Civil IjJ Engineering Library UNIVERSITY OF CALIFORNIA DEPARTMENT OF CIVIL ENGINEERING BERKELEY, CALIFORNIA ELECTRIC WELDING ELECTRIC WELDING BY ETHAN VIALL EDITOR AMERICAN MACHINIST Member American Society of Mechanical Engineers, Society of Automotive Engineers, American Institute of Electrical Engineers, Franklin Institute, American Welding Society. Author of Manufacture of Artillery Ammunition, United States Artillery Ammunition, United States Rifles and Machine Guns, Broaches and Broaching, Gas-Torch and Thermit Welding. FIRST EDITION THIRD IMPRESSION McGRAW-HILL BOOK COMPANY, INC. NEW YORK: 370 SEVENTH AVENUE LONDON: 6 & 8 BOUVERIE ST., E. C. 4 1921 .1 A Engineering Library l^U l^jf^^ COPYRIGHT, 1921, BY THE McGRAW-HILL BOOK COMPANY, INC PRINTED IN THE UNITED STATES OF AMERICA PREFACE Few fields afford a greater opportunity for study to the mechanic, the student, or the engineer, than that of electric welding. Arc welding, with its practical, every-day, shop appli- cations for repair and manufacture, is in some respects crowding closely into the field in which the gas-torch has seemed supreme. With the development of mechanical devices for the control of the arc, the range of application to production work has greatly increased. Resistance welding presents in its various branches some of the most interesting scientific and mechanical problems to be found anywhere. Spot-welding butt-welding line-welding all occupy a particular place in our manufacturing plants today, and new uses are being constantly found. In the gathering and arranging of the material used in this book, particular care has been taken to classify and place various subjects together as far as possible. This is not only convenient for reference purposes, but enables the reader to easily compare different makes and types of apparatus. In most cases, the name of the maker of each piece of apparatus is mentioned in the description in order to save the time of those seeking information. No time or pains have been spared in the endeavor to make this the most comprehensive book on electric welding equipment and practice, ever published. Every possible source of informa- tion known to the long-experienced editor has been drawn upon and properly credited. It is hoped that this book will prove a permanent record of electric welding as it is today, and also be an inspiration and source of information for those engaged in practice, research or development. ETHAN VIALL. New York City, November, 1920. 742296 CONTENTS PAGE PREFACE v CHAPTER I ELECTRIC WELDING HISTORICAL 1- 8 The Two Classes of Electric Welding The Zerner, the Ber- nardos, the Slavianoff, the Strohmenger-Slaughter and the LaGrange-Hoho Processes Early Methods of Connecting for Arc Welding Early Resistance Welding Apparatus First Practical Butt-Welding Device DeBenardo Spot-Welds The Kleinschmidt Apparatus Bouchayer 's Machine Principle of the Harmatta Patent The Taylor Cross-Current Spot-Welding Method. CHAPTER II ARC WELDING EQUIPMENT 9- 27 What Electric Arc-Welding Is Uses of B.C. and A.C. Schematic Layout for an Arc- Welding Outfit Carbon Electrode Process Metallic Electrode Process Selection of Electrodes Relation of Approximate Arc Currents and Electrode Diam- eters Approximate Current Values for Plates of Different Thickness Illustrations of the Difference Between the Carbon and the Metallic Arc Methods Electrode Holders Sizes 'of Cable for Current Face Masks Selecting a Welding Outfit Eye Protection in Iron Welding Operations The Dangerous Rays Properties of Various Kinds of Glass. CHAPTER III DIFFERENT MAKES OF ARC WELDING SETS 28- 46 General Electric Compound- Wound Balancer-Type Arc Weld- ing Set The Welding Control Panel Connections for G.E. Welding Set Data for Metallic-Electrode Arc Butt- and Lap- Welds Carbon-Electrode Cutting Speeds The Wilson Plastic- Arc Set Panel for Wilson Welding and Cutting Set Wilson Portable Outfit The Lincoln Outfit Westinghouse Single- vii viii CONTENTS PAGE Operator Outfit The U.S. Outfit The "Zeus" Outfit The Arcwell Outfit Alternating-Current Arc-Welding Apparatus G.E. Lead Burning Transformer. CHAPTER IV TRAINING ARC WELDERS 47- 65 Use of Helmets and Shields The Welding Booth Welding Systems The Electrode Holder Arc Manipulation Arc Formation Fusion of Electrodes Maintenance of Arc Control of Arc Travel Weaving Arc and Fusion Character- istics Polarity Length of Arc Stability Overlap and Penetration Heat Conductivity and Capacity Expansion and Contraction of Parent Metal Contraction of Deposited Metal Welding Procedure Electrode Current Density In- spection Terminology. CHAPTER V CARBON-ELECTRODE ARC WELDING AND CUTTING 66- 80 Currents Used with Carbon Arc Carbon and Graphite Elec- trodes Shapes and Size of Electrodes Filler Material Proper Welding Position Arc Manipulation Characteristics of the Arc Polarity Arc Length Building up Surfaces Fused Ends of Filler Rods Flanged Seam Welding Weld- ing Non-Ferrous Metals Applications of Carbon-Arc Weld- ing Cutting Data on Cutting Steel Plates Cutting Cast- iron Plates Cutting Cast-Iron Blocks. CHAPTER VI ARC WELDING PROCEDURE 81-108 Resume of Welding Instructions Filling Sequence Welding Two Plates The Back-Step Method Welding a Square Patch Quasi-Arc Welding Typical Examples of Arc Weld- ing Examples of Tube Work Locomotive Work Welding Calculations Strength of Welds Stresses in Joints Inspec- tion of Metallic-Electrode Arc- Welds Good and Bad Welds- Electrode Diameters for Steel Plate Variation in Weld Strength with Change in Arc Current Effects of Short and Long Arcs Heat Treatment Effects of the Chemical Com- position of Electrodes Physical Characteristics of Plates Chemical Analysis of Specimens The Welding Committees Electrodes. CONTENTS ix CHAPTER VII PAGE ARC WELDING TERMS AND SYMBOLS 109-126 Definitions of Strap, Butt, Lap, Fillet, Plug and Tee Welds The Single V, Double V, Straight, Single Bevel, Double Bevel, Flat, Horizontal, Vertical and Overhead Weld Tack, Caulking, Strength, Composite, Reinforced, Flush and Concave Welds Symbols for Various Kinds of Welds. CHAPTER VIII EXAMPLES OF ARC-WELDING JOBS 127-170 Work on the German Ships Seventy Cylinders Saved Without Replacement The Broken Cylinders of the George Washing- ton Cylinders of the Pocahontas General Ship Work Locomotive Work Repair on a Locomotive Frame Built-Up Pedestal Jaw Repaired Drive Wheel Flue and Firebox Work Side Frames and Couplers Amount Saved by Weld- ing Training of Welders Welded Rails and Cross-Overs Built-Up Rolling Mill Pods Repaired Mill Housing Welded Blow-Holes in Pulley Method of Removing Broken Taps Electric Car Equipment Maintenance A Large Crankshaft Repair Welding High-Speed Tips onto Mild Steel Shanks An All- Welded Mill Building Speed of Arc Welding. CHAPTER IX PHYSICAL PROPERTIES OF ARC-FUSED STEEL 171-190 Preliminary Examinations of Arc Welds Method of Prepar- ing Test Specimens Arrangement of the Welding Apparatus The ' ' Paste ' ' Used for Coated Electrodes Composition of Elec- trodes Before and After Fusing Relation Between Nitrogen- Content and Current Density Appearance of Specimens After Test Tensile Properties of Electrodes Results of Tests on Fifty Specimens Mechanical Properties of the Are-Fused Metal Dependence of Physical Properties on Soundness Macrostructure Discussion of the Results of the Tests Com- parison of the Bureau of Standards and the Wirt-Jones Tests. CHAPTER X METALLOGRAPHY OF ARC-FUSED STEEL 191-213 General Features of the Microstructure of the Electrodes Used Microscopic Evidence of Unsoundness Characteristic "Needles" or "Plates" Plates Probably due to Nitrates X CONTENTS PAGE Relation of Microstructure to the Path of Rupture Effect of Heat Treatment Upon Structure Persistence of "Plates" After Annealing Thermal Analysis of Arc-Fused Steel Summary. CHAPTER XI AUTOMATIC ARC WELDING 214-238 The General Electric Automatic Arc Welding Machine The Welding Head Set-Up for Circular Welding Set-Up for Building Up a Shaft Diagram of Control of Feed Motor- Some Work Done by the Machine Repaired Crane Wheels Welded Hub Stampings Welded Rear Axle Housings Welded Tank Seam The Morton Semi-Automatic Machine Methods of Mechanically Stabilizing and Controlling the Arc Examples of Work Done by the Morton Machine The G.E. Electric- Arc Seam Welding Machine. CHAPTER XII BUTT-WELDING MACHINES AND WORK 239-275 Resistance Welding Machines Butt-Welding Machines Cur- rent Used in Butt-Welding How the Secondary Windings of the Transformer are Connected Typical Butt- Welding Ma- chine with Main Parts Named How the Clamping Jaws are Operated Annealing Welds Portable Wire Welding Ma- chines Examples of Butt-Welding Jobs Welding Copper and Brass Rod Welding Aluminum Typical Copper Welds T-Welding Welding Band Saws Automobile Rim Welding The " Flash- Weld "Welding Heavy Truck Rims Welding Pipe The Type of Clamp Used for Pipe The Approximate Current Used for Pipe Welding The Winfield Butt-Welding Machines Cost of Butt- Welds The Federal Butt-Welding Machines Welding Motor Bars to the End Rings Welding Valve Elbows on Liberty Motor Cylinders An Automatic Chain-Making Machine Electro-Percussive Welding How the Machine is Made Uses of Percussive Welding Power Con- sumed and Time to Make a Percussive Weld. CHAPTER XIII SPOT- WELDING MACHINES AND WORK 276-323 Spot'- Welding Three Desirable Welding Conditions Welding Galvanized Iron and Other Metals Mash Welding Details of Standard Spot-Welding Machines Foot-, Automatic-, and CONTENTS xi PAGE Hand-Operated Machines Examples of Spot-Welding Work Form and Sizes of Die-Points for Spot-Welding The Win- field Spot-Welding Machines Machine for Welding Auto- mobile Bodies The Federal Spot-Welding Machines The Federal Water-Cooled Die-Points Rotatable Head Two-Spot Welding Machine Automatic Machine for Welding Channels Automatic Pulley Welding Machine The Taylor Cross-Cur- rent, Spot-Welding Machines Automatic Hog-Ring Machine A Space-Block Welding Machine Combination Spot- and Line-Welding Machines Spot-Welding Machines for Ship Work A Large Portable Spot-Welding Machine Duplex Welding Machine A Powerful Experimental Machine Portable Mash- Welding Machine for Square or Round Rods Cost of Spot Welding. CHAPTER XIV WELDING BOILER TUBES BY THE ELECTRIC RESISTANCE PROCESS. . . . 324-342 How Boiler Flues are Held for Welding How the Tube Ends are Prepared Scarf- Weld Straight Butt-WeldFlash Weld Use of a Flux How the Work is Placed in the Jaws to Heat Evenly Electric and Oil Heating Compared Kind of Machine to Use Flash Welding Welding in the Topeka Shops of the Santa Fe Railroad The Way the Work Heats Up The Final Rolling. CHAPTER XV ELECTRIC WELDING OF HIGH-SPEED STEEL AND STELLITE IN TOOL MANUFACTURE 343-364 The Machines Used to Weld Tools Welding High-Speed to Low-Carbon Steel Examples of Welded Tools Jaws for Special Work How the Parts are Arranged for Welding Clamping in the Jaws Insert Welding Jaws for Stellite Welding Jaws for Stellite Insert Welding The Vertical Type of Welding Machine Making a ' ' Mash-Weld ' ' Jaws for Mash-Welding Grooving the Pieces to be Welded Current Consumption for Various Jobs Sizes of Wire to Use. CHAPTER XVI ELECTRIC SEAM WELDING 365-381 The Process of Seam Welding Kind of Machine Used De- tails of the Roller Head Thomson Lap-Seam Welding Ma- chine Welding Oil Stove Burner Tubes Jig for Welding xii CONTENTS PAGE Automobile Muffler Tubes Jig for Welding Large Can Seams Jig for Welding Bucket Bodies Jig for Welding Ends of Metal Strips Together Flange Seam Welding Jig for Welding Teapot Spouts Approximate Current for Six- Inch Seam for Various Thicknesses of Sheet Metal Size of Wire to Use in Connecting up a Welding Machine. CHAPTER XVII MAKING PROPER RATES FOR ELECTRIC WELDING, AND THE STRENGTH OF WELDS 382-399 Reasons for Misunderstanding Between User and Producer The Metering Proposition Energy Consumption of Resistance Welding for Commercial Grades of Sheet Iron Effect of Clamping Distance Between Electrodes Upon Time and Energy Demand The Load Factor Maximum Demand Power Factor Strength of Combination Spot and Arc Welds Spot Welding Tests on Hoop Iron Strength of Spot-Welded Holes Plates Plugged by Welding Tested Plates Tensile Tests of Plates Plugged by Spot-Welding Strength of Mash- Welded Rods Strength of Resistance Butt-Welds Elementary Electric Information What is a Volt? What is an Ampere? What is a Kilowatt? What is Kva? ELECTRIC WELDING CHAPTER I ELECTRIC WELDING HISTORICAL All electric welding may be divided into two general classes arc welding and resistance welding. In each class there are a number of ways of obtaining the desired results. Arc welding is the older process, and appears to have been first used by de Meritens in 1881 for uniting parts of storage batteries. He connected the work to the positive pole of a current supply capable of maintaining an arc. The other pole was connected to a carbon rod. An arc was struck by touching the carbon rod to the work and withdrawing it slightly. The heat generated fused the metal parts together, the arc being applied in a way similar to that of the flame of the modern gas torch. Of the several methods of arc welding, there are the Zerner, the Bernardos, the Slavianoff and the Strohmenger-Slaughter processes, as well as some modifications of them. The different methods are named after the men generally credited with being responsible for their development. The LaGrange-Hoho process is not a welding process at all, as it is merely a method of heating metal which is then welded by hammering, as in blacksmith work. It is sometimes called the " water-pail forge." The Zerner process employs two carbon rods fastened in a holder so that their ends converge like a V, as shown in Fig. 1. An arc is drawn between the converging ends and this arc is caused to impinge on the work by means of a powerful electro- magnet. The flame acts in such a manner that this process is commonly known as the electric blowpipe method. The Zerner process is so complicated and requires so much skill that it is practically useless. A modification of the Zerner process, known - > %o<, v " i O S. 2 ELECTRIC WELDING as t be ; 'voltex process," uses carbon rods containing a small percentage of metallic oxide which is converted into metallic vapor. iThris vapor increases the size of the arc and to some extent prevents the excessive carbonizing of the work. This process, however, is about as impractical for general use as the other. The Bernardos process employs a single carbon or graphite FlG. 1 The Zerner Electric ' ' Blow-Pipe. : rod and the arc is drawn between this rod and the work. A sketch of the original apparatus is shown in Fig. 2. This is commonly called the carbon-electrode process. In using this method it is considered advisable to connect the carbon to the negative side and the work to the positive. This prevents the carbon of the rod from being carried into the metal and a softer weld is produced. In the Slavianoff process a metal electrode is used instead UNIVERSITY OF CALIFORNIA DEPARTMENT OF CIVIL ENGINEERING ELECTRIC WELDIN^H^^%IfA^ AL|FOR(X: , | ^ of a carbon. This process is known as the metallic-electrode process. The Strohmenger-Slaughter, or covered electrode, process is similar to the Slavianoff except that a coated metallic elec- "<<(fffff'///^///&//&/S/////////////////////////W FIG. 2. Original Bernardos Carbon Electrode Apparatus. Grid Rheostat Circuit Breaker, FIG. 3. Arc Welding Circuits as First Used. trode is used. In this process either direct or alternating cur- rent may be used. Some of the early methods of connecting up for arc welding are shown in Fig. 3. The LaGrange-Hoho heating process makes use of a wooden tank filled with some electrolyte, such as a solution of sodium ELECTRIC WELDING or potassium carbonate. A plate connected to the positive wire is immersed in the liquid and the work to be heated is connected to the negative wire. The work is then immersed in the liquid. When the piece has reached a welding temperature it is removed and the weld performed by means of a hammer and anvil Resistance Welding. The idea of joining metals by means of an electric current, known as the resistance or incandescent process, was conceived by Elihu Thomson some time in 1877. JMMA/^^ FIG. 4. First Practical Electric Butt Welding Device, Patented by Elihu Thomson, Aug. 10, 1886. Little was done with the idea from a practical standpoint for several years. Between 1883 and 1885 he developed and built an experimental machine. A larger machine was built in 1886. He obtained his first patent on a device for electric welding Aug. 10, 1886. The general outline of this first device is shown in Fig. 4. The first experiments were mostly confined to what is now known as butt welding, and it was soon found that the jaws used to hold the parts heated excessively. To remedy this water-cooled clamping jaws were developed. ELECTRIC WELDING HISTORICAL 5 FIG. 5. Plates "Spot Welded" by Carbon Arc. FIG. 6. The DeBenardo Carbon Electrode Spot Welding Apparatus. FlG. 7. The Kleinschmidt Apparatus, Using Copper Electrodes. 6 ELECTRIC WELDING Closely following the butt welding came other applications of the resistance process, such as spot, point or projection, ridge and seam welding. Percussive welding, which is a form of resistance welding, was developed about 1905. Since spot weld- ing is such an important factor in the manufacturing field today FIG. 8. Bouchayer's Spot Welding Machine, Using Duplex Copper Electrodes. the evolution of this process, as indicated by the more prominent patents, will be of considerable interest: Fig 5 shows plates spot welded together by means of the carbon arc. This was patented by DeBenardo, May 17, 1887, Pat. No. 363,320. The claims cover a weld made at points only. The darkened places indicate ELECTRIC WELDING HISTORICAL 7 where the welds were made. Fig. 6 shows the apparatus made by DeBenardo for making "spot welds," as they are known today. He patented this in Germany, Jan. 21, 1888. Carbon electrodes were used. This patent was probably the first to cover the process of welding under pressure and also for passing the current through the sheets being welded. The German patent number was 46,776 49. The apparatus shown in Fig. 7 is known as the Kleinschmidt patent, No. 616,463, issued Dec. 20, 1898. The patent claims cover the first use of pointed copper electrodes and raised sec- tions, or projections, on the work in order to localize the flow of the current at the point where the weld was to be effected. \L FIG. 9. Principle of the Harmatta Process, Using Copper Electrodes. Considerable pressure was also applied to the electrodes and work by mechanical means. Fig. 8 shows diagrammatically Bouchayer's spot welding machine, patented in France, March 13, 1903, No. 330,200. He used two transformers, one on each side of the work. Duplex copper electrodes were used, and if the transformers were con- nected parallel one spot weld would be made at each operation. If the transformers were connected in series two spot welds would be made. Fig. 9 illustrates the principle of the Harmatta patent, No. 1,046,066, issued Dec. 3, 1912. This is practically the same as the DeBenardo patent, No. 46,776 49, except that copper elec- 8 ELECTRIC WELDING trodes are used. However, it is under the Harmatta patent that a majority of the spot welding machines in use today are made. Fig. 10 illustrates the principle on which the Taylor patent is founded. This patent was issued Oct. 16, 1917, No. 1,243,004. It covers the use of two currents which are caused to cross the path of each other in a diagonal direction, concentrating the heating effects at the place of intersection. a S b 7 FIG. 10. The Taylor Cross-Current Spot Welding Method. From the foregoing it will be seen that spot welds, as this term is now understood, can be produced in a number of ways, none of which methods are identical. As a matter of fact, spot welds can be produced by means of the gas torch or by the blacksmith forge and anvil, although these methods would not be economical. CHAPTER II ARC WELDING EQUIPMENT Electric Arc Welding is the transformation of electrical energy into heat through the medium of an arc for the purpose of melting and fusing together two metals, allowing them to melt, unite, and then cool. The fusion is accomplished entirely without pressure. The heat is produced by the passage of an electric current from one conductor to another through air which is a poor conductor of electricity, and offers a high resistance to its passage. The heat of the arc is the hottest flame that is obtainable, having a temperature estimated to be between 3,500 and 4,000 deg. C. (6,332 to 7,232 deg. P.). The metal to be welded is made one terminal of the circuit, the other terminal being the electrode. By bringing the elec- trode into contact with the metal and instantly withdrawing it a short distance, an arc is established between the two. Through the medium of the heat thus produced, metal may be entirely melted away or cut, added to or built up, or fused to another piece of metal as desired. A particularly advantageous feature of the electric arc weld is afforded through the concentration of this intense heat in a small area, enabling it to be applied just where it is needed. Direct-current is now more generally used for arc welding than alternating-current. When using direct-current, the metal to be welded is made the positive terminal of the circuit, and the electrode is made the negative terminal. Regarding alternating-current it is obvious that an equal amount of heat will be developed at the work and at the elec- trode, while with direct-current welding we have considerably more heat developed at the positive terminal. Also in arc weld- ing the negative electrode determines the character of the arc, which permits of making additions to the weld in a way that is 9 10 ELECTRIC WELDING not possible with alternating-current. Inasmuch as the work always has considerably greater heat-absorbing capacity than the electrode, it would seem only reasonable that the direct-current arc is inherently better suited for this work. Two systems of electric arc welding, based on the type of electrode employed, are in general use, known as the carbon (or graphite) and the metallic electrode processes. The latter Circuit Breaker rrn Grief Resistors Electrode Courtesy of the Westinghouse Co. FIG. 11. Simple Schematic Welding Circuit. process is also sub-divided into those using the bare and the covered metallic electrodes. A simple schematic layout for an arc-welding outfit is shown in Fig. 11. The Carbon Electrode Process. In this process, the nega- tive terminal or electrode is a carbon pencil from 6 to 12 in. in length and from J to 1 in. in diameter. This was the original process devised by Bernardos and has been in more or less general ARC WELDING EQUIPMENT 11 use for more than thirty years. The metal is made the positive terminal as in the metallic electrode process in order that the greater heat developed in this terminal may be applied just where it is needed. Also, if the carbon were positive, the tendency would be for the carbon particles to flow into the weld and thereby make it hard and more difficult to machine. The current used in this process is usually between 300 and 450 amp. For some special applications as high as from 600 to 800 may be required, especially if considerable speed is desired. The arc supplies the heat and the filler metal must be fed into the weld by hand from a metallic bar. The class of work to which the carbon process may be applied includes cutting or melting of metals, repairing broken parts and building up materials, but it is not especially adapted to work where strength is of prime importance unless the operator is trained in the use of the carbon electrode. It is not practical to weld with it overhead or on a vertical surface but there are many classes of work which can be profitably done by this process. It can be used very advantageously for improving the finished surface of welds made by metal electrodes. The carbon electrode process is very often useful for cutting cast iron and non-ferrous metals, and for filling up blowholes. The Metallic Electrode Process. In the metallic electrode process, a metal rod or pencil is made the negative terminal, and the metal to be welded becomes the positive terminal. When the arc is drawn, the metal rod melts at the end and is automatically deposited in a molten state in the hottest portion of the weld surface. Since the filler is carried directly to the weld, this process is particularly well adapted to work on vertical surfaces and to overhead work. If the proper length of arc is uniformly maintained on clean work, the voltage across the arc will never greatly exceed 22 volts for bare electrodes and 35 volts for coated electrodes. The arc length will vary to a certain degree however, owing to the physical impossibility of an operator being able to hold the elec- trode at an absolutely uniform distance from the metal through- out the time required to make the weld. It is very essential that the surfaces be absolutely clean and free from oxides and dirt, as any foreign matter present will materially affect the success of the weld. 12 ELECTRIC WELDING When using a metallic electrode, the arc which is formed by withdrawing it from the work, consists of a highly luminous central core of iron vapor surrounded by a flame composed largely of oxide vapors. At the temperature prevailing in the arc stream and at the electrode terminals, chemical combinations occur instantaneously between the vaporized metals and the atmospheric gases. These reactions continue until a flame of incandescent gaseous compounds is formed which completely envelopes the arc core. However, drafts created by the high temperature of the vapors and by local air currents tend to remove this protecting screen as fast as it is formed, making it necessary for the welder to manipulate the electrode so that the maximum protective flame for both arc stream and electrode deposit is continuously secured. This can be obtained auto- matically by the maintenance of a short arc and the proper inclination of the electrode towards the work in order to com- pensate for draft currents. Selection of Electrodes. The use of a metallic electrode for arc welding has proved more satisfactory than the use of a carbon or graphite electrode which necessitates feeding the new metal or filler into the arc by means of a rod or wire. The chief reason for this is that, when the metallic electrode process is used, the end of the electrode is melted and the molten metal is carried through the arc to be deposited on the material being welded at the point where the material is in a molten state produce*! by the heat of the arc. Thus a perfect union or fusion is produced with the newly deposited metal. Wire for metallic arc welding must be of uniform, homogene- ous structure, free from segregation, oxides, pipes, seams, etc. The commercial weldability of electrodes should be determined by means of tests performed by an experienced operator, who can ascertain whether the wire flows smoothly and evenly through the arc without any detrimental phenomena. The following table indicates the maximum range of the chemical composition of bare electrodes for welding mild steel: Carbon trace up to 0.25% Manganese trace up to 0.99% Phosphorous not to exceed 0.05% Sulphur not to exceed 0.05% Silicon not to exceed 0.08% ARC WELDING EQUIPMENT 13 The composition of the mild steel electrodes, commonly used, is around 0.18 per cent carbon, and manganese not exceeding 0.05 per cent, with only a trace of phosphorus, sulphur and silicon. The size, in diameter, ordinarily required will be 1 / 8 in., 5 / 32 in., and Vie i n - an N ZzfFuse CRIOOO Starter JftA J To LI Al Electrode To Work probably grounded Series Field Shunt Field Series Field Com 'IT?. Field Armature Co mm. Generator Field Armature FIG. 24. Balancer and Control Panel Connections for General Electric Constant-Energy Constant-Arc Set. the generator, the field being connected to oppose the main field. This feature provides the current control by which six steps are obtained of the approximate values of 50, 70, 90, 110, 130 and 150 amp., which enables the operator to cover a very wide range. DIFFERENT MAKES OF ARC WELDING SETS 31 In addition, if intermediate current values are required, they can be obtained by means of the generator field rheostat. A small reactor is used to steady the arc and current both on starting and during the period of welding. Arc welding is usually done on metal which is grounded and this is especially unavoidable in ship work, where the ship struc- ture is always well grounded. Since successful operation requires that the positive terminal be connected to the work the supply circuit should be safely grounded on the positive side. Where a 125-v., d.c. supply system is not available, standard Thickness FIG. 25. Carbon Electrode Cutting Speeds for Different Thicknesses of Plate. "MIC" or "MCC" sets are furnished to supply power at 125 v., the motor being either 3-phase, 60-cycle, 220, 440 or 550 v., or d.c,, 230 or 550 v., and in three capacities, 5J kw., 7 kw., and 15 kw. With each motor generator set there is supplied a panel containing generator field rheostat and motor starter, which may be mounted beside the balancer panel. A diagram showing the balancer and control panel is shown in Fig. 24. The constant energy arc-welding equipment supplies, to the arc, practically constant energy throughout the welding range for metallic electrode welding only. If the arc is lengthened slightly the voltage increases and the current decreases, the total 32 ELECTRIC WELDING f "3^ ' o^ o O rH CO TjH -^ 3 E-t >-> CM* co* co* o - O b^ O* O V 1 ^n 1 1 i V 10 GO 05 1- " pq 3 g ^ S o ** j ti c 1 05 S o , CO C3 o o o o o 'be CO CO CO CO CO C M H 03 ^ 1 ft O2 . cS p o .GO CO CM O rH /5 w *o ^ o ^ 2 d 0* rH CO* rt* ^ EH w (V rH TH rH CM C-l Q | 's 1 1 S PQ g 0) . H 5 j N *2 n5 >- I || o P ^ h ^ 05 |JW o* ^ rH CO GO ^ f? H co' co' co* co' ^2 3 rji Si J3 METALLIC I 5 5 Ul ..J ^ o> n a ^ ^ H k> -M . ^ M b- fl CM -M ^o CS o rH "* ^ *> "c3 co* o ^ ">-' *C 1 a cc rt ft fl te ^ r3 r H^ J. S i> t> ^ TH o iq GO o g r ^ d c-i 01 cc EH CO Q PH 1^ Q 13 T^ 1 . ^ '5c 3 f*< 05 S o S H rH Tfl S a S ^ "52 V, V. &JD CO rH CG M S "^ \ >s \ \ ^ . ^1 *^ rH ^Q rH rH r , 03 Q PH QO rH ij M _, ""^ ^ W [fc 3 co H *a GO r^ QO Cl GO ^ s !5 y their method the FIG. 62. Diagram Illustrating Back-Step Method. plates are fastened together by light tack welds about 8 in. apart along the whole seam. The operator then makes a com- plete weld between the first two tacks as described in the preceding paragraph, and, skipping three spaces, welds between the fifth and sixth tacks and so on until the end of the seam is reached. This skipping process is repeated by starting be- tween the second and third tacks and so on until the complete seam is welded. The adoption of this method permits the heat, in a restricted area, to be dissipated and radiated before addi- tional welding is performed near that area. Thus the weld is made on comparatively cool sections of the plates which keeps the expansion at a minimum. ARC WELDING PROCEDURE 85 Another method very similar to the preceding one, is known as the back-step method, Fig. 62, in which the weld is performed in sections as in the skipping process. After the pieces are tacked at intervals of 6 in. or less for short seams, the arc is applied at the second tack and the groove welded back complete to the first tack. Work is then begun at the third tack and the weld carried back to the second tack, practically completing that section. Each section is finished before start- ing the next. Fig. 63 shows the procedure of welding in a square sheet or patch. Work is started at A and carried to B completely welding the seam. In order that work may next be started at the coolest point, the bottom seam is completed starting at />, finishing at C. The next seam is A to D, starting at A. D FIG. 63. Diagram Illustrating Square Patch Method. The last seam is finished, starting at 5, and completing the weld at C. Alternating- Current Arc Welding. Direct current has been used for arc welding because of the fact that it possesses cer- tain inherent advantages that make it especially adaptable for this class of work. However, the use of alternating current for arc welding has found a number of advocates. When employing this form of energy, use is made of a trans- former to reduce the distribution voltage to that suitable for application to the weld. Inasmuch as the arc voltage is obtained directly from the distribution mains through a transformer, the theoretical effi- ciency is high compared with the direct-current process which requires the introduction of a motor- generator or resistor or 86 ELECTRIC WELDING both. The efficiency of the a.c. equipments now on the market ranges from 60 to 80 per cent. The transformer, however, is designed to have a large leakage reactance so as to furnish stability to the arc, which very materially reduces its efficiency when compared with that of the standard distribution trans- former used by lighting companies. It is difficult to maintain the alternating arc when using a bare electrode though this difficulty ic somewhat relieved when use is made of a coated electrode. Quasi Arc Welding. The electrodes used in quasi arc weld- ing are made by the Quasi Arc Weldtrode Co., Brooklyn, N. Y., and are known as "weldtrodes." A mild-steel wire is used with a very small aluminum wire running lengthwise of it. Around the two is wrapped asbestos thread. This asbestos thread is held on by dipping the combination into something similar to waterglass. Either a.c. or d.c. may be used, at a pressure of about 105 volts, with a suitable resistance for regulating the current. The company's directions and claims for this process are : "The bared end of the weldtrode, held in a suitable holder, is connected to one pole of the current supply by means of a flexible cable, the return wire being connected to the work. In the case of welding small articles, the work is laid on an iron plate or bench to which the return wire is connected. Electrical contact is made by touching the work with the end of the weldtrode held vertically, thus allowing current to pass and an arc to form. The weldtrode, still kept in contact with the work, is then dropped to an angle, and a quasi-arc will be formed owing to the fact that the special covering passes into the igneous state, and as a secondary conductor maintains electrical connection between the work and the metallic core of the weldtrode. The action once started, the weldtrode melts at a uniform rate so long as it remains in contact, and leaves a seam of metal fused into the work. The covering material of the weldtrode, acting as a slag, floats and spreads over the surface of the weld as it is formed. The fused metal, being entirely covered by the slag, is protected from oxidation. TRe slag covering is readily chipped or brushed off when the weld cools, leaving a bright clean metallic surface. In welding do not draw the weldtrode along the seam, as it is burning away all the time, and therefore it is ARC WELDING PROCEDURE 87 only necessary to feed it down, but do this with a slightly lateral movement, so as to spread the heat and deposited metal equally to both sides of the joint. Care must be taken to keep feeding down at the same rate as the weldtrode is melting. On no account draw the weldtrode away from the work to make a continuous arc as this will result in putting down bad metal. The aim should be to keep the point of the weld- After After After After After Before After Before After After Before After _____ After After TIG. 64. Typical Examples of Prepared and Finished Work. => trode just in the molten slag by the feel of the covering just rubbing on the work. By closely observing the operation, the molten metal can easily be distinguished from the molten slag, the metal being dull red and the slag very bright red." The weldtrodes are supplied ready for use in standard lengths of 18 in., and of various diameters, according to the size and nature of the work for which they are required. 88 ELECTRIC WELDING Typical Examples of Arc Welding. The examples of weld- ing shown in Figs. 64, 65 and 66 are taken from the manual issued by the Wilson Welder and Metals Co. They will be found very useful as a guide for all sorts of work. Fig. 64 Before After Before 33 After Before Before After After After W" After After Before Before After After After FIG. 65. Examples of Tube Work. shows miscellaneous plate or sheet jobs, Fig. 65 shows tube jobs, while Fig. 66 gives examples of locomotive-frame and boiler-tube welding. As a basis for various welding calculations the following data will be found of use: On straight-away welding the ARC WELDING PROCEDURE 89 ordinary operator with helper will actually weld about 75 per cent of the time. The average results of a vast amount of data show that an Great care must be exercised in the preparation of the frames for welding, and that the proper heat valu9 and we I din a metals ~pe employed tor the different character or material in the frames^ to be welded Before p ^ Before Welding 1 Weld ing After Welding --) U^^"* J ta*- -^-^' - . . -j In welding flues by the Electric Arc process, the flue sheet and flues Pur ruciy t'C vnrica A' inccr cf/rrerenr conairions. ine proper heat value to emplcy jnd amount of me fa/ to apply must be determined in each case. FIG. 66. Examples of Electric Welding of Locomotive Frames and Boiler Tubes. operator 'can deposit about 1.8 Ib. of metal per hour. This rate depends largely upon whether the work is done out in the open or in a special place provided in the shop. For outside work such as on boats, an operator will not average 90 ELECTRIC WELDING in general more than 1.2 Ib. per hour, while in the shop the same operator could easily deposit the 1.8 Ib. stated above. This loss in speed for outside work is brought about largely by the cooling action of the air and also somewhat by the added inconvenience to the operator. The value of pounds per hour given above is based on the assumption that the work has been lined up and is ready for welding. On the average 70 per cent of the weight of electrodes is deposited in the weld, 12 per cent is burned or vaporized and the remainder 18 per cent is wasted as short ends. Other figures prepared by the Electric Welding Committee show the possible cost of a fillet weld on a |-in. plate, using a motor generator set and bare electrodes to be as follows: Average speed of welding on continuous straight away work 5 ft. per hour Amount of metal deposited per running foot 6 Ib. Current 150 amps, at 20 volts = 3 kilowatts. Motor generator eff . 50 per cent = 6 kw. -f- 5 equals 1.2 k.w.h. per 1 ft. run 1.2 k.w.h. at 3 cents per k.w.h. equals 3.6 cents per ft. Cost of electrode 10 cents per pound and allowing for waste ends, etc., equals 7.2 cents per ft. Labor at 65 cents per hour equals 13.00 cents per ft. 23.8 cents per ft. Suggestions for the Design of Welded Joints. From an engineering point of view, every metallic joint whether it be riveted, bolted or welded, is designed to withstand a perfectly definite kind and amount of stress. An example of this is the longitudinal seam in the shell of a horizontal fire-tube riveted boiler. This joint is designed for tension and steam tightness only and will not stand even a small amount of transverse bending stress without failure by leaking. If a joint performs the function for which it was designed and no more, its designer has fulfilled his responsibilities and it is a good joint economically. Regardless of how the joint is made the design of joint which costs the least to make and which at the same time performs the functions required of it, with a reasonable factor of safety, is the best joint. The limitations of the several kinds of mechanical and welded joints should be thoroughly understood. A bolted joint is expensive, is difficult to make steam- or water-pressure tight, but has the distinguishing advantage that ARC WELDING PROCEDURE 91 it can be disassembled without destruction. Bolted joints which are as strong as the pieces bolted together are usually imprac- ticable, owing to their bulk. Riveted joints are less expensive to make than bolted joints but cannot be disassembled without destruction to the rivets. A riveted joint, subject to bending stress sufficient to produce appreciable deformation, will not remain steam- or water- pressure tight. Riveted joints can never be made as strong as the original sections because of the metal punched out to form the rivet holes. There is no elasticity in either riveted, bolted or fusion- welded joints which must remain steam- or water-pressure tight. Excess -material is required in the jointed sections of bolted or riveted joints, owing to the weakness of the joints. Fusion-welded joints have as a limit of tensile strength the tensile strength of cast metal of a composition identical to that of the joined pieces. The limit of the allowable bending stress is also set by the properties of cast metal of the same composition as that of the joined pieces. The reason for this limitation is that on the margin of a fusion weld adjacent to the pieces joined, the metal of the pieces was heated and cooled without change of composition. Whatever proper- ties the original metal had, due to heat or mechanical treatment, are removed by this action, which invariably occurs in a fusion weld. Regardless of what physical properties of the metal used to form the joint may be, the strength or ability to resist bending of the joint, as a whole, cannot exceed the correspond- ing properties of this metal in the margin of the weld. Thus, assuming that a fusion weld be made in boiler plate, having a tensile strength of 62,000 pounds. Assume that nickel-steel, having a tensile strength of 85,000 Ib. be used to build up the joint. No advantage is gained by the excess 23,000 Ib. tensile strength of the nickel-steel of the joint since the joint will fail at a point close to 62,000 Ib. If appreciable bending stress be applied to the joint it will fail in the margin referred to. The elastic limit of the built-in metal is the same as its ultimate strength for all practical purposes, but the ultimate strength is above the elastic limit of the joined sections in commercial structures. In spite of the limitations of the fusion-welded joint it is 92 ELECTRIC WELDING possible and practicable to build up a joint in commercial steel which will successfully resist any stress which will be en- countered in commercial work. The fundamental factor in the strength of a welded joint is the strength of the material added by the welding process. This factor depends upon the nature of the stress applied. The metal added by the welding process, when subject to tension, can be relied on in commercial practice to give a ten- sile strength of 45,000 Ib. per square inch. This is an average condition; assuming that the metal added is mild steel and that the operation is properly done, the metal will have ap- proximately the same strength in compression as in tension. When a torsional stress is applied to a welded joint ' the resultant stress is produced by a combination of bending, ten- sion and compression, as well as shear. The resistance of the metal to shear may be figured at 8 /io its resistance to tensile stress. The metal added by the welding process, with the present development in the art of welding, will stand very little bending stress. A fusion-welded joint made by the elec- tric-arc process must be made stiffer than the adjacent sections in order that the bending stress shall not come in the joint. An electric weld, when properly made, will be steam- and water-pressure tight so long as bending of members of the structure does not produce failure of the welded joint. Little is known at the present time in regard to the resist- ance of an electrically welded joint to dynamic stress, but there is reason to believe that the resistance to this kind of stress is low. However, owing to the fact that in most struc- tures there is an opportunity for the members of the structure to flex and reduce the strain upon the weld, this inherent weak- ness of the welded joint does not interfere seriously with its usefulness. A few tests have been made of high-frequency alternating stresses and it has been found that using the ordinary wire electrode the welded joint fails at a comparatively small num- ber of alternations. This is of little importance in most struc- tures since high-frequency alternating stress Ms not often encountered. Stresses in Joints. The accompanying cuts show a number of typical joints and the arrows indicate the stresses brought ARC WELDING PROCEDURE 93 FIG. 67. Joints Designed to Overcome Certain Stresses. 94 ELECTRIC WELDING to bear on them. The proper way to weld each example is plainly shown. In A, Fig. 67, it will be noted that a reinforcing plate is welded to the joint to make the joint sufficiently stiff to throw the bending outside the weld. B shows a joint in straight tension. Since no transverse stress occurs the heavy reinforcing of A is not required. Just enough reinforcing is given the joint to make up for the defi- ciency in tensile strength of the metal of the weld. C shows another method of building up a joint that is in FIG. 68. Plate and Angle Construction. straight tension. It Should be noted that in both B and C as much reinforcing is placed on one side of a center line through the plates as is placed on the other. The original form of lap joint such as is used in riveting is shown at D. The method shown for welding this joint is the only method which can be used. It cannot be recommended because such a joint, when in straight tension, tends to bring the center line of the plate into coincidence with the center line of the stress. In so doing an excessive stress is placed on the welded material. E shows the construction used in certain large tanks where ARC WELDING PROCEDURE 95 a flanged head is backed into a cylindrical shell. The principal stress to be resisted by the welded joint is that tending to push the head out of the shell. The welding process indicated in the figure will successfully do this. Owing to the friction between the weld and the shell, the outer weld would be suffi- cient to hold the weld in place for ordinary pressure. For higher pressures the inside weld should be made in addition. FIG. 69. Pipe Heading and Firebox Sheet Work. F and G show another method of welding a flanged head to the cylindrical shell. These methods are preferable to the method indicated in E. G represents the recommended practice. Fig. 68 shows a plate and angle structure which might be used in ship construction. The particular feature to notice in the welding practice indicated, is that the vertical plates do not reach the entire distance between the horizontal plates. 96 ELECTRIC WELDING This is merely a method of eliminating difficulties in welding the plates to the angle. A in Fig. 69 shows a method of welding a head into a cylindrical pipe. The thickness of the head should be ap- proximately twice the thickness of the wall of the pipe. The extra thickness plate is to gain sufficient stiffness in the head to make the stress on the welded material purely shear. The pressure from the inside tends to make the head assume a hemispherical shape. This would place a bending stress on the welded material if the head were thin enough to give at the proper pressure. B shows a method of welding a crack in a fire-box sheet. The thin plate backing introduced at the weld makes the operation very much easier for the operator and produces the reinforcing of the water side of the fire-box sheet which is most desirable. INSPECTION OF METALLIC ELECTRODE ARC WELDS Determining the character of welded joints is of prime importance, says 0. S. Escholz, and the lack of a satisfactory method, more than any other factor, has been responsible for the hesitancy among engineers of the extensive adoption of arc welding. To overcome this prejudice it is desirable to shape our rapidly accumulating knowledge of operation into an acceptable method of inspection. Manufactured apparatus is practically all accepted on the basis of complying with a process specification rigidly enforced in conjunction with the successful reaction to certain tests applied to the finished product. Riveting impairs the strength of the joined plates, yet with a proper layout an^d intelligent inspection the completed structure possesses certain definite characteristics which do not require further verification. The inspector of a finished concrete structure is practically help- less, and the weakest sort of construction "may be concealed by a sound surface. With careful supervision, however, the physical properties of the completed structure can be reliably gaged to the extent that the use of concrete is justified even in ship construction. With this in view, electric arc welding is susceptible to even better control than obtain in either of these structural operations. ARC WELDING PROCEDURE 97 The four factors which determine the physical character- istics of the metallic electrode arc welds are: Fusion, slag content, porosity and crystal structure. Some of the other important methods that have been sug- gested and used for indicating these characteristics are: 1. Examination of the weld by visual means to determine (a) finish of the surface as an index to workmanship; (b) length of deposits, which indicates the frequency of breaking arc, and therefore the ability to control the arc; (c) uniformity of the deposits, as an indication of the faithfulness with which the filler metal is placed in position; (d) fusion of deposited metal to bottom of weld scarf as shown by appearance of under side of welded joint; (e) predominance of surface porosity and slag. 2. The edges of the deposited layers chipped with a cold chisel or calking tool to determine the relative adhesion of deposit. 3. Penetration tests to indicate the linked unfused zones, slag pockets and porosity by (a) X-ray penetration; (b) rate of gas penetration; (c) rate of liquid penetration. 4. Electrical tests (as a result of incomplete fusion, slag inclusions and porosity) showing variations in (a) electrical conductivity; (b) magnetic induction. These tests if used to the best advantage would involve their application to each layer of deposited metal as well as to the finished weld. This, except in unusual instances, would not be required by commercial practice in which a prescribed welding process is carried out. Of the above methods the visual examination is of more importance than generally admitted. Together with it the chipping and calking tests are particularly useful, the latter test serving to indicate gross neglect by the operator of the cardinal welding principles, due to the fact that only a very poor joint will respond to the tests. The most reliable indication of the soundness of the weld is offered by the penetration tests. Obviously the presence of unfused oxide surfaces, slag deposits and blowholes will offer a varying degree of penetration. Excellent results in the testing of small samples are- made possible by the use of the X-ray. However, due to the nature of the apparatus, the 98 ELECTRIC WELDING amount of time required and the difficulty of manipulating and interpreting results, it can hardly be considered at tlie present time as a successful means to be used on large-scale production. The rate that hydrogen or air leaks through a joint from pressure above atmospheric to atmospheric, or from atmospheric to partial vacuum, can readily be determined by equipment that would be quite cumbersome, and the slight advantage over liquid penetration in time reduction is not of sufficient importance to warrant consideration for most welds. Of the various liquids that may be applied kerosene has marked advantages because of its availability, low volatility and high surface tension. Due to the latter characteristics kerosene sprayed on a weld surface is rapidly drawn into any capillaries produced by incomplete fusion between deposited metal and weld scarf, or between succeeding deposits, slag inclusions, gas pockets, etc., penetrating through the weld and showing the existence of an unsatisfactory structure by a stain on the emerging side. A bright-red stain can be produced by dissolving suitable oil-soluble dyes in the kerosene. By this means the presence of faults have been found that could not be detected with hydraulic pressure or other methods. By the kerosene penetration a sequence of imperfect struc- ture linked through the weld, which presents the greatest hazard in welded joints, could be immediately located, but it should be borne in mind that this method is not applicable to the detection of isolated slag or gas pockets nor small, disconnected unfused areas. It has been shown by various tests, however, that a weld may contain a considerable amount of distributed small imperfections, without affecting to a great extent its characteristics. If a bad fault is betrayed by the kerosene test it is advis- able to burn out the metal with a carbon arc before rewelding under proper supervision. By the means of sandblast, steam or gasoline large quantities of kerosene are preferably removed. No difficulty has been encountered on welding over a thin film of the liquid. Electrical tests, by which the homogeneity of welds is determined, are still in the evolutionary stages, and many diffi- culties are yet to be overcome before this test becomes feasible. ARC WELDING PROCEDURE 99 Some of these difficulties are the elimination of the effect of contact differences, the influence of neighboring paths and fields, and the lack of practicable, portable instruments of suffi- cient sensibility for the' detection of slight variations in con- ductivity or magnetic field intensity. No simple tests are plausible, excepting those which involve subjecting the metal to excessive stresses for determining the crystal structure. Control of this phase must be determined by the experience obtained from following a prescribed process. The inspector of metallic arc electrode welds may consider that through the proper use of visual, chipping and penetrating tests . a more definite appraisal of the finished joint may be obtained than by either riveting or concrete construction. The r FIG. 70. Typical Arc- weld Scarfs. operation may be still further safeguarded by requiring rigid adherence to a specified process. Good results are assured if correct procedure is followed. Haphazard welding can no sooner produce an acceptable product than hit-or-miss weaving will make a marketable cloth. It is only logical that all the steps in a manufacturing opera- lion should be regulated to obtain the best results. As it is most welders consider themselves pioneers in an unknown art that requires the exercise of a peculiar temperament for its successful evolution, and as a result welding operators enshroud themselves in the halo of an expert and do their work with a mystery bewildering to the untutored. Once in a while, due we might say to coincidences, these " experts" obtain a good weld, but more often the good weld may be attributed to the friction between slightly fused, plastered deposits. In common with all other operations metallic electrode aix, 100 ELECTRIC WELDING welding is really susceptible to analysis. Regardless of the metal welded with the arc the cardinal steps are: (1) Prepara- tion of weld; (2) electrode selection; (3) arc-current adjust- ment; (4) arc-length maintenance, and (5) heat treatment. Sufficient scarfing is involved in the preparation of the weld, as well as the separation of the weld slants, so that the entire surface is accessible to the operator with a minimum amount of filling required. When necessary to avoid distortion and internal stresses, owing to unequal expansion and contrac- tion strains, the metal is preheated or placed so as to permit the necessary movement to occur. Various types of scarfs in common use are shown in Fig. 70. The electrode selection is determined by the mass, thickness FIG. 71. Good and Bad Welds. and constitution of the material to be welded. An electrode free from impurities and containing about 17 per cent, carbon and 5 per cent, manganese has been found generally satis- factory for welding low and high carbon as well as alloy steels. This electrode can also be used for cast-iron and malleable-iron welding, although more dependable results, having a higher degree of consistency and permitting machining of welded sections, can be obtained by brazing, using a copper-aluminum- iron-alloy electrode and some simple flux. Successful results are obtained by brazing copper and brass with this electrode. The diameter of the electrode should be chosen with reference to the arc current used. A great many concerns have attempted welding with too ARC WELDING PROCEDURE 101 low an arc current and the result lias been a poorly fused deposit. This is due largely to the overheating characteristics of most electrode holders, or using current value, .and thus- leading the operator to conclude that the cufreril^iise'd is nV excess of the amount that is needed. '^;^ :> A, Fig. 71, shows a section through one-half of air exposed joint welded with the proper current, and B the effects of too low a current. The homogeneity and the good fusion of the one may be contrasted with the porosity and poor fusion of ZOO Amperes Arc Current. FIG. 72. Diameters for Welding Steel Plate. the latter. These surfaces have been etched to show the char- acter of the metal and the welded zone. The approximate values of arc current to be used for a given thickness of mild-steel plate, as well as the electrode diameter for a given arc current, may be taken from the curve in Fig. 72. The variation in the strength of 1-in. square welded joints as the welding current is increased is shown in Fig. 73. Notwithstanding that the electrode development is still in its infancy the electrodes available are giving satisfactory results, but considerable strides can yet be made in the duc- tility of welds, consistency in results and ease of utilizing the process. The maintenance of a short arc length is imperative. A nonporous, compact, homogeneous, fused deposit on a 1-in. 102 ELECTRIC WELDING square bar from a short arc is shown in Fig. 74, A, and in B is shown a porous, diffused deposit from a long arc. Top .yiews of* these , welds are shown in Fig. 75. A short arc is SO 100 ISO Amperes Arc Current. zoo FIG. 73. Variation in Weld Strength with Change in Arc Current. PIG. 74. Sectional Views of Short and Long Arc Deposits. usually maintained by a skillful operator, as the work is thereby expedited, less electrode material wasted and a better weld obtained because of improved fusion, decreased slag content ARC WELDING PROCEDURE 103 and porosity. On observing the arc current and arc voltage by meter deflection or from the trace of recording instruments, the inspector has a continuous record of the most important factors which affect weld strength, ductility, fusion, porosity, etc. The use of a fixed series resistance and an automatic time-lag reset switch across the arc to definitely fix both the arc current and the arc voltage places these important factors entirely beyond the control of the welder and under the direc- tion of the more competent supervisor. Heat Treatment and Inspection, The method of placing the deposited layers plays an important part on the internal strains and distortion obtained on contraction. It is possible that part of these strains could be relieved by preheating and FIG. 75. Top Views of Welds Shown in Fig. 74. annealing as well as by the allowance made in preparation for the movement of the metal. The heat treatment of a completed weld is not a necessity, particularly if it has been preheated for preparation and then subjected to partial annealing. A uniform annealing of the structure is desirable, even in the welding of the small sections of alloy and high-carbon steels, if it is to be machined or subjected to heavy vibratory stresses. The inspector, in addition to applying the above tests to the completed joint and effectively supervising the process, can readily assure himself of the competency of any operator by the submission of sample welds to ductility and tensile tests or by simply observing the surface exposed on cutting through the fused zone, grinding its face and etching with a solution of 1 part concentrated nitric acid in 10 parts water. It is confidently assumed, in view of the many resources at the disposal of the welding inspector, that this method of 104 ELECTRIC WELDING obtaining joints will rapidly attain successful recognition as a dependable operation to be used in structural engineering. EFFECTS OF THE CHEMICAL COMPOSITION OF METALLIC ARC WELDING ELECTRODES In order to ascertain to what extent the chemical analysis of an electrode affected the welded material in metallic arc welding, says J. S. Orton, two electrodes R and W were chosen of widely different chemical analyses, each 0.148 in. in diameter. The R electrode was within the specifications of the Welding Research Committee except that the silicon content was a little high. The analyses were as follows: R wire C . 17 Mn 57 P 007 S 028 Si 14 W wire 0.39 1.01 0.005 0.024 12 The silicon content was rather high, but inasmuch as it was fairly constant in both electrodes the results are com- parative. A deposit was made on a |-in. plate by means of a metallic arc, the welded section being approximately 1 ft. long, 6 in. wide and 1 in. thick. The welding machine used was of a well-known make, with a constant voltage of 37 volts at 130 amperes. The plates used for depositing the first layer were machined away and two test bars were made from each elec- trode, composed entirely of welded material. The ends were rough-machined and about 4^ in. in the middle of the specimens were finished carefully. The physical characteristics of the plates are as shown in Table V. TABLE V. PHYSICAL CHARACTERISTICS or PLATES Tensile Strength 57,300 Elastic Limit 43,400 Elongation 8.0 KA Brinncl 15 3 2 56 050 50 500 6 5 9 JF-1. . . . 76 200 64000 75 13 2 72,650 60,260 5 5 7 1 After these bars were pulled, chemical analyses were taken at various points to get the values given in Table VI. ARC WELDING PROCEDURE 105 TABLE VI. CHEMICAL ANALYSES OF SPECIMENS E-l C 0.12 Mn 0.23 P 0.012 s 0.019 Si 0.10 2 0.09 0.24 0.016 0.014 0.08 3 0.11 0.26 0.014 0.020 0.08 W-\ 023 84 0.014 012 02 o , 0.20 0.80 0.014 0.014 0.05 3 0.20 0.88 0.013 0.013 0.02 Photographs of the different fractures are shown in Fig. 77. W-l, which gave the highest tensile strength, shows 100 per cent, metallic structure with a silky appearance. R-l shows a coarse intergranular fracture. R-2 shows a brittle, shiny crystalline fracture with a slag inclusion at the lower left-hand and upper right-hand corners of the bars. W-2 FIG. 76. Fractures of Test Specimens. shows partial crystalline and partial silky fracture. At the extreme right there is a portion which is not welded. This is probably the reason why W-2 did not pull as much as the other. Undoubtedly, next to the chemical analysis, the quan- tity of slag in the weld has the biggest bearing on the tensile strength. The structure of the test specimens is shown in the micro- photographs of Fig. 77. In making these photographs, no attempt was made to make a complete microanalysis of the two different specimens, but rather it was intended to show the general difference in structure between the two different types of electrode. All of these photographs were taken at 150 diameters except the last two, which were taken at 100. Photograph R-1A shows the general structure of the plate welded with the R electrode. This photograph shows a large- 106 ELECTRIC WELDING ARC WELDING PROCEDURE 107 grain growth and columnar structure which are characteristic of electric welds. Photograph Wl-A shows the general struc- ture of the plate welded with the W electrode. This shows comparatively small-grain structure. The structure seems to be much better than that of Rl-A. Photograph Rl-B shows a portion of a test specimen which was cut out of plate Rl and bent to an angle of 10 deg. It is interesting to note here the opening up of the welded material adjacent to slag inclu- sions. Photograph Wl-B shows a portion of a small specimen cut out from sample Wl and bent to an angle of 10 deg., the same as in the case of Rl-B. The welded material is opening up but not in the same degree nor around the slag inclusions as in the corresponding photograph Rl-B. Photograph Rl-C is a profile of the fracture of the Rl sample after bending through an angle of 15 deg. Photograph Wl-C shows the Wl sample after being bent through an angle of 17 degrees. It seems just as important to specify the chemical composi- tion of the electrode used in metallic arc welding as it is to specify the chemical composition in ordering any other type of steel. Chemical composition seems to affect the physical properties in electrodes as well as other steel. An excess of manganese seems to be needed in electrodes. The relation between the carbon and manganese of an elec- trode should be approximately one to three. High-carbon manganese wire tends not only to improve the weld on account of the amount of carbon and manganese in the welded material, but also on account of the type of structure which this wire lends to the deposited metal. There is a smaller amount of oxide and slag inclusions with a high-carbon manganese wire than with a comparatively low- carbon manganese wire. WELDING COMMITTEE ELECTRODES After an exhaustive series of tests the Welding Committee drew up the following tentative specification for electrodes intended to be used in welding mild steel of shipbuilding quality : Chemical Composition. Carbon, not over 0.18 per cent; manganese, not over 0.55 per cent; phosphorus, not over 0.05 108 ELECTRIC WELDING per cent; sulphur, not over 0.05 per cent; silicon, not over 0.08 per cent, Sizes : Fraction of Inch Lbs. Per Foot Foot Per Lb. Lbs. Per 100 Ft. 1/8 0.0416 24 4.16 5/32 0.0651 15.35 6.51 3/16 0.0937 10.66 9.37 Allowable tolerance 0.006 plus or minus. Material. The material from which the wire is manufac- tured shall be made by any approved process. Material made by puddling process not allowed. Physical Properties. Wire to be of uniform homogeneous structure, free from segregation, oxides, pipes, seams, etc., as proven by micro-photo graphs. This wire may or may not be covered. Workmanship and Finish. (a) Electric welding wire shall be of the quality and finish known as "Bright Hard" or "Soft Finish." "Black Annealed" or "Bright Annealed" wire shall not be supplied. (6) The surface shall be free from oil or grease. Tests. The commercial weldability of these electrodes shall be determined by means of tests by an experienced operator, who shall demonstrate that the wire flows smoothly and evenly through the arc without any detrimental phenomena. CHAPTER VII ARC WELDING TERMS AND SYMBOLS In order to aid the standardization of the various types of joints and welding operations the practice recommended by the Welding Committee of the Emergency Fleet Corp., for FIG. 78. Standard Symbols Eecommended by the Welding Committee of the Emergency Fleet Corporation. STRAP FIG. 79. ship work, is given. The symbol chart is shown in Fig. 78 and the application of special terms and symbols is individually shown in Figs. 79 to 112 inclusive. 109 110 ELECTRIC WELDING FIG. 79. A Strap weld is one in which the seam of two adjoin- ing plates or surfaces is reinforced by any form or shape to add strength and stability to the joint or plate. In this form of weld the seam can only be welded from the side of the work opposite the reinforcement, and the reinforcement, of whatever BUTT Fie, 80. shape, must be welded from the side of the work to which the reinforcement is applied. FIG. 80. A Butt weld is one in which two plates or surfaces are brought together edge to edge and welded along the seam thus formed. The two plates when so welded form a perfectly LAP FIG. 81. flat plane in themselves, excluding the possible projection caused by other individual objects as frames, straps, stiffeners, etc., or the building up of the weld proper. FIG. 81. A Lap weld is one in which the edges of two planes are set one above the other and the welding material so applied as to bind the edge of one plate to the face of the ARC WELDING TERMS AND SYMBOLS 111 other plate. In this form of weld the seam or lap forms a raised surface along its entire extent. FIG. 82. A Fillet weld is one in which some fixture or member is welded to the face of the plate, by welding along FILLET FIG. 82. the vertical edge of the fixture or member (see welds shown and marked A ) . The welding material is applied in the corner thus formed and finished at an angle of forty-five degrees to the plate. FIG. 83. A Plug weld is one used to connect the metals by PLUG FIG. 83, welding through a hole in either one plate A or both plates B. Also used for filling through a bolt hole as at C, or for added strength when fastening fixtures to the face of a plate by drilling a countersunk hole through the fixtures and applying the welding material through this hole, as at D, thereby fasten- ing the fixture to the plate at this point. 112 ELECTRIC WELDING FIG. 84. A Tee weld is one where one plate is welded vertically to another as in the case of the edge of a transverse bulkhead A, being welded against the shellplating or deck. This is a weld which in all cases requires exceptional care and can only be used where it is possible to work from both sides FIG. 84. of the vertical plate. Also used for welding a rod in a vertical position to a flat surface, as the rung of a ladder C, or a plate welded vertically to a pipe stanchion B, as in the case of water closet stalls. FIG. 85. A Single "V" is applied to the "edge finish " of a plate when this edge is beveled from both sides to an SINGLE "V FIG. 85. angle, the degrees of which are left to the designer. To be used when the "V" side of the plate is to be a maximum "strength" weld, with the plate setting vertically to the face of adjoining member, and only when the electrode can be applied from both sides of the work. ARC WELDING TERMS AND SYMBOLS 113 F IG . 86. Double "V" is applied to the "edge finish " of two adjoining plates when the adjoining edges of both plates DOUBLE "V SPACE ANY THICKNESS '/ 8 ' FlG. 86. beveled from both sides to an angle, the degrees of which are left to the designer. To be used when the two plates are to be "butted" together along these two sides for a maximum STRAIGHT SYMBOL z : NOTES BELOW FIG. 87. "strength" weld. Only to be used when welding can be per- formed from both sides of the plate. FIG. 87. Straight is applied to the "edge finish" of a plate, when this edge is left in its crude or sheared state. To be SINGLE BEVEL FIG. 88. used only where maximum strength is not essential, or unless used in connection Avith strap, stiffener or frame, or where it is impossible to otherwise finish the edge. Also to be used 114 ELECTRIC WELDING for a "strength" weld, when edges of two plates set vertically to each other as the edge of a box. FIG. 88. Single Bevel is applied to the edge finish of a DOUBLE BEVEL FIG. 89. plate, when this edge is beveled from one side only to an angle, the degrees of which are left to the designer. To be used for "strength" welding, when the electrode can be applied DECK PLATING FlG. 90. from one side of the plate only, or where it is impossible to finish the adjoining surface. FIG. 89. Double Bevel is applied to the edge finish of two adjoining plates, when the adjoining edges of both plates are ARC WELDING TERMS AND SYMBOLS 115 beveled from one side only to an angle, the degrees of which are left to the designer. To be used where maximum strength is required, and where electrode can be applied from one side of the work only. FIG. 90. Flat position is determined when the welding material is applied to a surface on the same plane as the deck, allowing the electrode to be held in an upright or vertical position. The welding surface may be entirely on a plane with the deck, or one side may be vertical to the deck and welded to an adjoining member that is on a plane with the deck. Horizontal position is determined when the welding material is applied to a seam or opening, the plane of which is vertical to the deck and the line of weld is parallel with the deck, TACK FIG. 91. allowing the electrode to be held in an inboard or outboard position. Vertical position is determined when the welding material is applied to a surface or seam, whose line extends in a direc- tion from one deck to the deck above, regardless of whether the adjoining members are on a single plane or at an angle to each other. In this position of weld, the electrode would also be held in a partially horizontal position to the work. Overhead position is determined when the welding material is applied from the under side of any member whose plane is parallel to the deck and necessitates the electrode being held in a downright or inverted position. FIG. 91. A Tack weld is applying the welding in small sections to hold two edges together, and should always be specified by giving the space from center to center to weld and the length of the weld itself. No particular "design of weld" is necessary of consideration. 116 ELECTRIC WELDING A Tack is also used for temporarily holding material in place that is to be solidly welded, until the proper alinement and position is obtained, and in this case neither the length, space, nor design of weld are to be, specified. FIG. 92. A Caulking weld is one in which the density of CAULKING FIG. 92. the crystalline metal, used to close up the seam or opening, is such that no possible leakage is visible under a water, oil or air pressure of 25 Ibs. per square inch. The ultimate strength of a caulking weld is not of material importance neither is the "design of weld" of this kind necessary of consideration. FIG. 93. A Strength weld is one in which the sectional STRENGTH FIG. 93. area of the welding material must be so considered that its tensile strength and elongation per square inch must equal at least 80 per cent of the ultimate strength per square inch of the surrounding material. (To be determined and specified by the designer.) The welding material can be applied in any number of layers beyond a minimum specified by the designer. The density of the crystalline metals is not of vital im- ARC WELDING TERMS AND SYMBOLS 117 portance. In this form of weld, the " design of weld" must be specified by the designer and followed by the operator. JP IG> 94. A Composite weld is one in which both the strength and density are of the most vital importance. The strength must be at least as specified for a " strength weld," and the density must meet the requirements of a "caulking weld" COMPOSITE FIG. 94. both as above defined. The minimum number of layers of welding material must always be specified by the designer, but the welder must be in a position to know if this number must be increased according to the welder's working con- ditions. FIG. 95. Reinforced is a term applied to a weld when the top layer of the welding material is built up above the plane REINFORCED FIG. 95. of the surrounding material as at A or B, or when used for a corner as at C. The top of final layer should project above a plane of 45 degrees to the adjoining material. This 45 degree line is shown " dotted" in C. This type is chiefly used in a "strength" or "composite" kind of weld for the purpose of obtaining the maximum strength efficiency, and should be speci- fied by the designer, together with a minimum of layers of welding material. 118 ELECTRIC WELDING FIG. 96. Flush is a term applied 'to a weld when the top layer is finished perfectly flat or on the same plane as on the adjoining material as shown at D and E or at an angle of 45 degrees when used to connect two surfaces at an angle to each other as at F. This type of weld is to be used where a maximum tensile strength is not all important and must be FLUSH FIG. 96. specified by the designer, together with a minimum number of layers of welding material. FIG. 97. Concave is a term applied to a. weld when the top layer finishes below the plane of the surrounding material as at G, or beneath a plane of 45 degrees at an angular con- nection as at H and J. To be used as a weld of no further importance than filling DOTTED LIVES SHOW THE FLUSH ! FIG. 97. in a seam or opening, or for strictly caulking purposes, when it is found that a minimum amount of welding material will suffice to sustain a specified pound square inch pressure with- out leakage. In this ''type of weld" it will not be necessary for the designer ordinarily to specify the number of layers of material owing to the lack of structural importance. COMBINATION SYMBOLS FIG. 98 shows a strap holding two plates together, setting vertically, with the welding material applied in not less than three layers at each edge of the strap, as well as between the plates with a reinforced, composite finish, so as to make the welded seams absolutely water, air or oil tight, and to ARC WELDING TERMS AND SYMBOLS 119 attain the maximum tensile strength. The edges of the strap and the plates are left in a natural or sheared finish. This type of welding is used for particular work where maximum strains are to be sustained. FIG. 99 shows a strap holding two plates together hori- STRAP WELD, REINFORCED, COMPOSITE OF THREE LAYERS, VERTICAL, STRAIGHT, ISP*:, PLATE ERTICAL WELD STRAP w, PLATE FIG. 98. zontally, welded as a strength member with a minimum of three layers and a flush finish. Inasmuch as the strap neces- sitates welding of the plates from one side only, both edges of the plates are bevelled to an angle, the degrees of which are left to the discretion of the designer. The edges of the STRAP WELD. FLUSH, I83HOF) STRENGTH OF 3 LAYERS, HORIZONTAL, FLAT AND OVERHEAD. DOUBLE BEVEL Fie. 99. strap are left in a natural or sheared state, and the maximum strength is attained by the mode of applying the welding material, and through the sectional area per square inch exceed- ing the sectional area of the surrounding material. FIG. 100 represents two plates butted together and welded 120 ELECTRIC WELDING flat, with a composite weld of not less than three layers, and a reinforced finish. A strap is attached by means of overhead tacking, the tacks being four inches long and spaced eight inches from center to center. In this case, the welding of the plates of maximum strength and water, air or oil tight, STRAP, TACK, OVERHEAD. 8' CENTER TO CENTER 4' LONG, BUTT, REINFORCED COMPOSITE OF 3 LAYERS, FLAT. STRAIGHT. OVERHEAD XVELD FIG. 100. but the tacking is either for the purpose of holding the strap in place until it may be continuously welded, or because strength is not essential. All the edges are left in their natural or sheared state. FIG. 101 represents a butt weld between two plates with the welding material finished concaved and applied in a mini- BUTT WELD. CONCAVE. CAULKING OF 2 LAYERS. FLAT. STRAIGHT FIG. 101. mum of two layers to take the place of caulking. The edges of the plates are left in a natural shear cut finish. This symbol will be quite frequently used for deck plating or any other place where strength is not essential, but where the material must be water, air or oil tight. FIG. 102 is used where the edges of two plates are vertically ARC WELDING TERMS AND SYMBOLS 121 butted together and welded as a strength member. The edges of adjoining plates are finished with a "double vee" and the minimum of three layers of welding material applied from each side, finished with a convex surface, thereby making the sectional area per square inch of the weld greater than that BUTT WELD. REINFORCED. STRENGTH OF 3 LAYERS. VERTICAL, DOUBLE VEE. FIG. 102. of the plates. This is a conventional symbol for shell plating or any other members requiring a maximum tensile strength, where the welding can be done from both sides of the work. FIG. 103 shows two plates butted together in a flat position where the welding can only be applied from the top surface. It shows a weld required for plating where both strength and 93F BUTT WELD. FLUSH, COMPOSITE OF 3 LAYERS. FLAT. DOUBLE BEVEL. Fie. 103. watertightness are to be considered. The welding material is applied in a minimum of three layers and finished flush with the level of the plates. Both edges of the adjoining plates are beveled to an angle, the degrees of which are left to the discretion and judgment of the designer, and should only be used when it is impossible to weld from both sides of the work. 122 ELECTRIC WELDING FIG. 104 shows the edges of two plates lapping each other with the welding material applied in not less than two layers at each edge, with a concaved caulking finish, so applied, as to make the welded seams absolutely water, air or oil tight. LAP WELD. CONCAVE. CAULKING OF 2 LAYERS, OVERHEAD AND FLAT STRAIGHT OVERHEAD WELD FIG. 104. The edges of the plates themselves are left in a natural or shared finish. Conditions of this kind will often occur around bulkhead door frames where maximum strength is not ab- solutely essential. FIG. 105 is somewhat exaggerated as regards the bending LAP WELD, REINFORCED. STRENGTH OF 3 LAYERS AND TACKING, 18' CENTER TO CENTER, 6 LONG, VERTICAL, STRAIGHT. 'I FIG. 105. of the plates, but it is only shown this way to fully illustrate the tack and continuous weld. It shows the edges of the plates lapped with one edge welded with a continuous weld of a minimum of three layers with a reinforced finish thereby giving a maximum tensile strength to the weld, and the other ARC WELDING TERMS AND SYMBOLS 123 edge of the plate, tack welded. The tacks are six inches long with a space of 12 inches between the welds or 18 inches from center to center of welds. In both cases, the edges of the plates are left in a natural or sheared state. PLUG AND LAP WELD, STRENGTH OF 3 LAYERS FLUSH. FLAT, OVERHEAD, HORIZONTAL. FLAT WELD 3 2 1 FIG. 106. FIG. 106 shows a condition exaggerated, which is apt to occur in side plating where the plates were held in position with bolts for the purpose of alinement before being welded. The edges are to be welded with a minimum of three layers of welding material for a strength weld and finished flush, PLUG AND FILLET WELD, REINFORCED, STRENGTH OF 3 LAYERS, FLAT, SINGLE BEVEL AND STRAIGHT. FIG. 107. and after the bolts are removed, the holes thus left are to be filled in with welding material in a manner prescribed for strength welding. The edges of the plates are to be left in a natural or sheared state, which is customary in most cases of lapped welding. 124 ELECTRIC WELDING PIG. 107 shows a pad eye attached to a plate by means of a fillet weld along the edge of the fixture, and further strengthened by plug welds in two countersunk holes drilled in the fixture. The welding material is applied in a flat position for a strength weld with a minimum of three layers FILLET WELD. REINFORCED. COMPOSITE OF 3 LAYERS, FLAT. VERTICAL AND OVERHEAD. STRAIGHT. PIG. 108. and a reinforced finish. The edges of the holes are beveled to an angle, which is left to the judgment of the designer, but the edges of the fixture are left in their natural state. This method is used in fastening fixtures, clips or accessories that would be subjected to an excessive strain or vibration FILLET WELD, FLUSH, STRENGTH OF 3 LAYERS FLAT. STRAIGHT. FlG. 109. FIG. 108 shows a fixture attached to a plate by means of a composite weld of not less than three layers with a reinforced finish. The fixture being placed vertically, necessitates a com- bination of flat, vertical and overhead welding in the course of its erection. Although a fixture of this kind would never ARC WELDI'NG TERMS AND SYMBOLS 125 be required to be watertight, the composite symbol is given simply as a possibility of a combination. FIG. 109 represents a fixture attached to a plate by a strength fillet weld of not less than three layers, finished flush. TEE WELD. FLUSH. STRENGTH OF 3 LAYERS. FLAT. SINGLE VEE. FIG. 110. The edges of the fixture are left in their natural state, and the welding material applied in the corner formed by the vertical edge of the fixture in contact with the face of the plate. FIG. 110 illustrates the edge of a plate welded to the face of another plate, as in the case of the bottom of a transverse TEE WELD. REINFORCED. STRENGTH OF 3 LAYERS. VERTICAL. SINGLE VEE. FIG. 111. bulkhead being welded against the deck plating. To obtain a maximum tensile strength at the joint, the edge of the plate is cut to " single vee" and welded on both sides with a strength weld of not less than three layers, and finished flush. This would be a convenient way of fastening the intercostals to 126 ELECTRIC WELDING the keelsons. In this particular case, the welding is done in a flat position. FIG. Ill shows another case of tec weld with the scam set- ting in a vertical position, and the welding material applied from both sides of the work. The edge of the plate is finished with a " single vee" and a minimum of three layers of welding material applied from each side, finished with a convex surface, thereby making the sectional area, per square inch of the weld, greater than that of the plate, allowing for a maximum tensile strength in the weld. FIG. 112 represents an example of the possible combination STRAP AND TEE WELD, FLAT, REINFORCED, TACK, 12' CENTER TO CENTER, 6' LONG, SINGLE BEVEL, OVERHEAD, STRENGTH OF 3 LAYERS, FLUSH FIG. 112. of symbols. An angle iron is tack welded to the plate in the form of a strap or stiffener, though in actual practice, this might never occur. The tacks are spaced twelve inches from center to center, and are six inches long, and applied in a flat position, with a reinforced finish. As the strap prevents welding the plate from both sides, the edge of the plate is beveled, and the welding material applied for strength in not less than three layers in an overhead position and finished flush. Note that in specifying tack welds, it is essential to give the space from center to center of weld, and length of weld by use of figures representing inches placed either side of the circumscribing symbol of the combination. CHAPTER VIII EXAMPLES OF ARC-WELDING JOBS Probably no mechanical job ever attracted more general attention than the repair of the German ships seized by us when we entered the World War. Even the mechanically minded Germans repeatedly declared that repairing was an impossibility, but the American engineers and mechanics showed the Hun that he had, as usual, vastly over-rated his own knowledge. One big factor in making the Hun so positive in this case, was his utter ignorance regarding the possibilities of arc welding but he learned and in the teaching many others were also enlightened. The work necessary on these German ships, of course, in- cluded much besides welding of the broken castings, but the welding work was of primary importance. The principal ships on which this welding work was done were the : Grqss Class of U. S. Name German name I.H.P. Tonnage Vessel Aeolus Grosser Kurf urst 8,400 13,102 Transport Agamemnon Kaiser Wilhelm IT .... 45,000 19,361 Transport America America 15,800 22,621 Transport Antigne Neckar 5,500 9,835 Transport Covington Cincinnati 10,900 16,339 Transport George Washington. George Washington .... 21,000 25,570 Transport Huron Friedrich der Grosse. . 6,800 10,771 Transport Leviathan Vaterland 90,000 54,282 Transport Maclawaska Koenig Wilhelm II.... 7,400 9,410 Transport Martha Washington Martha Washington . . . 6,940 8,312 Transport Mercury Barbarossa 7,200 10,984 Transport Mt. Vernon Kronprinzessin Cecelie. 45,000 19,503 Transport Pocahontas Prinzess Irene 9,000 10,983 Transport Powhatan Hamburg 9,000 10,893 Transport President Grant . . . President Grant 8,500 18,072 Transport President Lincoln. . President Lincoln 8,500 18,168 Transport Savannah Saxonia 2,500 4.424 Repair Shop Susquehanna Ehein 9,520 10.058 Transport Philippines Bulgaria 4,200 10,924 Shipping Bd. 127 128 ELECTRIC WELDING The total gross tonnage of the ships named was 288,780 tons, and the welding work was done by the Wilson Welder and Metals Co. of New York, using their "plastic-arc" process. Seventy Cylinders Saved Without Replacement. In all, there were thirty-one ships interned in the port of New York. Of these thirty-one ships, twenty-seven were German and four Austrian. Of the German ships, two were sailing vessels and four were small steamers which the Germans had not taken pains to damage materially. This left twenty-one German ships whose engines and auxiliaries were damaged seriously, ranging in size from the "Vaterland," the pride of the Ham- burg-American Line, of 54,000 tons, to the "Nassovia," of 3,900 tons. On the cylinders of the twenty vessels of German origin, not counting for the moment the turbine-driven "Vaterland," there were no less than 118 major breaks which would have entailed the renewal of some seventy cylinders if ordinary practice had been followed. In fact, such was the recommenda- tion of the surveying engineers in their original report. To any engineer familiar with the conditions at that time in the machine shops and foundries in the vicinity of New York, also in the drafting rooms, the problem of producing seventy cylinders of the sizes required by these vessels would seem almost impossible, and it is pretty well established that some vessels would have had to wait nearly two years for this equipment. It must be remembered that few drawings of these engines were available, and those in many cases were not discovered until months after the repairs had started. Therefore, it would have been necessary to make drawings from the actual cylinders, and competent marine engine draftsman not already flooded with work did not exist. The cylinders of fifteen vessels were successfully welded, while those of six were repaired by fitting mechanical patches, or, in other words, eighty-two of the major breaks were repaired by welding and thirty-six by mechanical patches. It was not until July 12 that the final decision was made placing the transport service in the hands of the Navy and designating what ships were to be transferred from the control of the Shipping Board to that of the Navy Department. How- EXAMPLES OF ARC-WELDING JOBS 129 ever, the first two large ships, the "Friedrich der Grosse," now the " Huron," and the "Prinzess Irene," now the "Poca- hontas," were ready for sea on Aug. 20, in spite of the fact that the engines on these vessels were among the worst damaged of them all, the " Irene" having the whole side of the first intermediate valve chest broken out on each engine, the side of the high-pressure cylinder on each engine destroyed, and other smaller breaks, which, under ordinary methods, would have necessitated the renewal of four cylinders. The "Fried- ^ rich der Grosse ' ' had the following breaks : Broken valve chest Z 2: of high-pressure cylinder of each engine (valve chest cast in Q 5 one with the cylinder), flanges knocked off both valve chest t and cylinder covers, steam inlet nozzles knocked off both first ^ . intermediate valve chests and walls between the two valves O in each check broken out, also steam inlet nozzles on both U- Jj second intermediate valve chests broken off. ^ These two vessels were the first in which straight electric O welding was used, that is, where patches were not bolted to 0) \r the cylinder walls. U iu Method of Repair. The nature of some of the breaks in castings is shown by the accompanying photographs, which ? IE were taken at various stages of the work. A, Fig. 113, shows the break in the starboard high-pressure cylinder of the North German Lloyd steamer " George Wash- ington. ' ' This break was effected by drilling a row of holes about an inch apart and knocking the piece out with a ram. To prepare this for welding it was necessary to chisel off the surface only roughly, build a pattern of the break, cast a steel piece from the pattern, stud up the surface of the cast iron of the cylinder with a staggered row of steel studs f in. in diameter, projecting | in. from the cylinder, bevel the edge of the cast piece, place the piece in position as shown in B, and make the weld. When completed, the appearance of the work is as it appears in C. The broad belt of welded metal is due to the laying of a pad of metal over the rows of studs previously noted. It cannot be too strongly insisted that tests have shown con- clusively that the weld can be properly made without this pad ; that is, if the approximate strength of the original metal is all that is desired in which case the studding of the metal is 130 ELECTRIC WELDING a a EXAMPLES OF ARC-WELDING JOBS 131 unnecessary. But the work in these particular cases was of vital importance, due to the uses to which the vessels were to be put when in service, and also it was appreciated that this exhibition of a new application of the art in the marine engineer- ing world required that the demonstration be satisfying, not only to the mind of the engineer, but to the eye, and ear, and when any engineer looked at that band of metal and sounded it with a hammer, he could not be but satisfied that the strength was definitely there and that the method of padding could be used in most of the situations which would arise. This at least was the effect upon all the engineers who saw the actual work. The metal was laid on in layers in such a manner as to take care of the contraction in cooling. Each successive layer was cleaned with a wire brush before the next layer was put on. It is in the keeping of the successive layers clean and in the laying on of the metal so as to take care of the con- traction that the operator's ability comes in fully as much as it does in the handling of the apparatus. The 'cylinders were not removed, but were repaired in place. Thus the work of fitting was reduced to a negligible quantity, and the refitting of lagging was not interfered with by projections, other than the f-in. pad, which is laid over the studs for extra strength. It will also be noted that these repairs can be undertaken at any place where the vessel may be lying, cither at her loading dock or in the stream, since such apparatus may be carried on barges, which can be placed alongside and wires run to the work. In this work a part consisted of the caulking of the surface of the welds which prevents porosity and also locates any brittle spots or places where poor fusion of metal has been obtained. This permits the cutting out of the bad places and replacing with good metal. The tool used was an air caulking hammer operated at 110 Ib. air pressure. Strength of Cast-iron Welds. Capt. E. P. Jessop, U. S. N., personally tested many welds for tensile strength in which cast iron was welded to cast steel, and in but one case was there a failure to obtain practically the original strength. This case was due to an inexperienced operator burning the metal, and was easily detected as an inferior weld without the strength test being applied. 132 ELECTRIC WELDING Much has been said about the effect of the heat of welding, upon the structure or strength of cast iron, and in this particular instance the Navy engineer who had direct charge of this work, made experiments to note if there were any deleterious effects on the iron resulting from the action of the weld and reported as follows: ' ' Scleroscopic investigation of the structure of the welds shows only a very slight vein of hard cast iron at the line of the weld, shot through with fingers of gray cast iron, while behind this area there was no heat effect whatever. The metal thus deposited was easily workable with ham- mer and chisel, file or cutting tool. Another very important feature is that with the use of the low voltage and absolute automatic current control of the Wilson system, there is a minimum of heat transmitted to the parts to be welded, this being practically limited to a heat value absolutely necessary to bring the electrode and the face of the metal to be welded into a semi-plastic state, thus insuring a perfect physical union, and in accomplishing this result neither of the metals suffers from excessive heat, and there is absolutely no necessity for pre-heating. Neither are there any adverse results from shrinkage following the completed work owing to a minimum amount of heat being transmitted to the repair parts, thus avoiding the possibility of distortion of parts through uneven or excessive shrinkage strains that are very common where pre-heating is necessary or excessive heat is used for fusing metals." A, Fig. 114, shows the damage done to the first intermediate cylinder of the U. S. S. ' * Pocahontas, " formerly the "Prinzess Irene. " The damage to this cylinder, it will be noted, was more destructive than to that of the ' ' George Washington, ' ' rendering the repairs much more difficult. B shows the steel section in place ready for welding, with the surfaces properly V'd out and with a staggering row of steel studs adjacent to the welding edge of the cylinder section. C shows the complete job with the extra band or pad of metal completely covering the studs on the cast-iron section. These bands or pads of metal are peaned or worked over with a pneumatic hammer to insure protection against porosity of metal. Had either or both of these cylinders been fractured on the lines shown of the cast-iron sections, and none of the parts removed, then the surfaces or edges of all lines of fracture would have been V'd out, and the weld made of the two cast- iron surfaces in the same manner that the cast steel was welded to the cast-iron cylinder proper. EXAMPLES OF ARC-WELDING JOBS 133 I r Jl 3d {3 I g . o "7^^; c g | c c "^ .- >>u w 5 |= ' ti- ! ffllii i g ^s> : 8 S> > II f.Si 11 at 09 a a * 22 gg bfc gjg, i'i s 22 s c e ecscs cc c e c' c s M! la 1 1111 O/o; .- .s.s.s.s P^gl |1111 5 U-> U. fc, U OU*> fN OOOOO OO 1 H B SSjr 0-ca 174 ELECTRIC WELDING permit a tension specimen (0.505 in. diameter, 2 in. gage length) to be machined out of it. Although the opinion is held by some welders that the properties of the metal of an arc-weld are affected materially by the adjacent metal by reason of the interpenetration of the two, it was decided that the change of properties of the added metal induced by the fusion alone was of fundamental importance and should form the basis of any study of arc-welding. The method adopted also per- mitted the use of larger specimens with much less machining -Weight HO V. DC. A tfj us to hie Wa ter ffheos fat ?/o Sodium Chloride Solution feet Welding ToblQ . FIG. 149. Arrangement of Apparatus for Welding. than would have been possible had the metal been deposited in the usual form of a weld. In the first few specimens prepared (ten in number) the metal was deposited by a series of '"headings" inside a l^-in. angle iron. The tension specimens cut from the deposited metal were found to be very inferior and entirely unsuitable for the study. This was largely on account of the excessive overheating which occurred as well as the fact that a relatively "long arc" was necessary for the fusion in this form. Because of the very evident inferiority of these specimens, the results of the mechanical tests made are not given in the tables. The method of deposition of the metal was then changed to PHYSICAL PROPERTIES OF ARC-FUSED STEEL 175 that shown in Fig. 150. This method also had the advantage in that the amount of necessary machining for shaping the specimens for test was materially reduced. The block of arc- Side View 12"- >, End View FIG. 150. Method of Formation of the Blocks of Arc-Fused Metal. fused metal was built up on the end of a section of J-in. plate of mild steel (ship plate) as shown. When a block of sufficient size had been formed, it, together with the portion FIG. 151. Block of Arc-Fused Metal with Tension Specimen Cut from It. Approximately Half Natural Size. of the steel plate immediately beneath, was sawed off from the remainder of the steel plate. The tension specimen was turned entirely out of the arc-fused metal. No difficulty what- ever was experienced in machining the specimens. Fig. 151 176 ELECTRIC WELDING shows the general appearance of the block of fused metal as well as the tension specimen turned out of it. In general in forming the blocks, the fused metal was deposited as a series of " beads" so arranged that they were parallel to the axis of the tension specimen which was cut later from the block. In two cases, for purposes of comparison, the metal was deposited in " beads" at right angles to the length of the specimen. In all the specimens, after the deposi- tion of each layer, the surface was very carefully and vigor- ously brushed with a stiff wire brush to remove the layer of oxide and slag which formed during the fusion. There was found to be but little need to use the chisel for removing this layer. Two types of electrodes were used as material to be fused. These differed considerably in composition as shown in Table IX, and were chosen as representative of a "pure" iron and a low-carbon steel. The two types will be referred to as "A" and "B" respectively in the tables. They were obtained in the following sizes: Y 8 , 5 / 32 , 3 / 16 an d V 4 i n - ("A" electrode 5 /ic in. ) . It was planned to use the different sizes with the follow- ing currents: Y 4 in. 75, 110 and 145 amp.; 5 / 32 i n - 145, 185 and 225 amp.; 3 / 16 in. 185, 225 and 260 amp.; y 4 in. ( 5 /ie in.) 300 amp. The electrodes were used both in the bare condition and after being slightly coated with an oxidizing and refractory mixture. For coating, a "paste" of the follow- ing composition was used: 15 g. graphite, 7.5 g. magnesium, 4 g. aluminium, 65 g. magnesium oxide, 60 g. calcium oxide. To this mixture was added 120 c.c. of sodium silicate (40 deg. Be.) and 150 c.c. of water. The electrodes were painted on one side only with the paste. The quantity given above was found to be sufficient for coating 500 electrodes. The purpose of the coating was to prevent excessive oxidation of the metal of the electrode during fusion and to form also a thin protective coating of slag upon the fused metal. Tension specimens only were prepared from the arc-fused metal. It is quite generally recognized that the tension test falls very short in completely defining the mechanical proper- ties of any metal; it is believed, however, that the behavior of this material when stressed in tension is so characteristic that its general behavior under other conditions of stress, PHYSICAL PROPERTIES OF ARC-FUSED STEEL 177 o CA r^* T I/-N r*4 IN f^ o oo rM fM f^ ^ffs N^O -*Air> ^ So ^ ooo :S op < do : odd lo oo D! 2 a* i S o o S 5 d o 85 o : ddo id : dd : 55 : d dd i^tx f ^'*^ oO * O ~ -O So . ooo - oo 000 o -00 ddo o -dd " 8 8 o 2 i II I II I I! II ^ ii 1 lllll lllr! 1 |j go i c CO A2 110 49,850 36,600 25.000 60 65 108 A3 145 51.950 36,250 30,000 80 13 114 fl fc 145 185 47.550 48.100 , 6.0 8.0 74 87 108 104 A9 A 225 45,500 '.:. . 80 96 io< A4 A 185 50,600 33,750 29,500 5.5 13 5 105 A5 A 225 49,150 36,250 22,000 7.0 10 102 A6 A 260 50.950 33,750 28,800 10.5 12.0 107 AIO A 300 46,670 ..... 12.0 119 104' Covered Electrodes AD2 j 110 51,250 35,000 25,600 9.5 II. 103 AD2-D no 43,000 23.000 50 90 AD3 MS 51,100 33,750 25,000 8.5 10*5 110 AD3-D 145 46,250 . .. 24.250 7.0 12 AD7 A 145 41,750 60 66 99 AD7-D A 145 46,950 25.500 80 94 AD8 A 185 44,620 6.5 58 103 AD8-D A 185 43,600 23,250 6.5 9 AD9 A 225 46,900 95 10 1 96 AD9-D A 225 41,550 25,500 5.0 6.5 AD4 A 185 51,200 35,000 30,000 10.5 10.5 101 AD4-D A 185 45,700 25,500 8.5 11.5 AD5 A 225 48.600 35,000 30,000 7.0 10 '96 AD5-D A 225 46.250 23,750 11.5 12 AD6 A 260 47,500 34,500 31,500 9.0 9.0 97 AD6-D A 260 50,700 8.0 28 105 ADIO A 300 45,900 8.5 11.5 98 Bare Electrodes B2 i 110 52,650 37,000 27,000 7.5 7.5 1 14 B3 i 145 54,500 36,000 27,000 12 5 12 106 B4 A 145 46,450 33,500 26,000 50 7 102 185 49,600 34,250 27,000 75 9 108 B6 A 225 49,500 30,500 28,000 90 5 110 j? 185 47,550 28,500 75 11.5 95 B8 A 225 42,900 18,750 75 16.2 101 B9 A 260 47,500 21,500 12 13 5 102 Covered Electrodes BD2 } 110 49,050 33,750 27,500 9.0 12.0 100 BD2-D 110 44,400 20,000 6.5 9.4 BD3 145 52,100 34.300 30,500 12 5 16 116 BD3-D 145 50,850 23,500 13 17 5 BD4 A 145 48,130 31,000 30,500 8.0 10 toi BD4-D A 145 41,750 21,000 6.0 95 BD5 A 185 49,086 31,730 29,000 12 5 130 97 BD5-D A 185 47,100 22,500 II 12 5 BD6 A 225 45,500 30,500 25,000 8 5 10 5 95 BD7 A 185 49,950 24,500 II 5 21 5 98 BD7-D A 185 51,150 23,750 14 5 19 5 BD8 A 225 41,500 17,850 6.0 12 7 99 BD8-D(?) A(?) 225(?) 48,750 21,250 12.5 16 BD9 A 260 46,350 24,000 10.0 15 99 Bare Electrodes Cl A 175 48,650 32,650 23,000 12.0 19 1 C2 A 175 45,200 32,400 23,000 7.5 16.6 85 * 175 175 49,720 32,650 54,500 32,500 25,000 25,000 90 110 13.6 17 5 118 C5 A 175 50,900 32,500 24,000 15 23.0 109 C6 A 175 50.500 33.500 23.000 12 16 PHYSICAL PROPERTIES OF ARC-FUSED STEEL 181 1 [1 M P o 1^1 oo m o o ->o **** ' * {j JjflC g c oo -t- oo < m r>i rx ^-vO Jj-~ o o o o o o o oo i " B > -S < Q M 8 \ CCOQ ~ -5 S \ 1 1 hJ ^g I* 5 "5 !r 5 1 5 s - <0 -o 1 1 A -o -o . -o> O oo * ~r s 3 S 1<3 1 ^ - "* " ^* ^ ++ 2J : c ^ PERTIES AND HARDNESS OF FlFTY SPE ARRANGED TN ORDER OF A Tensile Properties Yield Point . Elongat Lb. Sq.ln. pei JO O o m * -o ! " 'O o in ** -V ! O O ..... o . O o 'O BS ' "^ Q "^ 1 5 *3 ^ ' **- :co - Av 35,000 Av. 33.250 Av. 7.9 lectrodes used (Table III) = 110 amps, and 145(1) amps. & tn. rft'am. 145 P o H M ^1 "5^-5o--rC*'- S . ' S - :5 : : i< lo|l ft S g C-3 v lfTVTg g -- 35! ^j * ^ S 3; 5 S 5 5 S : :? S 1* goc; ^ooooooooo o oooo 5Jl|S J, HH H S m E^ <: ll < si o ^ ^ ^~ o oS ^~ C9* H 182 ELECTRIC WELDING In Table XII are given the results of the mechanical tests made upon the tension specimens which were turned out of the blocks of metal resulting from the fusion of the elec- trodes. The specimens listed, C x , C 2 . . . . C c are the six which were prepared outside the Bureau and submitted for purposes of comparison. It was stated that they were prepared from bare electrodes 5 / 32 in. diameter of type "B," containing 0.17 per cent carbon and 0.5 per cent manganese. As an aid for more readily comparing the mechanical prop- erties of the two types of arc-fused metal "A" and "B," the results have been grouped as given in Table XIII. The characteristic appearance of specimens after testing, illustrating their behavior when stressed in tension till rupture occurs is shown in Fig. 153. These represent two views of the face of the fracture, one in which the line of vision is perpendicular to the face, the other at an angle of 45 deg., together with a side view of the cylindrical surface of the specimen. The features shown are characteristic of all the specimens tested, though in some they were much more pro- nounced than those shown. The fracture of the specimen in all cases reveals interior flaws. In some of the specimens, however, these are microscopic and of the character to be discussed in a subsequent chapter on Metallography. Although many of the specimens (from the results of Table XII) appear to have a considerable elongation, it is seen from Fig. 153 that the measured elongation does not truly represent a prop- erty of the metal itself. It is due rather to interior defects which indicate lack of perfect union of succeeding additions of metal during the process of fusion. The surface markings of the specimen after stressing to rupture are very similar to those seen in the familiar "flaky steel." Resulting Physical Properties Depend Essentially on Sound- ness. It appears from the results above that, as far as the mechanical properties are concerned, nothing was gained by coating the electrodes. The results show no decided superiority for either of the two types of electrodes used. This may be expected, however, when one considers that the two are rendered PHYSICAL PROPERTIES OF ARC-FUSED STEEL 183 practically the same in composition during fusion by the burn- ing out of the carbon and other elements. The results of the tension tests upon the "C" series of FIG. 153. Characteristic Appearance of Tension Specimen After Test. At top, face of fracture, viewed normally. Middle, fractured end of specimen, viewed at an angle of 45 deg. At bottom, cylindrical surface of specimen. Mag- nification, X 2. specimens which were made outside of the Bureau and sub- mitted to be included in the investigation, show no marked difference between these samples and those prepared by the Bureau. In all cases the results obtained in the tension test 184 ELECTRIC WELDING are determined by the soundness of the metal and do not necessarily indicate the real mechanical properties of the material. The results of the hardness determinations do not appear to have any particular or unusual significance. The variations are of the same general nature and relative magnitude as the variations observed in the results of the tension test. In general the higher hardness number accompanies the higher tensile values, though this was not invariably so. As previously noted, specimens were prepared for the purpose of showing the relation between the direction in which the stress is applied and the manner of deposition of the metal. The metal was deposited in the form shown in Fig. 151, except that the "beads" extended across the piece rather than lengthwise, hence the "beads" of fused metal were at right angles to the direction in which the tensional stress was applied. The results of the tension tests show that these two specimens (AW t and AW 2 ) were decidedly inferior to those prepared in the other manner as shown in Table XIV. TABLE XIV. MECHANICAL PROPERTIES OF ARC-FUSED METAL DEPOSITED AT EIGHT ANGLES TO LENGTH OF SPECIMEN Specimen Ult. Strength, Lb. Sq. In. Proportional Limit, Lb. Sq. In. Elongation in 2 in. (per Cent) Bed. of Area, per Cent AW 1 40,450 22,500 6.5 8.5 AW 2 39,500 22,500 4.0 3.0 Macrostructure. The general condition of the metal result- ing from the arc-fusion is shown in Figs. 154 and 155, which show longitudinal median sections of a series of the tension bars adjacent to the fractured end. The metal in all of these specimens was found to contain a considerable number of cavities and oxide inclusions, these are best seen after the surfaces are etched with a 10 per cent aqueous solution of copper-ammonium chloride. In many of the specimens the successive additions of metal are outlined by a series of very fine inclusions (probably oxide) which are revealed by the etching. There appears to be no definite relation between the soundness of the metal and the conditions of deposition i.e., for the range of current density used nor does either type PHYSICAL PKOPEKTiES OF ARC-FUSED STEEL 185 FIG. 154. Macrostructure of Arc-Fused Metal, Type A. Medial Longitudinal sections of the tension bars indicated were used XII) ; etching, 10 per cent aqueous solution of copper-ammonium chloride. nification, X 2. From top to bottom in order: Ai)6 A electrode; iMo in., covered, 260 amp. A5 A electrode; iMe in., bare, 225 amp. A6 A electrode; Me in., bare, 260 amp. A3 A electrode; i in., bare, 145 amp. A4 A electrode; Me in., bare, 185 amp. AD2 A electrode; i in., covered, 110 amp. (Table Mag- 186 ELECTRIC WELDING of electrode used show any decided superiority over the other with respect to porosity of the resulting fusion. In Fig. 156 Fie. 155. Macrostructure of Arc-Fused Metal, Type B. Medial longitudinal sections of the tension bars indicated were used (Table XII) ; etching, 10 per cent aqueous solution of copper-ammonium chloride. Magnification, X 2. From top to bottom in order: B4 B electrode; %2 in., bare, 145 amp. B5 B electrode; %z in., bare, 185 amp. B2 B electrode; J in., bare, 110 amp. B3 B electrode; | in., bare, 145 amp. BD6 B electrode; 5 /te in., covered, 225 amp. BD4 B electrode; %'i in., covered, 14T> amp. is shown the appearance of a cross-section of one of the blocks of arc-fused metal prepared outside of the Bureau by skilled PHYSICAL PROPERTIES OF ARC-FUSED STEEL 187 welding operators. The condition of this material is quite similar to that prepared by the Bureau. The microscopic study of the material to be discussed in a subsequent chapter also revealed further evidence of unsound- ness in all three types, "A," n B" and "C." Discussion of Results. In any consideration of electric-arc welding it should constantly be borne in mind that the weld- FIG. 156. Macrostructure of Arc-Fused Metal, Type C. Specimen Cl (Table XII), cross-section of the block of arc-fused metal from which the tension bar was turned; etched with 5 per cent alcoholic solution of picric acid. Magnification, X 1.7. metal is simply metal which has been melted and has then solidified in situ. The weld is essentially a casting, though the conditions for its production are very different from those ordinarily employed in the making of steel castings. The metal loses many of the properties it possesses when in the wrought form and hence it is not to be expected that a fusion weld made by any process whatever, will have all the proper- ties that metal of the same composition would have when in the forged or rolled condition. A knowledge of the char- 188 ELECTRIC WELDING acteristic properties of the arc-fused iron is then of funda- mental importance in the study of the electric-arc weld. The peculiar conditions under which the fusion takes place also render the metal of the weld quite different from similar metal melted and cast in the usual manner. It is seemingly impossible to fuse the metal without serious imperfections. The mechanical properties of the metal are dependent there- fore to an astonishing degree upon the skill, care and patience of the welding operator. The very low ductility shown by specimens when stressed in tension is the most striking feature observed in the mechanical properties of the material as revealed by the tension test. As explained above, the measured elongation of the tension specimen does not truly indicate a property of the metal. Due to the unsoundness, already described in the discussion of the structure, the true properties of the metal are not revealed by the tension test to any extent. The test measures, largely for each particular speci- men, the adhesion between the successively added layers which value varies considerably in different specimens due to the unsoundness caused by imperfect fusion, oxide and other inclu- sions, tiny enclosed cavities and similar undesirable features. The elongation measured for any particular specimen is due largely, if not entirely, to the increase of length due to the combined effect of the numerous tiny imperfections which exist throughout the sample. That the metal is inherently ductile, however, is shown by the behavior upon bending (later to be discussed) in the microstructure of bent specimens. The formation of slip-bands within the ferrite grains to the extent which was observed is evidence of a high degree of ductility. It appears, however, that the grosser imperfections are sufficient to prevent any accurate measurement of the real mechanical properties of the metal from being made. The conclusion appears to be warranted therefore that the changes of composition which the fusion entails, together with the unusual features of micro- structure which accompany the composition change are of minor importance in determining the strength, durability and other properties of the arc weld. In arc-fusion welds in general, the mass of weld-metal is in intimate contact with the parts which are being welded so that PHYSICAL PROPERTIES OF ARC-FUSED STEEL 189 it is claimed by many that because of the diffusion and inter- mingling of the metal under repair with that of the weld, properties of the latter are considerably improved. The com- parison shown in Table XV somewhat supports this claim. The nearest comparison found available with the Bureau's specimen are some of those of the welds designated as the "Wirt- Jones" series reported by H. M. Hobart. These welds were of the 45 deg. double-V type made in ^-in. ship plate; the specimens for test were of uniform cross-section iXi in., the projecting metal at the joint having been planed off even with the surface of the plates and the test bars were so taken that the weld extended transversely across the specimen near the center of its length. The electrodes used were similar to those designated as type "B" in the Bureau's investigation. TABLE XV. COMPARISON OF WELDS WITH TESTS OF ARC-FUSED METAL PREPARED UNDER SIMILAR CONDITIONS. . Bureau of StanrlorHa TXTtw* T^^> a tie C ^ 1. i ' 4- 1 J e |4 Cu.S I a *F c H ga K B gfl i c o c in" 5 6* 85** d S""" * 110 52.650 75 1 110 45,800 8 Jt 110 49.050 90 1 115 58,200 14 i* no 44,400 6 5 I 115 59,400 13 Average 48,700 7 7 1 120 53,700 7 A 145 46,450 5 120 57,600 8 ft 11 145 A' 45 48,130 41,750 8 Average 150 54,940 60,900 10 8 Average 45.440 6 3 A 155 62,600 11 5 A 185 49.600 7 5 A verage 61.750 9.8 A 185 49.086 12 5 H 175 59.800 9 A 185 47.100 11 Average 48.395 10 3 * Electrodes were used in bare condition. t Electrodes were coated a this column were used bare s previously described , those not so designated in Since the specimens used in work described in the fore- going sections were prepared in a manner quite different from the usual practice of arc-welding, no definite recommendations applicable to the latter can be made. It appears, however, from the results obtained that the two types of electrodes used i.e., "pure" iron and low-carbon steel should give very similar results in practical welding. This is due to the changes which occur during the melting so that the resulting fusions are essentially of the same composition. The use of a slight 190 ELECTRIC WELDING coating on the electrodes does not appear to be of any material advantage so far as the properties of the resulting fused metal are concerned. Since the program of work as carried out did not include the use of any of the covered electrodes which are highly recommended by many for use in arc welding, particularly so, for " overhead work," no data are available as to the effect of such coatings upon the properties of the metal resulting from fusion. Although all of the specimens used in the examinations were made by the use of direct current, it appears from the results obtained with a consider- able number of welds representing the use of both kinds of current, submitted for the preliminary examinations which were made, that the properties of the fused metal are inde- pendent of the kind of current and are influenced primarily by the heat of fusion. Any difference in results obtained by welding with alternating current as compared with those obtained with direct current apparently depends upon the rela- tive ease of manipulation during welding rather than to any intrinsic effect of the current upon properties of the metal. CHAPTER X METALLOGRAPHY OF ARC-FUSED STEEL The same authors responsible for the description of the investigations at the Bureau of Standards, given in the previous chapter, also furnished the data given in this chapter: Fusion welds evidently are fundamentally different from other types of joints in that the metal at the weld is essentially a casting. A preliminary study of a considerable number of specimens welded under different conditions confirmed the impression that the arc-fusion weld has characteristics quite different from other fusion welds. In the present study, of which both the previous chapter and this one form a part, two types of electrodes, a "pure" iron called "A" and a mild steel called "B," were used, in the bare condition, and also after receiving a slight coating. With these were included a set of similar specimens prepared outside of the Bureau by expert welding operators. During the fusion the composition of the metal of the two types of electrodes is changed considerably by the " burning-out " of the carbon and other elements, the two becoming very much alike in composition. A very considerable increase in the nitrogen content occurs at the same time, as shown by chemical analysis. The mechanical properties of the arc-fused metal as measured by the tension test are essentially those of an inferior casting. The most striking feature is the low ductility of the metal. All of the specimens showed evidence of unsoundness in their structure, tiny inclosed cavities, oxide inclusions, lack of intimate union, etc. These features of unsoundness are, seemingly, a necessary consequence of the method of fusion as now practiced. They determine almost entirely the mechanical properties of the arc-fused metal. The observed elongation of the specimen under tension is due to the combined action of 191 192 ELECTRIC WELDING the numerous unsound spots rather than to the ductility of the metal. That the metal is inherently ductile, however, will be shown by the changes in the microstructure, produced by cold-bending. By taking extreme precautions during the fusion, a great deal of the unsoundness may be avoided and the mechanical properties of the metal be considerably im- proved. The specimens described, however, are more repre- sentative of actual present practice in welding. General Features of Microstructure. For purposes of com- parison the microstructure of the electrodes before fusion is shown in (1) and (2), Fig. 157. The "A" electrodes have the appearance of steel of a very low carbon content ; in some cases they were in the cold-rolled state ; all showed a consider- able number of inclusions. The "B" electrodes have the struc- ture of a mild steel and are much freer from inclusions than are those of the other type. It is, undoubtedly true, however, that the condition of the arc-fused metal with respect to the number of inclusions is a result of the fusion rather than of the initial state of the metal. It is to be expected that the microstructure of the material after fusion will be very considerably changed, since the metal is then essentially the same as a casting. It has some features, however, which are not to be found in steel as ordinarily cast. The general type of microstructure was found to vary in the different specimens and to range from a condition which will be designated as "columnar" to that of a uniform fine equi- axed crystalline arrangement as shown at 3 and 4, Fig. 157A. This observation held true for both types of electrodes, whether bare or covered. In the examination of cross-sections of the blocks of arc-fused metal, it was noticed that the equi-axed type of structure is prevalent throughout the interior of the piece and the columnar is to be found generally nearer the surface i.e., in the metal deposited last. It may be inferred from this that the metal of the layers which were deposited during the' early part of the preparation of the specimen is refined considerably by the successive heatings to which it is subjected as additional layers of metal are deposited. The general type of structure of the tension bars cut from the blocks of arc fused metal will vary considerably according to the amount of refining which has taken place as well as METALLOGRAPHY OF ARC-FUSED STEEL 193 the relative position of the tension specimen within the block. In addition it was noticed that the columnar and coarse equi- axed crystalline condition appears to predominate with fusion at high-current densities. FIG. 157. (1) "A" Electrode, 5 / 32 -in. Diameter. Annealed As Received. (2) "B" Electrode, 3 /i 6 -in. Diameter. Cold-Drawn. Picric Acid Etching. 3 FIG. 15 7 A. (3) Columnar Structure of B 2 . X 66 - Five Per Cent Picric Acid Etching. (4) Equi-axed Structure of AD 3 . X 200 - Two P er Cent Alcoholic HN0 3 Etching. Microscopic Evidence of Unsoundness. In all of the speci- mens of arc-fused metal examined microscopically there ap- pear to be numerous tiny globules of oxide as shown in Figs. 158 to 160. A magnification of 500 diameters is usually neces- sary to show these inclusions. In general they appear to have 194 ELECTRIC WELDING Is !2 'o <& f t^ W -2 HH M x-v e d O) t 02 ct METALLOGRAPHY OF ARC-FUSED STEEL 195 no definite arrangement, but occur indiscriminately through- out the crystals of iron. A type of unsoundness frequently found is that shown in (5), (6) and (7), Fig. 158; this will be referred to as "metallic- globule inclusions." In general these globules possess a microstructure similar to that of the surrounding metal, but are enveloped by a film, presumably of oxide. It seems prob- able that they are small metallic particles which were formed as a sort of spray at the tip of the electrode and which were deposited on the solidified crust surrounding the pool of molten metal directly under the arc. These solidified particles ap- parently are not fused in with the metal which is subsequently deposited over them i.e., during the formation of this same layer and before any brushing of the surface occurs. By taking extreme precautions during the fusion, a great deal of this unsoundness may be avoided and the mechanical properties of the metal may be considerably improved. Characteristic "Needles" or "Plates." The most char- acteristic feature of the steel after fusion is the presence of numerous lines or needles within the crysals. The general appearance of this feature of the structure is shown in (8) to (11), Fig. 159, inclusive. The number and the distribution of these needles were found to vary greatly in the different specimens. In general, they are most abundant in the columnar and in the coarse equi-axed crystals ; the finer equi-axed crystals in some specimens were found to be quite free from them, although exceptions were found to this rule. In general, a needle lies entirely within the bounds of an individual crystal. Some instances were found, however, where a needle appeared to lie across the boundary and so lie within two adjacent crystals. Several instances of this tendency have been noted in the literature on this subject. The needles have an ap- preciable width, and when the specimen is etched with 2 per cent alcoholic nitric acid they appear much the same as cementite i.e., they remain uncolored, although they may appear to widen and. darken if the etching is prolonged con- siderably. The apparent widening is evidently due to the attack of the adjacent ferrite along the boundary line between the two. The tendency of the lines to darken when etched with a hot alkaline solution of sodium picrate, as reported 196 ELECTRIC WELD1NU FIG. 159 (8 to 11). Characteristic "Needles" or "Plates" X 375. (8) BDg etched with 5 per cent picric acid in alcohol. (9) Specimen BD 8 after using for thermal analysis, re-heated in vacuo to 900 deg. C. four times. Picric acid etching. (10) Same as (9) except etched in hot alkaline sodium picrate solution. (11) Specimen of welded joint between slip-plate. Additional very small needles are noted. Etching: 2 per cent HNOs in alcohol. METALLOGRAPHY OF ARC-FUSED STEEL 197 by Comstock, was confirmed; (10) illustrates the appearance when etched in this manner. The needles are sometimes found in a rectangular grouping i.e., they form angles of 90 deg. with one another. In other cases they appear to be arranged along the octahedral planes of the crystal i.e., at 60 deg. to one another. This is best seen in specimens which have been heated, as explained below : In some of the specimens certain crystals showed groups of very fine short needles as in (11). The needles comprising any one group or family are usually arranged parallel to one another, but the various groups are often arranged definitely with respect to one another in the same manner as described above. Similar needles have been reported in articles by S. W. Miller. An attempt was made by Dr. P. D. Merica to determine whether the so-called lines or needles were really of the shape of needles or of tiny plates or scales. An area was carefully located on a specimen prepared for microscopic examination, which was then ground down slightly and repolished several times. It was possible to measure the amount of metal removed during the slight grinding by observing the gradual disap- pearance of certain of the spherical oxide inclusions the diameter of which could be accurately measured. By slightly etching the specimen after polishing anew it was possible to follow the gradual disappearance of some of the most prominent needles and to measure the maximum " depth" of such needles. It was concluded from the series of examinations that the term "plate" is more correctly descriptive of this feature of the structure than "line" or "needle." The thickness of the plate i.e., the width of the needle varies from 0.0005 to 0.001 mm. and the width of the plate ("depth") may be as great as 0.005 mm. The persistence of the plates after a regrinding of the surface used for microscopical examination may be noted in some of the micrographs given by Miller. The authors are not aware, however, of any other attempt to determine the shape of these plates by actual measurements of their dimensions. Plates Probably Due to Nitrates. The usual explanation of the nature of these plates is that they are due to the nitrogen which is taken up by the iron during its fusion. Other sug- 198 ELECTRIC WELDING gestions which have been offered previously attribute them to oxide of iron and to carbide. The suggestion concerning oxide may be dismissed with a few words. The plates are distinctly different from oxide in their form and their behavior upon heating. It is shown later that the tiny oxide globules coalesce into larger ones upon prolonged heating in vacuo; the plates also increase in size and become much more distinct (see (32), (34) and (36), Fig. 166). In no case, however, was any inter- mediate stage between the globular form and the plate pro- FiG. 160. (12) Specimen AD 3 , Etched with 2 Per Cent Alcooholic Nitric Acid. Shows Pearlite Islands, " Needles" and Oxide Inclusions. X 750. duced such as would be expected if both were of the same chemical nature. Regarding the assumption that they are cementite plates, it may be said that the tendency during fusion is for the carbon to be "burned out," thus leaving an iron of low carbon content. In all the specimens, islands of pearlite (usually with cementite borders) are to be found and may easily be distinguished with certainty. The number of such islands in any specimen appears to be sufficient to account for the carbon content of the material as revealed by chemical analysis. In some cases the peariite islands are associated with a certain type of " lines " METALLOGRAPHY OF ARC-FUSED STEEL 199 or "needles" such as are shown in (12), Fig. 160. These needles, however, appear distinctly different from those of the prevailing type and are usually easily distinguished from them. The fact that the plates found in the arc-fused metal are identical in appearance and in behavior (e.g., etching) as those found in iron which has been nitrogenized is strong evidence that both are of the same nature. (13) Fig. 161 shows the appearance of the plates produced in electrolytic iron by heating it for some time in pure ammonia gas. These plates behave in the same characteristic manner when etched with hot sodium picrate as do those occurring in arc-fused 13. mr 14 FIG. 161. (13) Characteristic Structure of Electrolytic Iron Heated in KH 3 at 650 Deg. C. Two Types of Nitride Plates. Etched with 2 Per Cent Alcoholic HNO 3 . X 375 - FIG. 161. (14) Arc-Fused Iron Produced in CO 2 Atmosphere. Type "A," Ysa-in. Electrodes, 150 Amperes. Etched with 5 Per Cent Picric Acid in Alcohol. X 375 - iron i.e., they darken slightly and appear as finest rulings across the bright ferrite. The fact that the nitrogen content of the steel as shown by chemical analysis is increased by the arc-fusion also supports the view that the change which occurs in the structure is due to the nitrogen. The statement has been made by Ruder that metal fused in the absence of nitrogen i.e., in an atmosphere of carbon dioxide or of hydrogen does not contain any plates and hence the view that the plates are due to the nitrogen is very much strengthened. In (15), Fig. 162, the appearance of specimens prepared at the Bureau by arc fusion of electrodes of type "A" in an atmosphere of 200 ELECTRIC WELDING METALLOGRAPHY OF ARC-FUSED STEEL 201 carbon dioxide is shown. The microscopic examination of the fused metal shows unmistakable evidence of the presence of some plates, although they differ somewhat from those found in nitrogenized iron and in metal fused in the air by the electric arc. Evidently they are due to a different cause from the majority of those formed in the iron fused in air. For convenience, in the remainder of the discussion the "plates" will be referred to as "nitride plates." Relation of Microstructure to the Path of Rupture. The faces of the fracture of several of the tension specimens after testing were heavily plated electrolytically with copper so as to preserve the edges of the specimens during the polishing of the section and examined microscopically to see if the course of the path of rupture had been influenced to an appreciable extent by the microstructural features. In general, the frac- ture appears to be intcrerystallme in type. Along the path of rupture in all of the specimens were smooth-edged hollows, many of which had evidently been occupied by the "metallic globules" referred to above, while others were gas-holes or pores. Portions of the fracture were intracrystalline and presented a jagged outline, but it cannot be stated with cer- tainty whether the needles have influenced the break at such points or not. (16) shows the appearance of some of the fractures and illustrates that, in general, the "nitride plates" do not appear to determine to any appreciable extent the course of the path of rupture. The behavior of the plates under deformation can best be seen in thin specimens of the metal which were bent through a considerable angle. Results of examination of welds treated in this manner have been described by Miller. Small rec- tangular plates of the arc-fused metal, approximately 3 / 32 in. thick, were polished and etched for microscopic examination and were then bent in the vise through an angle of 20 deg. (approximate). In (18) to (21), Fig. 163, inclusive are given micrographs illustrating the characteristic behavior of the material when subjected to bending. For moderate distortion the nitride plates influence the course of the slip-bands in much the same way that grain boundaries do i.e., the slip-bands terminate usually on meeting one of the plates with a change of direction 202 ELECTRIC WELDING so that they form a sharper angle with the plate than does the portion of the slip-band which is at some distance away (18). When the deformation is greater the slip-bands occur on both sides of the nitride plate, but usually show a slight variation in direction on the two sides of the nitride plate (19) ; this is often quite pronounced at the point where the plate is crossed by the slip-band. In a few cases evidence a 19 2O FIG. 163. (18 to 21) Behavior of " Nitride Plates" During Plastic De- formation of the Iron. Specimen RD 2 , Etched with 2 Per Cent Alcoholic, Nitric Acid Before Bending. X 500. of the "faulting" of the plate as a result of severe distortion was noted (20). This was a rare appearance, however, because of the nature of the metal, and is not shown in (21).. On account of the inclusions and other features of unsoundness of the metal, rupture occurs at such points before the sound crystals have been sufficiently strained to shoAv the character- istic behavior of the plates. Other micrographs show the beginning of a fracture around one of the "metallic globule" METALLOGRAPHY OF ARC-FUSED STEEL 203 inclusions before the surrounding metal lias been very severely strained. For this reason the influence of the plates on the mechanical properties of the crystals cannot be stated with certainty. It would appear, however, that on account of the apparently unavoidable unsoundness of the metal, any possible influence of the nitride plates upon the mechanical properties of the material is quite negligible. Some of the same specimens used for cold bending were torn partially in two after localizing the tear by means of a saw cut in the edge of the plate. The specimen was then copper plated and prepared for microscopic examination, the surface having been ground away sufficiently to reveal the weld-metal with the tear in it. The nitride plates did not appear to have determined to any extent the path taken in the rupture produced in this manner. Effect of Heat Treatment Upon Structure. With the view of possibly gaming further information as to the nature of the plates (assumed to be nitride), which constitute such a characteristic feature of the microstructure, a series of heat treatments were carried out upon several specimens of arc-fused electrodes of both types. Briefly stated, the treatment consisted in quenching the specimens in cold water after heating them for a period of ten or fifteen minutes at a temperature considerable above that of the Ac 3 transforma- tion; 925, 950 and 1,000 dcg. C. were the temperatures used. After microscopical examination of the different quenched specimens they were tempered at different temperatures which varied from 600 to 925 deg. C. for periods of ten and twenty minutes. The samples which were used were rather small in size, being only -J in. thick, in order that the effect of the treatment should be very thorough, were taken from test bars A 2 , A 3 , AD 10 , B 2 , B 6 and B 9 . These represented metal which had been deposited under different conditions of current den- sity, as shown in Table X. No plates were found to be present in any of the specimens after quenching. (22) Fig. 164 shows the appearance of one of the quenched bars, a condition which is typical of all. The structure indicates that the material comprising the plates had dissolved in the matrix of iron and had been retained in this condition upon quenching. The needle-like striations within the individual grains are char- 204 ELECTRIC WELDING METALLOGRAPHY OF ARC-FUSED STEEL 205 acteristic of the condition resulting from the severe quenching and are to be observed at times in steel of a very low carbon content. (23) shows the appearance of one of the "A" elec- trodes (V 32 i n -) quenched in cold water from 1,000 deg. C. Some of the crystals of the quenched iron also show interior markings somewhat similar in appearance to the nitride plates (24). These are, however, probably of the same nature as the interior tree-like network sometimes seen in ferrite whicli has been heated to a high temperature. The striations were found to be most pronounced in the specimens of arc-fused metal which were quenched from the highest temperatures, as might be expected. Braune states that nitride of iron in quenched metal is retained in solution in the martensite. The same may be inferred from the statement by Giesen that "in hardened steel, it (nitrogen) occurs in martensite." Ruder has also shown that nitrogenized electrolytic iron (3 hr. at 700 deg. C. in ammonia) after being quenched in water from tem- peratures 600 to 950 deg. C. shows none of the plates which were present before the specimen was heated. The sets of specimens (A 2 , A 6 , AD 10 , B 2 , B 6 and B ) quenched from above the temperature of the Ac 3 transforma- tion were heated to various temperatures, 600, 700, 800 and 925 deg. C. In all cases the specimens were maintained at the maximum temperature for approximately ten to fifteen minutes and then cooled in the furnace. (25) to (30), Fig. 165, inclusive summarize the resulting effects upon the structure. Heating to 650 deg. C. is not sufficient to allow the plates to redevelop, but in the specimens heated to 700 deg. C. a few small ones were found. The effect is progressively more pro- nounced with the increased temperature of tempering, and in the material heated to 925 deg. C. they are as large and as numerous as in any of the arc-fused specimens. The heating also develops the islands of pearlite which are not always to be distinguished very clearly in the simple fused metal. The work of Ruder shows that nitrogenized iron which has been quenched and so rendered free from the nitride plates behaves in a similar manner upon heating to temperatures varying from 700 to 950 deg. C. ; the plates reappear after a heating for fifteen minutes at 700 deg. C. (or above), followed by a slow cooling. The similarity in behavior of the two is a 206 ELECTRIC WELD1JMG FlG. 165. (25 to 30) Effect of Heat-Treatment of Arc-Fused Iron. All etched with 2 per cent alcoholic HNOs. X 450. (25) Specimen ADio as deposited. (26) Same after quenching from above HCa and reheating to 650 deg. C. No "plates" have formed. (27) Specimen ADio after quenching from above HCs and reheating to 700 deg. C. "Plates beginning to reform. (28) Specimen 69 after quenching from above ACs and reheating to 800 dep. C. (29) Specimen B2 after quenching from above ACa and reheating to 925 dep. C. (30) Specimen Au after quenching from above ACa and reheating to 925 deg. C. METALLOGRAPHY OF ARC-FUSED STEEL 207 FIG. 1G6. (31 to 36) Effect of 6-hr. Heating at 1000 Deg. C. in Vacuo. All etched with 2 per cent alcoholic HNO 3 . X 450. (31) Initial structure of AD2. (32) ADs after heating. (33) Initial structure of B*. (34) B 4 after heating. (35) Initial structure of Aio. (36) Aio after heating. 208 ELECTRIC WELDING further line of evidence that the arc-fused metal contains more or less nitrogenizcd iron throughout its mass. Plates Remain After Long Annealing. The persistence of the nitride plates was also studied in specimens heated at 1,000 deg. C. in vacuo for a period of 6 hr. A set of specimens (one each of test-bars AD 2 , A 3 , AD 6 , A 10 , B 2 , B 4 , B., and BD 5 ) was packed in a Usalite crucible, and covered with alundum "sand" ; this crucible was surrounded by a protecting alundum tube and the whole heated in an Arsem furnace. A vacuum, PIG. 167. (37) Effect of Pronounced Heating Upon the Structure of Arc-Fused Iron. Specimen ADio was heated for 6 hr. in vacuo at 1000 deg. C. The micrograph represents a section of the specimen at one corner. The oxide and "nitride plates" have been removed in the exposed tip of the thread. Etching, 2 per cent alcoholic solution of nitric acid. X 150. equivalent to 0.2 mm. mercury, was maintained for the greater part of the 6-hr, heating period; for the remainder of the time the vacuum was equivalent to 0.1 to 0.2 mm. mercury. The specimens were allowed to cool in the furnace. Ruder has stated that 1 hr., heating in vacuo at 1,000 deg. C. was sufficient to cause a marked diminution in the number of plates in both arc-wefd material and nitrogenized iron and that at 1,200 deg. C. they disappeared entirely. The results obtained are shown in (31) to (36), Fig. 166, METALLOGRAPHY OF ARC-FUSED STEEL 209 inclusive. In contradistinction to Ruder 's work the plates arc more conspicuous and larger than before, the oxide specks are larger and fewer in number. Many of the " plates" appear to have been influenced in their position by an oxide globule. It would appear that the conditions of the experiment are favorable for a migration of the oxide through an appreciable distance and for a coalescing into larger masses. (32), (34) and (36) all show some cementite at the grain boundaries which resulted from the "divorcing" of pearlite. The oxide is eliminated entirely in a surface layer averaging approx- imately 0.15 mm. in depth. Only in projections (right-angled corners, sections of threads of the tension bar, etc.), was there any removal of the nitride plates by the action of the continued heating in vacuo. This is shown in (37), Fig. 167, which illus- trates the removal of the oxide inclusions also. No evidence was found that the small amount of carbon present in the arc-fused metal is eliminated, particularly beneath the surface. (6) Fig. 158 illustrates an interesting exception to the rule that the nitride plates are flat. In the metallic and globular inclusion shown the plates have a very pronounced curve. The general appearance suggests that the " metallic globules" solid- ified under a condition of "constraint" and that this condi- tion still persists even after the 6-hr, heating at 1,000 deg. C. which the specimen received. Several of the specimens which were heated in vacua (6 hr. at 1,000 deg. C.) were analyzed for nitrogen. The results are given in Table XVI. TABLE XVI. CHANGE IN NITROGEN CONTENT UPON HEATING Average Nitrogen Content, Wt. of per Cent Sample Before After Heating Loss Specimen in Gr. Heating in Vacuo. per Cent A 3 1.39 0.127 0.062 51 B 4 60 124 078 37 BD 5 1.62 0.140 0059 57 B 5 . 1 16 121 054 55 The fact that the specimens lose nitrogen upon heating (although the amount remaining is still many times the nitrogen-content of the metal before fusion), coupled with the fact that the " nitride plates" are larger and more con- 210 ELECTRIC WELDING spicuous after heating than before, suggests very strongly that these plates are not simple nitride of iron. The method used for the determination of nitrogen gives only the "nitride" nitrogen, hence a possible explanation for the change in nitrogen content is that it has been converted into another form than nitride and may not have been eliminated from the specimen. Thermal Analysis of Arc-Fused Steel. In order to throw further light on the nature of the plates j( nitride) found in the metal after fusion in the arc, the thermal characteristics of the electrode material before and after fusion as revealed by heating and cooling curves were determined. Samples of a 3 / 16 -in. electrode of type "A" and of the specimen A, which resulted from the fusion were used as material (composition in Tables IX and XII.) TABLE XVII. THE THERMAL CHARACTERISTICS OF ARC-FUSED IRON H C 1 1 I I So Ao2, Maximum Deg. C. Beginning Maximum, Deg. C. " 1 Maximum Temp., Deg. C. Time Above As,- Min. Beginning 1 Maximum. Deg. C. ' Co 1 1 Maximum, Deg. C. ^ Unfused Electrode 15* 768 892 910 918 960 896 893 879 766 765 897 911 916 960 895 891 879 766 Arc-Fused Metal t 0. 14 764 .... 847 874 960 28 847 838 820 764 13 764 849 876 985 42 847 836 822 764 13 764 . . 844 870 960 29 847 837 821 765 13 766 850 874 1.035 256 848 835 816 764 * Heated at rate of 16 dcg. C. per sec., cooled 0.15 deg. C. per sec. for other specimens, the rate of cooling equaled the rate of heating. t The same specimen was heated four times in succession, as shown. (Fig 38) In Fig. 168 are given the curves obtained which show the characteristic behavior of the arc-fused metal upon heating. The commonly used inverse-rate method was employed in plot- ting the data ; the details of manipulation and the precautions necessary for the thermal analysis have already been described. In Table XVII are summarized the data shown graphically in the last cut. The principal change to be noted which has resulted from METALLOGRAPHY OF ARC-FUSED STEEL 211 O J I tS M/ I - s- EH * o J> ?l ^ E m S ? 212 ELECTRIC WELDING the arc-fusion of the iron is in the A 3 transformation. This is now very similar to the corresponding change observed in a very mild steel (e.g., approximately 0.15 per cent carbon). That the difference in the A 3 transformation of the arc-fused metal as compared with that of the original electrode is not due to an increase in the carbon content is evident from the lack of the sharp inflection of the A a transformation ("pear lite point ") which would, of necessity, be found in a low carbon steel. No evidence of the A x change was observed for the arc-fused iron within the range of temperature, 150 to 950 deg C. The change in the character of the A 3 transformation is without doubt to be attributed to the influence of the increased nitrogen-content of the iron. The specimen was maintained above the temperature of the A 3 transformation for a total period (four heatings) of 6 hr., the maximum temperature being 1,035 deg. C. The transformation apparently is unaffected by the long-continued heating, thus confirming the results described in the preceding section. In discussing the properties of steel nitrogenized by melting it in nitrogen under pressure, Andrews states that it was found possible to extract almost entirely the small quantities of nitrogen by heating a specimen at 1,000 deg. C. in vacuo for periods of 1 to 6 hr. The metal used contained 0.16 per cent carbon and 0.3 per cent nitrogen. Thermal curves are given to show that there are no critical transformations in the material; the nitrogen suppresses them. They gradually reappear, however, as the nitrogen is removed by heating the material in vacuo at 1,000 deg. C. Several days' heating was required, however, to obtain an entirely degasified product, the carbon being removed also. A further statement is made that a steel of 0.6 per cent carbon content containing 0.25 per cent nitrogen can be brought back to the normal state of a pure steel only by several weeks' heating in vacuo. The results of the thermal analysis add considerable con- firmatory evidence to support the view that the plates existing in the arc-fused metal are due to the nitrogen rather than to carbon. Summary. Microscopic examination of bent pieces of arc- fused metal show that the metallic grains are inherently ductile, METALLOGRAPHY OF ARC-FUSED STEEL 213 even to a high degree. Grosser imperfections, however, are entirely sufficient to mask this excellence. The view that the characteristic features observed in the structure of the arc-fused iron are due to the increased nitrogen content is supported by several different lines of evidence. These include the likeness of the structure of the material to that of pure iron which has been "nitrogenized," the similarity in the behavior of both arc-fused and nitrogenized iron upon heating, the evidence shown by thermal analysis of the arc-fused metal, together with the fact that, as shown by chemical analysis, the nitrogen content , increases during fusion, while the other elements, aside from oxygen, decrease in amount. The characteristic form in which oxide occurs in iron, together with its behavior upon heating, renders the assumption that the oxide is responsible for the plates observed in the material a very improbable one. Judged from the results obtained, neither type of electrode appears to have a marked advantage over the other. The use of a slight protective coating on the electrodes does not appear to affect the mechanical properties of the arc-fused metal materially in any way. The specimens were prepared in a manner quite different from that used ordinarily in electric-arc welding and the results do not justify any specific recom- mendations concerning methods of practice in welding. CHAPTER XI AUTOMATIC ARC WELDING The automatic arc welding machine, made by the General Electric Co., Schenectady, N. Y., is a device for automatically feeding metallic electrode wire into the welding arc at the rate required to hold a constant arc length, says H. L. Unland in a paper read before the American Welding Society. Under these circumstances the electrical conditions are kept constant and the resulting weld is uniform and its quality is thereby improved. It is possible with this device to weld at a speed of from two to six times the rate attained by skilled operators welding by hand. This is partly due to the stability of the welding conditions and partly due to the fact that the elec- trode is fed from a continuous reel, thus eliminating the chang- ing of electrodes. The automatic welding machine is adaptable to practically any form of weld from butt welding of plates to the depositing of metal on worn surfaces such as shafts, wheels, etc. Everyone who has made any investigation of electric arc welding has noted the wide variation in results obtained by different welders operating, as nearly as can be determined, under identical conditions. This also applies to the operations of a single welder at different times under identical conditions. These variations affect practically all factors of welding such as speed of welding, amount of electrode consumed, etc. When indicating instruments are connected to an electric welding circuit, continual variations of considerable magnitude in the current and voltage of the arc are at once noticed. Consider- able variation was found some years ago in the cutting of steel plates by the gas process and when an equipment was devised to mechanically travel the cutting torch over the plate a series of tests to determine the maximum economical speed, gas pressure, etc., for the various thickness of plate were made. 214 AUTOMATIC ARC WELDING 215 The result was that the speed of cutting was increased to as much as four or five times the rate possible when operating under the unsteady conditions incident to hand manipulation of the torch. Further, the gas consumption for a given cut was found to be decreased very greatly. As a result of many experiences an investigation was started to determine what could be done in controlling the feed of the electrode to the electric arc in a metallic electrode welding circuit. An electric arc is inherently unstable, the fluctuations taking place with extreme rapidity. In any regulating device the sensitiveness depends on the percentage of variation from normal rather than on the actual magnitude of the values, since these are always reduced to approximately a common factor by the use of shunts, current transformers, or series resist- ances. The characteristics of practically all electric welding circuits are such that the current and voltage are inter-related, an increase in one causing a corresponding decrease in the other. Where this is the case it will generally be found that the percentage variation of the voltage from normal when taken at the customary arc voltage of 20, will be approximately twice the percentage variation in current. Further, an increase in arc voltage, other conditions remaining the same, indicates that the arc has been lengthened, thus giving the metal a greater opportunity to oxidize in the arc with a probability of reduction in quality of the weld. The automatic arc weld- ing machine utilizes the arc voltage as the basis for regulating the equipment. The rate of feeding the wire varies over a wide range, due to the use of electrodes of different diameters, the use of different current values, etc., caused by details of the particular weld to be made. The simplest and most reliable method of electrically obtaining variations in speed is by means of a separately excited direct current motor. Thus the operation of this equipment is limited to direct current arc welding circuits, but these may be of any established type, the variations in characteristics of the welding circuits being taken care of by proper selection of resistors, coils, etc., in the control. The Welding Head. The welding head consists essentially of a set of rollers for gripping the wire and feeding it to the arc. These rollers are suitably connected through gearing 216 ELECTRIC WELDING to a small direct-current motor, the armature of which is con- nected across the terminals of the welding arc. This connec- tion causes the motor to increase in speed as the voltage across the arc increases due to an increase in the length of the arc and to decrease in speed as the voltage decreases, due to a shortened arc. A small relay operating on the principle of a generator voltage regulator is connected in the field circuit of the motor which assists in the speed control of the motor as the arc voltage varies. Rheostats, for regulating and adjust- ing the are voltage, are provided by means of which the equipment can be made to maintain steadily an arc of the desired length and this value may be varied from over twenty to as low as nine volts. No provision is made in the machine for adjustment of the welding current since the automatic operation is in no way dependent on it. The welding current adjustment is taken care of by the control panel of the welding set. This may be either of the variable voltage or constant potential type but it is necessary to have a source of constant potential to excite the fields of feed motor. It may be possible to obtain this excitation from the welding circuit, but this is not essential. The voltage of both the welding and constant potential circuits is immaterial, provided it is not too high, but these voltages must be known before the proper rheostats can be supplied. On account of the great variation in conditions under which this welding equipment may be used it is provided with a base which may be bolted to any form of support. It may be held stationary and the work traveled past the arc or welding head may be movable and the work held stationary. These points will be dictated by the relative size of the work and the head and the equipment which may be available. Provision must be made for traveling one or the other at a uniform speed in order to carry the arc along the weld. In the case of straight seams a lathe or planer bed may be utilized for this purpose and for circular seams a lathe or boring mill may be used. In many cases it will be found desirable to use clamping jigs for securely holding the work in shape and also to facilitate placing in position and removing from the feeding mechanism. In Fig. 169, the welding head is shown mounted on a special AUTOMATIC ARC WELDING 217 device for making circular welds. The work table is driven through a worm and worm gear by means of a separate motor. FIG. 169. Special Set-Up of Machine for Circular Welding. The welding head may be led along the arm by means of the handwheel, and it may be tilted at an angle of 45 deg. 218 ELECTRIC WELDING both at right angles to the line of weld and also parallel to the line of weld. Fig. 170 shows the building up of a shaft, the work being mounted on lathe centers and the welding head placed on a bracket clamped to saddle. Fig. 171 shows a simplified diagram of the control of the feed motor. In this cut A is the regulating rheostat in the motor field circuit controlled by the arc voltage regulator G; B is the adjusting rheostat in the motor field circuit j F FIG. 170. Set-Up for Building up a Shaft. indicates the feed motor field winding; M the feed motor wind- ing; D is the resistance in the motor armature circuit to adjust the speed when starting the feed motor before the arc is struck. The open-circuit voltage of the welding circuit is ordinarily considerably higher than the arc voltage. This resistance D is short circuited by contactor X when the arc is struck. The arc voltage regulator G maintains constant arc voltage by varying the motor field strength through resistor A. The regulator is adjusted to hold the desired voltage by the rheostat AUTOMATIC ARC WELDING 219 C. Permanent resistance E is in series with the over-voltage relay H, to compensate for the voltage of the welding circuit. Over voltage relay H holds open the coil circuit of the regulator G until the electrode makes contact in order to protect the coil from burning out. Observation of indicating meters on the control panel show that the current and voltage are practically constant, but it should be remembered that all indicating meters have a certain amount of damping which prevents observation of the varia- tions which are extremely rapid or of small magnitude. The resultant value as read on the instrument is the average value. Oscillographs taken with short arcs show that notwithstanding the fact that the indicating meters show a constant value, a Ammeter FIG. 171. Simplified Diagram of Control of Feed Motor. succession of rapid short circuits is continually taking place, apparently due to particles of the molten wire practically short- circuiting the arc in passing from the electrode to the work. This is indicated by the fact that the voltage curve fell to zero each time, and accompanying each such fluctuation there was an increase in the current. It was found that with the shorter arc the frequency of occurrence of these short-circuits was considerably higher than was the case when the arc was increased in length. To all appearances the arc was absolutely steady and continuous and there was no indication either by observation of the arc itself or of the instruments that these phenomena were occurring. Some Work Performed By the Machine, The principal field for an automatic arc welding machine is where a consider- 220 ELECTRIC WELDING able amount of welding is required, the operations being a continuous repetition of duplicate welds. Under these condi- tions one can economically provide jigs and fixtures for facilitating the handling of the work and the clamping. Thus can be reaped the benefit of the increased speed in the actual welding which would be lost if each individual piece had to be clamped and handled separately. Examples of different jobs done with this machine, using various feeding and holding methods, are shown in the accom- panying cuts. Fig. 172 is a worn pulley seat on an electric motor shaft built up and ready to be re-turned to size. It is possible to build up pulley and pinion seats, also worn bearings, without removing the armature or rotor from the m Fie. 172. Worn Motor Shaft Built Up. shaft and in practically all cases without removing the wind- ings due to the concentration of the heat at the point of the weld. On shafts of this kind, 3 to 4 in. in diameter, the figures are: current 115 amp.; arc voltage 14; electrode 3 / 32 in. in diameter; travel, 6 in. per min. ; rate of deposit about 2.1 Ib. per hour. Similar work on a 14-in. shaft where the flywheel seat 21 in. long was turned undersize, was as follows: metal about Vie i n - deep was deposited over the undersize surface, using current, 190 amp.; arc voltage 18; electrode -J in. diameter; travel 4 in. per min.; rate of deposit, about 2 Ib. per hour; welding time, 16 hr. ; machining time, 4 hr. Fig. 173 shows worn and repaired crane wheel flanges. These are easily handled by mounting on a mandrel in a lathe, AUTOMATIC ARC WELDING 221 and placing the welding machine on a bracket bolted to the cross-slide or the saddle. On wheels of this type 22 in. in diameter, the time taken to weld by hand would be about 12 hr. and by machine 2 hr. ; machining time 4 hr. ; approximate cost by hand welding $9; by machine $4. i FlG. 173. Worn and Repaired Crane Wheels. FIG. 174. Welded Automobile Hub Stampings. Fig. 174 is an automobile wire wheel hub stamping, to which a dust cover was welded as shown. Joint was between metal 1 / 16 and Vie in. thick. Current 100 amp.; arc voltage, 14 ; travel 10 in. per min. ; electrode 3 / 32 in. diameter. 222 ELECTRIC WELDING Fig. 175, welded automobile rear-axle housing, 3 / 16 i n - thick ; current 120 amp.; arc voltage 14; travel 6 in. per min. ; elec- trode diameter 3 / 82 in. Fig. 176, welded tank seam; metal -J in. thick; current 140 arnp. ; arc voltage 14 ; travel, 6 in. per min. ; time for welding ten tanks by hand, 4 hrs. 40 min. ; by machine, 2 hrs. FIG. 175. Welded Bear- Axle Housing. Tables XVIII and XIX give an idea of the speed of welding which may be expected, but it should be borne in mind that these figures are actual welding speeds. It is necessary to have the material properly clamped and supported and to have it travel past the arc at a uniform speed. In some cases the FiG. 176 Welded Straight Tank Seam. figures given have been exceeded and under certain special conditions it may be desirable to use lower values than those given. TABLE XVIII. SEAM WELDING Thickness in Inches 0.040 1/16 1/8 3/16 Amperes Speed, Inches Per Minute 45 to 50 20 to 30 50 to 80 15 to 25 80 to 120 6 to 12 100 to 150 4 to 6 AUTOMATIC ARC WELDING 223 TABLE XIX BUILDING UP (WHEELS OR SHAFTS) Diameter or Electrodes, Speed, In. per Lb. Deposit Thick., In. Dia., In. Amperes Min. Per Hour Up to 1" V- 60 to 90 11 to 13 1.04-1.56 Up to 3" % 90 to 120 6 to 8 1.59-2.1 Over 3" V- 120 to 200 4 to 6 2.5 -4.5 A SEMI-AUTOMATIC ARC-WELDING MACHINE A paper on " Welding Mild Steel," by H. W. Hobart, was read at the New York meeting of the American Institute of Mining and Metallurgical Engineers in 1919. In discussing this paper Harry D. Morton, of the Automatic Arc Welding Co., Detroit, brought out some interesting things relating to Automatic Arc Welding: "The generally accepted theory of the electric arc is that part of the electrode material is vaporized, and that this vaporous tube or column forms a path for the electric current. As a result of the vaporous character of the current path, all arcs are inherently unstable; and the maximum of instability is no doubt found in that form of arc employed for metallic-electrode welding purposes. We here have, in conjunction with the natural instability characteristic of all arcs rapidly fusing electrode materials and the disturbing effect of the constant passage through the arc of a large quantity of molten metal to form the weld. This molten metal must pass through the arc so rapidly that it will not be injured or materially contaminated; otherwise the weld will be useless. Prima facie, the combination of these unfavorable conditions would seem to justify fully the skepticism of most electrical engineers as to the possibility of affecting such control of the metallic arc as to permit of uniformity and continuity in welding results. In addition, there is another and more important factor, and one that seriously mitigates against this desired uniformity and continuity; namely, the personal equation of the operator. The consensus of opinion, so far as is known to the writer, seems to be that about 95 per cent, of the welding result is dependent on the skill of the operator and that at least six months' practice is necessary to acquire reasonably satisfactory proficiency. "As the result of thousands of observations of welds produced auto- matically (wherein the personal equation is entirely eliminated), the writer inclines toward the theory that the molten electrode material passes through the arc in the form of globules; and that where |-in. electrode material is employed with a current of about 150 amp. these globules are deposited at the rate of approximately two per second. The passage through the arc of each globule apparently constitutes a specific cause of instability in addition to those existent with slowly consumed electrodes. This hypothesis seems to be borne out by ammeter records, typical specimens of which appear in Fig. 177, together with the fact that the electrode 224 ELECTRIC WELDING fuses at the rate of about 0.20 in. per see. Moreover, the globules appear to be approximately equal in volume to a piece of wire 0.125 in. in diameter and 0.10 in. long. ' ' Assuming this theory to be correct, to maintain a uniform arc length in manual welding, the operator must feed the electrode toward the work FiG. 177. Typical Ammeter Charts of Operation of Morton Automatic, Metallic-Electrode Arc- Welding Machine. Average Time about 1 Min. 45 Sec. at the rate of 0.10 in. upon the deposition of each globule; in other words, 0.10 in. twice per second, a synchronism beyond human attainment. Simultaneously with such feeding, the arc must be moved over the work to melt the work material, distribute the molten electrode material, and form the weld. Inasmuch as the effect of the arc is highly localized. AUTOMATIC ARC WELDING 225 it is reasonable to suppose that different parts of the welding area present relatively wide variations in respect to temperature, fluidity, and conduc- tivity of the molten mass controlling factors not within the ken of the human mind. The situation is further complicated by the facts that neither the welding wire nor the work material is uniform in fusibility or in conductivity, and that the contour of the work varies continually as its surface is fused and the molten metal is caused to flow. The belief is general that a very short arc is productive of the best welding results; but it is an arc of this character that makes the greatest demands on the skill of the operator, for there is always the danger that the electrode will actually contact with the work and destroy the arc. ''As the fusing energy of the arc varies widely with fluctuations in the arc length and as the uniformity of the weld depends on the constancy and correctness of this fusing energy, it seems remarkable that operators are able ever to acquire such a degree of skill as to enable them to produce welds that are even commercially satisfactory. Further, so far as the writer is informed, there is no means, other than such as would be destructive, for determining whether a completed weld is good or bad. The logical solution appeared to be the elimination of the personal equation and the substitution therefor of means whereby tendencies toward variations in the arc would be caused automatically to correct themselves, just as a steam engine, through the action of its governor, is caused to control its own speed. Methods of Mechanically Stabilizing and Controlling the Arc. Our efforts for a number of years have been directed toward stabilizing and controlling the metallic arc, and applying such stabilizing and controlling means to two general lines of welding machinery: (1) Machines for automatically feeding the electrode wire, with reference to the work, and producing simultaneously therewith -relative movement between the wire and the work, and (2) what, for lack of a better term, might be called a semi-automatic machine, in which the feeding of the electrode and the control of the arc are accomplished automatically but the traversing of the electrode with reference to the work is manually effected by the operator, permitting him the exercise of judgment with reference to the quantity of metal to be deposited in various parts of the groove. The automatic machine has been in successful operation for a long period and the semi- automatic machine for about five months. While the goal was not attained without many difficulties and a great expenditure of time and money, the results have been surprisingly successful. ' ' Because of the lack of any definite data as to what actually occurs in this form of arc, or why it occurs, due, no doubt, to the impossibility of differentiating between phenomena that are characteristic of the arc and phenomena due to the personal equation of the welder, it seemed logical that the initial step should be to so environ the arc that it would not be subject to erratic extraneous influences, to the end that reasonably definite determinations might be substituted for scientific speculation. In the design and construction of the machines, great care was exercised to minimize the possibility of mechanical defects that might lead to 226 ELECTRIC WELDING erroneous conclusions. Starting with the assumption that the work could only be based on open-minded observation of the behavior of the arc under machine control, an automatic welding machine was built in which was incorporated the greatest possible number of adjustable features, in order that, if necessary, it might be possible to wander far afield in the investigations. This adjustability has proved invaluable in that it has permitted logical, consistent, and sequential experimenting over a very wide range of conditions. Working under these favorable circumstances, there were soon segregated a few clearly demonstrable facts to serve as a foundation for the structure, which has since been added to, brick by brick, as it were. 1 * Efforts have been directed toward the practical rather than the scientific aspect of the subject. The operation of the automatic machines has brought to light many curious and interesting phenomena, some of which appear to negative conclusions heretofore formed which have been predicated upon observations made in connection with manual welding. It is hoped that these and other phenomena, which can thus be identified as purely arc characteristics, will be the subject of profitable scientific investigation when time is available for this purpose. ' ' In the five forms of machines made in the course of the development, the welding wire is automatically fed to the arc; and, in the first four machines, the relative movement between the work and the welding wire is automatically and simultaneously effected. Early in his investigations, the writer concluded that a substantial equilibrium must be maintained between the fusing energy of the arc and the feeding rate of the welding strip; and it soon became evident that if the welding strip is mechanically fed forward at a uniform rate equal to the average rate of consumption with the selected arc energy, this equilibrium is actually maintained by the arc itself, which seems to have, within certain circumscribed limits, a compensatory action as follows: When the arc shortens, the resistance decreases and the current rises. This rise in current causes the welding strip to fuse more rapidly than it is fed, thereby causing the arc to lengthen. Conversely, when the arc lengthens, the resistance increases, the current falls, the welding strip is fused more slowly than it is fed, and the moving strip restores the arc to its normal length. 11 While this compensatory action of the arc will maintain the necessary equilibrium between the fusing energy and the feeding rate under very carefully adjusted conditions, this takes place only within relatively narrow limits. It was very apparent that, due to variations in the contour of the work, and, perhaps, to differences in the fusibility or conductivity of the welding strip or of the work, the range of this self -compensatory action of the arc was frequently insufficient to prevent either contacting of the welding strip with the work or a rupture of the arc due to its becoming too long. The problem that arose was to devise means whereby the natural self-compensatory action of the arc could be so greatly accentuated as to preclude, within wide limits, the occurrence of marked arc abnormalities. There was ultimately evolved, by experiment, such a relation between the fusing energy of the arc and the feeding rate of the welding strip as to AUTOMATIC ARC WELDING 227 give the desired arc length under normal conditions; and tendencies toward abnormalities in arc conditions, no matter how produced, were caused to FIG. 178. Piloted Cup Automatically Welded by Metallic-Electrode Arc Process to Tube to Form 75-MM. Shrapnel Shell. Analysis of Electrode Material: Silicon, 0.02 Per Cent; Sulphur, 0.013 Per Cent; Phosphorus, 0.07 Per Cent; Manganese, Trace; Carbon, 0.07 Per Cent; Aluminum, 0.038 Per Cent. FIG. 179. Piloted Cup Automatically Welded by Metallic-Electrode Arc Process to Tube to Form 75-MM. Shrapnel Shell. Analysis of Electrode Material: Silicon, 0.03 Per Cent; Sulphur, 0.049 Per Cent; Phosphorus, 0.008 Per Cent; Manganese, 0.31 Per Cent; Carbon, 0.28 Per Cent. bring into operation compensatory means for automatically, progressively, and correctively varying this relation between fusing energy and feeding 228 ELECTRIC WELDING rate, such compensatory means being under the control of a dominant characteristic of the arc. In their ultimate forms, the devices for effecting the control of the arc are simple and entirely positive in action, making discrepancies between fusing energy and feeding rate self -compensatory throughout widely varying welding conditions. For instance, the shrapnel shell shown in Fig. 178 was automatically welded with wire differing greatly in chemical constitution from that used on the shell shown in Fig. 179 (see analyses), yet no change was made in either the mechanical or the electrical adjustments. The radically different welding conditions were compensated for solely by the operation of the automatic control. The electrode materials used for the shells shown in Figs. 180 and 181 FIG. 180. Piloted Cup Automatically Welded by Metallic-Electrode Arc Process to Tube to Form 75-MM. Shrapnel Shell Analysis of Electrode Material: Silicon, 0.02 Per Cent; Sulphur, 0.032 Per Cent; Phosphorus, 0.008 Per Cent; Manganese, 0.20 Per Cent; Carbon, 0.18 Per Cent. differed so greatly from those employed respectively in welding the shells shown in Figs. 178 and 179 that a change in the relation between fusing energy and feeding rate had to be made manually. After this adjustment was made, the shells were welded with their respective electrodes, which varied widely in their chemical constitution, without further manually changing either the mechanical or the electrical conditions. "In a recent test of the semi-automatic machine, shown in Fig. 182, successful welds were made under the condition that the impressed voltage of the welding generator was changed throughout a range of from 50 to 65 volts, without necessitating any manual adjustment. The only observable effects of the wide variations in the supply voltage were slight differences in the arc length. In short, the compensatory action of the control has proved effective over a wide range of welding conditions, not only as to AUTOMATIC ARC WELDING 229 the electrical supply and chemical constitution of both electrode and work materials, but also as to extensive variations in the contour of the work and in many other particulars. This makes it seem apparent that the machines do not represent merely successful laboratory experiments but are suited to the requirements of actual commercial welding. "One particularly interesting observation resulting from the experiments is that the angle of inclination of the electrode with reference to the work is very important. An angular variation of 5 deg. will sometimes determine the difference between success and failure in a weld. About 15 deg. from the perpendicular works well in many cases. In welding some materials, the electrode should drag, that is, point toward the part already welded rather than toward the unwelded part of the seam. FlG. 181. Piloted Cup Automatically Welded by Metallic-Electrode Arc Process to Tube to Form 75-MAI. Shrapnel Shell. Analysis of Electrode Material: Silicon, 0.04 Per Cent; Sulphur, 0.016 Per Cent; Phosphorus, O.OfiS Per Cent- Manganese, None; Carbon, 0.24 Per Cent. "While it has been customary in some welding systems to provide means whereby extra resistance is inserted in series with the arc at the instant of the initial contact which starts the flow of current, the resistance being automatically cut out upon the striking of the arc, experience with the automatic machines indicates that this is quite unnecessary. "Early in the experiments, it was noted that in many cases there was a decidedly marked affinity between particular electrode materials and particular work materials. A slight change in either element affects the degree of this affinity. While it has invariably been possible to contiol and maintain the arc and weld continuously, in some instances incom- patibility between electrode material and work material has been productive of interesting phenomena. For instance, the combination of work material (steel of about 0.45 per cent, carbon content) and the particular electrode 230 ELECTRIC WELDING material used in Fig. 178 produced an arc that was remarkably quiet and free from sputtering. Throughout the weld, this arc was suggestive of the quiet flame of a candle or lamp, the erratic behavior that we are accustomed to associate with the ordinary metallic arc being absent. The effect is reflected in the uniform deposition of the welding material. "On some classes of work materia 1 Bessemer wire, which some authorities claim cannot be used in metallic-electrode arc welding, produces an arc FIG. 182. Morton Semi-Automatic Metallic-Electrode Arc-Welding Machine. The Electrode is automatically fed to the arc, which is automatically maintained while the machine is manually moved along the groove to be welded. and a weld very satisfactory in appearance. On other work material, the Bessemer wire arc is violently explosive. These explosions are accompanied by quite sharp reports and the scattering over some considerable distance of globules of molten metal frequently s / 82 in. or more in diameter. Under certain other conditions, apparently growing out of incompatibility between the work material and the electrode material, the oxygen flame accompany- ing the arc gyrates very rapidly about the arc, producing an effect sug- gestive of the ' whirling dervish. ' AUTOMATIC ARC WELDING 231 "From both the practical and the scientific points of view, the writer has experimented quite extensively with varying combinations of work material and electrode material. Throughout all the differences in arc conditions, many of which palpably accentuate the natural inclination toward instability, the control has so operated as to justify the expression 'the arc persists.' "Generally speaking, the Swedish and Norway iron wires seem to produce more quiet arcs and, possibly, a more uniform deposition of electrode material, than do wires of other classes. These welds may perhaps be found to be slightly more ductile than those made with wires of other chemical composition. On the other hand, these soft wires, although un- doubtedly of relatively high fusibility, do not, for some reason, seem to produce an arc that cuts into some work material as deeply as might be desired, nor as deeply as do the arcs formed with certain other kinds of wire. Considered from every angle, the writer is disposed to regard the Roebling welding wire as the best he has thus far tested for use on mild steel. The wire produces a reasonably quiet arc which seems to cut into the work to more than the ordinary depth, while, at the same time, the electrode material is fused with more than average rapidity thus increas- ing the welding rate. ' ' While scientists will no doubt ultimately arrive at the correct hypothesis for solving the problem of why one combination of electrode material and work material is productive of better results than can be obtained with another combination, the writer's conclusion is that, with the data at present available, the determinations must be made by actual experimenting having in mind the qualities desired in the particular weld, such as ductility, tensile strength, elongation, and elastic limit. Inasmuch as it is possible, with the automatic machine, to maintain arc uniformity with practically any kind of electrode material and to produce welds which, under low magnification, at least, appear to be perfect, and which respond favorably to ordinary tests such as bending, cutting and filing, it is reason- able to conclude that proper selection of electrode material will be productive of perfect welds on any kind of work material. To date, no steel has been tested on which apparently satisfactory welds could not be made. High- speed tungsten steel has been successfully welded to cold-rolled shafting, using Bessemer wire as electrode material, as is shown in Fig 183. Ordinary steels varying in carbon content from perhaps 0.10 to 0.55 per cent, have been welded with entire success. "Because of the fact that the complete welding operation has been automatic and may be continued for a considerable length of time, say 5 min., an exceptional opportunity has been afforded for close concentration upon the study of the appearance of the arc. What seems to occur is that the molten metal in the crater is in a state of violent surging, sug- gestive of a small lake lashed by a terrific storm. The waves are dashed against the sides of the crater, where the molten metal of which they are composed quickly solidifies. The surgings do not seem to synchronize with nor to be caused by the falling of the globules of molten metal into the crater, but seem rather to be continuous. They give the impression 232 ELECTRIC WELDING that the molten metal is subjected to an action arising from the disturbance of some powerful force associated with the arc such, for instance, as might result from the violent distortion of a strong magnetic field. Alto- gether, the crater phenomena are very impressive; and the writer hopes ere long to be able to have motion pictures made which, when enlarged, should not only afford material for most fascinating study, but also throw light upon some of the mysterious happenings in the arc. So far, electrode wires I in. in diameter have been chiefly used in the machines. Successful welds have been made with current values ranging from below 90 to above 200 amp., at impressed voltages of 40, 45, 50, FIG. 183. Tungsten High-Speed Eing Automatically Welded by Metallic- Electrode to Cold-Kolled Core to Form Milling-Cutter Blank. 55, 60, 65 and 80. Under these varying conditions, the voltage across the arc has been roughly from 16 to 22. The machines have thus far been run only on direct current. Inasmuch as it is possible, by electrical and mechanical adjustments, to establish nearly any arc length that may be found to be most desirable for a particular class of work, and as the control system will maintain substantially that arc length indefinitely, the fully automatic type of machine is nearly as certain in operation as a lathe, drilling machine, or any other machine tool. ' ' The tool shown in Fig. 182 weighs about 10 Ib. The operator draws the tool along the groove to be welded at such a rate as will result in the deposition of the quantity of metal required to satisfactorily effect the weld. This tool is intended for use in the many restricted spaces en- AUTOMATIC ARC WELDING 233 countered in ship welding, which would be relatively inaccessible to a fully automatic machine. In its use, the skill required by the operator is reduced to a minimum. After one man had practised with the welding tool for not more than 2 hr., the opinion was expressed that it would require six months to train a welder to such a degree of proficiency as to enable him to make a weld equally good in appearance. "Mr. Hobart, says 'There is always a matter of a 0.10 in. or more between the end of the welding rod and the work.' While undoubtedly it is difficult, if not impossible, to maintain in manual welding an arc shorter than this, the writer has frequently, with the automatic machines, made continuous and strikingly good welds with arcs of much less length. In fact, in some cases there has been continuously maintained an arc so short that there hardly seemed to be any actual separation. The writer FIG. 184. No. 11 Gage Steel Tubing Automatically Welded by Metallic- Electrode Arc Process at the Rate of One Foot per Minute. has even wondered whether, under these conditions, there was not a close approach to casting with a continuous stream of fluid metal acting as the current conveyor in lieu of or in parallel with the usually assumed vapor path. The work that has been done indicates that under automatic control much shorter arcs can be utilized than have hitherto been deemed possible, and with probable marked gain in quality of work in some instances; also, that there is much to be learned as to the mode of current action and current conduction in such an arc. "With the automatic machine, black drawing steel 0.109 in. thick has been welded at the rate of 22 in. per minute. A Detroit manufacturer welded manually with oxy-acetylene at the rate of four per hour a large number of mine floats 10 in. in diameter, made of this material. The automatic machine made the welds at the rate of forty per hour. Liberty 234 ELECTRIC WELDING motor valve cages 2| in. in diameter have been welded to cylinders in 36 sec., as against about 5 min. required for manual welding. No. 11 gage steel tubing, shown in Fig. 184, has been welded, with an unnecessarily FIG. 185. Two i-in. Ship Plates Automatically Welded by Metallic- Electrode Arc Process to Form Lap Joint. FlG. 186. Two -in. Ship Plates Automatically Welded by Metallic- Electrode Process to Form Butt .Joint. heavy deposit of metal, at the rate of 1 ft. per minute. The productive capacity of the machines so far made has been from three to ten times that of manual welding methods, depending on the thickness of the work AUTOMATIC ARC WELDING 235 material; the difference in favor of automatic welding varies inversely as such thickness. The writer is now designing an improved type of machine for use especially on heavy work, with which machine it is expected to be able automatically to lapweld -in. ship plates, in the manner shown in Fig. 185, at the rate of 15 ft. per hour. One of the largest shipbuilding concerns in the United States reports that the general average of all its manual welders on this class of work is from 1 ft. to 18 in. per hour. Other specimens of automatic welding on ship plates are shown in Figs. 186 and 187. * ' Bare wire only has been used in the automatic machines ; and the results obtained seem to indicate that the covering of the electrodes is an expensive superfluity. If the chief advantage of the covered electrode lies in the ability of the operator to maintain a very short arc, an arc equally short and possibly shorter can be continuously maintained by the automatic machine using bare electrodes. "No attempt has thus far been made to use the automatic machines FIG. 187. Two i-in. Ship Plates Automatically Welded by Metallic-Elec- trode Arc Process, Showing First of Three Layers to Form Lap Joint. on overhead work. The welds made with the fully automatic machine have been of three kinds, the usual longitudinal form, annular about a horizontal axis, and annular about a vertical axis. "As far as the maintenance of arc uniformity and the apparent character of the welds are concerned, the writer has repeatedly welded with wire showing evidence of pipes and seams, as well as with rusty wire and with wire covered with dirt and grease. In this connection it may be said that no pains is ever taken to remove rust, scale, or slag from the work material even where welds are superimposed. Apparently under uniform conditions of work traverse, arc length, and electrode angle of inclination, such as are possible in the automatic machine, impurities vanish before the portion of the work on which they occur reaches the welding area of the arc. "The writer is fully convinced that with the use of the automatic machine, ductility, like other physical properties in the weld, can be con- trolled "by proper selection of electrode wire, in conjunction with electrical 236 ELECTRIC WELDING AUTOMATIC ARC WELDING 237 and mechanical adjustments best suited to the particular purpose in view. Automatic welds have repeatedly been made on 5 /i 6 -in. mild steel which, when subjected to a 90-deg. bend, showed a marked extrusion of the welded material but no sign of fracture. When the welded pieces are cut with a hacksaw, it is very unusual to be able to note any difference in cutting qualities between the unwelded and the welded parts. "While the automatic machine has not been used on metal less than 0.109 in. thick, it is fair to presume that, with proper adjustments, entirely satisfactory results can be obtained on much thinner work particularly if the nature of the work is such as to permit of the use of a chill. The best method in welding very light metal seems to be to use a small electrode, a relatively low current, and a high rate of work traverse. In this way welding conditions may be controlled to almost any desired extent, because FIG. 189. How the Metal Edges Are Welded. the heating action of the arc can be modified, its effect intensely localized, and the edges to be welded subjected to the fusing action for as brief a time as might be found necessary to prevent burning of the metal. These conditions, which seem to be requisite in order to successfully weld very thin material, cannot be met by the manual welder. It is here that the deficiencies incident to the personal equation become most apparent. A very slight variation in arc length or the least hesitancy in moving the arc over the work will almost certainly result in its being burned through. In short, this class of welding calls for a coordination of faculties and a delicacy of manipulation beyond the capabilities of the most skillful manual electric welder. Therefore this work is usually done with the oxy-acetylene flame, wherein fusing conditions arc far more easily controlled than is possible in manual metallic electrode arc welding." SHEET METAL ARC- WELDING MACHINE The machine shown in Fig. 188 is used by the General Electric Co., Schenectady, N. Y., for arc-welding corrugated steel tank work. The seams are 116 in. long, and the arc 238 ELECTRIC WELDING is applied by means of a tapered carbon pencil 6 in. long, J in. in diameter at the large end and -J in. at the arc end. This concentrates heat where wanted. No metal is supplied to the weld, as the arc is employed simply to fuse the upturned edges as shown in Fig. 189. The metal welded is Vie and 3 / 32 in- thick. The speed on 1 / 16 -in stock is 5 1 / 2 in. per minute with a d.c. current of 45 amp., and 75 volts. On 3 / 32 -m. stock the speed is the same but 70 amp. and 75 volts d.c. current is used. CHAPTER XII BUTT-WELDING MACHINES AND WORK Aside from, arc-welding machines, which have already been described, electric welding machines may be all included under one head Resistance Welding Machines. These may be divided into butt-, spot-, seam-, mash- and percussive-welding classes. The first three are sometimes, for manufacturing pur- poses, used in combinations in the same machine, such as a spot-and-seam machine or a butt-and-spot-welding machine, and so on. This does not mean that these different methods of welding are carried on at the same time, but that a welder can do work on the same machine by simply shifting the work, or a part of the fixture. In butt-welding, alternating current, single phase, of any commercial frequency such as 220, 440 or 550 volts, 60 cycles, is commonly used. Lower voltages and lower frequencies can be used, but they add to the cost of the machine. The machine can be used on one phase of a two-phase or a three-phase system, but cannot be connected to more than one phase of a three-phase circuit. Direct current is not used because there is no way of reducing the voltage without interposing resist- ance, which wastes the power. As an example, a d.c. plating dynamo will give approximately 5 volts, which will do for certain kinds of welding, but for lighter work, less current is needed. If resistance is used to reduce the current this resist- ance is using up power just as if it were doing useful work. The voltage at the weld will run from 1 to 15 volts, depending on the size of the welder and work. To obtain this low voltage, a special transformer inside the machine reduces the power line voltage down to the amount required at the weld. The transformer is placed within the frame of the machine, as shown in Fig. 190. The secondary winding of the transformer is connected to the platens by means of flexible copper leads. 239 240 ELECTRIC WELDING From the platens the welding current travels to the work clamps and through them to the pieces to be welded. As the parts to be welded are brought into contact a switch is thrown in and the current traveling across heats the ends of the work and when the proper welding heat is reached the operator WORK STOP ClAMP ADJUSTMENT CLAMP JAW WITH STEEL DIE I/COPPER DIE CLAMP RELEASE CLAMP LOCKING Fie. 190. Principal Parts of a Butt-Welding Machine. pushes the two parts together and the weld is completed. Since the current value rises as the potential falls in the secondary circuit, and since the heating effect across the work is directly proportional to the current value it will be easily seen why a transformer is necessary to produce a heavy current by lower- BUTT-WELDING MACHINES AND WORK 241 ing the line potential. Due to the intermittent character of the load, there is no standard rating for welding transformers, and different makers frequently give entirely different ratings for their machines. However, regardless of the rating capacity in kilowatts, there can be very little difference in the actual amount of current consumed unless an especially bad FIG. 191. Butt-Welding Machine with Work in Jaws. transformer design is used. To heat a given size stock to welding temperature in a given time requires an approximately invariable amount of current. The machine just illustrated, is shown at a slightly different angle and with two pieces of rod in the jaws, in Fig. 191. This is the Thomson regular No. 3, butt-welding machine. It 242 ELECTRIC WELDING FIG. 192. Details of Foot-Operated Clamping Mechanism. PlG. 193. A Hand-Operated Clamp. FIG. 194. Toggle-Lever Clamp for Bound Stock. BUTT-WELDING MACHINES AND WORK 243 has a capacity of rod from j to J in. in diameter or flat stock up to Y 4 X2 in., in two separate pieces, or rings of Vie-in, stock and not less than 2 in. in diameter. Hoops and bands up to Yi 6 Xl 3 /4 in. and not less than 9V 2 * n - diameter when held below the line of welding, may also be welded. With jaws specially made to hold the work above the line of welding a minimum diameter of 4 in. is necessary. This machine will produce from 150 to 200 separate pieces, 150 to 300 hoops, or 300 to 400 rings per hour. The lower dies are of hard drawn copper with contact surfaces lVsX2 in.X2V 16 in. thick. FIG. '195. Clamping Device for Heavy Flat Stock. Standard transformer windings are for 220, 440 and 550 volts, 60 cycle current. Current variation for different sizes of stock is effected through a five-point switch shown at the left. Standard ratings are 15 kw. or 22 kva., with 60 per cent power factor. The dies are air cooled but the clamps to which the dies are bolted are water cooled. This type of machine occupies a floor space 40X33 in., and is 53 in. high. The weight is 1,750 Ib. A close-up view of the treadle-operated clamping .jaw mechanism is given in Fig. 192. The method of operating the clamping jaws differs accord- 244 ELECTRIC WELDING ing to the size of the machine and the work that is to be done. On some of the smaller machines the type of hand- operated clamp shown in Fig. 193 is used. On other machines, intended to handle round stock principally, the toggle lever clamp shown in Fig. 194 is used. For very heavy flat stock, the hand-lever clamping mechanism, shown in Fig. 195, is used. On some of the machines used on small repetition work the clamps and switch are automatically cam-operated as shown in Figs. 196 and 197. The first machine is a bench type used FIG. 196. A Cam-Operated Machine. for welding on twist drill shanks, and the second machine is used for welding harness rings. These jobs are, of course, merely examples as the machines are adapted for all sorts of the smaller welding jobs. Spring pressure, toggle-lever or hydraulic pressure arc used to give the final "shove-up" accord- ing to the machine used or weight of stock being welded. In welding hard steel wire of over 35 per cent carbon content, it is necessary to anneal the work for a distance of about 1 in. on each side of the wold. This is due to the fact BUTT-WELDING MACHINES AND WORK 245 that the wire on each side is rendered brittle by the cooling effect of the clamping jaws. To accomplish this annealing, all the small Thomson machines used for this work are equipped with a set of V-jaws outside of the clamping jaws, as shown in front in Fig. 198. The wire is laid in these V's with the FIG. 197. Automatic-Operated Machine Welding Harness Rings. weld half way between and the current is thrown on intermit- tently by means of a push button until the wire has become heated to the desired color, when it is removed and allowed to cool. The annealing of a small drill is shown in Fig. 199. The process of welding and annealing 12 gage, hard steel wire, 246 ELECTRIC WELDING FIG. 198. Machine Equipped with Annealing Device. FIG. 199. Annealing a Small Drill. BUTT-WELDING MACHINES AND WORK 247 requires about 30 sec. when done by an experienced operator. Copper and brass wire are easily welded in these same machines. The machine shown will weld iron and steel wire from No. 21 B. & S. to J in. i n diameter and flat stock up to No. 25 B. & S.Xi in. wide. Production is from 150 to 250 welds per hour, the actual welding time being 1| sec. on J-in. steel wire. The clamps are spring-pressure, with adjustable tension released by hand lever. The standard windings are furnished for 110, 220, 440 and 550 volts, 60 cycles. Five variations are made possible by the switch. The ratings are 1 kw. or 3 kva., with 60 per cent power factor. The weight is 120 pounds. For use in wire mills where it is desired to weld a new reel of wire to the end of a run-out reel on the twisting or braiding machines, it has been found convenient to mount the machine on a truck or. small bench on large casters. This enables one to move the welder from one winding machine to another very easily, to splice on new reels of small wire, the electrical connection to the welder being made by flexible cord, which is plugged into taps arranged at convenient points near each winding machine. It is also desirable to mount on this same bench a small vise in which to grip the wire to file off the burr resulting from the push-up of the metal in the weld. The average time required to weld, anneal and file up a 16-gage steel wire with this bench arrangement is only about one minute. The only preparation necessary for welding wire is that the stock be clean and the ends be filed fairly square so that they will not push by one another when the pressure is applied. In connection with welding wires and rods up to J in. in diameter, Table XX will be found very handy. For sizes from J to 2J in. the reader is referred to Table XXVI. Examples of Butt- Welding Jobs. while, as a rule, it is only necessary to have clean and fairly square ends for butt- welding in some cases where small welding is to be done it has been found best to bevel or V the abutting ends. This is more apt to be the case with non-ferrous metals, however, than with iron or steel. A notable example in the larger work is in the scarfing of the ends of boiler tubes when butt-welding is done. This phase of the question has apparently not been given the attention it deserves, and some cases where welding 248 ELECTRIC WELDING 0001 *H 'M 'X ui spi9M 0001 iad uoi^dums - u o o uotpas jo Gi I- co co * ur poi jo "Bi(j S t- OO OOaOG1 TjH OCOCOlOlOl^QOi I ' CO t^ * Ol C5 t^ ^ r- 1 * IO CO -i t^ t-n co 01 OOC5'-l'-'CO 5 g Q ii iM 2222N8cowi82eoi2Mi2igiiog8 ^ ||i| ^ i i i i ?a 1-1 O> t^ O 00 * 00 C5 CO -H 10 Tt< cN'*'O'*tOCOiOt~00 M 9 Hi* Id lOt* OP Q *4 M9 1 * M ^<0 COCOCOO w 3 ^ t~ oo oo t t^ ^ ej** *> o co j w

a5t>-OOOOO5-*tOO'n-l>'i-'t>- OO'-"-<^^iNC < 3CO'1<'-OOOOO'* < iOt-C>CN(NCDCOO^COt-O 02 W to ^ U DH n , 1 ii &g 00 rf -^ 00 CO CN Ci 1-1 O 00 O Tj< O frl CO CO CO * <* ELDING 1 >> $> i 2 i! i(5 IO O 00 O CO CO O5 CO 00 1-1 1-1 --H rH . O 00 -H - U MPT ION Q Nominal Internal ^ ^ ^ s; V J' t-l 1-1 1-1 > i Actual External 10 >0 >C 1C ^Ht>.-flOl-l O t^- t^ ^ 10 . 50 00 O CO OOiCO 00 iH i-H l-li ICJ (NCO't* (XIMATE ">? H Diametera Actual Internal 1C ^ -H (M 5O --i (M t CO 10 N 00 ociiN'fco >o r^O5co i-i o> >o M CM *# O t- O5 (M t" Oi CO 00 CO --H ^-i T-i N N CO 1 a, Nominal Internal \00\, f \00\N\,( l Vd\N \N \IN ri\ -5\ e*S \ C*S fi\ \ ^i\ \ i-H rH r* IM W CO CO H X 1 .^a OOOO^OCOTf >OM< -fh-O h-5O > Actual External iO O 'O 'O IO O'*t^'i <| o^ CO r- i^. ^lOCOOOOCO C005CO OOlO 10 _; ^ _; ^ '' HEADS WITH INSERT POINTS FIG. 227. Spot-Welding Machine for Heavy Work, with Parts Named. having over 15-in. throat, by a brass cover, insulated on the inside from the copper by a coating of asbestos sheet. The sliding head of the machine which carries the upper die-holder is a hollow steel plunger, sliding in a cast-iron head, which bolts to the body of the machine and on which arc mounted the control-switches. The pressure is applied by a toggle-motion above the plunger, actuated both by a swiveled hand-lever on top of the head, which may be swung into any 280 ELECTRIC WELDING position through an arc of 260 deg., and a foot-treadle at the base, which also may be swung in an arc of 30 deg. This enables the operator to control the machine by hand or foot from any position around the front of the machine. The current-control can be set to work automatically with the downward stroke of the upper die. In this case the pres- sure at the die-point is through an adjustable spring-cushion in the hollow cylinder-head. The current is automatically turned on after the die-points have come together on the work by further downward pressure of either lever. With the ap- plication of final pressure, to squeeze out any burnt metal as the weld is forced together, the current is automatically turned off. When working on pieces where more pressure is required to bring the parts together before welding than can be effected by the spring-cushion without turning on the current, it is possible to set a plug in the head of the machine so that direct connection is obtained from the hand-lever to the upper die-point while the foot-treadle still operates through the spring-cushion and with the automatic current-control. When it is desired to secure maximum pressure, the plug in the head can be set again so that both the hand-lever and the foot-treadle give direct connection to the die-point, the current being controlled by a push-buttom on the outer end of the hand-lever. The regular line of spot-welding machines of different makes, operate on 110-, 220-, 440- and 550-volt, alternating cur- rent. A welding machine of this kind can only be connected to one phase of an a.c. circuit. The transformer must be made to furnish a large volume of current, at a low voltage, to the electrodes. For further transformer details, the reader is referred to the article on butt-welding. The Thomson Foot-, Automatic-, and Hand-Operated Machines. The machine shown in Fig. 228 is representative of the Thomson line of small, foot-operated spot-welding machines. These are intended for use on light stock where but little pressure is required. The die-holders are water- cooled, arid the lower horn bracket allows the horn to be adjusted up or down for the use of various kinds of holders. The automatic switch and adjustable throw-in stop are plainlv shown at the back of the machine. SPOT-WELDING MACHINES AND WORK 281 The model is made in several sizes. The first size will weld from 30 to 16 B. & S. gage galvanized iron or soft steel, or to 24 gage brass. It will mash-weld wire from 14 gage to in. in diameter. Its throat depth is 12 in.; the lower horn drop clearance is 9 in.; size is 22X45X51 in. high; net weight FIG. 228. The Thomson Light Manufacturing Type Spot-Welding Machine. is 825 lb. ; full load rating is 5 kw., or 8 kva. The largest machine of this particular series, will weld 26 to 7 gage, B. & S., galvanized iron or soft steel, or 18 gage brass; it will mash-weld 10-gage to f -in. diameter wire ; has an 18-in. depth of throat; is 28X60X56 in. high; weighs 1,550 lb. and full load rating is 15 kw. or 25 kva. 282 ELECTRIC WELDING On repetition work, where the operator has to work the foot-treadle in rapid succession for long periods, it is very tiresome. For such work, power-driven machines similar to the one shown in Fig. 229 are made. These machines are sup- plied either with individual motor drive or pulley drive, as desired. The control is effected through the small treadle shown. The regular foot-treadle is used while setting up dies, FIG. 229. The Thomson Semi- Automatic Type Spot-Welding Machine. etc. If the operator desires to make but one stroke, he depresses the shorter treadle and immediately releases it, whereupon the machine performs one cycle of operation, automatically turn- ing on the current, applying the pressure, turning off the current, and stopping. A \- to -J-hp. operating motor is used according to the size of the machine. Otherwise the capacity of the various sizes is the same as in the regular foot-operated SPOT-WELDMG MACHINES AND WORK 283 FIG. 230. A Thomson Heavy-Duty Spot- Wei ding Machine. FIG. 231. Spot- Welding a Sheet Steel Box, 284 ELECTRIC WELDING machines. The lower horn and upper arm may be of either style illustrated. The machine shown in Fig. 230 is a hand-lever operated machine, although supplied with a foot-treadle which can be FIG. 232. Showing How the Horn and Welding Points May Be Set. swung back out of the way when not needed. This machine is typical of the Thomson designs used for the heavier run of commercial work. On the various sizes, the capacity for spot-welding is from 22 B. & S. gage galvanized iron or steel FlG. 233. Welding Small Hoe Blades to the Shanks. up to No. gage, or to 14 gage brass. Mash-welds may be made on from -J- to f-in. diameter wire. The throat capacities run from 15 to 51 in. and the lower horn adjustment is from 12 to 24 in. The smallest size is 28X62X75 in. high and the SPOT-WELDING MACHINES AND WORK 285 FiG. 234. Welding Stove Pipe Dampers. FlG. 2b5. Mash- Welding Lamp Shade Frames. 286 ELECTRIC WELDING FlG. 236. Butt-Welding Attachment for a Spot- Welding Machine. FlG. 237. Welding Galvanized Iron Pipe. SPOT-WELDING MACHINES AND WORK 287 largest size 28X98X75 in. high. The weights run from 2,335 to 3,225 and the full load ratings from 20 to 40 kw. or 35 to 67 kva. Various shaped horns, dies and other equipment are furnished to meet special demands. Examples of Spot- Welding Work. In connection with the Thomson machines, the welding of the corners of a sheet-steel box is shown in Fig. 231. The illustrations in Fig. 232 show how the lower horn is raised for welding side seams and dropped for welding on the bottom of a box. The welding of small hoe blades to the shanks is shown FIG. 238. Welding 12-Gage Iron for Guards. in Fig. 233. These are welded at the rate of 840 per hour, the shanks being bent afterward. Stove-pipe dampers are welded as shown in Fig. 234, and wire lamp-shade frames are mash-welded as shown in Fig. 235. Ordinary wire and sheet- metal oven gratings or racks, with seven cross-wires welded to the end pieces, have been made at the rate of 100 racks per hour, or 1,400 mash-welds. On certain kinds of wire work, it is desirable to butt-weld, and for this purpose the attach- ment shown in Fig. 236 is used. In general, however, where any amount of this kind of work is to be done, it is better 288 ELECTRIC WELDING to employ a regular butt-welding machine of the small pedestal or bench type. The spot-welding of galvanized ventilating pipe is shown in Fig. 237, and in Fig. 238 is shown the welding of 12 gage sheet steel machine guards. In this illustration the operator is using the foot-treadle which leaves his hands free to manipulate the work. In Fig. 239 the operator is welding gas-stove parts and the foot-treadle is thrown back out of the FIG. 239. Welding Stove Parts, Using a Swinging Bracket Support. way. A special bracket is employed to hold the work. The joints of this bracket are ball-bearing, making it very easy to swing the work exactly where it is wanted to obtain the spot-welds. POINTS FOR SPOT WELDING The form of spot-welding points shown in Fig. 240, says A. A. Karcher, has been developed by the Challenge Machinery Co., Grand Rapids, Mich., with gratifying results. Fig. 241 shows a typical weld and indicates the neatness, slight dis- SPOT-WELDING MACHINES AND WORK 289 FIG. 240. Form of Points for Spot Welding. FIG. 241. Spot Weld Showing Slight Discoloration and Freedom from Flash. 290 ELECTRIC WELDING coloration of the metal and entire freedom from flash either on the outside or between the parts. In one view the dis- colorations give an erroneous impression of the existence of bosses on the face of the metal, which is actually flat except for the depressions at the points of the welds. The shape of the points would lead one to expect that the small projections would require a lot of attention to keep them in shape. Experience shows, however, that this is not the case, as the points actually lengthen slightly and occasion- ally have to be filed down. Even when a weld is made close to the edge the operation is quicker and consumes less current. A little practice in determining the correct amount of current to use is all there is to learn in handling these points. SIZES OF DIE-POINTS FOR LIGHT WORK The data on the size of die-points in Fig. 242 arc given on the authority of Lucien Haas, and may be considered good Rounded Points FIG. 242. Sizes of Die Points for Light Work. general practice. These points are intended for welding two pieces of the same gage and material. On certain kinds of heavy spot-welding work circular metal disks are placed between the plates in order to localize the current and to provide good contact. In other cases, projec- tions are made in one or both of the plates. These latter, of course, necessitate a mechanical or press operation, previous SPOT-WELDING MACHINES AND WORK 291 .Welding Preware On Completion of Heating before Welding Presiure ia Applied After Completion by Arc Welding, for Calking Purpoie TIG. 243. The Tit or Projection Method of Welding. FIG. 244. Winfield Sliding Horn Spot- Welding Machine. 292 ELECTRIC WELDING FIG. 245. Winfield Heavy-Duty Machine with Adjustable Table. FlG. 246. Winfield Portable Spot- Welding Machine. SPOT-WELDING MACHINES AND WORK 293 to welding. Heavy plate work is shown in Fig. 243. At the upper left are shown plates as commonly arranged for welding. Next to this is a plate with a projection under the upper die- FIG. 247. Winfield Portable Machine with Swivel Head. point. A steel plunger is used in the lower die to give the needed pressure after the metal is heated. This saves crushing or distorting the soft copper. In the lower right-hand corner 294 ELECTRIC WELDING FIG. 248. Small Winfield Bench Machine. FIG. 249. Winfield Machine with Suspended Head for Welding Automobile Bodies. SPOT-WELDING MACHINES AND WORK 293 is shown a ridge or tit weld, after the seam has been arc- welded. The Winfield Machines. The machines made by the Win- field Electric Welding Machine Co., Warren, Ohio, comprise a varied line for every conceivable spot-welding purpose. In general, Figs. 244 and 245 may be taken as typical of their FIG. 250. Convenient Setting of Machine for Sheet Metal Work. light and heavy spot-welding machines. Fig. 246 shows a very convenient form of portable machine. In Fig. 247 is shown a much heavier portable machine with swiveling head, and in Fig. 248 is a small bench machine that is exceedingly useful for light work. 296 ELECTRIC WELDING A very interesting machine is shown in Fig. 249. This has the entire head suspended from the ceiling, so that work, like the automobile body shown, may be worked under it. FIG. 251. Federal Welding Machine with Universal Points. This machine is in use in the plant of the Herbert Manu- facturing Co., Detroit. A good way to place a machine for some work is shown in Fig. 250. This is employed in the shop of the Terrell SPOT-WELDING MACHINES AND WORK 297 Equipment Co., Grand Rapids, Mich., in the manufacture of steel lockers, steel furniture and the like. Federal Welding Machines. A feature of the spot-welding machines made by the Federal Machine and Welder Co., War- ren, Ohio, are the "universal" welding points used on most of their output. The principle will be instantly grasped by FIG. 252. A Few Positions of the Universal Points. referring to Fig. 251. Some of the different positions possible are shown in Fig. 252. Another feature of these machines, is the use of the type of water-cooled points shown in Fig. 253. The welding point is copper and it is attached to the holder in such a way that the water flows within half an inch of the actual welding contact. 298 ELECTRIC WELDING In general form, size and capacities, the Federal line does not differ materially from the machines already shown. PIG. 253. Federal Water-Cooled Points. FEDERAL ROTATABLE HEAD TWO-SPOT WELDING MACHINE The rotatable head two-spot, air operated welding machine, shown in Fig. 254, a 60-in. throat depth and is guaranteed to weld from two thicknesses of 24-gage up to two thicknesses SPOT-WELDING MACHINES AND WORK 299 of 8-gage steel stock. Twelve welds per minute may be made in the latter size. The machine is built with a 4 kva. welding transformer in the upper and lower rotating heads. Primaries are in parallel while the secondaries are in series, so that two spot welds must be made at the same time. The welding electrodes or points are 1J in. in diameter, are carried in water-cooled holders, and are so arranged that FIG. 254. Federal Rotatable Head Two-spot Welding Machine. welds from 3 to 8 in. apart may be made. The ends of each set of welding points can be separated a maximum of 5 in. The heads can be rotated through an angle of 90 deg. to permit welding at different angles on the stock being handled. Four air cylinders are used, each operating an independent point. The air control is hand operated and so arranged that an initial air line supply pressure of 80 Ib. will give from 300 to 700 Ib. pressure between the points during the heating period. A second step on the air control makes it possible 300 ELECTRIC WELDING to apply 1,200 Ib. pressure between the points for the final squeeze. The air is exhausted into the reverse side of the cylinders to withdraw the points. The regulating transformer supplies power to the welding transformer in eight voltage steps. FEDERAL AUTOMATIC SPOT-WELDER FOR CHANNELS The machine shown in Fig. 255 was made for spot-welding two rolled steel channels together to form an I-beam. It is FIG. 255. Federal Channel Welding Machine. capable of welding two spots at a time on two pieces of material J in. thick, at the rate of 60 welds per min. The two welding transformers are for 220 volts primary, and are air cooled. Four copper disks are used for welding contacts. These are securely bolted to bronze shafts to insure good elec- trical connections. The secondaries of the welding trans- formers are connected to the brass bearings of these shafts, com- pleting the welding circuit. The welding current is controlled by auto transformers SPOT-WELDING MACHINES AND WORK 301 in the primary circuit in eight equal steps from 65 per cent to full line voltage. The welding disks can be adjusted to handle from 4 to 16 in. channels. Simultaneous spot welds from 4 to 12 in. apart may be made. A variable speed motor is used to control the feeding of the work through the machine at from 25 to 60 ft. per min. AUTOMATIC PULLEY WELDING MACHINE The machine shown in Fig. 256 was made to weld the ring section of pressed-metal pulleys, known as the filler, to the FIG. 256. Automatic Electric Pulley Welder. rim itself. This ring, or filler, not only acts as a stiffener for the rim, but is the part to which the outer ends of the spokes are attached. In welding, one-half of a pulley rim is locked by means of a chain-clamping device to a rotating carrier, with the filler and spokes in place as shown. An adjustable mandrel on the 302 ELECTRIC WELDING carrier insures the proper distance between the center of the pulley and the rim face. Duplicate welding sets operate on each side of the filler, and spot weld intermittently as the work is automatically indexed around. The mechanical part of the machine is motor driven, and with the work in place, the machine will properly space and weld around the filler until it reaches the end, when it auto- matically trips. The points are water cooled and will make FIG. 257. Taylor Cross-Current Spot-Welding Machine. about 60 welds per minute. These welding points can be set to weld within 2^ in. of the center of the mandrel or supporting shaft, and have a maximum distance adjustment of 12 in. between them. The automatic indexing or feeding device is so arranged that welds from to 3 in. or more apart may be made. Pulleys from 12 in. up to 5 ft. in diameter may be handled, all the necessary adjustments being easily and quickly made to accommodate the various sizes. SPOT-WELDING MACHINES AND i^ORK $03 This machine occupies a floor space of about 30X66 in., weighs about 3,500 Ib. The Taylor Welding Machines. While the machines made by the Taylor Welder Co., Warren, Ohio, differ radically from others on the market, in that they employ double electrodes and cross current, the forms of the machines are about the same as those previously shown. An automatic belt-driven machine of the lighter type, is shown in Fig. 257. It may FIG. 258. Taylor Heavy-Duty Machine. be operated by the foot-treadle also when desired. This machine has a capacity up to two -in. plates. The horns are water-cooled and the adjustable points are locked in with a wrench as shown. Fig. 258 shows a heavier type of machine. This has a capacity of two j-in. plates ; overhang is 36 in. ; distance between copper bands and lower horn, 6 in. ; base, 26x42 in.; extreme height, 72 in.; greatest opening between welding points, 3 in. ; weight about 2,400 Ib. The transformer is 35 kw. and there is a ten-step self-contained regulator for 304 ELECTRIC WELDING controlling the current. This firm makes other sizes and styles of machines, to meet all the demands of the trade. The general principle of the cross-current welding method employed in these machines is illustrated in Fig. 259. Two separate currents are caused to flow in a bias direction through the material to be welded. A high heat concentration is claimed for this method. In operation, the positives of two separate CROS ENT SPO ING FIG. 259. Diagram of the Current Action in a Taylor Machine. welding currents are on one side of the material and the negatives on the other, with the co-working electrodes of each set so that the current travels diagonally across. An advantage claimed is that the electrodes on each side of the material may be set far enough apart to allow of the insertion of some hard material which will take the pressure instead of the softer copper welding points. These hard dies may be operated independently of the copper ones and make it possible to weld SPOT-WELDING MACHINES AND WORK 305 heavier material without crushing the copper die points, as these need to be pressed together only enough to give good FlG. 260. Automatic Hog-Ring Machine. FIG. 261. Partial Rear View of Hog-Ring Machine. electrical contact with the work. The process is also unique in that it can be operated with a multiphase circuit without 306 ELECTRIC WELDING unbalancing the lines, whicli is not the case with any spot- welding machine employing a single current. Some Special Welding Machines. An automatic machine for forming and mash-welding 11 gage wire hog rings, at the FIG. 262. Close-Up of Front of Hog-Ring Machine. rate of 60,000 per day, is shown in Fig. 260. This machine takes wire from two reels and turns out the complete hog rings. A partial rear view is shown in Fig. 261. A close-up of the front of the machine, with two hog rings lying on the platen, is given in Fig. 262. SPOT-WELDING MACHINES AND WORK 307 A machine in use in the punch press department of the General Electric Co., Schenectady, N. Y., is shown in Fig. 263. This machine welds small spacers to the iron laminations for motors and generators for ventilating purposes, and hence is FIG. 263. General Electric Space-Block Welding Machine. called a " space-block welder." A number of these machines are in use in this plant, and they are capable of welding 60 spots per minute when working continuously, not allowing for time to shift the stock. A combination spot- and line-welding machine, used in the 308 ELECTRIC WELDING General Electric Co.'s shops, is shown in Fig. 264. This is employed for welding oil switch boxes up to -J in. thick. As shown, the machine is fitted with a fixture for holding the boxes while line-welding the seams. A separate fixture is put FIG. 264. Combination Spot- and Line-Welding Machine, Set Up for Line- Weld ing Can Seams. on for spot-welding work. A seam 6 in. long can be line- welded on this machine. Another combination machine, used in the same shops, is shown in Fig. 265. This machine carries both the spot- and the line-welding fixtures at the same time. Fig. 266 shows the machine from the line-welding side. As shown, the SPOT-WELDING MACHINES AND WORK 309 machines are ready for welding straight plates. Machines of this kind should find a considerable field where it is desired to tack seams before line welding them. These machines have FIG. 265. A Combination Machine from the Spot- Weld ing Side. a capacity of 20 kva., and will weld up to 3 / 16 in- thick, and seams 18 in. long. Line welding machines, as developed in the Schenectady plant, comprise a transformer with a one turn secondary, through which a heavy current is delivered at low voltage to the material through the medium of a stationary jaw and roll- 310 ELECTRIC WELDING ing wheel. Both the jaw and wheel are water-cooled and pressure is applied to the wheel the same as to a spot-welding tip. A small revolving switch mechanically geared to the driving motor and welding wheel operates a set of contactors FIG. 266. Machine from the Line- Weld ing Side. or solenoid switches to throw the power on onco a second, the power being on f of a second, and off f of a second. The mechanism is synchronized so that during the f of a second the power is on, the welding wheel is rolling, and during the SPOT- WELDING MACHINES AND WORK 311 remaining f of a second the wheel is stationary under pressure while the soft metal is solidifying, thus completing the weld. Spot- Welding Machines for Ship Work. During the World War, welding of all kinds took huge steps forward. Spot- welding developed at least as much as any other kind. Writing in the General Electrical Review, J. M. Weed says: The machines to be described are two portable welders, one with 12-in. reach and the other with 27-in. reach, for use in the fabrication of structural ship parts, and one stationary machine with 6-ft. reach designed for welding two spots at the same time on large ship plates. A preliminary survey of the structural work in shipbuilding indicated that about 80 per cent of this work could be done by a machine of 12-in. reach, and that a 27-in. reach would include the other 20 per cent. Since both the weight of the machine and the kva. required for its operation are about 33 per cent greater for the 27-in. reach than for the 12-in., it seemed advisable to develop two machines rather than one with the longer reach. These machines were to a certain obvious extent patterned after the riveting machines, which they were intended to replace as will be seen from Fig. 267. They are necessarily considerably heavier than the riveting machines, but like these they are provided with bales for crane suspension, for the purpose of carrying the machines around the assembled work or parts to be welded. The maximum welding current available- in these machines, with a steel plate enclosed to the full deptn of the gap, is about 37,500 amperes, with the maximum applied voltage of 534 volts at 60 cycles. Reduced voltages, giving smaller currents, are obtained in six equal steps, ranging from 534 down to 267 volts, from the taps of the regulating transformers furnished with the machines. This wide range of voltage and current was provided in order to meet the possible requirements for a considerable range in thickness of work, and for experimental purposes. Tests have shown, however, that the machines will operate satisfactorily on work of thicknesses over the range on which they are likely to be used when connected directly on a 440-volt, 60-cycle circuit, with no regulating transformers. Two plates -in. thiek are welded together in spots from 1 in. to 1 in. in diameter, in from 12 to 15 seconds. Thicker plates require more time and thinner plates less time. The welding current under these conditions is about 31,000 amp.; the primary current is about 600 amp. for the 12-in. machine and about 800 amp. for the 27-in. machine, the corresponding kva. at 440 volts, being 265 and 350 respectively. Since the reactance of the welding circuit is large as compared with the resistance, the voltage necessary for a given current, and conse- quently the kva. necessary for the operation of the machine, is almost proportional to the frequency. Thus, these machines operate satis- 312 ELECTRIC WELDING factorily from a 25-cycle circuit at 220 volts, with the advantage that where the power-factor is from 30 to 40 per cent at 60 cycles, it is from 60 to 75 per cent at 25 cycles, and the kva. required at 25 cycles is about one-half that required at 60 cycles. The maximum mechanical pressure on the work for which those machines are designed is 25,000 Ib. This is obtained from an 8-in. air cylinder, with an air pressure of 100 Ib. per square inch, acting through a lever arm of 5 to 1 ratio. Lower pressures on the work are obtained with FlG. 267. Portable Spot-Welding Machine with 27-in. Throat Depth. Capable of Welding Two Plates In. Thick in Spots 1 In. in Diameter. Made by the General Electric Co. correspondingly reduced air pressures. A pressure-reducing valve is pro- vided for this purpose, and also a pressure gage for indicating the pressure on the machine side of the valve. The pressure required to do satisfactory welding depends upon the thickness of the plates. It is necessary that the areas to be welded should at the start be brought into more intimate contact than the surrounding areas, in order that the current may be properly localized, and the heat SPOT- WELDING MACHINES AND WORK 313 generated in the region where it is needed. It is therefore necessary, on account of irregularities in the plate surface, that the pressure should be great enough to spring the cold plate sufficiently to overcome the irregulari- ties. The pressure which will do this with heavy plates is ample for effecting the weld after the welding temperature is reached. It should be explained in this connection that the rate of heating at the surfaces to be welded depends largely upon the contact resistance, and consequently upon the condition of the plates and the pressure used. If the plates are clean and bright, and the pressure high, the rate of heating with a given amount of current is slow and the welding efficiency is poor. This makes it difficult to weld heavy plates if they are clean, since, as stated above, it is necessary to use large pressure with heavy plates to insure a better contact of the areas to be welded than that of surrounding areas. It is much easier to weld plates which carry the original coat of mill scale, or a fairly heavy coating of rust or dirt, affording a considerable resistance which is not sensitive to pressure. If this resistance is too great, the necessary current will not flow, of course, but if the scale is not too heavy it has little effect upon the current, the high reactance of the welding circuit giving it practically a constant current characteristic and making the rate of heating proportional to the resistance within certain limits. The scale melts at about the welding temperature of the steel, and is squeezed out by the high pressures used, permitting the clean surfaces of the steel to come together and effect a good weld. A gage pressure of about 70 lb., giving 17,500 Ib. pressure upon the work, has been found to give good results under these conditions in ^-in. plates. Both the mechanical pressure and the current are transmitted to the work in these machines through heavy copper blocks or welding electrodes. The shape of the tips of these electrodes is that of a very flat truncated cone. The severity of the conditions to which the tips of the electrodes are subjected will be understood when it is considered that the current density in the electrode material at this point is approximately 60,000 amp. per square inch, and that this material is in contact with the steel plates which are brought to the welding temperature, under pressures of 15,000 to 20,000 lb. per square inch. It must be remembered, also, that copper, which is the best material available for this purpose, softens at a temperature considerably lower than the welding temperature of steel. The difficulty of making the electrode tips stand up under the conditions to which they are subjected has, in fact, constituted the most serious problem which has been met in the development of these machines. The shape of these electrodes gives them every possible advantage in freely conducting the current to and the heat away from the electrode tips, and in giving them the mechanical reinforcement of the cooler sur- rounding material. However, it has been found necessary to reduce, as far as possible, the heat generated at the tips of the electrodes by cleaning the rust and mill scale from the surfaces of the plates beneath the elec- 314. ELECTRIC WELDING trodes. The most convenient way which has been found for ooing this is by means of a sand blast. The bodies of the electrodes are also internally water-cooled by a stream of water flowing continually through them. Still, after all of these things have been done, a gradual deformation of the tip of the electrode will occur, increasing its area of contact with the work, and thus reducing the current density in the work and the pressure density below the values needed for welding. This would make it necessary to change electrodes and to reshape the tips very frequently, and the total life of the electrodes would be short on account of the frequent dress- ings. An effort has been made to overcome this difficulty by protecting the tip of the electrode by a thin copper cap, which may be quickly and cheaply replaced. As many as 160 welds have been made with a single copper cap, y J6 in. thick, before it became necessary to replace it. Un- fortunately this does not entirely prevent the deformation of the electrode tip, but it stands up much better than it does without the cap. Another method which has been tried for overcoming this trouble is by making the tip portion of the electrode removable, in the form of a disk or button, held in place by a clamp engaging in a neck or groove on the electrode body. While this protects the electrode body from deformation and wear, the tip itself does not stand up so well as does the combination of electrode and cap, where the tip of the electrode is not separated from the body. Some electrodes have been prepared which combine the features of the removable tip and cap. These give the advantage of a permanent electrode body, and the removable tip with the protecting cap stand up better than the unprotected tip. Some interesting features were introduced in the design of the trans- formers which are integral parts of these machines, owing to the necessity for small size and weight. Internal water cooling was adopted for the windings, which makes it possible to use current densities very much higher than those found in ordinary power transformers. The conductor for the primary windings is -in.X2'i n ' copper tubing, which was obtained in standard lengths and annealed before winding by passing it through an oven which is used for annealing sheathed wire during the process of drawing. No difficulty was found in winding this tubing directly on the insulated core, the joints between lengths being made by brazing with silver solder. The entire winding consists of four layers of thirteen turns each in the 12-in. machine and three layers of thirteen turns each in the 27-in. machine. The U-shaped single-turn secondaries were slipped over the outside of the primary windings in the assembly of the transformers. These were constructed of two copper plates each in. thick and 6| in. wide, which wero bent to the proper shape in the blacksmith shop, and assembled one inside the other with a ^-in. space between them. Narrow strips of copper were inserted between the plates along the edges, and the plates were brazed to these strips, thus making a water-tight chamber or passage for the circulation of the cooling water. SPOT-WELDING MACHINES AND WORK 315 At 31,000 amp. the current density in these secondaries is about 6,200 amp. per square inch, the corresponding densities in the primary windings being about 7,000 for the 12-in. and 9,000 for the 27-in. machine. In case these machines are started up without the cooling water having been turned on, the temperature rise in these windings will be rapid, and in order to avoid the danger of burning the insulation, asbestos and mica have been used. The copper tubing was taped with asbestos tape, and alternate layers of sheet asbestos and mica pads were used between layers of the primary winding, and between primary and secondary and between primary and core. Space blocks of asbestos lumber, which is a compound of asbestos and Portland cement, were used at the ends of the core and at the ends of the winding layers. The complete transformer, after assembly, was impregnated with bakelite. The result is a solid mechanical unit which will not be injured by temperatures not exceeding 150 deg. C. Several welds could be made without turning on the cooling water before this temperature would be reached. The transformers are mounted on a chamber in the body of the frame. The long end of the U-shaped secondary runs out along the arm of the frame and bolts directly to the copper base upon which the bottom electrode is mounted. The short end connects to the base of the top electrode through flexible leads of laminated copper, to permit of necessary motion for engaging the work. The copper bases upon which the electrodes are mounted are insulated from the frame by a layer of mica, the bolts which hold them in place being also insulated by mica. The cooling water for these machines is divided into two parallel paths, one being through the primary winding, and the other through the secondary and the electrodes in series. Separate valves are supplied for independent adjustment of the flow in the two paths. The resistance of ordinary hydrant water is sufficiently great as to cause no concern regarding the grounding or short-circuiting of the windings through the cooling water, although it is necessary to use rubber tubing or hose for leading it in and out. Some pieces of %y(2-in. machine steel were welded in seven seconds with a current of 33,000 amp. They were afterward clamped in a vise and hammered into U-shapes. Small pieces were sheared from the seam where two - in. plates had been welded together in a row of spots. The pieces of the plates were then split apart with a cold chisel in one case, and an effort was made to do so in the other, with the result that one piece of plate broke at the welds before the welds would themselves break. Such tests as these show that the welds are at least as strong as the material on which the welds were made. Some samples of the x2-in. stock welded together in the same manner were tested by bending in an edgewise direction, thus subjecting the welds to a shearing torque. The ultimate strength calculated from these tests was in the neighborhood of 65,000 Ib. per square inch. These tests showed also a very tough weld, the deflection being almost 45 deg. in some cases before the final rupture occurred. The maximum load occurred with a deflection of from 3 to 5 316 ELECTRIC WELDING deg. with a very gradual reduction in the load from this time till the final rupture. The Duplex Welding Machine. The machine shown in Fig. 268 was developed for the application of electric welding as a substitute for riveting on parts of the ship composed of large-sized plates, which may be fabricated before they are assembled in the ship. The specification to which it was built stated that it should have a 6-ft. reach and should be capable of welding together two plates f in. thick in two spots at the same time. A machine capable of doing this work, with a 6-ft. gap, is necessarily FIG. 268. Duplex Spot-Welding Machine. Made by the General Electric Co. 6-ft. Throat Depth, and Capable of Welding Together Two Steel Plates | In. Thick, in Two Spots 1 In. in Diameter. so heavy as to preclude even semi-portability, and no effort was made in this direction. With the welding circuit enclosing a 6-ft. gap, and carrying the very heavy current necessary to weld f-in. plates, the kva. required would be very large. A great reduction in the kva. and at the same time a doubling of the work done, is obtained in this machine by the use of two trans- formers as integral parts of the machine, and two pairs of electrodes, thus providing for the welding of two spots at the same time. The transformers are mounted in the frame of the machine, on opposite sides of the work, and as near to the welding electrodes as possible, so as to SPOT-WELDING MACHINES AND WORK 317 obtain the minimum reactance in the welding circuit. The polarity of the electrodes on one side of the work is the reverse of that of the opposed electrodes, thus giving a series arrangement of the transformer secondaries, the current from each transformer flowing through both of the spots to be welded. The bottom electrodes are stationary, and the copper bases which bear them are connected rigidly to the terminals of their transformer, while the bases which carry the top electrodes are connected through flexible leads of laminated copper, to permit of the motion necessary for engaging the work. Previous tests with an experimental machine had shown that, to suc- cessfully weld two spots at the same time in the manner adopted here, it is necessary that the pressures shall be independently applied. Otherwise, due to inequalities in the thickness of the work, or in the wear and tear of the electrodes, the pressure may be much greater on one of the spots than on the other. This results in unequal heating in the two spots. The resistance and its heating effect are less in the spot with the greater pressure. The two top electrodes in this machine were therefore mounted on separate plungers, operated by separate pistons through independent levers. The pressures obtained in this machine with an air pressure of 100 Ib. per square inch, are 30,000 Ib. on each spot, giving a total pressure of 60,000 Ib. which must be exerted by the frame around the 6-ft. gap. The necessary strength is obtained by constructing the frame of two steel plates, each 2 in. thick, properly spaced and rigidly bolted together. The use of steel in this case is easily permissible on account of the restricted area of the welding circuit and its relative position, resulting in small tendency for magnetic flux to enter the frame. However, the heads carrying the electrodes, being in close proximity to the welding circuit, were made of gun metal. The two air cylinders are mounted on a cast-iron bed-plate in the back part of the machine. The levers connecting the pistons to the electrode plungers, which are 7 ft. in length, were made of cast steel, in order to obtain the necessary strength. The maximum welding current for which this machine was designed is 50,000 amp. This current is obtained with 500 volts at 60 cycles applied. The distance between the electrode bodies for this machine is fixed at 8 in., center to center, but the distances between the centers of the tips may be easily varied from 6 in. to 10 in. by shifting the tip from the center of the body toward one side or the other. Provision has been made for shifting the electrodes on their bases to positions 90 deg. from those shown in the picture, thus spacing the welds in a direction along the axis of the machine instead of traverse to it. The transformers are insulated and cooled in the same manner as those in "the semi-portable machines. The windings are interlaced in order to obtain minimum reactance, the primary being wound in two layers of 14 turns each, one inside and the other outside of the single turn secondary. With 50,000 amp. in the secondaries of these transformers, the current 318 ELECTRIC WELDING in tne primary is 1,800. The respective current densities are 7,000 and 9,000 amp. per square inch. The kva. entering the transformers on this basis, the two primaries being in series on 500 volts, is 450 for each transformer. This machine also has been provided with a regulating transformer for applying different voltages to give different values of welding current, FIG. 269. General Electric Co.'s Experimental Spot-Welding Machine. Current Capacity 100,000 Amp. Pressure Capacity 36 Tons. Has Welded Three Plates, Each 1 In. Thick. and with a panel carrying the necessary selector switches and contactor. The maximum voltage provided by this regulating transformer as at present constructed is 440. If it is found that the current obtained with this voltage is not sufficient for the heaviest work which it is desired to do with this machine, the maximum voltage may be changed to 500. The kva. entering the transformers of 440 volts will be approximately 350 each, instead of 450. SPOT-WELDING MACHINES AND WORK 319 In order that this machine may be operated from any ordinary power circuit, it will be necessary to use a motor-generator set provided with a suitable flywheel. This will eliminate the bad power-factor, distribute the load equally on the three phases, and over a much larger interval of time for each weld, thus substituting small gradual changes in power for large and sudden changes. On account of the high reactance the welding current will remain practically Constant as the speed of the motor-generator set falls away, thus favoring the utilization of the flywheel. The total maximum power drawn from the circuit with this arrangement would be about 100 kilowatts. FIG. 270. Portable Machine for Mash- Welding Square or Bound Rods. A Heavy Experimental Spot- Welding Machine. The machine shown in Fig. 269 was built in 1918 by the General Electric Co., in order to investigate the possibilities of welding plates from. in. up. Three plates each 1 in. thick have been welded with it. The machine is provided with a 2,000-kva. transformer, having a capacity of 100,000 amp. at 20 volts. Hydraulic pressures up to 36 tons are obtained at the elec- trodes. Motor-generator sets of 500- and 6,000-kva. capacity 320 ELECTRIC WELDING were used. From the nature of the service, it was apparent that some form of cooling was needed at the contact points. It was found, however, that it was impossible to water-cool the points sufficiently to give a reasonable life to the electrodes if they were kept the same diameter for any distance from the work. In consequence heavy masses of copper were placed FIG. 271. Lorain Machine for Spot- Welding Electric Rail Bonds. as close to the points of contact as practicable. By doing this it was possible to have a very large cooling surface at the top of the electrode and by passing water through this part at the time of welding and between welds, the joints were kept cool enough for all practical purposes. A portable machine for making mash-welds for splicing or attaching round or square rods cross-wise, is shown in Fig. SPOT-WELDING MACHINES AND WORK 321 270. This was made by the General Electric Co., for ship- yard use. A big machine for spot-welding electric railway bonds, is shown in Fig. 271. This is made by the Lorain Steel Co., Johnstown, Pa. It will weld two plates 18 in. long and 3 in. wide by 1 in. thick, each plate having three raised "welding bosses. ' ' Pressure as high as 35 tons is obtainable and current up. to 25,000 amp. may be used. Spot- Welding Data. It is difficult to give definite costs for spot welding, as much depends on the operator. A careless or inexperienced operator will waste more current than a good one, and various conditions of the metal being worked on will make a considerable difference at times. However, the information given in Table XXIII, which is furnished by the Winfield Electric Welding Machine Co., will prove of value as a basis for calculations. Tables XXIV and XXV will also be useful to use in connection with the measurement of the thickness of sheets, and in comparing different gages. TABLE XXIII. SPOT-WELDING POWER AND COST DATA Gauge Number Thickness of Sheets in Fractions of an Inch Thickness of Sheets in Decimals of an Inch K. w. Required H. P. Required Time in Seconds to Make a Weld Cost 1000 Welds at one Cent per K. W. Hour 30 Vso .0125 3.0 4.2 -.25 .002 28 74 .0156 4.0 5.6 .3 .003 24 7*0 .0250 5.0 7.0 .45 .006 20 Vso .0375 6.5 9.2 .6 .011 18 Yso .0500 8.0 11.3 .8 .017 16 V* .0626 9.5 13.5 1.0 .026 14 Ye* .0781 10.0 14.2 1.3 .036 12 v .1093 12.0 17.0 1.6 .052 11 % .1250 13.0 18.5 1.7 .061 10 Ye* .1406 14.0 19.9 1.8 .070 9 Ya 2 . .1562 15.0 21.3 1.9 .079 8 "/64 .1715 16.0 22.7 2.0 .088 7 Yl6 .1875 17.0 24.1 2.1 .099 6 */ m .2031 18.0 25.6 2.2 .110 5 Y32 .2187 19.0 27.0 2.4 .124 4 M / .2343 20.0 28.4 2.7 .148 3 % .2500 21.0 29.8 3.0 .174 As the cost of current varies in different places, we have figured the current at one cent per K. W. hour to give a basis for calculating the cost. Multiply the cost of current given above by the rate per K. W. hour you pay and you will have your cost per 1000 welds for current. 322 ELECTRIC WELDING 10 t- IQ 01 *0 (X 0*0 O I- GO CO N 00 i - t~ CO COOitOOCD^t^COOO"* CD < t *S t O O O S c^ *O 00 -i 3a O O CD r-t l> 00 I^H ( GO O 1> < to GO -" 00 1-1 *0 00 -<* t- GO CO 5 Z fc Lr.J CHAPTER XIV WELDING BOILER TUBES BY THE ELECTRIC RESISTANCE PROCESS About 1912 the resistance, or Thomson, process of electric welding was first tried out in a locomotive shop for the purpose of replacing the oil-furnace welding equipment in safe-ending boiler tubes up to 2^ in. in diameter, says P. T. Van Bibber, in the American Machinist. At the present time, in shops in different parts of the country where electric welding machines have been installed, one will find many enthusiastic " boosters" for this process. It is to these users that we are indebted for the information contained in this article and for the benefit of those who are unfamiliar with this adaptation of resistance welding, an endeavor has been made to cover all the details possible. In using the resistance type of machine for welding safe- ends onto locomotive-boiler flues, the old tube and the new safe-end are gripped securely in heavy copper jaws with the ends to be joined held in alignment. As these ends are pressed together a large volume of current from the secondary winding of the transformer is passed through them. Since the junction of the abutting ends is the point of greatest resistance to the electric current, the greatest heating effect is there and, usually, on a 2^-in. tube it requires only about 15 sec. to secure a perfect running or welding heat. A slight push-up by the pressure device on the welding machine sticks the two parts together solidly enough so that the tube can be removed to the mandrel of a rolling machine, exactly as is done when welding by the oil-furnace method, and the weld is then com- pleted in a few seconds by rolling down the joint. Since it is always necessary to scarf the ends of a tube and new safe-end before welding by the oil-furnace method, the first question that the practical boiler-shop man 324 WELDING BOILER TUBES 325 will ask is, How much preparation is needed for electric resist- ance welding? The first step in any method is to clear the tube from heavy scale, if in use under bad water conditions, by rolling in a large tumbling barrel. After this, the tubes are cut to the desired length to remove the old end that is to be replaced by the new section. In some shops it is the practice to never allow more than one or two welds in a tube, which means that after removing the second time, the tube must be used in a shorter boiler than before. This procedure is carried out until the tube can only FlG. 272. Machine for Cutting Off Flues. be used for small switching locomotives if it lasts that long after which it is scrapped. By this method, only one length of tube is bought new, which is that required for the longest boilers. In other shops the writer found tubes with many welds, showing that the safe-ending was continued in order to main- tain the same length each time until the tube was worn out, when it was replaced by a new one of the required length. This latter method necessitates buying several lengths new but in localities where the water is not very hard on tubes, it prevents a tube from going to the scrap pile as long as there 326 ELECTRIC WELDING is any good in it. After cutting off the old tubes, as shown in Fig. 272, which represents a common type of machine for this purpose, the tubes are next scarfed, or cut off square, according to which method of welding is to be employed. If a scarf weld is to be used, the old tube is generally ////7////////////////y/w///^ ^Y////////7///////////////. Old Tube Hew End ' 7/////////////////////////S Y////////////////////////// FIG. 273. Ends Prepared for Scarf- Weld. Fie. 274. Bolt Threading Machine Made Into a Scarfing Machine. beveled on the outside at an angle of from 45 to 60 deg., according to the length of scarf desired, about as shown in Fig. 273. The bevel is wholly a matter of personal opinion for just as good welds can be made with a 30-deg. scarf as when one of 60 deg. is used. WELDING BOILER TUBES 327 One type of machine used for scarfing is shown in Fig. 274. This has been rigged up from an old bolt-threading machine. The jaws shown at the left are for gripping the old tube which is then fed into a revolving chuck by means of the handwheel. This chuck contains the necessary cutters for forming the desired bevel on the outside of the tube end. The jaws on the right-hand side of the same machine grip the new short ends as they are fed onto a revolving tapered reamer, which cuts a scarf from the inside. In some shops, the scarfing is done on an old lathe with special fixtures, but the remodeled bolt-threading machine seems to offer the most efficient proposi- tion for, with this type of machine, it is possible for one man to scarf over 60 tubes and ends per hour. Old Tube End FIG. 275. Ends Prepared for a Straight Butt-Weld. If a straight butt-weld is to be made instead of scarfing the ends to be joined, they are cut off squarely, as shown in Fig. 275. This is done in an old pipe-threading machine, or a lathe, so that when placed in the welding machine, the abutting ends will be in contact practically all the way around their circumference. Although this last method of preparing work may sound shorter than scarfing, nevertheless, from actual observation of both methods in different shops, the former is faster by nearly two to one. After preparing the ends for welding, if the tubes have not already been tumbled to remove all scale, which usually leaves the outside surface quite bright and clean, it is necessary to grind the surface of both old tube and new ends back to a distance of about 8 in. in order to secure a good electrical 328 ELECTRIC WELDING contact between the tube metal and the copper jaws of the welding machine. There are three distinct methods of welding boiler tubes, which are called butt-, scarf- and flash-welding, the latter producing the same effect as a scarfed joint when completed. In the straight butt-weld, the ends to be joined are first brought firmly together by means of the pressure device on the welding machine, and the current is then turned on. There is always some point around the circumference of the tube which starts to heat first, due to the impossibility of making the two ends to abut with the same pressure at all points of their contacting surfaces. However, the heat will gradually become uniform all around the circumference before the welding temperature is reached. The current is maintained through the tubes until the joint reaches a good running heat, as evidenced by a " greasy" appearance of the surface, when the pressure is applied sufficiently to push up the hot metal about -J in. which partly completes the weld. The jaws are then released and the tube is immediately thrust onto the mandrel of the rolling apparatus, which is described further on, and the bulge at the joint, caused by the pushing up of the hot metal, is rolled down until the joint is of the same diameter as the original tube. This rolling-down operation, in addition to reducing this bulge of the tube, also forces a complete union of the plastic metal of the two pieces, thereby completing the weld. From this it may be seen that in welding boiler tubes, the welding machine is only used for a heating device to supplant the oil furnace, requiring only sufficient pressure to stick the ends together to hold it while removing work to the rolling machine where the welding is finished. In the scarf weld, the beveled end of the old tube is pushed into the chamfered end of the new piece and the current then turned on the same as in making the butt-weld just described. Due to the "feather" edge of the short new piece, it is often necessary to apply the current intermittently until the joint is well heated all around the circumference; otherwise points of the sharp edge, which come in contact first with the opposite member, will be burned off before the heat is evenly distributed around the tube. Owing to the expanding effect of the scarfed WELDING BOILER TUBES 329 ends, it is not necessary to apply so much pressure as with the butt-weld when the metal is plastic in order to stick the pieces together before rolling down. With either of the above welds, it is necessary to give the old tube more projection beyond the copper clamping jaws than is given the new short piece. This is because the wall thickness of the old tube has been slightly reduced by wearing away in service and if the two parts were given the same projection, the end of old tube would heat much more rapidly than that of the new piece since its resistance to the electric current would be greater, owing to the reduced sectional area. It is always necessary for the heat to form uniformly in each Old Tube New End FIG. 276. Ends Prepared for a Flash- Weld. of the abutting ends or one will burn away before the other reaches the plastic stage. In making a flash-weld, not so much preparation is required as for the two other methods just described; hence it is a much cheaper job and yet, from all tests made so far, it is the only type of joint which is always 100 per cent perfect when considering the number of defective welds in any lot of tubes. The old tube is cut off the right length in a machine, which has a cutting wheel so beveled as to give an angle of 30 deg. from the vertical on the end of the tube, as shown in Fig. 276. The new ends are bought direct from the tube manufacturers with both ends cut square and the surface cleaned well so that there is no preparation needed on the new pieces. After cutting off the. old tube it is only necessary to grind it on the outside about 8 in. back from the end to insure good electrical contact. The old tube is placed in the 330 ELECTRIC WELDING clamps with about 4 in. of projection and the new end with about 3 in. The current is turned on first and the pressure is then applied very slowly and steadily to bring the abutting ends into contact. As soon as they meet, a small arc or ' * flash ' ' is formed which commences to burn away the points of metal coming into contact first. This flashing is continued until the abutting ends are arcing all the way around the circumference and by this time the sharp edge of the old tube, although somewhat burned away itself, has burned its way into the square-cut end of the new piece. A sudden application of more pressure stops the flashing and the joint then quickly attains the running or welding heat as in the butt- or scarf-welding method. The ends are now shoved together and as the current is turned off, the end of the old tube will have forced itself into the end of the new piece sufficiently to form a scarf-weld when rolled down in the rolling machine. Using a Flux. From statements made by every operator interviewed, the use of flux does not help the welding in any way; yet it is used in each shop because it clears up the surface of the metal when the plastic stage is reached and enables the operator to judge the appearance of the heat more easily. The writer is confident that if a new operator were to be broken in on a welding machine, he would soon be able to correctly judge the right welding heat of the metal by its appearance without any flux, as there are many pipe shops using electric-welding machines for making joints in long coils, where flux was never heard of. Each railroad shop uses a slightly different kind of flux, but generally this material is nothing more than a common yellow clay, streaked with quartz formation, which has been pulverized and thoroughly dried out before using. There are several methods and machines employed in the various shops for rolling down and completing the weld after heating the joint properly. One of the simplest machines in use is shown in Fig. 277. It consists of a power-driven mandrel slightly smaller than the internal tube diameter, above which is a power-driven roller. This roller is held a short distance above the mandrel by a spring. When the hot tube is thrust onto the mandrel, the upper roller is brought firmly down onto the outside surface of the joint by pressure on a foot treadle WELDING BOILER TUBES 331 located under the table on which the device is mounted. The pressure is maintained until the joint has been rolled down to outer tube size. The main disadvantage of this style of apparatus is that the speeds of the roller and the mandrel must be in the correct ratio so as to not allow any slip on either inner or outer surface of the tube, otherwise the tube will roll unevenly and when finished will have a thicker wall on one side than on the other. However, this is the earliest form of rolling machine used with the electric-welding method and FIG. 277. Simplest Form of Boiling Machine. is still giving fairly satisfactory service in two well-known shops today. Another type, which is more elaborate but more positive, is a three-roller machine, shown in Fig. 278. The mandrel here is stationary and the three idling rollers, being mounted on a power-driven head, continually revolve around it. After inserting the tube, which is also held stationary, pressure is applied by means of a hand lever which closes the three rollers in toward the center of the mandrel and the joint is rolled down by the surface pressure of the three rollers revolving around it. In order to still further insure uniform rolling, the tube is turned slightly on the mandrel three or four times 332 ELECTRIC WELDING during the rolling operation since the mandrel is slightly smaller than the tube and if the latter were to be held in only one position, a difference in wall thickness on one side might result. Rolling machines of the types just described are sometimes located in direct alignment with the jaws of the welding machine, so that after obtaining the proper heat, it is only necessary to release the jaws and shove the hot tube directly FIG. 278. The Three-Boiler, or Hartz Type, Machine). onto the mandrel. If the three-roller type is being used, the tube is held stationary by locking one jaw of the welding machine. When a new position on the mandrel is desired the jaws are released and the tube allowed to turn slightly with the friction of the revolving rollers. Another method is to have the rolling machine in back of the welding machine so that when the correct heat is obtained, the tube is lifted out of the jaws by the operator's assistant WELDING BOILER TUBES 333 who shoves it onto the rolling mandrel, leaving the operator free to get the next tube lined up in the machine for heating. In this last method, the assistant must act quickly so as not to allow the joint to cool down before the rolling, as he cannot transfer the tube from the welding to the rolling machine as quickly as the operator could shove it forward onto the mandrel as first mentioned. As to speed in welding, the writer observed that the same production could be obtained in different shops by either method of locating the rolling machine ; hence it is purely a matter of space available around the welding machine, and local opinion. A third way of handling the rolling down is to have the rolling machine built onto the welding machine, as shown in Fig. 279. In this particular apparatus, the mandrel is made long enough to permit welding in to a distance of 10 ft. from the joint, so as to reclaim old short tubes by making a new long one with a joint in the middle. This reclaiming of tubes has proved to be perfectly practical, having been forced in one locomotive shop during the war due to the inability to obtain new tube stock. The mandrel is power driven as well as the upper roller, while the two lower rollers are idlers. After obtaining the welding heat, it is only necessary to move the tube about one foot to bring the joint onto the rollers. A clutch at the rear end is then thrown in to revolve the mandrel and upper roller, and pressure is applied through the latter by means of an air cylinder mounted above it. While being rolled the tube is allowed to revolve freely in the open jaws of the welding machine. The rear end of the tube is supported on idling rollers. After the rolling-down process, which is the same as has always been used with the oil-furnace method of welding, the tubes are subjected to the annealing and end-swaging processes. They are then usually tested hydrostatically for possible leaks and stacked away ready for assembling in the boiler. The percentage of leaks is less than 5 per cent in any shop, and in one shop they are so sure of their welding that the tubes are not tested until completely assembled in the boiler when the latter is subjected to a hydrostatic test as a complete unit. This particular shop uses the flash-weld method and has never 334 ELECTRIC WELDING I I be WELDING BOILER TUBES 335 had a defective joint since the welding machine was installed over four years ago. Merits of Electric and Oil Heating. When asked to com- pare the electric welding with the oil-furnace method on boiler tubes of any size, one of the oldest users of the former replied that there was "no comparison." Using oil it was never possible to average over 30 or 40 welds per hour on tubes up to 3 in. with one furnace and one gang. This meant that the tube shop was always behind the rest of the repair depart- ments and working overtime a great deal in order to catch up. Fuel oil will vary greatly in different lots as well as under different atmospheric conditions, so the oil furnace itself is a constant source of aggravation and calls for continual adjust- ing, which means an interruption in production while the fire is regulated. As to production with an electric-welding machine, the average output on tubes up to 3 in. in diameter, taken from all shops using this process, will run 60 completed welds per hour, requiring one operator and a helper at the machine and a third man to prepare the work for welding. In the days of piecework, in some of the shops, records show that the maximum number of small tubes turned out in any shop, with the same number of men, was 125 per hour or a little better than one tube every 30 sec. and this could be kept up for two hours at a time without greatly tiring the men. This speed was obtained by three different shops, each using a different style and arrangement of rolling-down apparatus, which shows that all of the methods outlined previously in this article are equally fast. On welding superheater tubes at the reduced section, where the diameter at the point of weld is about 4f in., the production will run about 10 to 20 welds per hour, although better time has been made on piecework. By comparing these figures with the oil-furnace welding production, even under the best of working conditions, nothing further need be said as to the speed of the electric process. As to cost, there are no figures available later than 1916, which of course would be much lower than at the present day, but by comparing costs of both methods at that time, taking into consideration upkeep, labor, cost of heat either way and 336 ELECTRIC WELDING cost of time lost by making adjustments or repairs to either apparatus, the electric costs per 1,000 tubes welded, is about one-third that of the oil-furnace method. The only wear on the welding machine is the surface of the copper dies or jaws which grip the pieces and this is so slight as to only require smoothing off a few times a week. The machine docs not cost anything for heating energy except when the weld is being made and it is always ready for action as soon as the operator has placed the work in the jaws. Hence there is no delay in starting up the fire in the morning or after lunch hour nor from the fire balking at any time during the welding. The replacements on welding machines in all the shops visited by the writer could be easily covered by $100 during the last six years. In recapitulating the three methods of electric welding flues, it is safe to say that the flash-weld, which produces a scarfed joint when finished, takes the lead for simplicity of preparation, speed of actual welding and reliability as to percentage of failures in any lot of tubes. Next to this comes the straight scarf-weld, which requires machining of the ends before welding but insures a good joint after welding although occasionally a small leak will show up on the first hydrostatic test. As stated before, the per- centage of leaks is very low with this type of weld and practically negligible with the flash-weld. The butt-weld, which was originally employed in all the shops, is now only used in one shop in the whole country, prob- ably due to the difficulty in making a perfect weld each time as compared to the ease of making a scarf weld. However, this one shop claims very high efficiency with a butt-weld, both as to tensile strength, which will average over 85 per cent of original tube section, and as to tightness of the joint under pressure. The principal objection offered by most shops against butt- welding is that should the weld prove tight under pressure, but still be a weak joint mechanically, it might break apart in service. This has happened in a few cases, allowing the tube to drop down in the boiler and subjecting the engine crew to the danger of scalding. With a scarf-weld, which generally shows a tensile strength equal to that of the original WELDING BOILER TUBES 337 tube, due to the area of the weld, should the tube not be welded strongly as just cited and a break should occur inside the boiler, the scarf would prevent the tube from pulling away from its end and only a slow leak could result. This some- times actually happens with oil-furnace welded tubes. The Kind of Machine to Use. As there are different styles and sizes of welding machines being used at the present time on flue-welding, the writer will endeavor to specify special characteristics that should be sought when selecting a machine for this class of work, which is different from any other pipe- welding job. The machine should be constructed to be as efficient electrically as possible ; that is, the clamping jaw should be as close to the transformer as is practical in order not to Copper Jaws Recess*, -"s. / \ Join fl Recess, 1 _., JrLJ A 1 i Contact A i A i ^Contact OldTube' Wew Tube End View Top View FIG. 280. Eecessed Copper Clamping Jaws. have large inductive losses caused by the large gap due to the long secondary leads widely spaced. The fewer the joints between the secondary loop of the transformer and the copper jaws which grip the tube, the less chance will there be for resistance losses that cut down the heating effect gradually as oxides form in the joints or by dirt collecting from allowing them to become loose. Although the jaws should be long to permit thorough water cooling, it is only necessary to grip the pipe over a length of about 2 in. This length is bored out to exactly fit around the tube as shown in Fig. 280. The pressure device does not need to be as heavy as would be used on the same welding machine for joining ordinary pipe or solid stock, since the squeezing together of the plastic metal 338 ELECTRIC WELDING is really done in the rolling machine. For fastest operation the clamping jaws should be operated by air cylinders so that only a slight movement of two valves is necessary to lock or unlock the tube in the jaws. For welding up to 3-in. size tubes, a machine of 30-kw. rating ought to be large enough to stand constant use. Any form of toggle lever or screw-wheel pressure device, which permits the operator to stand close to the work will be suitable, as not over 1,000 Ib. effective pressure is required on this size of work to stick the ends together sufficiently hard for placing in the rolling machine. To handle up to 5f-in. superheater tubes, a machine of about 75-kw. rating should be employed. For its pressure device, an air cylinder or hydraulic apparatus may be used to best advantage so as to secure up to three or four tons' maximum effective pressure. For ordinary butt- or scarf-welding, a hand-operated oil jack may be used, although trouble has been experienced in the past with this type of pressure device due to sticking of the valves at critical times, often spoiling a weld. Flash- Welding. For flash-welding, a toggle lever or hand- screw wheel on small machines and an air cylinder or hydraulic pressure device on large machines must be used, to effect a slow steady forward movement of the movable jaw in order to maintain the arc of the flashing, yet to have available a quick reverse to break the parts away should they stick too soon from too rapid movement of the pressure device. In small shops, it is advisable to install a 75-kw. machine to handle all sizes of tubes up to the largest superheater. If the shop is large enough to keep a small machine busy all the time on tubes up to 3 in., it will no doubt pay to install in addition, a large machine just to handle the superheater tubes as well as any overflow lot of small tubes. While the large machine will handle any size, it is not so rapid in operation on small tubes as the smaller one, and the bulk of flue-welding is on small tubes, less than 10 per cent of the total being represented by the larger sizes for superheaters. WELDING BOILER TUBES 339 WELDING IN THE TOPEKA SHOPS OF THE SANTA FE RAILROAD Supplementing the foregoing, we give the following extract from an article published in the American Machinist, June 8, 1916: In order to give the gripping jaws of the welder good, clean contact the ends of the pieces are ground on the outside for about 6 or 7 in. back from the ends, the operator simply FIG. 281. Close-Up Showing Inside Mandrel. revolving the tube end against the grinding wheel. The ground pieces are sorted out into suitable lengths to form full-length flues when two pieces are butted together, keeping in mind that only two welds are allowed to a flue. The butt-welding machine itself is practically as received, but the inside mandrel and outside rolls, together with the driving mechanism, were added in the shop after considerable experimenting. Without these the method would be a failure. A close-up view of the machine, from the back, is given in Fig. 281. This shows the mandrel A that works inside the 340 ELECTRIC WELDING FIG. 282. Flue Parts Beady for Welding. FiG. 283. Flue Eiids Just Beginning to Heat. WELDING BOILER TUBES 341 FIG. 284. Almost Hot Enough for Welding. FIG. 285. Rolling Out the Upset Metal. 342 ELECTRIC WELDING flue as the outside is rolled between the three rolls after the parts have been heated and butted together. The action of the mandrel and rolls is to take out the upset and give a weld that is smooth on the outside and with very little extra metal inside. The gripping jaws are water-cooled, and the operating air cylinders are plainly shown. Fig. 282 shows two parts of a flue in place in the jaws and illustrates how it is slipped over the mandrel. It will be observed that the mandrel does not extend far enough beyond the rolls to interfere with the welding or become heated from the current passing between the jaws. As it is impossible always to have the two parts to be welded of the same thick- ness, the setting of the pieces in the jaws must be done with judgment. If one piece is thinner than the other and they were both set in the jaws the same distance out, the thin one would burn before the thick one was hot enough to weld properly. To avoid this, a thick and a thin piece are placed about as shown at A and B. In this case the thick one is at A and the thin one at B. As the thick one is in closer to the jaw, it will heat faster. The thin one, being set out farther, gives practically the same amount of metal for the current to heat. The result is an even heating and a perfect weld. Fig. 283 shows two pieces the reverse of the ones just shown. As the work gradually heats, it looks as in Fig. 284. At the proper heat, the operator butts the work together to form the weld, which leaves a considerable amount of upset. He then shoves the tube along over the mandrel until the weld is be- tween the rolls, when he throws in the clutch and brings down the upper roll. The work spins between the rolls, as shown in Fig. 285 and the result looks almost like a new tube. CHAPTER XV ELECTRIC WELDING OF HIGH-SPEED STEEL AND STELLITE IN TOOL MANUFACTURE The cost of solid high-speed cutting tools is high. At the same time their remarkable cutting qualities make them a necessity in up-to-date shop practice. The electric process of butt-welding has made it possible to obtain all the advantages of a solid high-speed cutting tool and yet at a cost that is not a great deal higher than the ordinary tool-steel product. Stellite, -which has recently become more widely known, has been rather limited in its use owing to the fact that it cannot be machined, and it has been thought .by many that it could not be successfully joined to any other metal for holding it. This has limited its use to special forms of toolholders, which are often very clumsy in getting into difficult corners on special shapes. The electric process of butt-welding has made it possible to join Stellite bits of any common size and shape to a shank of ordinary steel, giving all the advantages of a solid cutting tool and yet employing only a small amount of the Stellite metal just where it is needed for cutting. The Thomson welding process consists of passing a large volume of electric current at a low pressure through the joint made by butting two pieces of metal together. The electrical resistance of the metals at the contacting surface is so great that they soon become heated to a welding temperature. Pres- sure is then applied mechanically and -the current turned off, thereby producing a weld. The metal is in full view of the operator at all times instead of being hidden by the coal of a forge or by flame in an oil furnace. No smoked glasses or goggles are required any more than would be if welding by the forge method. Due to the way the metal is forced together there is no oxidation such as there would be in an open fire and therefore no welding compound is ordinarily required. 343 344 ELECTRIC WELDING It is this feature alone which makes it possible to weld high- speed steel and Stellite, the former being very difficult to weld by the forge method and the latter practically impossible. "With this process of electric welding the heat is first developed in the interior of the metal. Consequently, it is welded there as perfectly as at the surface. When welding with other methods, however, the outer surface is heated first and very often the interior part does not reach welding heat, the result being an imperfect weld. There is no blistering or burning of the stock when welding electrically, whereas it certainly requires a very expert welder indeed to secure the proper heat on high-speed steel in a forge fire without burning at some point. The process is the most economical known, due to the fact that no energy in the form of heat is being wasted in heating more of the material than is required to make a weld and as soon as it has been completed the current is turned off so that the machine then is not using up any energy what- ever. The operator has complete control of the current at all times so that he can obtain any color desired on the metals, where are always visible, and waste by accidental burning of metal is reduced to a minimum. The only preparation of stock necessary for welding by this process is that when very rusty or greasy it should be thor- oughly cleaned, as the presence of either rust or heavy grease affords poor contact with the copper clamping jaws, retarding the flow of electricity and seriously reducing the heating effect. It is often asked if the electric current has any effect on the welded metal. This question arises from the fear that there may be some mysterious condition connected with electricity that will change the characteristics of the metal, particularly of high-speed steel or Stellite. The answer is, of course, in the negative, as the only effect of the electric current is to heat the metals being welded. The rapidity of work will depend largely on the operator, the size and shape of the pieces to be welded and the size of machine being used, as there is a wide range in welding time between heavy pieces requiring careful alignment in the clamp- ing jaws and light pieces which can be rapidly and easily handled. Welding High-Speed to Low-Carbon Steel. In tool welding ELECTRIC WELDING OF HIGH-SPEED STEEL 345 there are various kinds of welds to be made, which require different designs of holding jaws and often two distinct types of welding machine. Three butt-welding machines shown in Figs. 286, 287, and 288 are especially suitable for welding drills, reamers or other FIG. 286. Thomson 10-A6 Butt -Welding Machine. tools that can be made up of a combination of high-speed and low-carbon steel. The machine shown in Fig. 286, known as the 10-A6 machine, will weld iron or steel rods from J to J in. in diameter, or an equivalent cross-section in squares, rectangles or flats. An operator can make from 50 to 200 welds per hour, "'.cording to the size and nature of the work being handled. 346 ELECTRIC WELDING The clamps are of the horizontal operating type, adjustable for different sizes of stock as well as for horizontal alignment of the work. A close-up view of the left-hand clamping mechanism is shown in Fig. 287. The jaw blocks are water cooled and have a maximum movement of 1J in. by means of the hand-operated clamping levers. There is also a possible f-in. adjustment of both front and rear jaw blocks. Stops are provided for backing up the work. There are four copper jaws to a set, two being used on each clamp. These jaws are FIG. 287. Closeup View of Left-Hand Clamp. 2y z in. square by 1 7 / 1C in. thick. The pressure device for forcing the heated ends of the work together is a hand-lever- operated toggle movement, which enables the operator to "feel" his work. This toggle device gives a movement of 1 in. to the right-hand jaw. The maximum space possible between the jaws is 3J in. There is an automatic current cutoff mounted on the machine. The standard windings are for 220, 440 and 550 volt, 60-cycle alternating current. The current variation for different sized stock is effected through a five-point switch ELECTRIC WELDING OF HIGH-SPEED STEEL 347 mounted on the machine. Standard ratings are 15 kw., or 25. k.v.a., with 60 per cent, power factor. This size of machine covers a floor space 43X57 in., is 65 in. high and weighs about 1100 pounds. The machine shown in Fig. 288, or the No. 6 machine, is for heavier work, its capacity being from to 1 in. in diameter on iron or steel rods, or the equivalent in other shapes. Its production is from 50 to 125 welds per hour. The maximum jaw opening is 3 in. ; the four jaws are of hard-drawn copper, 2JX2f in. and 1 in. thick; toggle-lever movement 1^ in.; FIG. 288. No. 6 Butt-Welding Machine. maximum space between jaws, 4 in.; current standards are the same as for the previous machine. There are 10 points of current variation for different sized stock, effected through double-control switches mounted on the machine. Standard ratings are 30 kw. or 45 kva., with 60 per cent, power factor. The jaws are air cooled, but the copper slides to which the jaws are bolted, as well as the secondary copper casting of the transformer, are water cooled. It occupies a floor space 22X44 in. and the height to center line of the jaws is 37^ in. The weight is 3100 Ib. Its operation is practically the same as the first machine described. Another machine of very similar characteristics is shown 348 ELECTRIC WELDING in Fig. 289. This is known as the Special 5-D machine and is intended for the use of makers of small taps and twist drills up to f in. in diameter. It has very accurate adjustments on FIG. 289. Special 5-D Machine. the clamps and special jaws with steel inserts to prevent wear. To use these, however, requires that the pieces to be welded must be finished to uniform size so as to accurately fit the jaws in order to conduct the current properly. FIG. 290. Stellite-Tipped Roughing Drills. The machines shown in Figs. 286 and 288 are not only good for welding the steels mentioned, but also for Stellite work, samples of which are shown in Fig.- 290, since the com- ELECTRIC WELDING OF HIGH-SPEED STEEL 349 monly used bits of this metal are within their range. The hand-lever toggle action is quicker and is better suited to this work than the hydraulic-pressure device used on some of the larger machines. In welding twist drill or reamer blanks, such as shown in Fig. 291, not over } in. in diameter, it has been found practical FIG. 291. Twist-Drill Blanks Just Welded. to use a pair of jaws on each side that will handle all work from the smallest up to the J-in. size. These jaws are made as shown in Fig. 292. The two rear, or movable, jaws on each side of the machine are flat faced, while the front, or stationary, jaws, have a V-groove cut in them just deep enough to give clearance for the smallest size of stock to be handled in contact Round Stock being welded MOVABLE: DIE STATIOtMRY DIE Section Through Dies and Work FIG. 292. Copper Jaws for Various Sizes. with the face of the opposite jaw. The work is held in the jaws with a three-point contact, which has been found to be sufficient for stock of this size, although it is not to be recom- mended for larger work, since not enough current could be carried into the pieces without applying pressure sufficient to squeeze the work into the surface of the copper jaws. This would soon spoil all accuracy of alignment of the V-grooves. 350 ELECTRIC WELDING In this connection it may be well to mention that a welding machine is not a micrometer and the welding of finished pieces is not recommended in commercial production, although such welding is done right along for special jobs. By "special jobs" is meant the putting on of an extension to a drill, tap or small reamer and the like. In welding high-speed to low-carbon steel the low-carbon steel sliould project approximately twice as far out from the jaws as the high-speed steel does in order to equalize as much as possible the heating of the two pieces. Where a tool is to be made with a head larger than the shank, as shown at A, Fig. 293, holding copper jaws should HIGH- SPEED STEEL I WELD J LOW-CARBON STEEL (A) LOW CARBON STEEL (a) JAWS OFWELDER,GROOVED /^, FOR ROUND STOCK ( D j End View FIG. 293. Copper Jaws for Holding Large Heads and Small Shanks. be made as shown at D. In work of this kind the dimension B should always be about one-half of the diameter of C. The same rule holds good with this type of tool blank when placing it in the jaws as with steel of the same relative size ; that is, the low-carbon steel should project about twice as far from the jaws as the high-speed steel since the high-speed steel has the higher resistance and lias a tendency to become plastic sooner. To still further reduce its tendency to heat up quickly, the resistance should be reduced as much as possible by having the jaws as good a fit for the high-speed piece as it is possible to make them. Where different sizes are to be welded it is advisable to have special holding jaws for each separate size of high-speed steel head, although the low-carbon steel pieces may be held in V-grooved jaws made up to hold several sizes. ELECTRIC WELDING OF HIGH-SPEED STEEL 351 352 ELECTRIC WELDING This is the practice of some of the largest makers of reamers and large drills. The actual use of the machines shown for the work outlined is simplicity itself. The work is placed in the respective jaws and securely locked in place by pulling forward the two levers shown projecting upward on each machine. In addition to the grip of the jaws the work is kept from any possible slip by means of stops against which the outer ends of the work are butted. With the work solidly in place the operator pulls FIG. 295. Close-up of Machine with Work iu Jaws. on the pressure lever at the right of the machine until the ends of the work are in firm contact. He then turns on the current by means of a push button conveniently located in the pressure lever, and when the proper heat is reached, which is judged by the color, the push button is released. This shuts off the current and the operator then applies full pressure and the weld is made. The maximum capacity of the largest of the three machines described is 1 in. round or its equivalent in other shapes. For larger work a machine similar to the one shown in Fig. 294 ELECTRIC WELDING OF HIGH-SPEED STEEL 353 is used. This is known as a No. 9 butt-welding machine, and its capacity is from to 1J in.; the output is from 50 to 100 welds per hour ; the maximum jaw opening is 1 J in. ; the four hard-drawn copper jaws are 3 in. high, 3J in. wide and 1 in. thick; the pressure device is a 5-ton hand-operated hydraulic oil jack ; maximum movement with jack, 2 in. ; maximum move- ment with one stroke of jack, ^ in. ; maximum opening between jaws, 4 in. ; standard windings the same as for the previous machines; standard ratings, 40 kw. or 55 kva., with 60 per FIG. 296. Steps in the Making of a Large Reamer. cent, power factor ; width of machine, 27 in. ; length, 60 in. ; height, 46 in. ; weight, 3900 pounds. A closeup of this machine, with a large reamer blank in the jaws, is shown in Fig. 295, and progressive steps in the making of the reamer are shown in Fig. 296. The high-speed steel piece is 3 in. long by If in. diameter, and the machine- steel piece is 6 in. long. Two other machines (10-B and 40- A2 models) of this type suitable for heavy tool welding may be mentioned. They are made with a capacity of from to 1J and from 1 to 2 in. 354 ELECTRIC WELDING The first of these has a hand-operated pressure device capable of exerting a pressure of 12 tons and it weighs 7800 Ib. The second has a pressure device which receives its initial pressure FIG. 297. A Welded and a Finished Lathe Tool. from an external accumulator, which gives an effective pres- sure of 23 tons; it weighs 8000 Ib. and is 64X105X48 in. high. The Welding of Other Than Round Tools. The welding .WELD r- \HIGH-SPEED \ STEEL LOW-CARBON STEEL FlG. 298. How the Parts Are Arranged for Welding. of tools similar to the ones shown in Fig. 297, intended for lathe or planing-machine tools, may be done in any of the foregoing machines. The cutting parts may be of either Stellite End View (a)D/ BLOCKS or WELDER FIG. 299. How the Parts Are Clamped in the Jaws. or high-speed steel. This kind of welding is usually employed by manufacturing concerns in their own toolrooms in order to use up odd bits of high-priced steel or Stellite. The pieces arc ELECTRIC WELDING OF HIGH-SPEED STEEL 355 prepared about as shown in Fig. 298. Jaws for holding work of this kind are outlined in Fig. 299. Another way to make tools for lathe or planing-machine work is outlined in Fig. 300. This method may often be employed when the one just given could not. As can be seen, HIGHSPEED STEEL WELD- LOW CARBON STEEL FIG. 300. Method of Preparing for an Insert Weld. in order to properly support the high-speed steel piece, the low-carbon steel shank is milled away to form a recess for the reception of the high-speed steel bit. The welding can be done on any of the machines shown provided the parts are not of too great cross-section. The method of recessing the copper clamping jaws is clearly shown in Fig. 301. j Co) iws recessed to hold pieces a a a 1& r <-\ ,-> Jtt-: ICAK ?ED ' ST T SH, j >BON- r .EL W/t a Top View of Work Meld Vertically BIT Ph he i yon sss Pieces resting on bottom of recess Front View of Rear Jaws and Work FIG. 301. Jaws Used for Holding Work in Insert Welding. The perfect success of a welded high-speed tool depends not only on the correct welding but also upon the correct treatment after the welding itself has been accomplished. It is easily seen that if a piece of high-speed steel is welded to a piece of ordinary carbon steel and the joint allowed to cool 356 ELECTRIC WELDING fairly quickly in the air strains will be set up at the joint for the reason that the high-speed steel in cooling so quickly, both metals become hardened more or less but to a different degree. Hence if the weld is subjected to any great strain under these conditions it will break either at the joint or close by , due to the strain. It is therefore very evident that immediately after welding a piece of high-speed steel to carbon steel the work should be immediately put into some sort of furnace to be annealed. The amount of time that the tools should be left in the furnace for thoroughly heating through and the amount of time required to allow the pieces to cool down to room temperature depend entirely upon the size and Sfa tionaryja ws, only recesseq / * "A _______ _ _ 1 -i 1 HIGH- . SPEED BIT CARBON f STEEL "' SHANK Top View of WorK Held Horizontally Piece resting on bottom ofrec&ss V T End View of WorK in Right-Hand Jaws FIG. 302. Jaws Used for Stellite Butt Welding. character of tool being made. However, the annealing of any piece of any size requires that' the work be left in the furnace heated to at least a dull cherry red for a few hours and allowed to cool very slowly in the furnace. If a welded tool is not properly annealed before machining much difficulty is often experienced from hard spots being encountered in the machining of the pieces, which of course is more or less disastrous to the cutting edges of the tools being used in the machining process. The best method of hardening high-speed steel tools after the welding and machining depends also greatly upon the shape and size. Welding Stellite, Although the welding of the various ELECTRIC WELDING OF HIGH-SPEED STEEL 357 grades of Stellite is not difficult there is a certain knack in the welding and also in the clamping of the stock which must be fully acquired to produce satisfactory results. The welding should be done in a horizontal butt-welding machine with a quick-acting hand-lever pressure device. In butt-welding round drill stock or rectangular tool stock the pieces should be held as shown in Fig. 302. It will be noticed that the projection of the Stellite beyond the copper jaws is very short indeed while the projection of the carbon-steel J aws re<. hold pn essec/ >ces. \ to j , HIGH---' SPEED BIT CARBON- STEEL SHANK Top View of Work held Vertically SHANK ,,r V- r Pieces resi on bottot recess mg n of Front View of Rear Jaws and Work FIG. 303. Jaws Used for Stellite Insert Welding. piece is comparatively long. This is because Stellite has a very high resistance compared with the carbon steel. Since in this work the heating effect varies directly with the resist- ance of two metals the heating in the Stellite should be retarded as much as possible by surrounding it almost completely with the copper jaws. The correct amount of projection of the carbon steel will have to be determined by experiment in each case after observing with each setting of two pieces which has the tendency to heat the fastest. In welding in cutting bits of Stellite by the insert-weld method the pieces should be held as shown in Fig. 303. 358 ELECTRIC WELDING It will be seen from this cut that the copper jaws holding the small bit nearly surround it and at the same time back up the piece to take the pressure of the squeezing up of the FIG. 304. Vertical Type of Welding Machine. stock. The opposite jaws holding the carbon-steel shank do not have to grip very much of the metal but they serve to back it up to receive the force of the pressure. In the welding itself the current is applied intermittently, ELECTRIC WELDING OF HIGH-SPEED STEEL 359 as the Stellite usually has a tendency to heat very rapidly, until the carbon steel is fast approaching the plastic state. The current is then held on steadily and the instant the Stellite metal "runs," the pressure lever is given a quick jerk as thje current is turned off. It will be found that with a good weld there is scarcely any push up of the stock and very little of the FIG. 305. Making a "Mash" Insert Weld in a 20-AV Machine. metal flows out at the joint, requiring little grinding, if any, to finish the tool. Unlike high-speed steel Stellite requires no further heat treatment or attention of any kind if it is welded correctly. When it is taken out of the welding machine the tool is ready for use at once after grinding off the resulting burr. 360 ELECTRIC WELDING Where large numbers of tools of the lathe and planing- machine types are to be made, such as shown in Fig. 300, the highest production can be obtained by using a Vertical FIG. 306. Large 40-AV Vertical Machine. type of welding machine built on the lines of the one shown in Fig. 304. This machine (10-AV model) has a capacity of two pieces with contact areas between 0.40 and 0.30 sq. in. for pieces with a total thickness of f to 1J in. The production is 35 to 85 tools per hour, depending on the size; the upper and lower ELECTRIC WELDING OF HIGH-SPEED STEEL 361 jaws are of hard-drawn copper 1|X2J in. and If in. thick; the jaw blocks are water cooled; the machine has a current variation through a five-point " switch for different sizes of stock; standard windings are for alternating current 220 440 and 550 volt, 60 cycles; standard ratings, 15 kw. or 25 kva. with power factor of 60 per cent. ; the pressure device is hand operated, giving a movement of 2f in. ; maximum space between jaws, 3| in.; floor space occupied, 21 X 53 in.; height, 75 in.: weight, 1200 pounds. A larger machine (20-AV model) of the same type in opera- tion is shown in Fig. 305. This machine gives a maximum area of contact ranging from 1| to 1 sq. in. on pieces with a total thickness from 1 up to 2 in. ; production is from 50 to 75 welds per hour ; there is a throat clearance of 10 in. ; the copper jaws are 2X3 in. and 1| in. thick; pressure is by hand-toggle FIG. 307. Jaws and Work Arranged for a "Mash" Weld. lever and spring cushion; current control, as in the other machines, is by push button in the lever operating through a magnetic wall switch; the jaw blocks are water cooled; standard ratings are 30 kw. or 50 kva'. with 60 per cent, power factor; weight, 2200 pounds. Another still larger machine (40-AV model) is shown in Fig. 306. Except for its size it is but little different from the two just described, the main difference being the hydraulic- pressure device, which gives an effective pressure of 5 tons. This machine has a maximum contact area of 3 sq. in. and will weld pieces from 1-J.to 3 in. total thickness; production, 15 to 50 welds; throat depth, 6J in.; jaws, 2X4XH in. thick; maximum movement of upper jaw block, 2 in. ; movement with one stroke of lever, f in. ; space possible between jaws, 3 in. ; standard ratings, 60 kw. or 86 kva. with 70 per cent. 362 ELECTRIC WELDING I I ERj -i.-2 si s 11 si t .s 3 g 1 rod, o s = S'B i oooooooooooooo o wo 00*0000000000000000000 O) <-iO) GOQ4 0*O OC01>'-HO> 0* CO CO TJ rf< O <& O5 t~ t>W5lO*$l ^(XCHO Oi O I-H i-< (N <^cowiS<5coaoooeocoooojcoj ' rn' ^H rH i-< i-H 0< QOOi " CO i i GO *o t- 05 1> o co t s, I 3 ; s-S i ELECTRIC WELDING OF HIGH-SPEED STEEL 363 Size of Swi and Fus.e Size of Switch and Fuses .2 ze of and K.V.A. emand v cx a a a '6 S S S S S S o o o o o o Mi O O M* J S c/i c/i c/i' c/5 c/i c/i c/ ^ <^5 08 P3 PQ pQ M 00 00 O 364 ELECTRIC WELDING power factor; size, 34X60 in. by 79 in. high; weight, 3600 pounds. For welding tools on these machines the relative thickness of the two parts should be about that shown in Fig. 307. Under ordinary conditions the dimension A should be about one-third of B in order to have the point of the weld nearest the jaw in contact with the high-speed steel, so that the heating effect FIG. 308. Pieces Grooved to Make Better Welds with Less Current. will be lessened and its fusion point retarded until the low- carbon steel has a chance to heat up properly. In order to obtain the best results tools wider than 1 in. and with a recess longer than 1J in. should be grooved as shown in Fig. 308. This reduces the section in actual contact, thereby requiring less current, is easier and quicker to heat and assures a better weld over the entire area of contact. In order to assist those who have tool or other butt-welding to do some useful data are given in Table XXVI. In Table XXVII is given the proper size of copper wire to use to connect up the various machines mentioned for tool welding. CHAPTER XVI ELECTRIC SEAM WELDING Seam or line welding is the process of joining two over- lapping edges of sheet metal for their entire length without the application of any solder or spelter along the joint. In the Thomson process of lap-seam welding, the heat is produced by passing a large volume of electric current through the edges to be welded by means of a copper roller on one side of the joint and a copper track or horn underneath. In any electrical path, wherever high resistance is interposed, heating will result, and the higher the resistance to the current, the greater will be the heating effect. In the electric lap seam welding machines, the copper roller and horn are good con- ductors and the joint between the edges of the metal to be welded is the point of highest resistance. On this account it is evident that the greatest heating effect will be at that point. As the roller passes over the joint, heating the stock to a plastic state beneath it, pressure is applied by springs on the roller which forces the two edges together as fast as they are heated. Since 20 B. & S. gage or lighter metal heats very rapidly, the pressure and heating can be effected at the same instant of contact by the roller, and it is possible to weld as fast as 6 in. per second. The only preparation necessary for seam welding is that the stock must be absolutely clean, that is, free from any traces of rust, scale, grease, or dirt, if a tight, well-appearing joint is desired. If it is not necessary for the joint to be tight, it will not be necessary to have the stock so clean, although heavy scale or rust will obstruct the passage of current, so that little or no heating effect can be secured under these conditions. In welding sheet brass of 22 to 30 B. & S. gage, to secure a perfect joint the metal should be carefully pickled and washed to remove all traces of grease and tarnish which tend to prevent 365 366 ELECTRIC WELDING the passage of current across the joint of the edges. The metal should be welded soon after pickling, as, no matter how carefully it may be washed, oxidation is always sure to start very shortly after the brass has been removed from the pickling acid. Steel, to be successfully seam welded, should not have a carbon content of over 0.15 per cent., for a higher carbon steel than this has a tendency to crystallize at the point of weld, due to the rapid cooling of the welded portion from the sur- rounding cold meta! After welding, the joint will be found to be about one-third thicker than the single thickness of the metal. It is possible, by applying more pressure, to reduce this finished thickness still more, but it wears more on the copper roller to do so. In welding brass, a soft, annealed metal should be used, for although hard-rolled brass can be welded, it does not force the two edges together very much and the finished joint under these conditions is almost twice the original metal thickness. However, with a soft, annealed brass the finished joint will be not over a third greater than the single metal thickness, and by applying sufficient pressure can be reduced down to be not over 10 per cent, thicker. The principal advantage of electric seam welding is that no spelter and no flux are required, the metal itself forming its own cohesive properties, which allows great speed in produc- tion. The greatest efficiency of a seam welding machine lies not only in its welding qualities but in the use of a suitable jig to properly hold the work. The jig used should be made so as to enable the operator to place or remove the work in the shortest possible time, since the welding itself is very fast compared with any other known method of making a con- tinuous joint. In order that their seam welding machines may operate in every installation with the highest efficiency possible, the Thom- son Electric Welding Co., Lynn, Mass., build them standard only up to a certain point and then design a special holding jig to best fit the work to be done in each individual case. The amount of lap allowed in making lap seam welds is usually about twice the single sheet thickness of the metal. The operation of a lap scam welding machine is very sim- ELECTRIC SEAM WELDING 367 pie, once the machine is set for any given piece of work for which a special jig has been built. After placing the piece in the jig and securely locking it there, the operator depresses a foot-treadle which throws in a clutch and starts the copper roller across the work. By the proper setting of adjustable control-stops on the control-rod at the top of the machine, the current is automatically turned on as the roller contacts FIG. 309. Model 306 Lap Seam Welding Machine. with the overlapping edges of the piece to be welded and is automatically turned off when the roller reaches the end of its stroke; another stop reverses the travel of the roller and brings it back to the starting position. The control-stops may be adjusted to turn the current on or off at any point along the stroke of the roller for doing work with a seam shorter than the maximum capacity of the machine. The roller stroke may be also shortened so that the complete cycle of operation 368 ELECTRIC WELDING will be accomplished in the shortest space of time on seams shorter than maximum seam capacity of any machine. In order to keep the copper roller from overheating in action, water is introduced through its bronze bearings on each side. This same water circulation, also passes through the under copper horn or mandrel and then through the cast-copper secondary of the transformer, so that the machine can be operated continually, 24 hours per day if desired, without overheating. Lap Seam Welding Machines. The lap seam welding FIG. 310. Details of Welding Roller Head. machine, known as Model 306, shown in Fig. 309 will weld a seam 6 in. long in soft iron or steel stock up to 20 gage in thickness, or brass and zinc up to 24 gage thick. This machine will make from 60 to 600 welds per hour, depending on the nature of the work and the quickness with which the pieces can be placed in and removed from the jig. The copper horn is water-cooled and has an inserted copper track on which the work rests. The upper contact consists of a copper roller 6^ in. in diameter, mounted on a knockout shaft sup- ELECTRIC SEAM WELDING 369 ported in water-cooled bearings. Pressure is exerted on the copper roller by means of a series of springs on each side which are adjustable to give the proper tension for various thicknesses of stock. Current control is automatic through a magnetic wall switch carrying the main current. The latter is controlled from a mechanical switch which is thrown in or out by the action of the roller-carrying mechanism as it starts FIG. 311. Thomson No. 318 Lap Seam Welding Machine. and completes the stroke for which it is set. Standard wind- ings are for 220-, 440-, and 550-volt, 60-cycle, alternating cur- rent. Current variation for different thicknesses and kinds of stock, is effected through a regulator which gives 50 points of voltage regulation. A variable-speed J-hp. motor gives a wide variation in the speed with which the roller may be fed over the work. The standard ratings for the machine are 15 kw. or 25 kva., with 60 per cent, power factor. This 370 ELECTRIC WELDING machine covers 32X96 in. floor space, is 68 in. high and weighs 2750 Ib. A close-up view of the type of roller-carrying head used on all the lap seam welding machines, is shown in Fig. 310. In this view the roller is shown operating between the clamping bars of a special holding jig on the horn. As the roller itself occasionally requires smoothing off around its contacting sur- face, its bearing has been designed to knock out quickly so FIG. 312. Large Size, No. 324, Lap Seam Welding Machine. that removal and replacement of the roller is very simple and easy to accomplish. The cleaner the stock being welded is kept, the longer a roller will operate without requiring smooth- ing off, as dirt and scale on the stock cause a slight sparking as the roller passes along, which tends to pit up its contact surface. The machine shown in Fig. 311, known as Model 318, is a larger and heavier machine than the one previously described ELECTRIC SEAM WELDING 371 and will weld a lap scam 18 in. long on the same gages of metal quoted. Another very similar but smaller machine (Model 312) is also made for welding seams up to 12 in. In Fig. 312 is seen a considerably larger machine, Model 324, capable of welding a lap seam up to 24 in. in length. The production is from 30 to 120 welds per hour. The machine covers a floor space of 36X90 in., is 72 in. high, and weighs 3500 Ib. All other specifications are the same as given for Fig. 309. Examples of Holding 1 Jigs. The machines shown may be fitted with numerous forms of holding jigs from the simple FlG. 313. Oil Stove Burner Tubes Before and After Welding. bar clamps shown on the horns in Figs. 311 and 312, to various more complicated forms, some of which may be mounted on the knee below the horn or bolted direct to the face of the machine column. The small oil stove burner tubes shown in Fig. 313 lend themselves nicely to the seam welding process. Cylindrical pieces such as the shell tubes for automobile mufflers shown in Fig. 314, need a rather elaborate holding jig. A machine fitted up for this work is shown in Fig. 315. To insert a muffler shell into this jig the hinged end is swung outward and downward; the two halves of the holder are spread apart by pressing down on the left-handle treadle; the shell is then 372 ELECTRIC WELDING thrust into the holder; the treadle is released, which allows the holder sides to be pressed in by the springs and hug the muffler shell around the horn of the machine, with the edges overlapping enough for the weld; the end gate is then closed and the welding roller started over the seam. The principal function of the gate is to hold the muffler shell square in the jig and prevent it behig pushed out by the welding roller. FIG. 314. Seam Welded Automobile Muffler Tubes. A jig for holding large cans is shown in Fig. 316. The side clamps of this jig are operated by means of the lever shown at the left. An end gate, shown open, is used in the same way as in the muffler shell jig. Work of this kind is of course much slower than with a smaller jig, yet it is faster than by any other process of closing the scams. ELECTRIC SEAM WELDING 373 Bucket bodies are held as shown in Fig. 317. The holding jig is made to slide in a channel bolted to the machine knee. The jig is slid back clear of the horn and, with the gate in the flaring end open, the bucket blank is inserted. The gate FIG. 315. Holding Jig for Automobile Muffler Tubes. is then closed by means of the handle, the jig and work is pushed over the horn to a stop, and the weld is made as usual. Another application of seam welding, is to use it for welding the ends of strip stock together, end to end, so as to facilitate continuous passage of the strip through the dies of a punch press. A machine fitted up for this work is shown in Fig. 318. 374 ELECTRIC WELDING The ends of the two strips to be welded are inserted in the jig from opposite sides and the edges brought together. The pieces are then clamped by means of the two levers shown in front of the jig, which operate eccentrics over the clamping FIG. 316. Holding Jig for Large Sheet Metal Cans. plates. The welding roller is then run over the ends as in other work of this kind. Flange seam welding differs from lap seam welding in that instead of the metal being lapped a slight fin or flange is formed along the edges of the metal parts, the flanges being welded together and practically eliminated in the process. This ELECTRIC SEAM WELDING 375 Class of welding is especially adapted to the manufacture of light gage coffee and teapots spouts or similar work. A machine built especially for flange seam welding, known FIG. 317. Jig for Holding Bucket Bodies. as Model 26, is shown in Fig. 319. The work being done is the welding of the two halves of teapot spouts. In the operation the two halves of the spout are clamped securely in a special copper jig, Fig. 320, which has been carefully hand-cut to 376 ELECTRIC WELDING fit the halves of the spout perfectly on the entire contacting area. The jig is pushed around on the flat copper table, which constitutes the top of the welding machine, so that the seam of the edge to be welded is allowed to ride along the small FIG. 318. Jig for Welding Ends of Metal Strips Together. power-driven copper roller which is mounted on a vertical shaft, as illustrated in Fig. 321. The halves which are welded by this process must be blanked out by special steel dies to give the correct amount of fin or flange on each edge. This ELECTRIC SEAM WELDING 377 fin is heated to the plastic stage by contact with the roller and the slight pressure applied not only forces the metal of the two fins to cohere but also forces the projection into a level with the outer surface of the spout, thus giving a finished job direct from the welder which is smooth enough without FIG. 319. Machine for Flange Seam Welding. any grinding to be ready for the enamelling or agate-coating process. The secret of success of this work lies wholly in the proper preparation of not only the copper holding-dies, but also the steel flanging and forming dies. A finished spout, just as it 378 ELECTRIC WELDING ga-. a. w a, s Jig -pig JO S{BCDl3aQ sS oooot^t*'otou> o* s 0-c CO CO < c PS o 6 fc S3 o .5 5 i a ~ ^O G^ 00 't' &< CO CO CO CO 66666 380 ELECTRIC WELDING FIG. 320. Jig for Holding Teapot Spouts for Welding. FIG. 321. Diagram of Flange Seam Welding Operation. FIG. 322. A Finish Welded Teapot Spout. ELECTRIC SEAM WELDING 381 comes from the welding machine, is shown in Fig. 322. The welded seam is barely visible. In order to assist those who have welding jobs to do, to calculate the current cost on various jobs, Table XXVIII is given. This table shows the approximate current consumption, and multiplying the rate given by the local rate charged, the cost of 1000 welds can be easily ascertained. Table XXIX is very convenient for ascertaining the size of copper wire needed to connect the different machines men- tioned to the main source of current supply. CHAPTER XVII MAKING PROPER RATES FOR ELECTRIC WELDING AND THE STRENGTH OF WELDS The uncertainty which seems to exist regarding electric welding rates among central-station interests, says S. I. Oesterreicher in Electrical World, is no doubt due to the indif- ference of the welding industry, which during a long period in the past did not assist those affected by the rates as much as its unquestionable duty would have suggested. While welding installations of only comparatively small sizes had to be considered say from 25 to 100 kva. no great harm was done by such tactics to either interest. However, with the installation of large equipments and the operation of large unit welding machines, central stations suddenly experienced disturbances upon their lines and in their stations, which were anticipated but partly and were blamed entirely upon the welding equipment. Thus, to protect themselves, central-station interests launched into a partially retroactive policy, greatly to the detriment of the welding industry as a whole. Since welding installations of several thousand kva. capacity are not unusual, it is proper that all points of doubt should be considered as broadly and fairly as possible, and a far- reaching co-operative policy inaugurated. The revenue from such large installations may easily reach several thousand dollars a month. It is therefore obvious that, from a purely commercial standpoint, a welding load is a very desirable constant source of income to the central station. Looking at the reverse side, it should be recalled that cen- tral-station engineers, on account of past sad experiences, had jumped to the following conclusions : 1. That a welding installation is a very unreliable metering proposition. 382 MAKING PROPER RATES FOR ELECTRIC WELDING 383 2. That it has a poor load factor. 3. It has a constantly fluctuating load varying between extreme limits, and 4. It has a bad power factors The first important point is, no doubt, the metering. The time-honored opinion on one side that, due to the short period involved, an integrating wattmeter does not respond quickly enough, is contradicted by the claim on the other side that the deceleration of the meter disk compensates for the lagging acceleration. As far as the writer is aware, not the slightest positive proof has been offered to support either contention. Considering for instance a 200-volt, 300-amp., single-phase, two-wire wattmeter, whose disk at full load makes 25 r.p.m., and assuming -the total energy consumption to be integrated within 0.2 second, it will be found that to register correctly the meter disk has to travel about 0.08 of a revolution. It is scarcely possible that by merely looking upon a meter disk any one could guess within 100 per cent the actual travel during such a short time interval. A stop watch will scarcely be of any assistance; neither will a cycle recorder with an ammeter and voltmeter check be of any value, since no instru- ment is of such absolute dead beat as to come to rest from no load to full load within 0.2 second. Such methods therefore are of no value in ascertaining the behavior of a wattmeter under sudden intermittent heavy loads. The next step of the metering proposition was to take the rated energy consumption of the welding machine as given by the manufacturer, assume a certain load factor, calculate from these data the energy consumption, correct for the power factor and check the answer periodically on the meter dial. The result obtained on the meter was usually a constantly varying, lower energy consumption than calculated, and no doubt this was the cause of the great distrust of the meter. This method is worse than no check at all, and it is so for the following reasons : 1. The energy consumption at a welder depends upon the welding area of the metal, but is not a proportionate variable. That is, all other factors being the same, two square inches of a certain weld do not consume twice as much energy as one square inch does. Fig. 323 shows this fact plainly. It 384 ELECTRIC WELDING is also of common knowledge that on a spot welder the area of the weld varies from weld to weld just as much as the electrode contact area does. Assuming an electrode at the start as Vie in- diameter at the tip, after about 200 welds it might be anything from J in. to 5 /i 6 in- diameter, thus gradually increasing its contact area anywhere from 75 per cent to 175 per cent. 2. On butt welders the energy consumption does not depend 2000 2680 1600 1600 1 1400 c ~!000 E 300 600 X S- w 400 200 t\ / 24 ie 2144 v> 75 1876 3 V vo 1608 i" c 1340 * 0. 1072 | E 804 x *45 butt-welding machines, 247, *257 spot-welding machines, *292, *293 machine with 27-in. throat depth, *312 Position, correct, for using carbon arc and filler rod, *69 Positions of the universal spot-weld- ing points, a few, *297 412 INDEX Posture and equipment of arc welder, *49 Potts Co., John, electrodes made by, 13 Power factor in resistance welding, 383 rates, making proper, 382 required for percussive welds, 272, *273 Pressure required for flue welding, 337 heavy spot welding, 312, 313, 317 Principal parts of a butt-welding machine, *240 Projection allowed in welding boiler tubes, 330, *337, *340 method of welding, *291 Properties, mechanical, of twelve good arc welds, 173 , , inferior arc welds, 173 of arc-fused metal deposited at right angles to length of speci- men, 184 , tensile, of electrodes, 179, 180 Protecting the eyes in arc welding, 23 Pulley spot-welding machine, auto- matic, *301 Pulleys repaired by arc welding, *149, 150 Qualities of various kinds of glass, 25 Quasi arc welding, 86 , speed of, 168 Quasi Are Weltrode Co., 86 weltrodes, how to use, 86 Rail bonds, spot-welding, *320, 321 ends, built up cupped, 146 Railroad arc welding work, 145 Railway Age, 143 Rate of arc weldipg, 146 Rates, making proper power, 382 Rays, the infra-red, 23 , ultra-violet, 23 , visible light, 23 Reamer, steps in making a large, *353 * ' Recession, ' ' meaning of, 63 ' ' Re-entrant angle, ' ' meaning of, 63 Reinforced weld (arc), definition of, 117 Relation of arc current and electrode diameter, *13, 14 microstructure to the path of rupture in arc fused metal, 201 nitrogen content and current density, 178, 179 Removing broken taps, 150 Repairing crane wheels, 221 Resistance welding, 4 , energy consumption of, 384 machine, 239 Rims, automobile, butt-welding, 252, *255, *259 Ring welded to core with arc weld- ing machine, *232 Rivets in a ship, number of, 136 Rods, strength of mash- welded, *394 Roebling's Sons Co., John A., elec- trodes made by, 13 Roll housing, repaired, *148 Rolling boiler tubes, *331, *332, *334, *341 out upset metal on flue ends, *341 Rotatable head two-spot welding machine, 298, *299 Rotor ring butt-welding work, 263 Rovvdon, Henry S., 171 Ruder, W. S., reference to, 199, 205, 208, 209 Rules, general, for arc welders, 146 S Saw, butt-welding a band, *251, 252 Scarf angle for arc welding, 60 "Scarf, " meaning of, 63 INDEX 413 Scarf-weld, boiler tube ends pre- pared for, *326 welding boiler tubes, 336 Scarfing machine, a, *326 Scarfs, typical arc weld, *99 Schematic welding circuit, *10 Screens for arc welding, *19 Seam, automatic arc welded tank, *222 , flange, welding, 374, *377, *380 , flanged, welding with carbon arc, *75, 76 welding by the resistance process, 365 , current consumption for, 378 , details of roller head for, *368 machines, *367, *369, *370, *373, *374, *375, *376, *377 , material to use for, 365, 366 , speed of, with automatic arc machine, 222 Sectional view of carbon arc built-up weld, *73, *75 Selecting a welding outfit, 21 Self-contained portable welding set, Lincoln, *37 Semi-automatic arc-welding machine, 223, *230 Shaft, building up a, with an auto- matic arc welding machine, *218 , built up motor, *220 Shafts, worn armature, built up, *153 Shearing strength of butt- and spot- welds, 397 Sheet iron and steel, thickness and weight of, 322 , energy consumption in weld- ing, 384 metal arc-welding machine, *236, 237 work, convenient set-up for spot-welding, *295 steel box, spot-welding a, *283 Shell, cup for, welded by machine, *227, *228, *229 Shells, motor, repaired, *154, *156 Shields, hand, for arc welders, *15, *19 Ship parts, welded, examples of, *137 plates automatically arc-welded, *234, *235 work, spot-welding machines for, 311, *312, *316, *318, *319 Ships, German, names of, 127 Shops of the Santa Fe E. K., 339 Short and long arc deposits, *55 welding arc, *54 Single bevel, definition of, as ap- plied to edge finish, 114 "V," definition of, as applied to edge finish, 112 Size of cable for arc welding work, 18 carbon-electrode, 68 electrode for metallic arc welding of steel plate, 101 electrodes, 13, 14 Sizes of die-points for spot-welding, *290 electrodes used in automatic arc-welding machines, 220, 221, 222, 223, 232, 238 wire to use for connecting up different sizes of butt-weld- ing machines, 363 Slavianoff process, the, 1, 2 Sliding horn spot-welding machine, *291 Slip-bands, 188, 202 Smith, J. O., 134 Society of Naval Architects, 168 Solutions, etching, for steel, 185, 186, 187, 193, 194, 196, 198, 199, 200, 202, 204, 206, 208 Space-block spot-welding machine, *307 Stability of arc, 54 Special set up of arc welding ma- chine for building up a shaft, *218 machine for circular arc welding, *217 414 INDEX Speed of arc travel, 51 welding, 90, 167 automatic arc welding ma- chine, 220, 221, 222, 233, 238 building up shafts or wheels with automatic arc ma- chine, 223 cutting with the carbon arc, 78, 79, 80 < carbon electrode, 31 deposit per hour in arc weld- ing, 89 percussive welding, 272, *273 Quasi- Arc welding, 168 seam welding with auto- matic arc machine, 222 spot-welding, 321 welding boiler tubes, 333, 335 with metallic electrode, 32 Split-gear made solid, *160 Spokane & Inland Empire K. R., re- claimed wheels on, 157 Spot- and line-welding machines, combination, *308, *309, *310 welded holes, strength of, 392, *393 material that can be, 277 Spot-welding apparatus, first forms of, *5, *6, *7, *8 machines and work, 276, *278, *279, *281, *282, *283, *284, *285, *286, *287, *288, *291, *292, *293, *294, *295, *296, *299, *300, *301, *302, *303, *305, *307, *308, *309, *310, *312, *316, *318, *319, *320 , details of standard, 278 for ship work, 311, *312, *316, *318, *319 patents, *5, *6, *8 power and cost data, 321 tests on hoop iron, *390, 391 Spraragen, William, 167 Square patch arc welding method, *85 Stalls, individual, for arc welders, *20 Steel etching solutions, 185, 186, 187, 193, 194, 196, 198, 199, 200, 202, 204, 206, 207, 208 , melting, in nitrogen under pres- sure, 212 plates, rate of cutting, with the carbon arc, 80 seam welding, 366 wire butt-weld, *250, 252 Stellite insert welding jaws, *357 , jaws used for welding, *356 -tipped roughing drills, *348 , welding, 343 Steps in the making of a large reamer, *353 Stove parts, spot-welding, using swinging bracket support, *288 pipe dampers, spot welding, *285 Straight, definition of, as applied to edge finish, 113 Strap weld (arc), definition of a, 110 Stratton, Director of the Bureau of Standards, 171 Strength of arc deposited plates, 104 weld, variation of, with change of arc current, 102 welded joints, 91 welds, 140 cast iron welds, 131 resistance welds, 389 weld (arc), definition of, 116 of welded joints, 135 Stresses in arc welded joints, 92, *93 , result of ' ' locked-in, ' ' 62 Strip welding jig, *376 Strohmenger-Slaughter process, the, 1,3 Structure of arc deposited metal, *105, *106 electrolytic iron, character- istic, *199 Studies in overlap and penetration, *56 INDEX 415 Studs, use of, in arc welding, 129, ?130, *133, 144, 155 Successful welds, reason for, 138 Summary of the results of the study of the metallography of arc-fused steel, 212 Supervision of arc welders, 145 Surface, building up a, with the carbon arc, *72 Suspended head spot-welding ma- chine, 294 Swinging bracket support for spot- welding work, *288, *292 Swivel head, portable spot-welding machine, *293 Symbols, combination arc welding, 118, *119, *120, *121, *122, *123, *124, *125, *126 used in arc welding, 109 "Tack," meaning of, 63 Tack weld (arc), definition of, 115 Tank, corrugated steel, welding by machine, *236, 237 , how edges of, are welded, 237 seam, welded straight, *222 Tanks, arc welded, *137 Taper of carbon-electrode, 68 Taps, method of welding broken, to remove from hole, *149 , removing broken taps, 150 Taylor cross-current spot welding process, *8 spot-welding machines, *302, *303 Welder Co., 303 Teapot spout, a finish welded, *380 welding jig, *380 Tee weld (arc), definition of, 112 Tensile properties of electrodes, 179, 180, 181 Tension specimen, appearance of, after test, *183 Terminology, a brief, 63 Terms, elementary electrical, 398 Terrell Equipment Co., 296 Test blocks, formation of, for arc- fused metal, *175, 176 Tests, the Wirt-Jones, on arc welds, 189 Thermal analysis of arc-fused steel, 210 characteristics of arc-fused iron, 210, *211 Thickness and weight of sheet iron and steel, 321, 322 Thomson butt-welding machines, *240, *241, *244, *245, *246, *251, *254, *256, *345, *347, *348, *351 Co.'s tests on butt- welds, 395 spot welds, 391 Electric Welding Co., 366, 395 , Elihu, 4 foot-, automatic-, and hand-oper- ated spot-welding machines, 280 seam-welding machines, *367, *369, *370, *373, *374, *375, *376, *377 spot-welding machines, *226, *227, *281, *282, *283, *284, *285, *286, *287, *288 vertical mash welding machines, *358, *359, *360 Three-roller boiler tube machine, *332 Thum, E. E., 396 Time required to cut with the car- bon arc, 78, 79, 80 Tit or projection method of welding, *291 Topeka shops of the Santa Fe Kail- road, 339 Tool parts arranged for welding, *354, *355, *357, *361, *364 , grooving to aid in welding, *364 room butt-welding machine, *261 welding, the insert method of, 355 Tools, butt -welding, *350, *352, *354 Training arc welders, 47, 145 416 INDEX Transformer of butt-welding ma- chine, 239, *240 Truscon Steel Co., 397 Tube rollers, boiler, *331, *332, *334, *341 welding machine with built-on rolling device, *334, *339, *340, *341 -welding set for butt-welding work,'*263 work, examples of, *88 Tubes, boiler, pressure required for welding, 337 , , ready for flash weld, *329 Tubing automatically arc-welded, *233 Tungsten ring machine arc-welded to cold rolled core, *232 T-welding, 252 Twist-drill blanks just welded, *349 Two-spot welding machine with ro- tatable head, 298, *299 Typical ammeter charts of operation of Morton arc welding ma- chine, *224 examples of prepared and fin- ished arc welding work, *87, *88, *89 - light spot-welding machine, *278 Types of welding outfits, 21 U United Traction Co., shop repair work of, 150 Universal spot-welding die-points, *296, *297, *298 Unland, H. L., 214 Unsoundness of arc-fused metal, microscopic evidence of, 193 Uealite crucible, 208 Using a flux for flue welding, 330 U. S. Light and Heat Co., 40 portable a-c, motor- generator set, *39, 40 Van Bibber, P. T., 324 Vertical mash welding machines, *358, *359, *360 position defined as applied to ship work, *114, 115 seam welding, 62 Volt, what is a, 398 Voltage, effect of, on arc welds, 1G9 Voltex process, the, 2 Vulcan Iron Works, repair of large crankshaft by, 162 W Wagner, 167, 169 Wanamaker, E., 84 Warping of parent metal caused by deposit contraction, *59 Water-cooled die-points for spot welding, *281, *283, *284, *286, *288, *291, *296, *297, *298, *302, *303 ''Water-pail" forge, the, 1, 3 Watt, what is a, 398 "Weaving," meaning of, 64 of arc, 52 Weed, J. M., 311 Weight of sheet iron and steel, 322 Welded and riveted joints, *389, 391 automobile hub stampings, *2'2l rear axle housing, *222 Welder, points for the, to learn, 49 Welding boiler tubes by the electric resistance process, 324 booth, 48 Committee electrodes, composi- tion of, 107 the Emergency Fleet Corpor- ation, 90, 104, 107, 109, 134 high-speed to low-carbon steel, 344 Mild Steel, paper on, 223 other than round tools, 354 - pipe coils, *256 rotor bars to end rings in a special butt-welding machine, 263 Stellite, 356 INDEX 417 Welding valve elbows on Liberty motor cylinders, *268 Welds, arc, the Wirt-Jones tests on, 189 , good and bad arc, *100 showing poor and good fusion, *60 , terms and symbols for arc, 109 ' ' Welt, ' ' meaning of, 64 Weltrodes, composition of, 86 , sizes of, 87 Westinghouse Electric and Mfg. Co., 38, 47, 66, 81, 269 single-operator portable welding set, *38 Wheel, car, repairs, *157, *158 Wilson two-arc "plastic arc'" weld- ing set, *33 - Welder and Metals Co., 33, 81, 88, 128 welding and cutting panel, *34 Winfield butt-welding machines, *257, *259, 260, *261 - Electric Welding Machine Co., 260, 295 spot-welding machines, *291, *292, *293, *294, *295 Wire to use to connect up seam welding machines, 379 Wiring diagram for percussive weld- ing, *271 Wirt-Jones arc weld tests, 189 Work clamps for butt-welding ma- chines, *242, *243, *257 Worn and repaired crane wheels, *221 motor shaft built up by auto- matic arc welding machine, *220 Z Zerner process, the, 1, *2 Zeus arc-welding outfit, *42 THIS BOOK IS DUE ON THE LAST DATE STAMPED BELOW AN INITIAL FINE OF 25 CENTS WILL BE ASSESSED FOR FAILURE TO RETURN THIS BOOK ON THE DATE DUE. THE PENALTY WILL INCREASE TO SO CENTS ON THE FOURTH DAY AND TO $I.OO ON THE SEVENTH DAY OVERDUE. 10m-7,'44(1064s) YC 33567 y 742296 . UNIVERSITY OF CAL'^ORNIA )EPA RTMENT OF CIVIL ENGINEERING BERKELEY, CALIFORNIA UNIVERSITY OF CALIFORNIA LIBRARY