C0EXRIGHT DEPOSIT. Digitized by the Internet Archive in 2011 with funding from The Library of Congress http://www.archive.org/details/gastorchthermitwOOvial GAS TORCH AND THERMIT WELDING ^[ ^ Vfk QraW'MlBook (h. 7n& PUBLISHERS OF BOOKS FOP^ Coal Age '^ Electric Railway Journal Electrical World -^ Engineering News-Record American Machinist ^ Ingenieria Intemacional Engineering 8 Mining Journal ^ Power Chemical 6 Metallurgical Engineering Electrical Merchandising GAS TORCH AND THERMIT WELDING BY ETHAN yiALL 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, and Electric Welding. First Edition McGRAW-HILL BOOK COMPANY, Inc. NEW YORK: 370 SEVENTH AVENUE LONDON: 6 & 8 BOUVERIE ST., E. C. 4 1921 ^r^2c'^- ^■ \ Copyright, 1921, by the McGRAW-HILL BOOK COMPANY, Inc. MAR "8 1921 0)CU608790 r^ PREFACE The beginner, the practical worker, the student and tlie engineer, will find in this book a great amount of information regarding gas-torch and Thermit welding practice and equip- ment. No toil or expense has been spared to gather material of real and lasting value. Shops have been visited and data and photographs obtained first hand. Practically every book on welding has been carefully scrutinized for helpful suggestions. The services of experts have been engaged to give the results of long practice and research in their special lines. Each and every plan known to the experienced editor has been employed to give the reader the highest possible grade of information. The historical references have been cut to the barest state- ment of facts as we have been able to obtain them, yet they are ample enough to give the inquiring mind the genesis of each class. Foreign methods and equipment have not been touched upon, except in a few instances, because such treatment would add too greatly to the bulk of this work, without adding an appreciable amount to its real value, since the methods and standard equipment here are, in general, far in advance of anything similar elsewhere. Great care has been taken to indicate the sources of informa- tion and to give the names and addresses of the makers of equip- ment shown. It is believed that every well known maker of this class of welding apparatus in the United States has been men- tioned at least once in these pages. This has not been done with any idea of advertising them, but because it is information every reader is entitled to have without the necessity of making a separate search for it. Of course no recommendations regarding the best apparatus to use are made in any case. As in any other line, improvements are being constantly made, but in regard to newly invented or unknown equipment the seller should be made to prove his case before an investment is made. Apparatus which does not meet vi PREFACE the present day requirements, soon drops out of sight. It is a good plan for a prospective purchaser of equipment to consult some well established firm which is not afraid to advertise its product in open competition. Such a firm will see that its equipment is properly installed and works .satisfactorily. Ethan Viall. New York City, November, 1920. TABLE OF CONTENTS PART I— GAS TORCH WELDING. CHAPTER I PAGE History and Uses of the Gas Torch 1- 8 Meaning of * ' Autogenous ' ' and the Term ' ' Gas Torch ' ' Welds — The Oxy-Acetylene Gas Torch — Used for Both Weld- ing and Cutting — Hydrogen Gas — Thermalene Gas — Blaugas — Drigas — Illuminating Gas — Benzine or Benzol Vapors — Ex- plosive Limits of Welding Gases — The Field of Gas Torch Welding and Cutting. CHAPTER II TuE Production of Welding Gases — Oxygen and Hydrogen. ... 9- 25 Oxygen by the Liquid Air Process — Oxygen and Hydrogen by the Electrolytic Method — General Principles of the Elec- trolytic Method — Details of the Davis-Bournonville Electrolyzer Cell — The International Oxy-Hydrogen Generator — The Levin Type of Generator. CHAPTER III Acetylene and Medium, or Positive, Pressure Generators.... 26- 40 Acetylene — Acetylene Cylinders — Acetone Injurious to a Weld — Estimating Amount of Acetylene — Types of Acetylene Generators — The Positive Pressure Generator — The Davis- Bournonville Tj'pes — The Buckeye Carbide-Feeding Mechan- ism — The Portable Pressure Type — Approximate Dimensions and Weights of Acetylene Generator Sets. CHAPTER IV Low Pressure Acetylene and Thermalene Generators 41- 53 Low Pressure Generators — The Oxweld Duplex Generators — Thermalene Generators — How the Cartridge is Packed — Action of the Thermalene Generator — Some Advantages of Thermalene. viii CONTENTS CHAPTER V PAGE Gas Torches Used for Welding 54- 73 Types of Torches — The Davis-Bomnouville Positive Pressure Torch— The Prest-0-Lite Torch — The General Welding Co. 's Torch — The Imjaerial Torches — Calculating Amount of Gas Used — The Rego Welding Torch — The Oxweld Low Pressure Torch — The Messer Torch — The Thermalene Torch. CHAPTER VI Gas Cutting Torches 74- 94 The Davis-Bournonville Cutting Torch — The Oxweld Cutting Torch — Cutting Data — The Messer Torch — The General Welding Co. 's Torch — Imperial Torches — Caibo-Hydrogeu Torches — Airco-Vulcan Combination Torch — The Rego Torch — • The Milburn Combination Torch— The Torchweld Torch — The Davis-Bournouville Underwater Cutting Torch. CHAPTER VII Gas-Pressure Regulators and Working Assemblies 95-115 Oxweld Oxygen Regulators and Gages — Other Regulators and Gages — Tank and Hose Colors — Regulator Adaptors — Con- necting Up an Outfit — The Two Types of Tank Connec- tions — Characteristics of the Oxy-Acetylene Welding Flame — Imperial Three-Way Outfit — Lighting the Oxweld Low Pres- sure Torch — Characteristics of the Oxy-Hydrogen Flame — Characteristics of the Hydrogen-Compressed Air Flame — Char- acteristics of the Oxy-Illuminating Gas Flame. CHAPTER VIII Gas Torch Welding and Cutting Outfits 116-130 Typical Oxy-Acetylene Cutting Unit — Manifolds — Complete Working Outfits — Back Pressure A'alves — Lead Burning — Method of Connecting Outfits for Various Gas Combinations — A Gas Flow Indicator. CHAPTER IX Learning to Weld with the Gas Torch 131-153 The Way to Hold a Torch — Torch Motion — Welding Two Plates — Allowing for Seam Contraction — Using the Welding Rod — Various Welding Jobs — Sources of Trouble — Built-Up Welds — Vertical Welds — Filling Up a Hole — Forming Bosses or "Putting on" Metal — Practicing on Gear Teeth — Welding Backward — Lead Burning. CONTENTS IX CHAPTER X PAGE Makixg Allowance for Expansion and Contraction 154-168 Action of Metal When Heated — Using Heating Torches — Cooling Work — The Wiederwax Preheater — Suggestions Ee- garding the Welding of Gratings and Pulleys — Automobile Cylinder Work. CHAPTER XI Welding Various Metals and the Fluxes Used 169-186 Properties of Metals — Conductivity and Oxidation — Vaporiza- tion of Substances— Separation of Elements — Welding Various Metals — Welding Aluminum — Filling a Large Hole — Brass and Bronze — Cast Iron — Cast Iron to Steel — Copper — Copper to Steel— Lead — Malleable Iron — Monel Metal — Nickel — Steel — Special Steels — Manganese Steel — Nickel Steel — Vana- dium Steel — Chrome Steel — Wrought Iron — Galvanized Iron — German Silver — V/hite Metal Castings — Silver — Gold. CHAPTER XII Examples o? Welding Jobs 187-220 Preparing for Welding — Examples of Special Jobs — Welding Broken Machine Tools — Preparation for Locomotive Frame Welding — Cooling Devices — Rail Bonding — Rudder Frame Welding — Large Engine Cylinder Work — Welding High Speed Steel Tips to Low Carbon Shanks for Shop Tools. CHAPTER XIII Welding Jigs and Fixtures 221-238 Holding Pipe for Welding — A Welding Table — V-Blocks for Holding Shafts — Jig for Holding Crankshafts — An Adjust- able Crankshaft Jig — Crankcase Devices — Motorcycle Mani- fold Jig — Sheet Metal Roller Jig — Sheet Metal Cjdinder Jig — Apparatus for Welding Cylinder Ends — Welding Poison Gas Containers — Liberty Motor Work — Fixtures for Motor Manifolds. CHAPTER XIV Welding Machines 239-256 The Duograph — How the Duograph Works — Drum Welding Machine — Light Seam Welding Machine — Machine for Weld- ing Oblong Seams — Tube Welding Machines— Tube Welding by the Oxy- Acetylene Process — Arrangement of Rolls on a Tube Welding Machine. X ~ CONTENTS CHAPTER XV PAGE Cutting with the Gas Torch 257-277 Correct Cutting Position — Starting a Cut — Examples of Good and Bad Work — Blowing a Hole Through a Plate — Using Various Devices — Flame Control — Making a Ladle Hook — Costs of Some Cutting Jobs — Cutting Cast Iron — Torches Made to Preheat the Oxygen— Cost of Oxy-Hydrogen Cutting. CHAPTER XVI Cutting Machines 278-294 Cutting with Hand Machines — The Radiagraph — The Railo- graph — Circular Cutting— The Magnetogi-aph — The Camo- graph — The Great Western Cutter — The Pyrograph — The Uni- versal Cutter — The Oxygraph. CHAPTER XVII Welding Shop Layout, Equipment and Work Costs. 295-313 The Equipment Necessary for a First Class Shop — Layout of the Oxweld Shop — Keeping Track of C^osts — The Oxweld Cost Form — The Imperial Cost Form — Carbon Burning — Safety Rules for Gas Torch Workers — U. S. Railway Administration Autogenous Welding Rules — Strength of Oxy-Acetylene Welds. PART II— THERMIT WELDING. CHAPTER I Thermit Welding: Its History, Nature and Uses 317-321 What Thermit Is — Temperature and Characteristics — IMastic and Fusion Methods — Kinds of Thermit Commonly Used — Plain Thermit— Railroad Thermit— Cast Iron Thermit. CHAPTER II Making Plastic Process Welds 322-332 Uses of the Three Varieties of Thermit— A Pipe Welding Out- fit—How the Mold is Used— Placing and Igniting Thermit- Removing the Mold— Cost and Strength of Pipe Welds. CHAPTER III Fusion Welding of Heavy Sections 333-357 Type of Crucible Used for Thermit Welding— Tapping a Crucible — Life of Lining Prolonged With Magnesia Tar- Crucible Ready for Baking— Thimbles— Application of Fusion Welding— Wax Pattern Molds— Ramming the Mold— Preheat- ing the Mold — Igniting Thermit — Amount of Thermit Needed for Welds — Locomotive Frame Work — Other Railroad Work. CONTENTS xi CHAPTER IV PAGE Welding Crankshafts, Mill Pinion Tektii, Etc 358-373 Y-Bloeks for Welding Crankshafts — Defects that Frequently Occur — Inaccuracy of Alignment Explained — How to Locate Minute Cracks — Welding New Teeth in "Large Pinions — Mak- ing a Wax Tooth Pattern — Another Method of Welding Pinion Teeth — Preheating Large Work. CHAPTER Y Welding New Necks on Large Pinions and Other Heavy Work 374-390 Two Methods of Working — Foundation and Heating Arrange- ments — Constructing the Mold for a Large Roll or Pinion — ■ Amount of Thermit Required — Treatment When a Cope is Used — Alternate Method — Marine Repairs. CHAPTER VI Rail Welding for Electric Systems 391-402 Rail Joint Work — Adjusting the Insert Between the Rails — ■ Adjusting the Mold — Preheating — The iJse of Thermit Addi- tions — The Grinding Machine. CHAPTER VII Welding Compromise Rail Joints 403-412 Kinds of Compromise Joints — Using a Rail Section for a Pat- tern—The Clark Joint— Modified Clark Joint— Welded Cross- over — Motor Case Work — Car Truck Work. CHAPTER VIII Welding Cast Iron and Other Parts 413-425 Thermit to Use for Cast Iron — Examples of Cast Iron Welds — - Welding High Speed to Machinery Steel — Cost of Thermit Ap- paratus — Preheaters for Thermit Work — Cost of Materials and Apparatus for Pipe Work. Index 426 PART 1-GAS TOECH AVELDING GAS TORCH AND THERMIT WELDING CHAPTER I HISTORY AND USES OF THE GAS TORCH. According to common usage, the term "autogenous weld- ing" is erroneously applied only to hot gas flame fusion welds. The gas combinations used for the production of the hot flame for welding or cutting are oxy-acetylene, oxy-hydrogen, oxy- thermalene or any combination that will produce sufficient heat, and is applied by means of a torch or blow-pipe. The welds so produced are strictly fusion welds, as no pressure or ham- mering is employed to effect the union. The word ' ' autogenous ' ' means "self-produced" or "self-generated," that is, joined with the same metal, and as such applies equally to hot gas flame, electric arc or thermit welds, although as just stated, the present custom is to apply the term generally to hot gas flame welds. However, owing to the wide field that the term "autogenous" really covers, and to the looseness with which it is often applied, we prefer to use the term "gas torch" in connection with Avelding and cutting by means of the hot gas flame. While the use of a blow-pipe or torch in some form was known to the ancients, the high temperature gas flame is a development of the last quarter of a century. The more com- monly known gas combination is oxy-acetylene. Acetlyene (CoHo) was discovered by Edmund Davy in 1836, but it re- mained only a laboratory gas until T. L. Willson of North Carolina and H. Moisson, the Frenchman, developed commercial 2 GAS TORCH AND THERMIT WELDING methods of producing calcium carbide (CaC^) in large quanti- ties in 1891-92. In 1895 Le Chatelier read a paper before the Paris Academy of Sciences in which he stated that : ' ' acetylene Inirned with an equal volume of oxygen gives a temperature which is 1000 deg. C. (1800 deg. F.) higher than the oxy- hydrogen flame. The products of the combustion are carbon monoxide and hydrogen, which are reducing agents." Further along he said: "this double property makes the use of acety- lene in blow-pipes of very great value for the production of high temperatures in the laboratory." This statement of Le Chatelier is especially noteworthy, since he set the ratio of the gases at equal volumes, and not at the theoretical proportion of 2^ volumes of oxygen to 1 of acetylene. The application of the oxy-acetylene gas torch to metallic welding dates experimentally from 1901, and industrially from 1903. Edmond Fouche, of Paris, who did considerable experi- menting in conjunction with Picard, is generally credited with having devised the first really practical and safe torch. In February 1904 Fouche sent two of his torches to Eugene Bournonville, of New York, with which the latter repaired a machine that was still in use years later. The Fouche and Picard torch first developed, used both oxygen and acetylene under high pressure. There proved to be serious objections to this, and Fouche next produced the low pressure or injector type of torch which employed only the oxygen under high pres- sure. Following these was the Gauthier-Ely positive pressure or medium-pressure type which used both gases under moderate and independent pressures. This type was later brought to the United States by Augustine Davis and Eugene Bournon- ville in 1906. During this year Bournonville designed the first acetylene pressure generator produced in connection with the oxy-acetylene process. In 1905 and 1906 considerable Avelding work was done but the process was handicapped by the inadequacy and poor quality of the oxygen then obtainable, and also by the im- perfect knowledge and technique necessary to good work. In 1902, Carl Linde patented in England a process for liquefying air and producing oxygen and nitrogen. In 1906 a plant for the production of oxygen by the Linde process was established in Buffalo, N. Y. From that time on, oxygen plants of various HISTORY AND USES OF THE GAS TORCH 3 kinds have constantly increased in number and the commercial production of oxygen of good quality has been a great factor in the development of gas torch welding. At first, operations were limited to the simplest repair work on iron or steel. As the apparatus was improved and the efficiency of the welders increased, the field widened. New uses have been found for the process and the range of metals coming within its scope has steadily expanded. It has its limitations, however, which will be pointed out elsewhere. Used for Both Welding and Cutting-. — In addition to weld- ing, the oxy-acetylene flame, as well as a number of others, is applicable to cutting. In fact so closely allied are welding and cutting in this field, that an operator is usually called upon to do both many times in a day's work. Cutting by means of an oxygen jet was first made commercially possible by Jottrand, who took his basic patent in 1905. Aside from the manual operation of welding or cutting torches, a large number of machines have been designed. These range from simple wheel or radius attachments for the torch itself, to huge automatic pipe making machines or others of a complicated nature. The oxy-acetylene flame consists of two parts, a small inner luminous "cone" which is bluish white in color, and a larger enveloping non-luminous flame. The temperature at the apex of the cone is estimated to be about 6300 deg. F. This heat is not surpassed by any burning gas with the possible exception of thermalene, for which 6500 deg. F. is claimed. For welding purposes the high efficiency of acetylene is due to its high carbon content and to the fact that it is endothermic, that is to say, heat-absorbing in its formation. Energy stored up in formation is given off again in the form of heat by the acetylene upon dissociation. It is calculated that of 1475 heat units in a cubic foot, 227 are due to the mere breaking up of the gas. "While theoretically two and one-half volumes of oxygen are needed to completely burn one volume of acetylene, the ratio in which the gases are employed in prac- tice is about one volume of oxygen to one volume of acetylene. The flame yielded by such a mixture is the correct one, or the so-called "neutral" flame. By increasing or decreasing the proportion of oxygen, flames known as either oxidizing or 4 GAS TORCH AND THERMIT WELDING reducing may be obtained, tbe appearance of the cone changing as the proportions are modified. While the use of oxy-acetylene for welding is more commonly known than any other combination, there are several gases, which when mixed with oxygen, will produce more or less satisfactory welds. Some of them are to be preferred to acetylene for certain cutting purposes. The better known gases are described as follows, it being understood that they are to be used with oxygen. Hydrogen gas is a chemical element which exists in nature in great quantities in various chemical combinations. The most common is its union with oxygen to form water (HoO). As a consequence, water is used as a basis for making both oxygen and hydrogen. Oxy-hydrogen welding was the first gas torch welding system employed, and it was used quite extensively until the introduction of the more advantageous system of weld- ing with oxy-acetylene. While hydrogen may be manufactured on the premises, it is also handled commercially in steel cylinders. In using this flame for welding there is an existing danger that the oxygen may unite with the metal causing it to be overheated or burnt. To prevent the burning of the metal, it becomes necessary to use a supercharge of hydrogen so that oxygen liberated within the flame will combine with the free hydrogen instead of with the metal. This, however, increases the size and decreases the temperature of the flame. The temperature of the oxy-hydrogen flame according to Kautny, can never go higher than the dissociation temperature of water, which is estimated at 2000 dcg. C. (3632 F.) For welding thin metal sheets hydrogen is practical on account of its comparatively low heat. The quality of the weld, however, decreases as the thickness of the metal increases. While theoretically only two volumes of hydrogen are required to one of oxygen, in actual practice when employing an oxy-hydrogen toreh, it is necessary to use four or five volumes of hydrogen to one of oxygen in order to insure a non-oxidizing flame. This in itself is a waste- ful process, since the maximum heat obtained is limited to the amount produced by combining two volumes of hydrogen to one of oxygen. For heavy cutting it is preferred to acetylene on account of its longer flame. It is also used extensively for lead HISTORY AND USES OF THE GAS TORCH 5 burning, preheating, soldering, brazing, annealing, special forg- ing or rivet heating and a number of other things. Tliermalene is one of the latest gases to be produced. It is the discovery of Linus Wolf, Zurich, Switzerland, and it is handled in this country by the Thermalene Co., Chicago Heights, 111. It is a combination produced by the decomposition of calcium carbide and hydrocarbon oils, the heat generated by the carbide being used to vaporize the oil. It is used for either welding or cutting. Blmigas is a liquid under pressure. It is the discovery of Herman Blau and it is made from gas oil, a product of the oil refineries. It probably has the low^est explosive range of any gas used for illuminating purposes, the range being about 4 per cent wiiile that of coal gas is about 13 per cent. Like coal gas, however, it is little used for welding, though sometimes used ft)r cutting. Blaugas is marketed in steel cylinders having the equivalent of 1300 cu.ft. of city gas, by the American Blaugas Corp., New York. Its largest field is for cooking and lighting purposes where coal gas is not readily obtainable. Owing to its portability it may be used to advantage for pre- heating work. Drigas is a light oil gas, which is a vapor under pressure. It is sold in steel cylinders of about 150 cu.ft. by the same concern handling blaugas. It is especially good in combination with oxygen for cutting metal from 1 to 12 in. thick, and is also considerably used for preheating. Its explosive range is about ^ that of coal gas, and it is non-poisonous and non- asphyxiating. Illuminating Gas (coal gas or water gas) can only be used for welding very thin pieces owing to the low temperature of the flame. It may, however, be used for preheating or cutting. Benzine or Benzol Vapors have the same properties, approxi- mately, as blaugas. The temperature is a little higher than that of illuminating gas, but much lower than acetylene. It is only used for welding under special circumstances. While a number of gases, which are used with oxygen, have been mentioned, only the production and use of hydrogen, acetylene and thermalene will be described, along with that of oxygen. Explosive Limits of Welding Gases. — In order to be ex- 6 GAS TORCH AND THERMIT WELDING plosive, a combustible gas or vapor must be mixed with a certain amount of oxygen or air, the proportions of the mix- ture ranging between certain limits depending on the char- acter of the fuel. Any figures showing these explosive limits of the gases can only be approximate at best, since so many things enter into the calculations, such as the purity of the gas, means of ignition, temperature, pressure, and so on. In general, the mixture that has just enough oxygen for com- plete combustion of the fuel gives the highest pressures and temperatures, and very nearly the highest speed of ignition. If the proportion of oxygen (air) is increased beyond, or decreased from, the theoretical proportion, the maximum pressures and temperatures are lowered and the speed of ignition decreases until at certain upper and lower limits the mixture ceases to be explosive, and only slow combustion can occur. The figures here given are believed to be a fair average of those given by the various authorities. The explosion is sup- posed to be caused by an electric spark, at atmospheric pressure and a temperature of about 65 deg. F. Acetylene — 3 per cent gas plus 97 per cent tiir to 55 per cent gas plus 45 per cent air, or a range of 52 per cent (one writer says 73 per cent gas plus 27 per cent air). Blangas — i per cent gas plus 96 per cent air to 8 per cent gas plus 92 per cent air, or a range of 4 per cent. Coal Gas — 6.5 per cent gas plus 93.5 per cent air to 19.5 per cent gas plus 80.5 per cent air, or a range of 13 per cent. Drigas — 4 per cent gas plus 96 per cent air to 8 per cent gas plus 92 per cent air, or a range of 4 per cent. Hydrogen — 10 per cent gas plus 90 per cent air to 66 per cent gas plus 34 per cent air, or a range of 56 per cent (one writer says 6 per cent plus 94 per cent to 72 per cent plus 28 per cent). Thermalcne — 12 per cent gas plus 88 per cent air to 30 per cent gas plus 70 per cent air, or a range of 18 per cent. The ignition temperatures of some of the gases, at at- mospheric pressure are: Acetylene, 760 to 820 deg. F. ; city gas, 1100 deg. F. ; hydrogen, 1075 to 1100 deg. F. According to McCormack, the cu.ft. per pound of gases M'as calculated for the specific gravity and found to be : Acety- lene, 14.8 cu.ft. ; coal gas, 24.3 cu.ft. ; hydrogen, 192.4 cu.ft. ; thermalene, 13.97 cu. ft. HISTORY AND USES OF THE GAS TORCH 7 The Field of Gas Torch Welding and Cutting. — In a general way, the field of the gas torch welding and cutting may be outlined as follows, though some of the applications enume- rated are more advantageously done by other methods. This is especially true with reference to the welding of heavy sec- tions which should, as a rule, be done with thermit. Airplane Construction. — Welding water jackets to cylinder, valve cages to cylinder, of manifolds (intake, exhaust, and cooling), flanges to the manifold connections, spark plug thimbles, tubular sections for frame, splice plates, sockets to frames, aluminum crank cases, water tank. Automobile Industry. — Welding rear axle housings, defective gears and pinions, manifolds, shafts, steering posts, automobile bodies (aluminum and steel), tubing used in wind shields, etc., crank cases, transmission cases, wheels which are made of stamped-out parts, mufflers, valve stems to valves, rims, repairing crank shafts, frames, extending frame to make a truck out of a car. Copper Plate. — Welding manifolds, flats, kettles, vats, tanks, copper, stills and chemical ware. Electric Raihvay. — Welding of bonds, worn boxes, riotor housings, building in teeth of defective pinions and ge:irs, reclaiming of broken trucks, welding air receivers on air-brake system, steel trolley wires, side frames. Forge Shop. — Welding ornamental iron, complicated parts. Foundries. — Steel foundry: Welding up of blowholes, porous spots, blocks, cutting of risers, gates and heads ; welding moldings which are cast in parts. Cast iron foundry : Reclaiming castings. Lead Burning. — Burning of connectors on storage batteries, bat- tery repairs, lead linings in vats, tanks, etc., lead-pipe joints. Piping and Gas Main WorJi. — Welding of steam, air, gas, oil, and water lines, welding for high pressure gas distribution, annnonia systems. Fittings, such as T's, Y's, S's, crosses, which are cut and welded on the job, meter connections for houses, traps, drip pots. Plate Welding.^— Anmwma receivers, generators, air receivers, tanks for oil, vacuum driers, digesters, vats, steam driers, tanks of all kinds which are to be subjected to heat and pressure, plate assembly work for gas manufacture by-products, recovery work, stills. Poicer Plant Maintenance. — Building up worn or broken parts, weld- ing of cylinders, pistons, valve chests, etc. Welding of steam lines, of pump castings broken in service. Repairing of flywheels. Railroad Repair. — Firebox repairs (including patches), replacing side sheets, welding in flues, cutting off rails, mud rings, welding cracked steam chest, valves, cross-heads, cylinders, building up worn pins, cutting out links, irregular shapes of steel, filling worn spots on wheels, welding spokes, cutting and welding up locomotive frames. Welding together parts of car seats, chair and window frames. Re- 8 GAS TORCH AND THERMIT WELDING claiming bolsters, couplings, slotting forged engine rods; building up frogs and diamond crossings, scrappings. building steel cars. Rolliny Mill. — General repair of engines, rolls, hot beds, plates, furnace equipment, fabricating open-beai"tli water jacket doors, re- claiming copper tuyeres, cutting up lost heats, cutting up "kindling" or scrap, bar stock, billets, plates. Sheet il/<:Yrt?.— Manufacture of metallic furniture, steel barrels, trans- former cases, range boilers, kitchen utensils, light air tanks, tubing, oil storage tanks. Shipyards. — Cutting of plates, channels, special sections, welding and reclaiming of broken parts of machinery and propellers, patching of hulls, stringers, building up of worn chocks. Small Arms Manufacture. — Reclaiming component parts, spot hard- ening of different parts, spot annealing. Structural Steel. — Cutting as applied to coping, splicing and fitting rails, channels, I beams and other sha]ies. Cutting holes for rivets, welding up misdrilled holes, cutting of all kinds of gusset splice plates, cutting wrecking, welding structural parts where riveting is not possible. CHAPTER II THE PRODUCTION OF WELDING GASES— OXYGEN AND HYDROGEN Oxygen is a gas which constitutes about 23 per cent by weight and 21 per cent by volume of the air we breathe, most of the other percentage being nitrogen, a gas which does not support combustion. Oxygen itself will not burn, but it is the greatest supporter of combustion known. It was probably discovered by Stephen Hales in 1727, though Priestly was the first to publish a description of it in 1774. The name ' ' oxygen ' ' was later applied to the gas by Lavoisier. Pure oxygen is colorless, odorless and tasteless. For welding work it is im- portant that the oxygen used be as pure as it is commercially possible to obtain it. The impurities which decrease its effi- ciency are usually hydrogen and nitrogen. There are three ways to produce oxygen commercially; by means of liquid air, by chemicals and by the electrolytic process. When oxygen is made by the liquid air process, there is a certain amount of nitrogen present. In the chemical methods, a number of impurities may cause trouble. By the electrolytic process, the impurity is hydrogen. Oxygen by the Liquid Air Process. — As a general rule, taking everything into consideration, it is far better for the average or small user to buy his oxygen from a reliable con- cern and not try to manufacture it himself. The oldest concern in this country making oxygen by the liquid air process is the Linde Air Products Co., with offices in New York City. Their cylinders are regularly furnished in two sizes of 100 and 200 cu.ft. capacity respectively. They are charged to a pressure of 1800 lb. at a temperature of 70 deg. F. Customers are furnished cylinders free and pay only for the oxygen. Empty loaned cylinders are exchangeable for filled ones at stations in practically every city of fair size in the country. A 50 ft. 9 10 GAS TORCH AND THERMIT WELDING size of cylinder is obtainable for those whose requirements are very limited. A number of other concerns supply elec- trolytic oxygen for the market, the 100 cu.ft. cylinders being about 85 in. in diameter and 48 in. high, weighing approxi- mately 122 lb. when filled. Tlie average purity of oxygen in cylinders is about 99 per cent. Since the production of oxygen by the liquid air process is only applicable to large installations any detailed descrip- tion of the method would be out of place here. It is sufficient to say that in general, the process consists of first reducing the air to liquid form by means of the combined action of high compression and low temperature, and then separating the oxygen and nitrogen of which it is composed, by taking advantage of the different boiling points of the two. Under atmospheric pressure the boiling point of very pure liquid oxygen is — 182.7 deg. C. ( — 296.9 deg. F.) and of very pure nitrogen — 195.5 deg. C. ( — 319.9 deg. F.). This means a dif- ference in the boiling points of 12.8 deg. C, or 23 deg. F. These respective boiling points will, of course, vary under different pressures, and various degrees of purity, but tlie difference between the two is sufficient to allow of the nitrogen lieing vaporized in suitable apparatus and carried away before the oxygen vaporizes. The Chlorate of Potash Process. — Where circumstances make the chemical production of oxygen advisable, the chlorate of potash method is probably the most satisfactory at the present time. In this process chlorate of potash (KCIO3) and manganese dioxide (MnOo) are mixed together in the proportion of 100 to 13 parts (about 8 to 1). This riixture is placed in a retort filled as full as possible to exclude air. The retort is then heated and the oxygen is driven off. As the oxygen gas passes off from the retort it is conveyed through a cylinder or vessel containing sodium hydroxide (NaOH) which removes most of the impurities. The oxygen is then piped to a gasometer from Avhich it may be used direct or pumped into cylinders. A pound of the mixture is said to' produce about 4 or 4^ cu.ft. of oxygen. The manganese diox- ide is unchanged during the process. It is used because it enables the chlorate of potash to more readily give up its oxygen and at a lower temperature than without it. THE PRODUCTION OF WELDING GASES 11 An oxygen generator working on the general principles just outlined is made by the Macleod Co., Cincinnati, Ohio. This firm makes both a stationary and a portable type. A stationary type is shown in Fig. 1 and a portable one in Fig. 2. The generator, which consists of a furnace with retort, a scrubber for holding the purifying solution, and a receiver Fig. 1. — The Buckeye Chemical Oxygen Generating Set. for the gas, is quite convenient for small shops or garages where a limited amount of oxygen is used. It is quite possible, however, to fill tanks from these generators for storage pur- poses and immediate use if needed. An oxygen generator may be used in conjunction with a cylinder of dissolved acety- lene, or with a separate acetylene generator. Where portable 12 GAS TORCH AND THERMIT WELDING j]fenerators are used for both oxygen and acetylene, it is ad- visable to have the outfits on separate trucks so as to decrease danger should any leaks develop. The Buckeye generator is so made that it is adaptable to the use of wood, coal, coke or charcoal for fuel, or can be fitted with gas, gasoline, alcohol or oil burners. The portable type is shown fitted with a gasoline burner. The generators are Fig. 2. — Buckeye Oxygen Generator Mounted on a Truck. tested to 2^ times the maximum working pressure of 300 lb. Safety devices are provided so that it is safe for practically unskilled tenders. These generators are made in three sizes, with a capacity of 40, 60 and 100 cu.ft. of oxygen per hour. The weight of the portable type will range around 650 lb. and of the larger stationary type about 1600 lb. THE PRODUCTION OF WELDING GASES 13 OXYGEN AND HYDROGEN BY THE ELECTROLYTIC METHOD The electrolytic process for the production of oxygen is more adapted to private installations than the liquid-air process. An electrolytic installation is flexible, and may be expanded so as to produce any commercial quantity of gas desired with very little attention. One big advantage of this process is that hydrogen is produced at the same time as the oxygen, and in many cases this hydrogen can be used to advantage for welding, cutting, or other purposes. As a rule, oxy-hydrogen for welding is less desirable than oxy-acetylene, but for some purposes, especially when there is an abundance of hydrogen available, it is very satisfactory. The heat produced by the oxy-hydrogen flame (about 2632 deg. F.) is considerably less than that of the oxy-acetylene flame (about 6300 deg. F.), consequently it is commonly em- ployed for welding thin metals, lead burning or other work Avithin its heat range. As a general rule, oxy-hydrogen is good for welding 16-gage steel, or thinner, but should not be used on steel over ^ in. thick. As hydrogen contains no carbon, the weld is softer than with acetylene. Cast iron up to f in. in thickness may be successfully welded, as may also aluminum crankcases or alloyed metals. For cutting, however, oxy- hydrogen has a wide field, especially for heavy work. General Principles of the Electrolytic Method. — In an ele- mentary form, decomposition of water may be effected by passing an electrical current between two metallic poles, or electrodes, immersed in water. By the admixture of acid or alkali, forming an electrolyte, the resistance of the water is lowered to allow a large current of electricity to pass, pro- portionately raising gas production. Simultaneously with the passage of current, decomposition of water into its components, oxygen and hydrogen, begins. Oxygen, exhibiting positive electrical properties, is formed on the positive pole or "anode" ; double quantity of hydrogen is formed at the same time on the negative pole or ''cathode." The gases are immediately available, and by interposition of a suitable diaphragm be- tween the poles, are kept separate and led to their proper receivers. 14 GAS TORCH AND THERMIT WELDING Tlic rapidity of decomposition, and consequently tlie amomit of gases evolved being in direct measui'C of the electrical current passing, there is alTorded convenient and economical means of producing commercial oxygen and hydrogen. The electrolytic solution increases in density as the action con- tinues. The volume of water dissociated is therefore replaced at regular intervals. Complete separation of the gases is desirable in order to insure their availability at high purity. This involves the use of a diaphragm, which, immersed in the solution, will allow passage of current between the poles and at the same time prevent mixing of gases. The production of oxygen and hydrogen being in respect to the amount of current passing, it is apparent that the voltage required to send the specified amount of electricity tlu-ough the electrolyzer is a measure of the efficiency of the apparatus, since, if the kilowatt-hour consumption is known, the gas pro- duction may be compared with it. Thus, there has been evolved the commonly accepted performance rating of any electrolyzer given in terms of cubic feet of gas produced per kilowatt-hour operation. The production of pure gases is very important. In the earlier types of water electrolyzers the requirements for pro- ducing gases of high purity were not understood, witli the result that means of purification of the gases after generation were necessary. Devices of this character have been found expensive to maintain and inefficient in action. Modern de- signs of electrolyzers are capable of delivering oxygen of about 99 per cent, and hydrogen of equal or greater purity, so that the need for external purifying means no longer exists. On delivery from the electrolyzers the gases are conducted separately to a pressure regulating device which imposes equal pressures on both oxygen and hydrogen, thus equalizing the pressures on each side of the separating diaphragm. The gases are then passed to their respective gas holders, in which they are collected and stored at a few ounces pressure, l^pon the nature of service of the gases will depend the size of the gas holders, and the method of compressor control. If it is desired to compress the gases into cylinders for shipment, as in the case of a commercial plant, large gas THE PRODUCTION OF WELDING GASES 15 holders are employed having capacity for, at least, a con- tinuous day's run of the electrolyzers. High-pressure com- pressors draw from these holders and discharge to a manifold to which the portable cylinders are connected. The pressure carried in the cylinders is usually 1600 to 1800 lb. per sq.in. Cylinders of 100 and 200 cu.ft. capacity will weigh about 85 and 150 lb. respectively. There are many oxy-hydrogen producing equipments in- stalled in industrial establishments, the gases being utilized in various portions of the works. In the Davis-Bournonville installations, gas holders of moderate size are employed, their rise and fall starting and stopping the compressor motors 'through automatic electrical control devices. The gases may be stored in stationary pressure tanks to a moderate amount, these being fitted with automatic regu- lators, so that when they are filled to capacity, the entire plant will be shut down. The gases are piped, where desired, through pressure lines, thus avoiding the replacement of empty cylinders. This method of installation is particularly desir- able for continuous welding and cutting operations, either by hand or mechanical means. Provision may also be made for charging portable cylinders for use in operations carried on at isolated points. Through the automatic control mentioned, the flexibility of an oxy-hydrogen generating and compressing equipment may be appreciated. The required amount of attendance being small, and needed only at regular intervals, continuous 24- hour operation of the equipment or intermittent service, if desired, is quite feasible and practicable. If maximum pro- duction is not desired, reducing the current passing through the electrolyzers will proportionately lower the volume of gas that is being generated. Details of the Davis Electrolyzer Cell.— For various reasons, electrolytic installations are made up of small units or cells, which may be combined in such a way as to produce any required amount of gas. Details of an electrolyzer cell are shown in Fig. 3. This type of cell is made by the Davis- Bournonville Co., Jersey City, N. J. The type illustrated provides current conducting areas and gas generating sur- faces amply proportioned to their requirements. There is suffi- 16 GAS TORCH AND THERMIT WELDING cient over-capacity to minimize electrical resistance and afford high working efficiency. Long life of the vital parts is also insured. Research has shown that a nickel-iron-alkali com- bination of elements employed for electrolytic dissociation of water is a very efficient selection from an electrical input and gas producing standpoint. Parts subject to deteriorating action of any character are constructed of special material and pro- tected by processes especially adapted to service requirements. Care has been exercised in the design so as to avoid com- plication of electrical and mechanical connections of small cross-section. Thus studs, bolts, busbars and their contacts are amply large for all purposes. These eleetrolyzers are manufactured in two sizes, operating on specified currents of 500 and 1000 amp. respectively. The dimensions of the respective cells are 54 and 6I-2- in. high, 13^ and 15J in. thick and 24^ and 36 in. wide. The licight given is from the bottom of tlie cell to the center of the highest horizontal tube, through which the hydrogen passes into the service pipe. In stating the production of gases evolved by dissociation of water the commonly accepted formula employed specifies production . of 7.93 cu.ft. of oxygen with double quantity of hydrogen per kilo-ampere-hours at normal temperature and pressure. The normal production of Davis-Bournonville elee- trolyzers may therefore, according to their booklet, be stated as follows : Noniiiil Hourly (Jlas rroduction Type Aiiiperagc Oxygen Hydrogen 5 500 3.96 cu.ft. 7.92 cu.ft. 6 1000 7.92 " 15.84 " at 20 deg. C. and 760 mm. barometer. The closed-cell type of construction adopted eliminates the absorption of carbon dioxide (CO2) by the solution exposed to the atmosphere in the open type of electrolyzer, and its consequent deteriorating effect upon the electrolyte and purity of gases. Electrical current passing tlu'ough the electrolyzer is converted almost entirely into chemical energy for producing oxygen and hydrogen. There being practically no action on the electrolyte employed as a conducting medium between the poles other than the dissociation of water, it is evident THE PRODUCTION OF WELDING GASES 17 that tlio electrical pressure or voltage required to send the specified amount of electrical energy through the apparatus is a measure of its efficiency. Referring now to the illustration, it should be kept in Fig. 3. — Details of the Davis Electrolyzer Cell. mind that the solution used is water with certain chemicals, such as sodium hydroxide (caustic soda) or potassium hydroxide (caustic potash), added to increase the conductivity. The reservoir in which the solution is placed, is divided by 18 GAS TORCH AND THERMIT WELDING a metal plate A. Anodes B are suspended on each side of this plate, and on these the oxygen forms. The cell itself is made of metal, and the walls of this, as well as the sides of the metal plate A, form the cathode or negative pole from which the hydrogen gas rises. To keep the oxygen and hydrogen separated, asbestos sacks C are so placed as to surround each of the two anodes. The oxygen generated passes up through the hard rubber tubes D connected to the pipe E. The hydrogen passes up tube F into pipe G. The current to the anodes is conducted through the positive busbar assembly H. The nega- tive busbar assembly, shown at /, is attached to and forms part of the cast-iron cover of the cell and connects with the center plate and the tank walls. Valves for the three gas tubes are indicated by J. As the current passes through the cell the entire solution is charged and this results in the freeing of oxygen at the anodes and hydrogen at the cathodes. Since these gases have no tendency to pass off anywhere except at the respective terminals in the cell, the asbestos curtain effectively keeps them separated. The pressure of the two gases, how- ever, must be kept the same or the one having the higher pressure will be forced through the fabric of the asbestos sacks and mix with the other gas. This is taken care of by having the gases from the pipes E and G pass through a combined flash-back and pressure regulator. The function of this device is to receive the gases ; regulate their pressure through a simple water seal which equalizes the gas pressures inside the elec- trolyzer; separate and return to the cell any alkali carried over ; provide means of replacement of water to the cell ; bypass gases to the air if the delivery lines become obstructed, and to prevent admission of any flame to the electrolyzer. The replacement of distilled water, as needed, is made through the reservoir K, which combines the replacement func- tion with that of a hydraulic governor automatically adjusting the inner level of the solution. Under operating conditions, the usual replacement of distilled water amounts to approxi- mately one gallon per 100 cu.ft. of oxygen and 200 cu.ft. of hydrogen. This replacement and the ordinary inspection usually given to the electrical apparatus is practically all the atten- tion required for a battery of cells. The International Oxy-Hydrogen Generator. — The elec- THE PRODUCTION OF WELDiNG GASES 19 trolytic cell shown in Fig. 4 and in further detail in Fig. 5, is made by the International Oxygen Co., New York. Each cell unit requires a floor space 4 X 40 in., and with the neces- l^'iG. 4.— Cell Made by the International Oxygen Co. sary pipe connections, a head room of about 6 ft. These cells are intended to be run on a normal amperage of 600 and a voltage of 2.2 each, using a caustic soda solution. The 20 GAS TORCH AND THERMIT WELDING possible range above and below the normal amperage is con- siderable, without injury to the cells. An equipment of their type 4-1000 cells can be operated with good economy over a current range of less than 200 up to 1000 amp., representing a production range of more than one to five. In actual figures this means that an installation giving 600 cu.ft. of oxygen OXYGEN STANDARD CAS TRAPS WATER FEED HYDROGEN CXYG CAS CHAM Pig. 5. — Some Details of the Cell Construction. and 1200 cu.ft. of hydrogen per 24 hours, at 600 amp., can by varying the current and without any alteration in the plant, be made to deliver from less than 200 cu.ft. of oxygen and 400 cu.ft. of hydrogen, to more than 1000 cu.ft. of oxygen and 2000 cu.ft. of hydrogen per 24 hours. Operating at 200 amp., the power consumption per unit of gas generated is 16 per cent less than the normal 600-amp. operation. When THE PRODUCTION OF WELDING GA8ES 21 operating at 1000 amp., the power consumption per unit of gas is 15 per cent more than at 600 amp. Local rates will largely govern the number of cells required for a given output of the gases. Where the cost of current Fig. 6.— a Group of I. O. C. Cells. is low, the plant can be economically run on current above 600 amp., the increased production per hour giving the re- quired amount with fewer cells, since the lower current cost justifies the slightly lower electrical efficiency. Where the 22 GAS TORCH AND THERMIT WELDING price of current is high, it will be advantageous to use current less than 600 amp., thus taking advantage of the higher elec- trical efficiency, but more cells will be needed. In general principles,, this make of cell resembles the one Fig. 7.— Suggested Layout of a 50-Cell Plant. previously described, though different in form. By referring to the illustration, it will be seen that a cell is made up of a tliin rectangular-shaped box frame to the sides of which are bolted two cast-iron plates or electrodes. The cavities formed between the center and side plates are divided by asbestos THE PRODUCTION OF WELDING GASES 23 fabric diaphragms, forming two chambers. The asbestos dia- phragms are clamped directly hy metal and are not held in place by either rubber or cement. In the upper part of the Fig. 8.— Suggested Layout of a 200-Cell Plant. east-iron frame are reservoirs for the electrolyte, from which it is fed to the two sides of the diaphragms. There are also two gas chambers at the top of the frame, which serve as gas traps and gas take-offs, as well as an automatic pressure- 24 GAS TORCH AND THERMIT WELDING controlling dcvico. At the l)Ottom of the frame are com- municating passages which permit the equalization of densities in the electrolyte. Tlie cells are constructed of material which make them practically indestructible. Each cell is provided with an eye-bolt to facilitate handling. The electrodes are provided with a large number of pyramidal projections which greatly increase the area of contact with the electrolyte and facilitate the release of the gases at the generating surfaces. At 600 amp. the production of a plant per 24 hours is 105 cu.ft. of oxygen and 210 cu.ft. of hydrogen per square foot of floor space. A group of cells is shown in Fig. 6. Two typical plant layouts are shofrn in Figs. 7 and 8. Fig. 7 is a layout for a 50-cell plant, with a normal capacity of 5760 cu.ft. of oxygen and twice as much hydrogen. This plant may be put in a corner of an existing building, and requires a space 22 X 24 ft. It is complete with all accessories except gas liolders and storage tanks. Fig. 8 is a separate plant of 200 cells, with all necessary accessories — such a plant as might be installed for public service in cylinders for gas users. Its main dimensions are 53 X 79 ft. and it has a normal capacity of 23,040 cu.ft. of oxygen and 46,080 cu.ft. of hydrogen per 24 hours. The Levin Type of Generator. — The generator made by the Electrolytic Oxy-Hydrogen Laboratories, Inc., Dayton, Ohio, and also of New York, is the design of I. H. Levin, after whom it is named. It is of the unit type, and is made up of a few standardized parts wdiich can be easily assembled. Details of one of the cells are shown in Fig. 9. From this it will be seen that in the main the general construction is the same as others on the market. One noticeable difference from those described, however, is that the electrodes are made independent of the casing, being separated from and securely fixed within the casing by specially designed blocks of asbestos. Each compartment has an independent water feed which also serves as a blow-off device to vent the gases under ab- normal conditions. The surfaces of both the anode and cathode are plated with cobalt, which is said to lower the over-voltage in excess of what the gas electrodes require. The cells are sent out entirely welded, and completely and rigidly assembled, so that they may be filled with electrolyte, THE PRODUCTION OF WELDING GASES 25 connected up, and put into service immediately. Each cell is 6\ in. thick, 25 in. wide and 30 in. high, and weighs 185 lb. AVith the ground supports, porcelain insulators and the piping system above, the total height is 4 ft. 8 in., which l)rings all the parts within the range of normal reach and vision. Even oxroEN offtake: pipe ■ HTDROaFN OFFTAKE PIPE Minimum space fo be connected ^'^n ^ rubber tubing W^i ■ OXYGEN SIGHT FEED INDICATOR HYDROGEN SIOHT FEED INDICATOR INSULATOR., FRAME FOR ASBESTOS ---■ ASBESTOS DIAPHRAGM ■ CATHODE- ASBESTOS INSULATORS PORCELAIN INSULATORS ' Floor Line Fig. 9. — Cross-Section of the Levin Generator Cell. a minimum aisle space of 30 in. will leave ample room for the removal or replacement of any cell. Each cell is intended to operate at a normal amperage of 250, requiring a little over Vio kw. per hour. A battery of 1000 cells will occupy a space 4J X 31 ft. and is claimed to produce 200 cu.ft. of oxygen and 400 cu.ft. of hydrogen per hour. CHAPTER III ACETYLENE AND MEDIUM, OR POSITIVE, PRESSURE GENERATORS Ac tyiene, which is the gas most commonly used witli oxygen for gas-torch welding, is produced by the reaction between calcium carbide and water. It is used because of its large carbon content and also because of its cndothermie properties, which means that it is heat absorbing and energy stored up in its formation is given off again upon dissociation. Calcium carbide and water produce acetylene and slacked lime, tlie formula being: CaC, + 2IL0 = CoHo + CaO(H„0). The calcium carbide when in contact with water is divided and the carbon of the carbide joins witli the hydrogen of the water to form acetylene gas. The calcium of the carbide unites with the oxygen of the water and forms slaked or hydrate of lime, as just stated. Since calcium carbide combines with water in all of its forms it must be protected from moisture in handling. For this reason it is shipped in sealed metal drums or cans com- monly holding 100 lb. From these drums the carbide is placed in generators for the purpose of liberating the gas. Like other gases used for welding or cutting purposes, acetylene may be purchased in cylinders, but it cannot be compressed directly into ordinary steel cylinders with safety. When compressed to as much as 30 lb. per sq.in., it becomes very unstable and liable to explode unless handled with ex- treme care. The heat generated by compression therefore makes this a dangerous process unless means are provided for cooling. The method used in order to compress the gas and make it safe to handle is to fill the cylinders with some porous substance and then fill them with acetone. Acetone is a liquid that has the property of absorbing acetylene about the way sugar does water, and acetylene in this state is commonly known as dissolved acetylene. As acetone takes up acetylene it in- 26 ACETYLENE AND MEDIUM PRESSURE GENERATORS 27 creases in bulk. So suppose a cylinder to be filled only with acetone and dissolved acetylene under considerable pressure. If the cylinder gas valve is opened and some of the acetylene drawn off, the acetone will shrink in volume and leave a place in the cylinder filled with undissolved acetylene gas under pressure. This free acetylene under pressure is very explosive, and easily set off from shock or heat. However, it has been found that the gas will not dissociate when finely divided, and advantage is taken of this, and the cylinder or tank is filled with porous material. For this purpose a mixture of asbestos, charcoal, kieselguhr, and a small amount of cement to hold it together, is packed into the cylinder or tank, providing a finely divided porous filling that prevents dissociation of the gas. After a cylinder has been completely filled with this mixture, slightly dampened, it is baked in an oven until the moisture is completely driven off. It is then exhausted of air and acetone is introduced into the cylinder. Acetylene is then forced into this prepared cylinder, by means of a specially cooled, multiple-stage pump, great care being exercised during the process. Each cubic foot of acetone will absorb 24 cu.ft. of acetylene for each atmosphere (15 lb.) of pressure. However, in actual practice, the quantity of acetone in a cylinder is usually so regulated that the cylinder will contain about 10 times its own volume of acetylene for each atmosphere of pressure that is on the gas. Cylinders are, as a rule, charged to 15 atmospheres pressure at 60 deg. F., so they contain 150 times their own volume when charged. Thus a cylinder that would hold 2 cu.ft. of water when empty will hold 300 cu.ft. of acetylene at 225 lb. pressure, 60 deg. F. The filling material must be so placed in the cylinder as not to settle when handled or shipped, since if it does, a danger- ous pocket will be formed filled with free gas. In handling these cylinders, it should always be kept in mind that the pressure increases as the temperature is increased. The best results are obtained by keeping the cylinders of dissolved acety- lene at about 60 or 65 deg. F. An average cylinder of about 100 cu.ft. capacity will weigh about 85 lb. and one of 300 cu.ft. capacity, about 220 lb. A Davis-Bournonville cylinder, 12 X 36 in., 225 cu.ft. capacity, will weigh, fully charged, about 180 pounds. 28 GAS TORCH AND THERMIT \A ELDING Acetone Injurious to a Weld. — Acetone is very injurious to a weld, and in using gas from a cylinder of dissolved acety- lene, care must be exercised not to use the gas too fast or acetone will be drawn off with it. The allowable rate of cylinder discharge is at the rate of one-seventh of the capacity per hour. That is, a 225-cu.ft. cylinder will supply 225 ^ 7 = 32 cu.ft. per hour. Under emergency conditions this rate may be con- siderably exceeded, but the practice is not economical and is liable to injure the weld. The amount of gas in any cylinder may be found by noting the empty or tare weight stamped on the cylinder, then weigh- ing the cylinder and multiplying the difference in pounds by 14^. In some cases the weight marking may be different, but in any case, each pound by weight of dissolved acetylene is calculated to produce 14^ cu.ft. of gas. To calculate the amount of acetylene used on any job, weigh the cylinder before and after and multiply the difference in pounds by the number just given. Knowing the cost of a cylinder of gas, the cost of the amount of gas used on any job may be easily found. Weighing, which has just been outlined, is the only ac- curate method of computing the amount of acetylene, since gage-pressure readings are affected by changes of temperature. However, gage pressures are a great convenience in estimating roughly how much gas remains in a cylinder. In the case of a 100-cu.ft. cylinder, each 15 lb. of pressure represents, ap- proximately, 6^ cu.ft. of gas, and in a 300-cu.ft. cylinder each 15 lb. of pressure represents, approximately, 20 cu.ft. of gas. There are three types of acetylene generators, the essential differences being the method of bringing the carbide and water into contact. These three methods are : dropping the carbide into the water ; allowing the water to rise slowly into the carbide ; dropping the water onto the carbide. The first is by far the most satisfactory method and the one generally used in the United States, and for this reason will be the only method described. The carbide-to-water generators are of two types, known as positive- or medium-pressure, generators which deliver the acetylene at a pressure of more than 1 lb. and not to exceed 15 lb., and low-pressure generators which deliver gas at a pressure of less than 1 lb. While there are a number of firms making the ACETYLENE AND MEDIUM PRESSURE GENERATORS 29 different types, only the better known makes will be described in detail, since the general principles arc the same. Modern generators of all types are automatic in their action, the feed being regulated by the flow of gas, and they are provided with an ample number of fool-proof device:; which render their opera- tion and handling safe. The Positive-Pressure Generator. — The first positive-pres- sure generator built was designed by Mr. Bournonville in 1906. The present type of pressure generator handled by the Davis- Bournonville Co., is made by the Davis Acetylene Co., Elkhart, Ind. These generators are made in several sizes to meet the demands of various sized shops, and they may be had in 25, 50, 100, 200 and 300 lb. carbide capacity. These numbers in- dicate the weight of the charge of Calcium carbide to be used, the number of gallons of water per charge, and also the number cf allowable cubic feet of gas generated per hour in accordance with the rules of the National Board of Fire Underwriters. The generators of 50 lb. and 100 lb. capacity are standard for repair shops and light manufacturing, while the larger sizes provide for more 'extensive and continuous use in large repair shops and metal-working industries. They are of heavy con- struction with powerful feeding mechanism. They are fre- quently installed as units to build up plants for any require- ments of the oxy-acetylene process. They all have automatic feed with independent power and the quantity of carbide re- maining in the hopper is constantly indicated. The acetylene gas is piped directly from the generator under the required pressure through service pipe lines to the welding stations and is regulated by means of reducing valves fitted with pressure gages to govern the proper working pressure. The size of carbide used in the Davis stationary generators, is designated as 1^ X f in., and it is estimated to produce ap- proximately 4^ eu.ft. of acetylene gas to a pound of carbide. The size of carbide just quoted has reference to the size of screen mesh it will go through or not, that is : 1^ by f in, means that the lumps will all go through a 1-|- in. mesh screen, but none through a f -in. mesh ; carbide quoted as ^ by jV in. size, will go through a ^-m. mesh screen, but not through a jV-in. mesh, and so on. This smaller size is estimated to yield about 4 eu.ft. of gas per pound of carbide. 30 GAS TORCH AND THERMIT WELDING The carbide-to-water generators are designed to hold a gallon of water for each pound of carbide capacity. In practice the carbide is fed into the water only in sufficient quantity to maintain the gas supply at the required pressure. The extreme limit of safety pressure for free acetylene is 30 lb. and in the Fig. 10. — Stationary Type of Positive-Pressure Acetylene Generator. positive-pressure types of generators the pressure limit is placed at 15 lb. When this pressure is reached, the various safety devices begin to act to prevent further generation and conse- quent increase in pressure. As the carbide is fed from the containing hopper it drops down into the tank of water beneath, ACETYLENE AND MEDIUM PRESSURE GENERATORS 31 FeeJing Diaphragm ffo/e -rp/pe nnsac/ I Pipe Tap-. Fig. 11. — Details of 300 Lb. Carbide Capacity Acetylene Generator. 32 GAS TORCH AND THERMIT WELDING and sinks to the bottom. In this way most of the gas is gen- erated close to the bottom of the water tank, and has to rise through a considerable body of water to reach the top. This not only serves to cool the gas but also washes out a large part of the impurities. The carbide used in these generators may be obtained from distributing points in nearly all of the large cities in the country, in sealed drums containing 100 pounds. The acetylene generator shown in Fig. 10 is the Davis standard stationary 300 lb. carbide capacity apparatus. It is 115 in. high, 42 in. in diameter, and weighs about 1000 lb. Details of this apparatus are shown in Fig. 11. In this cut, A is the hopper into which the carbide is introduced from the top. A feeding mechanism at the bottom of this hopper is run by the clock motor B, which is operated by means of the weight C. The feeding device consists principally of a rotating disk with inclined vanes which sweep the lumps of carbide off into the water below. There are two safety-pressure diaphragms for stopping the motor should the gas pressure become too high. The first diaphragm acts when the pressure reaches 15 lb. If this should fail to act properly, the second diaphragm will be actuated a little above 15 lb. pressure. This last dia- phragm operates a positive lock which effectually stops th-e motor and consequently the carbide feed. In addition to this, there is a safety valve at E which is connected to a pipe leading to the outside of the building, so that the gas can escape in safety. At I> is a funnel by which water is run into the tank. As the gas is generated it is carried through the pipe F to the bottom of the flash-back chamber G. This chamber is full of water and serves the double purpose of giving the gas a second vrashing, and acting as a water seal between the service pipe and the gas in the generator. From this chamber the gas passes up through the filter chamber H, which is filled with asbestos that removes suspended impurities. From this filter the gas passes out through valve / into the service pipe and thence to the torches. The handle J operates the agitator and the tank settlings are removed through the gate valve K. In addition to the regular stationary type, the Davis-Bournon- ville Co. makes what is called a "Navy Type." These comprise a complete equipment for use in ship-building yards, railway systems, and large industries. The installations provide acetylene ACETYLENE AND MEDIUM PRESSURE GENERATORS 33 SSHgSBSeWSKiWMSKHfflS^SSSiW^^ , ' Ph O 34 GAS TORCH AND THERMIT WELDING ACETYLENE AND MEDIUM PRESSURE GENERATORS 35 generation with two-pressure, air-excluding generators of large capacity , supplying acetylene under direct pressure to welding and cutting stations and an additional low-pressure supply for compression into portable safety cylinders, with purification sys- FlG. 14. — All-Steel Hand Truck for Holding Gas Cylinders and Welding Outfit. tern, and three-stage specially designed acetylene compressor. Three sizes of the navy type, two pressure, air-excluding acety- lene generators are manufactured, having a capacity of 100 lb., 200 lb., and 300 lb. of carbide each, and with normal output of 36 GAS TORCH AND THERMIT WELDING 100, 200 and 300 eu.ft. of acetylene per hour respectively. When greater capacity than 300 cu.ft. of acetylene per hour is desired, two or more generators may be connected serially with a special purifying and compressing plant. A navy type installation is shown in Fig. 12 and a suggested plant layout in Fig. 13. One of the special features of the navy installation is the low-pressure supply from which the gas is drawn for charging portable cylinders. This is done because in many places around a shipyard or other large plants, it would be practically impossible to pipe the gas or to use a portable generator. The cylinders used are, of course, previously prepared as already outlined. After charging they may be transported to the work in various ways, but for convenience they may be mounted on a small hand truck, like the one shown in Fig. 14. This truck holds an acetylene cylinder and an oxygen cylinder, together with the necessary accessories. The approximate dimensions and weights of the various types of generators made by the Davis-Bournonville Co. are given in Table T. The Buckeye Carbide-Feeding Mechanism. — The phantom view given in Fig. 15 illustrates the feeding mechanism used by the Macleod Co., Cincinnati, Ohio, on their pressure gen- erators. These are made on the same principle as those just described, but this illustration gives a clearer idea of the ar- rangement of the feeding device. The operating motor is driven by a weight which is wound up by means of the handle on top. The way the vanes are set so as to sweep the carbide lumps into the hole in the center of the bottom of the hopper, is clearly shown. Since commercial demands frequently make a portable type of acetylene pressure-generating apparatus desirable, such ap- paratus particularly suited to the requirements of large shops, ship or railroad yards and the like, is made by several firms. The Davis Acetylene Co. makes them in two sizes, of 25- and 50-lb. capacity. These generators are provided with a locking mechan- ism which prevents the operation of the feeding mechanism when the generator is being moved. The Portable Pressure Type. — The portable pressure type of generator made by the Oxweld Acetylene Co., Newark, N. J., and Chicago, 111., is shown in Fig. 16. These generators are ACETYLENE AND MEDIUM PRESSURE GENERATORS 37 1/5 1/5 en ui t/i i3 J2X1 X3-D •V O O O O O MT3 LO LO lO O O ^ ^ ^ ^ T:h " C3 1 1 mO "; " T ' T 1 rO <^ ^S ^ ^ ^ rt< rf X X X X >< o.ti O O O '^ -* ot 1— ( ^— I T-H T— 1 *-H 2 X X K X X O O O '^i '^ to t^ o o o ro CM cs ID \0 00 ^ T-H CN (Vtl CO ^— < *— ' .2-0- X X X X X XXX S« On \D 00 00 to CN ID "-^ Gu p ^xTi „ (« 10 u5 M XI -Q J2 X3 -D J3X5X1 ^^ o o o o o o o o O ID 00 o o \0 O iM >j,0 r<0 ID >0 O CN t— 1 o (/5 (/> ir> sO t^ O ■^ =6 j_ lU 4) ^ ^ ^ ^ c S 00 't O f^ CN tU ^ CN ro fO rf Q *j . . (fi rt a U o PP^ 01 X-g ^ e ^ o o < Z (/5 (/5 ,„ 1 -o J2 jn ">; 1 o o o capaci red) o o o ^ '^ ^ rC) rO r- 00 t-^ ^— ( XXX X X X X X 00 OO ID CD O O -^ OO m ■^ ID 00 CD CN ■.— 1 Tf XXX X X X X X ^ vO fO On O O rh O r*:! rf ID r- ^O CM ^ ro ^ Ui if) tfj ifi iji ifi ifi JDX)X1 JDX5X1 j:iX3 O O O o o o o o 00 O O O O -^ LO o \0 O CN On vO ■•-^ CS On y-i ^—1 ^ CN (/5 t/5 c/5 (/5 cfi -Q JD^ X5X1 JD J3 X 1 ^^ ""^ •-« — ' — * •— ' — " -—^ O O o o o o o o ID ID o O O OO o o ID 00 o ID ID CN ID ■^ r-l CN ^ ^ ^ ^ O -^ OM ID VO t> „ 00 (>4 Ol -^ XXX X X X X X ID \0 \0 *. ^ ^ ^ ^ *^ NO NO t^ O CN T-H T-H tV> 00 ^ ID ^ ^ ^ i c ' O NO O CN ^ r- O — CM CN rt' ID nO "" O VO CN ID nD NO r^ ro ^ t^ T-l ' u. u^ o 1 O 1- en ^ o ^ 2 Purifier, 24 x 36 X '^^ inches Compressor, 31 X 36 X '^^ inches Diameter, Weight, Inches Lb. 30 800 36 1000 42 1200 76 1500 Diameter, Weight. Inches Lb. 16 80 6 200 800 2500 CHAPTER IV LOW-PRESSURE ACETYLENE AND THERMALENE GENERATORS The use of a low-pressure acetylene generator, the pressure of which is limited to 1 lb. or less, makes it possible to use a gasometer for the storage of the generated gas, with its accom- panying advantage of absolute volumetric or pressure control. A phantom view of a typical low-pressure generator, made by the Oxweld Acetylene Co., is shown in Fig. 17. The carbide-feeding mechanism and tank closely resemble that of the pressure type, and with the exception of the gasometer and its mechanism, the rest of the apparatus resemble it also. The possibility of loss of gas through ordinary leakage in the line and connections is practically eliminated with the low- pressure system. The gas bell A provides storage for gas and effectively guards against the loss due to after-generation. Control of the carbide-feed mechanism is accomplished through the rise and fall of the gas bell, giving a very constant gas pressure. The movement of the gas bell gives a dual motor control, first through a carefully adjusted friction brake, which provides for gradual starting and stopping, and, second, through a positive jaw clutch. Special provision is made for shutting off the motor feed when the gas bell is in its lowest position, which might result from a severe leak in the line or connections. Operation of the generator is effected by a small but efficient weight-driven motor B, which is automatically started and stopped in proportion to the amount of gas being used. The motor weights C always lower approximately the same distance for each pound of carbide used, and constitute a reliable in- dication of the amount of carbide remaining in the machine at any time. The generator requires no attention whatever other than periodic recharging with carbide and water. By the use of a positive forced feed, it is impossible for more than the 41 42 GAS TORCH AND THERMIT WELDING proper quantity of carbide to be fed to the water at a time. In order to secure cool, and hence efficient, generation, it is necessary to have not less than one gallon of water capacity per pound of carbide charge. This requirement is met by all of these generators. A complete and efficient system of interlocking safety de- yr FILTER HYDRAUUC BACK-PRESSURE VALVE Fig. 17. — Sectional View of Oxvveld Low-Pressure Generator Showing Interior Arrangement and Direction of Flow of Gas. vices prevents mistakes in operation due to carelessness or for- getfulness when charging the generator. The hydraulic back- pressure valve Avith three distinct water seals prevents the pos- sibility of any oxygen entering the generator. The generator is equipped with an agitator D, which churns the residuum thoroughly, allowing it to flow out freely and LOW-PRESSURE ACETYLENE GENERATORS 43 quickly when the residuum outlet is opened for recharging. These low-pressure generators use the nut, or 1:^- X §-iii. size carbide. The Oxweld Duplex Generators. — The duplex form of gen- erator, shown in Fig. 18, is designed to meet the requirements of many industries, where continuous operation of the generator, without interruption for cleaning or repairing, is essential. It is made up of two of the low-pressure generators, both deliver- FlG. 18. — Oxweld Low-Pressure Acetylene Generator, Duplex Type. ing to the same gasometer. Either generator can be operated independent of the other, allowing one to be used while the other is being recharged. The low-pressure generators are made in a number of sizes, details of which are given in Table II. Eeputable manufacturers always supply detailed instructions as to the care and maintenance of their generators. These should be posted in a permanent place near the generator and should be strictly adhered to in setting up, starting, and refilling. 44 GAS TORCH AND THERMIT WELDING The following general prccantions should always be ob- served : In charging a generator never use anything but the hands or a wooden stick to poke the carbide, on account of the danger of causing a spark. Table II. — Oxweld Low-I*ressure Gkneratoe Sizes Carbide Capacity Single or Gas Capacity Cu Ft Approx. No. of DIMENSIONS Approx. Shipping Weight, Min. Width of Door in Duplex Per Blow- Com- Through Pounds Hour pipes Height Length Width plete Plant Which Gen. will Pass SO Single , 50 3 7' 4" 5' 7" 3' 2" 1400 29y," 50 Duplex 50 3 7' 4" 8' 10" 3' 2" 1800 29'A" 100 Single 100 6 7' 6" 6' 2" 3' 7" 1900 33" 100 Duplex 100 6 7'b" 9' 8" 3' 7" 2400 33' 200 Single 200 12 9' 4" 8' 0" 5'0" 2500 41" 200 Duplex .. . 200 12 9' 4" 13' 0" 5'0" 3500 41" 300 Single 300 18 10' 1" >8' 10" S'V 4100 48" 500 Duplex 300 18 10' I" 14' 1" 5' 4" 4800 48" 500 Single, ,.,., . 500 !0 ir7" 11' 3" 6' 6" 4800 62" 500 Duplex .. . 500 30 ir 7" 18' 0" 6' 6" 6500 62" As far as practicable do all charging, cleaning, adjusting, and manipulating of the generator by daylight. Avoid as far as possible the introducing of air into the generator, or the circulation of air through it. Never leave any opening in the machine open longer than is necessary and never have any two openings, such as the carbide charging door and the sludge valve, open at the same time. Keep the generator full of water at all times, even when not in use. AVhen the sludge is drawn off and the machine flushed out, do not do anything else until it has been refilled as directed in the instructions for recharging it. Never open the carbide filling plug, charge the machine with carbide, do any adjusting or manipulating, or make repairs, unless the generating chamber is full of water, as directed in the instructions for operating the generator. Renew the water in the generating chamber each time any carbide is placed in the hopper. Never run more than one charge of carbide into the generating chamber without refilling with fresh water. Do not experiment. If water in any chamber of the generator should freeze, do not attempt to thaw it with anything but hot water; and then examine the generator carefully for any damage which freezing may have caused. LOW-PRESSURE ACETYLENE GENERATORS 45 If the generator should ever require repair: First, remove, all carbide ; second, drain all water from the generating chamber, then refill the generating chamber as if for recharg- ing; third, disconnect the generator from piping and remove it to the open air, then fill all the compartments with water and thus force out all mixtures of gas and air before applying soldering irons or any tools that could cause a spark. See that the workmen removing or repairing the generator understand this thoroughly. Familiarize yourself thoroughly with the directions for operating your generator, which are furnished liy its makers ; and when recharging the machine follow them in the order in which they are given. Thermalene Generators. — In composition and method of pro- duction, thermalene differs from any gas previously used for welding purposes. It is a combination produced by the de- composition of calcium carbide and hydrocarbon oil, the heat generated by the carbide being used to vaporize the oil. It is the discovery of Karl Friedrich Linus Wolf, of Zurich, Switzer- land, and is handled in this country by the Thermalene Co., Chicago Heights, 111. Thermalene generation is a very unique proposition and was first fully described in the "American Machinist," Aug. 10, 1916. At present thermalene is principally used for welding or cutting metals, in which it is used together with the proper mixture of oxygen. The nature of the gas will be better un- derstood by first giving a description of the method by which it is produced. A phantom view of a generator is shown in Fig. 19. To prepare this generator for operation, water is poured in through funnel A until it will run out of pet-cock B, which fills the easing about two-thirds full. The generating mixture is carried in a tin can or cartridge that is inserted from the bottom into the cartridge chamber, as shown at C. Previous to inserting the cartridge, however, care is taken to pull down the cam lever D, which pulls up the rod and closes the water valve E. This prevents any water entering the cartridge chamber. After the cartridge has been inserted through the door in the bottom of the chamber, the door is closed and the bar F brought up and locked by means of the handwheel G. To generate gas the cam 46 GAS TORCH AND THERMIT WELDING lever on top is pulled up as far as it will go. This admits water from the water chamber H into the cartridge chamber through the center tube to the bottom of the cartridge, from where it begins to work upward. How the Cartridge Is Packed. — The cartridge is packed with Tig. 19. — Phantom View of Thernialene Generator. alternate layers of calcium carbide and crude oil mixed with sawdust. The water as it rises slacks the carbide and generates acetylene gas. The heat caused by the slacking of the carbide layer vaporizes the oil in the layer of mixed oil and sawdust immediately above the layer of slacking carbide and generates LOW-PRESSURE ACETYLENE GENERATORS 47 oil-gas. When the gas pressure as indicated hy the gage is up to 5 Ih., the valve J is opened, which lets the gas enter the hose leading to the storage tank or torch. The pet-cock at K is used to drain off the impurities, such as phosphor-hydrogen or ammonia, which have been separated from the carbide gas in the process of generation. At L is a check valve, which allows gas to escape from the cartridge chamber, but effectually pre- vents any return of either gas or water. As soon as a cartridge is exhausted it is easily removed, carrying all of the dirt and sludge with it. With the use of a storage tank, enough gas is carried to run a torch while a new cartridge is being put in place in the generator, if needed. The pressure limit for welding or cutting is 15 lb. When this pressure is reached, a regulating diaphragm at M automatically cuts off the water feed and stops the genera- tion of gas. This diaphragm is placed at the top of the generator, on the inside, where the gas pressure acts directly on it. The diaphragm is connected to the feed-valve rod in the center, so that as the diaphragm is forced upward by the gas pressure, the water valve is partly or wholly closed. It is so arranged as to be easily set for various desired pressures. As previously stated, the cartridges are made up of alternate layers of calcium carbide and oil-soaked sawdust. Now it is well known that the volume of lime into which calcium carbide is converted by the slacking process is greater than the volume of the carbide. Therefore, when carbide is packed tightly in cartridges, as desirable, the expansion is liable to burst the case, and in some instances might cause the cartridge to jam in the chamber, as well as interfere with the successful working. One of the objects then is to so make the cartridge as to allow for the expansion of the contents. Another is to prevent the oil from coming in contact with the carbide, as this would interfere with the action and decrease the gas output. To summarize, the cartridge components must be so arranged as to promote free action of the carbide; free action of the heat evolved from the carbide on the volatile substance ; to prevent action of moisture on the carbide layers not being acted upon purposely ; and to promote the free discharge of the generated and volatilized gases. All of these points, as well as a number of others, have 48 (!AS TORCH AND THERMIT WELDING been considered in the makeup of the cartridges, as shown in detail in Fig. 20. As made at present for commercial purposes, there are four sizes corresponding to the different outfits. The cartridges consist of a tin can of suital)le size, the smallest being 4^g in. in diameter and 8 in. high, weighing 6 lb. and having a gas-producing capacity of 25 cu.ft., sufficient to supply a welding torch from four to five hours. The largest size is 92- in. in diameter and 16 in. high, weighing 40 lb. This has Carbide 'Spacer Sandvsf covered wfih Cloih ■Screen ■k: c Fig. 20. — Details of Thernialene Cartridge. a gas production of 200 cu.ft., or eight times the capacity of the smallest size w^hich has been previously mentioned. Referring now to the illustration, a cylindrical screen A is first placed in the can to be filled and then a layer of carbide is placed in the bottom of the can around the screen tube. An unglazed carboard disk B is next placed on top of the carbide. A spacer T', made of thin metal bent so as to lie edgewise, is placed on top of the cardboard and then a disk of screen is put over it. Sawdust impregnated with crude oil in a cloth sack made of two cloth disks sewed together is laid on the screen. LOW-PRESSURE ACETYLENE GENERATORS 49 On this is another screen, then a spacer and a cardboard disk, and so on to the top, ending with a sack of oil, soaked sawdust covered with a piece of screen that has no hole in the center for the cylindrical tube. The end of the can is closed with a cover for handling, which is removed before placing in a generator. As previously mentioned, the water feeding in through the valve in the cartridge chamber drips down through the cyjin- drical screen tube and starts slacking of the lower layer of carbide, the heat of which vaporizes the oil in the sack above it. The cardboard disks, while strong enough to hold the layers firmly in place while dry, begin to soak up as soon as the feeding starts, and consequently become soft, so as to give way under the pressure of the expanding carbide, allowing it to be forced into the spacing between the carbide and the oily saw- dust. This space is so calculated as to keep the various layers firmly in place until the cartridge is exhausted. Another thing is that, as any liberated steam or water vapor must pass through the absorbent cardboard as well as the oil, before it reaches the next layer of carbide, action of the steam on the carbide above is prevented. This insures that the respective layers of carbide will not be acted upon until the water becomes level with them in turn. In consequence, a cartridge can remain in a generator a long time without becoming spent. The screen disks on top and bottom of each layer of oily sawdust furnish efficient volatilization and egress for the gas. By the method of producing thermalene, the heat evolved by the generation of acetylene is absorbed, at the place of generation, in the production of the oil gas. This utilization of heat serves to keep the temperature down, since the heat generated is used and dissipated by the latent heat of the oil, so that radiation and absorption by water is not necessarily depended upon. The gas combination that results passes out through the T-pipe ends and bubbles up through the water into the upper part of the generator. The low temperature causes the impurities to drain back into the chamber, from which they are easily removed. The layers of carbide and oily sawdust are so proportioned as to cause only the vaporization of the lighter oils, such as benzine, naphthalene, kerosene and the like. The temperature is not high enough to vaporize the tar oils, 50 GAS TORCH AND THERMIT WELDING as these cue Iieuvy and give a deposit of lampblack. These heavy oils are therefore not utilized, but remain in the cartridge. The temperature in the cartridge is maintained between 200 and 300 deg. C. (392 to 572 deg. F.), depending upon the rapidity with which the gas is being used and the amount which is generated and delivered. It is not intended that an actual boiling take place at any time, for if the temperature is too high there will be a vaporization of the heavy oils, causing de- posits in the pipes and also an increase in the impurities. The gases passing through the pipes cooled by the water of the generator are kept between 60 and 70 dog. F. In the passage through the cooled pipes the impurities are removed in the following manner : Acetylene has a comparatively high specific heat, so that its rate of cooling when passing through the pipes is low. The specific heat of oil gas is, however, only one-eighth of that of acetylene, so that its cooling effect will be eight times as great. Now if the two gases are passed together along a cooling surface, the temperature of low specific heat will de- crease rapidly and cause a rapid lowering of the temperature of the other gas. This causes a deposit of the vapors suspended therein. So this action results in the deposit of the sulphur, j)hosphorus and silicon compounds, and of the ammonia. In order, however to bring about this precipitation the temperature must be sufficiently low — that is, as stated above, between 60 and 70 deg. F. If the gas finally issuing from the pipes and water was too hot, the impurities would not be thrown out. In this process the acetylene and oil gas, generated and cooled, will combine in the pipes after the impurities are re- moved. It is important, however, that the impurities be removed and the product sufficiently cooled, since no real combination will take place at too high a temperature. For instance, if the water is above 180 deg. F. the oil burns and no proper com- bination results. The combined gas produced, named thermalene, possesses marked characteristics that distinguish it from oil-gas, from acetylene, and from the usual mixture of the two. The density is greater than air, being 1.1 taking air as unity. The issuing gas can be seen to sink when thrown through a sunbeam. The specific heat is low, being a little over one-eighth of that of acetylene. Thermalene liquefies at between 1400- and 1500-lb, LOW-PRESSURE ACETYLENE GENERATORS 51 pressure per square inch at room temperature, and in its liquid state is nonexplosive and stable. A very noticeable thing is the odor, which is not at all like the odor of either acetylene or of oil-gas, but is a soft, sweet smell, not strong or offensive in any way. The color is white, but with a predominating pro- oortion of the red and yellow parts of the spectrum. The Fig. 21. — Thermalene Outfit for Shop Use. maximum flame temperature, according to H. McCormack, pro- fessor of chemical engineering. Armour Institute of Technology, Chicago, 111., has been found to be about 6500 deg. F. The high density of the gas has a number of advantages. It has more body than acetylene and does not need so much oxygen. More- over, it mixes better with oxygen. It does not explode as 52 GAS TORCH AND THERMIT WELDING readily as acetylene, so can be mixed with greater proportions of air. In a Bunsen burner it is possible to mix as much as 32 per cent of air without causing a flareback. It can readily be turned down without causing a flareback, and it can be used with a Welsbach mantle to advantage. The upper and lower explosive limits are 12 per cent and 30 per cent. It averages approximately 13.97 cu.ft. per pound. Some Advantages of Thermalene. — When used for welding and cutting, thermalene has numerous good points. It does not Fig. 2:i. — Portable Thermalene Generator Outfit. require an excess of oxygen, and the flame, therefore, produces a soft weld, especially in cast iron. When welding it is notice- able that less sparks are thrown ofl than when using acetylene. It can be used at a lower pressure also, owing to its greater calorific value. Owing to the removal of the various impurities, there are no corrosive effects on fittings, nor poisonous effects. It is also for this reason that there is little or no danger of explosion. Neither does the spent cartridge give off explosive gases, for the reason that the gases liable to cause explosion are separated and drained off from the generator chamber. Cor- LOW-PRESSURE ACETYLENE GENERATORS 53 rosion of interior walls, due to water action, is prevented by the oil vapor which is always present and forms a protecting and scaling effect throughout. A thermalene generator may discharge its gas into a storage tank, or the gas may be used direct from generator to torch. A welding or cutting outfit, suitable for shop use, where it will not need to be moved frequently, is shown in Fig. 21. Here the gas is used direct from generator to torch. The smallest generator like the one shown is 10^ in. in diameter, stands 3 ft. 9 in. high, and weighs about 60 lb. An apparatus mounted on a truck is shown in Fig. 22. As a rule a No. 2 generator is used for this purpose. This size is 16 in. in diameter, stands 6 ft. high, and will produce 70 cu.ft. of gas with one charge, the generator weighing 225 lb. and the storage tank 120 lb. The largest size generators will produce 200 cu.ft. of gas per charge, and they may be coupled in batteries where a large amount of gas is desired. These can be arranged so that part of them can be recharged while the others are working, or storage tanks of sufficient capacity may be used to allow for recharging. CHAPTER V GAS TORCHES USED FOR WELDING The gas torches used for welding in the United States may be divided into two general types, known as medium-pressure and low-pressure torches. Each type is made in a number of sizes, and each size is usually provided Avith a number of in- terchangeable tips for producing flames of different size. The medium pressure torches arc also known as positive- pressure torches, and to avoid misunderstanding, they will be referred to as positive-pressure torches hereafter. In these torches, using acetylene and oxygen for welding, the acetylene pressure will range from 1 to 14 lb. and the oxygen pressure from 1 to 24 lb. per square inch, the pressure employed de- pending on the thickness of the metal being Avoided, the make of torch, and the tips used. The pressures given may even be exceeded in some exceptional cases. A sectional view of a typical positive-pressure welding head is shown in Fig. 23. The oxygen enters at A and the acetylene enters at B. The oxygen goes to the small chamber C and thence out through the center hole. The acetylene goes to the chamber D and also out through the center hole, the two gases starting to mix at the point E, and as they pass out through the channel F to the end of the tip, they are thoroughly mixed. In this illustration, the removable tip is indicated by G. This m.ake of tip has a conical seat and held in its place by means of the lock collar H. Made in this way, there are no threads on the tip itself, although the practice varies in different makes. The low-pressure torch is also known as the injector type. In this type of torch, the acetylene, or other gas, is supplied under a pressure of a few ounces up to 1 lb., but the oxygen may have a pressure of from 5 to 30 lb. per sciuare inch, accord- ing to the size of tip being used. In some cases the oxygen pressure may be either higher or lower than the figures given. 54 GAS TORCHES USED FOR WELDING 55 A sectional view of a typical low-pressure torch is shown in Fig. 24. In this torch the acetylene, or other gas, enters at A and the oxygen at B. The acetylene goes to the chamber C from which it is sucked by the oxygen pouring out through the nozzle at D, and it is carried along with the oxygen into the mixing chamber E in the tip of the torch. From this chamber the gases issue thoroughly mixed and ready for com- bustion. As they qualify for classification as either positive-pressure Figs. 23 and 24. — A Typical Positive-Pressure Welding Torch and a Low- Pressure or Injector Type of Welding Torch. or low-pressure types of torches, the various makes of each type differ principally in form, the general principles of action remaining the same. A few examples of the different makes of positive-pressure torches will be shown first, and these will be followed by others of the low-pressure or injector type. A standard form of a positive-pressure welding torch, made by the Davis-Bournonville Co., is shown in Fig. 25 and in detail in Fig. 26. A number of torches used for various purposes are shown in Fig. 27. In this illustration, A is a small lead-burning torch. B is a midget torch used for welding very light sheet 56 GAS TORCH AND THERMIT WELDING metal for manufacturing purposes, sucli us the seams of cooking utensils, aluminum ware, etc. It weighs about 8 oz., is 10 in. long, and may be used on sheets up to ^ in. thick. C is a small- size torch for metal from V32 to Vie ii^- tliick, weighs 18 oz., is 14 in. long and uses oxygen pressures, with different tips, of Fig. 25. — The Davis-Bournonville Style C Positive-Pressure Weldiug Torch. Carburetinq device which posH-ivelycino/ [infimai-eiy mixes the iwo gases in proper proportion _ . ^-. .OXYGEN ■ Conical \ jiMii nil f— t- Orounc^ i^^ ' Seaf"-^ Ox/(^en neec^/e Valve \ 'ACETYLENE '^The two gases strike together at right angles creating a vortex which insures intimate mixture Acetylene neeolle Valve The oliameters of the parts in the carbureting alevice are proportionect to each size of tip, to deliver proper volumes of gas for each size of flame proaucec^. ^"Luminous Cone of flaime Secondary reaction. Hydrogen and carban monoxide burn, fa/ FOR riACHINE WELDING niDCET TIP czzHi^iD OXY.-HYDRIC WELD TIP =S OXY- ACET WELD TIP c — Zl 1" LONG OXY- ACET WELD TIP nULTIPLE FLAME OXY -ACET WELD TIP C 1 \^l SMALL STYLE "CWELD TIP c=:a3D LARGE STYLE "C WELD TIP cmOZBDo L,-,RGE "C" WATER COOLED oi ff::^ SHALL't TIP FOR CITY GAS STYLE "G" WELD T;P W \ ^ c==XBD CBo OD HID C=ZZI33 c:^=a> FOR CI ■ ■■ GAS flULTIPLt FLAME CHICAGO STYLE Fig. 28. — Different Kinds of Tips Used with Davis-Bouinonville Welding Torches. and are used with style C torches. These pressures serve only as approximate guides and are not to be taken literally in practice. The Prest-0-Lite Torch. — The torch shown in detail in Fig. 29 is made l)y the Prest-0-Lite Co., Indianapolis, Ind., but is handled by the Oxweld Acetylene Co. This torch is fitted with a long stem through which the gases pass and are thoroughly mixed before issuing from the nozzle. The stem is fitted to the mixing chamber by means of a union nut, which permits the operator to point the welding tip in any direction, without changing his method of holding GAS TORCHES USED FOR WELDING 59 the torch. This is particularly advantageous for vertical and overhead welding. Both oxygen and acetylene inlets on the torch are fitted with fine-adjustment control valves. The one on the oxygen supply is so placed that the operator while working can make any slight necessary adjustment with the forefinger of the hand that grips the torch. The handle of the torch is fitted with anti-fireback chambers for both gases, filled Avith a special material through which it is impossible for a flame to pass. 'iABLE III. — Approximate Gas Pressures for Davis-Bournonville Style C Welding Torch, With Nos. 99 and 100 Tips Thickness Acetylene Oxygen Tip of Metal Pressure Pressure No. Inches Lbs. Lbs. 00 /Very\ I Light] 1 1 c 1 2 1 ^-M 1 2 2 M-^ 2 4 3 ■h-y% 3 6 4 y%-% 4 8 5 H-'A 5 10 6 %-% 6 12 7 A-'A 6 14 8 'A-Vs C 16 9 Vs-H 6 18 10 M-Up 6 20 11 / Extra\ \Heavy/ 8 22 12 8 24 The torch is easy to dismantle, as all parts are screwed together on metal-to-metal seats and no packing or solder is used at any joint. For extra heavy work, a special stem 22 in. long is fur- nished, and for close work a 5^-in. stem may be had in addi- tion to the regiilar size. The regular stem has seven tips, the largest four being of copper which will stand up against the intense heat radiated from the work better than any other metal. The smaller tips are of a special alloy. A similar torch is also made for very light work which weighs only 3 oz. complete. In the detailed view given of the torch, A is the hose nipple 60 GAS TORCH AND THERMIT WELDING through which the oxygen passes. At B is a set of 40 strainer cloths for the oxygen filter chamber; C is the oxygen needle valve; D is the acetylene hose nipple; E is the acetylene Complete Oxygen I Valve Assembly C ^40 Pieces Complei-e Acefy/ene valve Assembly Fig. 29.— Details of Presto-0-Lite Type H Welding Torch. Fig. 30.— Torches Made by the General Welding & Equipment Co. pxyc^en Tube ^ M C Acel-ylene Tube'' F F F Fig. 31.— Details of Torch Shown in Fig. 30. needle valve; F is the acetylene filter-chamber cartridge; G is the stem. The seven tips are indicated from IH to IH, and their openings can be determined from Table IV. This table is especially valuable in that the nozzle openings are shown in GAS TORCHES USED FOR WELDING 61 regular twist-drill sizes, furnishing an easy method of com- parison. The two tips given at the bottom of the table are extras, used for heavy work. The figures quoted in the table are based on straight work on steel plate, beveled when over I in. in thickness and welded without preheating. Table IV. — ArpiioxiiiATE Welding Results With Type H, Peesto-O- LrrE Touch Tip .No. Tip Drill Size Gas consumption Cu. Fl. per hour Thickness of Metal Blow-pipe pressures Lbs. per sq. in. Lineal feet welded per hour Oxygen Acetylene Oxygen Acetylene IH 69 3 to 4 3 to 4 ^■. to i;^, in. 2 to 3 2 to 3 30 to 35 2H 60 6 to8i 6 to 8 i'ii lo ii in. 2 to 3 2 to 3 24 to 32 3H 55 10 to 12J 10 to 12 I to 3=5 in 3 to 4* 3 to 4 12 to 16 4H 52 12 to 21 12 to 20 A 'o s'i in- 4 to 6 4 to 5 9 to 12 5H 49 18 to 28 18 to 26 1 to i"c in 5 (o 7 5 to 6 6 to 8 6H 44 24 to 40 24 to 38 I to -j'o in. 8 to 11 8 to 9 41 to 6 7H 35 35 to 54 35 to 50 I in. and up 10 to 15 10 to 14 2 to 3 •7J 35 35 to 54 35 to 50 ?. in. and up 9 to 13 9 to 12 Not used on plates *8J 31 40 to 60 40 to 55 I in. and up 9 to 14 9 to 13 Not used on plates The General Welding Co.'s Torch. — A welding torch made by the General Welding and Equipment Co., Boston, Mass., is shown in Fig. 30. Each torch is furnished with nine tips, affording a range equal to all ordinary welding jobs. In addi- tion, stems of different lengths may be had. In the illustra- tion, A is the body of the torch; B is the mixing chamber; C is a short stem, the use of which makes the entire torch 16 in. long; D is a medium stem, making the torch 22 in. long; E is a long stem, making the total length of the torch 30 in. Details of this torch are shown in Fig. 31. Here the acetylene inlet is shown at A and the oxygen inlet at B. The body of the torch is indicated by C, and D is the mixing chamber ; E is the stem, and F various shapes of tips. Imperial Torches. — Another long-stemmed torch is shown 62 GAS TORCH AND THERMIT WELDING in Fig. 32. This is made by the Imperial Brass Manufacturing Co., Chicago, and differs but little from the one just shown. The gas valves, however are placed at the forward end of the body. Like most of the other torches, these may be used not only for oxy-acetylene, but also for oxy-hydrogen welding work, special tips and regulators being made for the purpose. For oxy-acetylene, the gas pressures are approximately the same Fig. 32.— Tho Imperial Type B Welding Torch. as for other makes of torches. For oxy-hydrogen, the pressures used for various thicknesses of metal and different tips are given in Table V. However, neither the size of the nozzle holes nor the amount of gas used per hour are given. This firm also makes a three-way torch which in outward appearance does not differ from the one shown. It is intended to use a com- bination of acetylene, oxygen and hydrogen. The method used Table V. — Pressure of Gas Used in I>rpERiAL Oxy-Hydrogex WIELDING Torches Tl ickness of Metal Pressure, LI). Weldin- Tip No. tn be Welded, In. Oxygen PI ydrogen III 1/64 to 1/32 10 10 2H 1/32 to 1/lG 12 12 3H 1/lG to 1/4 15 15 4H \/\ to 1/2 20 20 r.i-i 1/2 in. niid up o- 25 is to couple the acetylene and hydrogen hose by means of a Y mixing valve from which the two gases are conducted by a single hose to the torch, the oxygen hose being coupled in the usual way. This concern does not recommend the welding of steel above \ in., or cast iron above '\ in. in thickness with oxy-hydrogen. For light sheet steel, and especially aluminum, oxy-hydrogen has some advantages, provided the hydrogen can be obtained at reasonable rates. The combination of oxygen- GAS TORCHES USED FOR WELDING 63 aectylciie-liydrogeu, however, has a claimed temperature of about 5000 deg. F., which is about half-way between that of oxy-hydrogen and oxy-acetylene. It is also claimed that this flame possesses all the advantages of both the double combina- tions. The same tips are used as for oxy-acetylene, and only a small percentage of acetylene is needed to give a cone-shaped flame of far greater visibility than that of the oxy-hydrogen flame. The low visibility and long flame of the oxy-hydrogen flame are always a handicap in welding to any operator used to employing the oxy-acetylene torch. The approximate pres- sures employed Avhen using the Imperial three-way outfit are shown in Table VI. Table YI. — 1'kessukes of Gas Used in Imperial Thkee-Way Welding Touches Oxygen, Acetylene and Hydrogen Oxygen, Welding Tip, Thi^ L'kness of ]Metal, Pressures, Lb. No. to be Welded, In. Oxygen. Acetylene, Hydrogen. IT 1/32 3 2 f^ 2T 1/16 5 3 3 3T 3/32 6 4 4 4T 1/8 7 5 5 5T 1/4 8 6 6 6T 3/8 9 7 7 7T 1/2 10 10 10 8T 5/8 12 12 12 9T 3/4 14 14 14 lOT 1 in . and over IS 15 15 Calculating Amount of Gas. — It should always be borne in mind, when consulting a table where only pressures are. given, that these pressures do 7iot signify the amount of gas used, and that such pressures apply only to the particular make of torch mentioned. It might be possible for two different torches to use gases at exactly the same pressures as far as the gages indicated, and yet have one of these torches use several times the amount of gases used by the other. In order to make it possible for a user to estimate the amount of either oxygen or acetylene his outfit is consuming, three methods of estimating are given here. These methods are taken from the Prest-0-Lite instruction book. Other methods have previously been given 64 GAS TORCH AND THERMIT WELDING in the descriptions of the different gases and their production. The Prest-O-Lite methods are : Method 1. — Weigh your acetylene cylinders before starting work. Weigh again after the job is completed. Note the difference in weight in ounces, and multiply by 0.9 ; result equals the cu.ft. of acetylene used. When UJiing Prest-(;)-Lite torches multiply the acetylene used in cu.ft. by 1.1 ; the result equals the eu.ft. of oxygen used. Method 2. — Take readings of oxygen cylinder pressure gage in atmospheres before and after. For 100-cu.ft. cylinders the difference of readings in atmospheres multiplied by 0.S3 equals the oxygen consumption in cu.ft. For 200-cu.ft. cylinders the difference in readings in atmospheres multiplied by 1.G7 equals the ((jnsumptiou of oxygen in cu.ft. For 2.50-cu.ft. cylinders the difference in readings in atmospheres multiplied by 2.08 equals the consumption of oxygen in cu.ft. When using Prest-O-IJte torches, multiply the oxygen used in cu.ft. by 0.91 ; the result equals the acetylene consumption in cu.ft. Method 3. — Measure drill size of orifice in torch tip. using standard drills for measuring. Area of orifice in sq.in. multiplied by S3 ecjuals the acetylene con- sumption in cu.ft. per minute. Area of orifice in sq.in. multiplied l)y 91 equals the oxygen con- sumption in cu.ft. per minute. Note the minutes the torch is iu use and use the above figures to estimate gas consumption. Remember, the acetylene consumption cannot I)e accurately esti- mated from the pressure gage readings. In order to simplify the calculations in Method 3, the areas of the various orifices made with numbered drills are given in Table VII. This table was calculated by K. II. Condit, man- aging editor of the American MacMnisf, for both square inches and square millimeters. The Reg-o Welding Torch. — The Rego welding torch is made by the Bastian-Blessing Co., Chicago. The claim is made for this torch that it will not flashback under any condition — even if the tip is immersed in the molten metal, or if the head and tip are heated to a cherry red. The elimination of the flashback is accomplished, not by check valves but by balancing the pressure of tJie gases used. The tips used are of one-piece nickel-copper composition. This gives a harder tip than copper alone. The gases are mixed in the tip as shown in the illustration Fig. 33. By means of an expansion chamber within the tip itself, the gases are reduced in velocity as they come from the GAS TORCHES USED FOR WELDING 65 Table VII. — Areas of Drills froji 1 to SO Size in Sq.In. and Sq.MM. Manufacturers Standard Size of Size of Drill Area in Area in Drill Area in Area in No. in In. Sq.In. Sq.Mm. No. in In. Sq.In. Sq.Mm. 1 0.22S0 0.04083 26.35 41 0.0960 0.007238 4.670 ^ 0.2210 0.03836 24.77 42 0.0935 0.006860 4.426 3 0.2130 0.03563 22.99 43 0.0890 0.006221 4.014 4 0.2090 0.03431 22.14 44 0.0860 0.005809 3.748 5 0.2055 0.03316 21.39 45 0.0820 0.005281 3.406 6 0.2040 0.03269 21.09 46 0.0810 0.005153 3.325 7 0.2010 0.03173 20.47 47 0.0785 0.004831 3.117 8 0.1990 0.03110 20.06 48 0.0760 0.004536 2.926 9 0.1960 0.03017 19.46 49 0.0730 0.004185 2.700 10 0.1935 0.02940 18.97 50 0.0700 0.003848 2.483 11 0.1910 0.02865 18.48 51 0.0670 0.003526 2.275 12 0.1S90 0.02806 18.10 52 0.0635 (».()( »:u 67 2.043 13 O.lSoO 0.02688 17.34 53 0.0595 (».(l()2781 1.795 14 0.1S20 0.02602 16.79 54 0.0550 0.002376 1.533 15 O.ISOO 0.02545 16.42 55 0.0520 0.002124 1.370 16 0.1770 0.02461 15.88 56 0.0465 0.001693 1.092 17 0.1730 0.02351 15.17 57 0.0430 0.0014.52 0.9368 IS 0.1695 0.02256 14.56 58 0.0420 0.001385 0.8930 19 0.1660 0.02164 13.96 59 0.0410 0.001320 0.8510 20 0.1610 0.02036 13.14 60 0.0400 0.001257 0.8115 21 0.1590 0.01986 12.81 61 0.0390 0.001195 0.7710 22 0.1570 0.01936 12.49 62 0.0380 0.001134 0.7316 23 0.1540 0.01863 12.02 63 0.0370 0.001075 0.6936 24 0.1520 0.01815 11.71 64 0.0360 0.001018 0.6568 25 0.1495 0.01755 11.32 65 0.0350 0.000962 0.6207 26 0.1470 0.01697 10.95 66 0.0330 0.000855 0.5516 27 0.1440 0.01629 10.51 67 0.0320 0.000804 0.5187 28 0.1405 0.015.50 10.00 68 0.0310 0.000754 0.4865 29 0.1360 0.01453 9.374 69 0.0292 0.000669 0.4316 30 0.1285 0.01296 8.361 70 0.0280 0.000615 0.3968 31 0.1200 0.01131 7.297 71 0.0260 0.000.531 0.3426 32 0.1160 0.01057 6.819 72 0.02.50 0.000491 0.3168 33 0.1130 0.01003 6.471 73 0.0240 0.000452 0.2916 34 0.1110 0.009677 6.243 74 0.0225 0.000398 0.2.565 35 0.1100 0.009503 6.131 75 0.0210 0.000346 0.2232 36 0.1065 0.008908 5.747 76 0.0200 0.000.314 0.2026 37 0.1040 0.008495 5.481 77 0.0180 0.000254 0.1639 38 0.1015 0.008092 5.221 78 0.0160 0.000201 0.1297 39 0.0995 0.007775 5.016 79 0.0145 0.000164 0.10.58 40 0.09S0 0.007543 4.866 80 0.0135 0.000143 0.09226 mixing chamber, and just before they issue from the end of the tip. This produces a ''soft" flame which melts the metal without blowing it away. 66 GAS TORCPI AND THERMIT WELDING The Oxweld Low-Pressure Torch. — The Oxwolcl low-pres- sure torch is of the true iujector type. One of this make of torch is shown in Fig. 34 and in detail in Fig. 85. In this torch the acetylene is drawn into the combining tube by the injector action of the high-pressure oxygen jet, in the proper quantity to form what is known as a neutral flame ; that is, one Fig. 33.— The Rego Welding Torch. not having an excess of oxygen or acetylene. A torch of this type may be used with either a low- or positive-pressure acety- lene system, although primarily designed for low-pressure acetylene, which means at a pressure of 1 lb. or less. Ten hard-drawn copper tips, shown in Fig. 36, are regularly sup- plied for this torch, and bodies may be had to hold the tips Pig. 34. — The Oxweld Low-Pressure Welding Torch. 3 3 13 1 T6> 64 > 32) 8) 16» 4' i 5. and ,2)8 at 67| or 90 deg., although the regular angle is 45 deg. The tips shown are numbered 2, 3, 4, 5, 6, 7, 8, 10, 12 and 15 and are intended for use on metal 5 in. thick and up, respectively. A very light torch, weighing 9 oz. and known as model G, is shown in Fig. 37. This is intended for very light sheet metal, instruments, jewelry and the like. In order to secure a torch of minimum weight, the oxygen and acetylene regulating valves GAS TORCHES USED FOR WELDING 67 have ))ecn removed from the body of the torch and incorporated ill a separate valve block which may bo fastened in any con- venient position near the operator. Another very light torch is shown in Fig. 38. This is suit- Oxygen Va/ve^, ,VcPi/ve Body Oxygen Tube^^ Oxygen Hose- Connection Acetylene Hose Conn. Oxygen Chamber 'Injector "Acetylene Chcpimber 'Mixing ChitPimber Acet-ylene Voilve Fig. 35.— Details of the Oxweld Welding Torch. able for metals up to \ in. thick. In addition to the usual practice of placing needle valves in the rear body, there is also incorporated in the torch head a needle valve which gives minute control of the flame with each size of tip. Three sizes of in- FiG. 36.— A Set of Oxweld Welding Tips. terehangeable copper tips are supplied. It is 10 in. long and weighs 10 oz. The Oxweld water-cooled machine torches are shown in Figs. 39 and 40. Fig. 39 is a single-jet, known as type W-8, 68 GAS TORCH AND THEIUMIT WELDING o o 3 o u a> Ph g "=" 2 CO P-,h:i .is *» %5 iOOOO'-iC0«0-^Tt< OOOOOOOO^Cl OOOOO'-iiNiO'-HCOO i-HCOO ^ (MCOOOlCOt^t-iOC^ t^. CO(N(NCO(N(NCOCCCO'*CO CO TJH CO 00 O •^ 00 CO ■* o CD .-( rH rH (N CC lO CO iOiOcOI>000(McDOOOI> CC^OOOOiOOll^COC^OS ■-I —< '-i(N CO O CO :^ C0(Nr^-<^i-(05t*'1*C0>— i-l 00O(M»O 'O I— 1 1— 1 1—1 o 00:-^ 0)5 gs \J Si 1; ■t^ Js^ ^ lA t\j "^vj J3 Si 'J- V7 ^1 5^ D )(IB OD s: kg §S m ^ 78 GAS TORCH AND THERMIT WIELDING liydi'ogx'ii. Torch C (No. 471) is a circle cutting torcli tittccl witli a 15-in. radius rod adjustable for various sizes of circles. It uses either oxy-acetyleiie or oxy-hydrogen for heating, and takes all standard size tips. It is the standard torch for nick- ing billets for breaking. It may be had fitted with special adjustable holder, rack and pinion, for machine cutting. Tahle X. — Gas Pressures Used With the Davts-Bournonville Style C Cutting Torches, Using Style 12 Tips Thickness Acetylene Oxygen Tip of Metal Pressure Pressure No- Inches Lbs. Lbs. 1 Vs 3 10 1 % 3 15 1 M 3 20 1 % 3 20 2 Va 3 10 2 Vi 3 20 2 M 3 30 2 1 3 35 3 1 4 30 3 1^ 4 40 3 2 4 50 3 3 4 60 4 3 5 60 4 4 5 70 4 5 5 85 4 6 5 100 5 6 6 90 5 7 6 100 5 8 6 125 5 10 8 150 Torch P, No. 640, is a machine cutting torch for use with the different cutting machines made by the Davis company. It is fitted with an electric switch and shut-off valve which auto- matically starts the cutting-machine feed wheii the oxygen cutting jet is turned on, without the necessity of readjusting pressures or the heating flame. It uses all standard-size cutting tips and special Oxygraph and Radiograph tips. A larger and GAS CUTTING-TORCHES 79 heavier torch of similar dosign and construction for oxy- hydrogen machine work is shown at E. .This is known as No. 1314. Larger sizes, or special torches used for cutting, are water-cooled. Tips of various styles, for different pur- poses, which may be used with the torches mentioned, are shown in Fig. 47. The approximate oxygen and acetylene pressures used in Fig. 48.— The Oxweld Cutting Torch, Model B. the Davis style C cutting torches, using style No. 12 tips, are given in Table X. As in welding, these pressures are only general guides for the make of torch mentioned, and the skilled operator usually adjusts his flame regardless of the Oxycjen Pre-heat-incj Ve/s— - Fig. 49.— Details of the Oxweld Cutting Torch, Model B. tables given, since a neutral heating flame is essential at all times for satisfactory results. Oxv/eld Cutting Torches. — An Oxweld low-pressure or in- jector type of cutting torch is shown in Fig. 48. This is their model B. Details are shown in Fig. 49. In this torch, the cutting jet is entirely surrounded by the preheating flame, as the oxy-acetylene is delivered through six openings arranged 80 GAS TORCH AND THERMIT WELDING in a circle around the orifice for the oxygen jet. This arrange- ment makes it possible for the preheating flame to always precede the cutting jet, no matter in what position the torch is held, or in wliatever direction the cut is made, be it hori- zontal, transverse, circular, elliptical, toward or away from the operator. In so woi'king, the operator does not have to sliift his postion or turn the torch. This is especially valuable in wrecking steel structures, removing risers from steel cast- ings, or cutting steel scrap, especially where places difficult of access are encountered. The preheating flame is produced in practically the same way as in tlie welding torch previously shown. There is a separate valve for controlling the oxygen to the preheater, wliich enables tlie operator to secure close adjust- ment and avoid waste of gas. The oxygen-jet valve is of the Fig. 50. — The Oxweld Cutting Torch with a Eivet-Head Cutting-Nozz]e. plunger type, which is so constructed that its movement pro- duces no tendency to deflect tlie cutting jet from the line of the cut. The location of the valve lever is on top of the handle and its motion is in the direction of the vertical center plane of the torch. The valve is held open for continuous cutting, when desired, by a simple but effective button-like latch, which may be instantly engaged or released by a slight movement of the thumb. The external nozzle is furnished with a copper tip. The internal nozzle is held in place by tightening the external nozzle. To remove the former it is only necessary to unscrew the external nozzle. This torch is regularly furnished with four interchangeable tips, for cutting up to 1 in. ; from 1 to 3 in. ; from 3 to 6 in. ; and from 6 in. up. A model-B torch fitted with a special rivet-head cutting nozzle is shown in Fig. 50. Another form, known as model GAS CUTTING-TORCHES 81 Fig. 51.— The Oxweld Cutting Torch for Ship Work. Fig. 52.— The Oxweld Staybolt Cutting-Torch. 82 GAS TORCH AND THERMIT WELDING C-6, is sliown in Fig. 51. This was made to meet tlie demand for a light, rugged and adaptable cutting torch for work on double bottoms and below decks of ships. In general, it closely resembles the other models. It weighs 2^ lb. and is 20 in. overall.. For cutting inner and outer shells of locomotive fireboxes, where length and slenderness is necessary, the staybolt cutting torch shown in Fig. 52 has been made. This is, however, merely a special form of the model B. The long "stem" is made up of three gas tubes as shown. Tliis torch is regularly furnished in 42, 54, 69 and 84 in. lengths to suit the needs of the user. Tlie 84-in. torch weighs 6^ lb. The head is set at an angle of 20 deg., which experience has shown is the more generally useful. All the regular nozzles can be used with this torch. In addition to the torches mentioned, the Oxweld Acetylene Co. makes straight-tipped machine cutting torches, which may be used on any of the cutting machines on the market. Like all other cutting torches, small guide wheels may be used for steadying the torch when cutting to straight, irregular or circular lines by hand. In Table XI are given the oxygen pressures and amount of gas consumption for various thicknesses of metal. The acety- lene pressure is the same as for welding, 1 lb. With the data given in tliis table, and knowing the cost of acetylene and oxygen, the approximate cost of gas for any given job may be calculated with a fair amount of accuracy and serve as a basis for price estimates. It must be kept in mind that old rusty metal, like boiler plate, will take much more gas than will clean metal. Other Cutting- Torches. — In order to give tlie reader an idea of the construction of some of the other well-known cut- ting torches, a few will be shown. Fig. 53 shows details of a cutting torch made by the Messer Manufacturing Co., Phila- delphia. This type of torch will use either medium or low- pressure acetylene, as it works on the injector principle which is independent of the acetylene pressure. The oxygen jet is operated by means of the lever shown on top. The wheel guides are adjustable so that the tip may be kept the proper distance from the work. ' GAS CUTTING-TORCHES 83 y. S*^ I gjiocoocoioot^ocoooooco •^fe O — < (N CO Tt< lO CO t^ 05 CO t^ (N l>(N CO O *J 3 OOOOOOOOO'-t.-i(N(N'!j.iOTj-H 00 OOCONCOb-0«OCOt^05«OOOiOC^C005CO'.rHTl.C0O«0i-i«DC<)05C000iOMO (NO500t^CC)ii3U3U3'-< "-i C^ N CO CO CO Tj* Tj* lo CO t>- 00 05 >-" CO lO t^- rt* c3 O o CI d o 'z, o d "3 . •I :^" E-« pnfh^C 2 in. 90.75 ^^2 2.84 27.9 3.25 8 11 in X 11 in. 242 112 2.16 — — 9 6 in. diameter 113 46 2.48 11.3 10.00 10 15 " 177 162 1.1 12.6 14.00 11 16 " " 200 200 1.0 14.3 14.00 12 18 '.'- " 254 250 1.0 15.9 16.00 13 20 " «' 314 300 1.05 17.3 18.00 86 GAS TORCH AND THERMIT WELDING Test No. 1 comprises ordiiaary work as it came along. Risers were not clean and especially the smaller risers had sand and holes in the core. The oxygen consumption contains all the waste changing from one piece to another. Test No. 2a was made on clean metal with oxy-acetylene and with a special pointed tip. Test No. 2b was the same as No. 2a but cut with oxy-hydrogen, a regular tip being used. Test No. 3 was made on a very clean riser. The operator could rest his hand very comfortably. The cut looked as if it was done by machine. Tests Nos. Jf-7. — Cuts were made on risers of a 20-ft. flywheel. Cleaning was only superficially done and operator was in a fair, but not ideal position. One riser showed a large blow-hole. The cuts were clean through. Test No. 8. — Two cuts were made on the same flywheel. Risers were well-cleaned, but the cranes could not be spared to bring the face of the risers into a horizontal line. They had to be cut diagonally and the operator had to bend so far over that with the first riser he lost his balance and fell and had to interrupt the operation, therefore, no time was taken. The maximum thickness of the cut was 13 in. The cuts were clean through. Tests No. 9-13. — Cuts were made on risers of circular shape. The number of sq.in. cut per cu.ft. of oxygen is in the average lower than with rectangular shapes, as it is too cumbersome to regulate the oxygen pressure according to the varying thickness. It does not pay to start with lower pressure at the beginning and increase it the more the cutter is nearing the center or full thickness of the metal. More- over, the cuts were made with a two-line cutting torch so that the preheating flames would have suffered with regulating the oxygen pressure in such wide limits. With the heavier cuts of 15 in., 16. in., 18 in. and 20 in. thickness, the principal consideration was to cut through rather than to get stuck, and not look too close to the oxygen consumption. The Imperial Brass Manufacturing Co., Chicago, makes the cutting torches shown in Fig. 55. These are of the positive- pressure type. A is a combination cutting and welding torch, the oxygen pipe and tip being detachable, so that the curve tip may be put on when the torch is wanted for welding. B is a cutting torch only. Either may be used for oxy-acetylene or for oxy-hydrogen. When using oxy-hydrogen, the respective pressures are given in Table XIII. The torches may also be used for the company's three-way gas system, which uses a combination of acetylene, hydrogen, and oxygen, as explained under welding torches. The acety- GAS CUTTING-TORCHES 87 Taklio XIII. — I'i{ks,sui:e.s fou Oxy-IIyduookn Cutting with Impekial Touches Tliirkiif.ss of W'roUiuiit Cutting Iron or Steel . Pressures ^ Ti]i to be Cut, In. Oxygen, Lb. Hydrogen, I.b. Ill i to 2 30 to 40 5 to 10 211 2 to 4 no to 70 10 to 15 811 4 to 6 80 to 100 35 to 20 4H G to 9 100 to 125 20 to 25 5H 9 to 12 125 to 150 25 to 30 These figures represent minimum and maximum pressures. For in- termediate tliicknesses use pressures in i)roportion. lene and hydrogen are mixed through a Y-valve and enter the torch through the same liose. Pressures when the gases are used in this way are shown in Table XIV. Table XIV. — Pressures When Using Three-Way Gas Sy'stem Tliiclaiessof Steel or Cutting Tip, No. , Wrought Iron to be Cut. In. Oxygen, Lb. Acetylene, Lb. Hydrogen, Lb. IT 1 to 2 30 to 40 5 5 to 10 Oip 2 to 4 50 to TO 5 10 to 15 3T 4 to G 80 to 100 10 15 to 20 4T 6 to 9 100 to 125 10 20 to 25 5T 9 to 12 and over 125 to 150 15 25 to 30 The Carbo-Hydrogcn Co., Pittsburgh, Pcnn., makes two models of the injector-type hand cutting torches, shown in Figs. 56 and 57, and in detail in Fig. 58. They also make straight -nozzle machine torches. 'Mechanical, guides, or wheels, may be had for attaching to the regular hand models, or to the straight-nozzle torches. Tips are furnished in a number of interchangeable sizes and shapes. They are made from a solid brass bar, with an outer shell of copper solidly attached with rivets. The preheating holes are arranged closely around the cutting orifice so that the flame cones do not have a tendency to melt edges of the cut. It is claimed that the tips remain cool and do not have to be dipped in water to keep them cool when used for long periods. The regular sizes of the tips are arranged for cutting from the thinnest metals up to 18 in. in thickness, or more with special tips. The head and 88 GAS TORCH AND THERMIT WELDING base castings of the torelies are of Tobiii bronze and the tul)es are of seamless drawn steel, surrounded by an aluminum or a fiber handle. A special swivel-point valve stem, instead of the usual solid needle point, is used for controlling the Fig. 55. — The Iiiipciial Cutting Torches Fig. 56. — Carbo-Hydrogen Model C Cutting Toioh Fig. 57. — Carbo-Hydrogen Model B Cutting Torch. combustion gas. All parts are easily removed for cleaning, and the injector may be taken out witli a pair of pliers. These torches are designed for use with oxy-carbo-hydrogen gas. Carbo-hydrogen is a fixed gas, permanent under all weather conditions. Since it does not solidify there is said to be no residue left in the tanks or cylinders at any time. Tt is clean, GAS CUTTING-TORCHES 89 easy to use, and safe, since it is combustible but not explosive within itself. It is a product of the destructive distillation of suitable hydro-carbons, and has a general analysis of 85 per cent hydrogen and 15 per cent light hydro-carbons. It is claimed that this gas has no tendency to harden the surface of the metal being cut. It is not a sensitive gas, and backfiring is rare. It is marketed in steel cylinders under 1800 lb. pres- BASE CASTING Fig. 58. — Details of Carbo-Hydrogeu Cutting Torches. sure and of about the usual capacity. The pressure is reduced to from 5 to 10 lb. for working purposes, 5 lb. being the pressure usually employed. In Table XV are shown the approximate number of feet cut per hour, the pressure of the oxygen and the amounts of the gases used per lineal foot of cut. The carbo-hydrogen pressure is 5 lb. in each case. Table XV. — Gas Consumption and Pressures When Using Oxy-carbo- HYDROGEN CUTTING-TORCHES. ThE CaRBO-HYDROGEN PRESSURE IS About 5 Lb. in Each Case Thickness Size of Lineal Feet Pressure of Cu. Ft. of Cu. Ft. of Car- of Steel Cut per Hour Cuttinp Oxy- Oxygen Used in Inches Tip by Hand tren in Pounds per Lineal Ft. of Cut LTsed per Lineal Ft. of Cut J4" 1 A 110 15 1 1 M" 2 90 25 Iri 1 34" 2 75 32 2J^ IH 1" 2 60 35 3 \% IK" 8 45 45 47^3 2J^ 2" 3 38 50 7 -iVz 3" 3 28 60 14 6 4" 3 A 18 75 26 9 5" 4 13 85 32 12 6" 4 11 100 40 13 7" 5 8 120 50 13 8" 5-A 7 140 64 14 9" 5-A ti 160 78 16 Above pressures can be increased at times on various grades of steel advantage. 90 GAS TORCH AND THERMIT WELDING The Rego cutting torch is made by the Bastian-Blessing Company, Chicago. Like the "balanced pressure" welding torch made by this concern, it is claimed that the cutting torch cannot be made to flashback while in use. Details of the mechanism are shown in Fig. 59. The tip is of nickel-copper composition and mates with a ground joint to which it is held REINFOKCED TUBES NO FLftSHBACK DING SLEKVE EEDLE VAUVeS Fig. 59. — Eego Cutting Torch. by a union nut. The mixing chamber is easily renewable. All valves are outside and easy to get at for repacking or re- grinding. The high pressure oxygen valve seat is metal to metal, and it is controlled by a powerful spring and is operated by a long lever acting on a plunger, like the valve of a gasoline motor. The operating lever is easily locked so there is no strain on the operator's hand. GAS CUTTING-TORCHES 91 Combination Torches. — Several companies make combina- tion welding and cutting torches. These usually consist of a B Vir:pif^fi£AriNO AND CUTTING Fig. 60. — Aireo-Vulcan Combination Cutting and Welding Torch. Fig. 61.— Milburn ''Cut-Weld" Torch. Fig. 61A.— Details of Milburn Torch. Fig. 61B.— Details of Torch Tip. cutting attachment for the welding torch. As a commercial proposition, such combinations are usually not to be recom- mended, but where an operator occasionally has to shift quickly 92 GAS TORCH AND THERMIT WELDING from welding to cutting, they may sometimes be used to ad- vantage. Along with their regular lines of welding- and cut- ting-torches, the Air Reduction Sales Co., N. Y., put out the combination torch shown in detail in Pig. 60. This is known as the Airco-Vulcan cutting and welding torch, and it well illustrates the general principles of this kind of a torch. The main body is that of the regular Airco-Vulcan welding torch. The regular welding tip is removed from A to receive the connection of the special tip B. This tip is connected at C to the high-pressure oxygen tube D. A combination valve is screwed into the torch body at E in place of the single oxygen valve used for welding. This valve has a passage at F for the preheating oxygen and one at G for the cutting oxygen that goes out through D. Tlie preheating flame surrounds the cut- ting jet as in other regular cutting torches, so that an operator Fig. 62.— Torchweld Gas Cutting-Torch. may cut circles or angles without altering tlie direction of the torch body to any extent. The Milburn Combination Torch.— The "Cut-Weld" torch made by the Alexander Milburn Co., Baltimore, Md., is shown in Fig. 61. This torch is made into either a cutting or a welding torch by merely changing the tips. The illustration shows a cutting tip in place and a welding tip just at the left. The regular size is 19 in. long and weighs 2f lb. Details of the torch, with a cutting tip in place, are shown in Fig. 61A. De- tails of a welding tip are shown in Fig. 61B. In tests in Washington before engineers of the Stone & Webster Corp. and several government officials, a 12 in. steel billet was cut through in 6^ min., which includes ^ min. for preheating. A test was made to determine its resistance to backfire and though the tip Avas nearly burned ofp, no flashback took place. A hole was also blown through a 5 in. steel billet in 40 sec. with no flashback. GAS CUTTING-TORCHES 93 Torchweld Gas Cutting-Torch. — The gas cutting-torch shown in Fig. 62 is macU' by the Torchweld Equipment Co., Fulton and Carpenter Sts., Cliicago, and is known as their style 15 MC. It is designed to use oxy-acetylene, oxy-hydrogen, or oxy-hydrocarbon gases, such as butane, calorene, and the like. Special tips, however, are needed for the various gas combinations. An 85-deg. torch-head angle is standard but 70, 50, 35 and straight heads can be furnished when desired. A one-piece cutting tip is used and the mixing chamber is just back of the torch head. A novel feature of the construc- tion is that an annular space is provided around the mixer in which a small amount of gases accumulate. Drill holes con- nect this space with the gas passage-way leading to the tip and, in case of backfire to the mixing chamber, the ignited mixture in the annular space is designed to blow out the back- Acetylene Tube Front HP HP \/a/^e Push Pop/., ^ HPPear Tube, . P'uq iPl/intve RocKer B.P Valine Pluncfer 'tl.PVali'ePlunqei r, a /-« T- hPMivSeafOnly i--One Piece Cuttincj Fp Fig. G2J Details of Torchweld Cutting Torch. fire and eliminate the hazard of flashbacks into the flexible connecting hose. All the gas-tight seats in tips, needle valves and connec- tions, are of the line-contact type : In other words, a convex surface is brought into contact against either a flat surface or another convex surface. A tight seating is thereby much more easily obtained than by using two flat surface contacts. One of the difficulties experienced with two-hose type cut- ting torches is the back pressure of the acetylene into the oxygen hose. Under certain conditions this results in the oxygen hose becoming filled with mixed gases w^hich ignite at tlie tip and a more or less serious flashback into the oxygen hose is unavoidable. The Torchweld back-pressure valve is claimed to prevent the acetylene from entering the oxygen hose, since a certain pressure on the oxygen is necessary in order to open this valve. 94 GAS TORCH AND THERMIT WELDING and as the acetylene pressure also tends to close the valve still tighter. Details of the construction of this torch are shown in Fig. 62A. CUTTING UNDER WATER WITH A GAS TORCH A number of torches have been developed for cutting under water. One of these has been successfully used at the Puget Sound Navy Yard, Washington State, for some time, and was made by putting a special hood over the tip of a regular Fig. 63. — Underwater Cutting Torch. A cutting oxygen at 65 lb. pressure ; B, preheating oxygen; C, acetylene 24 lb. pressure; D, compressed air at 100 lb. pressure. Davis-Bournonville cutting torch as shown in Fig. 63. This hood is pressed against the metal to be cut, and air at 100 lb. pressure forces back the water and protects the flame. An electrical device is used to light the torch under water. In one case a cut was made 22 ft. under water by a diver, who cut out a piece 19 in. in circumference in -J in. ship plate and rose to the surface in 6 min. Six in. per min. was the rate cut on plate 1 in. thick. It is claimed that this torch will cut down to 200 ft. under water. CHAPTER VII GAS-PRESSURE REGULATORS AND WORKING ASSEMBLIES ►Since the gas pressure required in a welding or cutting torch is normally considerably less than that of a generator or storage cylinder, some form of pressure reducer or regu- lator must be used between a torch and the source of gas supply. The regulator used must not only reduce the pressures to working amounts, but must keep the gases supplied to the torch at as constant a pressure as possible regardless of the variation in the pressures at the sources of supply. This will be understood when it is shown that, for example, oxygen at 1800 and acetylene at 225 lb. pressure per sq.in., taken from cylinders, must be mixed in a Davis-Bournonville positive- pressure torch at approximate, pressures of 14 and 6 lb. re- spectively, when welding steel plate ^ in. thick. The pressure in the cylinders will constantly decrease as the gases are used, but in order to keep a correct neutral welding flame the gases must be supplied to the mixing chamber of the torch at the approximate pressures of 14 and 6 lb., and keep close enough to these figures for long periods of time to produce the desired flame without continual adjusting of the valves. The required working pressures are determined by the thickness of the metal being operated upon, the make of torch, and the size of tip being used, as tables already given indicate, but the prin- ciple remains the same in any case. Oxweld Oxygen Regnlators and Gages. — The gas-pressure regulators used on welding and cutting apparatus are prac- tically all made on the same general principle and vary only in minor details of construction. An Oxweld oxygen regulator shown in Fig. 64 will serve to illustrate the construction in general. The principal parts of a regulator of this kind are the body proper, regulating or shut-off valve, diaphragm, 95 96 GAS TORCH AND THERMIT WELDING pressure-adjusting spring and pressure-indicating gages. As a general rule all regulators have two pressure-indicating gages, one on the intake or high-pressure line, and one on the outlet, or low-pressure line. The gage, however, on the low- pressure acetylene line is sometimes omitted when using a low- pressure, or injector, torch on account of the low pressure at which the acetylene is used. In the illustration given, a dust or protecting plug is shown screwed into the connecting nut on the intake tube. This is Fig. 64. — Details of Oxweld Oxygen -Pressure Regulator. of course removed vvdien attaching the regulator to the supply pipe or valve. The arrows indicate the flow of tlie gas when free to move from the intake to the outlet. Following these arrows it will be seen the gas enters the intake and flows into the vertical passage A where it goes upward to the high-pres- sure gage B, which indicates the pressure of the supply line. The gas also flows downward in the same passage until it reaches the monel-metal nozzle of the regulating valve at C. ir tlie screw D is turned to the left far enougli to prevent spi'ing E from forcing the diaphragm F inward against the sliding sleeve, then spring G will keep the seat H solidly GAS-PRESSURE REGULATORS 97 against tlie nozzle C and no gas will enter the body of the regulator beyond the passage A. However, if the screw D has been run inward far enough to put a tension on spring E the diaphragm F will be forced inward and the regulating valve will be held open. Gas will then flow into the diaphragm chamber / until the pressure of the gas against the diaphragm overcomes the pressure of spring E. This allows spring G to close the regulator valve and stop the flow of gas. The flow is not usually actually stopped when the torch is in use, since the flow of gas and the pressure of the spring E will be so balanced as to allow just enough gas to enter to keep the pressure practically constant in the outlet line. The farther the screw D is run in the more tension is put on the spring E and the diaphragm F, and consequently the higher will be the gas pressure in the outlet line to the torch. From this it will be seen that any desired pressure within the capacity of the regulator can be obtained, and maintained, in the outlet to the torch by simply adjusting the screw D. The diaphragm used on a regulator of this kind may be made of reinforced sheet rubber, phosphor bronze or other composition metal that will not corrode or break easily. The regulators used for other gases differ but little from those used for acetylene or oxygen, and often the same regu- lators may be used provided the pressures required are within the range of the regulator in question. An oxygen regulator for cutting work should be built heavier and deliver a larger amount of gas than one used for welding on account of the higher pressure required and greater gas consumption. In using acetylene from a pressure generator it is good practice to have an acetylene line regulator as well as one for each operator's torch line. The Oxweld oxygen gages used when welding are made to register from to 2700 lb. per sq.in. on the high pressure side and from to 60 lb. per sq.in. on the low-pressure side, as shown in Fig. 65. It will be seen, by examination, that the outer scale on the high-pressure gage shows the pressure in pounds and the inner scale indicates the percentage of gas in the cylinder. That is, for example, if the gage hand points to 600 lb. there would be approximately 35 cu.ft. of oxygen left in the cylinder, providing a 100-cu.ft. cylinder was being used. 98 GAS TORCH AND THERMIT WELDING If it was a 200-eu.ft. fylindor the amount loft would be ap- proximately 70 cu.ft. As has been pointed out elsewhere, these figures cannot be taken as showing the exact amount of gas in the cylinder except under certain conditions, but they are sufficiently accurate for all ordinary purposes. For cutting purposes the Oxweld oxygen regulator shown in Fig. 66, is fitted with the same gage on the high-pressure Fig. 65. — Oxweld Oxygen Welding Eegulator. side as for welding, but on the low-pressure side the gage registers up to 200 lb. per sq.in. Their acetylene regulator is only supplied with a 350-lb. gage on the high-pressure side, as shown in Fig. 67. This is because of the fact that the Oxweld torches use acetylene at about 1-lb. pressure at all times. However, if required, two gages may be used as in all other makes. GAS-PRESSURE REGULATORS 99 Other Reg'ulators and Gages. — A Davis-Bonrnonville oxygen regulatoi- with gages is sliown in Fig. 68. This indicates from to 3000 lb. per sq.in. on the high-pressure side and up to 400 lb. on the low-pressure side. On the dial of the high- pressure gage are three rows of figures. The outer row shows the pressure per sq.in. ; the middle row, the cubic feet of contents for both 100- and 200-ft. cylinders; the inner row indicates the cubic feet of contents for 250-cu.ft. cylinders at Fig. 66. — Oxweld Oxygen Cutting Regulator. various pressures. Details of a regulator used for acetylene are shown in Fig. 69. This is practically the same in construction as the oxygen regulator. The numbers shown are list numbers of the parts, and are very convenient for ordering broken or damaged parts at any time. The regulator acetylene gages register up to 400 lb. on the high-pressure side and up to 300 lb. on the low-pressure side. An oxygen regulator made by the General Welding and Equipment Co., attached to a cylinder is shown in Fig. 70. 100 GAS TORCH AND THERMIT WELDING At the right and almost opposite from where tlie regulator is attached, is a projection which is a fusible blow-off plug i-equired on all cylinders by the Interstate Commerce Com- mission, to provide for the escape of the gas in case the cylinder should be overheated and the pressure become so great as to be liable to cause an explosion. This illustration clearly shows tlie kind of valve that is used on an oxygen cylinder. It is completely covered with a metal cap screwed onto the threads Fig. 67. — Oxweld Acetylene Regulator. snown at the top of the cylinder. The cap protects the valve and prevents it being broken off or damaged when the cylinder is handled or shipped. In using gas cylinders under working conditions it is advisable to have them placed on a portable truck made for the purpose, or else fastened in some way so that they cannot be tipped over. This will often prevent needless damage to the apparatus and sometimes avoid serious accidents. It should always be kept in mind that gases under GAS-PRESSURE REGULATORS 101 fi-oni 225- to 1800-lb. pressure per square inch arc not to be trifled with. Tank and Hose Colors. — Oxygen cylinders of ditferent con- cerns do not have a uniform color, but are usually painted gray and green, i-ed, yellow or dark-green. Acetylene cylinders are generally painted black and have a plate on them giving Fig. 68. — Davis-Bournonville Oxygen Eegulator. the quantity of gas the tank contains. Practice also varies as to the color of hose used to connect to the torches. Common colors arc black hose for acetylene and red hose for oxygen, although sometimes oxygen hose is black and the acetylene red. In making all hose or valve connections, they must be carefully bloAvn out to remove dust or any foreign substance. This is especially important on new hose which is almost sure 102 GAS TORCH AND THERMIT WELDING to contain considerable bloom left from the vulcanizing. In addition to their specific color, acetylene cylinder valves are often threaded left hand, as a safeguard against making the wrong connections. 2426 A '^m^i^zam^^ ^g^jijgg 14 1-:'%V <^f^"=^ F I52C 2919 29'23,-'Vl 1'^ 2926^ "U_ f{ 2916 -- -il_ Fig. 69. — Details of Davis-Bouruonville Acetylene-Pressure Eegiilator. Id making oxygen connections it must be remembered that under no circumstances should oil or grease be used on the oxygen regulator or cylinder valve. This is highly important GAS-PRESSURE REGULAT()i;S 103 as oxygen under pressure eoining in eontael with oil or gi'ease causes spontaneous eombustion whieh might easily result in Fig. 70.— Regulator Attached to a Gas-Cylinder Valve. Fig. 71. — Regulator and Cylinder-Couiioction Adapters a serious aecident. If a lubricant of any kind is needed a little glycerine may be used. Regulator Adapters. — No make of regulator is so made as 104 GAS TORCH AND IHEi^MIT WELDIXr. to be regularly interehangeahle with all makes of gas cylinders, since the sizes and threads used on diffei'cnt makes of cylinder connections vary considerably. Foi- this reason adapters must be used in many cases. Some of these are shown in Fig, 71. Care should therefore be taken to make sure that the regulator will fit the cylinder connections properly, or that the right adapter is used. If a regulator connection or an adapter does not start readily, it should not be forced as it is probably the wi"ong diameter or the thread may be of the opposite kind — that is right- or left-hand. Also be sure that an adapter with a round or conical seat is not used on a tlat seat, nor a round seat on a conical one or one not nmde for it. Adapters are made of soft brass and careless handling will cause a leaky joint. In ordering adapters the make of regulator used should he specifically stated, and also tlie make of cylinder on which it is to be used, as well as whether it is for oxygen, acetylene, h.ydrogen or other gas. Connecting Up and Lighting the Torch. — In order to make perfectly clear to the reader how to connect up a welding apparatus for the first time, an Imperial welding outfit is shown in Fig. 72. Fii'st I'cmove the protecting cap from the oxygen cylinder, and then open valve A very slightly. This is to blow out any dust and to insure the free working of the valve after the regulator is attached, which otherwise might be injured by the sudden rush of gas into it. In doing this, stand on the side opposite from the opening so that the gas will blow away from yoii. Always keep this in mind when blowing out the valve on any cylinder. Now take the regulator and turn the handle B to the left until it turns freely, so as to be clear of the diaphragm. Next make sure the connection at C is clean and free from dii't and fits properly or has the right adapter, then screw it up using judgment with the wrench so as not to break anything. With the valve at D closed, slowly open the valve A as far as it will go, using some force with the hand to insure that it is really backed up against the gland solidly. This is to aid in preventing the high-pressure oxygen from escaping around the valve stem. AVhen the valve is fully opened, the gage E will indicate the cylinder pressure which on a new one will be close to 1800 lb. Now put on the oxygen hose at F and then turn the GAS-PRESSURE REGULATORS 105 handle B to the right until about 5 lb. are registered on gage G. Then open valve to D so as to blow any dirt or bloom out of the hose. The valve D is then closed and the hose connected to the torch at H. The valve / on the torch may now be closed, the valve D opened, and the handle B screwed in until the gage G registers the proper pressure for the proposed welding job. Fig. 72. — Imperial Welding Outfit Connected to Tanks. as indicated in the pressure table for the make of torch being used. The various connections should then be carefully gone over with soapy water to test for leaks. Never use a flame on the oxygen or any other gas tank even though oxygen alone is not inflammable. Assuming that the proper tip has been placed in the torch for the thickness of metal to be welded, the torch valve I may now be opened fully and the handle 106 GAS TORCH AND THERMIT WELDING screwed in until the gage G registers about 2 lb. over the pressure given in the tabk'. This is to allow for the variation in cylinder pressure as the gas is used. The torch valve / is next closed, and it is also well to close the valve i> as a safe- guard before attaching the acetylene hose. The acetylene regulator and tank are now connected up in exactly the same way, except that the acetylene tank valve /, must be only opened one full turn. (On one make of cylinder the dii-ections say two tui'ns, so the operator should read the dii'ections on the tanlv carefully.) The hose is con- nected at K and blown out to remove any dirt, care being taken that no flame is near. It is then connected to the torch at L, and tests are made for leaks as before with the valve M open and the torch valve N closed. The torch valve N is next slightly opened and the issuing gas lighted. The valve is then fully opened, and if the gage shows any appreciable drop, the handle P should be turned until the gage registers about 2 lb. above th(> amount shown by the table. The resulting flame from the Inirning acetylene will be long, white, smoky, and of comparatively low temperature. The torch valve N may then be manipulated until ihc pressure l^lows the flame from Vig to ^/4 in. away from the tip, the distance depending on the size of the tip being used. This can only be judged properly by experience. The oxygen may now be turned on slowly. The flame will gi-adually i-eduee in size, the outer end or envelope becoming less luminous and the part near the torch tip, known as the cone, assuming a clear outline without any ragged edges. Wlien this is obtained, turn off the oxygen slowly until a shadowy point shows from the cone. Then with extreme care turn on the oxygen again until this shadowy point just disappears. This is the so-called neutral flame, and is neither oxidizing nor carbonizing. From time to time, while at work, the operator should test the flame as just outlined, as a slight excess of oxygen pressure will not readily show in the flame and can only be detected by this method. It will be found in practice, as a rule after the pressures have been set on the gages, that all regulation necessary for the smaller sizes of tips may be made with the torch valve, but that on the regular sizes it is often advisable to readjust at the regulators. It will be well to repeat here, c.As-ri;Kssum'; rkculators 107 for the l)enc'(it of the beginner, that all indicated table pressures are only approximate and good only for the make of torch mentioned in connection with them. Characteristics of the Oxy-Acetylene Welding Flame. — The chart shown in Fig. 73 will serve to illustrate the looks of the oxy-acetylcnc flame as far as it is possible to do on paper: A shows acetylene turned on with sufficient pressure, so that it blows away from the tip. This space depends upon the size of tip being used. B shows oxygen partly turned on, united with the acetylene. The flame has hcgun to assume two different shapes and two different colors. The center A B Fig. 7'.i. — Characteristics of the Oxy-Acetylene Welding Flame. flame is white and is shaped somewhat like a rosebud. Not enough oxygen has yet been given the acetylene and the flame is called carbonizing. Such a flame will leave the metal brittle and hard. G is the neutral welding flame. The rosebud cone of th(! upper figure has become blunt, with no ragged edges and of a beautiful blue-white color. D is an oxidizing flame — ruinous to welding. This is obtained by turning on too much oxygen and the cone has become shorter, of a darker, dirtier blue, and is more pointed. Tliis view is exaggerated. The utmost care is necessary to guard against this flame. Even a slight excess of oxygen is detrimental, as it will "burn" the metal. To stop work temporarily, first close the oxygen valve in 108 GAS TORCH AND THERMIT WELDING the torch and then the acetylene valve. To stop work per- manently, first close the torch valves in the order just given, then screw back both regulator handles until they are free of the diaphragms. Then shut off the tank valves tightly. In case of a flashback, always close the oxygen valve in- stantly, then the acetylene valve, after which the torch head may be cooled in a bucket of water. It should always be kept in mind never to turji on the gas at the cylinder with the regulating screw tight, as this puts spring tension on the diaphragm and allows the gas from the cylinder to enter the body of the regulator very suddenly (because the plunger of the valve is away from the seat) and as the sudden pressure strikes the diaphragm, the plunger is thrown violently against the seat, often causing the seat to become cracked or broken. With the motor of an automobile racing, you wouldn't throw the gears in mesh for high speed direct from neutral and attempt to start away from the curb — not if you wanted to keep your automobile very long — yet turning on the oxygen with the spring tension on the regulator has about the same effect on the regulator. Bear in mind that the regulator is a steadying device — that the diaphragm is the balance between the higli pressure of the cylinder gas and the spring tension and that at all times the movement of this diaphragm should be slow — never violent. The low-pressure gage is a positive index of regulator trouble. If you are operating, say at 15 lb., and after shutting off the valve on the torch, the hand on the dial keeps moving to 25 or 30 or 40 lb. without stopping, it means that the seat is damaged — that the high pressure of the cylinder is leaking past the plunger of the valve and the regulator should be immediately sent back to the factory for repairs. Only by violating some of the rules previously given would you be likely to damage this seat ; but once damaged, it should be immediately repaired. It will be noticed that two acetylene tanks arc shown in Fig. 72. These represent the two types in common use. The one in the middle is the type furnished by both the Air Re- duction Sales Co. and the Commercial Acetylene Co., while the tank at the left is furnished by the Prest-0-Lite Co. In the first named the regulator stands out at right angles, and GAS-PRESSURE REGULATORS 109 in the other it stands up as shown. The valve in the Prcst- 0-Lite cylinder differs considerably from the others as will be seen in Fig. 74. In this illustration the valve is shown at A, the valve wrench at B, the packing nut of the valve at C, and the union nut by which tlie regulator is attached, at D. E is the high-pressure gage, F the low-pressure gage, G, the regulator, H the pressure-adjusting handle, 1 the outlet and J the hose nipple. Lighting the Oxweld Low-Pressure Torch. — The directions given by the Oxweld company for the lighting of their low- pressure or injector torches, differ slightly from the foregoing. Fig. 74, — Prest-0-Lite Acetylene-Regulator Assembly. so they will be quoted here, starting from where the gases have been turned into the high-pressure sides of the regulators, which is the same as already outlined : First, connect the oxygen hose from the oxygen regulator to the hose connection on the torch marked oxygen. Likewise connect the acetylene hose to the torch valve marked acety- lene. Then select the proper welding head or tip that is to be used according to the chart or table furnished, and screw it carefully into the torch. Turn on the oxygen by means of the handscrew of the oxygen regulator until the pressure on the small gage is as given on the chart. Be sure that when this is done the oxygen valve on the torch is open. Then close no GAS TORCH AND THERMIT WELDING tliis valve. Open the acetylene valve on tlie toi-cli. Then turn the liandscrew on the acetylene regulator to tlie riglit until acetylene is passing through the torch. Then close the acety- lene valve on the torch. ,The apparatus is now ready for use, and the gases are further regulated when necessary by adjusting the valves on the torch itself. Open the acetylene valve entirely. Open the oxygen valve slightly. Then light the gases. After lighting the gases, open the oxygen valve wide ; adjust the flame by turning the acetylene valve to the right until a neutral flame is produced. "When the job is finished and you want to shut off the torch for a short time, release or turn the handscrew on both oxygen and acetylene regulators to the left until the flame on the torch goes out. Then close the torch valves. When work is completed for the day and the apparatus is to be put away, first close the acetylene valve, then the oxygen valve of the torch. Then turn off the valves on hofli cylinders. Then open the valves on the torches until all the gas in the regulators and hose passes out of the torch into the air. Then turn the handscrew of both regulators to the left until loose. Then disconnect the oxygen and acetylene regulators from the cylinders. Each regulator has a dust plug which is to be put on its cylinder connection during all the time the regulators are not con- nected to the cylinders. Place the regulators and torches with wrenches, goggles, heads, and tips in their proper place so that they will be safe and protected from dust, dirt, and rough handling. Roll up the hose and put it in the case or tool box where it belongs. Chemistry of the Oxy-Acetylene Flame. — According to the Prest-0-Lite company, the chemistry of the oxy-acetylene flame is as follows: Acetylene (CoH.,) is composed of carbon (C) and hydrogen (H). On combustion, the carbon combines with oxygen to form carbon dioxide (COo) and the hydrogen combines with oxygen to form water vapor (HoO). This takes place in the following manner: When the gases issue from the torch into the welding flame, the acetylene immediately dissociates; in other words, it siplifa up into carbon find hydrogen which in combination with oxygen form respectively carbon dioxide and water vapor. In con- GAS-PRESSURE REGULATORS HI sequence of the high flame temperature (6300 deg. F.) the water vapor formed by this primary combustion is immediately dissociated into hydrogen and oxygen. The oxygen assists in the burning of the carbon while the hydrogen (which can only combine with oxygen at a temperature below 4000 deg. F.) passes away from the high-temperature zone and combines with the oxygen of the atmosphere at the outer blue part of Fig. 75.— Imperial Three-Way Gas Outfit. the flame, where the temperature is sufficiently low to permit it. The result of this is that the inner or welding cone of the flame is protected by a shield of free hydrogen which prevents loss of heat and also tends to protect the weld from oxida- tion. The temperature of the oxy-acetylene flame is approxi- mately 6300 deg. F., at the hottest part of the flame, which is the tip of the inner white cone. The effect of this tremendous heat at the point of treatment is to bring the metal very 112 GAS TORCH AND THERMIT WELDING rapidly to a molten state so that it flows together and mixes thoroughly with the proper quantity of metal added by the operator. The molten mass thus formed does not merely cement two pieces of metal together — it fuses them into one uniform mass. Characteristics of Other Gas Flames. — The way an Imperial three-way outfit is connected up is shown in Fig. 75. The procedure is along the same lines as outlined for the oxy- acetylene work. This combination of oxygen, acetylene and hydrogen gives a more visible flame and a sharper cone than oxy-hydrogen alone does. Only a small percentage of acetylene is necessary to give the sharper cone but the flame retains the clearness, beauty and good qualities of the oxy-hydrogen flame. The percentage of acetylene may be varied according to the thickness and character of the metal being welded, so that the degree of heat and amount of carbon can thereby be regulated to meet different conditions. The approximate pressures to be used for the three gases for average w^ork, will be found in a previously given table. The combination will produce a heat of about 5000 cleg. F. The oxy-hydrogen flame will produce a much softer weld than oxy-acetylene if properly used, but its lower heat and the fact that the cone is not concentrated in a sharjD needle point, which allows the heat to radiate more, are drawbacks wdien heavy welding is attempted. The low visibility of the oxy-hydrogen flame also makes it difficult to regulate properly, and an operator requires considerable experience before he can become proflcient in its use. As has been already men- tioned, however, its long flame makes it very desirable to use for the preheating flame in a cutting torch, especially on heavy, thick work. In welding with the oxy-hydrogen flame, the torch has to be held farther away from the work than with the oxy-acetylene torch on account of the longer and less concentrated flame. When a black spot appears in the weld it shows that the torch is being held too close. The Oxy-Hydrogen Flame. — The characteristics of the oxy- hydrogen flame are shown in Fig. 76. In this illustration, v/hich is as clear as a flame can be represented on paper, the different flames are outlined as follows : E shows the hydro- gen turned on Avith sufficient pressure so that it blows away GAS-PRESSURE REGULATORS 113 from tlio end of tlic tip. The distance will vary from about \/jy to V4 ii^- according to the size of tip and pressures used. F shows the oxygen turned on. A narrow, light-blue streak appears in the center of the hydrogen mantle. This is the desired neutral flame. G is an oxidizing flame that will burn the metal. The oxygen valve should be gradually closed until the excess of oxygen disappears. Where hydrogen and compressed air are used as is done in preheating work, light welding, or lead burning, the flame closely resembles that of the oxy-hydrogen flame. The appear- Jb/ue " Fig. 76. — Characteristics of the Oxy-Hydrogen Flame. ance of the hydrogen-air flame is indicated in Fig. 77. E shows the hydrogen turned on with pressure enough to blow the flame away from the tip, the distance being about the same as already given. I shows the compressed air turned on and a dark streak of mixed air and hydrogen appears in the center. This is the neutral flame. / is the oxidizing flame. In general the air pressure used for this flame is close to that where oxygen is used. The Oxy-IUuminating Gas Flame. — The flame produced by mixing oxygen and coal gas, or natural gas, is suitable only for lead burning, preheating, very light steel welding, light cast-iron welding, or the welding of light brass or aluminum. 114 GAS TORCH AND THERMIT WELDING The cliaiactei'lstich; are shown in Fig. 78. K shows the gas turned on full force enough to slightly blow the yellow flame DarA- FiG. 77. — Characteristics of the Hyilrogen-Compressed-Air Flame. Purple - - Blue- FlG. 78. — Characteristics of the Oxygen Illuminating Gas Flame. away from the tip. L is the neutral flame produced by turn- ing on the oxygen. The cone is narrow and about \ in. long, of a beautiful purple color in a pure-blue outer mantle. M GAS-PRESSURE REGULATORS 115 shows too much oxygen. The cone has turned a reddish color. The oxygen must be decreased until the sharp purple-colored cone appears. In using oxygen and illuminating gas, a water seal should be used on the gas line to assist in purifying the gas and to prevent the entrance of .any flame, or Oxygen which might form an explosive mixture. Where acetylene and compressed air are used, as is some- times done for certain preheating or welding jobs, the flame characteristics closely resemble the oxy-acetylene flame. In order to obtain the best results, special tips should be used in the torch for the different gas combinations described. These can usually be promptly supplied by the makers of any of the torches on the market. CHAPTER VIII GAS-TORCH WELDING AND CUTTING OUTFITS Cutting torches are lighted in the same way as are the welding torches. In most cases, however, the oxygen pressure to the preheating flame has to be several pounds higher, on account of the drop after the cutting jet is turned on. Tlie apparatus used for cutting is also set up in the same way as for welding, as will be seen from the typical Oxweld cylinder outfit shown in Fig. 79. Where large amounts of oxygen are used for cutting or welding, it is well to have a centralized source of supply so arranged that the flow of oxygen need not be interrupted at any time, and will be ample for all demands. For this pur- I^ose, the Oxweld company has designed the oxygen cylinder manifold shown in Fig. 80. Oxygen from this manifold may be piped anywhere in a plant exactly the same as the acetylene is. It eliminates the enormous amount of handling necessary in supplying full cylinders and removing the empty ones from each individual station. The manifolds are so arranged that each half operates independently, making it possible to provide an uninterrupted supply of oxygen. If desired, both halves may be operated in unison. These manifolds are made in four sizes to accommodate, 6, 10, 20 and 30 cylinders respectively. Each manifold will handle cylinders of either 100- or 200-cu.ft. capacity without any additional change or adjustment. They have two constant pressure regulators, one of which is for relief service. Acetylene cylinder couplers or manifolds are used for the same reason as are those for oxygen. A number of Prest-0- Lite acetylene cylinders coupled together is shown in Fig. 81. These are valuable where large-sized torch tips are used more or less continuouslv, since the capacity of the cylinders supplying acetylene should be at least seven times the hourly 116 GAS-TOKCH WELDING AND CUTTING OUTFITS 117 consumption. Where the hourly requirements are from 61 to 75 eu.ft. of acetylene, 5-WC or 2-WK size cylinders should be used. Complete Working Outfits. — A welding or cutting outfit may consist of the bare essentials, or be so complete as to OXYGEN REOULKTOR Oxygen Tank Valve Connection ^^ut Safety VaWe ACETYLENE REGULATOR Tank or High- -Pressure Gage Connecting Nut' Adapter Safety Valve-- Tank or High- Pressure Gage --Low-Pressure Gage "■-Handle Outlet Connection --Cutting Nozzle il^- -Torch Head TORCH Cutting ValveJ , -Lever -Handle , Oxygen "Valve ".Oxygen Hose Connection Acetylene Hose-' Fig. 79 Typical Oxy-Acetylene Cutting Unit. include everything that will be needed to care for any job that will come along. The outfits may also be of either the stationary or the portable type, or a combination of the two. If of the stationary type, there will naturally be included many things such as holding jigs and fixtures which do not 118 GAS TORCH AND THERMIT WELDING Fig. 80. — Oxweld Oxygen-Cylinder Manifolds. ' -ARRESTER- r^ *-h\ ■l — %..M^m'^^ Fig. 81. — Prest-0-Lite Acetylene-Cylinder Manifolds. GAS-TORCH WELDING AND CUTTING OUTFITS 119 ordinarily belong to the strictly portable type. Two typical lists of parts and material for the ordinary run of work, where gas cylinders are used, are here given. These lists are taken from the catalogue of the K-G Welding and Cutting Co., New York. OX\'-ACETYLENE CUTTING UnIT 1 cutting torch, standard size, witli four interchangeable tips of graduated sizes. 1 high-pressure oxygen regulator, with 3000-lb. gage for cylinder pressure, 300-lb. gage for working pressure and one reducing valve with connection for hose coupling and needle valve. 1 acetylene-pressure regulator with 400-lb. gage for cylinder pressure, 60-lb. gage for working pressure and one reducing valve with connection for hose coupling and needle valve. 25 ft. high-pressure, copper-covered oxygen hose. 25 ft. steel-covered gas hose. 1 pair colored goggles for operator. 1 pair fireproof gloves. 2 wrenches and flint lighter, instructions, etc. 1 leather instruction and memorandum book. OXY-ACETYLENE WELDING UNIT 1 welding torch, standard size, complete with eight interchangeable tips of graduated sizes. 1 low-pressure oxygen regulator, with 3000-lb. gage for cylinder pres- sure, 60-lb. gage for working pressure and one reducing valve with connection for hose coupling and needle valve. 1 acetylene-pressure regulator with 400-lb. gage for cylinder pressure, 60-lb. gage for working pressure, and one reducing valve with connection for hose coupling and needle valve. 25 ft. red corrugated-rubber oxygen hose. 25 ft. black corrugated-rubber gas hose. 1 pair colored goggles for operator. 1 pair fireproof gloves. 2 wrenches, 1 flint lighter, instructions, etc. 10 lb. cast-iron rods. 10 lb. Norway iron for welding. 2 lb. aluminum rods. 1 lb. cast-iron flux. I lb. aluminum flux. It, of course, is not necessary for a user to purchase two complete units if his work does not warrant it, as these may be judiciously combined. For instance, the welding unit may be used for either welding or cutting if a cutting torch and 120 GAS TORCH AND THERMIT WELDING a liigli-pressuro oxygen regulator are added. It is very con- venient where portable apparatus is used to a considerable extent over an extended territory, to have a suitable carrying case for the smaller parts. This may consist of a chest attached to the cylinder or portable truck, or of a hand case, such as shown in Fig. 82. This ease and outfit is sold by the Air Reduction Sales Co., New York, and holds everything neces- sary for immediate attachment to the cylinders and starting to work. It has been previously mentioned that it is not advisable Fig. 82. — Carrying Case for Welding and Cutting Outfit. to place gas cylinders so tliat tliey may be knocked over. If they have to be stood up near the work, it is best to chain them to a post or brace them up in some way. A portable tr-uck is, of course, the best of all where outfit must be moved about. Such a hand truck has already been shown. Prac- tically every firm making gas-torch supplies sells a similar one. A very impoi'tant part of any gas-torch outfit is a pair of suitable glasses or goggles. These should not be merely dark glasses, but the lens should be made expressly for work of this kind. Cheap glasses are dear at any price, as the result may be ruined eyesight from the intense glare of the hot metal GAS-TORCH WELDING AND CUTTING OUTFITS 121 or from injurious rays. The glasses may be of either the spectacle or the goggle form, but the lens should be the same in either case. Glasses may be obtained from practically any of the firms mentioned in the various descriptions. The goggle form of glasses has the advantage over those of the spectacle type in that they will better protect tlie eyes from flying or ghmcing particles of hot metal or sparks. It is well to have tlic colored lens protected by clear glass. A pair of goggles is shown in Fig. 83. These are of a very satisfactory form. They are light and all parts that come in contact with the skin sliould be made of fiber, metal, or something that is sanitary and easily sterilized. The guards may be made of aluminum screen or fiber. Never on any account use goggles with guards Fig. 83.— Goggles for Gas-Toreh Work. made of celluloid. If the glasses used can be employed for long periods of time without the eyes feeling ''dazzled," or if they can be removed without white spots appearing before the eyes, then they are all riglit. Otherwise get others that are better fitted for the work and your eyes.. A darker lens is usually used for welding than for cutting and sometimes it is advisable to use different glasses for different metals. These will enable the operator to see more clearly when the glare is not so intense. Fire-fighters often need to cut through steel or iron win- dow bars, shutters, steel plates or sheathing, in order to rescue imprisoned persons or to get at a fire advantageously. For this purpose the Davis-Bournonville Co. supplies a very com- pact apparatus, shown in Fig. 84. The metal case is 6^ in. 122 GAS TORCH AND THERMIT WELDING wide, 14 in. deep and 50 in. high. It contains a 40-cu.ft. cylinder of acetylene, a 50-cu.ft. cylinder of oxygen, and a complete cutting unit with extra length of hose. Handles on each side of the case provide means for easy carrying by two Fig. 84. — Emergency-Cutting Outfit. men, and these handles placed as shown, when in use, insure stability. The complete outfit weighs 125 lb. A complete two-station welding and cutting outfit of the stationary type, using an acetylene generator and oxygen GAS-TORCH WELDING AND CUTTING OUTFITS 123 'JSi be O o o I 124 GAS TORCH AND THERMIT WELDING cylinders, is shown in Fig. 85. This can be extended to include any number of individual stations, according to the size of the generator employed. It is advisable to place the acetylene generator in a separate room or building. The oxygen cylinders and regulators arc placed within easy reach of the workers. Back-Pressure Valves. — Hydraulic back-pressure valves Fig. 86. — Oxweld Low-Pressure Hydraulic Back-Pressure Valve. should always be connected to the individual acetylene pipes, as shown, to avoid the danger of a flashback or an explosive mixture entering the acetylene line. An Oxweld low-pressure back-pressure valve is shown in Fig. 86. In some quarters it is thought that such a valve is not needed where positive- pressure torches are used, but this is not true, as the danger GAS-TORCH WELDING AND CUTTING OUTFITS 125 is always present when the oxygen pressure is greater than that of the acetylene or which, through accident, may become greater. An obstruction in the nozzle of the torch, or clogging of certain passages is apt to force the oxygen into the acety- lene line. There are also other conditions which may cause a serious accident. While mechanical valves may help, they are not so reliable as hydraulic valves and should not be used on shop lines. In the Valve shown, the acetylene enters at A and bubbles out at B where it rises to the surface of the water seal, and normally goes out of the hose valve C. The depth of the water through which the acetylene bubbles is sufficient to cover the tube leading to the outside air. If there is a backward flow of oxygen, the pressure exerted on the surface of tlic water lowers its level, causing it to rise in the tube open to the air, and if continued, forces it out of the tube, thus opening a clear passage for the oxygen to the outside air at D. The acetylene inlet meanwhile is protected by a seal of water. To avoid the possibility of blowing the water out of the seal, the valve is so designed that in case of a blow-back the water automatically flows back to the body of the valve, renewing the seal. The high-pressure valve of this type is provided with a ball check seated under water, which effectually prevents an excess pressure from working backward into the generator. It is also supplied with a relief valve of ample size to carry off any excess pressure which may accumulate in the body of the hydraulic valve. Both the high-pressure and low-pressurp valves should be kept full of water up to the screw plug E which is provided to regulate the height of water when filling. Lead Burning. — Lead burning is practically the same as any other Avelding work, except that the melting point of lead (620 deg. F.) calls for a much smaller flame. The same outfits intended for lead burning may be used for welding jewelry, small metal parts of various kinds, or for brazing work in some cases. On account of the torch itself being made small and as light as possible, it is customary to have only an oxygen valve on it, the gas and oxygen valves being placed in a "bench block" placed about midway between the torch and the sources 126 GAS TORCH AND THERMIT WELDING of supply. For lead burning alone, it is usually advisable to use some gas combination giving less heat than oxy-acetylene, such as oxy-hydrogen, iiydrogen and compressed air, oxygen Fig. 88. — Oxy-Hydrogen Lead-Burning Outfit. and illuminating gas, or others. In order to assist the would-be user, several typical set-ups, taken from the Imperial Hand- book, will be shown. An oxy-acetylene combination is shown in Fig. 87. This GAS-TORCH WELDING AND CUTTING OUTFITS 127 shows the bench block A, screwed to the wall, which in many cases is the better way. However, it is well to have this block in easy reach of the operator. The oxy-hydrogen set-up is very similar, as shown in Fig. 88. A hydrogen compressed-air unit is shown in Fig. 89. A constant air-pressure regulator is shown attached to the air line. In Fig. 90 is shown an oxygen-illuminating gas combina- tion. A water seal is used on the gas line as a safeguard and to act as a scrubber to some extent. A very light outfit, intended for jewelers, dentists, or cx- FiG. 89. — Hydrogen and Compressed-Air Outfit. perimental laboratory work, is shown in Fig. 91. The bench block differs some from the ones just illustrated. The torch itself is almost as light and is about the size of a large fountain pen. "With the bench block close to the operator, the torch valve is not needed. This outfit is made by the Davis-Bournon- ville Co. A very convenient feature is the torch-holding clip, shown at the top of the bench block. This obviates the neces- sity of laying the torch down at any time, with its attendant danger of fire. A Gas Flow Indicator. — It is often desirable to measure the flow of gases used in a welding or cutting torch. For this purpose, the Hydrate Engineering Corporation, Buffalo, N. Y., 128 GAS TORCH AND THERMIT WELDING has produced the Hydrex Flow Indicator shown in Fig. 92. In Fig. 93 the principle on which it works is outlined. The gas enters the nipple a, as indicated hy the arrow. Thence it flows into the chamber c, up through the tube d and out the nipple h. As the gas passes into tube d it raises the plunger c. The greater the flow of gas the higher will the plunger be lifted. The disk / is suspended from plunger e and is visible through the gas tube g, so that the flow of gas is indicated on a scale calibrated in cubic feet per hour, reduced to normal Fig. 90. — Oxygen and Illuminating Gas Outfit. conditions for a gas flowing at a definite pressure and a definite temperature. The gas at the time it is measured, may flow at a known pressure and a known temperature, which do not coincide with those for which the calibration is prepared. In such a case the reading has to be converted into the proper volume, by applying those formulas which govern flow of gas through orifices. These calculations, of course, should not be necessary for a conveniently applicable apparatus. The elimination of com- putations is accomplished by the use of a chart, which permits GAS-TORCH WELDING AND CUTTING OUTFITS 129 the conversion of a reading for any pressure and temperature. This is practical in tlie laboratory where close observation and intelligent interpretation of the chart may be expected. For the ordinary shop work it is out of place. To make the instrument a practical and convenient shop and an all-around test apparatus, it was necessary to simplify the determination of the volume passing through the flow in- FiG. 91. — Manufacturing Jewelers' and Dentists' Welding Outfit. dieator. A pressure gage makes this possible. This pressure gage indicates factors instead of pressures in pounds per square inch. The reading on the flow indicator scale, multiplied by the factor, is the actual volume of gas, reduced to normal conditions, passing through the flow indicator. Thus, the sliop operator is relieved of all complicated mathematical considera- tions and he may concentrate his energies upon his work. 130 GAS TORCH AND THERMIT WELDING The inflnenoc of the temperature npon the reading is ap- proximately 1 per cent for each 10 deg. F. The influence of the pressure upon the volume is inverse to that of temperature, increasing temperature decreases the den- sity of gas, while increasing pressure will increase the density. Fig. 92. •iG. 93.— Details of Gas Flow Indicators. On a calibration made for 40 Ihs./in." gage pressure, the increase of one pound in pressure compensates for an approxi- mate increase of 10 deg. in temperature; and if the calibration is for 10 Ibs./in." gage pressure, an increase of | Ib./in.^ pressure will compensate for an increase of approximately 10 deg. F. CHAPTER IX LEARNING TO WELD WITH A GAS TORCH Directions as to how to handle a gas torch and to weld are of no use without actual practice — and lots of it. How- ever, the v/ould-be welder should have certain definite instruc- tions given him before he attempts to do any work. To be- come a first-class all-round welder requires long experience, a mechanical sense, and a liberal application of brains as a flux on every job. Learning to do simple, one-operation weldiug jobs, however, is comparatively easy if a competent instructor is to be had. If such an instructor is not to be had, the direc- tions given here will serve as a foundation upon which a fair knowledge of the work may be built. It is easier to learn to cut than it is to learn to weld, but as welding is the more im- portant of the two the method will be described first. It is taken for granted that the welder, following directions already given, is familiar with his apparatus, has the correct size of tip for the metal to be welded, knows how to light his torch and has ample gas for the work. In making a gas-torch weld, it is necessary that fusion penetrate entirely through the metal. In order to aid this the pieces are usually cham- fered or beveled with an air hammer, a grinder, or cold chisel. By beveling is meant the grooving or chamfering of the metal at the line of the weld, the depth of this groove or V being e(iuivalent to the thickness of the metal. Beveling is not required on castings or plates lighter than ^/g in. in thickness. From Vg in. to ^/jg in. in thickness a chamfer of 45 deg. on each piece, or a total angle opening of 90 deg., is about right. From ^/jg in. up to the maximum thickness weldable by the gas torch, an angle opening of from 60 deg. to 90 deg. is used, the angle being dependent somewhat upon the nature of the material and the location of the weld. On very thick metal a channel with parallel sides, beveled only at the bottom, is frequently used. 131 132 GAS TORCH AND THERMIT WELDING Under certain conditions it is advisable, but seldom economical, to use an oxygen cutting torch for beveling. In case this is done, care must be taken that all the oxide pro- duced on the surfaces cut by the torch is removed before welding. The beginner should start by welding strips of rolled iron or steel I in. thick and about 1^ X 6 in. These may be welded without the use of a welding rod. The pieces must be properly cleaned and free from scale, grease or dirt. The operator must wear suitable colored goggles, and should grasp the handle of the torch firmly. It is not good practice to hold it in tlio finu'ors, because it is impossible to manipulate Fig. 94.— The Way to Hold the Torch. the flame with as great regularity and control, nor will it be possible to do as heavy work without tiring. Occasionally, the hose is thrown over the man's shoulder. In this case the weight of the torch is suspended and held by the tubing, so that it is only necessary to impart the typical welding motion to the torch, which can usually be done by the fingers. The movement of the welding flame is hindered, however, and this method is therefore not recommended. It should be used only as a relief when the work is of long dura- tion and the operator's wrist and forearm become tired. The head of the torch should be inclined at an angle of about 60 deg. to the plane of the weld, as in Fig. 94. The LEARNING TO WELD WITH A GAS TORCH 133 inclination of the head should not be too great, because if it is the molten metal will be blown ahead of the welding zone and will adhere to the comparatively cold sides of the weld. On the other hand, the welding head should not be inclined too near the vertical, because in that case the preheating effect of the secondary flame will not be efficiently applied. There are certain cases, however, where the conductivity of the metal is such that it is not necessary to utilize this pre- heating. Also certain metals have the property of absorbing the gases of the flame. Consequently, in these cases it is best that the flame impingement be concentrated to as small an area as possible. Torch Motion. — The motion of the torch should be away Fig. 95. — Different Flame Movements. from the welder and not toward him, as closer observation of the work can be obtained and greater facility in making the weld will be experienced. AVhere very thin sheet material is being welded and it is not necessary to use a welding rod or wire, a weld may be produced by moving the torch in a straight line. It can ]-eadily be seen that this does not apply to welds which have been beveled, and which require the use of filling material, for in this case a swinging motion must be imparted to the torch to take in both edges of the weld and the welding wire a1 practically the same time. In comparatively light work a motion is imparted to the torch which will cause the incandescent cone to describe a series of overlapping circles, the overlapping extending in the 134 GAS TORCH AND THERMIT WELDING direction of the welding, as shown at A, Fig. 95. In order that the weld be of a good appearance this mnst be constant and regular in its advance. The width of this motion is de- pendent upon the size of the material being welded and varies according to the nature of the work. In some cases a move- ment like a figure 8 is used, as shown at B, but this is a rather complicated one for a beginner. In heavier work, if the system described were used, a great deal of the motion would be superfluous. Consequently, either an oscillating movement, or one in which the jet of the torch will describe semi-circular zig-zags, as at C, should be used. This confines the welding zone; and while the progress is not so fast, it is more thorough than the other system for this class of work. To the average beginner the regular control of these mo- tions is difficult, and considerable practice is required to be- come skilled. It is the regularity of these motions that pro- •^ WeM Finished . Here''-. Weld ~\ k-/i"-H<— /i"-*t k- 6" H Fig. 96.— Plates and Finished Butt Weld. duces the characteristic even-rippled surface of good gas-torch welding. The progress of a welder and the quality of his work can be determined to some extent by the skill with which he produces this effect. On the practice pieces of |-in. thick material, the operator should so manipulate the jet as to take in about \ in. on each side of the joint. The point of the cone of the flame should be held about Vio in. from the metal, but not actually touching it. On thicker metal the distance of the cone tip will need to be greater, or about \ in., depending on the size of the tip used. It is far better to have the torch too far away than too close, as a hole may be blown through the metal. Start to weld at the end of the joint, and as the metal commences to get red give the torch a swinging motion from side to side. Keep this up until the corners of the plates are run together clear through their thickness. Then go on until the entire length is welded. Do not move the torch any faster than necessary to give the metal a chance to run together LEARNING TO WELD WITH A GAS TORCH 135 properly. Be sure that the bottom edges of the plates are melted together before going ahead farther. The plates and the way finished weld looks are shown in Fig. 96. Care should be taken not to touch the torch tip to the metal, as this will obstruct the flow of gas and may cause it to backfire and burn inside the tip. In such a case, shut off the oxygen at once, then the gas, and cool the tip in cold water. Then relight according to previous instructions. Do not go back over the weld unless absolutely necessary. Do not run the hot metal on top of the cold metal. Do not leave blowholes, scale nor low spots in your weld. A blowhole is a bubble in the metal. It sometimes occurs alone, and other times there are several of them. It maks the metal look spongy s ABC Fig. 97. — Plates Welded Without Proper Divergence. or porous and is caused by not properly running the metal together or by leaving impurities in the weld. When metal is melted, a coating which flakes off is formed. This coating is called scale and is bluish steel-gray in color. Low spots are unfilled spaces in the metal caused by moving the torch too fast or unevenly, - The beginner will soon discover that two pieces laid with the edges close together will not weld properly, as they will at first diverge, as shown at A, Fig. 97. Then they will gradually draw together as at B when the weld is about half done. From this point on the edges will draw in until they overlap about as shown at C. There are two methods of overcoming this: The first is to **tack" the two strips at intervals, as shown at A, Fig. 98. This method, however, has the disadvantage of causing buckling or warping under certain conditions. This warping may in some cases be eliminated by 136 GAS TORCH AND THERMIT WELDING rolling or hammering after welding. The second, and more satisfactory method, is to diverge the plates, as shown at B. The amount of divergence is dependent to some extent upon the thickness and nature of the metal, but it is safe to say that this divergence should not be less than 2^ per cent, nor more than 6 per cent of the length of the weld. Some operators stick to the hard and fast rule of | in. per foot, which is a very satisfactory figure in general. However, to obtain the best results it is sometimes best to deviate from this, as prac- tice and experience dictate. In welding a sheet-metal cylinder, it is often necessary to insert a wedge or pin, as shown at C, Fig. 98. — Methods of Allowing for Scam Contraction. in order to obtain the proper separation and prevent over- lapping. On very thin sheets, it is usually advisable to flange the edges, instead of trying to butt weld with or without adding material from a welding rod. The flanges are turned as shown at D. The flanges may be tacked as for butt welding, or clamped with tongs, as shown. In regular manufacturing work, where the welding proceeds rapidly, it is common to have a helper move the tongs ahead of the welder. Any warping can, as a rule, be easily ironed out of the thin sheets. In some cases, the edges are lapped and both edges welded, either using a welding rod or not, according to the nature and uses of the parts being welded. It is very important that the beginner should test his work LEARNING TO WELD WITH A GAS TORCH 137 by bending the metal sharply at the weld and carefully in- specting for defects, which should be overcome on the next piece. The tendency of a beginner is to experiment on all sorts and thicknesses of metal, but he will progress faster if lie sticks to one kind and thickness until he masters it. The beginner, as well as the more experienced welder, should occa- sionally test his flame to be sure he is maintaining the proper neutral flame. This is done by turning off the oxygen until a shadowy point appears on the cone. The oxygen is then turned on again until this shadowy point just disappears into the cone. The reason for this testing of the flame, is that changes in temperature of the torch or variations in pressure may alter the flame without the operator's knowledge, and tlius injure the weld by cither carbonizing or oxidizing the metal, according to whether there is an excess of acetylene f<-— w Fig. 99. — Beveled-Edge Plates and'Flaine Movement. or of oxygen. If the tip should become clogged from any cause, it should be cleaned out with a soft wire or a piece of wood, taking care not to get the hole in the tip out of round. Using the Welding Rod. — In starting to use the welding or filling rod, it is just as well to begin to weld pieces similar in size to those used for the first lessons in welding, except that the pieces should be about | in. thick, and beveled on one side of the edges only. The amount of divergence is judged as in the case of the plain butt-welding work, as shown at A, Fig. 99. The flame movement is indicated at B. The welding rod should be held and inclined as shown in Fig. 100. In this position sufficient quantity of metal may be added at the right time. With the welding rod held in a vertical position or horizontal, the possibility of the addition of an excess of metal, part of which is not fused, is great. In adding this metal, care must be exercised that the edges of 138 GAS TORCH AND THERMIT WELDING the weld are in the proper state of fusion to receive it. If the metal is not sufficiently hot, the added material will merely stick to the sides and fusion will not exist. It is therefore necessary that, by the motion of the torch, fusion be produced at the edges of the weld equal with that of the welding rod. The usual faults of the beginner are failure to introduce the welding rod at the proper time into the welding zone, to hold the rod at the wrong angle, or to fuse either too little or too much of the rod. The filling material when melted should never be allowed to fall into the weld in drops or globules. When the proper time arrives to add it, the welding Fig. 100. — Using a Filling Eod in Welding. rod is lowered into the weld until it is in contact with the molten metal of the edges. When in this position the flame of the torch is directed around it, and thus fusion is produced. It is customary to add metal in excess of that of the original section, and round it over nicely. There are several very important reasons for doing this. First, the weld is reinforced and the strength is accordingly increased. Second, in case a finished surface is desired a sufficient stock must remain to allow for finish. Third, small pinholes or blowholes may be found just under the surface of the weld, which do not extend to any depth, and may be removed by filing or machining. LEARNING TO WELD WITH A GAS TORCH 139 In some cases the plates do not start to draw together until the weld has nearly reached the center. In a case of this kind it is good policy to slow down the welding. After the weld has been completed to the center, the plates will commence to draw together more rapidly, and in case the plates draw p'^'^'-'"\/^y$;^ 1 -t ] Bevel on metal less n i r ^ than I or ^n. and over ^^^'^ '"'J ^ f ^V""' 1- ., ■ 7 over4in. thick where 8 in. thick . ^ . , , , parts must be welded from one side only 1 R K- Sections over j in. thick Parts are beveled and welded on both sides Shaft beveled to chisel edge for welding Proper Tnethod of beveling shaft over 2 in. in diam.eter How filling material must be built up when welding different thicknesses Joint when welding convex end to steel cylinder I Weld concave head in steel cylinder Flat end to be welded into tube Flange to be welded Pipe butt weld ^^^^y_-z Lap weld Edge weld. Corner weld. First side to To have a deplh equal " to thickness of metal Angle corner weld. Large angle first to show through before finishing Fig. 101. — Various Examples of Welding Jobs. too fast, speed up on the welding until the proper distance is secured between the two plates. After the beginner has practiced until he can make a good weld on the plain and beveled plates, as suggested, he may practice on the forms shown in Fig. 101. These forms repre- sent most of the kinds he will encounter in the regular run of work in a job or general repair shop. 140 GAS TORCH AND THERMIT WELDING Sources of Trouble. — It will be well to repeat to some extent, the instructions and advice previously given, in order to emphasize the danger points: The first source of trouble in making a vi^eld is improper adjustment of the welding flame. If the flame is not adjusted properly the resultant weld will be inferior. The commonest fault is the presence of too much oxygen. In this case, unless the welder takes a great deal of care in removing the oxide by mechanical means, it will be incorporated throughout the weld. The presence of oxide prevents the thorough blending of the metal, and therefore decreases its strength. Failure to penetrate to the bottom of the weld is the cause of a great many defects. Tliis fault is not only that of a beginner, but also that of the more skilled operator. Very often the desire to complete a weld rapidly will cause the operator to hasten over the most important part of his work, which is to secure the absolute fusion of the edges at the bottom of the weld, before filling rod is added. This defect not only reduces the section of the weld, but also produces a line of weakness in case the weld is submitted to bending or transverse strains. When molten metal is added to metal which is not in fusion, a weld is not secured. The molten metal merely sticks to the cooler metal; this defect is common with careless operators. It may be caused by improperly beveling the pieces to be welded, by the faulty manipulation of the torch or by improper use of the welding rod. For the beginner it is at first difficult to distinguish the proper temperature at which to add the filling material. Usually he applies the filling rod before the edges of the weld are in fusion. The adhesion in this case occurs at both edges. Occasionally, one edge of the weld is in fusion, but the other is not, in which event the adhesion is restricted to one side. In some cases the edges of the weld are both at a point of fusion too soon. Under these conditions a film of oxide may exist on each edge. When a filling material is added, adhesion is produced with a film of oxide separating the edges and the added material. Quite often an operator, in applying -the welding rod to the weld, will concentrate his flame on the welding rod and the edges of the weld. As he plays the LEARNING TO WELD WITH A GAS TORCH 141 torch around the rod he will inadvertently force some of the molten metal ahead. The metal not being in the proper state of fusion, there will consequently be only a small area of adhesion. In welding cast iron, copper,- and to some extent steel, a very common fault of the beginner is that of forming blow- holes or porous sections in the weld. This can be overcome by close observation of the work while welding and by certain corrective means, the principal one of which is the use of proper fluxes and proper manipulation of the welding rod. ■ mid Here -^"^"^^^^^^^^/^/.^ saiioM thkjUuh cluep of blvll 1 - Melt bottom of V l-Add filling rod till V is half fiUed 3- Add filling rod till V is filled 4- Melt down edge of metal previously added and melt bottom of V 5- Add filling rod till V is half filled 6- Add filling rod till V is filled 7 - Proceed as in 4 Proceed in this manner till weld is completed Fig. 102.— Method of "Building Up" a Weld. It is needless to say that the existence of such defects in a weld seriously affect its ultimate strength. Occasionally, wields are encountered in which dirt or some foreign material is incorporated. This Avill cause porosity and an inferior weld, which could readily have been avoided by removing the material either before or during the execution of the weld. Built-Up Welds. — Where steel of considerable thickness is to be welded, the Oxweld company recommends the method illus- trated in Fig. 102. In the example selected the steel plates are f in. thick, beveled 45 deg. on each edge, making an in- cluded angle of 90 deg. In doing the work in this way, first melt the edges of the bottom of the V together for a length cf 1 in. Add the welding rod to this length until the V is about half filled. Be sure that the sides of the V are melted 142 GAS TORCH AND THERMIT WELDING when the rod is added. Then go back over this and fill up the V V32 ill- thicker than the original plate. AVhen this length of weld is done, melt the edges of the plates ahead down into the bottom of the V, and at the same time being sure that the end of the weld already finished is melted and flows into the botton of the V. Then add to this next section metal until a reenforcement of V32 i^- greater than the thickness of the plate is formed. Keep on in this way until the plates are welded. Near the finish of the weld it is necessary that the rod be given a slight swinging motion, similar to the torch. This is in order that the top of the V be entirely covered. Vertical Welds. — Where plates are to be welded with the seam in a vertical position, the same rule for the amount of divergence is used as for those in any other position. The weld should be started at the bottom and carried upward with- .■■We/d .here-. f i ( y- •aU<- >I/V k/vx 'finished 'Weld-- K— — •— IZ'-— >1 Fig. 103.— The Way to Fill Up Holes. out stopping until the weld is completed. Practice on work of this kind will give the welder experience in the control of the molten metal as no other kind of weld will, and he should put in considerable time on this work. Filling' Up a Hole. — A thing that every welder should learn as soon as he has mastered the simpler welds, is to fill up holes properly. A good way to learn is to take a piece of |-in. plate and drill three holes in it, \, \ and 1^ in. in diameter, as shown in Fig. 103. The beginner should commence with the smaller hole first. The weld should be started by melting down the top edge of the hole in one place. This will give a slight angle to one side. On the face of this angle metal is added. The hole is built up by adding metal continuously from the bottom to the top until it gradually closes up. The welding should be carried on around the hole, however, and should not be built up from one side only. When the hole LEARNING TO WELD WITH A GAS TORCH 143 is properly filled in, the metal should meet at the center. Proceed with the 5-in. hole and the 1^-in. hole exactly in the same manner. Turn the plate over and clean up the bottom side by melting the excess metal with the torch. Forming: Bosses or "Putting On" Metal. — The forming of bosses, building up missing parts, or putting on metal where needed, forms a very important part of a welder's work. Con- yteld ' Here'' |-K JZI Finished k- j'-H Fig. 104. — Building Up Bosses. sequently, the beginner should practice work of this kind as soon as he has mastered the ordinary run of welds outlined. He can begin by building up bosses an inch or so in diameter and 1 in. high on a steel plate, keeping at the work until he can produce a boss of fairly regular outline. He can then practice on square or rectangular bosses. Built-up bosses of this kind are shown in Fig. 104. Since the welder has already practiced vertical welds he should have little trouble in placing |<-3>1 ■F/nished We id - /2 Fig. 105. — Putting a Collar on a Shaft. his metal where it is wanted. Care, however, should always be taken to make sure that there is perfect fusion of the added metal and the plate before building up the boss. If a good weld to the surface of the plate is not made, the rest of the work is worthless. Be sure that all scale and dirt are worked out of the metal. Another type of built-up weld is shown in Fig. 105. In making a weld of this kind for the first time, take a piece of 2-in. shaft, 12 in. long and clean off the surface for about 3 in. at one end. Use a ^As-in. welding rod, and a No. 10 144 GAS TORCH AND THERMIT WELDING Oxweld tip, or its equivalent, and 21-lb. oxygen pressure. Place the flame of the torch on one spot of the surface until it is melted. Then add the welding rod. Add a layer of 1 in. wide and 3 in. long along the shaft. Make this layer ^ in. thick. When tliis strip is finished, weld another strip on top of it, starting at the end just finished. This gives a strip of added metal 1 in. wide, 3 in. long and ^ in. thick. When this is done, start another strij) at the side of this, being careful that the metal of the shaft is melting before the welding rod is added and also that the edges of the first two layers are at the same time melted down to the shaft. Proceed with the welding exactly the same as just described, adding strip /We/dHere-:^ .-Finished Weld-. ^ ,-\f\S\J\j Fig. 106.— Building Up Gear Teeth. after strip, side by side, until the end of the shaft is covered all around. Ilemember that the shaft must be melted before any metal can be added, that each layer must be melting before another layer can be added to it, and that each strip must be welded botli to the shaft and the strip next to it. AA^hen the shaft is completely covered, the end of the weld should be gone over with the welding flame, in order to clean it up and to be sure that a weld is produced at this point. Following the building-up work just outlined, it is a good thing to practice building up worn- or broken-out teeth in old or scrap steel gears, as shown in Fig. 106. Before a welder attempts to do any actual repair work on gears, how- ever, he should first learn more about expansion and contrac- tion, and the metliods of overcoming their effects. WELDING BACKWARD In an article published in tlie "Acetylene and Welding Journal," London, England, Capt. D. Richardson describes a method of welding which differs considerably from the gen- erally accepted American practice. LEARNING TO WELD WITH A GAS TORGH 145 In 191 G, M. IvouUeau, ;i Froiu-h ucetylene engineer, was seat by his firm to Italy in connection witli tlie manufacture of large welded pro- jectiles. The welds were being made by welders with little knowledge of the process and examination showed that from the number of de- fective welds obtained it would be difficult to get worse results. The welds were porous, adhesion was common, and the solidity of the joints was extremely bad. The supervision of the 1000 to 1200 welders distributed through seven or eight different works was a serious problem. The daily consumption of oxygen and carbide, at a cost of thousands of francs, in producing work of the type described made it an urgent economic problem to bring about improvement. Faced with these various problems, j\I. Roulleau, after experiment, introduced the method of welding backward into the various Italian workshops which came under his technical supervision. This change in method produced excellent results and on returning to France after a mission of three years in Italy, M. Roulleau collaborated with the Union de la Soudure Autogene with a view to a wider application of his method. Welding backward may be defined as the method of executing a weld in which the welding rod follows the torch as opposed to preceding it. Or again, it may be defined as a method in which the flame is inclined towards the welded portion. A fuller and perhaps better definition would be : The flame is inclined backwards and only undergoes a slight transversal motion in addition to its regular ad- vancement, the welding rod follows the flame and is given a movement, the end of the rod always being molten. It is claimed that, having acquired the method of welding backward, the welder will find it easier to execute welds and that the penetra- tion is always satisfactory ; adhesion is almost impossible ; the metal is sound; there is a diminution in the amount of oxide, and the metal is more ductile. Finally, that the speed of welding is greater than with the old method from which it will be gathered that there is economy in labor and gases. The economy in the consumption of filling material is of the same order as it is possible to reduce the beveling angle. APPLICATIONS OF METHOD This method of welding is applied mainly to steel plate above Vic in. in thickness. It should be used on all plates falling in the range of J to ^ in. in thickness. The process is particularly valuable for certain industries such as the manufacture of boilers, etc. Its application to mild steel has been specially studied. A number of experiments have shown that welding backward gives better results than the usual method when welding steels with a higher carbon con- tent—medium and hard. It is not satisfactory for aluminum welding and gives indifferent results when welding cast iron. On the other 146 GAS TORCH AND THERMIT WELDING hand, the first series of tests in using the method when welding copper and brass have proved satisfactory. Wlien comparing this metliod of executing welds with other methods, one might say that welding backward is a "more mechanical" method. From this it follows that more definite rules have to be observed in executing welds and it is advisable, especially for beginners, to strictly observe the rules laid down. These instructions, which are the result of investigation, may subsequently be modified and added to, but in the meantime they give good results and should be followed. For example, it will be noticed that for this method of welding, the edges of the two pieces of metal to be welded should be chamfered or beveled, so that when they are placed together, the two beveled edges form a V. Although the angles of the bevel can be reduced, beveling is still indispensable, even on thin material. PREPARATION OF MATERIAL The parts to be welded are prepared in the ordinary way, tacks, or short welds, at intervals of from 2 to 6 in., according to the thickness of material may be used. In the absence of jigs, the parts are supported so that the lower edges are in the same horizontal plane. The angle of the V formed by the two beveled plates should never exceed 90 deg., and, as already mentioned, for this method can be distinctly less. Beginners can gradually reduce the angle of the V from the previous standard of 90 deg. until they arrive at GO deg. for material I in. in thickness or above, and for material between J in. and i in. an angle of between 45 and 50 deg. can be used. The beveling should be carefully done and extend the full depth of the plates as partial beveling will produce defective welds. The power of a torch, in other words, the quantity of heat which it is capable of giving out in a given time, is generally measured by the number of cubic feet of acetylene burnt in an hour with the flame perfectly regulated. The rule which has been adopted for welding steel by the usual method of executing welds can be followed, namely, that the consumption should be about 5 cu.ft. of acetylene per hour for every Vie in. in thickness. So that for material '/le in. thick, a torch consuming about 15 cu.ft. per hour would be required, for I in. one consuming 30 cu.ft., and so on. However, it will be found that beginners obtain the best results by using a less powerful torch than indicated, whilst the more expert welder rapidly reaches the stage where he can advantageously use a more powerful one, as for example, for Vs-in. material a torch consuming 5.3 cu.ft.; for '/le-in. material one consuming 20 cu.ft. ; and for |-in. 40 cu.ft. SIZE OF ROD The choice of the right size of welding rod is of great importance. A rod that is too small melts too freely, has a tendency to burn, and is distributed badly. The defects of adhesion and oxide inclusions are LEARNING TO WELD WITH A GAS TORCH 147 common. If the wire is too large, tlie rate of welding is retarded, the molten bath is cooled, and it is difficult to add the metal uniformly. In both cases burning and overheating of the welding rod and material are likely to take place. The following table gives the sizes of rod for welding various thicknesses : Thickness Diameter of Wire to be Welded Decimal In Inches S. W. G. Equivalent Nearest 1/64 1/S 14 0.081 5 5/32 to 3/16 11 0.116 8 1/4 to 9/32 8 0.160 . 10 11/32 to 13/32 6 0.192 12 above 4 0.232 ir. POSITION OF TORCH The flame of the torch should be given a definite inclination. In certain cases and especially with expert welders, familiar with the Fig. 107. — Position of Torch and Eod for Backward Welding. backward method of executing welds, this inclination of the flame to the plane of the weld is very small, in other words, the flame is almost perpendicular to the weld. However, it has been found that an angle of 20 deg. gives the best results, that is to say, the angle between the nozzle of the torch and the perpendicular should be 20 degrees, the flame being turned backwards as shown in Fig. 107. Beginners should pay particular attention to obtaining and practically working at this angle of inclination. In welding backward it is the welding rod that is given a move- ment and not the torch. The torch is therefore held in such a manner that the flame advances along the bevelled faces with as great a 148 GAS TORCH AND THERMIT WELDING regularity as possible, the rate of niovenieut being in proportion to the speed of welding. A very slight transversal movement may be given to the torch to produce more rapid fusion of the two bevelled faces. The white cone of the flame should penetrate very deeply into the angle of the V as shown in Fig. lOS. If held too high as shown in Fig. 109 the melting at the bottom of the V is not sufficient, the size Fig. 108. — Position of Flame. Fig. 109. — Wrong Position. of the weld is unnecessarily increased, the metal near the surface is overheated and the speed of welding is diminished. The penetration of the white cone should be carefully observed if the advantages, economy and quality of welds, which can be obtained by welding backward are required. HOW TO HOLD THE ROD The melting of the metal is produced, as previously explained, be- hind the torch and not in front as is the common practice. This melting is not obtained by the welding cone of the flame, but by the additional heat contributed by the envelope of the flame, the torch being inclined towards the rear, in other words, towards the welding wire. The position of the welding rod and its movement should be closely followed. The rod is inclined to the line of welding, in the advancing direction, that is to say, in the opposite direction to the inclination of the flame. The best angle of inclination, between the weld and the rod, has been found to be 45 deg. for material about J-in. thick, and for thinner material, say J-in.. an angle of about 30 degrees. This in- clination is maintained whilst the welding rod is given its proper move- ment in the line of welding. This movement for the thicker material, say, about -}-in.. consists in alternately moving the molten extremity of the wire from one side to the other of the line of welding, as shown LEARNING TO WELD WITH A GAS TORCH 149 ia the small illustration in Fig.lOT. The movement for material less than this thickness becomes first of all ellipsoidal or gyratory, and then for materitil about ^-iu., and especially when the material is about Vis in., the movement is translated into a reciprocating one without any transversal movements, as shown in Fig. 110. In both of these cases the extremity of the wire remains continually in the molten bath. In order that the line of welding should present an homogeneous appearance, it is advisable to operate in the manner already laid down and with the same speed at the conunencement and completion of the work. If, say, one of the extremities of the weld is attacked too soon with the torch, free fusion and regular advancement are not obtained until after a certain time, with the result that irregularities are noticeable at the beginning of tho weld. To avoid this the plates should be preheated for a length of a few inches with the torch, so as to obtain, at the beginning, regularity, and a normal rate of welding. Fig. 110. — Rod Movement When Welding Thin Metal. The torch and the welding rod being held in the manner indicated, the cone of the flame is directed so as to well penetrate into the angle of the bevel, and the first molten bath is obtained by giving the torch a slight gyratory movement, immediately after which the extremity of the welding rod is introduced into the molten bath and the torch is then given its regular advancing movement. The welding rod, on the other hand, follows immediately after the flame, and describes a recipi'ocating or a more or less elliptical and longitudinal motion, according to the thickness of the metal, as indi- cated and shown in Figs. 107 and 110. taking care to always maintain the given angle of inclination. The weld is thus obtained in a normal and very continuous manner. Care must be taken to use a welding rod which satisfactorily fills the lines of welding, without excess or insuflicient addition of material. If necessary, the position of the torch is changed when the extremity of the weld is reached in order to obtnin a clean finish as is usual with the ordinary method of welding. The melted metal being attacked in the rear, as a result of the inclination of the flame, the bevelled faces are always well melted ; 150 GAS TORCH AND THERMIT WELDING the weld, is what is commonly said to be well penetrated and the defect of adhesion is almost impossible. However, it is advisable not to travel too fast so as to give the bevelled surface sufficient time to melt freely, otherwise candles of molten metal will appear on the underside of the weld as a result of the addition of too much heat at the bottom of the V. From the point of view of good penetration of the metal and the absence of the defect of adhesion, welding backward offers considerable advantage over other methods of executing welds and is capable of entirely eliminating these defects. Etching tests on welds obtained by this method show perfect join- ing between the metal of the plate and the added metal. The welds show distinctly less oxide inclusions than those obtained by the or- dinary methods, and are free from blowholes. In addition they possess ordinary hardness and the remaining mechanical properties are more regular. Bending tests have given good results. It is possible to fold the weld without starting a crack, which is a very good indication of excellent elongation properties and good penetration of the weld. The tensile strength of the weld is also greater than that of ordinary welds and improvement in the other mechanical properties is obtained. INSTRUCTIONS FOR LEAD BURNING The "Eveready" instruction book, issued by the Oxweld Acetylene Co., gives the following hints on lead burning: The size of the flame used in lead burning depends almost entirely upon the class of work. The 0-H3 and 1-H3 tips are used on very light sheet lead and similar work, the 2-H3 tip on heavy sheet and light storage battery work, and the 3-H3 and 4-H3 tips on general storage battery work. In all cases where lead burning is to be done, it is essential that the edges of the parts to be burned are first cleaned. Otherwise a film of oxide will form on the molten surfaces of the metal, which will tend to keep the metal from flowing together, slow down the work and quite possibly result in a poor joint. Clean the edges to be joined and also clean the surface a short distance back from the edges, either with a lead scraper or a wire brush. It is extremely difficult to burn lead which has been subjected to the action of a strong acid, such as the sulphuric acid used in storage batteries. Where it is possible to neutralize the acid by a solution of ammonia or sodium bicarbonate without getting any into the battery and injuring it, that method is allowable. It is decidedly better in all cases, however, to wipe dry the parts to be burned and then scrape them bright. The scraping will remove the layer of lead which has been affected by the acid and will insure a good joint. LEARNING TO WELD WITH A GAS TORCH 151 LEAD BURNING-STICKS OR WELDING RODS In some cases additional metal is fused in to completely fill the parts being burned or for reinforcing the joint. Where extra metal is added, the "burning sticks" or rods employed for this purpose are either pure lead or lead containing a percentage of antimony. Pure lead rods are preferable for working on sheet lead or for any part which may be subject to bending strains. Rods containing antimony are preferable where the work is to be threaded or where it must be rigid enough to withstand twisting strains, as for instance storage battery terminal posts. Hold the torch so that the flame is almost perpendicular to the surface of the work and the white cone almost touches the metal. Be careful, however, not to jab the tip of the torch in the molten lead, and under no circumstances hold the torch tip any closer to the work than may be necessary to play the tip of the inner cone of the flame upon it. In all lead burning it must be remembered that the melting point of lead is low and that as soon as it reaches the melting point it will flow rapidly and unless care is exercised it may get beyond the control of the operator. The chief thing to learn is to know when the lead is flowing properly and to lift the flame immediately from that part of the work so that no excess melting will be done. Should the metal start to run away, lift the torch and allow the work to cool before attempting to proceed. When adding metal the torch flame should be played simultaneously on the rod and along the edges of the work to be joined so that they will reach the fusion point at the same time. It does no good to deposit molten lead upon cold lead. All the lead must be melting, otherwise it will not fuse together. Do not allow the rod to touch the metal being worked upon, as it will probably stick and become firmly attached. LAP JOINTS In burning sheet lead it is always better, wherever possible, to lap the joints, that is, lay the edge of one sheet of lead i to ^ in. over the edge of the other. The overlap of both sheets must be thoroughly cleaned, not merely the edges of the sheets. After placing the sheets in position tap lightly with a wooden mallet along the line of lap to bring the two sheets together. Though lapped joints are sometimes burned without the use of burning sticks, they are not so strong as when a filler is used. In the former method the torch flame is merely played along the edge of the overlapping sheet. With a little practice, this class of joint can be made at high speed. BUTT JOINTS When the edges of the work are butted together (not lapped), it is possible to burn them together without the addition of metal. Although this makes a very neat appearing joint the use of the rol3 will insure 152 GAS TORCH AND THERMIT WELDING a stronger joint with less cliance of leaving unburned spots in the seam. For a butt joint, the sheets must be cut true and must lie true while being burned. Tapping along the line of burning with a wooden mallet about 6 in. ahead of the burn is desirable. In tacking, burn a small spot at each end of the seam and if it is a long one, burn small spots at about G-in. intervals to keep the edges from pulling apart. This is especially important on vertical seams, but it is also desirable on horizontal seams to prevent trouble that may be caused by the edges spreading. The movement or play of the torch flame is largely a matter of choice on the part of the operator. Some operators prefer a slight circular movement, progressing along the line to be burned, while others prefer to play the flame alternately from each side of the line of burning. For the beginner, the circular movement is probably the better. In burning horizontal seams lapped or butted joints may be used, and it is desirable to tack before burning. If extreme strength is desired, use the lap joint. In vertical seams lapped joints should be used, and should be tacked before burning. Start from the bottom and work upward. For overhead seams lapped joints should be used and should be tacked before burning. These seams require skill on the part of the operator, and considerable practice will be found necessary before good burning and neat resvilts are obtained. STORAGE BATTERY BURNING In battery repair work there are several operations that call for lead burning. It should be noted that great care must be used to see that the work is thoroughly scraped bright before burning and that all oxide and traces of acid are removed. For rods, use antimonial lead if certain that the plate connectors are made of antimonial lead ; if uncertain, use pure lead. Note: Antimonial lead after scraping has a silvery appearance as compared with the blue tinged color of pure lead. Antimonial lead is also much tougher and harder to scrape or pare with a lead scraper or knife. In burning plates to plate connectors, set up the plates in a burning rack or comb, which will provide for proper spacing and true align- ment. The lugs of the plates must extend above the top of the comb. Place the post in position before attempting to burn the lugs together on a lead strap. To burn, play the flame along the ends of the lugs and when they are molten add metal from the rod and form a strap connecting them all and the post. A comparatively large flame should be used to insure perfect joints because the plates must be fused perfectly to the strap and post. If a slotted plate strap or connector is used, set up the plates as described above with the lugs extending up into the slots. To burn, play the flame along the sides of the slots to bring them and the LEARNING TO WELD WITH A GAS TORCH 153 lugs to a melting point at the same time, then add metal from the rod to fill up the slots and flush the strap off smooth. BURNING CELL CONNECTORS OR TERMINALS TO TERMINAL POSTS The connectors should be tapped lightly with a small wooden mallet until they fit snugly around the terminal posts. To secure a good burn, it is necessary that tlie surface of the top of the terminal post be about | in. below the top surface of the cell connector or battery terminal. If necessary, the post should be cut off to insure this feature. To burn, play the flame on the top of the post and bring it and the inner wall of the connector to a molten state, forming a molten pool. To this add metal from the lead rod. As the pool fills up, be sure to watch that the metal on the inside wall of the con- nector flows into and with the added metal. Continue until the added metal is flush with the top surface of the connector. Then allow the connector or terminal to cool sufficiently so that the lead will not crumple when brushed, clean the top with a wire brush, and again apply the flame and add enough lead to smooth off and finish the job. It is sometimes impossible to burn on a connector or terminal in one complete operation, because the metal surrounding the cavity be- comes overheated. In sucli cases, stop work as often as the lead seems to be running too rapidly, and allow it to cool before proceeding. In burning on a terminal in which the end of a cable is imbedded, protect the rubber insulation on the cable with a strip of wet cloth, to avoid burning it. In battery repair shops it is often necessary to build up a terminal post which was drilled out when the battery was torn down. When building up a post, a mould should be used to hold the metal in place. This mould can be made of sheet metal and should be tapered so as to be easily withdrawn from the finished work. Be sure that the top of the post is in a molten state before adding lead, so that the post and the metal added will be solidly . fused. Unless this is done, the joint will be weak. LEAD BURNING DATA Approximate results obtained with Eveready Lead Burning Torch: Pressure Gas Consumption Oxygen Acetylene Oxygen Acetylene lb. per lb. per cu.ft. cu.ft. Size of Tip sq.in. sq.in. per hour per hour 0-H3 2 to 3 2 to 3 1 1 1-H3 2 to 3 2 to 3 2 2 2-H3 2 to 3 2 to 3 3 to 5^ 3 to 5 3-H3 3 to 4 3 to 4 6 to 10 6 to 9 4-H3 3 to 5 3 to 4 9 to 12 8 to 11 CHAPTER X MAKING ALLOWANCE FOR EXPANSION AND CONTRACTION Through his practice work, as already outlined, the be- ginner in welding has learned a little about the trouble that expansion and contraction may cause when proper allowance is not made. This was shown to some extent when he at- tempted to butt-weld two plates set close together. The remedy in that case was to allow for the contraction of the cooling metal and weld, by diverging the edges of the plates. From this example alone he can get a slight idea of the tre- mendous stresses often set up when, for instance, a broken spoke in a flywheel is welded Avithout proper allowance being made for the amount of expansion in heating and contraction in cooling. These stresses ynay be so great as to quickly cause other fractures, or be of such a nature as to cause the sub- sequent destruction of the wheel. Different metals conduct heat with varying degrees of speed, that of copper being much more rapid than that of steel. On this account the welder must know something of the characteristics of the metal he is working on in order to obtain good results. How- ever, except for the amount of expansion and contraction, the same general rules apply to all cases. It should be kept in mind at all times, that nothing can prevent this expansion or contraction of metals when heated or cooled, and that allow- ance in one way or another must always be made. Where the ends of a broken bar are butt-welded together the parts are free to expand as they are heated, unless rigidly held. Suppose, however, that they are free to move, then when heated the broken ends will move toward each other, pushing the two parts of the bar in opposite directions when the heated ends touch. After the weld is complete, the cooling will cause the metal to contract, drawing the two parts of the 154 ALLOWANCE FOR EXPANSION AND CONTRACTION 155 bar closer together. In somo cases the bar may be shorter than before, depending on the care and skill with which the weld was made. Owing to the fact that the parts of the bar are free to move no bad stresses are set up. Again suppose the bar happened to be part of a frame, as shown in Fig. 111. Then when heated at the break A, the ends could only move toward each other, in certain cases causing these ends to upset, or become thicker. After the weld was completed, the metal would start to contract, the tendency being to pull the cross ends in as shown at B and C. If the metal was ductile — that is, would stretch — it would probably actually bend in as suggested. Wrought iron or steel, for example, would Fig. 111. — Broken Frame with Preheating Zones Indicated. probably do this. Cast iron would probably break. Aluminum would break or bend, according to the alloy used. In any case it would be a poor job, no. matter how well the welding work itself was performed. The way to obtain a good job is to heat the frame at D and E, so that these side bars will expand as much as the middle one will while being welded. The contraction on cooling will then be the same on all three. Local heating like this is not always sufficient, and it is often necessary to heat the whole piece. Sometimes conditions are such that neither a part, nor the whole of a piece may be heated properly. We may then use a jack to open the break in the middle bar a short distance, make the weld, and then slowly loosen the tension on the jack as the metal contracts. Or we may 156 GAS TORCH AND THERMIT WELDING wrap wet cloths or wet asbestos or clay around the middle bar, close to the weld and keep cold water running on it while Fig. 112. — Preheating Furnace, Using Charcoal. welding. This simply holds the expansion to a limited area and should be employed only when no other method is possible. Undoubtedly the better method in nearly all cases is the pre- FiG. 113. — Preheating Furnace on an Iron Table. heating of the article or a portion of it, though in each case proper judgment must be exercised. The simplest way to preheat work is, as a rule, to build a temporary firebrick charcoal furnace, the form depending on ALLOWANCE FOR EXPANSION AND CONTRACTION 157 the shape of the work. Where a piece like the frame just men- tioned is to be heated all over, a furnace something like the one shown in Fig. 112 is very handy. Often an iron table and furnace like the one shown in Fig. 113 will serve for numerous repair jobs. USING HEATING TORCHES Where the nature of the work makes the use of charcoal unsuitable or impossible, a coal gas torch, kerosene torch, or Fig. 114. — Crude Oil or Kerosene Preheater in Action. Fig. 115. — Using Two Burners to Preheat a Large Gear. even the welding flame itself, may be used. The last, however, is too expensive to use except where absolutely necessary. As a 158 GAS TORCH AND THERMIT WELDING rule, a coal gas torch makes a very satisfactory means of pre- heating if it can be used. In using any preheating flame it is best to build up with firebrick or asbestos in such a way as to confine the heat where wanted. This also saves fuel. An Ox- Fig. 116. — A Gas and Preheat inf:^ Torch. weld preheater using any grade of fuel, crude or kerosene oil, is shown in action in Figs. 114 and 115. This has two burners, and has to be pumped so as to have about 25 to 50 lb. pressure to get good results. The large size weighs 110 lb. One big ad- vantage of a burner of this type is, that it may be carried any- FiG. 117. — Portable Electric Blower-Type Gas-Burning Preheater. where and used. Where a shop has a compressed-air system and illuminating gas, the type of torch in Fig. 116 will prove exceedingly satisfactory. In case a shop does not have a com- pressed-air system, but has gas and electricity, the apparatus shown in Fig. 117 may prove useful. This is made by the ALLOWANCE FOR EXPANSION AND CONTRACTION 159 Tyler Manufacturing Co., Boston. The motor driving the fan is of the universal type, operating on either alternating or direct Fig. 118. — Portable Preheater Mounted on a Stand. Fig. 119.— Iron Table With Firebrick Top. current. One motor will supply air enough for four regular size burners. Another torch is shown in Fig. 118. This also uses illu- minating gas, the air being supplied by means of an electric 160 GAS TORCH AND THERMIT WELDING driven fan. This device is made by the North American Manu- facturing Co., Cleveland, Ohio. It is claimed that from 500 to 2500 deg. F. may be obtained with this torch. For welding work of all kinds where proper alignment must be maintained, a table with a heavy cast-iron top is almost in- dispensable. Tables of this kind may be obtained from almost any of the supply firms. The Imperial Brass Manufacturing Co., Chicago, supplies a table with a firebrick top, as shown in Fig. 119. This kind of a table is very handy, as it enables the welder to construct firebrick furnaces of all kinds, and to so box in his work as to conserve all the heat possible. It also brings the work up to where he can work on it to the best advantage. COOLING WORK The cooling of steel or iron work after welding is often as important as the preheating. Some work must be annealed after it has cooled, by heating to a red heat and then slowly cooling again. Small parts may be buried in slacked lime, ashes or the like. Flat work may be laid in a sheet-iron box partly filled with lime and covered with sheet asbestos or more lime. In any case, the weld should be protected as much as possible from drafts. Where a firebrick furnace has been built up around some part, it may be closed in and the work allowed to cool as slowly as possible. The welder must use good judgment in all eases. It must not be thought that preheating is only necessary to take care of expansion and contraction, for while in small work this is often the only consideration, on large work it saves expense. By this it is meant that the use of charcoal, gas or other heating mediums is nuicli cheaper than to try to bring the parts to be joined up to a welding heat with the welding fiame alone. The way in which expansion and contraction will take effect often requires considerable study. If the work can be heated all over, this is often the best way. As already mentioned, this is often not possible, so in order to assist the beginner, a number of examples will be given showing just where certain jo])s should be preheated to get good results. A good thing to keep in mind as to where to preheat, is to imagine a wedge driven in at the break and note what places this action would put under strain. ALLOWANCE FOR EXPANSION AND CONTRACTION 161 THE WIEDERWAX PREHEATER A preheating furnace suitable for both preheating and slow cooling is shown in Figs. 120 and 121. This heater is made Fig. 120. — The Wiederwax Preheater. i **;■' * '\ eat^A^' Fig. 121. — Showing the iSlow Cooling Oven. by the Geist Manufacturing Co., Atlantic City, N. J., and is known as the "Wiederwax preheater. It has. eight gas burners 162 GAS TORCH AND THERMIT WELDING entering from each end and extending to the middle. This makes it possible to heat any section or all of the top, as desired. The top has parallel grate bars to support the work and the gas burners are buried in pieces of refractory, heat-retaining material, so that parts are heated with the use of a minimum amount of gas. After the welding is done, the work is placed in the oven underneath the heaters and slowly cooled. When using this heater or welding at any time, the work should be covered with asbestos sheet as much as possible. This concern also makes a floor-type of preheater for heavy work. SUGGESTIONS REGARDING THE WELDING OF GRATINGS AND PULLEYS S. "W. Miller, of the Rochester Welding Works, writing in the "American Machinist" says: It should be clearly understood by the welder that in a restrained weld, which is one entirely surrounded by metal, such t o B__ 1 '— A I Fig. 122.— A Restrained Weld. Fig. 123.— Grating to be Welded. as a crack in the center of a plate, it is impossible to get rid of both strain and distortion ; a moment 's thought will make this clear. Such a crack is shown in Fig. 122. When crack A is welded the metal between B and C is heated hotter than that between D and E, and F and G, and, being soft, the expansion crushes it; so when cooling occurs, B to C contracts more than the rest of the plate, which either causes a crack, or leaves a tensile strain in the plate. It is true that the strain can be re- lieved by annealing, but that leaves a distortion of some kind; if the distance BC be the same as originally, the thickness along BC must become less, and vice versa. If this is not clear, let the welder heat a spot in the center of a |-in. plate of steel, and see what happens. In a partly restrained weld, such as Fig. Ill, it is evident ALLOWANCE FOR EXPANSION AND CONTRACTION 163 that the method of preheating described produces expansion of the part to be welded in such a way that the opening of the crack is tlie same at all points in its lengtli. In other words, there is no variation in the width of the crack after expansion and before welding. This is an important principle to be noted in expansion and contraction problems; it may be worded as follows: A crack must always be opened the same amount at every point in its length, if both strain and distortion are to be avoided. Fig. 123 shows the condition of a grating when received at the welding shop. The quickest and cheapest way to weld this, if three men are available, is to double bevel all breaks; use three men to weld A, B and D at the same time, first on one side and then on the other; clamp the piece on a flat table to /Heating Zone-, Fig. 124. — Eesult of Improper Fig. 125. — Welding a Broken Heating. Pulley. keep it straight and let it cool. Then have one man heat at E with a torch, and let another weld at C as at the other points. If only one man is at hand, he should weld B first; then, keeping bar Z> hot, weld B; next, keeping both B and D hot, weld A; allow the piece to cool, heat E and weld as before. This brings out the principle that one should never finish a weld in the center of a piece, unless it is absolutely necessary ; always finish at the edge. It will be seen that the method just described keeps the sides of the cracks parallel, and that if the proper allowance is made for contraction, there will be no strains left in the welded piece. The matter may be made clearer by a study of Fig. 124, w-hich indicates the result of improper heating. Heating the corner H for crack B will lengthen sides A and C, the resultant of these being in the direction of D ; so that the upper part of 164 GAS TORCH AND THERMIT WELDING crack B will lie as in the dotted lines, and after welding and cooling there will be strains at E, F and (/ which will surely cause cracks in service, if not at once. With a smaller grating it is possible that there would be spring enough in the bars to avoid breakage; but there would be strains that could be avoided by other methods. The basic principle of all welding is to weld without having either strain or distortion in the finished piece. With regard to pulley welding such as shown in Fig. 125, the rim should be heated on both sides of the broken spoke to lengthen the rim and pull the crack A open, and the spoke should be kept as cold as possible. Also all spokes should be welded from both sides. The break at B should be opened by heating the spokes next to it so that when the crack is open enough, the edges of the break will be the same distance from the center; this of course means different amounts of lieat in the two spokes. AUTOMOBILE CYLINDER WORK The next example is a block of two cylinders, broken as indicated in Fig. 126. The first break to be repaired is in the water jacket at A. The second break is on the flange at B, and the third is on the water jacket as shown at C. The fourth crack D is on the inside of the water jacket, necessitating the removal of part E by drilling in order to get at it. To weld any of these cracks, except B, without preheating would break the casting from expansion w^hcn the flame was applied to it, and contraction when the weld cooled. The pre- heating in this case means the heating of the entire casting alike. To do this, build a furnace of firebrick, as shown at F. In order to give enough draft for the fire, the bottom bricks should be set about 1 in. apart. The cylinder block, with the cracks properly chipped and beveled, may then be placed on the first layer of brick and the walls built up around it. These walls should be so built as to allow about 6 in. between them and the cylinder. That is, there should be space enough allowed to turn the cylinder without knocking down the walls when ready to weld. About three or four shovelfuls of charcoal should he put around the cylinder and a little kerosene put on it and lighted. After the charcoal has become thoroughly lighted and ALLOWANCE FOR EXPANSION AND CONTRACTION 165 the cylinder slightly heated, more charcoal should be added until half the casting is covered. Then a piece of sheet asbestos ■^ ■^. - ' ^---r Tr^^-Tg-St= _ -^-- Air Holes / Fig. 126. — Broken Motor Cylinder Block and Preheating Furnace. should be put over the top, and a few holes punched in it for draft. Where the break is bad, leave the cylinder alone until 166 GAS TORCH AND THERMIT WELDING it reaches a dark red all over, then remove the asbestos and turn the cylinder so that the part to be v^^elded first is upper- most. Then replace the asbestos and cut a hole in it so that the break can be reached with the torch and welding rod. Never take the cylinder out of the fire to weld. In working on automobile cylinders it is usually unnecessary to preheat at all when welding lugs; in cases where it is neces- sary, only a very slight amount is needed — just enough so the cylinder cannot be handled without tongs. Usually ordinary jacket cracks can be welded at much less than a red heat. In fact if a red heat is used, in most cases the cylinder bores will warp badly. The important points in cylinder welding are a uniform soaking heat, not necessarily very high, welding on rising tem- perature, and slow cooling. Never weld an important break in a cylinder block without about three hours slow preheating, never allowing the fire to die down. Then pack it in loose asbestos in the fire, to insure slow cooling. Of course, there are exceptional cases where the break is so bad that high preheating is needed. Care must be taken not to let melted metal run down into the water jacket. Be sure to work out all dirt or scale, and do not leave any pinholes or blowholes. In order to prevent the bore of the cylinder from scaling, it should be coated with oil and a thin coating of graphite applied before the cylinder is placed in the furnace. After welding the graphite can be cleaned off with a piece of waste. For crack B, the cylinder seldom needs to be preheated, as previously mentioned. The inner weld C is made the same as crack A. Crack D, being on the inside of the water jacket, is treated differently. A portion of the outer wall of the water jacket is removed by drilling, as previously mentioned. The inner weld is made, after which the removed portion is replaced and welded. In order to hold the piece in place while being welded a cast-iron rod may be welded to it to serve as a handle. After the weld is finished, this rod is cut off. After the cylinder has been welded and cooled off, it should be tested to be sure that the weld is entirely tight. "Where it is possible, this should be tested with water pressure. If it is not possible to do this, the water jacket should be filled with ALLOWANCE FOR EXPANSION AND CONTRACTION 167 keroselie, because kerosene penetrates a crack or a pinhole faster than water. In case any leaks are found, the metal should be chipped out at that point, placed in the fire, and rewelded exactly as before. The next job, illustrated in Fig. 127, is typical of a large class. Work of this kind is usually in cast-iron, though occa- sionally of steel, or semi-steel. The piece shown is a cast-iron shear arm broken where the section is about 6 or 8 in. thick and 16 in. across. In preparing a casting of this size, the break is beveled at 60 deg. on each side, using a 12-lb. sledge and a handled chisel. It is then lined up and a preheating firebrick furnace built around the break, allowing sufficient space for the charcoal around the work. Then heat to a bright red well back from the break. Two welders should work on a job of this kind. ^■yf-^^.-Weld here Fig. 127. — Broken and Welded Shear Arm. each welding about half an hour or less, at a time. A large welding head, a long-handled torch and |-in. welding rod are used. Two rods should be used, welded end to end. There should be plenty of rods and plenty of flux. There should also be a good supply of asbestos sheet and a bucket of water. The sheet is to keep in the heat and to protect the welders, and the water is to cool the torches. After the weld is com- pleted on one side, and while the metal is still pretty hot, the casting is turned over. The movement will naturally destroy the firebrick furnace, which has to be rebuilt again, the casting heated up as before, and the weld made. In a job of this kind it is necessary to reinforce the weld on both sides. This rein- forcing should be about \ in. high. After the welding is all done, the piece is heated up to an even red heat all around the weld and allowed to cool very slowly. In welding in or building up gear teeth, as shown in Fig. 168 GAS TORCH AND THERMIT WELDING 128, it is seldom necessary to preheat. If the gear is a light one, say about 3 in. wide and with a rim depth of not over 1 in., the job can usually be done without preheating. However, pre- heating with some cheap fuel will, on large work, save the more expensive gases. In doing a job of this kind the greatest care must be taken to start properly. The work should be done so ■rinhhed Weld\ Fig. 128.— Method uf Filling in Gear Teeth. as to do away with as much machining as possible. Sometimes the tooth may be built up about as shown at A and finished by filing or otherwise. At other times it will be necessary to weld in on teeth each side of the one to be replaced, as shown at B. Again it may be possible to use carbon blocks and fill in as shown at C. In using blocks of this kind, care must be taken that there is ample room at the bottom for the root of the tooth being replaced. CHAPTER XI WELDING VARIOUS METALS AND THE FLUXES USED The first property to be considered in welding various metals is the melting point. Gas-torch welding is the joining together of two metal parts by fusion at the line of contact and in order to secure a perfect weld it is necessary that each part be melted, and the molten metal allowed to flow together and harden in this state of mixture. The approximate melting points and other properties of the metals and alloys commonly welded are given in Table XVI, taken from "Oxwelding and Cutting." When metallic bodies are subjected to an increase in tem- perature they expand the rate of this expansion, being closely known for each degree of rise in temperature. When the tem- perature is lowered a reverse action takes place, the bodies contract and the volume and linear dimensions decrease. This has been explained to some extent in the previous chapter, and examples given to show the effects. Each metal has its own co- efficient of expansion, which varies materially for the different metals. As seen from the table given, of the metals most com- monly welded aluminum expands the most, bronze and brass next, then copper, steel, and iron. Aluminum expands almost twice as much as iron or steel. Conductivity and Oxidation. — The conductivity of a metal i?- its property of transmitting heat throughout its mass. This property is not the same for all metals, and varies widely. It is commonly called thermal conductivity. It can be seen that if one metal conducts or transmits the heat from the torch flame more rapidly than another, it is necessary that allowance be made as to the method of handling the job, the size of the torch, and the nature of the preheating equipment used. In welding metals of high thermal conductivity it is neces- sary to use oversize tips — as in the case of copper where the 169 170 GAS TORCH AND THERMIT WELDING ^ o ■u c3 tn • , : O 6 d5 'f^-i 05 ■*^ J-< ^ ri § g 03 OJ 2 6°| Q +3 t- _s ie ^ •o :8 (N ^ ■•o • o :^ • o •Q '•^ * 02 c^ P5 •?? •CO •h* :B &> o ^ o o X O sz WELDING VARIOUS METALS AND FLUXES USED 171 melting point is low and the conductivity high. However, too large a flame is bad, because the operator will not be able to correctly place the mass of molten metal. On sheet work the proper flame will melt the metal to a width about equal to the thickness of the sheet. When welding heavy work the operator should be very care- ful not to blow a part of the molten metal on to the colder por- tions as it will make a defective weld at that point (called an "adhesion"). If this should occur, the flame should be played over this chilled portion until it is in fusion with the molten metal. Certain metals oxidize more rapidly than others. Oxidation is the reaction produced by the combination of oxygen with a metal. The weld may become oxidized by contact with the oxygen in the air and by the presence of excess oxygen in the welding flame. An oxide has none of the metallic properties of the metal from which it is formed. When present in a weld it seriously weakens it and it is therefore very necessary that it be avoided as far as possible. Some oxides are lighter than the metal itself, while others are heavier. Consequently, when a metal is reduced to a molten condition the oxide will either float on the surface of the liquid metal, remain suspended, or tend to sink toward the bottom. The melting point of oxides is in some cases higher, and in others lower, than that of the original metals. This point must also be considered in attempting to eliminate oxide from a weld. Some metals when molten also have the property of dis- solving a portion of the oxide, the extent of this solution being dependent upon the metal itself. When this is the case the oxide is retained in solution until the metal hardens, in some cases separating and producing a weakened weld, in others being retained permanently. Oxide may be dealt with in two ways. First, by taking means to prevent its formation, by the use of a neutral or re- ducing flame in the torch, or by the use of various cleaninr^ fluxes ; second, by eliminating the oxide after its formation with suitable fluxes^ which either dissolve or float it off, or by mechan- ically removing it by the manipulation of the welding rod or a paddle made for this purpose, 172 GAS TORCH AND THERMIT WELDING The subject of oxidation is one of vital importance to the welder, one that he should study thoroughly in order to become familiar with all its forms. Oxidation is the cause of a great majority of defective welds. There are also metals, that, when heated to the melting point, have the property of absorbing gases from the flame. "When the metals cool, the gases are released. In a great many cases the release of the gases occurs at a time when the metal is not sufficiently fluid to allow them to pass to the air. Consequently, small bubbles or blowholes are incorporated in the weld. Then, too, in the welding of various metals the force of the torch flame causes the molten metal to flow back away from it. When the flame is withdrawn the molten metal returns — similar to the action of any other liquid. At such times the return of the molten metal may be so rapid that small quantities of the gases become entrapped and remain in the weld as blowholes. This is a very common occurrence in cast-iron welding. In the case of the absorption of the gases by the hot and molten metal, the difficulties may be overcome by the use of proper protecting and cleaning fluxes and properly prepared welding rods. In the case of the gases being entrapped by the molten metal, this may be overcome by "working out" the gas by means of the torch and welding rod. Vaporization of Substances. — In the manufacture of metals substances are combined in amounts which determine the behavior and characteristics of the metal. In iron and steel there is a certain amount of carbon, silicon, manganese, phos- phorus, sulphur, etc. While these substances may be present in only very small quantities, yet their elimination, or presence in excess, may materially affect the mechanical properties of the metal. The high temperature of the welding flame may cause these substances to burn out or to volatilize. They can burn or oxidize directly in the oxygen of the atmosphere, in excess oxygen in the welding flame, or by the reduction of the oxide of the metal formed in melting. In the working of brasses, bronzes, or an alloy in which zinc is present, it is commonly observed that the zinc vaporizes and passes off as heavy white fumes in the form of zinc oxide. WELDING VARIOUS METALS AND FLUXES USED 173 It can be seen that when this occurs the zinc content of the alloy is materially reduced, and consequently the resultant weld will not have the same mechanical and chemical properties as the original metal. Special fluxes are provided to prevent this ; also, welding rods can be obtained which will either prevent the vaporization of the volatile substance or will replace it. Separation of Elements. — Alloys are uniform mixtures of metals. The fusion of the different elements composing an alloy is carried out at a certain fixed temperature. In the welding of metals of different kinds, it has been noted that when some of these alloys have been heated to the high temperatures pro- duced by the welding flame various substances separate or segregate and that it is impossible to secure a uniform weld. This segregation occurs quite frequently under the torch flame and also occurs in the manufacture of the metal. Under these conditions the difficulty of welding some alloys can be readily seen. Welding Various Metals. — With the foregoing facts and suggestions in mind we will now take up the various metals most frequently welded, and give directions that will apply in their special cases. The composition of various fluxes will also be given, but it should be remembered that the different acces- sory concerns can supply far more satisfactory fluxes than can be made in small lots by the individual user, and the welder should, where possible, buy the fluxes needed and apply them accord- ing to directions. This also applies to welding rods which should be bought from reliable concerns for certain specified jobs. For emergency work, where the proper rods are not available scrap material or wire may be used, but it is not good practice. A first-class welder who cares for his work and his reputation will use rods of the proper chemical composition for the work he has to do. For this purpose he should buy his rods from firms of e^ablished reputation, who are not afraid to advertise their output. Welding Aluminum. — Aluminum parts to be welded may be divided into two classes — those made of drawn or rolled aluminum and those which are cast. Rolled aluminum is usually 98 per cent pure or better, the main impurities being silicon and iron. Aluminum as pure as this is seldom used for eastings, since its strength is considerably 174 GAS TORCH AND THERMIT WELDING less than that of various alloys. Zinc in amounts ranging from 5 to 25 per cent, but usually about 10 per cent, was often used in the past, but the alloy was so brittle just below solidification that a large number of castings were defective owing to shrink- age cracks. A copper alloy is now more commonly used, the copper content being less than 15 per cent, 7 per cent probably being the favorite. This is not so strong at ordinary tempera- tures as the zinc alloy, but it does not have such a tendency to crack. This makes it much better for welding as well as for easting, especially on complicated work. Aluminum oxidizes easily in the air, especially at high temperatures, and in the latter condition the oxide coating is quite thick. This oxide melts at a nuich higher temperature (5000 deg. F.) than aluminum (1215 deg. F.) and as the oxide is of greater specific gravity (heavier) than molten aluminum, it will sink down into the metal when welding unless it is re- moved in some way. As the oxide is very persistant to the action of any acid or alkali, even at a high temperature, any flux used must of necessity be drastic in action and if carelessly used, exceedingly injurious to the aluminum weid. On this account any flux should be used with caution and any surplus removed as soon as possible. Table XVII.— Fl UXES FOR Welding Aluminum Chemicals Sodium Chloride Potassium Chloride Lithium Chloride 1 % . 30. . 45. . 15. . 7. . 3. 2 % 33.3 33.3 33.3 Formula N 3 4 % % 12.5 IG. G2.7 79. 20.8 . . . 4 5. UMBERS 5 • % 17. 83. G % G.5 56. 23.5 4. 10. 7* % 30. 45. 15 Sodium Fluoride Potassium Fluoride ~). Sodium Bisulphate Potassium Bisulphate Sodium Sulphate Potassium Sulphate 3. Aluminum Sodium Fluoride. * Recommended by the French Laboratories of the Autogenous Weld- ing Association. A flux is generally used in welding sheet aluminum where the puddling method cannot well be employed. More divergence WELDING VARIOUS METALS AND FLUXES USED 175 niu55t be allowed than for iron. Fluxes are usually composed of alkaline fluorides, chorides or other combinations as shown in Table XVII. However, these and other flux mixtures are only given for reference purposes, and it cannot be too strongly urged that all welders buy the fluxes used and follow directions in each case. Where a flux is used in welding aluminum, the edges and adjacent surfaces should be well scraped and cleaned as the flux is only intended to eliminate the oxide and not grease and dirt. In welding heavy sheets the edges should be beveled and in light ones the welding will be aided by flanging the edges about Vio ii^- Aluminum castings are handled a little differently from sheets or plates. As previously mentioned castings are of differ- ent composition. Since the metal has a low melting point, high conductivity, and becomes rather fragile previous to fusion, preheating and cooling must be carried out very carefully. The average aluminum casting is somewhat complicated in its design, hence the necessity of skillfulness in carrying it through the preliminary heating period. The use of ^ flux on aluminum castings has been abandoned by the majority of w^elders. In place of it they break down and remove the oxide by means of a paddle, which is also used to smooth off the surface of the weld after it is completed. In many cases it is an advantage when working on castings, not to bevel the edges. In most cases aluminum articles should be preheated to some extent before welding. In certain cases the playing of the secondary flame on the object will be sufficient ; in others a more thorough treatment is required, such as charcoal or coke. On the regular run of castings it is safest to preheat to about 500 or 600 deg. F., which on iron would correspond to a low red heat. In the case of a broken lug or piece of a flange, it is often really dangerous to preheat as it may cause the whole piece to collapse or distort. The beginner should also be very careful about shifting or turning a hot aluminum casting as it may get out of shape or crumble into pieces. Since the metal is so apt to crumble when hot it is advisable for the beginner, and often the expert, to back up the parts. This may be done by molding a backing out of asbestos fiber 2 parts and plaster 176 GAS TORCH AND THERMIT WELDING of paris 1 part, made into a thick paste with water. Have this mold about an inch thick and perfectly dry before setting in place. Fireclay may also be used in many cases to back up or support fragile parts. When the weld is completed the casting should be allowed to cool very slowly and evenly. The iron puddling rod should not be allowed to get too hot or oxide of iron will be formed and scale off, making a defective weld. Only a small amount of metal from the welding rod should be added at a time and this must be thoroughly stirred or "puddled" until a pool is formed that insures perfect fusion with the surrounding parts. Use the puddling rod to scrape off surplus metal while it is in a pasty condition. The beginner will find it a little difficult to manipulate the puddling and the welding rods alternately with the same hand but this becomes a habit with practice, and many do this by holding them between the fingers so that neither needs to be laid down. The property of conducting heat is greater in aluminum than in iron, but as the melting point is much lower, about the same size torch tip is used as for cast iron of corresponding thickness. Filling a Large Hole. — The Journal of Acetijle7ie Welding says that when the filling of a large hole is required a chill of galvanized iron is provided, backing up the hole and welding against this when filling the hole with aluminum. Galvanized •iron is preferable to any other material, such as tin or iron, since it peels away from the aluminum quite readily, and can therefore be easily removed after the weld has been completed. This is undoubtedly due to the zinc content of the galvanizing composition. The chief value of the use of the chill is that it causes the filler to cool and harden quickly, thereby preventing it from contracting after the weld is finished. It also prevents the heat of the weld from spreading, which might cause the job to crack back. The chill causes the added metal to cool almost as fast as it is connected to the edge of the break. After the weld has been finished and cooled, the chill can be removed by gently prying it away from the weld by means of a cold chisel. Brass and Bronze. — The composition of brasses and bronzes varies so widely that it is not good for a welder to use welding rods of the same composition for the general run of repair WELDING VARIOUS METALS AND FLUXES USED 177 work. However, rods of Tobin bronze or manganese bronze are very satisfactory for all-round work. Where it is important to match the color of the weld with that of the surrounding metal it is necessary to use special rods of practically the same composition as the welded metal, and also to use extra care with the torch. This latter will be better understood when the welder knows that several of the alloys such as zinc, tin, etc., used with copper to make brass or bronze, volatilize easily and in so doing change the character of the metal. These metals should be prepared in the same way as any other. They must be so placed as to not move during the welding. Fireclay may be used to back up pieces in danger of collapse. The end of the flame cone should not touch the metal, but should be kept some distance above it. If a white smoke rises, or, as in the case of bronze, the metal bubbles, remove the flame, as it indicates that too much heat is being used and some of the elements are passing off in vapor. Do not breathe this vapor as it is poison- ous. It is desirable to use a tip about the same size as for cast iron. A flux should be used though not too liberally. Calcined borax is good. Boracic acid is also good and, if used, may be applied by dipping the hot rod into the powder from time to time. The principal points to watch are not to heat too hot ; do not move the parts until well cooled ; do not use too much flux, and be sure to guard against caving in or distortion of the work by properly supporting it previous to heating. If the metal is porous on cooling it is a sure sign that too high a heat was used. Cast Iron. — When cast iron is molten it oxidizes very rapidly. The oxide which begins to form at a bright red heat, melts at a temperature of 2400 to 2450 deg. F. Since the metal itself melts at a temperature several hundred degrees below this, it can be seen that the oxide will not be melted at the same time as the metal. In order to break the oxide down and allow the metal to flow together a flux must be used. A properly formulated flux will dissolve the oxide and float it to the surface, so that it may be removed by scraping the molten surface with the end of the welding rod. The welder should tap the end against something to free it from oxide before con- tinuing to add it to the weld. A good flux for use in welding cast iron may be made up of 178 GAS TORCH AND THERMIT WELDING equal parts of earlionatc of soda (washing soda) and bicarbonate of soda (baking soda). There is practically no advantage in using the pure chemicals in this case as the commercial product, which may be obtained at the grocery store, will do as well as any. The two sodas should be thoroughly mixed, however, which may be done by running them through an old coffee mill several times, or thoroughly shaking and sifting in a sieve. The flux is applied, as in most other cases, by dipping the hot welding rod into it. Cast iron is quite fluid M^hen melted. For this reason it offers considerable difficulty where vertical or overhead welding is attempted. Also its fluidity causes it to entrap gases, dirt, and oxide. These may be removed by proper manipulation of the torch and welding rod. As the molten iron can be forced ahead of the weld very easily, adhesion to the cold metal will result, if the welding is not watched carefully. The silicon will volatilize to some extent in the molten metal and the lowering of the amount of this constituent will seriously increase the hardness of the metal. In order to compensate for this loss, a welding rod is used that contains from 2.75 per cent to 3.5 per cent silicon. The other substances such as sulphur, manganese, and phosphorus should be kept within rigid limits. The welding rod should be soundly cast, free from dirt, sand, scale, rust, etc. The welding flame should always be neutral. The flame should be applied to the weld at such an angle that the metal will not be blown ahead. Inasmuch as the metal is quite fluid when molten, the welding is carried on in a series of overlapping "pools" or puddles. The welding rod is applied by placing it in these pools and playing the flame around it. The welding is aided by continually "working" the rod in the weld in order that blowholes, dirt, scale, etc., will be forced out. The central jet of the flame should never impinge on the molten metal. It should be held Vs ii^- to ^Ao i^- from it. Occasionally it is necessary to remove a blowhole, in which case the hole is burnt out with the flame and then the metal is worked over with the welding rod. The working over of a weld should be avoided unless it is absolutely necessary. If it is necessary to do this the welding rod should be used always, for otherwise a portion of the silicon will be lost. WELDING VARIOUS METALS AND FLUXES USED 179 When the wold is finished and it is still hot, the accumula- tion of scale, dirt, flux, etc., on the surface should be removed by scraping with a coarse file or other tool. This is a superficial coating that, when cold, is very hard. As soft welds are nearly always desired, the casting should bo cooled slowly and evenly. Where the work is complicated or of heavy section, it is by all means best to reheat it to a good red heat and then allow is to cool slowly. In some cases where charcoal has been used it is sufficient to allow the casting to cool in the preheating fire, without the additional reheating. Cast Iron to Steel. — To weld east iron to steel, east-iron rods must be used as the welding material. The steel must be heated to the melting point first, as cast iron melts at a lower temperature. A very little flux should be used. Copper. — Copper usually is produced in an almost pure homogeneous form. The impurities are present in small amounts and are not affected materially by fusion. Copper is a good conductor of heat, and is very tough, ductile, and malleable. From these properties it would appear that it is easily welded. Unfortunately this is not true. There are few welders skilled in the handling of this metal. Copper has two very pronounced properties under the weld- ing flame. It absorbs gases very readily, notably carbon monoxide and hydrogen. These are released when the metal begins to solidify, with the result that they remain entrapped, producing a porous structure. Copper oxidizes very rapidly when undergoing fusion. The molten metal has the property of dissolving the oxide thus formed. It will take up such large quantities of it that the mechanical properties of the weld will be affected. In addition to these two peculiarities the tensile strength of copper de- creases rapidly as the temperature is raised, particularly from 500 deg. F. upward. The efifect of temperature is so severe that at 900 deg. F. the tensile strength is only 40 per cent of that at atmospheric temperatures. Because of this weakening under heat, the strains resulting from contraction in the weld during cooling must be carefully dissipated, otherwise the metal in the weld or adjacent to it will fail. A neutral flame should always be applied in welding this 180 GAS TORCH AND THERMIT WELDING metal. If an excess of acetylene is used the products of com- bustion are richer in those gases which are easily absorbed. If an oxidizing flame is used, the weld becomes saturated with the oxide. A larger size torch tip than the melting point of copper in- dicates is used because of the high thermal conductivity. Where possible, auxiliary heating, such as air-gas flames and charcoal fires, should be employed. This is done not only from the standpoint of economy, but it also aids greatly in the success of the weld. The torch flame should play on the weld in a vertical direc- tion. The metal when molten is quite fluid, and for this reason if the torch were applied at an angle the metal would be blown ahead, producing adhesion. Also, by applying the torch vertically the molten metal is protected from the oxygen of the atmosphere ])y means of the enveloping flame. Copper, if properly prepared and free from grease or dirt, does not need a flux. The factor that contributes most to the successful welding of copper is the use of a properly formulated welding rod. Such a rod will overcome, to a great extent, both the absorption of gases and the solution of the oxide. It is not considered prac- tical to remove the oxide in the weld by means of a flux, because it is dissolved in the metal. A welding rod is needed that has combined with it a reducing or deoxidizing agent. The reducing agent has a greater affinity for oxygen than copper, hence it combines with it and brings it to the surface in a fluid form. This material acts as a glaze or protecting coating for the molten metal beneath it, with the result that it tends to retard the absorption of the gases. Several materials, when added to a pure copper rod, have proved to be beneficial. The most prominent element at this time is phosphorus. It should be present in amounts not over 1.0 per cent ; otherwise the metal will be pasty and the weld will be weakened. Newly welded copper has only the strength of cast copper, but after welding, the grain of the weld and the metal ad.jacent can be improved by hammering at a low heat. Copper to Steel. — If it is desired to weld copper to steel, heat the steel to a welding heat, then place the copper in con- WELDING VARIOUS METALS AND FLUXES USED 181 tact. The metals will then fuse together. Take away the flame as soon as the copper flows properly. No flux is needed. Lead. — Lead can be readily welded. The process, how- ever, is usually known as "lead burning." The gas torch pro- vides a means of doing this work quickly and at a low cost. Skill in the manipulation of the torch is necessary, particularly on vertical seams. A light torch should be used. "When welding sheets or plates, proceed as in lead burning by other processes. The burned joint on a lead or block tin pipe line is not only a neat and permanent joint, but is all lead, and a block tin line is all block tin. The joints are fused together with the addition of enough metal of the same kind. If, for instance, a lead pipe line is to carry acid, the burned joints contain no solder which could be attacked by the acid. In preparing lead pipe for welding the two pipe ends are scraped clean for about an inch back, and are tapered slightly at the edges. It is not necessary to drive one pipe into the other. The two ends are merely placed in contact and welded or "burned" with the addition of more lead to fill up the joint. No flux, no grease and no "wiping" of any kind is needed. Malleable Iron. — The manufacture of malleable iron has made enormous progress during the last few years, and while formerly malleable iron was really an unknown quantity and might contain different mixtures, from white iron to hard steel, in the same casting, a great uniformity is now obtainable by adherence to strictly scientific methods. The Associated Manufacturers of Malleable Iron has set a standard of quality, to which all its members must adhere rigidly and castings procured from one of its members may be relied on to consist of a uniform and thoroughly high-class product. While formerly the welding of malleable iron was considered almost impracticable on account of the different structures in one and the same casting, it may now be welded with almost a cer- tainty of success, if the casting was made in accordance with the rules of the association. The break on malleable iron is prepared exactly the same as for any welding job, cleanliness in this instance being espe- cially desirable, since dirt tends to weaken the weld considerably. Allowance should be made for the effects of expansion and con- traction; malleable iron is less liable to break than cast iron. 182 GAS TORCH AND THERMIT WELDING since it is ductile, but will be distorted unless such provisions are made. Use for flux the same powder used for brass — that is : borax or a purchased mixture. As with cast iron, do not let the end of the cone touch the casting, but hold it just a little distance away. Watch the metal carefully and as soon as the metal begins to melt, add the filling rod, either Norway iron or malleable rod of the same grade as the casting. However, not all malleable castings are of the high degree just described. They were originally white cast iron, very brittle and hard. By heat-treatment the carbon content is changed, and instead of the brittle casting, it becomes ductile, fairly soft and changes to a darker color. Just how far into the body of the metal this change penetrates depends upon the size of the casting and the length of the heat-treatment, so that a malle- able casting, as it is generally called, may be steel on the surface, a semi-steel part way through and white cast iron at the core. Very small castings sometimes are steel all the way through and we may weld them without flux, using Norway iron or mild steel as the welding rod. In nearly all cases, however, it will be found that the casting, if not made to association specifications, is composed of different metals — if the break is examined, we can tell this by the differ- ent colors. It is obvious that such a casting cannot be welded, since it would be extremely difficult to determine just where one metal left off and another began. The practice of using cast iron as a welding rod on malleable castings is not a good one, since the bond is very brittle and in all cases where strength is desired we would better use manganese or Tobin bronze — in this way securing a brazed joint instead of a welded one, of a different color than the casting but with the factor of strength a big one. Watch the metal carefully and when the spot the flame is playing upon reaches a bright red heat, bring the bronze welding rod, which has previously picked up some borax, down upon this section, being careful that the cone does not come directly in contact with the bronze rod. Bronze melts at a lower tem- perature than malleable iron and with the iron at a bright red heat, and with plenty of flux used, it will be found that the bronze attaches itself to the iron. We must not, however melt WELDI^IG VARIOUS METALS AND FLUXES USED 183 any portion of the malleable iron and we must not play the cone directly on the iron or on the bronze. Monel Metal. — Technically, monel metal is an alloy of nickel and copper, containing about 67 per cent nickel, 28 per cent copper, and 5 per cent of other elements. This remaining 5 per cent consists partly of iron from the original ore and partly of manganese, silicon, and carbon introduced in the process of refining. It contains no zinc or aluminum. The alloy can be machined, forged, soldered and welded, both electrically and by the gas torch. In the automobile industry it is used for float A'ulves in carburetors because it combines hardness with non- corrodibility. Borax or boracie acid may be used as a flux. Nickel. — Nickel melts at 2600 deg. F. and when melted has the property of absorbing large amounts of various gases, espe- cially oxygen. The gases so absorbed remain when the metal cools, making it very porous. Nickel has also a great affinity for sulphur. It is often stated that nickel cannot be welded, but this is an error, although it is an extremely difficult metal to weld satisfactorily. Anodes used in nickel plating may be fused together without flux as the blowholes do not affect the conductivity to any appreciable extent. However, where a weld is wanted free from blowholes, the nickel pieces should be laid on a heavy plate heated bright red or white, and the nickel heated to a bright white with the gas torch. The joint should then be carefully hammered with a light hammer. Previous to heating the nickel should be freed from grease or oil and scraped well back from the weld. Steel. — Steel welding on a commercial scale should never be attempted until after the operator has proved to his own satisfaction that the weld is strong by welding together mild steel plates -I to :J in., sawing them through the weld to make sure that the material is really bonded and testing them by bending back and forth in a vise. Steel melts at 2500 to 2700 deg. F. When molten it is not extremely fluid. At dull red heat it begins to oxidize rapidly. The oxide, which melts at a temperature of several hundred degrees below that of the metal, remains at the surface and can be easily removed. A flux is not necessary, although some welders use a little borax. Close attention must be paid to the removal of the oxide, however, for its presence is very harmful. 184 GAS TORCH AND THERMIT WELDING It is a common fault to have layers of oxide in the weld, which cause a laminated structure that weakens it. Steel does not melt rapidly. It gradually comes to fusion, confined to small areas. Because of this, the weld is made up of small overlapping layers. The strength of the weld depends greatly on the thorough bonding of these layers to each other and to the beveled edges of the piece being welded. It is a common fault to force the metal ahead of the welding area and allow it to adhere to the cold sides of the beveled edges. A welding rod of over 99 per cent pure iron wire is com- monly used. Occasionally a nickel steel rod is used with good results on such work as crankshafts. A mild-steel rod is par- ticularly satisfactory on steel castings. Diameter of Thickness of Steel Welding Rod 1/8 in '. 1/16 in. 1/4 in. to 3/lG in 1/8 in. 1/4 in. to 3/8 in 3/16 in. 1/2 in. and up 1/4 in. Steel is very sensitive to the welding flame. An excess of acetylene tends to carbonize the metal ; an excess of oxygen tends to oxidize. Therefore, a neutral flame should always be used and should be tested frequently in order that it be kept in proper adjustment. Failures due to expansion and contraction are not numerous, because of the toughness and strength of the metal. If expan- sion and contraction are not properly taken care of, however, warping and buckling will surely take place, and internal strains will exist in the weld. These can be avoided by properly setting up the work and with proper preheating methods. The strength of a steel weld can be improved by mechanical treatment. Hammering is the most common method employed. After the welding has been completed, the entire weld should be heated to a bright red heat, and. the hammering carried on at this temperature. If the hammering is done at a lower temperature, the weld will be weakened instead of strengthened. The welder should always keep in mind that the higher the percentage of carbon in the steel the greater is the danger of burning the metal, with its consequent weakening effect. WELDING VARIOUS METALS AND FLUXES USED 185 Steel castings should be handled in a manner similar to cast iron. They may be preheated and prepared in the same way. C'ast steel as a rule has a percentage of carbon between that in mild steel and gray cast iron. As a filler, good results will be obtained if cast bars of the same material are used. If not available, use vanadium steel or Norway iron filler. While little work is done in welding high carbon or hard steel, the following instructions are given as a guide to the operator in case of necessity. Parts should be prepared for welding as for wrought iron or steel. Use a larger tip than for the same thickness of mild steel. For filling material where the parts are to be hardened, use ordinary drill rod. Drill rod is a hard steel which is used by tool manufacturers in the manu- facture of drills, reamers, etc. Ordinary mild steel cannot be tempered and this is often necessary when high carbon or hard steel parts have to be welded. Employ cast-iron flux. Execute the weld very rapidly as there is a tendency for the metal to burn easily and also to decarbonize; that is, to burn out the carbon, leaving the metal in poor condition. A very slight excess of acetylene in the welding flame may be advantageous. To weld high-speed steel to ordinary machine steel, the end of the high-speed steel to be welded must first be heavily coated with soft special iron. It can then be welded to ordinary machine steel without burning, but it takes an experienced Avelder to make a good weld of this kind. Special Steels. — There are many special or alloy steels used in the metal industry. The operator is often called upon to attempt welding on these. Many automobile and locomotive parts are made from special high-carbon steels, and often these castings or forgings undergo, during manufacture, special heat- treatments which are in many cases more or less of a secret process. It will be appreciated that welding with a high-tem- perature flame must necessarily counteract the effects that were produced by the heat-treatments, consequently, to make the part efficient it is essential that after welding, the piece be properly heat-treated by an operator skilled in such work. The services of such a man are rarely available, therefore, the results obtained when welding high-carbon alloy steels will be uncertain. Fortunately, however, many of the alloy steels used in practice are not high carbon and can be welded satisfactorily. 186 GAS TORCH AND THERMIT WELDING Manganese Steel (low carbon) is welded quite readily. The manganese acts as a deoxidizing agent ; that is, it counteracts the effect of burning the metal. If possible, use a tilling mate- rial of the same composition as the part welded. If this cannot be obtained use Norway iron. Nickel Steel (low carbon) can be welded without difficulty in exactly the same way as mild steel, but nickel-steel filling rod must be used. VoAiadium Steel (low carbon) — This is probably the most commonly used steel alloy. Very fortunately it is extremely easy to weld, and flows much more readily than ordinary mild steel. Weld as mild steel, but use vanadium-steel filler. Clirome Steel is in the class of mild or low-carbon steel and can be welded readily. "Weld as mild steel. Use a chrome- steel filler. Many chrome steels, however, are in the high-carbon or hard-steel class. Wrought Iron may be easily welded without a flux though a little borax or other flux is sometimes advisable. The same [general rules apply as for mild steel. Galvanized Iron cannot be wielded, since the iron is covered with, and to a greater or less extent impregnated with, a lower melting metal. German Silver, in many cases, is considered unweldable, due to its absorption of gases. For practically all commercial purposes, it may be bonded, using the same flux as for brass and a strip of German silver for the welding rod. Especial care must be given to expansion and contraction. White Metal Castings used for die molded purposes usually are composed of aluminum, tin and zinc in varying proportions, but nearly always with the lower melting metals in the larger proportion. While the castings have a good deal the same ap- pearance as aluminum, they are considerably heavier. They may be considered unweldable. Silver acts very similar to nickel and should be welded in the same way by heating and hammering. However, soldering usually answers all purposes. Gold welds very easily and the pure melting process is all that is needed. CHAPTER XII EXAMPLES OF WELDING JOBS The way a crack to be welded is Vd out or plates are beveled, has already been outlined, but it will be well to elaborate a little on the methods of doing this work. On steel or wrought iron, the beveling may be done with a gas cutting torch. On Fig. 129. — An Air Chisel May be Used Either for Grooving or Finishing. other metals, such as aluminum or cast iron, the gas cutting torch cannot be used, although the metal may be roughly melted away. Sometimes the work is of such a nature that the bevel may be ground, either with a stationary or a portable electric grinding machine. On cast iron, a sledge and a handled chisel is often the cheapest and quickest way, and in nearly every case it is superior to melting the metal away with a gas torch. A very satisfactory beveling tool for all-round shop work, is an electric or a pneumatic chisel such as shown in use in Fig. 129. This may also be used for taking off surplus metal after welding, although a portable electric grinding machine is usually preferable. 187 188 GAS TORCH AND THERMIT WELDING On work like the propeller blade shown in Fig. 130, the slots may be cut with a saw or a milling cutter and the pieces Fig. loU. — Propeller Blade Partly Beveled for Welding. Pig. 131.— Cylinder Grooved Out for Welding. left may be knocked off with a hammer. This bevel might also be chipped, ground or melted off, as the occasion or equipment at hand demanded or made advisable. EXAMPLES OF WELDING JOBS 189 An engine cylinder grooved out and ready for preheating is shown in Fig. 131. In a case of this kind the grooving may Pig. 132. — The Cylinder as Welded. Pig. 133. — Broken Automobile Cylinder. probably be best done by using a sledge and handled chisel for the easily reached parts, and a pneumatic chisel for the 190 GAS TORCH AND THERMIT WELDING rest. However, this largely depends on the size of the work and the judgment of the workman. Fig. 132 shows the cylinder welded and ready to be smoothed up. A badly broken four-cylinder block is shown in Fig. 133 and the repair in Fig. 134. The actual cost to weld this job was less than five dollars. The method of procedure has been previously described. In Fig. 135 a welder is shown working on a job while a helper is tending to the preheating of another. The method of weld- ing through a hole in a large sheet of asbestos not only keeps the heat in, Imt protects the operator as well. In Fig. 136 is shown a badly broken aluminum upper crank Fig. 134.— The Finished Weld. case and the repair. Work of this kind often comes to the shop that caters to the automobile trade. Another repair of interest to the garage man is shown in Fig. 137. The cost of putting in a new frame in this 5-ton truck would have been at least $600. The weld was finished and guaranteed for $25. The welding of a tire for a 15-ton truck wheel, 10 in. wide by 2f in. thick, is shown in Fig. 138. The preheating was done in a large blacksmith's forge. In this connection, the welder must get out of his head a very common idea that preheating is only needed to take care of expansion and contraction. It is just as valuable in its way, for saving expensive welding gas. This is the reason for preheating the large tire, since its shape and nature precludes any expansion of contraction troubles provided the welder has even ordinary skill. EXAMPLES OF WELDING JOBS 191 Fig. 135. — Welding and Preheating. Fig. 136. — Broken and Repaired Aluminnm Crank Case. 192 GAS TORCH AND THERMIT WELDING In the example shown in Fig. 139, which is a kettle 5 ft. 6 in. in diameter and 1^ in. thick, preheating is absolutely necessary in order to take care of the expansion and contraction. The crack was around the outlet and was 22 in. long. The •Ji^* Fig. 137. — Welding Frame of 5-Ton Motor Truck. Fig. 138.- Preheating and Welding Large Truck Tire. welding time was 1 hr. 45 min., in addition to the time it took to preheat the kettle the required amount. The welding of a 7-in. crankshaft for a 200-hp. internal- combustion engine is shown in Fig. 140. The finished weld is shown in the insert. The work was finished and the crankshaft put back in service inside of 30 hr. The section of the shaft added was oversize to permit machining for alignment. Pre- EXAMPLES OF WELDING JOBS 193 heating in this case saved a considerable amount of welding gas. The improvised furnace also made slow cooling possible. Fig. 139. — Preheating- and Welding a Large Kettle. Fig. 140 Welding a Large Crankshaft. Welding Broken Machine Tools. — The planing-machine bed, shown in Fig. 141, was cracked through on one side close to the housing boss. The job was finished without serious disalign- 194 GAS TORCH AND THERMIT WELDING ment, but under ordinary circumstances such a repair would not be recommended unless means were at hand for refinishing the ways and possibly other machined surfaces. As a war-emergency Fig. 141. — Weld on Large Planing-Machine Bed. Pig. 142. — Broken Punch-Press Frame. repair, however, it proved satisfactory. The redemption of a similar casting, damaged while still in the rough, might also be a money-saving proposition in some cases. The punch-press frame shown in Fig. 142, outside of its EXAMPLES OF WELDING JOBS 195 Fig. 143. — Welded Blowholes in Lathe Pan. Fig. 144.— Welded Crack in Lathe Bed.. Fig. 145. — Broken Press Frame with Breaks Beveled. 196 GAS TORCH AND THERMIT WELDING Fro. 146.— The Weldei ^f^;--::^' pt/>|ri < Afh Z2 HDjQb^ mr^ ft,- Block for Holding Blade "^a ^" U I I L (mcmtSTm) K"R:-^J \ \ctR 1 .>l^"k ->)l"k '" I j/ri S\rap(WCm£smL) 16 Threads per 'Ipch.U.S.Std Right Hand Contour of Blade after Welding |< - -- S- >l >\ VQZ50" y Rough grind t s"..^ 2L. Contour of finished Blade Clamp (COLO-ROLUOSTUL) B Fig. 172. — Jig for Welding High-speed Steel Tip to Machine-steel Body. made from two pieces of f X f-in. machine steel D and E. They are fastened with two clamps F, made of flat steel held by h- or f-in. U-bolts G. Wing nuts should be used if at all possible, as they are easily and quickly adjusted. 220 GAS TORCH AND THERMIT WELDING The reason for using the jig is that when the flame of the torch is directed on the places to be welded it heats the tips and the body of the blade very quickly and causes the steel in the blade to expand before the jig has time to get very hot. These two bodies, when heated until they run at the weld, are forced together by their expansion, resulting in a better weld. The work should be removed from the jig quickly and placed in powdered lime or bar sand, to prevent chilling and the formation of hard brittle spots in the weld, which are difficult to machine even after the regular annealing process. CHAPTER XIII WELDING JIGS AND FIXTURES Where a welding shop does a general line of work which includes everything that comes, there should be an ample assort- ment of drilled straps, angle irons, bolts, V-blocks, clamps, Fig. 173.— Table for Holding Welding Work. plates and the like. Good supplies of fireclay and plaster of paris are also very desirable for supporting or holding irregular work that is apt to collapse or get out of line. Many times, a table such as sho^vn in Fig. 173 will answer for certain jobs. 221 222 GAS TOPCH AND THERMIT WELDING Fig. 174.— Holding T'ipc lor Wi-lding, Fig. 175.— V-Blocks for Holding Shafts. WELDING JIGS AND FIXTURES 223 The top of this table is made of a "grated" slab of cast iron supported on a welded angle- and strap-iron frame. The slots provide means for the insertion of clamping bolts. A table similar to this can easily be made in any welding shop. Pipe welding is a very common and re-occurring job in most shops. Some rig up V-blocks, rollers or other devices, but the method shown in Fig. 174 is very good. It is simply a piece of angle iron placed on iron horses as illustrated. The ends of the pipe to be welded are cut square and the outside ground back for about two or three inches to remove rusty scale and dirt. On long pieces of pipe the grinding may be done with a portable electric grinding machine while the end of the pipe sticks out a foot or so from the end of the channel iron. The pipe in this case remains still and the grinding machine is moved around it as the operator stands in front of the pipe end. /Weld here Fig. 176.— Jig for Hokling Crankshafts. The short or easily handled pieces of pipe may be ground on a stationary grinding machine. The best part about using an angle iron is that the pieces of pipe to be welded are held in line while being tacked together. On ordinary sizes the welder will have no difficulty in turning the pipe as he welds. On heavy pipe some form of rollers will be found very convenient. A very simple way to weld straight shafts is shown in Fig. 175. Here the shaft simply rests on high V-blocks which keep il in line but do not interfere with expansion or contraction. A jig for holding a motor crankshaft broken in the shaft is shown in Fig. 176. The main part of the crankshaft is clamped to three V-blocks. The bases of these "V-blocks are grooved to fit over a tongue in the baseplate, so that they may be slid along in order to adjust them to various sizes of crank- shafts, and yet keep them in line. The V-block holding the 224 GAS TORCH AND THERMIT WELDING Pig. 177. — An Adjustable Ciankshaft Jig. Fig. 178. — Welding a Broken Web in the Jig. WELDING JIGS AND FIXTURES 225 short pieco to be welded on is made in the same way, but the piece should not be clamped in solidly but should be so held that it can move lengthwise. This may be done by clamping loosely or else haying the V-block free to move along the tongue The rigid clamping of all parts would cause distortion and springing of the crankshaft. The device shown in Fig. 177 is in use in the Oxweld shop. rtnllllilUlllM FiG. 179. — Aliimimim Crankcase Stiffened by Angle Iron. Fig. 180. — Angle Iron Applied to Another Job. Four V-bloeks are made to slide on bars, as illustrated, and may be clamped wherever desired. Each V-block carries its own clamping screw for holding the work. For ordinary shaft welding the table may be used in a horizontal position, as shown, but for welding breaks in webs the table may be tilted as shown in Fig. 178. This illustration also shows the use of a coal-gas and air torch to heat the work while the welder is using the welding torch. 226 GAS TORCH AND THERMIT WELDING Fig. 181. — Crankcase with Angle Iron and Bearing Mandrels in Place. ' ■ ' ^^'. 1 %J^Km^ \\ 1 V K jUmi^^ '-iiiiiiii^ff i w' jBt/^ml^B^^^^^^^BB^^^BMSM 1 i d B^ ' ""^MHirf"^ "^' - ■ 7*" ^ fc#**'* ^^^^< I^^^^^^I^BSMI^^^'^^^^S 12 g Fig. iSi;. — Motoreycie Manifold Welding Jig. WELDING JIGS AND FIXTURES 227 Crank eases or other automobile parts may be held in order to prevent distortion as much as possible, as shown in Fig. 179. In this case angle irons and short bolts with wingnuts are all that are needed. The patch to be welded in is shown tacked in place at A. ' Another application is shown in Fig. 180. The patch B has been tacked in two places ready for welding. Fig. 183.— a Welded Motorcycle Manifold. Fig. 184. — A Sheet-Metal Roller Welding Jig. In Fig. 181 both angle irons and mandrels are used in the bearings. These mandrels may be solid or of pipe to fit the bearings. Sometimes, where it is necessary to keep the bearings cool, a pipe with elbows screwed on each end may be clamped in. With the ends of the elbows up, the pipe may be filled with water. The Henderson Motorcycle Co. uses the jig shown in Fig. 228 GAS TORCH AND THERMIT WELDING Fig. 185. — A Welded Conveyor Roller. Pj(5. 186.— Larffe Sheet-Metal Cylinder Welding Jig. WELDING JIGS AND FIXTURES 229 182 to hold the parts of their exhaust manifolds while welding. The construction and operations are obvious. A welded mani- fold is shown in Fig. 183. Holding Sheet-Metal Cylinders. — A very simple welding jig is shown in Fig. 181. This consists of four castings: the Fig. 187. — Apparatus for Welding Ends in Cylinder or Tanks. base, two side pieces and the hollow mandrel. The cylinders welded are 6 in. in diameter and 8^ in. long, made of ^-in. plate. They are used to make the conveyor roller shown in Fig. 185. The seam is welded in three minutes. Another cylinder welding jig is shown in Fig. 186. This is in use in the Thermalene shop. The edges of the cylinder 230 GAS TORCH AND THERMIT WELDING WELDING JIGS AND FIXTURES 231 to be welded arc held up to the V-channel from underneath by a bar locked in place by bolts and large wingnuts. The speed for welding sheet metal will of course vary widely, but the following approximate results on sheet iron and steel are a fair average : Tliickness of Feet per metal hour 20 gage 40 18 35 16 30 14 24 12 22 10 20 . 8 18 The welding of steel barrels of about 30 to 35 gal. capacity, used for oil or gasoline, can be done by an operator of average skill at the rate of 16 to 18 per day. These barrels are made of 12-, 14-, or 18-gage sheet steel, and require one seam weld, two complete end welds, two bungs welded in and a reinforcing ring welded on each end. In welding the ends on cylinders or drums, the device shown in Fig. 187 is sometimes used. The work rests on a turn table which is rotated by the welder's foot. A supporting arm and a suspension spring assist the welder in holding the gas torch. The method of welding gas containers for war use with Oxweld apparatus is shown in Fig. 188. As shown at the right, the container bottoms are welded in while resting on rollers set on an inclined base in such a way as to present the work at the right angle. Fig. 189 gives a better view of the bung welding apparatus, and also shows the excellent method of suspending the torches when not in use. LIBERTY MOTOR WORK In describing work on the Liberty Motor in the Aynerican Machmist, May 29, 1919, H. A. Carhart, mechanical engineer of the Lincoln Motor Co., Detroit, first outlines some of the electrical welding and then says: "The water jacket is fitted to the cylinder, and the latter when assembled, is placed in a clamping fixture. Fig. 190. This fixture consists of a frame 232 GAS TORCH AND THERMIT WELDING WELDING JIGS AND FIXTURES 233 with two jaws which are placed around the jacket for holding while being tacked. The equipment used in this operation is a miniature-style Torch weld torch, equipped with a No. 3 tip and using ^/.^An. Norway welding wire. "After being tacked, the cylinder is placed in another fixture, and the bottom of the jackets are welded to the jacket flange on the cylinder. The fixture employed in this operation is a fork made of steel 1| in. wide, so shaped as to fit the top of the cylinder and thus keeping it in an inverted position. Weld- ing the bottom first was considered to be an advantage as the Fig. 190. — Tacking Jacket for Liberty Cylinder, heat applied at that point had a tendency to draw the jackets closer together. It was also considered advisable to leave an opening between the jacket halves about V^c in- to Vg, in. to take care of the contraction which the different jacket-welding operations tended to produce. ''The cylinder was next placed in a horizontal position on a cast-iron table, and the jacket-side seams were welded. The welds were started from the bottom proceeding upward to the port- holes. The cylinder was then placed in an upright position and the top seam completed. The equipment used in this 234 GAS TORCH AND THERMIT WELDING operation was a miniature Torchweld torch equipped with a No. 2 tip and ^/^^.-m. Norway welding wire. "The next operation was to wekl the jacket around the two spark plugs, valve guide and camshaft housing bosses. The fixture employed in this operation carries a pilot upon which the cylinder can be turned circularly. The equipment used is the sam6 as before except that a No. 3 tip was found best. Ftg. ]91.— Welding the Inlet Pipe. "Next came the welding of the jacket around the porthole flange. The fixture used in this operation consists of a standard with a revolving cradle to lay the cylinder in. The revolving motion of the cradle, together with the varying height of the standards, gives, the operator easy access to weld. The equip- ment used is a No. 1 Torchweld torch equipped with a No. 3 tip and Vi^-in. Norway welding wire. WELDING JIGS AND FIXTURES 235 "The next operation consists of welding both water pipes to the jacket and the jacket seams between the valve guide and porthole flange. This is shown in Figs. 191 and 192, both inlet and outlet being handled in the same fixture. This fixture is equipped with two devices for holding the water pipes in their proper position while being welded and is so arranged that it can be oscillated to suit the position in which the operator Fig. 192.— Welding the Outlet Pipe. is welding. The cylinder is located by a flat clamp which locates the flat on the cylinder bolt flange in proper relation to the pipe-holding devices. "The jacket seams are welded first, followed by the water outlet pipe. In these two cases, the cylinder is welded in an upright position. The locating clamp on the bolt flange flat is then released and the cylinder given a half-turn to the op- 236 GAS TORCH AND THERMIT WELDING posite flat. The inlet pipe is then placed in its holding device and welded with the fixture in a semi-horizontal position to suit the welder. ' ' The cylinder is then removed and the water pipes inspected for proper location. If it is passed, the cylinder is hammered with a rawhide hammer to remove scale and loosen any poor weld, after which it is tested for leaks. If leaks are found, they are repaired hy expert welders and returned for re-inspection." WELDING FIXTURES FOR MAKING MANIFOLDS Writing in the American Machinist for Mai'ch 25, 1920, C. C. Phelps says: "Several ingenious fixtures are employed w great Fig. 193. — Fixtures Used in Welding Liberty Motor Manifolds. advantage in manufacturing manifolds for the Liberty engines by means of the oxy-acetylene process at the plant of the Ireland & Matthews Manufacturing Co., Detroit, Midi. The fixtures were designed in accordance with plans furnished by the en- gineering department of the Oxweld Acetylene Co., Newark, N. J. "In assembling the manifold parts in the fixture. Fig. 193, the five branch inlets are first mounted on their respective pivots A in the bed of the fixture ; tlie trunk of the manifold is then WELDING JIGS AND FIXTURES 237 placed above and in contact with the branch lines, so that the openings in the trunk coincide with the ends of the branches, and, finally, the five hinged clips B are swung into position and clamped down on the assembled manifold by means of the hand clamps. The end of the trunk is bent to serve as one of the inlets and this end in turn is inserted over the end pivots. The fixture shown in Fig. 194 serves to hold the assembled parts in perfect alignment. "The fixture proper is suspended at the ends to permit com- % ra m bMDXSxmXCM Fig. 194. — Details of Fixture Proper. Assembled Manifold Indicated by Dotted Lines. Fig. 195. — Details of Swing Support for Manifold Welding Fixture. plete freedom of rotation, and the points of suspension arc so -located that the device will be in balance when containing the tubing. The support for the fixture. Fig. 195 is mounted on a pedestal in such manner as to allow rotation in a horizontal plane. Thus the operator is enabled to shift the work so that the torch flame can be applied in the most advantageous manner at all times. Fig. 196 shows the complete manifold. "During the war this company manufactured various kinds of tubes and manifolds for Liberty and Le Rhone engines, 238 GAS TORCH AND THERMIT WELDING bombs, gas shells, floats for the Navy and poison-gas tanks. When it is considered that the company had no welding equip- FlG. 196. — Coniplutcd Mauifokls for Liberty Eugiues. mcnt prior to May, 1918, great credit must be given to the inex- perienced girl operators and the equipment that produced such results. ' ' CHAPrER XIV WELDING MACHINES Gas-torch welding machines with automatic feed are used for a large variety of work, although straight-seam welding is the more common. In this latter class of work are included sheet-metal-cylinder side-seam welding and pipe or tube welding. A welding machine, known as the Duograph, is shown in Fig. 197. This machine was made by the Davis-Bournonville Co. and was especially designed for welding the seams of steel drums or containers, insuring a mechanical weld uniform in appearance and efficiency. It comprises a turret-top holding device with water-cooled arms and clamps for holding the steel drums in position, permitting the work being placed in position for welding on one set of arms while the work on the opposite set of arms is being welded. The turret top is then swung half around, the welded work removed and another job set up. The gas-torch carriage is moved forward at a fixed speed by power, belt driven, and is reversed by means of a hand- wheel when the weld is finished. Various speeds for different thicknesses of metal are obtained by the use of cone pulleys. The carriage is fitted with two torches — one above, the other below — as shown in Fig. 198, for welding both sides of the seam simultaneously. For very light welding, one torch only is re- quired. Water-cooled welding torches are used. The No. 1 machine will weld a 36-in. seam, and will take containers from 12-in. to 36-in. in diameter. The No. 2 machine welds a 54-in. seam. An average speed of welding of 18-in. per minute is obtained on 16-gage sheets. Fig. 199 is a close-up of a man putting a sheet-metal drum into position on one of the turret arms. Fig. 200 shows the drum clamped down and swung into place ready to be welded. This illustration gives a good idea of the operating mechanism. 239 240 GAS TORCH AND THERMIT WELDING WELDING MACHINES 241 Fig. 198. — Torch Arrangement on the Duograph. FiQ. 199. — Putting a Drum Onto a Turret Arm. 242 GAS TORCH AND THERMIT WELDING Fig. 200. — Drum in Position Eeady for Welding. Fig. 201.— The Finished Seam Weld. WELDING MACHINES 243 Fig. 201 shows the seam weki completed and ready to })e removed. A much simpler machine is shown in Fig. 202. The opera- tion of the feeding mechanism is obvious. A smaller, though very similar machine, is shown in Fig. Fig. 202. — Heavy Drum Welding Machine. Fig. 203. — Lioht Seam Welding Machine. 203. While the work shown in position is cone shaped, cylinders may be held as well. The machine shown in Fig. 204 is for welding bottoms onto tea kettles, cans, drums or other circular work. The machine is so made as to allow for a considerable range of adjustment for different sizes of work. 244 GAS TORCH AND THERMIT WELDING The last three machines mentioned were designed by Linus Wolf, of the Thermalene Co., Chicago Pleights, 111. A machine developed at the plant of the Edison Storage Battery Co., Orange, N. J., for welding bottoms in storage- battery cases, is shown in Fig. 205. This machine was first de- sci'ibed in the American Machinist, Aug. 10, 1911. The bottom Fig. 204. — Machine for Weldino- Circular Seams. to lie welded in is made of sheet steel with njiturned edges. A four-part expanding form is placed within the edges of the bottom and locked by turning down the screw shown in the center of the case. With the bottom and expanding form in place as shown, the ease is "shrunk" to it and sized by turning the eccentric lever A. The gas torch B, which is hinged at C, is then swung down into welding position and so set as to throw the flame correctly onto the upturned edges of the bottom WELDING MACHINES 245 and the case. The motor is then started and the feed thrown in by means of lever D. This lever operates a clutch on a shaft carrying a pinion meshing with the oblong gear on the bottom of the frame which Fig. 205. — Machine for Welding Oblong Seams. supports the case. This moves the frame and the seam to be welded along under the welding flame. When a corner is reached the trip E throws the lever F and slips the clutch G into contact with the upper teeth, increasing the speed of the driv- 246 GAS TORCH AND THERMIT WELDING iiig pinion so that the seam being welded moves at the same speed under the flame while turning the corner as while being driven along the straight seam. As soon as the corner has been turned the lever F is forced down by another trip and the sun-and- planet gearing / again comes into play, giving a slower move- ment as the straight part of the seam is fed under the flame. Fig. 206. — Single-Torch Tube AVeldina; Machine. For liringing the mechanism back to the starting point," the handle J is used. A single-jet tube welding machine made by the Thermalene Co., is shown in Fig. 206. Views of a double torch machine made by the same concern, are shown in Figs. 207 and 208. In a general way, these are typical of all machines designed to butt-weld formed tubes. WELDING MACHINES 247 TUBE WELDING BY THE OXY-ACETYLENE PROCESS Writing in the American MacMnist, Nov. 13, 1919, F. M. Smith, Chief Engineer of the Oxweld Acetylene Co., Newark, N. J., says: "Tubing, considered merely as a structural shape, Fig. 207.— Duplex Tube Welding Machine. has the greatest strength in compression and bending to be ob- tained from any shape of equal cross-sectional area. Likewise, it is one of the most convenient shapes known. In large work this fact has been applied for years in the use of round cast- iron columns and standard commercial wrought-iron pipe of vari- ous weiglits, but until the advent of welded tubing nothing but 248 GAS TORCH AND THERMIT WELDING drawn seamless tubing was available for weights lighter than pipe and its price has been prohibitive for ordinary purposes. "When the oxy-acetylene welding process offered a means of producing the substitute for drawn seamless tubing in the form of welded tubing of thin gage, numerous manufacturers were so well impressed with the possibilities of this line of work Fig. 208. — Another View of Duplex Machine. that they went into it on a large scale. As a result, the product of the oxy-acetylene tube-welding process is of such excellent quality and is produced so much cheaper than drawn seamless tubing that it is rapidly superseding the latter form of tubing for very many purposes. "Tubes welded l\y the oxy-acetylene process are almost in- WELDING MACHINES 249 variably butt-welded; that is, in the tubes as formed up, the square edges lie butt to butt. The heat is applied to the seam only and must be of an intensity to raise the edges immediately under the flame to the fusing point. The edges are then pressed together while in a molten condition and flow into each other, forming a true and homogeneous weld. The weld may be upset or reinforced, if so desired, to a greater thickness than the original tubing wall by compression of the seam at the point of weld. "As the uses of tubing are principally structural, the material from which it is made is usually low-carbon steel, although this is not used as universally as might be expected because old high-carbon rails make an excellent and comparatively cheap raw material for the manufacture of tubing by the hot formed process, where a smooth and polished surface is not required. "The low-carbon steel tubing may be made from either hot- or cold-rolled stock. By far, the largest amount is made from cold-rolled sheet on account of its smooth finish and good ap- pearance. Other metals can be machine-welded satisfactorily, but where they are ductile as in the case of brass and copper, it may be cheaper to make them by seamless drawing. * ' Low-carbon steel us usually purchased from the mills in the shape of strips or sheets which can be rolled and sheared ac- curately to the required dimensions. Gang shears reduce the sheet to strips of accurate width in one operation. The tubing is formed cold by first rolling it to semi-circular form and closing it in to the circular form either in a conical drawing die or by further reductions in size between rolls, the latter being generally the preferred process because it eliminates subsequent warping at the seam. "Rails from which high-carbon tubing is usually worked up are broken into the correct lengths to produce a 50-ft. skelp, heated in an ordinary heating furnace and broken down to rectangular section of the desired gage and width by a series of rolls of the usual rolling-mill type. The skelp is then passed, while still hot, through a pair of rolls which form the section to a half circle and force it through a conical die which draws in the edges to the final complete circular section required for tubing. "Owing to the inaccuracies of this type of mill, the gage and 250 GAS TORCH AND THERMIT WELDING width of the skelp, which latter dimension determines the diameter of the tubing, cannot be held to any close accuracy, and, consequently, this type of tubing is used only for purposes where exact size is not important, such as bedstead frames, handles for tools, and the like. "Machines for tube welding consist, in general, of a table with frame, or housings, upon which is mounted the feeding Fig. 209. — Feeding End of Tube-Welding Machine With Oxweld Multiple-Jet Gas Torch. mechanism and its drive ; above this, a gas-torch holder equipped with a horizontal adjustment at right angles to the seam, another in the direction of the axis of the torch and a hinged arrangement by which the torch may be lifted away from the work. "There are two general types of feeding mechanism, one of which, covered by the Lloyd patents, consists of a continuous or endless-chain arrangement. Two chains are required, one WELDING MACHINES 251 on each side of the tube, each link carrying a block with a groove in its face to fit the tube. Two corresponding blocks, when moved forward between rolls upon which the endless chains are mounted, catch the tube between jaws and carry it Fig. 210.— The Left Set of Eolls Feed the Tubing while the Spreader Disk Opens the Seam Slightly. The central rolls hold the tubing to specified diameter while welding and the right set of rolls act solely as guides. forward at a uniform rate of speed, at the same time com- pressing the seam the desired amount to secure a satisfactory weld. "How Rolls Are Arranged. — The other, and more generally used method, shown in Figs. 209, 210 and 211, employs two trains of rolls, three on each side, having grooved surfaces to 252 GAS TORCH AND THERMIT WELDING fit the tubing. The middle pair, known as the welding rolls, are chamfered away on the top side to allow the welding flame to work in between them. The forward and rear pairs are used merely as feed and guide rolls. In order that different sizes may be welded on the same machine, rolls having grooves of varying depth are provided to fit the various diameters of tubing to be welded. Provision is also made for opening and closing the rolls to get heavier or lighter welds. "Experience has shown that to secure an even and smooth seam, it is necessary to counteract the effect 'of expansion and RE/iR ROLLS WELD/NO ROLLS FORW/\RD ROLLS Elevoiti on Sec+ion A- A Fig. 211. — Diagram of Arrangement for Tube Welding. contraction by spreading the seam open Vg. in., more or less, about 6 in. ahead of the welding rolls. This is generally accom- plished by mounting a thin disk abreast of the forward guide rolls at such a distance that its lower edge will enter and spread the seam apart the necessary amount to secure this smoothness of weld. This disk also answers the purpose of aligning the seam under the flame. "The drive should be arranged to give different speeds in order that the machine may be used on different types of weld- ing, but should be of such pattern that when once the speed is set, it will remain closely uniform. WELDING MACHINES 253 ''There are two general types of oxy-acetylene torches in use, corresponding to the two schools of oxy-acetylene practice ; namely, the low-pressure and pressure types. The low-pressure type is very successful in welding operations owing to the 'softness' of the flame due to the low velocity of the heating gases. It also implies efficiency, as the gases passing over the metal at a low velocity have more time to give up their heat. The uniformity of the low-pressure flame is because the acety- lene is supplied from a gas bell at an even pressure. Fig. 212.— Four-Jet Type W-5 Water-Cooled Welding Torch. Fig. 213.— Oxweld Type W-8 Single- Jet Water-Cooled Welding Torch. "In the welding of heavier gages, more heat is required than for lighter work and this is secured by arranging a variable number of flames in a line progressively along the seam to be welded (see Fig. 212). The forward flames successively raise the temperature of the tube to such a point that the last flame will fuse the metal just abreast of the welding rolls which compress the fused edges together, forming a perfect weld. For welding tubing of 14 gage, as high as nine or ten flames have been successfully employed and with this number of flames a welding speed of 5 ft. per minute can be obtained. With such a concentration of heat, water-cooling is necessary to maintain 254 GAS TORCH AND THERMIT WELDING a uniform flame and oliminato backfiring or ignition of the gases within the tip. The torch proper is always water-cooled to secure uniformity of flame and for the comfort of the operator. For the heavier work, mentioned above, the possibilities range down to 22- or even 24-gage metal, upon which a single flame torch, Fig. 213, may be employed. Short samples of welded tubing are shown in Fig. 214. "After welding, if it is desired to secure a very exact diameter, either externally or internally, this is accomplished by drawing. The operation is performed on the draw bench which consists of a long frame, usually horizontal, at one end of which is a Fig. 214. — Samples of Tubing Welded by the Oxy- Acetylene Machine Welding Process. A — Cartridge case. D — Tubing as welded. C and D — All traces of welding eliminated by grinding and polishing. E and F — Wind-shield frame and wedge. die of chilled cast iron or hardened steel. This die must be of the correct diameter and smoothly polished to produce the correct finish. It may be cither externally or internally applied. When used on the inside of the tubing, it is known as a 'triblet.' The tubing is entered into the die or over the triblet ; the end is crushed to secure a grip for the drawing mechanism, which is located at the opposite end of the bench, and the tube is then drawn through the die producing the correct size and a very smooth finish. Suitable lubricants must be applied to the surfaces of the tubes during this process. "Where only a bright finish, without exact diameter, is re- quired, drawing is not employed — ordinary flashing or polish- ing methods being sufficient. If the tubing is to be painted, or WELDING MACHINES 255 for some other reason high polish is not necessary, it is often satisfactory, if a smooth weld is secured, to use the tubing just as it comes from the welding machine. ' ' The principal difficulties encountered in tube welding are to secure a uniform, continuous and neutral flame and to feed the tubing under the torch at a correct and uniform rate of speed at the most effective distance from the tip with the seam exactly in the flame. "The first difficulty is largely solved by using the proper torch. The second difficulty is mechanical and can only be eliminated by standardizing shop conditions. "The stock strip must be held to exact width if a uniform diameter and thickness of weld are to be produced. The tolerances are determined by experience, but once set should be strictly held to. Burrs on the sheared edges must absolutely be eliminated as these get into the seam and hold it open causing 'skips.' When the strip is formed into tubing, care must be taken to secure a very exact adjustment of the rolls and dies or otherwise the seam in the tube will assume a spiral shape, causing difificulties in guiding the seam exactly under the flame. If the flame does not play exactly upon the seam, only one side of the seam will be fused and the weld will 'skip' until the flame is correctly adjusted. Lost motion in the feed rolls and in the adjusting mechanisms of the torch holder will invari- ably aggravate this difficulty. "In the early days of this art, when manufacturers were content with speeds of 2 to 3 ft. per minute, it was possible for the operator to adjust his flame from side to side, follow the irregularities in the seam, and to slow down his machine when the Aveld showed a tendency to skip. With the speeds attained in modern production, from two to four times as fast as formerly, this is no longer possible. The operation progresses so rapidly that it is humanly impossible for the operator to adjust his machine to changing conditions and, in consequence, if the seam is not perfectly true the operator must necessarily fail to get a continuous weld and even in attempting to do so he will constantly slow down his machine, thus limiting the quantities of tubing produced. "The successful manufacturers have, therefore, applied the well-known principles of scientific management to this problem 256 GAS TORCH AND THERMIT WELDING and by studying the conditions under which the forming and welding are done have succeeded in so standardizing their product that the welder is not required to do anything but start the tube into the machine and keep the torch lit and the tip free from accumulations of slag and dirt. The tubing is re- quired to come absolutely uniform in diameter and straight in seam. The set of the rollers, speed of welding and pressures of the gases in the torch are adjusted by the foreman or tool setter in charge to predetermined standards which are not allowed to be changed by the operator. If, with the machine as delivered to the operator, he cannot secure good tubing, the machine is shut down, the trouble discovered and the proper remedy applied. "By the application of such methods as these and with the full cooperation of the manufacturer of oxy-acetylene apparatus, it is not remarkable that great progress in the development of the art of oxy-acetylene tube welding should have been made within so short a time." CHAPTER XV CUTTING WITH THE GAS TORCH The gas-torch cutting process consists of heating a spot of the metal to be cut to a good red heat and projecting on it a jet of oxygen. This causes the metal to burn away, a stream 01 slag running out of the kerf thus produced. Cutting is not melting, in the ordinary sense, although since the heating flame is the only visible agent, such might be the beginner's conclu- sion. It should be remembered that the heating flame is only used to make the metal hot enough to oxidize easily. Metals whose oxides have a lower melting point than the metal itself can be cut by the gas torch. Such metals are wrought iron and steel. Where the oxide has a higher melting point than the metal, cutting with the gas torch is either im- possible or not satisfactory. Such metals are copper, brass, aluminum, cast iron, etc. A big factor in successful cutting is to properly support the body and torch to as great an extent as possible commensurate with the steady forward movement of the torch. The position must be an easy one, as muscles under tension will cause vibra- tions and these are fatal to good cutting. An ideal position for an operator, is shown in Fig. 215, although in actual, every-day practice one usually has to be satisfied with less desirable con- ditions. Theoretically, with the cut once started the oxygen jet alone should be sufficient to keep up the combustion, as there is con- siderable heat generated in the process. However, the stream of oxygen is small and the burning metal confined to a very narrow slot, and scale, dirt, sand, blowholes and other things interfere to prevent the continuation of the cut of the jet with- out an accompanying heating flame. A cutting torch is lighted in the same way as for welding, except allowance must be made for the drop in the oxygen 257 258 GAS TORCH AND THERMIT WELDING Fig. 215. — Au Eii^y Cuttiiii; Positiori. Fig. 216. — Starting a Ciit With a Davis-Bournonvine Torch. CUTTING WITH THE GAS TORCH 259 pressure when the cutting jet is turned on. This allowance can be made by regulating the flame while the jet valve is open, which is done before starting to work. When the flame is adjusted, hold the torch as shown in Fig. 216, the left hand grasping it well toward the head and the right hand on the handle with the thumb- or fingers controlling the jet level valve. The metal to be cut may be a piece of Fig. 217.— Making a Clean Cut Thronah a Plate. heavy boiler plate, steel bar or structural steel. Rest the elbow, forearm or hand on the plate to steady the torch. It is usually best when cutting without a guide wheel, to arrange to cut either to the right or to the left rather than toward or away from the operator. However, an operator should learn to cut in any direction. "When it is possible, always start on the edge. Hold the flame on one spot until it is a nice red, then turn on the high-pressure oxygen jet. Hold the torch steady 260 GAS TORCH AND THERMIT WELDING with the luminous cone almost touching the metal, until the cut goes through. Sparks should show as in Fig. 217. If they fly, as in Fig. 218, the cut is not going through. In the cutting of plates, it is advisable to tip the torch head in the direction of the movement, once the cut has progressed a little. This rule does not apply in the case of blowing holes in metal where the nozzle must be tipped away from the slag Fig. 218. — Cut Not Going Through Properly. SO that no particles will impinge on the orifices or a back pressure be created on these orifices. If the metal is very thick, the oxygen pressure will have to be high. In beginning a cut of this type, it is necessary to blow the oxide out at the bottom before the cut has traversed very far into the body of the metal, otherwise, a pocket will be formed and it will be impossible to penetrate to the bottom of the metal. In cutting heavy material, success depends entirely upon the ability of the individual. The nozzle must be turned outward in preheating and must be carried inward with the CUTTING WITH THE GAS TORCH 261 tip gradually moving to a vertical position and finally forward as the cnt progresses. In blowing holes, as in Fig. 219, the metal must be blown away from the tip, and to accojnplish this Fig. 219. — Blowing a Hole Through a Plate. Fig. 220.— Cutting OflE a Eivet Head. it is advisable to begin with a very Avide kerf, produced by rapid movement of the torch sideways while carried away from the origin of the cut. In this way the oxygen penetrates deeper into the metal while the torch is moving, until, finally, the 262 GAS TORCH AND THERMIT WELDING oxygen emerges at the bottom, when the torch can be brought to a final cutting position and the metal cut in any direction. Rivet head cutting in shipyards is generally accomplished by means of a specially designed nozzle, which rests upon the plate so that the preheating jets and cutting jet will act at the base of the rivet head as shown in Fig. 220. In blowing out countersunk rivet heads, the same procedure must be followed as in blowing holes, but more precautions are necessary in order that particles of metal do not impinge on the preheating orifices and clog them or cause backfire. The "nicking of billets" became very common during the ti^iG. 221. — [Isinsr I^ollcrs and a Bar Guiilo. war. A narrow, shallow cnt is made on one side or around the circumference of a steel section, then the billet is snapped oft' ;it the nick in a press or hammer. When a cut must he ''easonably smooth, use wheel guides, if possi])le. If a straight line must be followed, a bar of metal may l)e clamped to the work as shown in Fig. 221. A good way to both guide the cut and support the operator's hand, when cutting ship plates is shown in Fig. 222. This principle may be applied to other wtu'k For cutting circles, a radius attaclimeiil is used, similar to the one shown in Fig. 22:) Tiiis device is made by the (*arho- liydrogen Co., Pittsburgh, Pa., but practically every torch manu- facturer makes something of the kind. CUTTING WITH THE GAS TORCH 263 The way llic cul on a 12-i]i. shaft kwks is shown in Fig. 224. This was cut with an Oxweld low-pressure torch. Tlie chalked Fig, 222. — Cutting Ship Plates. 'i 4 c ? i J ;| '■ Fig. 223. — Kadiiis Cutting Attachment for Straight-Tip Torch. arrows indicate a blowhole and a crack which materially re- tarded the cutting. This cut took 3 min. 27 sec. and about 75 cu.ft. of oxygen was used. A similar cut, under similar 264 GAS TORCH AND THERMIT WELDING conditions, bnt made without encountering any flaws in the steel, was made in )> iiiin. 10 sec. and 67 en. ft. of oxygen was used. On woi'k 1 in. thick or over, a slot of from Vi,; to ^ in. is about right. For thinner stock, or when using a machine, the slot may often be reduced to less than Vj,., in. by a skilled operator with special tips. Flame Control. — In woi-ldng liold the flame so that the end of the cone just clears the metal — do not attempt to plunge it Fig. 224.— a 12-Tii. Shaft Cut with a Gas Toix-h. down into the cut. When cutting two plates or more, or where there is a lap joint, remember that there is more or less of an insulation (air, dirt, etc.) between these plates and that the oxidation cannot be as fast as where only one thickness is cut. Remember that the flame does not do the cutting — therefore, work with the smallest flame possible — it means a neater cut. Keep the oxygen pressure as low as possible and yet maintain speed. A high pressure is spectacular and there are a great number of sparks, but it is not economical and a wider kerf is made. Do not use the torch with greasy gloves — a spark CUTTING WITH THE GAS TORCH 265 in combination with a leak on the oxygen supply will badly l)urn the hand. If a cut must be started in any place except on the edge, drill a hole or use a cold chisel and a hammer to roughen up the surface, the idea being to get an edge to quickly stai't oxidation. Making- a Ladle Hook. — As an instance of the many savings lliat may l)e obtained by the intelligent use of the gas-torch cutting process, the following will be of interest : At one of the shipyards scrap ship plates are cut into special shapes for building up large hooks like the one shown in Fig. 225. These hooks are used for handling large ladles in a near-by steel mill and have resulted in a great saving. H ■ HH ■ s^' HHi ■ m ^ 1 i^ g I » '*jj _^. H Q^^^ ^^v^^ 1 ^^^^i^^i Fig. 225.— Ladle Hook Made of Toreh-Cut Plates. The hooks are 8 ft. in total length and are made up of six layers of plates which run the full length, with four short layers, all securely held together with countersunk rivets. The four inner plates are each ^ in. in thickness. The two outer full-length plates are of ^-in. material. Adjoining the latter plates on either side is a half-length plate ^ in. in thickness. The hook proper is still further reinforced by two slightly shorter outer plates, each | in. in thickness. The plates are first marked with the aid of a templet to serve as a guide for the cutting torch. After cutting, they are assembled and riveted as shown in the illustration. A laminated construction of this sort is not only exceptionally strong, but 266 GAS TORCH AND THERMIT WPJLDING is a decided economy, as it makes use of what would otherwise be waste material. Firemen are frequently confronted with locked steel doors or barred windows. These readily yield to a properly applied cutting torch. Fig. 226 shows a fireman demonstrating how an Oxweld emergency cutting outfit may be used. The entire kit weighs 118 lb. A very wide field for the cutting torch is in reducing scrap to workable dimensions. The figures here given regarding the cutting of scrap, are taken from a bulletin issued by the Oxweld Fig. 226. — Fireman Demonstrating an Emergeney Kit. Co. Where costs are cpioled the estimates should bL' about doubled for present conditions (1920). An operator recently cut two twenty-ton steel fire boxes into sci'ap, prepared for the shears in twelve hours. More than 300 lin.-ft. of cut was made th]-ough |-in. plate (considering the mud ring and over-lapping plates). The total cost for oxygen, acetylene and labor was $24.10 per fire box — cost per ton $1.22. In another case a locomotive boiler was cut into scrap at ii total cost of $2.68, the numl)er of lineal inches cut totaled 210 through ^-in. plate, 9 through 3-in. plate and 172 through f-in. plate. This amount of cutting was completed in fifty- three minutes at a cost of 8 cents per foot for tlie various thicknesses. The foreman in charge of this job stated that the work done by one operator in one and one-half days would CUTTING WITH THE GAS TORCH 267 require the services of two men for, at least, a week, with ordinary working methods. A ten-ton boiler was reduced to scrap read}' for shears by one operator in nine hours at a total cost of $16.00 or $1.60 per ton. On another piece of work, the operator cut 78 ft. of 4-in. plate in two and one-half hours. One piece of this plate, 18 ft. long, was cut in 13 min. The cost of cutting the 78 ft. was $4.25 or $0,054 per foot through the. ^-in. plate at a rate of over 30 ft. per hour. A three-ton boiler averaged $2 per ton cut in one and one- half hours. The total length of cut equaled 60 ft. 4 in. It would have cost $3 to $4 per ton to cut this boiler by hand. A fourteen-ton boiler was cut at the rate of $1.23 per ton in nine hours and at a total cost of $17.38. One hundred and eleven feet eight inches of cut was made at the rate of $0,149 per foot. Three fire-box boilers weighing ten, twelve and fourteen tons respectively were scrapped at the average rate of $1.40 per ton. The users of this plant state that the apparatus enables them to cut into scrap five locomotives Avhere one was handled by the methods used before the Oxweld process M^as employed. Cutting steel car frames into scrap shows equally important savings in time and money. A five-ton car frame was cut in two and one-half hours at an average cost of $2 per ton. It was cut into 4^-ft. lengths through three and four thicknesses of plate in some parts of the frame. A record kept of cutting about 12,000 lb. of wrecked steel car frames shows a total cost of $8.10, or about $1.35 per ton. These frames were cut into 4|- to 9-ft. lengths, in five hours. CUTTING CAST IRON WITH THE GAS TORCH In a paper read before the American Welding Society, April 22, 1920, Stuart Plumley and F. J. Napolitan, of the Davis-Bournonville Co., outlined some of their experiments in relation to the cutting of cast iron with a gas torch. They said in part : While we are rather skeptical of the coiiimercial value of a cast- irou cuttiiiix torch, and are convinced that, financially, we shall never 268 GAS TORCH AND THERMIT WELDING be repaid for the expense of our experiments, yet there are un- doubtedly occasions when tlie cutting of cast iron wouhl be of great value. In oi'dinary scrap-yard work, it is so easy to lireak cast iron that it would hardly be economic to use the cutting torch as for steel. You are all aware, of course, of that application of oxygen cutting used largely in blast furnace practice, the opening of a "frozen" tap hole. You could not (luite reconcile this more or less conmion applica- tion of the process with the pet theory that cast iron could not be cut. One of the usual methods for releasing a frozen tap hole in a blast furnace is substantially as follows : A piece of |-in. iron pipe with a brass handle at least 10 ft. long is attached to a manifold of several oxygen cylinders. Oxygen is delivered through this pipe at a pressure of approximately 100 lb. per sq. in. A hole is started with a star drill or diamond point, until it is about 3 in. deep. The metal adjacent the hole is heated with a fuel oil burner or by other means. The end of the iron pipe is ignited and the composite stream of molten iron slag and oxygen caused to impinge against the frozen cast iron. A spectator to this performance of infernal fury, is readily con- vinced that the heat is not all due to the combustion of the wrought iron pipe, but that the cast iron is Ijurning with a violence equal to that of steel. This reaction inspired some inventors to incorporate a device in an oxy-acetylene torch for cutting cast iron, which would feed a steel wire between the cutting jet and the cast-iron piece being cut. Ignition of the wire carried a stream of molten slag on to the cast iron and it was hoped thus to propagate the cut. In a second process, a plate of steel of a definite and predetermined thickness, was placed on top of the cast iron. It was hoped that the slag incidental to the oxidation of the steel would exercise some influence over the cast iron and enable it to be cut. Unfortunately for those responsible for the exploitation of these devices, the inventors were more concerned with converting cast iron into iron oxide by means of the oxy-acetylene torch than they were in constructing a practical process and a practical tool. It was next proposed to simplify the reaction by supplying an apparatus with a mixture of pulverized slag and iron powder, and in fact a number of patents were issued covering various applications of such a device. Crude and elementary as such devices were, they actually produced combustion of the cast iron and went a long way in stimulating us in our endeavors to find a successful method. Experimental work was carried on with a torch having a good many different tubes leading to the head so that almost any com- bination of gases at varied pressures might be obtained. Mr. Napolitan evolved from these experiments interesting theories i)ertaining to the reactions which take place in cutting, together with their relation to success in cutting cast iron. He has noted these theories in a separate paper. We are presenting these thories to the members of the society CUTTING WITH THE GAS TORCH 269 for wliiil tlu'.v arc wortli. We ciiii jictually cut cn.st iron and we do it l).v ]»•< h< (liiiui fli( oxiKjcn. Ill the i)apei' prepared by Mr. Napolitan, he said: From tlie ease with whicli \\rouglit iron is cut we may conclude that an aggregate of ferrite comhines with oxygen witli greatest r.vidity, and permits the propagation of a cut with least interruption. As the carbon content is increased, there is a material change in the nature of the metal. In place of the preponderance of ferrite grains, we recognize the formation of cementite, and its union with some of the ferrite to form pearlite — tlie original mass of pro-eutectoid ferrite rapidly diminishing in prominence. As we should anticipate from the nature of pearlite, no material change is noticed in the performance of these alloys under the cutting torch. Of course, an ultra-pi-ecise consumption test would probably indicate a lowering of the efficiency coefficient, but from all appearances no unusual difficulty is experi- enced in cutting carbon steels up to about 80 to 90 point carbon. But here, a definite transition is indicated by a distinct laboring of the cutting torch. While the torch will begin to cut with practically the same effort, and proceeds to completion without interruption of un- usual delay, yet the kerf is wide and ragged and undeniably dis- tinguishable from that of a mild steel cut. It is recognized practice, now, to preheat the piece to be cut to a black or dull red heat, when the impediment, whatever it was, seems to have been entirely eliminated. But let metallography explain the sudden change of properties of the steel. As the carbon content of the hyper-eutectic steel was in- creased, the proximate mass of pearlite increased, and the pro-eutectoid ferrite correspondingly diminished in volume, until eventually a point was reached where all of the cementite and ferrite existed in the stratified or laminated relationship of pearlite. This state is recog- nized as existing where the carbon content is between 80 and 90 points — the approximate analysis of pearlite is yet undefined. As the carbon content is further increased, there appears a constituent that we know as pro-eutectoid cementite — in fancy, the cementite which has been ejected from the pearlite growth. It is circumstantial that the presence of this pro-eutectoid cementite is directly responsible for the increasing difficulty of our cutting. But why did preheating of the steel before cutting make such a remarkable difference in the results? To be sure, the rise in temperature might affect the stability of any martensite, troostite, or even sorbite that might have existed, but the temperature was too far removed from the ACa^i point to affect the characteristics of the pearlite. And surely the pro-eutectoid cementite was unchanged — and it was this same constituent that we blamed for the difficulty. Again, as the carbon content is substantially increased, an equivalent interference with cutting is api)arent. until, when the carbon content approaches 2.5 per cent, cutting becomes so labored as practically to cease, and no amount of preheating short of incipient fusion will 270 GAS TORCH AND THERMIT WELDING permit it to propagate. As you are aware, the metal is now termed "cast iron," and a micro-analysis indicates that in addition to the presence of a certain amount of pearlite and pro-eutectoid cementite, as well as certain foreign and, to our discussion, unol)trusive sub- stances, we i-ecognize the presence of the final and most stable state of carbon-graphite. The pearlite constituent exercises a favorable in- lluence upon the operation of cutting — ^and the pro-eutectoid cementite, while it impedes cutting, is readily compensated by a slight preheating — but the graphite presents an entirely new problem. We might digress from the subject enough to present some remarks that would prove the fallacy of at least one of the stereotyped ex- planations of why cast iron cannot be cut — that the melting point of the slag is appreciably higher than the melting point of cast iron. A micro-analysis of the structure of an average cast iron — and by average we refer to a gray cast iron of about three to four per cent carbon — would indicate a structure identical with ih:it of a hypothetical steel of the same carl)on content, except that some of the carbon seems to have been precipitated as graphite. But should that identical pour of cast iron have been cast against a cold iron mold, or otherwise chilled, the carbon would not have been pi'ecipitated as graphite and we should have had what we shall call a "chilled cast iron," or a "white cast iron," — and it would actually have been a hyper-eutectic steel. Sucli alloys are not uncommon in com- merce, and the fact tliat operators have lieen able to cut them with no extraordinary effort has lieen responsil)le for in- numerable false claims that cast iron has been cut. Unfortunately, the nomenclature of steels and irons is not clearly defined, and un- doubtedly a chilled cast iron is but an extension of the hyper-eutectic series. The melting point of an iron-carbon alloy is a constant of its composition, whether, in the solid state, the metal exists as a typical cast iron or as a steel. Long before the point of fusion, the carbon and the iron exist in one relationship, that of austenite. The condi- tions affecting the pouring of a melt of cast iron would determine the final state of its constituents — and we nught.as readily produce a gray cast iron or a chilled white cast iron — the carbon as graphite or the carbon as in cementite. In either event, the melting points of the resulting products would be identical. We agree that chilled cast iron can be cut with comparative ease. It is evident, then, that the melting point of slag is not resi)onsil)le for the difficulty encountei'ed in cutting cast iron. We had concluded that while the existence of pro-eutectoid cementite appreciably retarded cutting, the presence of but a comparatively small amount of graphite completely prevented cutting. The phenome- non, if it were true, is unique, for it would pre-suppose the incom- bustibility of carl)on. Science contradicts us innnediately. In fact, our own welding practice belies us. We might point to the reaction accompanying the removal of carbon from automotive cylinders by the oxygen method — or, leaving our immediate field, we might mention CUTTING WITH THE GAS TORCH 271 the explosive combustion of carbon in ordinary gun-powder. We are forced to conclude then that, far from retarding the combustion of the steel matrix, the graphite of cast iron should actually assist it. We investigated further to determine how much graphite influenced cutting. AVe obtained specimens of so-called malleable castings of the characteristic "black heart" structure. Such a structure is made in this country by the annealing of white cast iron in which all of the carbon exists in cvmentite or pearlite, the latter in some cases entirely removed. The treatment decomposes the cementite to precipitate the carbon in minute particles, differing from the graphite of gray cast iron in their extreme subdivision and uniform distribution throughout a ferrite matrix. In making a black heart casting, an oxidizing pack- ing is used in this country so that while the core is that of a black heart casting, the mass near the surface is ferrite. We removed this shell of ferrite so that our materials indicated, under the microscope, a uniform aggregate of ferrite and temper carbon. By preheating this piece to a dull red heat, it was cut with the characteristics of a high- carbon steel. Then we were satisfied that carbon as such did not prevent cutting, but that the physical state of that carbon was responsi- ble. As plates of graphite, cutting was prevented ; but as finely divided particles, cutting was scarcely impeded. Reconsidering our previous observations in the light of this de- velopment, we began to substantiate our first logical hypothesis. We found, to summarize, that ferrite permitted most readily to be cut. Pearlite with pro-eutectoid ferrite did not materially affect the condi- tions. A completely eutectic "composition first suggested a transitory stage. The existence of pro-eutectoid cementite retarding cutting ; but preheating of the piece to a red heat readjusted the conditions so that cutting was again as efficient as in the case of ferrite. As the com- paratively low temperature produced by preheating was insufficient to effect any change in the physical state of the constituents of the alloy, w-e were forced to conclude that the addition of heat units affected a definite constant, which we assumed was the heat of com- bustion of the iron, as the two forces were of like characteristics. Then a constant result from a variable made axiomatic the existence of a second variable. Our second variable then, we concluded, was the cooling effect of the stream of cutting oxygen, and a further thought suggested a third variable in the time of chemical reaction between the iron and oxygen. The preheating flames ignited the steel — the cutting oxygen produced combustion — and the pi-opagation of the cut was a natural consequence. But as the carbon content was increased, the speed of the reaction was materially lowered; however, the velocity of cutting oxygen to insui'e a continuity of oxygen and slag to the bottom of the cut, was a constant. Then, eventually, a point was reached where the rate of combustion between the iron and oxygen was so slow that the heat units liberated from the reaction were dissipated to such an extent as no longer to ignite adjacent masses of metal— and cutting ceased. By preheating the piece before 272 GAS TORCH AND THERMIT WELDING cutting, we add to the forces on the weakening sido of the equilibrium, and cutting once more obtained. The heat units so obtained com- pensated for the relatively less heat units liberated from the chemical combination of the iron and oxygen in a definite unit of time. While the pearlite and pro-eutectoid cementite are readily com- pensated, the graphite carbon effectively prevents cutting by the ordinary means. No addition of heat units short of incipient fusion, by pre- heating the object, restores the equilibrium. We cannot strengthen further one side of our equilibrium, but we have not attempted to affect the other side. We had made no attempt to reduce the cooling effect of the cutting oxygen. We therefore experimented in this direction, and found that we could so effectively preheat the cutting oxygen that we could restore the equilibrium without preheating the object. In regard to the foregoing it will be of interest to the reader to know that the following article was published in the July, 1919, issue of Autogenous Weldi7ig: Fig. 227. — Four Corners of Large Oast-Iron Stone Crusher Head Beveled with Cutting Torch for Welding. "Substantial progress has been made which shows that cast iron cutting with the torch is a practical commercial proposi- tion. Proof is shown in the two views of the four corners ol a large stone crusher head that were prepared for welding by beveling the edges with the torch as shown in Fig. 227. The job was accepted in our welding shop, with a promise of com- pletion in two days, but it was found that a much longer time would be required alone to bevel the edges by chipping. The Staff of the Engineering and Research Department of the Davis- CUTTING WITH THE GAS TORCH 273 Bournonville Co., which had been experimenting in cast iron, was appeaknl to, with the result tliat the four pieces were made ready for welding in less than one hour ! "Each corner piece represents a cut 4-2- in. thick and 17 in. long, with an area of 76 sq.in. The cuts were made in 6^ min. each, using 24 cu.ft. of oxygen and about 4 cu.ft. of acetylene. The cut surface produced was smooth and the edges were sharply defined, as is shown in the views. The kerf was about Vju-in. wide at the top and bottom — about the same as would be produced by cutting steel of the same thickness. The Fig. 228 — Cutting Torch Made to Preheat the Oxygen Cutting Jet. process was not one of melting, as the sharp edges prove — in fact the finish of the cut surfaces compares favorably with that of steel. After the cuts were started they were carried through to completion without a stop, and the pieces dropped apart of their own weight." Since it is known that the cutting of cast iron is principally accomplished by preheating the oxygen, attention is called to the fact that there have been cutting torches on the foreign market for several years so made as to preheat the oxygen cut- ting jet. One of these is shown in Fig. 228, the principle on which it is made being self-evident. Another torch which is a combination carbon electrode and 274 GAS TORCH AND THERMIT WELDING oxygen jot is shown in Fig. 229. This was patented by R. E. Chapman and J. W. Kirk in 1918. The construction is obvious as the electric current source and connections are shown at the left and .the oxygen tank and tuliing at the right. The carbon electrode has a hole in the center through which the oxygen jet is projected. As the temperature of the arc is considerably Fio. 229. — Carbon Electrode and Oxygen Jet Torch. higher than that of the gas torch the oxygen is highly heated as well as the metal. OXY-HYDROGEN CUTTING Elmer II. Smith, secretary of the Commercial Gas Co., writing in the Welding Engineer says : As I had never had the opportunity to get an accurate cost on oxy-hydrogen cutting, either through experience of others or through my own worlv I was very glad to tal^e the data on a recent jol) wliere I could get absolute facts and figures. The work consisted in splitting a number of steel plates l.S/lG-in. thick by -(> ft. 10 in. long and was done by one of our spring motor cutting torch carriers with our standard straight line torches with a No. 1 cutting tip made especially for hydrogen cutting. As cutting is a process of burning we convert the iron to iron oxide or FesOj, which means that the burned iron is composed of three parts of iron and four parts of oxygen. Since the atomic weight of one part of iron is oG as comitan'd to IG for oxygen, this means that the weight of the oxide would l»e composed of ;> X -"'^J i^^^it weights of iron and -IxlO unit weights of oxygen, or a ratio by weight of 16S parts of iron to G4 parts of oxygen. In other words, if we were to take 232 lbs. of slag produced by cutting it would contain 64 lbs. or about 700 cu.ft. of oxygen and IGS lbs. of iron. For means of comparison the following figures are set forth, showing CUTTING WITH THE GAS TORCH 275 what tlio jj^as coiirsuiiiptiou wovild be on this particular class of work if the cutting was done with theoretical efficiency. Fe304 = Fe IGS, O 64 (pts. by weight). 2G ft. 10 in. = 322 in. 322 in. — 13/lG-in. thick = 261.6 sq.in. of cut. 4 cuts = 1046.4 sq.in. 1046.4 sq.in. of cut 3/32 in. thick ^98.1 cu.in. metal removed. Steel weighs .2831 lb. per cu.in. 98.1 cu.in. weighs 27.8 lbs. Now if all the oxygen used had combined with the iron to do the cutting Ave would require 64/168 of 27.88 lbs. or approx. 10.6 lbs. of oxygen. As 1 lb. oxygen equals 11.209 cu. ft., the amount required theoretically is 118.81 cu.ft. In actual practice, however, we cannot obtain any such efficiency, as the cutting jet must be of sufficient volume and pressure to blow out the slag and clean the kerf of oxide. This means that an excess of oxygen must be used to do the more or less mechanical part of the work. In addition some oxygen must be supplied to coml)ine with the hydrogen to provide the heating jet. None of this oxygen is used in converting the iron to FejOi and represents a total loss as far as the theoretical figures go. Of course if we expected to get down to exact figures we should have an analysis of the steel to be cut and allow for the oxygen neces- sary to convert the carbon and manganese to oxides, but the proportion is so small as to be almost negligible. Now let us see what obtains in actual practice in machine cutting of ordinary steel ship plates free from rust but coated with the usual amount of thin mill scale. I set up the motor and after a short trial adjusted the oxygen pressure to 22 lbs. pressure and the hydrogen to 3 lbs. This oxygen pressure may appear a trifle high, but it must be remembered that it was not only efficiency in gas that was desired but also saving in time and there was too the friction of a 25-ft. length of hose to overcome. (I have often cut 1-in. steel with less than 10 lbs. pressure.) I then adjusted the speed of the machine to 14 in. per minute which was the maximum speed at which the cut w-ould clear itself perfectly. The pressure gages, previously tested for accuracy, showed 1780 lbs. on the oxygen and 1800 lbs. on the hydrogen drum. At the end of the run the gages showed 305 lbs. on the oxygen and 1175 lbs. on the hydrogen. As the drums were 200 cu.ft. capacity the consumption was 1475/9 = 164 cu.ft. oxygen and 525/9 z= 70 cu.ft. of hydrogen. As the oxygen for the heating jet was supplied from the same drum and through the same regulator as the oxygen for the cutting jet it was impossible to determine the amount used in the heating jet, but I would estimate that it was approximately 30 cu.ft. as it was much more than is used in any oxy-hydrogen welding flame. This would leave 134 ft. oxygen for the actual cutting operation, which is but 15.19 cu.ft. more than the theoretical amount required. The kerf was just a few^ thousandths of an inch under 3/32 in. which would throw the balance against the 276 GAS TORCH AND THERMIT WELDING cflicieiu-.v slightly, but eoniputiiig from the aliove ligures it is shown that the actual results are SS per cent of the theoretieul tigures. On account of the laclc of definite figures on cost of cutting with either the oxy-acetylene or oxy-hydrogen process on actual cutting jobs it might not be amiss to compute the foregoing figures to show the approximate cost per sq.in. of cut. I'urity of oxygen 99.5. I'urity of hydrogen 99.8. 1288 lineal inches 13/16 in. thick ==: 1046.4 sq.in. 164 cu.ft. oxygen @ l^c $2.46 70 cu.ft. hydrogen @ Ic 70 Total gas cost 3.1 7 Total oxygen cost per sq.in. cut .$.00235 Total hydrogen cost per s(i.in. cut 0007 Total gas cost per sq.in. cut .$.00305 Oxygen per lineal ft.... 1..53 cu.ft. Hydrogen per lineal ft. . .65 cu.ft. Total gas cost per lineal ft. cut $.02945 These figures cannot be obtained in free-hand cutting and could not be expected as the cutting jet must be maintained at a considerably higher pressure to overcome the unsteadiness on the part of the operator. The heating flame must also be larger in free-hand cutting, or the cut will frequently be lost requiring the operator to turn off the cutting jet and start ovei". From the figures given me on the use of oxy-acetylene cutting in which the oxygen was approximately 98 per cent pure, the cost of the cut per lineal ft. was Oc. However, I did not conduct the test with oxy-acetylene and am taking the results obtained by the foreman in charge. He stated that the pressure in the oxygen was 35 lbs. and on the acetylene 8 lbs. The resulting cut was not nearly so smooth nor regular as that produced by the oxy-hydrogen and in addition there was an accumulation of slag at the bottom of the cut which was entirely absent when hydrogen was used. We all know that if oxygen is diluted with a very small quantity of incombustible gas, such as carbon dioxide, its efficiency is greatly reduced and a decrease of 4 per cent in purity results in an increased consumption of 60 per cent in oxygen. But we may start out with a very pure oxygen in oxy-acetylene cutting and still have a high consump- tion, which would lead one to believe Hiat the oxygen of the cutting jet is diluted after it leaves the cutting tip with the products of com- bustion of the heating flame, which contains carbon monoxide, which is not present in the oxy-hydrogen flame. It was interesting to note that when the heating jet was adjusted to heat at about the same rate as oxy-acetylene in starting the cut the top of the kerf was very slightly melted over making a rounded CUTTING WITH THE GAS TORCH 277 edge, but that when the Ihiiiie wiis so adjusted that it required a half u iniiiute or more to start the cut tlie top edges of tlie kerf were sliarp and square as they were at the bottom. Altliough with the flame adjusted at a point requiring 30 seconds to start the cut, after it was once started no trouble was experienced by losing the cut and in every case the full 26 ft. 10 in. was cut without an interruption, except on the extreme end of one piece which had a thick scale of rust. Although it is seen that in this particular case the consumption of hydrogen is considerably less than half the amount of oxygen, my experience has not been quite so satisfactory in cutting scrap iron, that is scrap boilers, etc., where there is more or less rust and boiler scale to contend with and also the frequent interruption of the cutting operation in shifting from one piece to the next. In cases of this kind the heating flame must be of ample capacity to penetrate the scale and to start tlie cut almost immediately. It is also the custom to have the torch burn while moving about from one cut to the next. This increases the hydrogen consumption about 100 per cent more or less depending on the operator. The thickness of the metal to be cut also determines the proportion of hydrogen used. It is probably safe to say that the amount of both gases used will be practically doubled. If it would not complicate construction to the point of imprac- ticability it would be a step in the right direction to make a cutting torch with a valve control that would open first the valve controlling the combustible gas, then the oxygen for heating jet and lastly the cutting jet valve. CHAPTER XVI CUTTING MACHINES Wliere close cutting to a line or pattern with a gas torch is desired, some mechanical device must be provided for guiding the torch. A properly constructed machine saves time, mate- rial and gas. The more common mechanical devices in use are for feeding Pie. 230. — Cutting a Billet with au Oxweld Machine. the torch in a straight line. These are used to cut bars, bil- lets, boiler plate, armor plate and the like. An Oxweld straight-line cutting machine is shown in Fig. 230. An ordinary cutting torch is used in this case, and the end of a 12 X 12-in. billet has just been cut off. The feed screw may be turned from 278 CUTTING MACHINES 279 citlicr end by means of handwlieels, and means are provided for cross adjustment. Another device, made by the Davis-Bournonville Co., is shown in Fig. 231. The pieces, which were cut the long way, measured 15 X 13^ in. This machine has a liandwheel on one end of the feed screw and a cone pulley, for power drive, on the other. Unlike the device first shown this one is not mounted on legs but has a short section of I-beam for a base. On this account Fig. 231. — Another Straight-Line Cutting Machine. it may be placed on the object to be cut or laid on blocks or horses, as occasion demands. The device shown in Fig. 232 differs from either of the foregoing in that racks are used in place of lead screws. This is made by the Great Western Cutting and Welding Co., San Francisco, Cal. The heavy structural iron base is so made that it may be placed on the work, laid on blocks or horses or mounted on legs. Means are provided for adjusting the torch up or down or at an angle. Any of these machines may be used for cutting out marked square, rectangular, round or irregular shaped holes. Where 280 GAS TORCH AND THERMIT WELDING the metal is thick, it is often better, especially on repetition work, to drill a liolc tlirou^h for starting-. The Kadiagraph.— Tlic Radiagraph shown in Fig. 283 is made by the Davis-Bournonville Co. It is a motor-driven device, with oxy-acetylene or oxy-hydrogen cutting torch, adapted to cutting along straight lines or circles in steel plate from -^ in. to 18 or 20 in. in thickness, the speeds varying from 2 in. to 18 in. per minute, according to the thickness of the plate. For Fig. 232. — Portable Cutting Machine with Rack Feeds straight line cutting, it operates upon a parallel track, and for circle cutting, with a rod and adjustable center. The device consists principally of a three-wheeled carriage driven by an electric motor attached to the carriage, which may be connected to the ordinary lighting or power circuit, either d.c. or a.c, 110- or 220-volt circuit. An adjustable arm and torch holder provides for raising or lowering the torch while in operation, and for adjustment at an angle for bevel cutting. The adjust- able arm also permits of following an irregular line within a CUTTING MACHINES 281 Fig. 2'A'A. — Davis-Rournoiiville Radiagrapk Fig. 234.— Radiagraph Cutting Steel Plate at the New York Shipbuilding Yards. 282 GAS TORCH AND THERMIT WELDING variation of 3 in. on either side of a straight line. The cutting torch is connected by hose to the gas supply. The machine is portable, weighing approximately 50 lb. complete, and has proven an invaluable aid in steel cutting, greatly facilitating such work in shipyards and steel mills, several machines being employed advantageously in some of the larger plants. An example of some of the straight line cutting done by the Radiagraph is shown in P'ig. 234. Here the track has been Fig. 235. — Davia-Bournonville Raiiograph. laid on a heavy piece of ship plate and the torch is fed along at a uniform rate by the motor. The Raiiograph. — For cutting railway rails the device shown in Fig. 235 is used. This is clamped to the rail while it is in position on the roadbead if desired. The cutting torch may be mounted in a holder on either side of the rail. Each holder is carried by a slide. Attached to each holder is a roller which runs in contact with a cam formed in such a way that it pro- vides for maintaining the tip of the cutting torch at a uniform distance of about J in. from the surface of the work as the torch is fed around the rail. Feeding of the torches is accomplished by two handwheels which transmit motion through a set of CUTTING MACHINES 283 siiital)le gearing. In operation the torch is first applied at one side of the rail and fed over the line on which the cut is to be made, one-half of the base and head of the rail and the web l)eing cnt in this way. The torch is next removed from the holder and mounted at the opposite side of the rail, where it is again passed over the line of cut, with the result that the remain- ing half of the base and head of the rail is severed. A 9-in. traction rail can be cut off in about three minutes. Fig. 236. — Cutting Heavy Plate at the Schneider Works, Creusot, France. The cutting of heavy steel plate in the great Schneider Works, Creusot, France, is shown in Fig. 236. The portable devices are very similar to those used in the United States. In the background is a huge machine so made that it can be used to trim ends, square up a plate and cut angles or circles. The torch carriage is fed along by a lead screw run by a motor seen at the extreme right. A motor-operated device will raise 284 GAS TORCH AND THERMIT WELDING or lower the frame or give it a circular movement. The plates to be cut are run into position on small flat cars. Circular Cutting-.— The l-tacliagraph cutting- circles, is shown in Fig. 237. The work was 2^ in. thick and was cut at the rate of 6 in. per minute. Note the true circle and surface of the cut. The pieces were for a special type of heater for the Gov- ernment. The round piece, or flue sheet, is 30 in. in diameter and the ring, or flange, 45 in. outside diameter. Another device is tlie Holograph, shown in Fig. 238. It Fig. 237.— Radio,i;raph Used for Circular Work. is a device for cutting holes in the web of a rail, or in struc- tural iron, of not more than ^ in. thick. It is quickly attached and accurately adjusted. It pierces through the iron almost instantly, without any previous drilling, and will cut smooth round holes from 1 to 2 in. in diameter in from 30 to 60 sec.' It is particularly adapted for railroad work, and enlarging or cutting holes in building and bridge work. The Mag-netogTaph. — The Magnetograph shown in Fig. 239 was designed for mechanically cutting circles up to 12 in. diameter in steel plate in perpendicular position, such as cut- CUTTINC; MACHINES 285 tiii<>' poft liolcs in tlic side ])lal(>s ot: sliips. Stool plalo fi-om -] ill. up to several inches tliick is cut quickly, witli a finished Fig. 238. — Davis-Boumonville Holograph. Fig. 239. — Davis-Boiirnonville Magnetograph. and true surface, the movement of tlie oxy-aeetylene or oxy- hydrogen torch and flame being given by handwheel and gears. 286 GAH TORCUI AND THERMIT WELDING (Alt ting is accomplished at varying speeds according to thick- ness of plate, from 3 in, up to 20 in. per minute, oi' even faster on light plate. The device is constructed as much as practical of aluminum to obtain lightness, and is held firmly on tlie plate by means of three electromagnets, connected by wire to an electric circuit (direct current) or to battery. The Camograph.— The Camograph, Fig. 240, is an adapta- tion of the Holograph. It is of the same general construction, except that it is larger and has a wider range of work. It is fitted with a cam for each particular kind of work, and Tig. 240. — The Camograi^h. will cut almost any form desired, within the capacity of the machine. This machine requires special cams iov each operation. The Camograph No. 2, shown in Fig. 241, is a later develop- ment of the Davis-Bournonville Co. It is automatic in opera- tion and is used for cutting openings in steel plates that cannot be done conveniently or economically on a drilling machine. The torch is mechanically traversed over a fixed path and at a predetermined speed. The path followed is controlled by an internal cam at the top of the machine, the shape of which determines the shape of the opening being made, the double- jointed radial arm permitting universal movement of the flame CUTTING MACHINES 287 which perforates the steel. The principle of the cam guiding action is unique. The feed roller is magnetized by a powerful electromagnet, and is thus attracted to the inner face of the cam, the parts in contact being made poles of the magnet, one of which rotates and thus acts as a traction driver. The roller is driven by a sniall variable-speed motor through double worm Fig. 241. — Camograph No. 2. gearing, the magnetic attraction being sufflcient to cause it to travel along the face of the cam in a positive manner. Direct current is required owing to the magnetic feature, and the control consists of a double push-button switch for starting, stopping and also for energizing the magnet. Arrangement is provided whereby when the cutting oxygen is turned on the feed motion automatically starts. The nominal diameter of 288 GAS TORCH AND THERMIT WELDING the largest hole cut is 7 in., but openings other than circular, liaving one dimension much larger, may l)e provided for. All thicknesses of plate used on the largest marine boilers are readily cut with this machine. The macliine is 17 in. wide, 15 in. deep, 25 in. high, weighs 125 lb., and uses 110 volt, direct current. The Great Western Cutter. — The machine shown in Fig. 242 is made by the Great Western Cutting and Welding Co. FiQ. ^4:2. — (.ireut Western Cutter. It is designed to cut round, square or oval holes. Three master plates are furnished for holes of these shapes. By- turning the handle the torch travels around the inside of the form, to which it is held by the two coiled springs shown. The machine is simple and liglit. Extensions are furnished for cutting large holes. For odd-shaped holes extra plates are required. This machine is especially adapted for boiler shops, shipyards, etc., in cutting hand holes, manholes, fire-box door holes, and holes in tube sheets. CUTTINCx MACHINES] 289 Tlir maoliine shown in Fig. 243 is known as the Pyrograph and is made by the J)avis-lU)urnonvillc Co. The model sliown is not the latest, but well illustrates the general principles of the more impi-ovcd ones. It was designed primarily for boiler- shop use in turning flanged boiler heads or cutting openings for doors, manholes and the lilvc. In one shipyard boiler plant, flanged combustion chamber heads, f in. thick with a flange periphery of 27 ft., were trimmed and beveled to the calking angle in 30 min., exclusive of the setting up. Fig. 24;>. — Pyrograph Trimming and Beveling Boiler Flanges. As can be seen, the Pyrograph comprises a motor-driven carriage supported on a radial arm of* a length that provides for cutting the flange of a 9-ft. diameter boiler head at one setting. "While the largest diameter circle that can be cut at one setting is 9 ft., much larger work may be trimmed and beveled, inasmuch as the arm can be swung through a semi- circle of 20 ft. or a full circle of 20 ft. diameter, provided the shop conditions permit the arm to swing in a complete circle. Heads larger than 9 ft. diameter are reset as many times as may be found necessary to reach the flange all around. The radial arm construction is light but rigid, consisting 290 GAS TORCH AND THERMIT WELDING of two cokl-rollcd parallel round steel bars firmly tied together by end connections and intermediate spacer blocks, and sup- ported by a truss rod. The vertical cast-iron pivot member of the radial arm is mounted on ball bearings at the top and bottom, in order to insure the maximum ease of movement. The steel post around which the radial arm swings is adjustable vertically by means of a crank operating a rack-and-pinion gear. A dog and racket hold the post at any height within the limits of adjustment required. The colum-U has a broad flanged base which may be bolted Fig. 244. — Details of Pyrograph Feed Mechanism. to a cast-iron floor plate or a concrete foundation if required to be self-supporting, or the top of the post may be shackled to a column of the shop building and the base supported on an ordinary floor without an individual foundation. The carriage is supported on the radial arm by four grooved ball bearing rollers which provide for the easy radial move- ment required to follow the feed action freely. The carriage and the arm derive their movements from the feeding mechanism which operates directly on the part to be beveled, the flange part itself acting as the track and guide for the feeding mechanism, as shown in Fig. 244. CUTTING MACHINES 291 The torch is adjustably mounted on the carriage beneath the radial arm, and the tip may be directed at any angle required to cut to the desired calking angle. The flange to be trimmed and beveled is gripped between the three feeding rollers, two of which are small idlers on the side next to the torch while the driving roller, considerably larger, is located on the far side of the flange. The driving feed roller derives its motion from a small electric motor mounted on top of the carriage and driving through a reducing train of worm and bevel gears. Variations of speed are pro- vided by making the upper worm and worm gear replaceable with worm gears of different ratios. The following speeds are available: 12 in. in 70 sec, 12 in. in 90 sec. and 11 in. in 60 sec. The pressure on the feed rollers required to produce the traction necessary to traverse the torch and carriage is obtained from the weight of the torch, the slide rests on which it is carried and the frame to which the two idler feed rollers are attached. The frame carrying the slide rests and idler feed rollers is pivoted, and the weight forces the idler feed rollers against the side opposite the driving roller with sufficient pressure to traverse the carriage positively. The feed mechan- ism operates on any shape whether straight or curved, thick or thin. Flanged sheets are generally rough, presenting a more or less irregular contour, but this does not interfere with the carriage traverse and the torch action. The operator may interrupt the feed at any point by raising the frame, thus relieving the pressure on the feed roller. The driving roller and its shaft are protected by a shield of fireproof composition having a beveled flange at the bottom, on which the sparks and slag have no effect. The machine, once set, trims a flanged sheet evenly all around, provided the sheet has been properly leveled. OtherAvise it is necessary to chalk a line to be followed. In the plant of the New York Shipbuilding Corporation, three different combustible gases are used in cutting torches, namely, carbo-hydrogen, acetylene, and hydrogen. The com- bustible gas selected for different classes of work depends upon the thickness of the plates Avhich have to be cut. The range of thickness handled by the different gases is as follows: Up 292 GAS TORCH AND THERMIT WELDING to 3 in., earbo-liydrogcn ; 3 in. to 6 in., acetylene; and over 6 in., hydrogen. It will, of course, be understood that either of these gases is mixed with oxygen. A Universal Cutter. — A machine built somewhat along the lines of the Pyro graph, but a much more universal machine, has been developed for use in the shops of the Genei'al Electric Co., Schenectady, N. Y. This machine is sliown in Fig. 245. It can be set for automatically making circular, spiral, radial or tangential cuts. Its rate of feed can be varied from 1 to 72 in. per minute, according to the character and thickness of the metal. The base of the machine is provided with a Fig. 245. — Automatic Universal Cutting Machine. powerful electro-magnet to be used if the machine is placed on a rough or uneven surface and also to hold it in position when it is necessary to perform cutting operations on work held in a vertical plane. Ordinarily, the weight of the machine is sufficient to hold it steady. As shown, the machine is mounted on a truck for easy transportation, as it weighs 1,900 lb. The Oxygraph. — With the Oxygraph, steel plate from 1 in. to 15 in. or more in thickness is cleanly cut with a narrow, smooth kerf, along straight lines, sharp angles, or curves, according to drawing or pattern. The pantagraph principle is employed, with a motor-propelled tracing wheel, with which the lines of the drawing are followed and reproduced with CUTTING MACHINES 293 the cutting torch. Either the oxy-aeetylene or the oxy- hydrogcn cutting flame is used, with hose connection to the source of gas supply. The only power required is for revolving Fig. 246. — yingie-Torch Oxygraph. Fig. 247. — Oxygraph with Two Torches. the tracing wheel, and this is supplied by a small motor attached to the tracing head, which may be connected to the ordinary electric light or power circuit. A universal motor, either d.c. or a.c, 110- or 220-volt circuit, with rheostat and friction 294 GAS TORCH AND THERMIT WELDING governor is used. Tlic speed of cutting varies from 2 to 18 in. per minute, according to the thickness of steel being cut. One size of machine is applicable to small work and die cutting, within a cutting area of 16 in. square, a circle of 18 in., or a rectangular form 12X40 in. may be cut by extension of the tracing table. With this machine, a drawing or pattern double the size of the cut to be made is required, the drawing being placed on the tracing table shown at the right in Fig. 246. Another machine is made for larger, heavier work. It has a double pantagraph frame and is fitted with two cutting torches for making duplicate cuts at the same time, the position Fig. 248.— Cuttinq- Out a Large Slot. of the torches and tracing wheel being adjustable. The entire pantagraph frame may be moved backward from the table to allow placing of heavy plate with a shop ei-ane. This machine reproduces the cut of equal size with the pattern, or 1 to 1. A machine with two pieces of work and iDattern in place is shown in Fig. 247. Another practical application of the Oxygraph is shown in Fig. 248. The piece worked on is a fishing tool used for fishing out broken tulics in oil wells. This Oxygraph has a bed frame 30 in. wide l:)y 9 ft. long. The fishing tool is hollow, with walls 2^ in. thick, and weighs 900 lb. The total cut made of 21 lin. ft. was made in 21 min. or 1 ft. per minute. CHAPTER XVII WELDING SHOP LAYOUT, EQUIPMENT AND WORK COSTS The layout and equipment of a welding shop will, of course, vary with the class and amount of work handled, the capital available, and the personal opinions of the owner. One should, however, have enough equipment of a mechanical nature to insure the finishing of work in a reasonable time without too great an expense for labor. A first class workman can, when necessary, turn out a good job of difficult work with a single welding and cutting outfit; means for preheating which may consist of a few firebrick, asbestos and charcoal ; a chisel or two ; a good hammer and a few files. These are insufficient, however, where any amount or variety of work is to be handled economically and to the satisfaction of the ordinary run of patrons. A minimum amount of mechanical equipment should include a number of hand and handled chisels, several ham- mers and sledges of different weights, a portable electric grinder or at least a grinding stand, and a portable electric or a stationary drilling machine, or both. To this, for more extensive work should be added a pneumatic or an electric chipping hammer, a lathe, cranes, and possibly a portable or a stationary motor-cylinder grinding machine. Oil- or gas- burning i3reheaters are also almost a necessity in any case, while a gas-burning preheater of the table type, will save an enormous amount of time and trouble on the general run of gasoline motor work. Special grated iron welding tables, heavy surface plates and grids, iron blocks and straps and numerous other articles will need to be added as local requirements dictate. The shop layout for equipment will liave to conform to the building unless the shop is built purposely for the work. In this connection very few suggestions of any value can be made, 295 29G GAS TORCH AND THERA/IIT WELDING except that tlie shop manager should endeavor to so place his equipment as to cause the least running back and foi-tli possible. We will, for the benefit of our readers give tlie layout of a large shop doing nothing but Avelding work. This is the shop of the Oxweld Acetylene Co.^ Newark, N. J., and it was built expressly for this work. Allowance in position had to be made for the set directions of the railroad and street lines. Fig. 249.— -Exterior View of Oxweld Shop, Showing Crane Hoists. Fig. 249 shows the end of the Iniilding next to the railroad. The overhead track for the chain blocks is so placed as to be readily used for loading or unloading either cars or trucks. This is good, but a still better arrangement would be to extend the runway on into the shop itself and so save considerable rehandling in order to get the work to or from the welding floor. Fig. 250 is a floor plan showing the location of the various benches, lockers, machines, etc. WELDING SHOP LAYOUT 297 /jpunoj 298 GAS TORCH AND THERMIT WELDING WELDING SHOP LAYOUT 299 300 GAS TORCH AND THERMIT WELDING WELDING SHOP LAYOUT 301 111 Fiji^. 251 is shown a view of the shop just inside the nortlu'i'ii end. Tlu> doors sliown at the rifflit are tlie ones Working Order— Welding Shop JOB No . DATE _ _ tSI NAME - - ADDRE!)S PER HOUR. GAS. MATERIAL AND EXPENSES EXTRA. INVOICE S - SA ^ AMT. OF COMMIS SION S _ 91 SHOP TRANSIT GAS LABOR I *RT,CLE TOTAL P«,CE TOTAL COST %"%%". IcftYltne Dati ""-nT' HCU., HATE total] Chahcoal . PRIHKATINO Cast iaon ?/on'w%e *•■"« R°«:::: e.ONZK Bma» ■ Flux STIKk RSb TOTALS 1 1 A«BeiTO« INDIRECT CHAHG OlovK* Cahbon. Block Gogol C5 TOTAL OVER HEAD _ «..._ 1 ___.— .. Gas AMT. Phicb Totau O.T..M 5 BO.OO O.D EXPMCSS ACtT.L.NE S. _ 1 r s.,^ ,„ CARTAn BASE TOTAL COST-SHOP TRANSIT % COM. TO SALESMAN - S 1 "'"*""■ PROFIT FOR COMPAN v..._ - s....._, J Fig. 254. — Cost Keeping Form. that open ont nnder tlie crane shown in Fig. 249. This interior view in Fig. 251 gives a good idea of the lighting and the .302 (iv\s -roRcri and ^riij-:RMiT welding ventilators at the top for eari-ying away the fumes. Tlie air and acetylene pipe lines are shown, aiid in tlie left foregj'ouncl is illustrated in the way cylinders are chained. In the central foi'eground one of the workmen is chipping a casting with a pneumatic chisel. The opposite end of the shop is shown in Fig. 252. Here a portable crane is shown in the middle foreground. Suspended from it is a portable electric grinding machine. Just back of this is an electi'ic grinding stand. At the right, in the back- ground, is a "Wiederwax preheater and just in front of this is an iron preheating and welding stand with an operator at work at it. At the left in Fig. 253 is shown a number of welding tables with grated (tast-iron tops and welded angle- and strap-iron legs. Both the daylight and artificial lighting are excellent throughout the shoj). Probably no other shop would be built (exactly like this, as conditions differ so radically, but a careful study will reveal to the prospective shop man some of the things that will, or will not, apply to his particular case. Keeping Track of Costs.— No shop can succeed financially without keeping a close watch on cost of material, gas, labor, overhead, etc. The way this is done in the Oxweld shop will be seen by referring to the form shown in Fig. 254. This is so made as to cover both inside and outside jobs. These forms are made in duplicate on white and pink paper, so that a cai-bon of the original is made. These forms are for shop and office use only, and from them the customer's bill is easily made out. With forms of this kind, the entire data relating to any job may ])e had at any time hy reference to the files. Another form of cost card, suggested by the Imperial Brass Manufacturing Co., is shown in Fig. 255. This is not so com- plicated as the form just given, and will answer in many eases. The manager should not forget, however, to add to this the cost of overhead, which it is wise to make fairly high to allow for contingencies — say from 100 to 150 per cent. Carbon Burning-. — While carbon burning has nothing to do with welding, the ordinary welding shop is often called upon to do such work on account of having a supply of oxygen at hand. (!ai-bon in a motor cylinder is caused hy imperfect com- WELDING SHOP LAYOUT 303 bustion. It may be that the carburetor was not adjusted so as to give sufficient air, or it may be too much oil was uSed. The use of oxygen is the most practical and thorough way to remove this deposit. Oxygen gage, start ISOO lbs.=100 cu. ft. Oxygen gage, finish 900 lbs.= 50 cu. ft. Oxygen used 900 lbs.= 50 cu. ft. Acetylene used : 50 cubic feet @ .$0,025 ,$1.25 Oxygen used : 50 cubic feet @ .02 1.00 PREHEATING COST Charcoal Gas, i hour, 2 burners @ .GO .30 Kerosene LABOR (Preparing) : 1 hour 30 inin. @ .GO .90 LABOR (Welding) : 1 hour 30 niin @ .GO .90 LABOR (Finishing and testing) : 1 hour niin @ .30 .30 RODS : Lbs. Steel @ 15 Lbs. Cast Iron @ .10 1.50 Lbs. Bronze @ Lbs. Copper @ Lbs. Aluininiuii @ FLUX: 4 Cans Cast Iron.. . , » , . . . . o , . o o = . . . o , , . .(5) .50 .25 Total $G.40 REMARKS Fi(i. 255. — Suggestion for Cost (Jaid. A decar])onizing outifit is shown in Fig. 256. Here A is the oxygen tank valve, B the tank coupling, C the pressure gage showing the pressure at which the oxygen is delivered to the 304 GAS TORCH AND THERMIT WELDING "torch," D the regulating screw, E hose connection, F trigger valve, 6r hose connection and H the flexible copper tip. To use this outfit, connect it up as shown, then with the motor running shut off the gasoline and let the motor run down. If the engine is particularly dirty, it may be advisable to protect the carburetor and pan by placing some asbestos paper at points to prevent fires from flying sparks. Fig. 256. — Imperial Decarbonizing Outfit. Remove spark plugs from cylinders — not the valve caps. Crank the motor until the cylinder to be started upon has the piston at the top, with both valves closed. Set the pressure on the regulator at about fifteen pounds and partially depress the lever on the handle of the carbon burner. Use a Avax taper or drop a lighted match into the spark plug opening of cylinder, at the same time directing the copper WELDING SHOP LAYOUT 305 tube of the carbon Inii'iier at that point. This ignites the carbon, and it* it is not too dry, the oxygen shonUl thereafter be sufficient to completely consume it without again lighting it. At the start, particularly if the cylinder is oily, there will be some flame as well as considerable sparks. Hold the pressure down until the flame has practically disappeared, then press down the lever all the way and move the nozzle back and forth around the walls until sparks stop. Sometimes the cylinder is very dry and the carbon is rather difficult to burn. This can be more or less determined by the appearance of the spark plug. If it is dry, squirt about a teaspoonful of kerosene into the cylinder, spreading it over as large a surface as possible, to aid the burning. The copper tube is flexible and may be bent as desired to reach any portion of the cylinder. Actual contact with the carbon by the tube is not necessary to consume it — carbon burns in an atmosphere of oxygen after it is ignited. The only possible danger to the cylinder, valves or piston is a too high pressure of oxygen on an extremely oily cylinder — there would be considerable heat generated in this instance. Hold the pressure down, then, until the flames have gone and sparks only are being thrown out before fully opening the lever on the handle. When through cleaning, it is desirable to remove the valve cap and blow out any solid particles there may be present; these solid particles cannot be carbon, but may be pieces of iron, etc. The appearance of the cylinder will be considerably improved by swabbing off the top of the piston and valves with an oily rag. Carbon burning is a very practical solution of carbon de- posits — but care and horse sense must be used, though the process calls for no particular degree of skill. SAFETY RULES FOR GAS-TORCH WORKERS The following rules were adopted in 1920 by the Western Pennsylvania Division of the National Safety Council : Equipment Rules 1. All pressure tanks should be fitted with safety relief devices, and tanks not so equipped should not be used. o()(3 GAS TORCH AND THERMIT WELDING 2. The cqnipinenl sliould include a higli-pivssure gage to indicate the pressure on tlie taid;, a reducing valve, and a low-pressure gage to indicate the pressure on the torch. These should he assenihled as one unit and so arranged that they need not he separated when they are attached to, or detached fror.i, the tank. The two gages should have different-sized openings: one should have a right-hand thread and the other a left-hand thread so that they cannot he interchanged. There should he one of these units for the oxygen tank and one for the acetylene tank. 3. All pressure regulators should he equipped with a safety relief valve which will relieve the pressure from the diaphragm and low- pressure gage in case the high-pressure valve should develop a leak. 4. Wire-wrapped hose should not lie used. 5. The oxygen and acetylene hose should he of different color or the couplings should he stamped for identification purposes, so as to avoid interchanging the hose. 6. The torches should l)e of a type which will not hacktire. Rules for Opeeation 1. Under no condition should acetylene he used where the pressure is greater than 15 Ih. per square inch. 2. Special care should he given to the storage of oxygen and acet- ylene tanks. Acetylene is classed as an explosive, and only a limited number of containers should be stored in any one place. Oxygen tanks should be stored in a separate place from acetylene tanks. o. Oxygen and acetylene tanks should not be allowed to remain near stoves, furnaces, steam heaters or other sources of heat, and should not be exposed unnecessarily to the direct rays of the sun, as an increase in the temperature of the gas will cause a corresponding increase in the pressure within the tank. Any excess of heat may also soften the fusible safety disk with which the tank is provided, causing it to blow out and permitting the gas to escape. 4. Oxygen tanks should never be handled on the same platform with oil or grease which might find their way into the valves on the tanks. 5. Oxygen and acetylene tanks should never be dropped nor handled roughly, and should never be stood on end unless fastened so as to prevent them from falling over. 6. Tanks should not be handled by crane, either magnetic or mechanical. 7. All empty tanks should be marked plainly with the word "empty" and returned promptly to the storeroom. 8. An open flfime should never lie used for the purpose of discovering leaks in acetylene tanks. Tweaks can generally l»e detected ])y the odor of the acetylene gas, and their location can be determined by applying soapy water to the surface of the tank and watching for the soapy bubbles formed by the escaping gas. 9. No repairs to oxygen or acetylene tanks or equipment should be WELDING SHOP LAYOITT 307 inado or altciniilcd. All dclVcts should 1)0 reported ijroniptly to tlu' rorcniiin, and liy liiiii to the niaiiui'acturcr. Id. lA'akiii,u' act'tylonc tanks should not be used, but should l)i' placed in the oi)en air and all open lights be kept away from then). All leakin,u acetylene tanks should be reported promptly to the foreman and innnediately returned to the manufacturer. 11. All open liames should be kept away from any place where there is any possiliility of acetylene escaping;. 12. Care should l)e taken to protect the discharge valves of tanks from being bumped, as a jar may damage the valve and cause it to leak. i:>. Grease in contact with oxygen under pressure may cause spon- taneous ignition, (treat care should l)e taken not to handle threads or valves with oily hands or gloves, and gages should not be tested with oil or any other hazardous carbon. If a lubric;int must be used, the purest glycerine is permissible. 14. Gages, apparatus and torches requiring repairs should be sent to the manufacturer, and local repairs should not be attempted. Valve seats sliould never be replaced except by the manufacturer. 15. The use and operation of the pressure regulator or reducing valve on oxygen or acetylene tanks should be as follows: (a) Open the discharge valve on the tank slightly for a moment and then close it. This is to blow out of the valve any dust or dirt that might otherwise enter the regulator. (b) By means of the stud or nut connection on the regulator, connect the regulator to the discharge opening of the tank, (c) Release the pressure-adjusting screw of the regulator to its limit, (d) Open the needle valve slightly if there is one. (e) Open the discharge valve on the tank gradually to its full width, (f) Open the needle valve to its maximum if there is one. (g) Adjust the pressure-regulating screw until the desiivd pressure is shown on the low-pressure gage. 16. The discharge valves on the tanks should be opened slowly, and care should be taken to avoid straining or damaging them by the use of a hammer or the wrong kind of wrench. A special wrench should be made for use in opening these valves in case they stick. 17. When the operation of the cutting or welding torch is stopped for a short time, the needle valve on the regulator should be closed, or the pressure-adjusting screw should be released to keep the pressure off the hose. The torches should be opened momentarily to let the pressure out of the hose lines. IS. All tanks should be inspected at the close of the day's work. 19. Proper precautions should be taken to protect the hose from flying sparks. 20. All hose should be examined periodically at least once every week. This should be done by cutting the hose off at the end of the connection and examining it. In addition, after a few months' use, the hose should be cut off about two inches back of the connection and examined for defects. A defective hose should never be used. 308 GAS TORCH AND THERMIT WELDING 21. Special care sliould l>e taken to avoid tlie intercliange of oxygen and acetylene hose or piping, as this might result in a mixture of these gases that would he highly explosive. The practice of using right- and left-hand threads is recommended. 22. White lead, grease, or other similar substances should never he used for making tight joints. All joints and leaks in equipment should be made tight by soldering or In-azing. 23. The oxygen and acetylene valves at the base of the torch should he tested daily for leaks. 24. Where hydrogen or other gas is used instead of acetylene, the same precautions should be observed as for acetylene. 25. A lire extinguisher should be carried as regular equipment to be used in case of fire. 26. Men vising welding apparatus should wear suitable welding gog- gles for eye protection, having frames tliat are nonconductors of heat (not celluloid), side shields to protect against hot particles of metal, and lenses of proper color. 27. Operators' clothing should be fireproof. 28. If valves become frozen, they should be thawed by hot water, not by flame or hot metal rod. 29. Home-made generators should never be used, as they are unsafe. Only generators permitted by the Board of Underwriters should be used. 30. Where acetylene is used from generators or is piped through the plant, an approved water seal should be interposed between the generator and the piping system, and individual water seals should be placed at each blow-pipe. Water seals should be inspected daily without fail. 31. Portable generators should not be used inside the building. 32. Safety devices on tanks, generators, or apparatus should not be removed or tampered with. 33. In welding brass or bronze, injurious fumes may be given off, making it desirable to wear a respirator. 34. Smoking while on duty should be prohibited. 35. Electric lights in a generator house should be enclosed in vapor- tight globes protected by the regular guards. 36. Snap switches shoidd be placed outside of the generator house in a suitable place, pi'ovided the house is isolated. 37. Piping which is used to carry acetylene or hydrogen should be painted a distinctive color. 3S. The manufacturers should provide couplings for the hose which cannot be mistaken and put on the wrong hose. If the couplings could be made only with the proper connections, it would be impossible to make a mistake. 39. In storage houses where hydrogen or acetylene tanks are stored, the wiring should conform to the same rules as for the generator house, so that an explosion could not he caused by defective wiring or a In'eak in the bulb. WELDING SHOP LAYOUT 309 40. The valves on the piping sliould contain neitlier copper, brass, nor l)ronze. 41. In opening tlie outlet valve of a full tank, do not remove the regulator. 42. The operator should not stand in front of the gages when opening the discharge valves on the tank. If the pressure goes off suddenly, it may possibly destroy the gage and the glass, and parts will be blown out at the front. 43. A label should be placed on every tank of oxygen, stating that it should be kept away from grease. RULES FOR WELDING The following rules, adopted by the Committee on Standards for Locomotives and Cars, U. S. Railway Administration, for the purpose of preventing the abuse of autogenous welding for purposes for which it is not well adapted, have been sent to the regional directors by Frank McManamy, assistant director of the Division of Operation, with instructions to direct all roads to observe the rules in the construction or repair of locomotive boilers, so that any failures which may have been caused or contributed to by unrestricted or improper use of autogenous Avelding may be prevented. 1. Autogenous welding will not be permitted on any part of a locomotive boiler that is wholly in tension under working conditions, this to include arch or water bar tubes. 2. Staybolt or crown stayheads must not be built up or welded to the sheet. 3. Holes larger than 1-| in. in diameter when entirely closed b}- autogenous welding must have the welding properly stayed. 4. In new construction welded seams in crown sheets Avill not be used where full size sheets are obtainable. This is not intended to prevent welding the crown sheet to other firebox sheets. Side sheet seams shall not be less than 12 in. below the highest point of the crown. 5. Only operators known to be competent will be assigned to firebox welding. 6. Wliere autogenous welding is done the parts to be welded must be thoroughly cleaned and kept clean during the progress of the work. 7. "When repairing fireboxes a number of small adjacent 310 GAS TORCH AND THERMIT WELDING patches will not be applied, but the defeetive part of the sheet will be cut out and repaired with one patch. 1- - A?" *-| \< /O" J x^V 1 m /5 J2 /^ IT s U* /£>■ TMieHMesa'<"-P/.ATf use o'tv^^' ^/ s/ ABove Uro g' %2 sy AOoye'^To^- %? s/ ABOVe ifTO-C ■ %2 S2 ABoye Itf' %h S3 B£^o^e W/^tof/ve D ? — \ / — \ \^60*J \. D Fig. 257. — Kinds of Welds Tested and Examples Used as Welding Guides. 8. The autogenous welding of defeetive main air reservoir is not permitted. 9. Welding rods must conform to the speeihcations issued l)y the Inspection and Test Section of the United States Rail- WELDIXC SIIOl^ LAYOUT 311 I'oatL Adiiiiiiistralion for the various kiiicLs of work for which tliey arc pi-csci'ibcd. STRENGTH OF OXY-ACETYLENE WELDS The results of tests made by the Welding Committee of the Emergency Fleet (*orpoi'ation, on oxy-acetylcne welds of 4-r)2S 4-r)29 4-r)30 4-r.8i 4-r):^2 4-r)33 Tabi.k XVIII. — Strength of Oxy-Acktylenk Welds Mark Size- Kind of Weld ritimate Strength Strength Per Sq. In. Per Cent. Av. Per Cent. Break In 1 2 3 I" Sq. Butt 50600 50235 49795 50600 50235 49795 92. 91. 90. 91. Weld ]-m4 1-513 ]" Dia. .. 43315 44885 45160 55000 56900 57400 100. 100. 100. 100. '' 1-r.n i-r)i2 i-.no 2" Dia. " 15520 149515 153900 47500 46800 48200 86.5 85.2 87.5 86.4 » 2--)! 6 2-rii7 2-r)i8 iXU" R. Ang. 7515 7645 7915 40000 40500 42000 73. 73.5 76.3 74.2 Bar 2-.-)! 9 2-520 2-521 :1XH" '' 12605 12820 12150 33600 34000 32300 59.2 61.5 59.0 59.9 Weld B-525 3-52G 3-527 hX}¥' Butt 10890 10775 10935 58000 57500 68500 100. 100 100. 100. Bar Weld Bar 3-522 3-523 3-524 \xn" " 21460 22025 20785 57000 58600 53300 100. 100. 96.5 98.8 Bar Bar Weld ixn" ixn" Lap 10970 10725 10965 58500 57500 68600 100. 100. 100. 100. 21905 22085 21435 59300 58700 57000 100. 100. 100. 100. Bar Bar Bar Bar Bar Bar 5-537 5-538 5-539 .JXU"' Tank 2495 2210 2760 13300 11750 12106 24.6 21.4 22.6 Bar Bar Weld 5-534 5-.535 5-536 iXU" 4936 5675 5196 13100 15100 15600 23.8 27.4 24.7 26.: AVeld Bar Weld 312 GAS TORCH AND THERMIT WELDING various kinds ai-e given in Table XVIII. The 1 and 2 in. bars were of machine steel, turned to size before testing. The sheet was steel ship plate. The careful engineer will find some rather puzzling discrepancies in this table, but it is the best available at the present time. The key numbers of the various specimens will be found to correspond to the types of welds illustrated in Fig. 257. At the bottom of this illustration will be found examples used as guides for welding different joints, the thick- nesses of the plates and end-spaces being indicated in the table just above the examples. The arrows in the illustrations of the test welds indicate the direction of pull. STRENGTH OF OXY-ACETYLENE WELDED PIPE The Linde Air products Co., Buffalo, N. Y., report that they made some tests in their laboratory in 1920, to determine the strength of welded pipe. These tests were intended to prove to a large user of oil pipe from Kansas, that properly welded pipe will not break at the weld under pressure. According to the report made, the Linde engineers welded together two short sections of standard 3-in. iron pipe, threaded the ends and screwed on two standard cast-iron caps. When the cold water pressure test was applied to the breaking point, the top of one of the caps blew out, leaving the pipe and weld intact. The undamaged cap and the remaining portion of the broken cap were then removed and two extra heavy iron caps were screwed on. At a pressure of 6,200 lb. per sq. in. one of these caps let go, still without injury to the weld or the pipe. Again the uninjured cap and remnant of the broken one were taken off and extra heavy steel caps screwed on. This time the caps held, but the pipe split and ripped under the added pressure upon passing the elastic limit, tearing up to, and l^eing effectually stopped by, the weld which refused to give. The next test was made with 4-in. pipe. Two lengths were welded together, the ends threaded and two extra heavy standard caps screwed on. In this test one of the cap heads blew out at 4,400 lb., which gave a total end pressure on the cap of approximately 33 tons, proving that the broken cap was not in any respect defective. The weld was not impaired at WELDING SHOP LAYOUT 313 all. After this test it was suggested that an entirely new Avelcl with other pipe lengths of the same diameter be tried. Accordingly two more lengths of 4-in. pipe were welded, threaded and sealed, this time with extra heavy steel caps made to withstand a working pressure of 3,000 lb. of air. The pressure was applied and the pipe gave way in the threads at 4,200 lb. In all of the tests the welds held securely. PART 1I-THEK3IIT WELDING CHxiPTER I THERMIT WELDING: ITS HISTORY, NATURE AND USES The affinity of finely powdered aluminum for oxygen, sulphur, chlorine, etc., is such that it is utilized to, effect a reduction of metals from their respective oxides, sulphides and chlorides.. This was known for many years and is generally credited to Frederick "Wohler. About 1894 Claude Vautin found that when aluminum in a finely divided state was mixed with such compounds and ignited, an exceedingly high tem- perature was developed by the rapid oxidation of the aluminum. Since fine aluminum will not burn at a temperature below that of molten cast iron Vautin and others first heated the mixtures in a crucible. The result was that the initial temperature was so high at the moment of ignition that the reaction was explosive. Profiting by the experiments already made, Dr. Hans Gold- schmidt of Essen, Germany, discovered a method of igniting a cold mixture of fine aluminum and iron oxide by means of a barium-jDeroxide fuse which was set off by means of a storm match. His first discovery was made about 1895 or 1896 while trying to reduce chromium and manganese. Later magnesium powder or ribbon was used for ignition purposes, being set off in the same way. A mixture of a few pounds of the powders was found to burn quickly and the resulting tempera- ture was very high. The original patent for the reduction of metals, upon wdiicli all his following patents were founded, was granted March 16, 1897, the serial number being 615,700. Over 40 have been issued since and more are pending. About 1898 Dr. Goldschmidt made used of his reduction method to weld two pieces of iron together. From this time on the experiments developed and difficulties were overcome until a process was evolved for the commercial use of the reaction 317 318 OAS TORCH AND THERMIT WELDING for welding and oliici' purposes. Tlie proeess so developed was called tlie Thermit process. The company handling- the mixtures and apparatus Avas originally known as tlie Gold- sehmidt Thermit Co., but in 1918 the lumie was changed to the Metals and Thermit Corporation, New York. The present Thermit reaction is 8Al+3Fe.,O,=9Fe+4AL03. Expressed in weights this is 217 parts aluminum plus 732 parts magnetite=540 parts steel plus 409 parts slag, oi-appi'oximately 3 parts of aluminum plus 10 parts of magnetite will produce on combustion 7 parts of steel. The steel produced by the reaction I'epresents al)out one-half of the oi'iginal Thermit by weight and one-third by volume. Commercial Thermit is a mixture of finely divided aluminum and less finely divided magnetic iron scale. The aluminum is al)out like gi'anulated sugai- and the scale like coarse sand, the 7'atio ]jy weiglit being approximately three of iron scale to one of aluminum. According to tlie company just mentioned the average analysis of Thermit steel is : Carbon 0.0.1 to 0.10 Manganese O.OS to 0.10 Silicon 0.01) 1 ,, ().:;() Snliiluii- O.o:; to 0.04 riiosphonis 0.04 to O.O.j Ahiminuni 0.07 to 0.18 Of course to produce a steel of the foregoing composition the aluminum and ii-on scale nuist be very pure. For the mixture, scale from l>essemei' or open-hearth steel would prob- ably come close to meeting commei-cial demands. The average tensile sti'ength of a Thermit weld of the foregoing average composition is about 61,000 lb. per squai'e inch. This can be varied by adding other elements. The ehistic limit is slightly more than half this figure, or an avei-age of about 34,000 ]i()inids. Temperature and Characteristics.— While the temperature of the I'eaction is too high to be measured by a pyi'ometei' it can be calculated (,uite accurately theoretically and Pi-of. Joseph W. Richards in his hook "Metallurgical Calculations," gives it as 2694 deg. C., which is equal to 4881 deg. F. THERMIT WELDING: ITS HISTORY, NATURE AND USES 319 M. Fery, using liis radiation pyrometer, found tlie temperature of the stream of steel as it issued from the crucible to be 2300 deg. C. (4172 F.). Making allowance for the chilling effect of the crucible this is probably about right. Considering the melting point of steel to be about 1350 deg. C, Thermit steel is nearly twice as hot. There is absolutely nothing explosive about the present Thermit reaction and no danger is incurred in storing it or iiandling the material owing to the fact that it takes over 1300 deg. V. of heat to ignite it. It is for this reason that a special ignition powdei- must be used for starting the reaction. The ignition powder, however, must be kept away from heat, and in pai'ticulai' the box containing it should be tightly closed before the Thermit reaction takes place, so as to prevent any spark from dropping into it. All Thermit materials must be kept dry, for Thermit that has once become wet cannot be restored to its original condition by drying. Plastic and Fusion Methods. — There are two different methods of using Thermit for welding purposes. For con- venience these methods may be designated (1) the plastic method and (2) the fusion method. The first is used pi-incipally for welding together the ends of pipes. Here the pipe ends are first machined off so that they will fit snugly together. A mold is then placed around the ends, and Thermit is poured into tlie mold from a crucible. The Thermit mixture is first placed in the crucible and ignited by means of a small amount of ignition powder set off by a match. After the reaction the molten content of the crucible is poured into the mold and around the pipe ends. By pouring from the top of the crucible the slag enters the mold first and surrounds and coats the pipe ends and the inside of the mold and thus prevents the pipe ends from being burned through. This allows the Thermit to heat the pipe ends to a welding heat, after which they are forced together, causing a slightly upset welded joint. The second, or fusion metliod, is the more commonly used. In using this method a mold is also used to surround the parts to be welded, but the pai-ts must be preheated — usually to a red heat — in ordei- to prevent the Thermit being chilled by contact with the colder metal and causing an imperfect weld. The Thermit to be used is placed in a cone-shaped crucible 320 GAS TORCH AND THERMIT WELDING so made tliat the melted Thermit steel may be rim out of the bottom into the mold, thus preventing the slag from getting into the mold and spoiling the perfect fusion of the parts. In this method it is also necessary to have the parts to be v^^elded some distance apart in order to give the Thermit steel an opportunity to properly fuse the surfaces desired and produce a perfect union. The distance the parts are separated depends on the size and nature of the pieces and whether the weld is to be made on two separate pieces or is merely to weld a cracked place. This last method is especially adapted for welding together large heavy parts of considerable section, on account of the Thermit steel being produced and inti'oduced into the weld quickly in bulk and thereby resulting in but one contraction throughout the entire mass of metal. Provision to compensate for this contraction can always be made, so that when the metal cools there will exist practically no strains. "Welds requiring as high as 4000 lb. or more of Thermit have been completed successfully. A big advantage of the process is that huge welds may often be made without dismantling the machine or structure, thus saving an enormous amount of labor and time in many cases. Neither method, however, is commercially or practically adapted to welding very small sections or long seams in thin sections, which work can be better accomplished by the oxy- acetylene or electric welding processes. It should, however, be used for welding shafts when the break is in a journal. It is interesting to note in this connection that neither the oxy- acetylene nor the electric welding processes has proved practical for welding trolley rails together. All so-called welded joints l)y tliose methods consist merely in welding plates to the rails and the joint therefore is never really eliminated. The Thermit process, on the other hand, has proved extremely efficient and economical for this work, and thousands of Thermit-welded rail joints have l)een in service for years in all parts of the world. Kinds of Thermit Commonly Used. — For commercial weld- ing purposes there are now produced three varieties of Thermit known as : Plain Thermit. THERMIT WELDING: ITS HISTORY, NATURE AND USES 321 Railroad Thermit. Cast-iron Thermit. Plain Thermit is simply a mixture of aluminum and iron oxide, as previously stated, and is used in making pipe welds and welding necks on mill rolls and pinions where the Thermit is merely used as a heating agent to bring the pipe ends up to a welding temperature and the roll and pinion ends to a molten state. . Railroad Thermit is plain Thermit with the addition of I per cent, nickel, 1 per cent, manganese and 15' per cent, mild-steel punchings. This grade is used in connection with steel welds. Cast-iron Thermit is plain Thermit with the addition of 3 per cent, ferrosilicon and 20 per cent, mild-steel punchings, and is used, as its name implies, for welding cast-iron parts. CHAPTER II MAKING PLASTIC PROCESS WELDS Taking up now the various uses of the tlu-ee varieties of commercial Thermit we will first describe in detail the butt welding of pipe. This is done by the plastic process which is especially adapted to welding joints in the pipes in refrigerating plants and for high-pressure steam, hydraulic or compressed-air pipe lines. It is also applicable to the welding Fig. 1. — Pipe-Facing Machine Open to Receive Pipe. of superheater units for locomotives. The joints so welded are permanent, nonleakable and never require attention, and the original cost compares very favorably with that of the special mechanical joints of any type used for refrigerating or high-pressure purposes. All of the apparatus necessary for the work is easily portable, so that the woi-k may be done anywhere. In preparing pipe for butt welding it is first necessary to insure that the ends are cut square. If the pipe is threaded the threaded portions will have to be cut off. While the ends of the pipe may of course be squared by various mechanical 322 MAKING PLASTIC PROCESS WELDS 323 a 324 GAS TORCH AND THERMIT WELDING means tlie company handling Thermit makes a small portable machine for the purpose, which is much more satisfactory to use than anything else. This device is shown in Fig. 1. The operation of this machine is so obvious to any mechanic tliat further explanation of its operation and method of use would be needless, except to say that the crank handle by which< the cutter is rotated has a ratchet on it so that the cutter may be worked in close quarters. If after the pipe ends have been faced off they should become tarnished they should be brightened with clean, fine emery cloth or a fiat carborundum stone, but in no case should the faced ends be touched with a file or the fingers. Fig. 3. — Pipe Held in Clamps, Mold Partly Assembled. When the pipe ends have been properly faced off, the pipe is lined up so that the faced ends will butt squarely together and then the mold is put in place. In welding coils or bends suitable apparatus should be rigged up to keep the pipe in alignment. Where the pipes are close together in coils it is usually possible to spring out the pipe to be welded so as to permit the adjustment of the mold and clamps. A Pipe-Welding Outfit. — A complete outfit for welding pipe, less the facing machine, is shown in Fig. 2. In this cut, pieces of pipe are shown at A; welding portions of Thermit at B; cast-iron mold at C ; at D is the magnesia-lined crucible; E, clamps ; F, turnbuckles for clamps ; G, crucible tongs ; H, gloves; /, pins for tightening turnbuckle nuts; J, ignition powder; K, dark glasses, and L, wrench for tightening nuts MAKING PLASTIC PROCESS WELDS 325 of clamps. Ill order to make the procedure clearer a pipe-mold and clamp unit is shown in Fig. 3. Here the pipe is shown secui'ely clamped in place with the ends butting together. When putting this apparatus in place the clamps are first adjusted at an even distance from the ends of the pipe and securely clamped. The two tension bolts are then adjusted so as to bring an even bearing all around on the pipe ends, but care must be taken not to put such a tension on the pipe that it will buckle when heated. Just enough tension to hold the ends securely is all that is needed. It will be seen that v^hile a set of clamps may be used for a number of sizes of pipe a mold can only be used for the size for which it was made. How the Mold is Used. — The loAvcr part of the mold is Fig. 4.— Mold Fully Assembled for Weld. shown in place, but the top part of the mold is shown at the left ready to be put on. The slightly beveled recesses in tlie top and bottom parts of the mold are for the pouring gate. Before putting the top part of the mold in place care must be taken to see that the lower part is blocked up or held so as to be in close contact wdth the pipe. Tliis may be done by means of wedges, earth or any other means at hand that Avill stay in place during the process. AVith the top part of the mold in place, as shown in Fig. 4, the operator next pre- pares the Thermit for pouring. This is done by placing the crucible tongs on the ground convenient to the mold and then setting the crucible in the jaws of the tongs. It is very im.- 326 GAS TORCH AND THERMIT WELDING portant that the crucible ])e thoroughly dried before using, and if it is a new crucible it is advisable to burn a pound or so of Thermit in it and then pour the contents out on dry sand. This is the quickest and most convenient way of drying a crucible. The operator should have the handles of the tongs placed toward him convenient for pouring the metal when the reaction is completed. If double tongs are used, as some- times is necessary with large crucibles, the helper should stand opposite the operator in readiness to help pick up the crucible at the proper time, lilue or special glasses should be worn to protect the eyes from the glare of the reaction. As each size of pipe must liave its own size of mold so must the portions of Tlu'rmit be measured out in order to have the proper amount for a given size of pipe. This is taken care of by the concern making the Thermit, which puts it into bags, each containing a certain amount of Thermit suit- able for a given welding job. For all ordinary jobs this scheme makes it unnecessary for the operator to do any special calculating in order to know liow much to use in order to do the woi'k and not waste material. Placing- and Igniting- the Thermit. — With a bag containing the proper amount of Thermit tlie operator poui's about one- half into the crucible and the rest in a hand scoop for feeding into the crucible during- the reaction. Having therefore one part of this portion in tlie ci-ucible and the remainder in a scoop he places a half spoonful of ignition powder (barium peroxicTe) in one spot on top of the Thermit in the crucible. This is set off by means of a parlor match, whicli is applied immediately after striking and before the head is burned off, to the ignition powder itself or else to another match head set into the ignition powder. The ignition of the powder in turn starts the Thermit reaction. After the reaction is well started the operator adds the rest of the Thermit from the scoop, trying to keep about one-half of the surface of the molten material covered with unburned Thermit, and pouring in a steady sti'eam until all the Ther-mit is in the crucible. He sliould tlien immediately grasp the crucible with the tongs, obtaining a firm grip, and pour the contents into the mold, the slag entering first. The pour should be made as soon as the Thermit has reacted and MAKING PLASTIC PROCESS WELDS 327 the slag lias conu' to tlic top. The crucible and tongs are then set aside and a short time is allowed to elapse for the Thermit mass to bring the ends of the pipe to a welding heat. Shortly after pouring the operator should turn the tension nuts enough to keep a constant pressure to determine when the pipe begins to soften, lie should then wait for 10 to 20 sec., depending on the weight of pipe, and then force the ends together by means of four quarter turns (one complete revolution) of the nuts. It usually takes about 10 sec. from the time the pipe softens for it to reach a welding heat, or from 45 sec. to 1^ niin. from the time of pouring. Removing" the Mold. — After the clamps have been drawn up, the mold should be allowed to remain in place for 3 or 4 min. longei', after w'hich the clamps can be removed and the cast-iron mold knocked away from the pipe by means of a hammer. The Thermit steel and slag will come away from the pipe with the mold and can be knocked out of the mold afterward. Care must be taken in every case that a complete welding portion be used, as only the full measure of Thermit will give a good weld. It is advisable where joints are being welded in quantities to have several molds so as not to use molds continuously while hot. It is also advantageous to allow the mold with its contents of steel and slag to remain on the pipe for a considerable time before" removing. If they can be left 10 to 15 min. it is all the better. It has been found in practice that two men, one facing the pipe with the pipe-facing machine and the other doing the welding, can complete a weld inside of 10 min. and that it is a simple matter for them to make from 40 to 50 finished pipe welds a day. It must be understood from the foregoing that the slag that forms on top of the molten material in the crucible is poured into the mold first. As soon as this slag strikes the cold pipe and inner surface of the mold it forms a protective coating which prevents the superheated liquid steel Avhich flows in after it from coming in direct contact with either the pipe or the mold. The heat of the entire mass, however, serves to bring the pipe ends to the desired temperature. The 328 GAS TORCH AND THERMIT WELDING method of pouring will be undei-stood from Fig. 5. In this cut A shows the slag flowing into the mold and coating the pipe and inside of the mold; B shows the slag in the mold and the steel following, displacing the slag in the bottom part Fig. 5.— Pouring Slag an.l Steel into, the Mold. Fig. 6.— Mold for Welding Vertical Pipe. of the mold, and C shows the mold abont half full of steel, but a film of slag separating it from the pipe and mold. The foregoing instructions apply to welding pipe in a horizontal position. Vertical pipe, however, can be welded in MAKING PLASTIC PROCESS WELDS 329 essentially the same manner, bnt a special mold is required. This mold is constructed in such a way as to divide the Thermit-steel collar into two parts, so that it may be readily removed from the pipe on the completion of the weld. The mold also has a different type of pouring gate, as will be seen in Fig. 6. The same method used for welding pipe may be used for welding bars or rods of mild steel or wrought iron of various sizes and shapes, but this method is not applicable to welding cast iron or high-carbon steel. For the latter work another method must be used. Cost and Strength of Pipe Welds. — It may be of interest to compare the cost of a Thermit-welded joint with that of mechanically coupled pipe, and for this the reader is referred to Table I, which was taken from ' ' Reactions, ' ' which is the house organ of the Metal and Thermit Corporation. While accurate determination of the cost that will cover general practice is always difficult there is no doubt about the superiority from every standpoint of the Thermit-welded joint where a solid, leak-proof and especially strong coupling is wanted. For ammonia-pipe or similar installations the solid joint is one that every real engineer will recommend. As to the comparative strength of a Thermit-welded pipe joint the following is taken from a report made by Prof. Frederick L. Pryor of Stevens Institute July 10, 1914: "Two classes of pipe, standard weight and extra heavy, were tested, the sizes selected for each type being 1 in. and 1^ in. Six samples each of 1-in. standard, 1-in. extra heavy, ll-in. standard and 1^-in. extra heavy were selected for test, three being used for the bursting test and three for the tensile test. One specimen was left in its original form and two specimens were cut in half and joined together, one with a Thermit weld and one with standard-threaded couplings. One specimen of each size and type was subjected to tensile and bursting tests. Standard-Aveight couplings were used for standard pipe and extra heavy couplings for extra heavy pipe, and the welded speci- mens were put together by your process. The thickness of the material at the weld was afterward determined and found to be about 0.02 in. more than the thickness of the pipe for the 1-in. specimens, and about 0.075 in. more than the thickness of the pipe for the 1^-in. specimens. A number of pieces of pipe were measured to check the thickness with the accepted standard and eacli specimen was within the tolerance factor. "The tension tests were made in the usual manner, except that 330 GAS TORCH AND THERMIT WELDING a plug was inserted in eacli end of tlie pipe in order to assist tlie gripping action in the niaclaine. o S K H w ?r w < r ) H lO rn CO H K ^ n fe ffi H H <-i c M k'. W ■ W ' o od r^ r«^ o:> 1— ( bn O n fi O O Eh K <; •J H O" <; F-i OOOOrHTHC^eOTjl j;? u aj ^ ! 0*i OOOOOOOOO OOOOO'-iCNfO'* T-JCSCOOOOl-OtOiO ooooooo — -- u Hro.^^iO •M i-H fS (M fO PO ■"^ "The bursting test was made by pumping oil into the pipe under pressure, the actual pressure to force the plug into the cylinder being MAKINCi PLAyTIC PROCESS WELDS 331 deteruiiiied by the dimensions of tlie presstiiv plu^^ and the weight on tlie scale l)eain. "In tlie tension tests all the piin-s joined l).v couiilings ruptured in the root of the thread at the coupling, and the welded samples ruptured away from the weld with one exception. "In the bursting tests all samples, including those put together by couplings and welds, ruptured in the seam of pipe. The location of the ruiitures for both the tension and bursting tests are noted in Table IT." Table II. — ItEsui/rs oi-' Tension and Bursting Tests on Sckew-Coltpled AND Thermit-Welded Pipe Joints. DIMENSIONS IN INCHES Outside Diameter 1" Standard 1.315 IM" standard 1.660 I" extra, heavy 1 .315 l}4" extra heavy 1 .660 Inside )iameter 1.049 Thickness .133 1.380 .140 .951 .182 1.272 .194 Ulti- Yield mate Character Point, Tensile of Pipe (Actual), Strength, Lbs. (Actual), Lbs. 1" Standard Straight. . . 18,000 27,900 Coupling. . 17,250 20,320 Welded... . 16,500 26,130 Approximate Location of Rupture Burst- ing Pres- Approximate Loca- tion of Rupture sure, Lbs. Sq. In. Between grips 1 1 , 580 Center of pipe Root of thread at coupling 9 , 260 6" from coupling lYi" from weld 10, 560 4" from weld V Extra Heavy Straight... 19,200 34,730 Between grips 13,510 2" from end of pipe Coupling 18,770 Root of thread at coupling 13,310 3" from end of pipe' Welded. ... 23 ,jOOO 34,970 8" from weld 14,220 2" from end of pipe V/i" Standard Straight... 22,300 37,670 Between grips 9,440 7" from end of pipe Coupling... 21,380 28,500 Root of thread at coupling 8,050 7" from end of pipe Welded. ... 20,930 36,020 At weld 8,460 13/^" from end of pipe 1 J4" Extra Heavy Straight ... 29 , 380 Coupling... 29,100 Welded... . 27,800 50,620 Between grips 12,900 U 2" from end of pipe 29,100 Root of thread at coupling 12,770 2" from end of pipe 50,980^ 6" from weld 11,490 1" from end of pipe Professor Pryor also at about the same time made some vibj-atoi-y tests. The pipe selected \vas l^-in. extra heavy, 332 GAS TORCH AND THERMIT WELDING and each test piece was composed of 6-ft. lengths joined to- gether by the coupling or weld. The Thermit welds were made in the presence of Professor Pryor by a representative of the Thermit comjDany and were regular standard Thermit joints. These 12-ft. pieces with a joint in the center were subjected to a vibratory motion, the pipe depressed and raised 2 in. below and above the center line. The test pipe was filled with water under 22 lb. pressure in order to show the first failure of the material. Two tests were made of the pipe joined with extra heavy screwed coupling and one test of the welded joint pipe. The reason only one test was made on the Thermit-welded pipe was that the number of deflections on it was about 250 times the deflections made on the screwed coupling, with no sign of any deleterious effect. Both the screwed- joint specimens l)roke just outside the coupling in the root of the thread under 6160 and 3430 vibrations respectively. The Thermit specimen was vibrated 1,566,340 times, after which it was removed from the test and at the time no injury was apparent. The speed of these vibrations was about 225 vibrations per minute. CHAPTER III FUSION WELDING OF HEAVY SECTIONS The method, of welding heavy sections and castings, or the fusion method, differs considerably from that used for welding pipe joints. For one thing the parts to be welded must be preheated to a red heat and also another type of Figs. 7 and 9. — Sectional View of Thermit Automatic Crucible and Method of Lining It Fui. 7. — AA, magnesia stone; BB, magnesia thimble; C, refractory sand; D, metcl disk; E, asbostos washer; F, tapping pin. Fig. 9. — Method of lining a crucible — AA, magnesia stone; BB, luting of fire clay; C, cast iron crucible cone; D, layer of wrapping paper or newspaper. crucible is needed. The type of crucible used is shown in Fig. 7. This is a conical-shaped, sheet-metal receptacle, or shell, with an opening in the lower pointed end. In use this is suspended or supported above the gate of the mold by means of a tripod, bracket or other support. The metal receptacle 333 334 GAS TORCH AND THERMIT WELDING is lined witli magnesia tar, a liard-bui'nt magnesia stone being set in as shown at A. Tliis has a tiibnhir opening in it into which a small magnesia tliimble B is pressed. This thimble provides a channel through which the liquid Thermit steel is poured into the mold. The object of making the crucible so tliat it ^nay be tapped from the bottom is to prevent the slag from entering the mold, which is directly opposite to the procedure for welding pipe. The hole in the bottom of the crucible is closed previous to putting in the Thermit mixture by means of the tapping pin F, the asbestos washer E, the metal disk D and the refractory sand C. This sand is put up in small bags for tlie purpose by the company selling the mixture. When everything is ready the Thermit is put into Fig. 8. — Tapping a Crucible. the crucible and ignited exactly as described for pipe wielding. After the reaction the tapping pin is pushed up as shown in Fig. 8 and the molten steel allowed to run out into the mold. The crucible and the thimble tlirough which the metal runs after the reaction are two of the most impoi-tant factors in the whole process. The high temperature, together with the violent ebullition of the molten metal during the reaction, necessitates a lining that is not only mechanically strong but of a very high refractory substance. It has been found that magnesia-lined crucibles are the only ones that satisfy these conditions. Life of Lining- Prolong-ed by Patching- With Magnesia Tar. — As refractory as this material is, however, the crucibles that are used to any extent must be I'elined. Sometimes the life of a lining may be prolonged by patching with magnesia FUSION WELDING OF HEAVY SECTIONS 335 tar where needed and then baking it. While any good mechanic can scheme out ways to line a crucible in an emergency and may use fire clay on occasion the following method is given as the best way : The magnesia-tar lining material should be heated until it becomes plastic. A few handfuls are then placed in the bottom of the crucible shell and a magnesia stone imbedded in this material, as shown in Fig. 9, and centered over the hole. More magnesia tar is then rammed around the stone to hold it firmly in place. The cast-iron crucible cone should then be placed in position Avith the small projection set into the hole in the magnesia stone. The upper part is next centered in the shell by means of wedges inserted at equal distances along the circumference. The magnesia tar is then rammed into the space between the cone and the shell a little at a time and tamped hard. On the density or hardness of the lining depends the life of the crucible. Special iron tamping tools with flat ends should be used. The rammer should be pounded well with a good-sized hammer when ramming in the lining. Better still is a pneumatic bench rammer. Do not put in the material too rapidly, and let it be remem- bered that the better and more uniform the tamping the longer the crucible will last. As the mass nears the top the wooden wedges sliould be removed as the lining already in place will hold the cone in position. The Crucible Ready for Baking. — When completely filled and tamped a mark should be made with a piece of chalk on the cone and the point opposite to it on the lining, so that wdien the cone is withdrawn it may be replaced exactly as before. Then take the cone out, exercising care not to disturb the lining, place a layer of wrapping paper or newspaper over the tar lining, then replace the cone carefully, so that the marks previously made come opposite to each other. After this put on the crucible ring and lute carefully around the top with fire clay to protect the upper part of the lining from the heat in baking. It is also well to place damp fire clay around the l)ottom of the crucible and inside of the stone for the same purpose. The crucible is now ready for baking, and for this purpose 336 GAS TORCH AND THERMIT WELDING it should be placed in a suitable oven. The heat should gradually be raised until the cast-iron cone becomes red hot and should be held at that temperature until all the fumes stop coming off from the tar, after which it can be allowed to cool gradually before removing from the oven. If the crucible is baked too long the lining will appear crumbly and the life of the crucible will be very much shortened. Baking for too short a time will leave some of the tar in the lining and cause a violent Thermit reaction. When cool the clay luting may be removed, tlie cone taken out and the crucible is ready foi' use. Thimbles. — The portion that has to withstand the most severe strain of all is the part at the bottom of the crucible, or walls of the hole through which the metal is tapped. It has to stand the wash and pressure of the weight of the moving liquid metal and slag undei' great heat. The magnesia stone which is centered in the bottom of the crucible and around which the material for lining is packed has a tapered hole in the center. The tliimbles are of the same taper as the hole in the magnesia stone and are set into the latter. When the tliimble is used up (either through enlargement of hole or by splitting) it can be knocked out and replaced with a new one, so that the full life of the crucible may be utilized. Thimbles should be wrapped with one layer of uncreased paper before being placed in position. Since various amounts of Thermit must be used for dif- ferent sized welds the crucibles used must vary accordingly, although it is possible on occasion to use more than one crucible at a time for a given melt. This is not advisable, however, unless necessary. For lining these various sized crucibles the Thermit company makes magnesia stones and thimbles in certain sizes designated by numbers. The metal cones as well as crucibles of a given capacity are also numbered. All of the ordinary sizes, as well as the amount of magnesia tar needed, are shown in Tabl(> HI. The Care of Crucibles. — We have described in detail the construction of automatic crucibles to be used in connection with Thermit welding, and it might be well to include a few words on the proper care of these crucibles. They should be very carefully handled, as the lining is FUSION WELDING OF HEAVY SECTIONS r=; J^l M a-> wT .. Shippin Weigh of Cone Pounds O o LO lO o lo lo VD o o lO lO t^ <>4 lO 1^ o o ^^ 4) M lO LO ^-H lo OO as ■<* CV) CM CM c^ CD u itside .met€ Top, ches :^ :^ :^ :^ ^ t 00 o c^ T*' ^ o ID ■ OO o •* 3 ra ^ c ^_( CM CM (M ro D ■* o CN "0 lO ITi r>j ID t^ ■V— t ^_4 «^ CM ■* ID vo t--. o >. .. --^ Capacit in Pounds of R. R Thermi ■o 00 ID "0 tn o o O o o CM ^0 t--. Tt< ^-^ OO o CM CM •<# ^Ji tN <^ Tf "-) o t^ 00 OS o ^3 •^ N 3 d d d d d d d d d d w^U 2 2; 2 2 2 2 2 2 2 2 o OS O, (U > c 338 GAS TORCH AND THERMIT WELDING apt to erack or Jail out under rough IrcalnK'ul. It is also always important that they be stored in a dry place, as the lining, being porous, will absorjj moisture, and a moist lining will cause violent Thermit reaction. After a crucible has once been used, it is not necessary to clean it of the slag adhering to the inside, as this is a very refractory material itself and can do Jiothing but help preserve the crucible if left on. At the bottom, however, in the vicinity of the stone and thimble, the slag has to be removed so as to clear the opening of the thimble or permit of an old thimble being knocked out and a new tliimble inserted. Applications of Fusion Welding. — With the construction and method of using the automatic crucible in mind we will POyRINd GATE BASIN TO HOLD SLAG PREHEATING GATE @= Facing ksFire Sand, ks Fire Clay, IzOround Fire Brick H = Loam or Mixture of ?j sliarpSand^ 'j Fire Clay ^ = Iron Plu:,/d;iib ■■ J •■ zr-if-iX .. / - li; IS ' ^k " I ••„ ,14 '11' l/i ■■ / •■ Z4'/S''^/i • /P-?4-''iBolfsivithlNut l?-l6"''i'Bolti wr^h^/iut5 75- c Plate D'fo slide in Tor t>acl Tor l?"Molc( Si End For Box for IB'Mold I Req \,^, Back End Piece B' For Box for 16" Mole* I Req Boick End Piece'B" Tor Box for l?"Mold Fig. 11. — Design and Materials Kequircd for Standard Mold Box. it cannot be obtained in the vicinity it can be ordered from the Thermit company. Tliis material should be well riddled, nnxed dry and then moistened with just enough water so that it will pack well. Ramming the Mold.— In ramming up the mold place a small amount of molding material in the box and ram with a small rammer around the edges and working toward the center, keeping the mold level, and ram hard. Too much emphasis cannot be laid on this point, for in the construction of the mold depends the safety of the entire welding operation. See FUSION WELDING OF HEAVY SECTIONS 341 to it that the material is well rammed underneath the pattern. There should be a wall of molding material at least 4 in. thick between the wax pattern and the mold box at all points, as the Thermit steel is intensely hot and ample material must be provided to hold it. A wooden gate pattern for the pre- heating opening should be set at the lowest point of the wax pattern and leading out to the front of the mold box, where an opening is provided for it. Where the sections to be welded together are of the same size this preheating gate should be set directly in the middle of the lowest part of the wax pattern so as to heat both sides of the frame equally. Sometimes, H— ■ ...... 14' -1 mi ji 7i I*: PATTERN FOR POURING GATE IT W IS PATTERN FOR HEATING GATE h- jf PATTERN FOR RISER ' Fig. 12. — Wooden Patteins for Pouring Gate, Riser and Heating Gate of Alold. These Are Large Enough for Welds Up to 5 X '^ J^"- Larger Welds Eequire Proportionately Larger Patterns. however, it is necessary to weld a light frame section to a heavier one, in which case the preheating opening should favor the heavier section, which will require a longer time to heat than the light section. AVith the preheating gate provided for, set another wooden gate pattern directly above it and one inch away from the wax pattern and have it properly shaped for the pouring gate. Drawings for these various patterns are shown in Fig. 12. Be sure that the molding material is Avell rammed around these patterns so that it will not "cut out" under the blast of the preheater. 342 GAS TORCH AND THERMIT WELDING At the highest point of the wax pattern place the riser pattern. If there is more tlian one liigh point, place a riser pattern over each, as the function of a riser is to hold a supply of steel wliich will remain liquid for a considerable period of time, and take care of all shrinkage, so that when a "pipe" is formed, due to shrinkage, this pipe will appear in the riser and not in the weld. Also the riser acts as a depository for loose sand or other foreign matter that may be washed into k Z4" dOTTOMFRAML >| Section Through Trough Fig. 13. — Method Employed in Making Welds in Inaccessible Places. it by the Thermit steel in passing through the mold and pre- vents this material from clogging in the weld. It sometimes happens that welds are made at a point where a wooden riser pattern cannot be withdrawn conveniently. In such cases a piece of jacket-ii'on pipe may l)e used and left in the mold after ramming up. The Thermit steel will flow into this open- ing and simply melt the iron i^ipe and amalgamate with it. After the mold is all rammed up, hollow out on top so as FUSION A\l':LDIiNG OF HEAVY SECTIONS 343 to form a l)asiu in Avliieh tlu- slag' may collect so as not to overrun the mold box. Then vent the mold thoroughly by making holes with a vent rod nmde from 8 to 10 gage steel wire, so that all gases in the liquid metal will have a ciiance to escape, as shown in the typical mold. This is important. Now lightly rap the gate, riser and preheating opening patterns and draw them out carefully, wiping away any loose sand that might tend to fall into the holes. A molder's slick, trowel and lifter are very useful in this connection. Then cover the various openings so that nothing will fall into them and adjust the crucible in position with the bottom about 3 in. above and directly over the center of the pouring gate. "Where this cannot be done, construct a runner, as shown in Fig. 13, to lead the steel into the pouring gate of the mold. Preheating the Mold. — The mold is now ready for preheat- ing. Set the burner of the preheater so as to point into the heating gate of the mold and about 1 in. from the opening; then apply the blast. It is best to start easily at first, as too much of a blast would tend to "cut" the mold. The wax will burn out, leaving a perfect mold the shape of the wax pattern. Keep the heat going until the mold is thoroughly dried out and the parts to be welded are brought up to a good, red, workable heat such as would be required if the frame was to be hammered. AVhile the preheating is in progress the charge of Thermit and additions should be placed in the crucible, which is first plugged in accordance with the directions previously given. It is important to put in a few handfuls of Thermit first before dumping in the rest of the charge, so as not to disturb the plugging material. Mix the Thermit charge thoroughly before putting in the crucible. No ignition powder should be added until the Thermit charge is ready to be ignited. If the Thermit charge when leveled off comes closer than 2 in. to the top of the crucible or if the crucible has to be tipped slightly it is best to build up the crucible by means of a ring. This ring should be less in diameter than top of crucible so it can set in the crucible about 1 in. It should be from 8 to 10 in. high and made from ^-in.. stock. Lute with fire clay between ring and crucible. When it is assured that the frame is at a good workable 344 GAS TORCPI AND THERMIT WELDING heat (luickly remove tlie i)relieater and dii-eet it down the riser so as to blow out any sand or dirt that may be in the mold. If the riser is difficult of access direct the burner down the pouring gate. Then plug the preheating hole with a piece of fire brick ground to fit or an iron plug inserted as shown in Fig. 10. Back this up with several shovelfuls of molding material between the mold box and steel plate provided for the purpose and then pack the sand down hard with a rammer. This will prevent any possibility of the Thermit steel running out through the preheating opening. All heating apparatus should be removed to a safe distance while the Thermit reac- tion is in progress. IGNITING THE THERMIT Place one-half teaspoonful of ignition powder on top of the Thermit in the crucible (Thermit will not ignite from the heat of the preheater and the reaction cannot be started with- out ignition powder). Ignite this with a parlor match, apply- ing the same immediately after striking, or else ignite with a red-hot iron; this often ts the easier method. It is important that ample time be allowed for the completion of the reaction and for the entire fusion of the punchings, which are mixed with the Thermit. It is best to wait at least 35 sec. before tapping the crucible. This is accomplished by knocking up the tapping pin which sets in the bottom of the crucible, using for the purpose the tapping spade or a flat piece of iron l^Xi in. by 4 ft. Hold up the expansion on the parts with a jack or pre- heater until the metal in the weld has set and shrinkage commences to set in; then remove the jack or shut off the heat. This should be usually done about two or three hours after the weld is made, but depends largely on the size of the section and length of preheating. The mold should be allowed to remain in place as long as possible, preferably over night, so as to anneal the steel in the weld. In no case should it be disturbed for at least six hours after pouring. After removing the mold, drill through the metal left in the riser and pouring gate and knock these sections off, or else cut them off with an oxy-acetylene torch. FUSION WELDING OF HEAVY SECTIONS 345 Amount of Thermit Needed for Welds. — Tlie amount of Tliermit needed for welding sections of different sizes can be derived from Table IV, which contains the proper proportions of manganese, nickel and pimchings. These amounts are given on the supposition that the Thermit collar or reinforcement is made in accordance with the dimensions published in the table. Table IV. — Welding Portions for Welding Rectangular Sections. Width of Depth of Width of Thickness of Quantity of . Section Section Thermit Thermit Steel Railroad Thermit Steel Collar Collar at Center Required for Weld Inches Inches Inches Inches Pounds 3 2 4 1 40 3 2J^ 4 1 40 3 3 4 1 45 3 ^Vz 4 1 50 3 4 4 1 55 4 4 4 1 65 4 ^Vz 4 1 65 4 5 4 1 70 4 5J^ 5 1^ 75 4 6 5 iVi 75 4}^ Wi 5 IM 70 W2 5 5 IM 75 4H 5H 5 IM 75 43^ 6 5 1% 80 5 5 5 IH 75 5 51^ 5 IH 80 5 6 6 Wz 85 5 7 6 iVz 90 5H ^Vz 6 W2 85 5H 6 6 VA 90 5J^ 7 6 1^ 110 6 6 6 IM 100 6 6J^ 6 VA 120 6 7 7 1% 130 6H 6M 7 IY» 130 6J^ 7 7 1% 150 ^Vi 8 7 ik 160 7 7 7 IVs 155 It is better practice, however, to calculate the amount of Thermit needed for a weld from the weight of wax used in the pattern and it is advisable anyway to make this calculation as a check. "Where the quantity of Thermit is calculated from the wax great care should be taken to see that the entire space which is to be filled with Thermit steel is filled with wax so that not only the collar but the space between the sections is filled with wax. Then, by weighing the wax before and after the 346 GAS TORCH AND THERMIT WELDING completion of this operation, the difference will be the quantity of wax used, and this weight in pounds multiplied by 30 will give the proper amount of railroad Thermit for the Aveld. It is recommended that railroad Thermit be used in all cases, as it is ready mixed with 1 per cent pure manganese, I per cent nickel shot and 15 per cent mild-steel punchings. This has been found to give the best results for welding wrought iron and steel. For convenience railroad Thermit is supplied in waterproof paper bags holding 29| lb. of the mix- ture, so that one bag to the pound of wax is sufficient for a weld. This rule provides ample Thermit steel not only for the weld proper, but also for the pouring gate and riser. In case the user has only plain Thermit on hand he should then allow 25 lb. of plain Thermit to the pound of wax, and should mix with this amount of plain Thermit 1 per cent pure manganese, f per cent nickel shot and 15 per cent mild-steel punchings. In other words, to every 100 lb. of plain Thermit add 1 lb. pure manganese, 10 oz. nickel shot and 15 lb. mild- steel punchings. These punchings must be clean and free from grease or dirt of any kind and not more than f in. in diameter by -J in. thick. These rules apply only to welds requiring less than 300 11). of Tliermit. For welds requiring more than 300 lb. of Thermit the usual mixture takes 20 per cent mild-steel punchings with the other additions the same. If railroad Thermit is used add 3j lb. of punchings to each bag. In special eases it is some- times advisable to make up a special mixture in order to produce a Thermit steel of essentially the same analysis as the steel in the parts to be welded. Where the amount of Thermit calculated comes to ten bags or more, one of these bags may be dispensed with ; that is, instead of using ten bags use nine, or instead of using 20 bags use 18, and so on, as the smaller percentage of metal required for gates and risers makes it unnecessary to use so much of the mixed Thermit. Where it is desired to calculate in advance the amount of Thermit required for a w^'ld it is first necessary to estimate the number of cubic inches in the space to be filled with Thermit steel, i.e., the space between the ends of the sections to be welded together and the cubical contents of the Thermit-steel collar or reinforcement fused around the Aveld. Allow ^ lb, FUSION WELDING OF HEAVY SECTIONS 347 of railroad Thermit to the cubic inch, and this will be sufficient not only for the weld proper but will provide ample metal for pouring gate and riser. In estimating the cubical contents of the collar the simplest method is to multiply the width by the greatest thickness (i.e., the thickness at the middle part) ; then multiply this product by 0.7. This will give the average area of the cross-section of the collar. If this is then multiplied l)y the total length of the collar around the outside of the frame and if all measurements are taken in inches the result will be the number of cubic inches in the collar. },:zoheD.7 J ■ZorieVA; m ^^orie^m ■Jac/rwrD> T "^^g y^ idackforD,> jZorieEEr tjack here or here to Line up Fig. 14. — Method of Preventing Unequal Stresses When Welding Locomo- tive Frames Broken at Various Points. Fracture Location Remarks Zone A* Heat zone B to get -^iii in. expansion and hold 2 to 3 hours after welding. Preheater or basket fire.* Zone B* Heat zCfte A to get ?io in. expansion and hold 2 to 3 hours after welding. Preheater or basket fire.* Zone C, C, or Cj Jack ?i6 in. at C, C, or C,; keep jack in place 2 to 3 hours after welding and then remove entirely. Zone D or D, Jack ?i6 in. at D or D,; keep jack in place 2 to 3 hours after welding and then remove entirely. Zone E Cut out unfractured member of splice to clear collar. * When heating either Zone A or Zone B the adjiieent pedestal brace or braces should be ])ut in place before commencing to heat so as to distribute the expansion and not upset or distort the leg. Locomotive Frame Work. — The foregoing directions refer to the general run of wrought-iron or steel repairs, but with only slight variations the same method is followed for locomo- tive-frame work. Tlie principal difference is in placing the mold or allowing for contraction in various members and not in the use of the Thermit itself. In order to make it clear where stresses are liable to be set up in a locomotive frame the diagram shoAvn in Fig. 14 has been made. T>y a careful study of this and the application of the principles illustrated a welder should be able to figure out his work so as to produce satisfactory results. 348 GAS TORCH AND THERMIT WELDING The illustrations will be of assistance in planning the work on various parts of a locomotive frame. Fig. 15 shows how to place the mold, and jacks for welding a broken frame leg. With the pouring gate and risers as indicated they permit of POURING GATE Fig. 15. — Method Employed in Welding Locomotive Frame Broken in Leg. 'O , , „ P0UPIN6 A ■^ 3"h- (<■ 4" >l Section Through ThermitCollar This Section applies to Frames up fo4'S'ifor Dimensions of Collars on LarqerFrames see Table HEATING GATE Section on A- B Frame Ready for Mold JACK HERE OR HERE TO LINE UP Fig. l(i. — Method p]mployed in Welding Locomotive Frame Broken in .Liw. a good wasliing action for the Thermit steel, so that any slag or sand tliat might be in the mold will be carried into the risers. Fig. 16 shows how to weld a frame broken in the jaw. FUSION WELDING OF HEA\'Y SECTIONS 349 Fig. 17 sliOAvs how to wold a frame broken in the splice. In this it is best in making the repair not only to weld the broken sections together, bnt also to cut ont a piece abont 1X5 in. of the unbroken member, so the Thermit will flow entirely around the broken sections. By making the repair in this manner a good, strong job is assured, and if the bolt hole is welded up and the two members welded together future breakage at these particular points is practically eliminated. POURIHG GATE Fig. 17. -Method of Welding Frames Broken in Splice. Lower Drawing Shows How to Drill and Cut the Unbroken Member. The only objection that can be raised against this practice is the trouble of separating the members in case the splice is to be removed or in order to take out or renew a cylinder. This objection, however, is not serious because it is only neces- sary to drill a line of small holes where the parts are welded together and the member can then be removed. When replac- ing it is best to cut a keyway where the frame is cut out and then bolt together in the same way as when the frames were originally assembled. 350 c;as ix)Rcn and thermit welding Fig. 18 shows liow to weld locoiuolivc luiul I'ings without cutting' tlie slu'ets. This nietliod has proved cntii-cly satisfae- Tm-!£'''^i RISERS k NOT T0PqrMUDR/N6 -OPENED UP \ M ^HEATING GATE Section Showing Weld,Gates and Risers ^Tdiam holes,i"deep drilled into mudrin5 Side Elevoition Showing Weld Section Showing Prepared Mold in 01 Stcindcird Mold Box Fig. is. — Mold for Welding Mud Eings Without Cutting Sheets. Fig. 19.— Thermit Wehl on Mud RinP-. tory and many such welds liavo lioen completed and are givinp: good service. Typical Welds. — Fig. 19 is that of a finished mud-ring weld, in Avhich tlie sheets ai'e not cut. FUSION WELDING OF HEAVY SECTIONS 351 Fig. 20. — Fracture in Crosshead Cut Out for Welding. Fig. 21. — Weld Louipleted and Crosshead in Service. 352 GAS TORCH AND THERMIT WELDING Fig. 20 shows a fracture in a crosshoad cut out for welding. Fiff. 21 shows the weld completed and the part in service. Fig. 2-^. — Weld on Broken lioeker (Shaft Before Machinin^f Fig. 22 sliows a weld on a broken locomotive rocker shaft before machining. '•^s^ii*!!!'*' Fig. 2;?. — Repair on Broken Guide Yoke. Fig. 23 shows a repair on a broken guide yoke. Fig. 24 illustrates two welds in an engine splice. FUSION WELDING OF HEAVY SECTIONS 353 Fig. 25 is a repaired driving-wheel center. Fig. 26 shows details of a crucible holder for frame welds. No attempt has hccii made to make the list of ]-epairs on locomotive parts complete, but enough has been shown to serve as a guide for practically everything that is apt to confr-ont Fig. 24. — Two "Welds in Splice of "Frame. the practical man. For superheater work, or pipe work of any kind, the directions given luider the heading of pipe weld- ing will cover all that is necessary. As a sort of recapitulation of the foregoing directions, it will be well to keep the following "don'ts" in mind when getting ready for all locomotive Thermit-welding work. 354 GAS TORCH AND THERMIT WELDING Don't keep your material and appliances in a damp place. Better store them all in a good, dry room under lock and key, the foreman in charge of the Thermit work to have the key. Better still construct a tool wagon and keep all Thermit material in it. Don't start to make a Thermit weld unless you have all the necessary materials and appliances and the latter in good condition. Don't neglect to clean the frame tlioronghlv- Be sure to Fig. 25. — Weld on Driving- Wheoi Center. remove all the grease, paint, etc., and have as good clean metal as possible to work on. Don't neglect to take care of the contraction that is bound to be set up as the metal in the weld cools. If this cannot be allowed for by spreading the sections with a jack or other mechanical means heat the opposite unl)roken member with the other burner of a double-burner preheater fitted with a flaming-burner attachment. If this is not available hang a FUSION WELDING OF HEAVY SECTIONS 355 basket fire of charcoal or coke al)oiit the unbroken member and heat until proper expansion is obtained, holding up the heat for two or three hours after the weld is poured. It is advisable to expand the frame ^/m in. on the average. Cut out the frame along the fracture so as to make a clean opening 1 in. wide. Proportion your wax pattern to the size of the frames as shown in the table. O / ' loles; from eactiEnci- ZP AT TERNS MATERIALS^; ,^ , „ One Steel Bar - 4 ''^"-^ ij^g" (>j6' lon^ Three Set Screws 3'^ -i 6Wanted Finish all over Fig. 27. — Dcsiau of V-BIocks for Weldina" Crankshafts. ever, the fact remains that the parts of the shafts arc strongly forced apart so as to slightly tilt the V-blocks and raise the shaft out of line. Tliis may explain why the inaccuracy in alignment is not in the direction of the contraction of the weld but almost at right angles to it. In a great many cases tliis tendency to separate will shift the parts of the shaft horizontally as much as \ in. and as the V-blocks resist tliis they are tilted by the force and the shaft thrown out of line. It may be hours before any considerable contraction sets in, 362 GAS TORCH AND THERMIT WELDING and by this time the shaft has been permanently set out of line. Heating the opposite slab will slightly connteract this, but not sufficiently, because the heat conducted from tlie Thermit steel will expand one slab a great deal more than any possible preheating on the opposite slab. Crankshafts that are broken in such a way that they can be lined up with the throws in a vertical position will be almost as far out of line because the sudden expansion of the adjacent parts will have to shift part of the shaft and even sometimes lift it partly out of the V-blocks, and this force is being exerted througli molten or perhaps plastic metal so that a certain amount of upsetting will naturally take place. V-Blocks for Holding Shafts. — In order to overcome these important defects the special V-blocks shown in Fig. 27 will allow a horizontal motion after the mold is rammed. If then the proper allowance for contraction is made the shaft should come back into line because the force tending to separate tlie fracture will not be resisted and will be subsequently offset by an equal contraction. On the other hand such V-blocks will permit of watching the contractions of the shaft so that different allowances can be made on the next shaft if necessary. These V-blocks sliould be made in such a way that they will be divided in two parts horizontally. The upper and lower parts should each have divisions, accurately marked on them next to the dividing line, tlie central division being longer and heavier than the rest. When the two parts of the V-blocks are central on each other, accurately turned pins, preferably tapered, may be inserted in reamed holes passing through the two lugs so as to securely fasten them together. This locates accurately the central position where the shaft is to be lined up "in line." Where a horizontal contraction is to be allowed for, the pin should be left out of certain V-blocks and the parts of these V-blocks slightly shifted on each other if neces- sary. If the V-block pins are not in place the holding down bolts can be relied upon to hold the shaft in a desired position during the ramming of the mold. When the preheating is started these bolts should of course be removed and the shaft allowed to move freely. x\nother advantage of this type of V-block is that flat shims can be placed between the halves of the V-blocks to allow for different journal diameters instead WELDING CRANKSHAFTS, MILL PINION TEETH, ETC. 363 of placing the shims on ilic slanting face of the V-blocks. The thickness of the sliims will of course be just half the difference in the diameters of the journals. In allowing for contraction of a Thermit weld it must be remembered that the actual contraction of the small amount of Thermit steel in the space between the pieces is almost negligible, whereas the actual contraction of the weld may Pig. 2s. — Two-Throw Crankshaft — Fracture Cut Away for Welding. Fig. 29,— Two-Throw Crankshaft Welded— Repair Made in 72 Hours. vary from \/,,. to V^ in. This is due to the fact that during the preheating operation the ends of the pieces at the fracture expand or approach each other by the amount of the expansion of the adjacent parts by the preheating. For instance, if the fracture is opened up I in. to allow for the contraction and the expansion of the parts during the preheating approach each other almost ^/\ in (perhaps V'ei in. less) the parts should 364 GAS TORCH AND THERMIT WELDING bo almost exactly in line after welding. In welding large sec- tions slightly greater allowances for conti-action should be made than in smaller ones, because to bring the fracture to tlie proper heat takes a longer time and consequently the heat "soaks" further along the parts, causing a greater expansion and a greater tendency to close up the distance between the fractures. A large two-throw crankshaft previous to welding is shown Fig. oU. — Fracture iu Web (Jut Away for Wekliug a Crankshaft. Fig. 31.— Welded in 6i In. Crankshaft Broken in the Web, in Fig. 28. This same crankshaft after welding is shown in Fig. 29. A 6^ in. crankshaft broken in the web is shown in Fig. 30 and the finished weld in Fig. 31. How to Locate Minute Cracks in Crankshafts or Other Parts. — In the course of welding crankshafts and other im- portant work it is often found that while the part to be welded is broken clear through there are other minute hairline cracks WELDING CRANKSHAFTS, MILL PINION TEETH, ETC. 365 near by which are sure to give trouble later. It is probable that the strain thrown on the part when the break occurs is often sufficient to start these small cracks. They may also be caused by strains in the metal from improper treatment in the first place, and which may have been responsible for the first break. In any ease, however they may have been caused, the proper thing" to do is to locate these cracks and so weld the parts as to eliminate them. As they are many times so minute as to be invisible to the naked eye some other means must be found to locate them. A very efficient method is to paint the entire section with a mixture of whiting and alcohol. The whit- ing and alcohol should be mixed so as to form a good white paint, but not too thin. This dries quickly and becomes dis- colored by the grease or dirt in the very fine cracks, so that these cracks show up very distinctly. Since it is the oil or dirt in these cracks that causes them to show so clearly on the white paint it is not a good method to detect cracks in a new piece. The part to be painted should of course be cleaned of all the dirt and grease on the surface. It is a conservative estimate to say that probably one-third of all crankshafts will be found to contain additional cracks other than where the break is visible. If these are not found and remedied the chances are that they will develop into real breaks later. Welding- New Teeth in Large Pinions to Replace Teeth Broken Out. — The Tliermit process is coming into more and more general use in large steel works and rolling mills for welding teeth in heavy pinionS;, as it can be relied on to give a permanent, efficient and economical repair in the case of these very heavy sections. The following instructions cover a method which has been in use for several years, and if they are carefully followed a satisfactory repair is assured. Many pinions weighing up to 17 tons have been repaired in this way and are now doing service. The repairs usually consist of replacing teeth or parts of teeth which have broken out. They are peculiar in that the tooth is a comparatively small projection on an extremely heavy steel casting. For this reason, if the repair were at- tempted by the ordinary method, i.e., if the casting were pre- 366 GAS TORCH AND THERMIT WELDING heated at the wekl only as covered in previous instructions for making Thermit wekis, the heat would be carried away into the castings so quickly, especially during the interval of removing the preheating burner and tapping the crucible, that in most cases a poor weld w^ould result. Everything possible must therefore be done to conserve the heat at the weld, and to do this efficiently it is necessary that the whole pinion should be heated to a red heat. This may be done by bricking in the heaviest part and preheating it by means of oil or gas burners conveniently placed while the part to be welded is being preheated in the regular way. The Thermit company's flaming-burner preheater attachments are admirably adapted to this preheating work, as they give an extremely hot flame which may be adjusted to suit the conditions. (Vire should be taken, however, to bring up the heat slowly, as otherwise there is danger of cracking the pinion. In making all welds where a relatively small amount of Thermit steel is to be added to a heavy steel casting or where one or both of the parts to be joined is considerably heavier and larger than the Thermit steel part it is necessary to take special precautions to secure thorough amalgamation of the Thermit steel with the heavier part, especially at the extreme edges of the line of junction where in service the greatest strain will come. The slightest imperfection at this line of junction or extreme flber will cause a tear to start in service which will cause a fracture of the welded part. A perfect weld on this extreme fiber is made more difficult by the fact that the metal in the weld always shrinks a little more than the white-hot steel of the pinion due to the slight diffierence in shrinkage between molten steel and white-hot steel. It is necessary therefore that the fusion be obtained for a consider- able depth even at the extreme edge of the Thermit steel. Fusion at this point is more difficult because the heat of the Thermit steel comes from one side only and not from all sides as it does near the center of the weld. For all these reasons it is desirable to increase as far as possible the surface exposed to the Thermit steel in the width cf the weld. This at the same time produces edges or corners which melt more readily and thus aid in the fusion. These edges may be readily produced by cutting out a groove or WELDING CRANKSHAFTS, MILL PINION TEETH, ETC. 367 slot in the main body of the pinion at the center part of the root of tlie tooth broken out. This slot should be half the /"or f Pipe. I^- Length fosuif Height of MM W-3"P/pe i;^ ^ Preheater Hose ,Y Connection -£5pace between '>; Pip^s filled with § MoldingSand. ^/Plate '^CSteel Plate ^ Elevation W-Outer Pipe tu protect inner Pipe from Meat VentHoles ^ ^ X-DiometerafRoll Neck to see Roll Ad oXy- These Surfaces should NeckandTor[/ jj; ) be well coated with TEMPiET IN FORMING NEW TOOTH IN WAX Dftinq Plate V-U~^' -, / q, stronq Silica Wash chnnlH Pino \ y ■' shouldPIpe burn off DETAIL or BtrnNEH METHOD OFPREHEATINd SURFACE OF FRACTURE TYPE OF PEMOyABLE PATTERNS NECESSITATED BT PODS Fig. ^ k- Y \'Soltibr 'liftinq 'iPaftern 1 '|a i. r-'--^-'^"' K-— X---» PA TTERN FOR ROLL-NECK REPAIRING xn £E ■FLAME Preheating Cope if Necessary to use H Plcinof PartsB .'filled with fire Clay Hoolf to be used in 5and \ Case Pipe'A'should Core come loose from , Plate: when removinq Burner insert Hoolf in VentHoles folift '' out Plate PLUa FOR OVERFLOW OATE Two wanted J Bolt ■Pipe 32. — Designs for Patterns and Heating Apparatus for Eepairinj Steel Pinions. width of the tooth in depth and also in Avidth, i.e., if the tooth to be welded in is 6 in. wide at its root the slot should be made 3 in. wide by 3 in. deep. The most economical way to 368 GAS TORCH AND THERMIT WELDING cut tliis slot is to place the pinion on a planer and machine it out. The cutting of such a slot also serves to bring the line of junction between the Thermit steel and the metal of the pinion well into the body of the pinion so that a strong and efficient weld is assured. After the slot has been cut, the pinion in the vicinity of the weld should be carefully cleaned and then mounted vertically for the welding operation. In this mounting great care should be taken that the pinion is properly supported so that there will be no danger of its settling under the added weiglit of the mold box. This can be accomplished in the following manner: First dig a hole in the ground the proper size to receive the neck of the pinion. Then lay two T-rails across the top of the hole so that they will come underneath the shoulder of the pinion. If the ground is not sufficiently hard to properly support the T-rails steel plates can be placed underneath in order to prevent the rails from settling into the ground. Making- the Wax Tooth Pattern. — With the pinion properly supported in this manner the next step is to provide the wax pattern for the new tooth. This can best be done by con- structing a rough wooden box a little larger than the tooth in question. Place this against the pinion where the new tooth is to be added and lute around the edge of the box with fire clay. Next fill this box completely with molten wax. When the wax has set remove the box and shape to proper form by means of a templet as shown in A, Fig. 32. This templet should be made from J^-in. steel plate and the outline of the teeth cut into it by using three good teeth in the pinion as a guide. The center tooth, however, which will be the guide for the tooth to be welded in, should be cut V32 ii^- larger all around so as to allow for the contraction of the Thermit steel tooth. The two outside teeth of the templet engage with the teeth on each side of the wax pattern, and therefore wlien this templet is moved up and down it will cut the wax to proper shape and also assure that the new tooth is welded on in proper pitch. WELDING CRANKSHAFTS, MILL PINION TEETH, ETC. 369 ANOTHER METHOD One disadvantage of this method of making the wax core is that if the adjacent teeth arc considerably worn the new tooth will not conform to their shape unless the templet is juggled considerably when shaping the wax pattern. A newer method has recently been developed by F. N. Keithley and used with success. This method gives a east tooth of the same Tooth broken out here Scrape offl^ from Sand C Wax cut up info small Cubes Board over Shroud Rin^ Clay Luting Height of Tooth > Length of doard'AlessI/"^ gj^^^/ BOARD USED AS BOTTOM OF SAND CORE ^- 1^ J Fig. 33. — Recently Developed Method of Making Wax Tooth Pattern. approximate shape as the others in the pinion, even if con- siderably worn, which is an obvious advantage. Referring to Fig. 33 the broken tooth is slotted out as in the previous method and the adjacent teeth are cleaned and scraped. AVith the pinion in a horizontal position Avooden strips are fitted to the bottoms of the tooth spaces, as shown at the left in C. Lag screws are screwed into these for handling purposes. Further details of the strips are shown at E. A 370 GAS TORCH AND THERMIT WELDING mixture of two pai'ts building sand to one of tire clay is sifted through a No. 4 mesh riddk^ and moistened a little more than for ramming a mold. If this mixture does not draw well more fire clay may be added. The mixture is pressed between the model teeth on top of the board strips, as shown at the right in C. The mixture is rammed in firmly to a point f in. above the top of the tooth on the side for the wax pattern, as in- dicated at C and at F. The idea is to provide sufficient height of wax to allow for shrinkage. After the two parts are rammed they are lifted out and laid carefully on a board. One-fourth of an inch of material is then carefully scraped off of the side of each piece that does not come in contact with the wax, and the surfaces are slicked. This is to allow for shrinkage of both wax and Thermit steel. The two pieces are now placed in position as shown at D. Weights should be placed partly on the pieces and partly on the adjacent teeth to hold the pieces in place. The ends are then luted with fire clay and the space filled with small pieces of wax. The melted wax is then poured in, taking care not to have it too hot, as it will eat into the sand if it is. The mold parts in position and the wax poured are shown at D. If the pinion is shrouded the wax pattern for the shroud can be put on at the same time that the wax tooth is formed. It is only necessary to roll a clay rod about 1 in. in diameter and lay it against the pinion 3 in. away all around from the space cut in the shroud. Back this up with a board large enough to extend above the top of the tooth and lute as in- dicated at B. When the wax pattern is finished the mold box should be placed in position and securely clamped to the pinion, the clamps to be in a position so as not to come in contact with the fire when the pinion is being preheated. This mold box should be wide enough to take in two teeth on each side of the tooth to be welded. Now ram up the mold box, allowing for a preheating gate, a pouring gate and a riser in accordance with instructions already given. When this is completed construct a brick furnace around the exposed part of the pinion and about 2 in. away from the teeth. Next place a sheet-iron casing around the exposed neck on top. WELDING CRANKSHAFTS, MILL PINION TEETH, ETC. 371 This easing slionld bo 6 in. larger in diameter than the neck and about 4: in. higher. Now ram sand between the casing and the neck and cover the top with a layer of sand 4 in. thick. In this way the entire pinion is insulated. Preheating. — The next step is the preheating. Place a burner at the bottom of the brick furnace as shown in Fig. 34, and start with a very mild heat. This is to avoid lieating ElG. 34. — Mold Box, Brick Furnace and Crucible in Position and Pinion Being Preheated. the pinion too quickly, thus causing internal strains which might result in cracking the pinion. After the pinion has been thoroughly soaked with heat the fire can be increased to a good sharp heat so as to bring the entire pinion to a good blood red or about 1200 deg. Fahrenheit. While the heating is in progress, as shown in the rear view, Fig. 35, place an automatic crucible of the proper size to hold the Thermit charge in position over the pouring gate and 372 GAS TORCH AND THERMIT WELDING charge with tho wokling portion of Thoivmit. Tn case of very- large welds it is sometimes necessary to use two crucibles and provide two pouring gates in the mold. Repairs of this kind usually require anywhere from 350 to over 1000 lb. of Thermit. -O, Fig. -Rear View fShowiug i'relicating of Body of Piuion in Brick Furnace. In special cases it is advisable to make up a special steel mixture of essentially the same analysis as that of the pinion. Continue heating in the brick furnace initil tlie Thermit steel has cooled to about the same temperature as the body WELDING CRANKSHAFTS, MILL PINION TEETH, ETC. 373 of tlic pinion, then remove tlie burner from the furnace, take oft' a few of the top bricks and fill in betAveen the bricks and the pinion with dry sand, thereby protecting the pinion com- pletely from the air currents. Fig. 36. — Finished Weld, Showing Metal in Pouring Gate and Eiser. The pinion should be allowed to cool slowly in this mold for at least six or seven days so as to thoroughly anneal the metal in the entire piece. The mold can then be dismantled, the weld trimmed and the pinion will be ready for service. A pinion previous to trimming is shown in Fig. 36. CHAPTER V WELDING NEW NECKS ON LARGE STEEL PINIONS AND OTHER HEAVY WORK Frequent breakag'es of heavy pinions in steel plants have resulted in the development of a very ingenious adaptation of the Thermit process for their repair. Obviously the casting on of a new neck entirely out of Tliermit steel would be a very expensive operation and it would also be costly and diffi- cult to turn up a new piece of steel and weld it on to the original section, as the weld would be a very large one to make. Experience has shown, however, that the intense heat of the reaction can be utilized for the purpose of bringing the broken surface of the pinion to a fusing temperature, at which time a supply of liquid steel can be poured in from the ladle, and this will unite with the original body of the pinion to form a new neck thorouglily amalgamated with the rest of the piece. Briefly the operation consists in constructing a mold around the broken section so as to permit of casting on a new neck to replace the one broken off. The original section is then preheated to red heat by means of gasoline or oil burners, after which Thermit steel from a crucil)le is allowed to flow over the fractured surface to a depth of 1 in. This completes the heating operation and brings the surface of the roll to the melting point. A supply of liquid steel from a ladle is then tapped into the mold and allowed to wash through and overflow into an ingot mold so as not to be wasted. The overflow gate may then be closed and the mold filled to the top with steel. Detailed instructions for these various opera- tions follow, but it is recommended that if the process is to be used for the first time for such repairs an experienced engineer should be obtained to supervise the first welds and give personal instructions for executing this class of work. 374 WELDING NEW NECKS ON LARGE STEEL PINIONS 375 Tlic iiistruetions given here liavc been written more especially for the purpose of acting as a guide for a reference, and while we hope that there are sufficiently adequate and complete to enable anybody to make these welds, the personal supervision and instructions of an experienced engineer are much to be preferred. Two Methods of Working. — We give two methods for executing these repairs. The first method which follows is undoubtedly the safest and surest method to use, but it involves considerably more trouble and expense than the second method. We can recommend it strongly, liowever, and believe it would be to the interests of steel plants having much of this work to do to equip themselves properly to follow out this, method. Pig. 37. — Sawing Off End of Neck Previons to Welding on a New One. Before imdertaking a pinion repair the broken end should be cut off square, as shown in Fig. 37, so as to form a level surface when the roll stands in a vertical position. The object of this is to permit of a uniform covering of Thermit steel over the entire surface to be welded. If the break is in the pods, cut off 2 in. below the point where the pod joins the neck. If when the neck is cut off it should be found to contain any pipes or cavities these should.be bored out and steel plugs turned to a driving fit and driven into the cavities at least 5 in., care being taken that the plugs are driven in even with the surface on the end of the neck. Another and better method is to dry out the inside of the cavity by heating and then fill with liquid steel. This will eliminate any danger of the Thermit metal melting the plug 376 GAS TORCH AND THERMIT WELDING and running into the cavity, which might cause a violent and dangerous eruption of the steel. Clean off all dirt and grease at least 20 in. from point of weld. FROM FUCLTANK FROM FUEL TANK >«'~*-, ~ "■'~"\ -_^ in^ In m li ^^^<.'^^S y ■ ■ ■ ' ' ^' " "* -^ '"' Bmi^ Fig. 41. — Eoll Neck Welded to Large Steel Roll with Pods Cast in. Treatment When a Cope is Used. — If a cope is used clean off the surface of the top of the mold and set on the cope,~ clamping it securely, and then fill the cope to the height desired I'iG. 42. — New Neck Welded to Large Steel Pinion. In this Case the Pods Were Milled Afterward. with steel and again cover over wdth dry sand or powdered charcoal. If the body of the roll or pinion has cooled to any extent it would be desirable to again preheat it. After the body 382 GAS TORCH AND THERMIT WELDING of the pinion is sufficiently preheated cover the pit to make it as nearly airtight as possible, so as to cause the roll or pinion to cool slowly. By doing this further annealing is unnecessa;'y. Fig. 43. — Worn Pods Bnilt I^^p with Thermit Steel. The Repair Consisted of Four Wehls Made Simultaneously, using Two Pouring Gates and Two Crucibles. Fig. 44. — Building Up Worn Pods by Means of Three Thermit Welds, Pouring Gates Were Connected at Top and Bottom to Insure Equal Distribution of Metal. After the pinion is sufficiently cool (usually about 48 hours or more) remove from tlie mold and machine to size. In some steel plants it is considered preferable not to use WELDING NEW NECKS ON LARGE STEEL PINIONS 383 a cope but to build the mold all in one piece to the height of the new neck. This method simplifies the making of the mold and the pouring of the weld, but sometimes complicates the operation for the following reasons: When riser patterns are withdrawn it is a little more diffi- cult to remove loose sand from the mold. In preheating it is not so easy for the operator to see what he is doing. There are times when these necks will be as much as 5 ft. high, which makes it a little unhandy to work around the mold. There are numerous arguments on both sides of the ques- tion, but we feel that either method will give good results. For short necks, however, a cope can probably be dispensed with without introducing any difficulty. Two welding jobs just as they came from the molds are shown in Figs. 41 and 42. The first is a neck welded onto a large steel roll with the pods cast in. The second one shows a new neck welded to a large steel pinion. In this last case the pods were milled out afterward. Two other welding jobs are shown in Figs. 43 and 44. These both illustrate the repair or replacing of worn pods on heavy steel mill pinions. Alternative Method.- — While the preceding directions cover the welding of pinions under what might be considered ideal conditions it is not always possible to do the work in this way, and where such is the case we M^ould recommend that the following directions be followed, as they represent a simpler method, yet one which has always resulted in satis- factory repairs. Patterns, mold box, runners, etc., should be constructed in accordance with directions given for the previous method. In these repairs great care should be exercised in supporting the pinion so that there is no danger of its settling under the added weight of the mold and the steel which is poured into it. If it is not desired to go to the expense of constructing a special i^it as outlined in the previous method a satisfactory and economical way is to dig a hole in the ground about 8 ft. in diameter and of sufficient depth to receive at least f of the entire length of the pinion. Cover the bottom of the hole by laying a double flooring of 2-in. planking, being careful that 384 GAS TORCH AND TPIERMIT WELDING I lie planks in one layer run in opposite directions to those of the other layer. On top of this place a steel plate in order to distribute the weight of the pinion over the entire floor area. Such a founda- tion has ahvays proved adequate and is not expensive. ection A-A l^ Foundatioh +o suit condition of r^-^ ground etc In all cases see that '■ mold IS supported on roll independent of -Floor Size and number ofrnold boxes to suit roll. Mo/e fo be plu(^qecl eifter '~~^' Ihermitsfeel IS washed ouf RUNNER TO INGOT MOlD •■'y//yy/////y:^^/^}/y}////y/y'^/y^ Fig. 45. — Alternate Method of yuppoitiug Roll or Pinion to i>e Iicpaired. Set the pinion in the hole so that the surface to be welded '\i level and fill in all around the pinion with dirt, ramming hard to hold the pinion permanently in position. Dig a second hole alongside of tlie buried pinion to receive the ingot mold. Tins should be at such a distance fi-oni the pinion that a suitable runner for the overflow steel can easily WELDING NEW NECKS ON LARGE STEEL PINIONS 385 be placed. Tlie top of the ingot mold should, of coiirse, be lower than the top of the roll or pinion neck. If the neck is broken off: close to the body of the pinion it is absolutely necessary to provide arrangements for pre- heating the body protruding above the ground in order to avoid shrinkage strains. A simple way to do this is to build up a brick furnace and heat in accordance with directions relating to casting of teeth in large pinions, and more of the pinion body should protrude above the ground than shown in Fig. 45. If, however, there is one foot or more of neck pro- truding from tlie body of the pinion, it is not absolutely neces- sary to preheat the rest of the j)inion. THE MOLD BOX Tlie mold box should be constructed with heavy steel flasks or a substitute made of at least ^J-in. plate and should be supported entirely on the roll and independent of the ground. Fig. 46. — Finished Weld on Anchor Davit of U. S. S. " Ohnnpia. " With the mold box adjusted in place ram up witli good molding material in accordance with the previous directions and then draw out the various wooden patterns. Nails or chaplets should be driven into the sand so as to hold it firmly in place. It is advisable to coat the mold Avith a good steel wash. AVhen the mold is completed the preheating of the body of the pinion should be started if the repair is of such a nature r,s to require this preheating. If this heating is not 386 GAS TORCH AND THERMIT WELDING necessary start preheating on top of the neck as shown in Fig. 32-B. If the body of the pinion is heated the heating of the neck shouki not be started until tlie pinion approaches a red heat. Fig. 47. — Anchoi- of the Morgan Yacht "Corsair''' Rciiaivcd with Tlicrniit. Fig. 4N. — Weld jn VVlu-ol Nhat'f Of steamer "Nashville" on the Cumbcrlanil River, Made in 1912. Heat the surface on top of the neck to a good red heat, timing the operation so that it will be red hot at tlie time the openhearth steel is tapped out. WELDINCx NEW NECKS ON LARGE STEEL PINIONS 387 Fig. 49. — Weld on lO-in. Wheel Shaft of Steamer ' ' Osceola, Made at Jacksonville, Fla., in 1915. Fig. 50.— Weld on Sternpost of Tug No. 32, Made September, 1911. 388 GAS TORCH AND THERMIT WELDING AVhilo tlic prolicating of the neck is progressing set an automatic crucible, size 7, as shown in Fig. 45, and charge it so that it will be ready when needed. If possible it is best to support the crucible with a crane so that it can be quickly removed after it has been tapped. The procedure is then the same as described for the previously given method. After cooling, strip the mold and machine the parts to proper size. It is sometimes desirable where tlie l)ody of the pinion has been preheated to continue this heating after the weld Fig. 51. — Stcrnpost Weld on the "William Henry Mack," July, 1912. is completed. This can easily be done by again igniting the burners directed into the brick furnace. After the body is sufficiently preheated remove the burners and fill in between the bricks and the pinion with dry sand so as to cause slow cooling. By doing this further annealing is unnecessary. As previously mentioned, a cope may or may not be desirable, and this is left to the judgment of the operator. Marine Work. — The general principles to be followed in making marine repairs are the same as for any other repairs of a similar size and nature, so no detailed description need WELDING NEW NECKS ON LARGE STEEL PINIONS 389 be given. Ilowever, it will be of interest to know the exact natui'e of some of the more common repairs made on anchors, \\ heel shafts, sternposts or the like, so a few views of some of the actual repairs are shown. A big point in favor of the Thermit process in cases like the ones given is the short time necessary for the ship to be laid up. In many cases little or no dismantling is necessary. In a number of stern shoe welds on lake steamers the sizes of the parts at the break have Fig. 52. — Another Sterupost Weld. S. S. "Corumia" of the Canadian Lake Transportation Co., Made in 1907. been from 11X16 in. to 10X20 in. or more, and the average time required has been about 36 hours complete. On a large number of ocean-going vessels the welds made have been on sections of larger dimensions than those quoted. In Fig. 46 is shown a weld on the anchor davit of the fam- ous U. S. S. "Olympia." Fig. 47 shows a repair on the anchor of the Morgan yacht "Corsair." Fig. 48 is a repair on the Avheel shaft of the ''Nashville," a Cumberland River steam- boat, made in 1912. Fig. 49 shows a repair on the wheel shaft 390 GAS TORCH AND THERMIT WELDING of the river steamer ''Osceola," welded at Jacksonville, Fla., in 1915. Fig. 50 shows a sternpost weld on a tug, made in 1911. Another sternpost weld is shown in Fig. 51. Still another very similar weld is shown in Fig. 52. These are sufficient to give the reader a good idea of the application of the process to this class of work. CHAPTER VI RAIL WELDING FOR ELECTRIC SYSTEMS Pr()l)ably no problem in recent years has received more consideration at the hand of traction companies than the snl)ject of track construction and maintenance. Progress along this line has been continuous and expense has not been spared to obtain the best roadbed possible ; heavier rails are being used and particular attention is paid to the foundation, ties and drainage ; everything is of the best material and work- manship until it comes to joining the rails together, and here there usually develops the weak point of the entire system. This refers to the mechanical joints in common use. No matter how much care is taken in applying splice bars mechanical discrepancies and disadvantages cannot be overcome ; the rail sections are seldom uniform and the American Society for Testing Materials has adopted specifications which limits a maximum difference in height to ^/g^ inch. Variable height in new rails is often unavoidable, as joint plates fitting one rail perfectly may not fit its neighbor. There are many reasons why the mechanical joint fails to fulfill its function, which are not necessary to enumerate here. On the other hand the advantages of a welded rail joint, where circumstances will permit, are plain. AVhile there are several systems by which rail joints may be welded, at the present time the Thermit method is the only one that absolutely eliminates the joint. It also has advantages in the modified joint-welding woi-k. These advantages will become clear as our description of the actual processes develops. Briefly, the weld is accomplished by pouring the superheated steel obtained from the Thermit reaction into a mold surrounding the rail ends at the joint. This fuses ^^^th the base and web of the rail as well as with the lip and one side of the head. An insert cut from a rolled section of similar analysis to that of the 391 392 GAS TORCH AND THERMIT WELDING rail itself is placed Ijetween the heads at the running face, and the lower part of tliis insert is melted into the Thermit steel. The mold is so constructed, however, that the head of the rail and the top part of the insert are not melted but are merely heated to a welding temperature, so that when the Thermit metal begins to cool and contract, thus drawing the rail ends together with tremendous force, the squeezing action on each side of the insert thoroughly butt welds it into the head. When this has been accomplished the running face is ground to a true surface and the surplus metal ground out of the groove. The weld obtained in this way is so perfect that it is practically impossible to detect its location after the rails have been paved in. In practice the welding is best done after the ties have been concreted. If the welding is done before the concreting it is somewhat difficult to keep the rails in perfect alignment and to proper surface after the temporary splice bars have been moved. The rails are spaced f in. apart and must be thoroughly cleaned for a distance of 4 in. from each end. This can be accomplished by using a small steel-wire brush. The ends of the rail heads must next be cleaned and where necessary filed smooth so that when the insert is fitted the maximum amount of contact surface will be obtained. In other words the insert must be made to fit accurately. Following this the rails are brought to proper surface and alignment, care being taken to keep the rail ends a trifle high, so that when a straightedge about 30 in. long is centered over a joint there will be a space of about ^/oo in. between the ends of the straightedge and the surface of the rail directly under it. This slight raising of the rail ends has been found necessary in practice, as it assures proper alignment and surface after grinding. PLACING THE INSERT The joint is now ready for the placing of the insert, as shown in Fig. 53. Various thicknesses of inserts must be kept on hand to fit the variation in gap brought about by temperature changes and other causes. With the inserts in position two part sand molds are applied, as shown in Fig. 54. these are rammed in advance on a foundry squeezing machine over a wooden pattern, the dividing line of the molds being RAIL WELDING FOR ELECTRIC SYSTEMS 393 Fig. 53. — Adjusting Insert Between Bails. Fig. 54. — Adjusting Two-Part Mold to Eails. 394 GAS TORCH AND THERMIT WELDING in the center of the web. A squeezing machine being used for this purpose is shown in Figs. 55 and 56. The pattern is so constructed as to form an opening around the rail ends in the shape of a collar into which the Thermit steel is poured. This collar is usually about 3 in. Avide and Fig. 55. — Eaniniing Molds on a Squeezing Machine. Wooden Patterns and Sheet-Iron Mold Box in Foreground. varies from ^ in. to | in. in tliickness, depending on the rail section to be welded. Pouring gates and risers are also pro- vided for in the mold in a similar manner to those already described. Recent improvements in the construction of the molds used RAIL WELDING FOR ELECTRIC SYSTEMS 395 have been perfected to guard against runouts. A bead pro- trudes around certain parts, so that when the two halves are clamped together it is compressed between them and forms a safeguard against tlie liquid steel running out. Before clamping the molds to the rails two cords of asbestos Fig. 56. — Cutting Heating Gate in Mold. Finished Mold and Wooden Pattern Shown in Foreground. soaked in molasses are applied around the contour of the rail, one on each side of the joint, as shown in Fig. 57, and placed in such a position that they will come just inside the outer edge of the mold box. The molasses is sufficiently sticky to make these cords adhere tightly to the rail and form a very 396 GAS TORCH AND THERMIT WELDING efficient luting all around the outside edge between the mold and the rail. As an additional precaution a small amount of fireclay is blown into the mold through the pouring gate by Fig. 57. — Applying Asbestos and Molasses Strips to Rails Previous to Placing Mold. Fig. 58. — Fijial Luting Process, Blowing Powdered Fire Clay into the Mold. means of compressed air, as shown in Fig. 58, while the other openings are closed temporarily by inserting wooden plugs. Should there be any opening around the edges of the mold RAIL WELDING FOR ELECTRIC SYSTEMS 397 the fire clay escaping tlirough that opening will he caught by the molasses on the asbestos cords and the opening auto- matically sealed. When the air-pressure gage indicates that the mold is tight the wooden plugs are removed and all surplus fire clay blown out through a small blowout gate provided in the lower part of the mold. Preheating. — The rails are now ready for preheating, and must be brought to a bright-red heat. For this purpose a special portable heater is used. The opening in the mold through which the rails are lieated Fig. 59. — Mold and Crucible Clamped in Position Eeady for I'rehealing. Box Containing Additions Set in Eiser of Mold. is situated about two-thirds from the top, so that the flame strikes the lower portion of the web of the rail and has the effect of heating the entire rail section uniformly. This pre- heating also accomplishes two other objects, as it bakes the mold and at the same time heats up a can of additions Avhich are added to the Thermit in the crucible to improve the quality of the steel produced. This can of additions is placed on top of the mold in a special receptacle provided for the purpose in the riser opening, as sliown in Fig. 59. As soon as the rai^.s are red hot, the mold thoroughly dried, and the can of 398 GAS TORCH AND THERMIT WELDING additions heated to the proper temperature, the hurner is with- drawn and the heating gate plugged witli a sand core. The burner is then directed down tlie pouring gate in order to restore any heat that may have been lost during the plugging operation. The mold is now ready for pouring, and the crucible is placed in position, the reaction started and the melt tapped, just as described for other work of this character. The steel enters the mold on the lip side of the rail and thoroughly melts and amalgamates with the lip as well as with Section through Rail Joint O'fter Pouring Fig. 60.— Sectional View of Eail Weld. the web, base and one side of the head. No Thermit steel is allowed to touch the running surface of the rail, as the steel insert closes up the space between the rail ends except for a small opening of about |- in. from the outside of the head. The under part of the insert is thoroughly fused with the liquid steel and becomes a part of the fusion weld, as shown in Fig. 60. As explained previously, when the metal in the weld begins to cool it produces sufficient pressure on each side of the insert, due to contraction, to thoroughly butt weld it in position so that at the end of the operation the entire rail is welded into »• RAIL WELDING FOR ELECTRIC SYSTEMS 399 one homogeneous mass. A finished weld is shown in Fig. 61 and in Fig. 62 is shown a number of operations going on at once. One of the principal advantages of this type of weld is Fig. 61. — Fiuisbed Tlicnnit Fully Welded insert Joiut. that it depends on a predetermined chemical reaction which is always uniform. The success of the reaction is assured by the fact that every Thermit-welding portion is weighed out Fig. (52. — Preheatiug Eails, Drying Molds and Heating Thermit Additions, All in One Operation, Four Joints Being Heated at Once. separately for each rail section to be welded and in just the proper quantity to obtain a perfect weld on that section. The placing of the additions in a sheet-iron container is a fairly recent improvement and has resulted in not only obtain- ing a higher grade of steel in the weld than was possible 400 GAS TORCH AND THERMIT WELDING formerly but also permits of obtaining a greater quantity of steel from the Thermit used. This in turn permits a weld to be made with less Thermit than formerly and has considerably reduced the expense of the welded joints. THE USE OF ADDITIONS The scheme of employing the additions as outlined is to use plain Thermit, and instead of mixing the additions of manganese, nickel and other materials for improving the quality of the Thermit steel throughout the Thermit itself these additions are now put up in the sheet-iron container which is placed on top of the mold during the preheating and is heated red hot by the waste gases. It is then placed in the center of the crucible and is melted down by the heat of the Thermit reaction. This method enables the production of more steel from a given amount of Thermit without any sacrifice of heat, with the result that the cost of the welding portion of Thermit is considerably reduced. After the weld has been poured the mold should he left tmdisturhed for a few hours; in fact, the longer it is allowed to cool the better, as the metal in the weld lias time to become properly annealed. The mold boxes can then be removed, the metal left in the risers and pouring gates cut off by nicking Avith a hammer and chisel and knocking off. All sand, etc., is cleaned from the rail in the vicinity of the mold and the joint is ready for grinding. The Grinding- Machine. — As the steel insert is left a trifle higher than the rail section this excess metal must be ground off together Avith the excess metal left in the groove and on the outside of the head where the riser is removed. This grinding can best be accomplished by means of tlie machine shown in Fig. 63. It possesses the advantage that the weight is concentrated over the grinding Avheels so that a deeper cut can be taken. It also has attachments permitting of grinding in the groove of the rail and on the gage. If only a very few joints are being welded, however, the grinding can be very satisfactorily accomplislied with a flexible-shaft grinding machine. In fact one of the great advantages of RAIL WELDING FOR ELECTRIC SYSTEMS 401 the Thermit process of rail welding is that a few joints can be welded almost as economically as a large number, and trac- tion companies can do the work themselves wherever they please, and when the work is properly organized a gang of nine men can weld from 35 to 40 joints in a nine-hour day. The simple and portable character of the welding outfit, including the grinding machine, is a strong point of merit, as one service car will carry all the equipment or it can be hauled on its oAvn wheels. Fig. 63. — Rail-GriiuUng; Machine Derailed Under Its Own Power. The rail-grinding machine referred to was originally dc-- signed by the Thermit company for grinding Thermit-welded rail joints, but has proved very efficient for grinding out corrugations, pounded joints, and in fact anything that is required for a rail-grinding machine on an electric-railway system. An important feature is tlie derailing device which permits the removal of the machine from the path of traffic very quickly. The concentration of the weight over the grind- ing wheels has already been mentioned. By this arrangement deep or light cuts may be taken Avithout danger of the grind- ing wheels chattering. The truck frame carries two axles. 402 GAS TORCH AND THERMIT WELDING One axle is fitted with ordinary 22-in. ear wheels and the other has eccentrically mounted 13-in. wheels. By means of sliding adjustments the wheels may be spaced on their axles to fit road gages of from 4 ft. 8^ in. to 5 ft. 2^ in. The grind- ing-wheel brackets have both vertical and horizontal movement to allow for adjustment to the work. The bracket also may be rotated so that the face of the grinding wheel may be adjusted to the surface to be ground. By a special arrange- ment of the motors used the machine can be driven forward at a rate of 2 or 4 ft. per minute and 6 to 12 ft. per minute on the return. A reversing switch permits speeds to be used in either direction. Where it is desired to grind out a depression a special arrangement is used by which an accurate curve is obtained. The grinding wheels run about 1833 r.p.m., which corresponds tc a peripheral speed of 6719 ft. per minute on a new 14-in. wheel. The grinding-wheel motors are rated at 3^ hp. and tlie propelling motor at If hp., and they are designed to operate on 500 to 550 volts. The derailing device operates by power and is quickly brought into use. The complete machine weighs about 6000 pounds. CHAPTER VII WELDING COMPROMISE RAIL JOINTS There is very often a demand for a compromise joint to be supplied quiclvly when some bolted or cast-welded joint has failed. An inexpensive shop-welding outfit enables traction companies to weld their own compromise joints in a few hours, and these will give the same results in service as a regular Thermit rail weld. When a large number of compromise joints are to be made on the same rail section it is advisable to have patterns and mold boxes made especially for the purpose. Where only tAvo or three are to be welded, however, the work can be done by the "wax method," which is the method used in all general repair work. To assist in the aligning and surfacing of the rails when two short lengths are to be welded together it is advisable to provide a suitable bed to which the rails can be bolted. Two stringers running about 10 in.X6 in.XJO ft. long, on Avhich four wooden or steel tics can be bolted, answer this purpose very well. The two center ties should be spaced about 18 in. in the clear and the second tie spaced to take care of the shortest length of rail to be welded. It is best to imbed the surfacing bed in the ground up to the top of the stringer. To hold the rails to the ties long bolts can be used in place of track spikes. These bolts should be allowed to project through the face of the tie a sufficient amount to take the smaller of the two rails. Allowance must also be made for a U-elamp, or bridge clip, one end of which will bear on the base of the rail and the other on the tie, or in the case of a small rail, will rest on a spacing block. To bring the smaller rail up to the surface of the high rail a filling block is placed under the base, and to obtain accurate surfacing, shims or old hack-saw blades may be used in addi- tion. With the rails spaced | in. apart and accurately adjusted, 403 404 GAS TORCH AND THERMIT WELDING the U-elamps are bolted down tight and the insert fitted as described for making ordinary rail welds. If patterns and mold boxes have been provided in advance the work proceeds in the same way as for ordinary rail welding, bnt if the wax method is to be used about 3 lb. of wax should be broken into small pieces, placed in a pan and heated until entirely melted. The wax is then allowed to cool until it becomes plastic. It may then be shaped by hand around the rail ends in the form of a collar. USING A RAIL SECTION FOR A PATTERN In cases where five or six compromise joints are to be welded between the same rail sections considerable time and trouble can be saved by using a short length of each rail section as a pattern. These should be about 8 in. long, butted together and fastened by tacking with the oxy-acetylene-weld- ing process so that they will be held together securely. Mold boxes of sheet iron in two halves can be cut to fit with the oxy-acetylene cutting flame so that they will fit the sections to approximately ^/jc in. all around. Each half of the mold box will then have a different section cut in each side. A wax collar is formed around the joint of the two pieces of rail in the regular way as described above and the one-half oi tlie mold box is laid on the ground back down, placing the pattern made by tacking the two rail sections together on this half of the mold box. Then thin pieces of sheet iron are laid on top of this half to obtain a parting when the other half of the mold box is placed on top. The other half may then be placed in position and rammed up to the height of the riser with molding material, using a wooden pattern for the riser opening. "When this has been completed the whole mold box, pattern and all, is turned over and the bottom half of tlie mold l)ox rammed, inserting a wooden riser pattern in a similar way to the first half. This ])ottom half is then rapped slightly and lifted from the pattern. Then by rapping the pattern it may be lifted from the other half of the mold box. After removing the riser pattern the molds are ready to be placed on the rails to be welded, but before doing so the space between the lip and the ball of those rail sections should be WELDING COMPROMISE RAIL JOINTS 405 rammed flusli with molding material, "When the boxes are adjusted they should be luted carefully with fireclay around the outside edges between the mold and the rail to guard Fig. 64. — Welded Compromise Joint Between T-Eail and Grooved Eail. Fig. Gu. — Welded Compromise Joint Between Two T-Eails, Showing False Lip. against any run out of Thermit steel. The joint can then be preheated and poured in tlie regular way. This method will be found to save considerable time over waxing each joint separately, and the two short-rail sections 406 GAS TORCH AND THERMIT WELDING can be used as a pattern for any number of molds. If both right-hand and left-hand compromise joints are required the same pattern and mold boxes can be used by simply discon- necting and rearranging for either right or left. A welded compromise joint between a grooved and a T-rail is shown in Fig. 64. Where a compromise joint is to be welded between two T-rails as shown in Fig. 65 the same method can be used except that in this case it is necessary to arrange for a false lip to be cast out of Thermit steel similar to tlie lip of a grooved rail. In other words the Thermit-steel collar must be carried around each side of the head and on one side must be shaped in a corresponding manner to the lip of a groove rail. This is necessary because when the metal begins to cool and con- tract there must be an equal shrinkage force on each side of the insert extending to the top of the head tending to draw the rails together, otlierwise the insert will not be thoroughly butt-welded into the head. THE CLARK JOINT Shortly after the development of the Thermit rail-welding process; Charles H. Clark, chief engineer of the Cleveland Fig. 66. — Complotod Clark Joint. Railway Co., perfected a joint known as the ^' Clark joint," which has proved exceedingly successful in Cleveland and other Eastern cities where many thousand joints have been installed. In its original form it consisted of a combination of splice WELDING COMPROMISE RAIL JOINTS 407 bars and Thermit steel, it being Mv. Clark's opinion that the head of the rail conld be supported by using plates that would Fig. 67 —Open Mold aud Crucible in Position for Making Clark Joint. Fig. 68. — Completed Modified Clark Joint, Showing Weld of Base. come under the ball of the rail. Furthermore, in order to hold the rail rigid he considered it important that there should 408 GAS TORCH AND THERMIT WELDING be no play in the bolts, so the holes in the plates and rails were drilled round and machine bolts used after reaming for a drive fit. In order to keep the bolts and plates from working- loose and to afford bonding between the rails a Thermit-steel shoe was cast around the base as shown in Fig. 66. In practice the rails and splice bars are drilled with holes Vi6 ill. less in diameter than the bolt to be used. The splice Fig. 69. — Section through Modified Clark Joint. It will be ISTotiecd that- the Lower Part of the Kail and Fish Plates are entirely Amalgamated. bar is then applied in the ordinary way and held in place by a couple of temporary bolts, a drift pin being driven into one hole each side of the joint to keep the rails in position. The remaining holes are then reamed with straight-end cutting reamers, after which the machined bolts are driven and tight- ened up in the usual manner. After preheating the rail ends the Thermit steel is run into an open mold surrounding the lower part of the rails as illustrated in Fig. 67. WELDING COMPROMISE RAIL JOINTS 409 In tlio latest type of Clark joint rivets arc substituted for the machined bolts, the riveting being accomplisbed by a pneumatic riveter suspended from the rear end of a flat car carrying an air compressor. A modification of the Clark joint, shown in Fig. 68, has been adopted with marked success by the United Railways and Electric Co., Baltimore, and is also being used on other properties. The object of the modification was to obtain a larger weld Fig. 70. — Appliances in Position for Welding Third Rail. of the base, and in order to do this the Thermit steel was poured into an inclosed mold box instead of into an open mold and the rail ends were preheated to a red heat with the molds in place before the Thermit charge was ignited. Furthermore, the design of the fish plates is somewhat changed, these being of special design 1 in. in thickness and 32 in. long and being so formed as to fit snugly the contour of the head and base of the rail. At the same time they provide a minimum amount of space betAveen the web of the rail and the vertical sides 410 GAS TORCH AND THERMIT WELDING of tlie fish plates. The eliannel bars and rails are of the same kind of steel (liigh carbon) and both are punched at the mill with ten IVic-ii^- holes, spaced 3 in. centers and beginning 2 in. from the end of tlie rail. Fig. 71. — A Welded-Up Cross-Over. Fig. 72. — Motor Case with Broken Lug Previous to Welding. The joint has been applied thus far exclusively for 7-in. girder groove rails weighing 103 lb. per yard. These 7-in. girder sections are undercut by the manufacturers ^/jg in. so as to provide a space of ^ in. at the base when the rail heads are butted. This jprocedure more effectively enables the WELDING COMPROMISE RAIL JOINTS 411 Thermit steel to weld the rail and fish plates into a solid mass at the joints, as shown in the section, Fig. 69. Welding- the Third, or Conductor, Rail. — The welding of the third rail has been carried on more extensively abroad than in the United States. This is especially true of France, where several thousand joints have been welded for the Metropolitan Eailway and others in the neighborhood of Paris where this method of bonding is now standard practice. In making these welds in France the base and flange only of the rail was welded Avith Thermit steel. As a great deal of third rail, both here and abroad, is used Fig. 73. — Motor Case Welded and Eeady for Service. in tunnels and subways, it is not necessary to make any provi- sion in such cases for expansion and contraction, it having been found from practical experience that the temperature changes seldom exceed 25 or 26 deg. F. These are the figures that were determined by experiment in the subways controlled by the Metropolitan Eaihvay of Paris. In cases, however, where the third rail is laid in open stretches of track where it will come under tlie full influence of atmospheric changes in temperature it is of course necessary to provide suitable means for taking care of the expansion and contraction, and this can be easily done by installing an expansion joint at regular intervals. Welding third rails has advantages that need hardly be 412 GAS TORCH AND THERMIT WELDING enlarged upon, providing, as it does, for a nniform electrical conductivity of the rail in question and a method of bonding which will not deteriorate. A Thermit outfit in position for welding a tlurd rail is shown in Fig. 70. A very simple way to make cross-overs of any desired form is shoAvn in Fig. 71. Tlie rails are cut, shaped and then Thermit welded together and then the surfaces are ground. Tig. 74. — Weld on Broken Tinek Frame. The result is as smooth and solid a cross-over as it is possible to make. This is suggestive of many other similar uses. The same outfit required for welding compromise joints can be used most advantageously for welding motor cases and truck frames. The process offers special advantages for such repairs, OAving to the fact that the collar, or reinforcement of Thermit steel, which is fused around the weld may be made heavy enough to insure against future breakage. In Fig. 72 is shoAvn a broken motor case, and in Fig. 73 a welded one. A welded truck frame is shown in Fig. 74. CHAPTER VIII WELDING CAST IRON AND OTHER PARTS The Tliormit process, while adapted to the weldmg of cast iron, cannot be nsed on all cast-iron welds owing to the diffi- culty in many cases of allowing properly for the shrinkage of the metal in the weld when cooling. The Thermit steel contracts twice as much as the cast iron, so that in certain constructions shrinkage strains will be set up in the weld causing cracks. Tliis difference in shrinkage often makes it impractical to weld long cracks in thin sections. As a general rule we should say that if the length of crack is more than eight times the thickness of the material a Thermit weld should not be attempted, because on account of the difference in shrinkage along the line of the fracture small hair-line cracks will appear perpendicular to the line of the fracture. These cracks, however, being perpendicular to the line of the weld are often therefore of little consequence and do not interfere with the strength of the weld. It is also not usually feasible to weld cracks in cast-iron cylinders, pots, kettles and similar castings. Where there is a clean break between two sections or where the section to be welded can be completely cut through and can be separated a sufficient amount to allow for the contraction in the weld, and also where the length of the weld is not more than eight times the thickness of the material, a Thermit weld would be entirely practical and can be made in exactly the same way as outlined for the welding of wrought-iron and steel sections. In cases where the crack can be opened up mechanically or by lieating a parallel part to a dull red a weld may be made, but it must be remembered that expansion gained in this way must be a little more than the expansion of the parts next to the weld during the preheating. Care should be taken in preheating for cast-iron welds not 413 414 GAS TORCH AND THERMIT WELDING to heat tlie sections too hot ; a dull red is sufficient. One should be careful to keep the heat going until the mold is thoroughly dried out. The mixture of Thermit for the weld should he different than for wrought iron and steel, and for this purpose the special mixture known as cast-iron Thermit is recommended. This consists of plain Thermit with which is mixed 3 per cent ferrosilicon and 20 per cent mild-steel punchings, i.e., to every 100 lb. of plain Thermit is added 3 lb. of ferrosilicon and 20 lb. of punchings. This gives the best results on cast iron and produces a homogeneous metal in the Avelcl. Welds on cast iron are a little more difficult to machine than welds on wrought iron and steel, as the metal along the line of junction of the Ther-mit metal and the cast iron is apt to be a ti'ifle hard due to the absorption of carbon from the cast iron. This objection is not a serious one, however, and hundreds of welds have been completed with the most satis- factory results. Examples of Cast-iron Welds. — In order to show the pos- sibilities of welding various cast-iron pieces with Thermit a few examples taken from actual practice are given. Fig. 75 shows how a new jaw was burned onto the frame of a heavy shear, in the shops of the Kaleigh Iron Works Co., Raleigh, N. C. The job was done in 1906 and the machine is still in service. This machine was designed for shearing l-|X6-in. bars, producing an enormous strain on the jaw^s. A large part of the corner of the lower jaw, weighing about 75 lb., broke off. It then became a question of getting a new frame or burning on a new jaw corner, which would require the fusing of a surface about 1 sq. ft. in area in order to obtain a thorough union. It was finally decided to try Thermit. The surface of the break was chiseled off for the reason that a cast-iron break is usually glazed with graphitic carbon. The mold was then put in place and well luted with fire clay where there was any danger of leakage. Moist sand was also rammed around the mold and other parts as a further precaution. The surface of the fracture and the stock around the jaw was then heated through the gates and risers by means of gas jets, in order that the casting might be thoroughly heated and all moisture driven out. The gas torches were kept in action WELDING CAST IRON AND OTHER FARTS 415 several houi's. A crucible containing 175 lb. of Thermit Avas then put ill position for tapping into the mold. After the reaction tlie Tliermit steel was run into the mold and at once fused the entire surface of the fracture. In the meantime a Fiu. 75. — New Jaw Burned to Frame Castiiis^ of Heavy iShear. ladle of molten cast iron containing about 600 lb. of metal was held in readiness and was superheated by means of a Thermit semi-steel can. As soon as possible after pouring the Thermit, this cast iron was poured into the second gate, shown at the front of the jaw, and this forced the Thermit steel out of the 416 GAS TORCH AND THERMIT WELDING mold after it had served its purpose in bringing the surface of the jaw to a Avelding or fusing heat. The result was a new corner of cast iron burned onto the old jaw. The illus- tration shows the two gates and four risers before they were trimmed off. In Reactions for the fourth quarter of 1917 W. J. Musick, blacksmitli foreman of the St. Louis shops of the Missouri Pacific R.R. wrote : We recently had one of our ste;un liannners break tlirouLcli both sides of the frame and through the throat, the fracture Ix'in.i; Gl in. long. This hammer was so badly broken that it seemed as thou-h it were doomed for the scrap pile. A new . team hammer was ordered, but in the meantime it was decided to try to repair the old one with Thermit. The weld was made and a successful repair was accomplished, resulting in saving a considerable amount of money. ^^'e have also Avelded a cast-iron engine bed for the Helmbacher Rolling Mill Co. The !)cd of this engine is 32 in. high and 12 in. across the top. The mill was only shut down 24 hours while the repair was being made. Master Mechanic George M. Stone, writing in the same , publication, says : I think your readers will be interested in the accompanying illus- trations of Thermit welds which have been made by me in the shops of the Chicago, Rock Island and Pacific Railroad, Chickasha, Okla., and I would like to call particular attention to the welding of valve seats on locomotive cylinders, which I consider an exceptionally good piec liHi^PW L 1 1 --^ i 'f ' ^ l^i^iMii HHHi 1 "*! , i; 1 , t 1 i 1 1 li 1 ^ Fig. 81. — Sheet-Iron Mold Box and Wooden Pattern. and the inner cylinder. After it is removed from the outer cylinder and the wax hardens, enough of it is trimmed away between the blades to allow for chip space. The wax matrix with the blades in place is then rammed up in a mold. The inner cylinder is then warmed slightly and removed. A steel shank is next inserted and centered correctly. The wax is now melted out and the Tliermit steel run in, welding the blades securely to the steel shank. A wax matrix, with blades and inner cylinder in place, is shown in Fig. 82. In this illustration the wax has been cut away between the blades and the assembly is ready to be put into the mold box and WELDING CAST IRON AND OTHER PARTS 421 be rammed up. When the work is cool it is centered and the blades ground for size and clearance. A helical reamer with inserted high-speed steel blades is shown in Fig. 83. This reamer was made by T. 0. Martin, Fig. 82.— High-Speed Steel Cutters Held in Wax Pattern with Hollow Steel Core Fig. S3. — Helical Inserted-Blade Reamer Made by the Thermit Process. blacksmith foreman of the Illinois Central R.R. shops at Jack- son, Tenn. Preheaters for Thermit Work. — As practically all welds in ordinary practice, except those on pipe, require preheating 422 GAS TORCH AND THERMIT WELDING it is well to use heaters made for the purpose wherever possible. In some shops gas- burning torches supplied with compressed air may be used. Many shops, however, have neither gas torches nor compressed air. This method is not practicable on outdoor welds. Crude-oil heaters should not be used at all, on account of their tendency to deposit carbon or other matter on the surfaces to be welded, thereby causing imperfect welds. In order to make the preheating work as easy and convenient as possible the Metal and Thermit Corporation makes the preheaters here shown. Fig. 84 shows two kinds, a single and a double burner. For infrequent or small jobs the single burner, which may be fitted with a flaming burner also, will probably answer the purpose. Where a number of Table V. — Cost of Thekmit and Apparatus fofv General Welding Gross Lb. Shipping Weiglit Cost Railroad Thermit (.50-lb. boxes only ) 67^ $17.50 I'lain Thermit (50-lb. boxes only) . ,59 17.00 Cast-iron Thermit (.50-lb. boxes or.ly ) 70^ 17.50 Ignition powder (i-lb. cans) .45 Yellow wax, per pound . . . .35 Punchings, per pound .025 Special molding material (300 lb. ) 340 4.00 Fire clay (300 lb. net) 340 8.50 Fire brick, per barrel (300 lb. net ) 340 4.00 Kiln-dried silica sand (.300 lb. not ) 340 3.50 Single-burner preheater 200 50.00 Double-burner preheater 225 75.00 Flaming-burner attachment 3.00 Magnesia stones, No. 1 .15 Magnesia stones. No. 3 .20 Magnesia thimbles, No. 1 .10 Magnesia thimbles, No. 3 .15 Magnesia tar, about 400 lb. net 450 .06 I*lugging< material, No. 2 package .10 Automatic crucibles (with ca:is and rir.gs) 40 4.00 Automatic crucible, No. 2 60 5.50 Automatic crucible, No. 5 1,50 11.00 Automatic crucible, No. 10 775 60.00 Cast-iron relining cone, No. 1 50 5.00 Cast-iron relining cone, No. 5 1.50 12.00 Cast-iron relining cone, No. 10 600 40.00 Tripods, Nos. 1 to 7, weights 11 to 05 lb .$2.50 to 9.00 WELDING CAST IRON AND OTHER PARTS 423 Fig. 84. — tSingle-Bunier and Double-Burner Preheaters, Using Either Gasoline or Kerosene. Fig. 85 —Rail Prelieater That Will Heat Four Joints at Once. 424 GAS TORCH AND THERMIT WELDING welds are to be made, however, tlie double-burner apparatus should be selected. In these illustrations A is the place to attach the hose from the compressed-air supply ; B is the valve for regulating the pressure on the surface of the fuel; C is a tube which runs within a few inches of the bottom of the tank; D is the needle valve which controls the fuel to the burner; E is the air pressure control to the burner; i^ is a check valve which prevents back fire ; G are torches or burner pipes ; // is a flaming burner. The small tank on the left side is a water separator for the compressed-air supply. Table VI. — Cost or Materials and Appliances for Pipe Welding Standard Weight Pipe. PRICE OF WELDING POBTION9 Price of Mold Inside Diam- eter, loches 100 Less Over 100 and Le33 than 500 Over SCO and Less than 1000 1000 More Price and Size of Crucibles Price and Size of Tongs Price and Size of Clamps Price and Size of Pipe Facing Machines H $0.36 $0.32 $0.27 $0 19 $0.75 No. 2] No. 2 No. 1] No. 11 H .44 .40 .36 .26 .75 2 2 1 -CO .78 .90 .56 .74 .86 .55 .69 .84 .39 .60 .75 1.00 1.25 1.50 2 2 2 $1.75 2 2 2 $2.00 $20 00 $35.00 2 1.03 .99 .97 .90 1.75 2 2 2H 3 1.50 2.16 1.46 2,12 1.43 2.09 1.35 2.06 2.00 2.25 31 3i 3.00 3 3 2.50 3H 4 3.06 4.63 3.02 4.59 2.99 4.56 2.96 4.50 2.50 2.75 41 4J ^ 4.75 4 4 3.25 2 2J > 25.00' 2 2i 60.00 EXTRA HEAVY PIPE U 1 IK VA 2 2H 3 3H 4 $0.45 $0.42 SO. 37 $0.29 $0.75 .54 .43 .42 .34 ,75 .72 .67 .62 .56 1,00 .90 .86 .84 .75 1,25 1.14 1.10 1.05 .98 1.50 1.78 1 75 1.71 1,65 1,75 2.94 2.90 2.86 2.80 2.00 4.23 4.20 4.16 4.10 2,25 5.43 5.40 5.36 5.10 2,50 6.22 6.18 6.14 6.08 2,75 1.75 3.00 4 75 5 7,50 21 2 2i$2,00 2 2J 3 4: 2.50 3,25 5 4 50 $20.00 25.00 $35 00 60.00 DOUBIiE EXTRA HEAVY PIPE Vi $0,93 $0.90 $0.86 $0.81 $1.50 No-2|$1.75 No. 2] No. 1 No. 1 H 1.08 1.04 1.02 96 1.75 2 $2.00 1 1.20 1.16 1.14 1.08 2.00 l\ 3-00 l\ 4 3} 2.50 VA 1.49 2.14 1.43 2.08 1.41 2.03 1.36 1.94 2.25 2.50 $20.00 $35 00 2 3.82 3.73 3.71 3.68 2.75 2H 7.30 7.27 7.24 7.20 3.00 1} 7-50 1} 4-50 3 10.60 10.57 10.54 10.50 3.50 ' Fig. 85 shows a four-burner portable apparatus used largely for rail-welding work. It carries its own air compressor, WELDING CAST IRON AND OTHER PARTS 425 which may be run by attachmg to a trolley wire or to some other olectric-ciirrent supply. All of these burners use either gasoline or kerosene. Cost of Thermit Welds and Apparatus. — There are many factors winch enter into the calculation of the cost of Thermit welding and the apparatus. Where a single weld is to be made and the shop man has to buy the apparatus, materials and do the work himself, the cost will naturally be higher than where several welds are to be made or where he can hire it done. There are so many places now making a specialty of Thermit welding that in ordinary circumstances it is usually better to have them do the work on large jobs than for inex- perienced men to undertake the work. Data for the appropriate cost of various jobs have been given in tables and specifications throughout the article, but in order to give those responsible for repair or other welding work, as exact information as possible on which to base their calculations, the accompanying tables are included. These are taken from the price list of the Metal and Thermit Cor- poration, published June 15, 1918, and of course are subject to changes. These quotations are f.o.b. Jersey City, N. J. Table V gives prices for general welding materials, and Table VI for pipe work. It will be noted that some of the quotations in this last table do not exactly agree with figures given in the table of comparative costs of Thermit welded and mechanically-joined pipe, but it should be borne in mind that the comparative table gives averages only, and is also subject to variations in cost of labor and materials. INDEX A Acetone, 26 — , capacity of, for acetylene, 27 — injurious to weld, 28 — , nature of, 26, 27 Acetylene and Welding Journal, 144 — , cubic feet per pound, 28 — cylinder filling material, 27 , capacity of, 27 — , danger point of, 26 — , discovery of, 1 — , estimating amount of, in cylin- der, 28 — , explosive limits of, 6 — gas from pound of carbide, 29 — generator, Davis-Bournonville "Navy type," 32 — , heat units in, 3 — , ignition temperature of, 6 — manifolds, *118 , Davis positive-pressure, de- scription of mechanism, 32 , positive-pressure, *30 — — , details of 300 lb. size of, *31 , low-pressure type, phantom view, *42 , Oxweld portable pressure tjTpe, 36, *39 repairs, 45 sets, dimensions and weights of, 37 — generators, capacity of, 29 , low pressure, 41, *42, *43 , Navy type, size of, 35 , positive-pressure, capacity, 29 , pressure limit of, 30 , standard rating, 29 , the three types of, 28 ■ , -types, 28, 29 Acetylene plant layout, *34 , "Navy type," *33 — positive-pressure generator, 29 — ■ pressure generator, first, 2 — , production of, 26 — , specific gravity of, 6 Action of cutting torch, 257 Adaptors for regulator and cylinder connections, *103 Additions, use of Thermit, 400 Air chisel for welding work, *187 — - Reduction Sales Co., 92 — screen for cooling, 201, *202 Airco-Vulcan combination welding and cutting torch, *91 Alexander Milburn Co., 92 All-steel welding truck, *35 Allowance for expansion and con- traction, 154 Aluminum gear case, repair of, 205, *207 — oxide, melting point of, 174 — , preheating, 175 —, purity of, 173, 174 — sodium fluoride, 174 — , welding, 173 fluxes, 174 American Blaugas Corporation, 5 — Machinist, 208, 215, 231, 236, 244, 247 — Society for Testing Materials, rail specifications of, 391 — Welding Society, 267 Amount of Thermit to use, 345 — ■ — ■ — used in roll welding, 379 Anchor welds, *385, *386 Angle iron used in welding, *225, *226 Apparatus, cost of Thermit, 425 427 428 INDEX Applications of Thermit fusion weld- ing, 338 Areas of drill holes, 65 Armour Institute of Technology, 51 Asbestos and molasses strips, use ot, 395, *396 Assembly for welding and cutting, *105, *109, *111, *117, *12G, *127, *128 Automatic crucible for Thermit, *333 Automobile cylinder, broken, *189 — — welding, 164, *165 "Autogenous Welding," 1 Autogenous Welding, 272 B Back-pressure valves, *124, 125 Backward welding, 144, *147, *148, *149 Baking Thermit crucible, 335 Barium peroxide for igniting Ther- mit, 326 Bastian-Blessing Co., 64, 90 Battery, storage, burning, 152, 15;! Benzine vapor, 5 Benzol vapor, 5 Bethlehem Shipbuilding Corp. Acetylene plant, *33 Beveling boiler flanges, *289 Bisulphates of sodium and potas- sium, 174 Blaugas, discovery of, 5 — , explosive limits of, 6 — , — range of, 5 — , makers of, 5 ■ — ; method of selling, 5 Blau, Herman, 5 Blooming-mill housing Thermit weld, *417 Blowing a hole_ through a plate, *261 Boiling point of liquid oxygen, 10 — nitrogen, 10 Bond, rail, *204, *205 — welding outfit, *205 Bosses, forming, *143 Bournonville, Eugene, 2 Brass and bronze welding, 176 Brennan, A. F., 216 Bronze and brass welding, 176 Buckeye carbide feeding mechan- ism, 36, *38 — oxygen generator, *11 — ■ portable oxygen generator, *12 Building up a weld, *141 Burning battery cell connectois, 153 — , lead, data on, 153 — out carbon, 302 Butt joints, lead, 151 — -welding plates, *134 Calcium carbide, 2, 26 — — , how handled, 26 Calculating amount of Thermit, 346 welding gases, 63, 64 Calmbach, G. M., 200 Camograph cutting machine, *286, *287 Capacity of Oxweld generators, 40 oxygen cylinders, 9 Carbide, amount of acetylene pro- duced, 29 — feed, Buckeye, 36, *38 — , size of, 29 Carbo-Hydrogen Co., 87, 262 — cutting torches, 87, *88, *89 Carbon burning, 302 outfit, *304 — • electrode and oxygen jet torch, *274 — monoxide, 2 Card for cost keeping, *301, *303 Carhart, H. A., 231 Carnegie Steel Co., 208 Carrying case for cutting or weld- ing outfits, *120, *122 Cartridge, Thermalene, 46, 47, *48, 50 Cast, aluminum, purity of, 173, 174 — iron, cutting, 267 , samples of cut, *272 Thermit, composition of, 321 to steel, welding, 179 welding, 177, 178 , with Thermit, 413, *415, *4]7, *418, *419 INDEX 429 Chain links, building up, 205, *207 Chapman, R. E., 274 Characteristics of welding flames, *107, *113, *114 Charging an acetylene generator, 44 Chemical oxygen generator, *11 — symbol lor acetylene, 1 calcium carbide, 2 Chemistry of the oxy-acetylene flame, *107, 110 Chlorate of potash oxygen genera- tors, size of, 12 process, amount of oxygen produced, 10 for oxygen, 10 Chlorides of sodium, potassium, lithium, 174 Chrome steel welding, 186 Circular cutting, 284 City gas, ignition temperature of, 6 Clark, Charles H., 406 — joint, the, *406 , the modified, *407, *408, 409 Cleveland Eailway Co., Thermit welded joints for, 406 Coal gas, explosive limits of, 6 , specific gravity of, 6 Collars, building up, *143 Colors of tank and hose, 101 Combination torches for cutting and welding, *91, 92 — welding and cutting torch, Mil- burn, *91, 92 Commercial Gas Co., 274 Compromise rail joints, welding, 403, *405 Conductivity and oxidation, 169 Connecting up and lighting the torch, 104, *105, 109 Containers for poison-gas, welding, *230, *232 Contraction and expansion, 154 Cooling devices, 201, *202 — oven, Wiederwax, *161 — work, 160 Conveyor roller welding, *227, *228, 229 Copper, flux for, 180 — to steel, welding, 180 — welding, 179 Corsair, welded anchor of, *386, 389 Corunna, sterupost weld on, *389 Cost keeping form, *301, *303 — , — track of, 302 — of cutting, 266, 267 oxy-hydrogen cutting, 276, 276 Thermit apparatus, 422 pipe welds, 329, 330, 424 welds, 424, 425 welding large cylinders, 211 per foot, 204 Cracks, how to locate, 364 Crank case, broken and repaired,. *191 welding, *225, *226 Crankshaft welding jig for Thermit work, *361 — repair, 192, *193 — welding, *223, *224, 225 Crankshafts, welding with Thermit, 359, *363, *364 Crane, portable, for welding shop, *299 — , trolley, and hoist, *296 C, E. I. & P. R. R. shop work, 416 Crosshead welding with Thermit, 351 Cross-over rails welded with Ther- mit, *410 Crucible, baking Thermit, 335 — holder for locomotive work, *355 — , lining Thermit, *333, 334 — , Thermit automatic, *333 — , tapping Thermit, *334 Crucibles, details of Thermit, *333, 337 Crude oil or kerosene preheater, *157 Cumming, J. R., 202 Current required for separatiHfT oxv gen and hydrogen, 16, 20 Cut, size of, made by a cutting torch, 75 Cutting a rivet head, *261 — action of a gas torch, 74, 75 430 INDEX Cutting and welding outfits, 116, *117, 119, *120, *122, *126, -n27, *128, *129 — cast iron, 267 —, circular, 284, 285, 288, 289, 292 — , cost of, 266, 267 — data, 85 — , learning how to do, *258, *259, *260 — machines, *278, *279, *280, -281, *282, *283, *284, *285, *286, *287, *288, *289, *290, 291, *292, *293, *294 — , manifolds for, *118 — speed of gas torch, 83 — steel risers, 85 — tests, data on, 86 — tips, *84 , Davis-Bournonville, *77 ■ — tools, Messer, *84 — torch. Airco-Vulcan, *91 , first, 3 — • — for ship work, Oxweld, *87, 82 , how the, acts, 257 made by General Welding and Equipment Co., *84 — — , Milburn, *91, 92 , Eego, *90 , rivet-head, 80 , staybolt, *81, 82 that preheats oxygen, *273, *274 , Torchweld, *92, *93 , underwater, *94 — torches, 74, *75, *76, *79 ■ , earbo-hydrogen, 87, *88, *S9 — — , Davis-Bournonville, *75, *76 , gas pressures used, 78 , Imperial, 86, *88 , machine, *76, 78 , Oxweld, *79 — under water, 94 , depth of, 94 — unit, typical, *117 — with a guide, *262 oxy-hydrogen, 274 the gas torch, hand, 257, *258, *259, *260, *261, *262, *263, "266, 267, 269 "Cut-weld" torch, Milburn, ^91, 92 Cylinder, amount of acetylene in, 28 — , automobile, broken, *189 , welded, *190 , welding, 164, *165 — connection adaptors, *103 — grooved for welding, *188 — , Liberty, tacking jacket, *233 — , preheating low-jiressure, *210 — pressures for acetylene, 27 — welded, *189 — welding, a remarkable job of, 208, *209, *210, *211, *212 , cost of, ^11 — , — • low-i^ressure, *211 — ■, wrecked low-pressure, *209 Cylinders, acetylene, filling material, 27 — , — , temperature of, 27 — for oxygen and hydrogen, capac- ity of, 15 , weight of, 27 — , motor, removing carbon from, 302 — , sheet-metal, jigs for welding, ■*227 *228 *229 231 —,—, welding, *240, *241, *242 D Data on lead burning, 153 Davis-Bournonville Co., 15, 239, 267, 272, 279, 280, 286 cutting machines, *279, *280, *281, *282, *284, *285, *286, *2S7, *289, *290, *292, *293* *294 — torches, *75, *76 , gas pi-essures for, 78 Duograph, 239, *240, *241, *242 hand truck for welding out- fit, *35 " Navy type ' ' acetylene gen- erator, 32 — — positive-pressure acetylene generators, 29 INDEX 431 Davis-Boiirnonville underwater cut- ting toieh, *94 water-cooled welding torches, *57 welding torches, 55, *56, *57 — Acetylene Co., 29 — acetylene generator, size of, 32 -^, Augustine, 2 — electrolyzer cell, 15 , details of, *17 Davy, Edmund, 1 Decarbonizing motor cylinders, 302 Dentist's torch, *129 Details of Thermit mold box, *340 Discovery of acetylene, 1 oxygen, 9 Dissociation temperature of water, 4 Driers, acetylene, Oxweld, size, 40 Drigas, 5 — , explosive limits of, 6 — , method of selling, 5 — , explosive range, 5 Drill hole areas, 65 Drums, sheet-metal, welding, *240, *241, *242, *243 Duograph, the, 239, *240, *2'^1, *242 E Edison Storage Battery Co., 244 — welding machine for oblong seams, *245 Electric blower tj^e of preheater, • *158, *159 Electrical properties of oxygen and hydrogen, 13 Electrolytic hydrogen, purity of, 14 — method, principles of, 13 — oxygen, 10 , purity of, 14 — Oxy-Hydrogen Laboratories, Inc., 24 Electrolyzer cell, Davis, 15 -, description of, 18 ,—, details of, *17 International, *19 — , current used, 20 — , details of, *20 — , principles of, 22 Electrolyzer cells, space for battery of, 25 ■ — , details of Levin, *25 ■ — , Levin, 24 — , — , principles of, 24 — plant layout, *22, *23 Electrolyzers, currents used in, 16 — , Davis, sizes of, 16 — , gas capacity of, 16 Elements, separation of, 173 Emergency cutting outfit, *122 — Fleet Corporation tests on strength of oxy-acetylene welds, *310, 311 Endothermic acetylene, 3 Estimating amount of acetylene in cylinder, 28 Eveready instruction book, 150 Examples of welding jots, 187 methods, *139 Expansion allowance on locomotive frame, *200 — and contraction, 154 Explosive limits of acetylene, 6 blaugas, 6 • — • coal gas, 6 drigas, 6 hydrogen, 6 thermalene, 6 — welding gases, 5 — range of blaugas, 5 • drigas, 5 Equipment of welding shop, 295, 296, *297, *298, *299, *3G0 — rules, 305 Feeding mechanism. Buckeye car- bide, 36 Fery, F. M., 319 Field of gas-torch welding and cut- ting, 7 Filling rod, using the, 137, *138, *147, *148, *149 — up a hole, 142 First acetylene pressure generator, 2 — cutting torch, 3 — uses of the gas-torch, 3 432 INDEX First welding gas-torch, 2 Fixtures for welding, *219, 221, *222, *223, *224, *225, *226, *227, *228, *229, *230, *232, *233, *234, *235, *23G, *237, *238. Flame characteristics, *107, *113, *114 — , oxy-acctylene, 3 —,—, chemistry of, *107, 110 Fluorides of sodium, potassium, aluminum-sodium, 174 Flow, indicator for gas, 128, *130 Flux for aluminum castings, 175 ■ brass and bronze, 176 cast iron, 177 copper, 180 Fluxes for welding aluminum, 174 ■ — used in welding, 169 Fouche, Edmond, 2 Frame, rudder, ready for welding, *201 —,—, repair, *208 — , welded locomotive, *199 — welding, locomotive, *200 Fuel used in chemical oxygen gen- erators, 12 Furnace for preheating large pinion, *372 — , preheating, on iron table, *156 — , — , using charcoal, *156 Fusion and plastic welding with Thermit, 319 — welding, application of Thermit, 338 of heavy sections with Ther- mit, 333 G Gate patterns for Thermit molds, *341 Gauthier-Ely, 2 Gear case, repair of, 205, *207 — teeth, welding, *144, *168 ,— in, *197 Generating plant, "Navy type" acetylene, Bethlehem, *33 Generator, charging acetylene, 44 — , details of positive-pressure, *31 Generator, positive-pressure, sta tionary type, *30 — repairs, acetylene, 45 — sizes, Oxweld low-pressure, 44 Generators, acetylene, pressure limit of, 30 — , — , the three types of, 28 —,—, types, 28, 29 — , Davis acetylene, sizes and weights, 37 Gages for gas j)ressures, 95, *96, *98, *99, *100, *101, *102 Galvanized iron welding, 186 "Gas Torch," 1 Gas, acetylene, amount from pound of carbide, 29 — capacity of Davis electrolyzers, 16 — consumption in cutting, 83 lead burning, 153 -of carbo-hydrogen cutting torches, 84 — cutting torches, 74, *75, *76, *79 — flow indicator, 127 — pressure in lead burning, 153 regulators, 95, *96, *98, *99, *100, *101, *102 — - — used in Thermalene welding torches, 72 — pressures for cutting torches, 78 Davis-Bournonville welding torches, 59 Imperial oxy-hydrogen torches, 62 three-way welding torches, 63 Oxweld torches, 68 — — welding torches, 68 — Prest-0-Lite welding torches, 61 — Thermalene welding, 72 welding torches, 59, 61, 62, 63, 68, 72 used in cutting, 83 — torch, field of, 7 welding and cutting outfits, 116, *117, 119, *120, *122, *126, *127, *128, *129 INDEX 433 Gas toreli welding speed, 61, 68 — torches used for welding, 54, *55, *56, *57, *60, *62, *66, *67, *G9, *70, *71, *72 Gases, calculating amount of weld- ing, 63, 64 — , explosive limits of, 5 — , ignition temperatures of, 6 Gasometer, Oxweld, size and capac- ity, 40 General Electric Co., 292 — Welding and Equipment Co., 84 welding torch, *60, 61 Generator, Oxweld acetylene, port- able tj-pe, *39 Generators, Oxweld duplex, *43 — , — , sizes of, 40 — , Thermalene, 45, *46, *51, *52, 53 German silver welding, 186 Gold welding, 186 Goldschmidt, Hans, 317 — Thermit Co., 318 Goggles for gas-torch work, 120, *121 Grating, welding, 155, *162, 163 Great Western cutter, *288 Cutting & Welding Co., 279, 288 Grinding machine for rail work, 400, *401 , use of, 187 Grooved cylinder ready for welding, *188 Grooving with an air chisel, *187 Guide for cutting, *262 — , yoke welding with Thermit, *352 Guides for welding, *310 H Hales, Stephen, 9 Hastings, G. A., 215 Heat of Thermit, 318, 319 — - units in acetylene, 3 Heating, improper, *163 — torches, *157, *158, *159 , using, *157 Helmbacher Eolling Mill Co., 416 Henderson Motorcycle Co., 227 High speed tips, welding, 213, *214, 215, *216, *217, *219 welded to machinery steel with Thermit, 418, *419, *420 History and nature of Thermit, 317 Hole, filling a large, 176 — , blowing a, through a plate, *261 Holder, crucible, for locomotive work, *355 Holding the gas torch, *132 Holes, filling, 142 Holograph cutting machine, *285 Hooks, making large, *265 Hose, color of, 101 Howard, H., 201 Hydrate Engineering Corp., 127 Hydrex gas flow indicator, 128, *130 Hydrogen and acetylene flames compared, 13 oxygen flame, heat of, 13 , rate of electrolytic pro- duction, 16 — by the electrolytic method, 13 — compressed air flame character- istics, *114 — cylinders, pressure of, 15 , size and weight of, 15 — electrolytic, purity of, 14 — , explosive limits of, 6 — gas, 4 — , ignition temperature of, 6 - — , specific gravity of, 6 Ignition temperature of acetylene, 6 city gas, 6 — — — gases, 6 Igniting Thermit, 317, 326, 344 Illinois Central E. E. shops, 421 Illuminating gas, 5 Imperial Brass Mfg. Co., 86, 259, 260 — cutting torches, 86, *88 — decarbonizing outfit, *304 434 INDEX Imperial preheating torch, *158 — three-way gas outfit, *111 welding torch, 63 — welding torch, *62, (53 Improper heating, *163 Injector type gas torch, 54, *55 Inlet pipe, Liberty motor, welding, *234 Insert rail welds, 392, *393, *396, *397, *398, *399 Instructions for lead burning, 150 International cell, capacity of, 20 -- cells, group of, *21 — Oxygen Co., 19 generator, 18 Ireland & Mathews Mfg. Co., 236 Iron table with firebrick top, *156, *159 Jaw, locomotive frame, welding with Thermit, *348 Jeweler's torch, *129 Jig for welding tool tips, *219 Jigs and fixtures for welding, *219, 221 *222 *223, *224, *225, *226, *227, *228, *229, *230, *232, *233, *234, *235, *236, *237, *238 Jottrand oxygen jet cutting patent, 3 Journal of Acetylene Welding, 176 K Kautny, 4 Keeping track of costs, 302 Keithley, F. N., 369 Kerf of a cutting torch, 75 Kerosene preheater, *157 Kettle, welding a large, 192, *193 Kinds of Thermit, 320 Kirk, J. W., 274 Knuckle machine repaired with Thermit, *419 Ladle hooks, making large, 265 Lap joints, lead, 151 Lathe bed repairs, *195 Lavoisier, 9 Layout of acetylene plant, ' ' Navy type, *34 welding shop, 295, *297 Lead burning, 125, 126 data, 153 , gas consumption in, 153 , — pressure used in, 153 — — instructions, 150 — — outfits, *126 • sticks, or rods, 151 — welding or "burning," 181 Leaks, testing for, 105 Learning to weld with a gas-torch, 131 Le Chatelier, 2 Le Ehone motor, 237 Levin, I. H., 24 — generator for oxygen and hydro- gen, 24 Liberty motor manifold work, *236, *237, *238 work, 231, *233, *234, *235 Lighting low-pressure torch, Oxweld, 109 — the torch, 104, 109 Lincoln Motor Co., 231 Linde Air Products Co., 9, 312 — , Carl, 2 — process, 2 Lining Thermit crucible, *333, 334 Links, chain, building up, 205, *207 Lithium chloride, 174 Liquid air process, oxygen by the, 9 Lloyd tube welding patent, 250 Locating cracks, 364 Locomotive crosshead, welding with Thermit, *351 — frame, heating zones on, for Thermit welding, *347 — — leg, welding, with Thermit, *348 splice, welding with Thermit, *349, *353 , welded, *199 — — welding, *200 with Thermit, *347 - — mud ring with Thermit, *350 INDEX 435 Locniiiotivr vnokcr shaft, weldini;- with 'riuMiiiit, ''''.'i5'2 — wheel weliling with Thermit, *o54 Low-pressure acetylene generators, operation of, 41 torch, lighting the, 109 — — welding torch, 54, *55 torches, Oxweld, *66, *67, *69, *7() Ludwick, Herbert V., 218 M Maehine cutting torches, *76, 78 — for circular seams, *244 facing pipe ends, 322 welding oblong seams, *245 — tools, welding, 193, *194, *195, *196 — , torches for welding, *253 — welding torches, "57, *64, *7() Machinery steel welded to high speed steel with Thermit, 418, *419, *420 Machines for cutting, *278, *279, *280, *281, *282, *283, *284, *285, *286, *287, *288, *289, *290, 291, *292, *293, *294 welding, 239, *240, *241, *242, *243, *244, *245, *246, *247, *248, 249, *250, *251, 252 Macleod Co., The, 11 — Co. 's cai-bide feed, 36, *38 Magnesia stone thimbles for Ther- mit crucibles, *333, 336 Magnesium powder for igniting Thermit, 317 Magnetograph cutting maehine, 284, *285 Making allowance for expansion and contraction, 154 Malcher, L. M., 208 Malleable iron welding, 181 Manganese dioxide, uses of, 10 — steel welding, 186 Magnesia tar used for Thermit crucible, 334 Manifold welding, *226, *227, *230, *232, *236, *238 Manifolds, *118 Martin, T. O., 421 McCormack, calculations of, for specific gravity, 6 McManamy, Frank, 309 Melting points of various metals, 170 Messer cutting tools, *84 — Mfg. Co., *70, *71, 82 — welding torch, *70 Metals and Thermit Corporation, 318, 422, 425 — , commonly welded, properties of, 170 Metropolitan Railway of Paris, rails welded for, with Thermit, 411 Miller, S. W., 162 Modified Clark joint, *407, *408, 409 Moisson, H., 1 Mold for pipe welding, *324, *325, *328 welding high speed and ma- chinery steel with Thermit, "419, *420 — box, details of Thermit, *340 — , ramming Thermit, 340 — , Thermit, for heavy sections, *33S Molds, Thermit, *323, *324, *325, *328, *338, *342, *348, *349, *350, *367 — , wax-pattern, *338, 339 Monel metal welding, 183 Motor case welding with Thermit, *410, Mil — cylinder welding, 164, *165 Motorcycle manifold welding, *226. *227 Movement for welding, *133, *137, *147, *148, *149 Mud ring, locomotive, welding with Thermit, *350 Musick, W. J., 416 N Nail machine repaired with Ther- mit, *418 436 INDEX Napolitan, F. J., 267, 269 Nashville, wheel shaft weld on, *386, 389 National Safety Council rules for gas-torch users, 305 ' ' Navy type ' ' acetylene generator, 32 generators, size of, 35 — plant layout, *34 Necks on pinions, welding new, 374, *376, *378, *380, *381, *382, *384 Neutral flame, 3 New York Shipbuilding Yards, work in, *281 Nickel steel welding, 186 — welding, 183 "Nicking" billets, 262 Nitrogen and oxygen, separation of, 10 ■ — , boiling point of, 10 North American Mfg. Co., 160 — — ■ preheater, *159 O Olympia, welded anchor davit of, *385, 389 Operation rules, 306 Osceola, wheel shaft weld on, *387, 390 Outfits for welding and cutting, 116, *117, 119, *120, *122, *123, *126, *127, *12S, *129 Outlet pipe. Liberty motor, weld- ing, *235 Oval hole cutting machine, *288 Oven, Cooling, Wiederwax, *161 Oxidation and conductivity, 169 Oxide, how to deal with, 171 Oxweld Acetylene Co., 150, 208, 209, 236, 247 — — Co.'s portable pressure-type acetylene generator, 36, *39 • — cutting data, 83 ■ — cutting machine, *278 torches, *79 • — duplex acetylene generators, *43 — gasometer, size and capacity, 40 Oxweld low-pressure generator sizes and capacities, 44 torch, *66, *67, *69, *70 — outfit, a complete, *123 — preheating torches, *157 — pressure type acetylene gener- ators, 40 — regulators, 95, *96, 97, *98, *99, *100 — rivet-head cutting torch, *80 — sheet-metal welding torch, *69 — torches, gas pressures for weld- ing, 68 — water-cooled welding torches, *69, *70 Oxy-acetylene flame characteristic, *107 temperature, 3 — welds, strength of, 311 Oxygen-acetylene-hydrogen welding pressures, 63 — , amount obtained from cholorate of potash, 10 — and hydrogen electrolyzer plant, *22 rate of electrolytic pro- duction, 16 — — nitrogen in the air, 9 — , separation of, 10 — by the electrolytic method, 13 liquid air process, 9 — , chlorate of potash process, 10 — cylinder pressure, 9 — cylinders, sizes of, 9 — , discovery of, 9 — , electrolytic, purity of, 14 — generator, chemical, *11 , International, 18 — illuminating gas flame character- istics, *114 — jet cutting patent, 3 — , Linde method patent, 2 — , liquid, boiling point of, 10 — manifolds, *118 — , purity of, 10 — regulator, details of, *96, 97 — , using, for carbon burning, 302 INDEX 437 Oxygraph cutting machine, 292, *293, *294 Oxy-hydrogeu cutting, 274 — cutting pressures, 86, 87 — flame characteristics, *113 , temperature of, 4 — for cutting, 13 — welding flame, uses of, 13 pressures, 62 torch, *71, *72 Patents on Thermit, 317 Patterns for Thermit mold gate and riser, *341 Phelps, C. C, 236 Picard, 2 Pinion, a Thermit welded, *373 — , preheating, for Thermit welding, *371 — teeth, replacing with Thermit, 365, *367 Pinions, welding new necks on, 374, *376, *378, *380, *381, *382, *384 Pipe facing machine, *322 — mold for welding vertical pipe, *328 — welding, *222, 223, *228 — — materials, *323 mold, *324, *325, *328 outfit, 324 with Thermit, cost of, 424 — welds, cost of Thermit, 329 , strength of, Lindc tests, 312 Pit for Thermit welding roll and pinion necks, *376 Pittsburgh Steel Co., 418 Planer bed repair, *194 Plant layout for electrolyzers, *22, *23 Plastic and fusion welding with Thermit, 319 — process welds, 322 Plate; speed of welding, 204 Phimley, Stuart, 267 Pneumatic chisel for welding work, *187 Pods, building up, 205, *206 Poison-gas containers, welding, *230, *232 Portable electric blower type of pre- heater, *158, *159 — pressure type acetylene gener- ator, 36 Position for cutting, *258 Positive-pressure acetylene gener- ator, 29 — ■ • , first one made, 29 Potassium bisulphate, 174 — chloride, 174 — fluoride, 174 — hydroxide, uses of, 17 — sulphate, 174 Pouring Thermit into pipe mold, *328 Preheater, electric blower type of, *158, *159 — , North American, *159 —, Tyler, 158 — , Wiederwax, *161 Preheaters, *156, *157, *158, *159, *161 — , gasoline and kerosene, for Ther- mit work, *423, 424 Preheating and welding method, *191 — aluminum, 175 — for Thermit welding, *371, 42'1, *423 rail welds, 397, *399 — motor cylinders, *165 — Thermit mold, 343 — torches, Oxweld, *157 — zones indicated, *155, *162, *163, *165 Preparing Thermit mold, 343 Pressure gages, gas, 95, *96, *97, *98, *99, *100, *101, *102 — , gas, for welding torches, 59, 61, 62, 63, 68, 72 — of carbo-hydrogen for cutting, 89 gas in lead burning, 153 oxygen in cylinders, 9 Pressures, gas, in cutting, 83 • — • used for underwater cutting, 94 in three-way gas system, 87 . 438 INDEX Prest-0-Lite tool welding practice, *216 — welding torch, 58, *60 Priestly, 9 Production of oxygen, 9 ■ welding gases, 9 Propeller blade beveled for welding, *188 Properties of metals commonly welded, 170 Pryor, Frederick L., 329 Paget Sound Navy Yard, 94 Pulley, welded in 12 places, "198, 199 - — , welding, *163 Pump, building up worn parts of, 205, *206 Punch press frame repairs, *194, *195, *196 Purifier, acetylene, Oxweld, size, 40 Purity of oxygen, 10 "Putting on" metal, 143 Pyrograph cutting machine, *289, *290, 291, *292 Quantity of Thermit to use, 345 wax used for Thermit weld ing, 346 R Rack-feed cutting machine, *280 Radiagraph cutting machine, 280, *281 Radius cutting attachment, *263 Rail bonding, 204, *205 — grinding machine, 400, 401 — joints, compromise, welding, 403, *405 — preheater for heating four joints at once, *423 — specifications for difference in height, 391 — welding for electric systems, 391 — weld patterns, *394, *395, 404 — welds, insert, 392, *393, *396, *397, *398, *399 Railograph cutting machine, *282 Railroad Thermit, composition of, 321 Railway Administration welding rules, 309 — , K. C. Southern, 200 Raleigh Iron Works Co., 414 Ramming Thermit mold, 340 Ecactions, 416 Reamer, inserted blade, made by Thermit process, 419, *421 Rego cutting torch, *90 — welding torch, 64, *66 Regulator attached to gas-cylinder, *103 — connection adaptors, *103 Regulators, acetylene, *100, *102 — , gas, Davis-Bournonville, *101, *102 —, Oxweld, 95, *96, *97, *98, *99, *100 —, oxygen, 95, *96, 97, *98, *99, *101 Restrained weld, *162 Richards, Joseph W., 318 Richardson, Capt. D., 144 Riser patterns for Thermit molds, *341 Risers, cutting steel, 85 Rivet cutting, *261 Rivet-head cuttmg-torch, Oxweld, *80 Rochester Welding Works, 162 Rocker shaft welding with Thermit, *352 Rod, size of, for steel, 184 — , using the welding, 137, *138, *147, *148, *149 — , welding, for cast iron, 178 Rolled aluminum, purity of, 173 Roller, welding sheet-metal, *227 Roll neck welding with Thermit, *367 — pods, building up, 205, *206 Rolling mill base repaired with Thermit, *418 Rolls and pinions, welding new necks on, 374, *376, *378, *380, *381, *382, *384 INDEX 439 Rolls, arrangement of, for tube weld- ing, 251, *252 Koot & V'auilervoort Engineering Co., 218 Roulleau, M., 145 Rules, equipment, 305 — for operation, 306 welding, U. S. Railway Ad- ministration, 309 Rudder frame ready for welding, *201 repair. 208 Rules, safety, for gas-torch workers, 305 S Safety rules for gas-torch workers, 305 Schneider works, cutting heavy plate in, *283 Seam, circular, welding, 244 — contraction, allowing for, *135, *136, *137 — , oblong, welding, *245 Separation of elements, 173 Shaft welding, *222, 223 with Thermit, V-blocks for, *361, 362 Shear arm, welding, *167 — jaw burned on with Thermit, *415 Ship plate cutting, *263 Shop layout, 295, *297 Silver, German, welding, 186 — welding, 186 Sizes of oxygen cylinders, 9 Smith, Elmer H., 274 — , F. M., 247 Sodium bisulphate, 174 — chloride, 174 — fluoride, 174 — hydroxide, use of, 10 — — , uses of, 17 — sulphate, 174 Sawing oft" end of roll neck previous to Thermit welding, *375 Special steel welding, 185 Specific gravity of gases, 6 — heat of various metals, 170 Speed, carbo-hydrogen cutting, 89 — , cutting, of gas torch, 83 — , machine, for welding tubes, 255 — of cutting, 263, 273 — — — steel risers, 85 underwater, 94 — ■ ■ — ■ machine cutting, 280, 283, 284, 286, 289, 291, 292, 294 — ■ ■ — ■ oxy-hydrogen cutting, 275, 276 plate welding, 204 — — welding steel and sheet iron cylinders, 231 — with a gas torch, 61, 68 Splice, locomotive frame, welding with Thermit, *349, *353 Spreader disk for tube welding, *251, *252 Square liole cutting machine, *288 Standard Parts Co., 417, 419 Starting a cut, *258, *259 Staybolt cutting torch, Oxweld, *81, 82 Steel, high speed and alloy, welding, 185 — — — , welding, 213 — , Thermit, composition of, 318 — to cast iron, 179 copper, welding, 180 — welding, 183 Sternpost welds, *387, *388, *389 Stone & Webster Corp., 92 — , Geo. M., 416 Storage battery burning, 152, 153 Straight-line cutting machines, *278, *279, *283 Strength of oxy-acetylene welded pipe, 312 welds, 311 Thermit welds, 318, 329, 331 — — welded tank, 203 Sulphates of sodium and potassium, 174 T Table for welding work, *221 — , iron, with firebrick top, *156, *159 Tables, welding, *300 440 INDEX Tank and hose colors, 101 —, welded, *203 — , — , strength of, 203 — welding jig, *229, 231 Tapping Thermit crucible, *334 Temperature of oxy-acetyleue tlanie, 3 oxy-hydrogen flame, 4 ■ Thermit, 318, 319 Thermalene flame, 3 Tensile strength of various metals, 170 Tested, kinds of welds, *310 Testing for gas leaks, 105 Tests of Welding Committee, *310, 311 Thermalene, advantages of, 52 — cartridge, 46, 47, *48, 50 — Co., 5, 45, 244, 246 — , composition of, 45 — , discoverer of, 5, 45 — , explosive limits of, 6, 52 — , first description of, 45 — flame, temperature of, 3, 51 — gas welding pressures, 72 — generators, 41, 45, *46, *51, *52, 53 — , makers of, 5, 45 • — , production of, 49 — , properties of, 51, 52 — , specific gravity of, 6 Thermit additions, use of, 400 — , amount of; used for roll and pinion work, 379 — , — to use, 345 — crucible, 333 — crucibles, details of, *333, 337 — , history of, 317 — , igniting, 317, 326 —, kinds of, 320 — molds, *323, *324, *325, *328, *338, *342, *348, *349, *350, *367 — patents, 317 — pipe welding, 322 — plain, railroad and cast-iron, 320 — plastic-process welds, 322 — reaction, 318 Thermit steel, composition of, 318, 320, 321 • — , temperature of, 318, 319 — , the two methods of using, 319 — welding ' ' don 'ts, ' ' 353 — welds, strength of, 318, 329, 331 Theoretical proportions of oxygen and acetylene, 2 Thimbles, magnesia stone, for Ther- mit crucibles, 333, 336 Third rail welding, *409, 411 Three-way gas system, pressure used m, 87 — welding torch. Imperial, 63 Tip, preparing to weld, *214, *216, *217, *219 Tips, cutting, *77 — for welding torches, *58, *60, *67, *70, *71, *72 — , welding on high speed steel, 213, *214, 215, *216, *217, *219 Tire, welding, for truck, 190, *192 Tool welding, 213, *214, 215, *216, *217, *219 practice of Eoot & Vander- voort Engineering Co., *217, 218 Tooth pattern, making a wax, *367, 368 Torch arrangement on welding ma- chine, *241, *243, *244, *245, *251, *252 — , carbon electrode and oxygen jet, *274 — , cutting, action of, 257 — , — With the hand, 257, *259, *260, *261, *263, *266, 267, 269 — , gas and air preheating, *158 — , how to hold the gas, *132 — motion, *133, *137, *147, *148, *149 Torches, combination for welding and cutting, *91, *92, *93 — , cutting, 74, *75, *76, *79 —, heating, *157, *158, *159 — , water-cooled machine, for weld- ing, *253 *258, *262, INDEX 441 Torchwold cutting torch, *93 — Equipment Co., 93 Trouble, sources of in welding, 140 Truck, car, welded with Thermit, *412 — , motor, welding, *192 Tube, samples of welded, *254 — welding machines, *246, *247, *248, 249, *250, *251 Tyler, Mfg. Co., 159 Types of acetylene generators, 28 welding torches, 54, *55 Typical oxy-acetylene cutting unit, *117 Tyler, preheater, *158 U United Railways and Electric Co., Thermit welded joints for, 409 Universal cutting machine, *292 V V-blocks for shaft welding with Thermit, *361, 362, *363, *364 Valves, back-pressure, *124, *125 Vaporization of substances, 172 Vanadium steel welding, 186 Vautin, Claude, 317 Vertical pipe, welding mold, *328 — welds, 142 Volumes of oxygen and acetylene used, 3 oxy-hydrogen, 4 W Waclark "Wire Co., 417, 418 Water-cooled cutting torches, Davis- Bournonville, *76 welding torches, Davis-Bourn- onville, *57 torches, Oxweld, *69, *70 — , dissociation of, 4 — jacket, tacking, for Liberty motor, *233 Wax-pattern molds for Thermit welding, 338, *339 — tooth pattern, *367, 368 Weiderwax, preheater, *161 Weight of oxygen cylinders, 10 Thermit apparatus, 422 various metals, 170 Weld, restrained, *162 Welding aluminium, 173 - — • and cutting, field of, 7 outfits, 116, *117, 119, *120, *122, *126, *127, *128, *129 preheating method, *191 — backward, 144, *147, *148, *149 — Committee tests, *310, 311 Welding Engineer, 200, 213, 274 — gases, estimating amount, 63, 64 , explosive limits of, 5 — jigs and fixtures, 221 — jobs, examples of, 187 — machines, 239, *240, *241, *242, *243, *244, *245, *246, *247, *248, 249, *250, *251, 252 — motion, *133, *137, *147, *148, *149 — outfit on hand truck, *35 — ■ portions for Thermit welding of rectangular sections, 345 — rod, using the, 137, *138, *147, *148, *149 • — shifts on large work, 213 — shop layout, 295, *297, *298, *299, *300 — speed with a gas torch, 61, 68 — torch, Airco-Vulcan, *91 , first, 2 • — - — , low pressure, 54, *55 , Messer, "70, *71 , Milburn, *91, 92 made by General Welding & Equipment Co., *60, 61 , positive-pressure, 54, *55 , Prest-0-Lite, 58, *60 , Rego, 64, *66 — — , Thermalene, *71, *72 — torches, 54, *55, *56, *57, *60, 442 ■ INDEX *62, '66, *67, *69, *7(), *71, Wheel, driviug, welding with Ther- *72 mit, *354 Welding torches, Davis-Boiunonville, — shaft welds, *;)86, *3S!) 55, *56, *57 White metal welding, 186 , Imperial, *62, 6;! IVillium Ilciiri/ Mack, Sternpost , machine, *57, *69, *70 weld on the, *3S8 , Oxweld low-pressure, *66, Willson, T. L., 1 *67, *69, *70 Wohler, Frederick, 317 , t^-pes of, 54, *55 Wolf, Linus, 5, 45, 244 — various metals, 169, 173 Wrought iron welding, 186 Wolds, l)uilt-up, *141, 142