Hardening, springs, 205-207. steels, 24, igg, 200. cast, ig3, 194. crucible, 144. tool, 200-210. Temper, tin, 239. Tensile tests : Aluminium bronze, 307, 309. antimony, 252, 446. area reduction. See sectional reduc- tion, beech, 637. belts, 687-691. birch, 637. bismuth, 446. blanks for records, 360. boxwood, 637. brasses, 289, 296-299, 457, 461, 462. elastic limit, 461, 462. formulae, 476. bronzes, 276, 282-287, 448-455, 458, 480. formulae, 474, 476. strain diagrams, 454. manganese, 303-306, 450. ordnance, 448, 449, 458, 481. variations with temperatures, 494, 529-531- carbon, effect of ; formula, 376, 369- 371, 373- , cast iron, 15, 369, 371, 374-376, 389. variations with temperature, 494, 495- cedar, 637. copper, 15, 226, 227, 232, 282, 296, 429, 436, 446, 452, 453, 461, 481, 528, 529. area reduction, 452, 461. cast, 232. elastic limit, 436, 452, 461. temperature modifications, 528, 529. wire, 232. density relations, 3S4. elastic limit, brass, 461, 462. Tensile tests : elastic limit, copper, 436, 452, 461. elongation, brass, 461, 462. copper, 436, 452, 461. tin, 441, 453. zinc, 462. fir, 637. gold, 446. variations with temperature, 532. greenheart, 637. hardening elements increase it, 197. heat reduces, 197. • See temperature, holly, 637. horse chestnut, 637. iron, 446, 491, 492. admirality tests, temperature varia- tions, 492. armor plate, Lowmoor, 187. boiler plates, 119, 363, 364. bridge plates, 373. cast irons, 374-376, 389. compressive resistances compared, 389- ordnance, 375. Salisbury, 378. different elements affect, 197. flange, 363. forged, large size, 427, 428. formulas, carbon percentages, 371. ingot, 369, 370. Lowmoor, 187. tank, 363, 364. temperature modifications, 528, 529. wire, 123, 366, 367. annealed, 497. cables, 51 1, wrought, 15, 123, 197, 358, 362, 368, 369, 427, 428. Edgemoor Iron Co. , formula, 365. temperature modifications, 4jjo- 492, 528, 529. Kalchoids, 476, 478, 484, 485, 487, 488. temperature modifications, 530, 531. lance wood, 637. larch, 637. INDEX. 755 Tensile tests, lead, 15, 247, 444-446. manganese bronze, 450. metals, 15, 494, 495. mortars, 587. Muntz metal, 457. nickel, 446. ordnance bronze, 282-287, 448, 458, 481. palladium, 446. variations with temperature, 532. phosphorus copper, 429. iron, 427, 428. plate iron, 119, 363, 364. sectional reduction, boiler plate, 364. brass, 461, 462. bronze, 452. chilled, 535, 536. copper, 452, 461. plate iron, 364. tin, 452. wrought iron, 362, 364. zinc, 443, 462. shock, effects, 492, 493, 496. silver, 446, 532. size modifications, 365, 366. steels, 15, 369, 497. annealed, 370, 497. Bessemer, iii, 356, 363, 365. carbon in, 369-371. 373- crucible, 372. tool, 374. sterro-metal, 458, 465. temperature modifications, 197, 492- 496. 530-532. 534-536. test pieces, forms and sizes, 365-367. tin, 15, 236, 287, 437, 440, 441, 446, 452. woods, 637. working metal, effects, 367. zinc, 15, 299, 442, 443, 446, 462. Terbium, 267. Teredo-navalis, 606, 607. Ternary alloys, 310, 312. See Kalchoids. Terne plates, tin, 131, 239. Terras in hydraulic cements, 583. Tertiary rock, 51. Testing blanks, 360, 361. methods, 358-361. Test-pieces, forms and dimensions, 353- 359. inspection of tested pieces, 417-419, 424-426. Texture iron, 198. See Fracture. Thallium, 267. cost, 268. electrical resistance, 19, 20. melting point, 36. Thermal conductivity, brass. \^See Heat conductivity.] 20. metals, tables, 17, 18, 26. Thermal. See Heat. Thickness, pipes, cylinders, flues, 79, 384, 388, 394. Thomas, basic dephosphorizing process, i65._ Thomson, investigations of bronzes, 282- 287. Thorium ; oxide, 267. Thurston, age effect upon metals, 498- 500. analyses, copper, 431-436. tin, 237, 436-441. annealing of wire, effects, 497. See Blair, blanks, forms for records of tests, 359- 362. brasses, investigations of properties, 459-463. bricks, resistance of, 580. bronzes for bearings, 279-287, 448- 455- investigation of properties, 274. maximum, 482-488. prolonged stress, effect, 542-546. carbon percentages, tenacity formulae, 370, 37t. cast iron, in compression, 390, 391. formulae for tenacities, 376, Salisbury, tests, 375. strain diagrams, 378, 391. temperature, effect of strength, 494- 496. ;s6 INDEX. Thurston, compressive resistances, cast irons, 390-391. effect of speed, 432. woods, 630. copper, analyses tests, 431-436. elasticity moduli, 434. oxygen absorption, 225-226. decrease of resistance with time, 55°- 556. elasticity of pine, effect of prolonged stress, 657. moduli of copper, 434. elastic limits of irons, 375, 376, 522- 526. variations, 520, 522-526, 556-561. endurance iron wire under static load, 513- forms and proportions of test pieces, 353, 357, 358. for records of tests, 359-362. formulae for shafts, steam engine, 410. tenacities of irons and steels, 370- 371. fracture of test pieces, 424-428. heat. See temperature, iron. See all entries, kalchoids, investigation of properties, 464-488. lead, strength, 445. magnesium lamp, 262, 263. signal apparatus, 262, 263. maximum stress for non-ferrous metals, 562. , mercury bath for tempering, 200. nickel tenacity, 446. non-ferrous metals maximum stresses and safety factors, 562. . orthogonal strains, 526, 527. oxygen absorption of copper, 225, 226. pine, effect prolonged stress, 657. planning researches, 473, 479. prolonged stress, effect upon pine, 657. proportions of test pieces, 353, 357, 358. punching of wrought iron, 408. Queensland tin analysis, 237. rate of set of metals, 516-520. Thurston, records of tests, blanks, 359- 362. safety factors for non-ferrous metals, 562. Salisbury cast irons, 375. sets of metals, rate, 516-520. temperature modifications, 492. shafts for steam engines, formula, 410. shearing, wrought iron, resistance, 405. speed. See velocity, steel, carbon percentages and tenaci- ties, 370. tempering in mercury bath, 200. strain diagrams, 410-426. cast irons, 378. in compression, 391. stresses, maximum, non-ferrous metals, 562. prolonged effect upon bronze, tin and zinc, 542-546. pine wood, 657. unintermittent and heavy, effects, 548-550. temperature, effect upon strength of cast irons, 494-496. modifications of sets, 492. tenacity effects upon, 494, 495. tempering steel in mercury bath, 200. tenacities, formulae for cast irons, 376. iron and steel, 370-371. test blanks for records, 359-362. pieces, fracture inspections, 424- 428. proportions, sizes, 353, 357, 358. time, effects upon bronze, 546, 548. elastic limits, 320, 522-526. prolonged stress on bronze, tin, zinc, 542-546. varying resistances, 432, 514-520. - wire, 512-514. tin, analyses, 436-441. Queensland, 237. tests, 436-441. effects of prolonged stress, 542-546, torsional, 441. transverse, 439-441. torsional, metals, 410-426. INDEX. 717 Thurston, torsional, woods, coefficients, 655, 656. transverse tests, tin, 439-441. woods, 646, 647. unintermitted and heavy stress effects, 548-550. variations of elastic limits, 556-561. compressive resistances, with speed, 432. resistances with velocity, 434, 515- 520. velocity effects. See Time and Varia- tions, wire, effects of annealing upon strength, 497. resistances to wire drawing, 122. woods, compressive resistance of pine, 630. prolonged stress, effect, 657. torsional coefficients, 655, 656. tranverse resistances, 646, 647. See other entries, wrought iron, punching and shearing resistances, 405. tensile tests, 362, 363. zinc, effects of prolonged stress, 542- 546. tests, 442-444. Tile, 105. Tile copper, 222. Tilted steel, shear steel, 140. Timber, air seasoning, 599, 600. brash wood, 604. characteristics of good, 603, 604. climatic effects, 604, 605. cutting. See Felling, decay, 605, 606, 634. See preservation, defects, 605. dry rot, 603. elastic limits, 636. felling, 635. seasonfor, 599. white oak, 623. yellow pine, 610. marine animals which attack, 606, 607. measurement, 634. Timber, nomenclature, 602, 603, 605. oil seasoning, 601, 636. pine, qualities, 612. preservatiqn, 605, 606, 670-686. seasoning, 509-602, 635. air, 599, 600. hot air, 600. oil, 601, 636. steaming, 600. shrinkage, 601, 602. soil influences, 604, 605. strength, 635-669. trees, 597-634. varieties, 607-634. water contained, 601. See Wood. Timber hitch, 696. Time, brass tests, 540, 548, 551. bronze tests, 546, 548. continuous stress, effects, 512-514. copper tests, 540. decrease of resistance of metals with, 548-551. deterioration of metals, 229, 230. ductility of copper, 229, 230, steel, 511, 512. elastic limit, 520, 522. of bronze, 556-561. wrought iron, 520, 522-526, 556, 561. See Flow, rupture, 539, 562. strength affected, 539, 562. wire cables after use, 511. transverse tests, 555. See Variations, wire tests, 512-514. Tin, 234-239. Alabama deposits, 239. alloys, 239. with aluminium, electric resistance, 18. antimony and lead, 312. zinc, 312. copper. See bronze, and ircSn, 302, 303. iron, 177. 758 INDEX. Tin, alloys with iron and copper, 302, 303. lead, 329. electric conductivity, 18. for organ pipes, 311. zinc, 311, 312. heat expansion, 33. analyses, 237, 239, 437. ■with antimony and lead, 312. antimony and zinc, 312. area reduction under tensile stress, 452. atomic weight, 29, 446, Banca, 234, 239. analyses, 237. tests, 439, 444. See Bronze, casting, 437. in cast iron, 178. cold struts formed in casting, 437. compressive resistance, 438, 446, 452. copper alloys. See Bronze, and iron alloys, 302, 303. cost, 269. deflections, 439-441, 452, 554, 555. ductility, 24, 25, 236, 237, 287, 452. elasticity moduli, 439, 441, 446, 452. elastic limits, 438, 441, 432. depressed with intermittent strain, 552. electric conductivity, 18, 19, 287 elongations, 441, 452. enamel, 237. <, expansion. See heat, 30, 33, 34, 237. foil, 237, 238. fusibility, 287. hardness, 17, 287. heat conductivity, 17, 18, 287. expansion, 30, 33, 34, 237. iron alloy, 177. and copper, 302, 303. latent heat, 37, 237. lead. See alloys. electric resistance, 18. malleability, 25, 236, 287. melting point, 35, 36, 237. pipe, 238. Tin plate, 128-131, 238, 239. prolonged stress, effects, 542, 544. Queensland, analysis, 237. effect of time upon re^stance, 548-" 552. tests, 439-441. resilience, torsional, 441, 452, 453. resistance, affected by time, 548-552. sets, transverse tests, 439. sheet. See plate, solder, 33, 331. sound conductivity, 446. specific gravities, 22, 23, 26, 236, 287, 440, 446, 482. specific heat, 26, 27, 29, 237. strength, 436-441. affected by heat, 237. prolonged stress, 542—544. time, 540. varied velocities, 542-544. structure, 287. tenacity, 15, 236, 287, 437, 440, 441, 446, 452. See ductility, elongation, 441, 452. time tests, effect on resilience, 548-552. torsional tests, 441, 452, 542-545. transverse tests, 439, 440, 441, 452, 554, 555- zinc alloys, 311, 3T2.» and antimony, 312. Tinned copper, 227. T-iron, 117. Titanium, 267. cost, 269. fusibility, 36. hardness, 17, steel, 177. Tobin's alloy, 487, 488. Toggle and sheet bend, 696. Tombac, 289, 296, 297, 330. • Tooth-axed stones, 590. Tools for cutting metals, temper, 208. Tool steels. See Steels, carbon in, 135, 193, 194. chipping tests, 374, compressive tests, 374. INDEX. 759 Tool steels, drilling tests, 374. elasticity moduli, 347. magnetism, igg. melting points, igg. planing tfests, 374. slotting tests, 374. specific gravities, 374. strain diagrams, 422-424. strength, summary of tests, 374. tempers, 208. tensile tests, 374. torsional tests, 374, opp. 415, 423-425, 408-411. turning tests, 374. Topaz, hardness, 573. Torsion of shafts, 408-411. Torsional strains, 344. Wertheim's laws, 657. Torsional tests by autographic machine : brass, 461-463. bronze, 452-454, 545, 546. copper, opp. 415, 435, 436, 452, 461, 540. iron, cast, 374, opp. 415. 4g4, 4g5. cold-rolled, opp. 415, 421. malleableized, opp. 415, 425, 426. wrought, American high grade, 413- 420. special grade, 413-320. English, best, 413-420. low grade, opp. 415, 4g4. Swedish, opp. 415-417. kalchoids, 473-475. 484. 485- record blanks,- 361. steel, Bessemer, American, opp. 415, 421, 424, 425. English, opp. 415, 420. cast, 4g4. crucible, American medium, opp. 415- English " German," opp. 415, 423-425. German, 4g4. medium, 4g4. shear, English double, 415, 423. Siemens-Martin, opp. 415, 420. American, opp, 415, 420. Torsional tests by autographic machine, steel, soft, opp. 415. spring, American, opp. 41 5, 423. tool, 374. American, opp. 415, 424, 425. English, opp. 415, 424, 425. annealed, opp. 415, 423. tin, 441, 452, 542-545. woods, 654-657. zinc, 444, 462, 545. Townsend, flow of metals beneath a punch, 508-510. Tracing cloth, 693. Transverse tests, 403, 404. blanks for records , 360. brasses, 403, 461-463, 560, 561. bronzes, 451-453. 545, 546. 553-556. copper, 432, 433, 436, 452, 461, 463, 545, 546. dead loads, 545, 546. formula, 403-405, 650-654. iron, 402-405. stones, 575. tin, 439-441. 452, 454, 455. 545, 546. 554. 555- woods, 646-654. wrought iron, 552, 553, zinc, 462. Trees, 597-634. See Timber, Wood. Tubes, brass, 294, 317. copper, 22g, 233. lead, 247. See Pipes. tin, 238. Tulip wood, uses, 660, 663. Tungsten and alloys, 267. in cast iron, 178. fusibility, 36. specific gravity, 22, heat, 27. in steel, 151, 267. Turpentine, 618-620. for paints and varnishes, 670-673. Turgite, 51, 52. See Hematite, brown. Turned work, kalchoids, 334. 76o INDEX. Turning tests, tool steels, 374. tools, temper, 208. wood for, 659. Turquoise, hardness, 573. Tutenag composition, 330, Twaite, corrosion of iron and steel, 213, 214. Type metal, 38, 330. U. Ultimate strength, definition, 341. Underpoling, copper ore reduction, 2ig. Unit, heat, 37. United States Board, investigation of brasses, 459-463. bronzes, 282-287, 451-455- zinc, 443, 444. Universal mill, 115. Uranium, 267. cost, 269. fusibility, 36. Valton, effect of temperature upon iron, 493, 495- Valves, india rubber, 694. kalchoids, 332. Vanadium, cost, 268. Vaporization. See Volatilization. Variations in properties. See Carbon, Cast iron, Copper, Ductility, Elasticity Moduli, Elastic limits. Heat, Iron, Kalchoids, Malleability, Metals, Phosphorus, Specific gravity, Steel, Strain, Strain diagrams, Stress, Sulphur, Tempering, Tensile, Thurston, Time, Tin, Torsional, Transverse, Woods, Wrought iron. Varnishes, 670, 675. Velocity of rupture, effect on resistance, 514-516, 539-562. Verde-antique, 570. Vermillion, 259. Vicat, effects of continual stresses, 512, 513- Viscosity of metals. See Flow. Volatilization of antimony, 253, arsenic, 264. bismuth, 253. by salt, 54. copper, 35, 215. iron, 197. lead, 247, 249. mercury, 258-260. metals, 24. zinc, 241, 242. Voltaic deposition of crystals, 26. inducing corrosion, 213, 214. resistance. See Electric conductivity. Vulcanite, 257. combined with aluminium, 257. Vulcanized rubber, 694. W. Wabbler, rolling machinery, 113. Wade, cast copper, tenacity, 429. casting at different temperatures, 535, 536. ordnance bronze, specific gravities and tenacities, 282-287, 448, 449. remelting effect upon strength, 377. Wagner, definition calcination, 54. Walnut beams, deflection, 651. black, tree, 628. wood, 637. crushing resistance, 638. shearing resistance, 646. torsion resistances, 655, 656. uses, 659. elasticity and toughness, 659. uses, 659, 660. white tree, 628. wood, crushing resistance, 638. shearing resistance, 646. transverse resistance, 647. » Ward discovers cobalt, 15. Warm blast, charcoal furnace, 58-60. Washing ores, 9, lo. tin, 235. Wash pot, tin plate manufacture, 130. Watchmakers' brass, properties, 299. INDEX. 761 Water for annealing steel, 210. blast furnace, 81. hoists, 77-80. in ores, 154. latent heat, 37. pipes, cylinders, flues, strength, 79, 384-388, 393, 394. tempering bath, 200, 208, 209. Waterproof glue, 670. Watt i-beam, 404. Watts, bronzes, properties, 282-286, cubic expansion of solids, 34. properties of metals, 26. Weapons of bronze, Mexican and Peru- vian, 41. Weather exposure of metals, 498-500. Weathering ores, 87. Weems, separating constituents of brass by pressure, 292. Weidermann, brasses, electric and heat conductivities, 296-299. bronzes, properties, 282-287. copper properties, electric and heat conductivities, 282, 296. zinc, properties, 287, 299. Weight camphor wood, 633. charcoal, 67. coal, 67. coke, 67. See Density. fir wood, 614. formulae for estimating, 20, 21. hematite ores, 67. leather belting, 687. lime — limestone, 67. magnetic ore, 67. metals, table, 22, 23. oak, white, 623. pine, white, 608. spathic ores, 67. 51?^ Specific gravity. Welding of iron, 170. Cort's early patents, 48. nickel, 255. platinum, 262. steel, 199, 204. Weld iron, 132. Weld iron, strength affected by tem- perature, 490, 491. See Wrought iron. Weld steel, 133. classification, 136. Welds, strength of, 384. Wellington, resistance of woods to pene- tration and extraction of nails, 668. Wandel, analyses of ancient iron, 40. Wertheim, elasticity moduli of copper, 434- temperature variations, 533. wrought iron, elasticity, 415. palladium, tenacity, 446. properties of metals, 446, 447. tin, strength, 445. torsional resistance, laws, 657. velocity of rupture, effect upon resist- ance, 514. West Indian ebony, transverse strength, 647. Wet drawing, wire making, 121. Wet process in copper extraction, 223. metal reduction, 7. puddling, 90-105. rot, 606. Weyranch, formulae for tenacities of iron and steel, 371. Wharton, nickel analysis, 256. manufacture, 255. Wheel teeth, 660, 661. White alloy. See Manganese-bronze, ash tree, 627. beech tree, 625. brass, 463. composition, 330. cast iron, 90. See Cast iron, characteristics, 82. classification, 133. fuel and influence of, 58. fusibility, 198. properties, 79. specific heat, 189, igg. cedar, 615, 616. durability, 659. .shearing resistance, 646. 762 INDEX. White copper alloy, 309, 310. Chinese, 330. fir, 612, 613. lead, 247, 252. effect of presence of copper, 322. paint, 671-675. metal alloys, 329, 330. oak, 621. American, tenacity, 637. chestnut, 624. crushing resistance, 638, 642. shearing resistance, 646. tenacity, 637. transverse strength, 647. uses, 660, 65i. pine, 608. columns, strength, 641. crushing resistance, 638. deflection, 651. elasticity limit, 636. shearing resistance, 645, 646. tenacity, 637. torsion coefficients, 655, 656. resilience, 656. uses, 659. turpentine, 618. walnut tree, 628. wood, crushing resistance, 638. shearing resistance, 646. wood, crushing resistance, 642. glued joints, 669. transverse strength, 647. uses, 660. ^ Whistle, composition alloys for, 329. Whitwell hot-blast stove, 74, 75, Whitworth, hydraulic compression of castings, 168. Wiedmann and Franz, heat conductivi- ties of metals, 18. Williams, W. M., annealing steel with boiling water, 210. O., fractures of iron at different tem- peratures, 492-493. Willow wood, 637. crushing resistance, 638. deflection, 651. elasticity limit, 636. Willow wood, transverse strength, 647. Wilmot, form of test pieces affecting resistances, 355, 356. tenacity of Bessemer steels, 356. Wilson, analyses of ancient alloys, 272. Winding drum, 77. Wind shakes in timber, 605. Wire, 120-125. brass, 293, 298, 319, 330. formulae for elasticity modulus, 532, 533- wire drawing, 316. copper, 226, 232, 233. formula elasticity modulus, 532, 533. drawing, 120-124. gauges, 123, 124. iron, 367. annealing, effect upon tenacity, 497. elasticity modulus, formula, 532, 533. tenacity, 367. annealing effects, 497. telegraph, 366. time and use, effects upon tenacity, 497. lead, 282. mill. See Rolling mill, nickel, 255. phosphor-bronze, 281. See Rolling. Woestyn and Gamier, molecular specific heat, 28. Wohler, aluminium, discovery, 257. separation, 15. cubic crystal in cast iron at high tem- perature, 504. rupture by repeated stresses, 561. Wolfram, steels, hardening, 208. WoUaston isolation of palladium, etc., 15, 33- Wood, beams. See Deflection, engineering purposes, and construc- tion, 658-661. for carpentry, 659. joiners' use, 659. lathe work, 650. patterns, 659. turning, 659. INDEX. 763 Wood colors, 662, 663. crushing resistance, 638, 639, 642-644. deflection, 650-654. detrusion, 644-646. dyeing, Boucherie process, 683. elasticity limits, 636, elastic kinds, 659. floor beams, Woodbury's formula, 652- 654. grain, 659, 661, 662. hold upon nails, screws and spikes, 667-669. impermeable. See Preservation, incombustible. See Preservation. Lanza, crushing resistance of woods, 642. ornamental, 659, 661-663. pillars. See Columns, preservative processes, 670-6S6. prolonged stress, effects, 657, 658. safety factors, 636. shearing resistance, 644-646. stiffness. See Deflection, strength, 635-669. strong kinds, 660. tenacity, 637. tools to cut, temper, 208. torsional resistance, 646-654. tough kinds, .659, 660. transverse resistance, 646-654. See Timber, Trees. Woodboring worms and insects, 606, 607, 676. Woodbury, loads for floor beams, 652- 654- Wootz, or Indian steel, 145. Working brass, 293, 294. copper, 5, 226, 229-233, 537, 538. iron. See Drawing, Forging, Hammer- ing, Rolling, steel, 200-210. Worms, wood-boring, 606, 607, 676. Wright, methods of ore-reduction Ta- ble, 7. Wrought copper, tenacity, 232. iron, 39-51. 87-131- analyses, 188, 189. Wrought iron, area reduction under ten- sile stress, 362, 364. See Bessemer. boiler plates, tensile tests, 364. boiling, 90-105. British, strength, 392. classifications, 132-136. compressive resistance, 392, 393. copper, effect of presence, 177. corrosion, 2 1 1-2 14. crystallization by use, 501, 502. cylinders and flues, 393, 394. de-carbonizing processes, 87. deflections, 552, 553. elasticity moduli, 347, 362, 368. elastic conductivity, 197. limits, 362, 367, 375, 376,522, 526. elongation, 362, 368, 392. expansion by heat, 197. extension. See elongation. flange iron, boiler plate, 119. forged, tenacity, 363. forge irons, 179. fractures, appearance, 417-419. hardness, 16, 197. heat conductivity, 197. interrupted strains, 557. See Iron, magnetic ores, 57. malleability, 24, 197. manufacture, 7-9, 44-49, 83-131. melting point, 35, 36, 197. microscopic appearance, 505, 506. See Plate iron. pneumatic processes, 186. See Bessemer. properties, 170, 171, -196-198. See Manufacture. puddling, 7, 90-105. re-worked, tenacities, 368, 369. sets, tensile, 362. transverse, 552, 553. shapes, 1 17-120. shearing resistance, 405, 406. sheet iron, 131. See Plate iron. Siberian, 177, 178. 764 INDEX. Wrought iron. See Siemens-Martin, specific gravity, 23, 197. heat, 27, 197. specular ores, 57. strain diagrams, torsional, 413-419. interrupted, 520, 522, 557. structure, 411. microscopic, 505, 506. Swedish, compressive resistance, 392- tank iron, tenacity, 363, 364. tensile tests, 15, 197, 358, 362, 363, 368, 427-428. at different temperatures, 490-492. tools to cut, 208. torsional tests, 413-420, 494. transverse tests, 552, 553. weight, 23. weld iron, 132. strength affected by temperature, 49°. 491- See Wire. Wych elm tree, 627. X. Xanthosiderite, 51, 52. See Hematite, brown. X-iron column, 401. Yellow brass for brazing, 328. bronze, composition, 330. dip turpentine, 618, 619. fir, 611. oxide, tin, 237. pine, 608, 609. wood, crushing resistance, 638, 642- 644. durability in dry construction, 659. elastic limit, 636. elasticity, 659. lightness, 659. prolonged stress, 657, 658. Yellow pine wood, resilience, torsional, 656. safe loads, 649. shearing resistance, 646. strength, 659. tenacity, 637. torsion coefficient, 655. resilience, 656. transverse resistance, 647. Yew tree, longevity, 598. Yorkshire irons, 106. Yttrium, 267. Zebra wood, 660, 663. Zinc, 239-246. alloys, 240. antimony-tin, 312. See Brass. copper-iron, 302. tin, 311, 312. area reduction under tensile stress 443, 462. atomic weight, 29, 243, 446. bronze, addition to, 321. bronzing liquids, 337. coating. See galvanizing, chloride for wood preservation, 676,- 677, 685, 686. compressive tests, 442, 443, 462. copper, presence, 16. See Brass, cost, 269. crystallization, 27. deflection, 444, 462, 463. distillation, 33, 241, 242, 245. ductility, 25, 240, 243, 299, 462. elasticity moduli, 441, 442, 444, 462. elastic limit, 442, 462. elongations, 462, 463. ^ expansion by heat, 30, 34. electric conductivity, 19, 26, 299. voltaic element, 245. flow during tests, 545. fusibility, 299, hardness, 17, 299, 573. INDEX. 765 Zinc, heat conductivity, 17, 26, 299. effects, 243. expansion, 30, 34. historical, 4, 239, 240. malleability, 24, 240, 243-245, 299. melting point, 35, 36, 245. paint, 674, 675. preservatives, 245, 674-677, 684-686. resilience, 462. rest under stress, 545. reverberating furnace, 54, 243. sets, 443. silicates, 6, 240, 246. specific gravity, 22, 23, 26, 462. heat, 26, 27, 29. Zinc stresses, slow, rapid, rest, 545. structure, 299. sulphate for wood preservation, 684. sulphides, 6, 240. tensile tests, 15, 299, 442, 443, 446, 462. torsional tests, 444, 462, 545. transverse tests, 444, 462, 463. vaporization, 35, 241, 242, 245. voltaic element, 245. weight, 23. white, as a paint, 674, 675. wood preservation, 245, 674-677, 684- 686. Zirconium, 267.