TT 
 
 2U 
 B5 
 
 UC-NRLF 
 

 
 LIBRARY 
 
 OF THE 
 
 UNIVERSITY OF CALIFORNIA. 
 
 Class 
 

 
 Autogenous 
 
 of Metals 
 
 THE BOILER MAKER 
 
 17 Battery Pi 
 
 1 Qftft 
 
Autogenous Welding 
 of Metals 
 
 A description ot the application of Auto- 
 genous Welding to the manufacture of 
 Tanks ; Gasometers, Receptacles for Liq- 
 uids or Gases, with or without pressure ; 
 Steam and Hot Water Boilers, Kettles ; 
 Small Boats; Automobiles; Piping, either 
 steel, copper or brass ; and Coils of all 
 kinds ; and also its application to Repair- 
 ing old or new Castings injured through 
 such defects as blow- holes, cracks, etc. 
 Its application to the Manufacture of 
 Steel, Brass, Bars and Plates, and to the 
 Destruction of Metals, Structures, etc. 
 
 Translated from Reports of the National School of Arts and Trades 
 of France 
 
 By L. L. BERBER, M.E 
 
 L*AAIIV* 
 
 F THE 
 
 N1VERSITY 
 
 NEW YORK 
 
 THE BOILER MAKER 
 
 17 Battery Place 
 1908 
 
GENERAL 
 
 Copyright, 1908 
 THE BOILER MAKER 
 
Autogenous Welding of Metals. 
 
 CHAPTER I. 
 
 HIGH TEMPERATURES FOR INDUSTRIAL PURPOSES OBTAINED BY 
 MEANS OF BURNERS. 
 
 DEFINITIONS. 
 
 The so-called oxyacetylene welding of metals consists in the 
 assembling, by means of more or less complete melting of 
 metallic pieces of the same nature, the surfaces of which are 
 brought in contact at a high temperature, without interposition 
 of a different metal from that constituting the pieces. 
 
 By extension, the name "oxyacetylene" has been applied to 
 welding made between pieces of different metals which can, 
 however, form a resisting alloy. On the other hand, the in- 
 terposition of another neutral metal capable of forming a 
 more or less perfect assembling between the two similar or 
 different metals in presence constitutes a "brazing," and tin 
 soldering is among that class. 
 
 The melting points of the various metals being very dif- 
 ferent, the oxyacetylene wdding of a certain number of 
 metals (lead, for instance) has been easily accomplished for a 
 number of years, while its application to iron, steel, brass, etc., 
 has been more difficult and required entirely different pro- 
 cesses. 
 
 We shall consider these latter processes only, and if we 
 were to refer to the definition of oxacetylene welding we 
 would have to establish the following classification: 
 
 Welding with a blacksmith's forge. 
 
 i 
 
 17928.3 
 
Hydrothermal welding. 
 
 Electrical welding. 
 
 Oxyhydric blow-pipe. 
 
 Oxyacetylenic blow-pipe. 
 
 Illuminating gas and oxygen blow-pipe. 
 
 However, the processes using gas burners affording numer- 
 ous advantages to the ordinary work of the various industries, 
 only the following processes are actually to be considered : 
 
 Oxyhydric burner (oxygen and hydrogen). 
 
 Oxyacetylenic burner (oxygen and acetylene). 
 
 Oxygas burner (oxygen and illuminating gas). 
 
 COMPARISON BETWEEN THE VARIOUS TYPES OF BLOW-PIPES. 
 
 The oxygen and hydrogen used in the oxyhydric burner may 
 be obtained from the tanks in general use in the trade, in 
 which these gases are compressed under a pressure of 300 to 
 i, 800 pounds per square inch. They may also be generated on 
 the spot by the electrolysis of water, and then compressed 
 under a pressure of 300 pounds, to be then distributed to the 
 burners through appropriate pipe lines. 
 
 In general, this latter principle is not applied because of the 
 costly apparatus, the experienced personnel and the expensive 
 maintenance which it requires. Moreover, the decomposition 
 of the water into its elements and the fact that the hydrogen 
 and oxygen pipe lines must of necessity run close to each 
 other, constitute a source of danger and render impossible the 
 production of these gases in the consumer's plant. 
 
 As a rule all the oxyhydric burners in use are supplied from 
 ordinary tanks containing the. gases under pressure. The 
 actual average prices are: oxygen, 4 cents per cubic foot; 
 hydrogen, about I cent per cubic foot. 
 
 For Oxyacetylenic burners commercial oxygen is used ; the 
 acetylene comes from one of the following sources : Dissolved 
 acetylene in tanks, where the gas is dissolved in acetone, 
 impregnating a porous material and under an average pressure 
 of 150 pounds ; acetylene generating apparatus, producing the 
 gas on the spot under a pressure of about 10 pounds when- 
 
 2 
 
ever required. The actual average prices are as follows: 
 Oxygen, 4 cents per cubic foot ; dissolved acetylene, 2 cents 
 per cubic foot. The cost price of acetylene produced on the 
 spot by generating apparatus is about I cent per cubic foot. 
 
 Oxygas blow-pipes are supplied with ordinary commercial 
 coal gas from the distributing pipes of special tanks. In the 
 following calculations illuminating gas has been reckoned at 
 the rate of $1.25 per thousand cubic feet : 
 
 Oxy acety- 
 lene Oxyhydric Oxygas 
 Mixture. Mixture. Mixture. 
 
 Temperature obtained by com- 
 plete combustion of the mix- 
 ture about 6,219 4> I 56 3^3 2 
 
 Number of B. T. U. obtained 
 by complete combustion of i 
 cubic foot of combustible gas. i ,5 70 . 290 . 61 6 . 
 
 Quantity of pure oxygen theor- 
 etically necessary for the com- 
 plete combustion of i cubic 
 footof combustible gas. cu. ft. 2 -5io 0.620 0.923 
 
 Quantity of pure oxygen fed to 
 the blow-pipe per cubic foot 
 of combustible gas (results of 
 experiments) cu. ft. 1-300 0.250 0.670 
 
 Respective quantities of the 
 gases to be fed to the blow- 
 pipe to obtain 1,000 B. T. U. 
 
 cu. ft. A\666 H 3 .8 4 6 Gi.Si 
 
 O .866 O .96 O i. 21 
 
 Cost price of 1,000 B. T. U. 
 (according to above prices): 
 
 Dissolved Ac cts. 4 . 797 
 
 Generator Ac cts. 4.13 7.69 5.066 
 
 The above figures will be very useful in making comparisons 
 hereafter. 
 
From the point of view of their use the oxyacetylene weld- 
 ing processes may be classified as follows : 
 
 Processes admitting of the easy transportation of the welding 
 apparatus to the places where the work is to be done ; dissolved 
 acetylene welding; oxyhydric welding. 
 
 Processes necessitating the transportation of the piece to be 
 welded to a fixed welding station; oxyacetylene welding, the 
 acetylene being supplied by a generator ; illuminating gas 
 welding. 
 
 PORTABLE OXYACETYLENE WELDING APPARATUS. DISSOLVED 
 ACETYLENE. OXYHYDRIC. 
 
 (a) Comparative Cost Prices of the Two Systems. 
 
 The cost price of 1,000 B. T. U. can easily be established, 
 knowing the price of the gases used and the quantity of prac- 
 tically pure oxygen fed to the blow-pipe for the combustion of 
 i cubic foot of combustible. Figuring on 4 cents per cubic 
 foot for oxygen, I cent per cubic foot for hydrogen, and 2 
 cents for dissolved acetylene per cubic foot, it will be found 
 that 1,000 B. T. U. (acetylenic, dissolved acetylene) cost 4.797 
 cents ; 1,000 oxyhydric B. T. U. cost 7.69 cents. 
 
 But the cost of workmanship must also be taken into ac- 
 count in the fixing of the cost price of the welding. In con- 
 sequence of the lower temperature of the oxyhydric mixture 
 the time required to melt the metal is naturally longer with 
 this process than with the acetylene, and the result is that for 
 a given amount of work the cost of workmanship is higher. 
 The following curves (Fig. i and 2), obtained from the 
 results of numerous trials, clearly show this difference. 
 
 The superiority of each one of the constituting elements of 
 the cost price of the acetylenic mixture over that of the 
 oxyhydric mixture is also apparent, and, greater still, in the 
 total of this price. Thus experience has shown .that the cost 
 of welding sheet metal of ^4 inch thickness by the dissolved 
 acetylene process is only half of that of the oxyhydric mixture, 
 (fr) Comparison of the Two Processes from the Points of 
 View of Easiness of Handling and Applications. 
 
 4 
 
When buying an apparatus, however good and economical 
 its principle may be, it is very important to consider the ques- 
 tion of how easily the necessary force of men will be trained 
 to its use, and to be able to indicate the well-defined mode of 
 its regulating. In the oxyacetylene process of welding, more 
 
 2 3 4 6 6 7 8 9 
 
 Thickness of sheets to weld In millimeters. 
 FIG. 1. 
 
 10 
 
 than in anything else, the regulating is of supreme importance, 
 as the quality of the finished work is essentially dependent on 
 it, because an excess of either the combustible or of the sup- 
 porter of combustion may cause an alteration of the metal and 
 its properties. 
 
 In the course of a recent dispute the partisans of the oxyhy- 
 dric mixture announced the opinion that the oxyacetylene 
 blow-pipe produced brittle welds. The inaccuracy of this 
 
 n 
 
9.00 
 
 8.00 
 
 Curve of total cost, gas and 
 workmanship. 
 Curve of cost in gas, 
 (workmanship not included.) 
 
 345 67 8 9 10 
 
 Thickness of sheets to weld in millimeters. 
 
 FIG. 2. 
 
assertion was demonstrated, particularly at the Society of the 
 Civil Engineers of France (see official report of the meeting 
 of March 16, 1906), and with the corroboration of extensive 
 actual trials it had to be admitted that the oxyacetylenic as 
 well as the oxyhydric mixture gave perfect results, the 
 defects noted with one or the other systems being due to a 
 defective regulation of the flame. 
 
 The combustion of the oxyhydric mixture produces a flame 
 of very small illuminating power, the purple and red color of 
 which it is almost impossible to describe. In the proportions 
 required to obtain a perfect weld, the flame of the oxyacety- 
 lenic mixture is characterized by a white central part very 
 neatly outlined, surrounded by an almost colorless flame. The 
 least variation in the proportion of the combustible gas and the 
 gas supporting the combustion, immediately modifies the shape 
 and the color of this central part, noticeable even to the most 
 inexperienced eye. This remarkable property of the oxy- 
 acetylenic flame is applied to the regulating of the blow-pipes, 
 and enables anybody to discover at a glance the proper or 
 defective composition of the mixture. 
 
 Unfortunately, this sure and practical method of regulating 
 cannot be applied to the oxyhydric mixture, which has no 
 well defined characteristic flame, and the excess of one or the 
 other gases causing variations of color that the eye cannot 
 detect. Undoubtedly, at first, the oxyhydric blow-pipe, placed 
 into the hands of a few very smart workmen, enabled them 
 to practically obtain excellent results. But as its use became 
 more general it was very soon found that in a number of 
 cases there was no welding, but simply that two damaged 
 parts of metal were merely stuck together. The problem to 
 be solved was the automatic supply of the gases in the required 
 proportion : the high pressure under which the tanks were 
 filled when put into service had to be taken into account, as 
 also the decrease of said pressure in the course of the con- 
 sumption of gas; and, finally, the variable flow required. 
 
 In the instruments then invented a regulating valve was set 
 in motion through the elasticity of reciprocating parts and 
 
 7 
 
springs. This valve was raised only a few tenths of a milli- 
 meter, and its almost capillary orifice was subject to frequent 
 partial obstructions. The working of such systems is most 
 uncertain, and the results obtained are by no means up to the 
 expectations. At the present time the problem has not as yet 
 been solved, and this difficulty of regulating the flame is cer- 
 tainly one of the reasons which retard the development of 
 oxyhydric welding. 
 
 The temperature of 6,000 degrees, produced by the oxyacety- 
 lene blow-pipe, was at first considered as an inconvenience by 
 those who thought that the metal to be soldered would itself 
 reach that temperature, which was much higher than its melt- 
 ing point, and that its properties would consequently be altered. 
 But this criticism is absolutely without foundation, because 
 when welding the metal is not in the same condition as if it 
 were in a furnace subjected to the action of the blow-pipe, 
 and the melting part, being in contact with a part that is not 
 yet melted, has not the same temperature as the flame. In the 
 case of a metal plate, for instance, if the action of the blow- 
 pipe is kept up longer than is necessary, the melted metal 
 falls in drops, making a hole in the plate, but its temperature 
 is not raised considerably over the melting point. The 6,000 
 degrees for which the acetylene blow-pipe was blamed was, on 
 the other hand, used by the acetylenists as an argument in its 
 favor. The quickness in reaching the melting point renders 
 possible the welding of thick plates and pieces of great bulk, 
 such as steel castings. This cannot be done with the oxy- 
 hydric blow-pipe, which does not cause the melting of the 
 metal on account of the low temperature of its flame, and of 
 the greater relative losses of heat by conductibility which are 
 the result thereof. The oxyacetylene blow-pipe, by its rapid 
 action, reduces to a minimum the heating of the parts adjoin- 
 ing the weld, and with it close brazings and many other repairs 
 may be made that would otherwise be rendered impossible by 
 the extensive softening of the metal under the slower action 
 of other processes. 
 
 Experience shows that for the average thickness of l /4 inch 
 
to Y$ inch the welding operation that can be performed with 
 a 50 cubic foot tank of liquid acetylene (weighing 55 pounds), 
 and a 70 cubic foot tank of oxygen (weighing 50 pounds), will 
 require at least 495 cubic feet of hydrogen and 125 cubic feet 
 of oxygen by the oxyhydric process, or four tanks of hydrogen 
 weighing 90 pounds each and one 55-pound tank of oxygen. 
 In other words, where a io5-pound oxyacetylene apparatus 
 will do the work without any taking apart of pieces, a 415- 
 pound oxyhydric installation will be necessary, and will re- 
 quire the hydrogen injector to be taken off and put back three 
 times and the oxygen injector once. 
 
 In cases where the metallurgical plants are not near to the 
 establishments producing the gases required it will be neces- 
 sary, in the selection of the apparatus, to take into account 
 the important difference in the number of B. T. U. obtained 
 per cubic foot of material which exists between the oxyacety- 
 lene and oxyhydric processes, and which considerably affects 
 the cost price of welding through the cost of transportation 
 of the tanks. ^ Conclusions. 
 
 On the whole, the dissolved acetylene process, while on an 
 equal footing with the oxyhydric process in point of expense 
 of installation and maintenance, is very superior from the 
 following points of view, which are the only ones to be con- 
 sidered in ordinary work : 
 
 Price of the gases used. ^ 
 
 Cost of workmanship. ^ Production being equal. 
 
 Cost of transportation. J 
 
 Easy regulation of the flame. 
 
 Ability to weld large pieces. 
 
 On the other hand, the oxyhydric process is to be preferred 
 to the oxyacetylene method in the welding of very thin plates 
 (less than 1/32 inch), which requires less ability on the part 
 of the workman, because its action is not as rapid, the fusing 
 speed of the mixture not as high. This speed, indeed, with 
 an oxyacetylene blow-pipe operated by an inexperienced man, 
 may cause holes in the piece that is being welded, which, 
 however, it is easy to repair. 
 
INSTALLATIONS OF STATIONARY WELDING APPARATUS. 
 
 ist. Oxyacetylene supplied by a generator. 
 2d. Oxygas Illuminating gas supplied by the ordinary dis- 
 tributing pipes. 
 
 (a) Comparative Cost Prices of Both Systems. 
 
 A preceding table has shown that the cost price of 1,000 
 B. T. U. obtained by what shall hereafter be termed "gen- 
 erator acetylene," is 4.13 cents, while the cost price of 1,000 
 oxygas B. T. U. is 5.066 cents, taking as a basis the following 
 prices: Acetylene, 2 cents per cubic foot; illuminating gas, 
 $1,25 per 1,000 cubic feet; oxygen, 4 cents per cubic foot. 
 
 The two following diagrams, Nos. 3 and 4, show the results 
 of many experiments and the economy of the oxyacetylene 
 mixture. 
 
 (b) Comparison Bctiveen the Two Processes from the Points 
 of View of Easy Handling and Applications. 
 
 The oxyacetylene flame is undoubtedly much easier to 
 regulate than that of illuminating gas ; however, with practice 
 the oxygen flame can be regulated. For this reason we shall 
 .not take into account the more or less easy regelating of the 
 flame, as we have done in comparing the oxyhydric and oxy- 
 acetylene mixtures. 
 
 The oxygas welding can, of course, only be used in cities 
 where there are coal-gas plants. On the contrary, the oxy- 
 acetylene welding may be installed anywhere. 
 
 The expense of installing an acetylene welding plant is 
 necessarily greater than in the case of coal gas, owing to the 
 price of the generator, which is not to be considered in the 
 case of a coal-gas plant. 
 
 This advantage of coal gas is counterbalanced by the incon- 
 veniences resulting from the low temperature and the great 
 volume of the flame : impossibility of practically welding 
 thicknesses of 1/6 to 1/5 inch, considerable losses of heat 
 through radiation, which, for ^ inch thickness, are a great 
 nindrance to the workman. 
 
 Besides, it must be remembered that coal gas for illuminat- 
 
 10 
 
Time necessary to weld per meter, 
 in minutes. 
 
 H^ tt-?t<3COCO'*-4-CJl 
 
 cncci l>' S e? S 55 S 
 
 
 
 
 . xi 
 
 \e^T 
 
 r// 
 
 ^3^ 
 
 ^ 
 
 
 *x 
 
 4^1 
 
 
 
 
 ^ 
 
 ^ 
 
 ^x^ 
 
 
 
 
 
 ^ 
 
 e^ 
 
 ^$ 
 
 pi> 
 
 
 
 
 
 
 ^ 
 
 
 
 
 
 
 
 
 r 
 
 
 
 
 
 
 
 
 
 1234567 89 10 
 
 Thickness of sheets to weld in millimeters. 
 Pig. 3* 
 
 st per meter to weld, in francs, 
 
 P M N- to 10 00 
 
 S 8 8 g 8 S 8 
 
 ~ Curve 
 workn 
 
 of total cost, (gas and 
 lanship.) 
 of cost of gas, (workmanship / 
 eluded.) ,/ 
 
 not in 
 
 TA 
 
 // 
 
 / 
 
 . .*/ 
 
 / 
 
 
 ? 
 
 C 
 
 & 
 
 T 
 
 s* 
 
 
 
 .0.20 
 
 
 
 
 
 
 
 
 
 123456789 10 
 Thickness of sheets to weld in millimeters 
 
 Fig. 4 
 
 11 
 
ing purposes contains many impurities, very variable in nature 
 in consequence of the combustibles used in its manufacture. 
 These impurities have a great influence on the quality of the 
 work, and are such as to preclude the certainty that the welds 
 made by this process will withstand any pressure, even if the 
 greatest caution is observed. 
 
 (c} Conclusions. 
 
 On the whole, and if the above curves of cost prices are 
 taken in consideration, the oxyacetylene process is more ad- 
 vantageous than the oxygas, except, however, as regards the 
 cost of installation. Hence the oxygas process should only 
 be preferred to oxyacetylene in cases where the cost of in- 
 stallation has to be taken into account, where the plates to be 
 united are not thicker than 1/6 to 1/5 inch, and where the 
 quality of the weld obtained is of secondary importance only. 
 
 CONCLUSIONS. 
 
 The above has shown the advantages, in portable appa- 
 ratus, of dissolved acetylene over hydrogen, and in stationary 
 apparatus, of generator acetylene over coal gas. 
 
 When should "pressure" acetylene or "generator" acetylene 
 be used? 
 
 It is evident that liquid acetylene imposes itself in all cases 
 where the possession of a portable apparatus is necessary for 
 work outside of the shop, or in the shop on pieces difficult to 
 handle, and also when the work to be done is of short duration. 
 
 The smaller initial expense and the easy, perfect control of 
 the consumption are also in favor of dissolved acetylene for 
 experiments, studies and regulation of new manufactures. 
 
 On the other hand, the cost price of the work will become 
 the most important factor in steady manufactures, and prefer- 
 ence will then be given to generator acetylene, if the plant has 
 enough room for the installation of the generator. 
 
 The generator will also be available in shops not provided 
 with a practical system of lighting, or if acetylene is desired as 
 an emergency system of illumination, as, for instance, when the 
 motors are out of commission in shops electrically lighted. 
 
 12 
 
CHAPTER II. 
 
 DESCRIPTION OF OXYACETYLENE WELDING PLANTS. 
 
 With reference to the various parts of the apparatus the 
 welding installations may be grouped in two classes : 
 
 ist. The combustible gas and the gas supporting the com- 
 bustion enter the blow-pipe under rather high pressures (in 
 general 7 to 15 pounds per square inch), in order to insure to 
 each sufficient velocity at the mouth of the pipe. This refers 
 to the following installations : Oxyhydric, dissolved acetylene 
 and generator acetylene under pressure. 
 
 2d. One of the gases, generally the combustible, is obtained 
 from a source where the pressure is only a few ounces, and 
 the other gas must flow under a pressure sufficient to draw 
 out the first and insure a proper velocity to the mixture at the 
 mouth of the blow-pipe. Such are the installations of oxygas 
 and generator acetylene without pressure. 
 
 INSTALLATIONS USING BOTH GASES UNDER PRESSURE. 
 
 In this class we can make two sub-divisions : 
 
 (a) Both gases come from tanks where their pressure is 
 higher than that needed in the blow-pipe (oxyhydric and dis- 
 solved acetylene apparatus). 
 
 (&) One of the gases is produced under a pressure nearly 
 equal to that in the blow-pipe (oxyacetylene under pressure 
 from generator). 
 
 (a) Oxyhydric and Dissolved Acetylene Apparatus. 
 
 The hydrogen and oxygen used in these processes come 
 from seamless steel tanks, in which they are under a pressure 
 of 150 pounds per square inch. The acetylene is contained 
 
 13 . 
 
in steel tanks, completely filled with a porous matter soaked 
 in acetone, in which the gas is dissolved under a convenient 
 saturation pressure of about 150 pounds per square inch. 
 
 These gases, to be used in the blow-pipe, must therefore be 
 brought to a lower pressure ; this result is attained by the use 
 of regulating valves mounted on the tanks. In the actual 
 work these regulating valves are used in conjunction with 
 
 14 
 
gages, constantly indicating the pressure of the gas in the 
 tank, and consequently how much gas is left in it. 
 
 The various valves and gages used for oxygen, hydrogen 
 and dissolved acetylene are almost similar, and differ only in 
 the metal of which they are made, but their, principle is the 
 same. 
 
 The instrument, Fig. 5, described below is composed of a cyl- 
 indrical box divided in two parts. In one of them A is' the 
 gage, in the other B the regulating valve. The gas coming by 
 a tube C passes through a chamber D lined with a filtering 
 matter, stopping the dust that might be driven in by the gas ; 
 this chamber has two orifices ; one in connection with the gage 
 tube, the other, incompletely closed by a needle point E, and 
 through which passes the gas going into the expansion cham- 
 ber. On this needle point E rests a lever F connected with 
 the center of an elastic diaphragm G. 
 
 Under this diaphragm is a spring which constantly main- 
 tains the needle point in close contact with its seat when the 
 pressure on top of the diaphragm is equal to the atmospherical 
 pressure without regard to the pressure of the gas in the tank. 
 On top of the diaphragm is a second spring /, the tension of 
 which is regulated at will by means of a screw K. Under 
 normal working conditions the pressure of the gas on the 
 diaphragm and the pressure of the spring counterbalance each 
 other. It follows that the pressure . of the expanded gas 
 depends only on the tension of the spring, and is independent 
 or nearly so of the variation of pressure in the flask. 
 
 A safety valve L placed on the expansion chamber limits 
 the pressure in that chamber to a maximum, which a defective 
 working of the needle point E might tend to exceed. A needle 
 point cock M, placed at the end of this chamber, connects the 
 regulating valve and the blow-pipe, and is used to stop the 
 flow of gas at any time without being compelled to close the 
 cock of the tank itself. 
 
 Emanating from the regulating valves, the gases are brought 
 through rubber tubes of convenient diameter to the blow-pipe, 
 where the mixture takes place, to be ignited at the mouth. 
 
 15 
 
In the type of installations now discussed both the com- 
 bustible gas and the supporter of combustion arrive in the 
 blow-pipe under very nearly equal pressures sufficient to insure 
 a speed of the mixture at the mouth of the blow-pipe slightly 
 greater than that of the spreading of the flame through this 
 mixture. Under these conditions, if the working of the appa- 
 ratus was .regular and perfect, there would be no fear of a 
 back draft inside of the blow-pipe. Relying on this theo- 
 retical impossibility of a back draft, under normal working 
 conditions, a number of manufacturers offered, until very re- 
 cently, oxyhydric blow-pipes having no safety appliances to 
 
 FIG. 6. 
 
 avoid a return of the flame inside of the apparatus. But if 
 for any reason (defective working of a regulating valve, 
 insufficient pressure in a tank, smashing or folding of a 
 rubber hose, contact of the end of the blow-pipe with the pipe 
 to be welded, etc.) the speed at the mouth became for an 
 instant inferior to that practically necessary, the blow-pipe 
 became red hot, burning the workman's hand, and the flame 
 ran back as far as the rubber tubes, which were burned; the 
 valves had to be instantly closed and everything put in shape 
 again. 
 
 In the oxacetylene blow-pipe of the Dissolved Acetylene 
 Company the safety appliance is located inside of the appa- 
 ratus ; the back draft is avoided by a particular construction, 
 by which the mixture of acetylene and oxygen convenient for 
 the welding flame takes place in a very small space near the 
 mouth of the blow-pipe. From this mouth to the source of the 
 gases the quantity of acetylene in the mixture decreases. The 
 explosive wave, coming in contact with parts of gas of de- 
 
 16 
 
creasing explosive power, slackens rapidly. The result is a 
 very reduced localization of back drafts. This appliance is 
 also completed by the interposition, in the supply pipe of 
 acetylene, of a porous screen, which effectively stops the flame, 
 a result that could not be obtained by a cushion of wire gauze, 
 contrary to general belief. 
 
 (&) Oxacetylene Installations with Generator Producing 
 Acetylene Under a Pressure of About 7^2 Pounds. 
 
 The blow-pipes employed in this case may be simply mixers 
 of gas, as above, with the condition, however, that the pres- 
 sure of the acetylene supplied by the generator be sufficient to 
 insure to the mixture a speed at the mouth superior to that 
 of the spreading of the flame (a condition which is generally 
 realized in these instruments). But the installation is, how- 
 ever, different from the others. If we suppose that for some 
 reason one of the gases, through an excess of pressure, has a 
 tendency to enter the flask containing t*he other gas, it is 
 easy to understand, by referring to the description of the above 
 regulating valves and manometers, that the construction of this 
 apparatus renders impossible the connection of one of the 
 tanks with the other. 
 
 In the case of installations with generator producing acety- 
 lene under a pressure of 7^ pounds the acetylene regulating 
 valve and manometer do not exist, and nothing would stop 
 the oxygen coming out of its tanks to get into the acetylene 
 generator. In order to avoid this possibility of accident it is 
 necessary to interpose in the acetylene pipe, between the blow- 
 pipe and the generator, a safety hydraulic interrupter. (See 
 Technologic Bulletin, December, 1903, page 1,348.) 
 
 INSTALLATIONS USING ONE OF THE GASES TO DRAW OUT THE 
 OTHER. 
 
 The acetylene produced by ordinary generators for lighting 
 purposes and the coal gas distributed in the cities are under 
 a pressure of only a few ounces of water. 
 
 The oxygen always comes from tanks where its pressure is 
 150 pounds per cubic foot. 
 
 17 
 
A regulating valve and manometer, placed on the oxygen 
 tank, reduce the pressure of this gas to 15 -.pounds, under 
 which it is fed to the blow-pipe. The latter is provided with 
 an injector, the diameter of which changes according to the 
 quantities of oxygen required. The oxygen, acting as motor, 
 brings the combustible gas in, and in this way it is possible to 
 
 regulate the speed of the flame. This is the principle of all the 
 blow-pipes of this class, the difference consisting of more or 
 less perfect details of manufacture. 
 
 The first blow-pipe using acetylene without pressure has 
 been invented by Mr. Ed. Fouche (August, 1877). 
 
 On account of the considerable difference between the pres- 
 sures of the oxygen and the combustible gas, and in order to 
 
 18 
 
avoid a flowing back of the oxygen in the pipes of the com- 
 bustible through any accident, it is necessary to interpose on 
 these pipes, and as near the welding point as possible, an 
 
 FIG. 8. 
 
 hydraulic safety valve. This appliance, Fig. 8, is composed of 
 a central tube A for the combustible gas, terminated at its 
 lower extremity by a bell-shaped casting, on the circumference 
 
 19 
 
of which are numerous holes through which the gas escapes, 
 reaching the blow-pipe through the tube S\ The valve chamber 
 is filled with water to the level of a gage cock R, and is also 
 provided with a safety tube B, the lower end of which is 
 slightly below the normal water level. The top is connected 
 with a basin communicating with the air outside by holes in 
 the cover. 
 
 In case of an accidental flowing back of the oxygen in the 
 apparatus, the water rises in the tubes A and B, and the section 
 CD of the safety tube B is uncovered ; the gas escaping outside 
 and no flowing back can occur in the central tube, which re- 
 mains immersed in water. 
 
 The water carried away is collected in the upper basin, and 
 falls back to the bottom. 
 
 BLOW-PIPES. 
 
 As previously said, all the blow-pipes used in these kinds 
 of installations are identical as to principle, the only difference 
 between them being the details of manufacture, rendering their 
 working more or less perfect and their manipulation more or 
 less safe. 
 
 The safety appliance placed on the acetylene pipe, avoiding 
 all back drafts of the flame, is one of the most important parts 
 of the welding tools. 
 
 The French company for the dissolved acetylene process 
 uses a porous material. In the Fouche system the acetylene 
 goes through a series of very long and very thin tubes. In 
 other cases an accumulation of metallic gauzes is resorted to, 
 which is more dangerous than efficacious, and others solve the 
 problem by the complete absence of the safety appliance. 
 
 Although it is not our intention to describe the various 
 systems of blow-pipes, we shall draw attention to two systems, 
 still unknown because they are very new. 
 
 ist. Warming up of the gaseous mixture before its inflamma- 
 tion. 
 
 The insufficient temperature of combustion of coal gas led 
 to appliances for heating it. The company using the com- 
 
 20 
 
pressed gases, to attain this end, warms up the oxygas mix- 
 ture by means of a flame, bringing to red heat a coil through 
 which the mixture passes. It is evident that this disposi- 
 tion insures a higher temperature of combustion, but its 
 slow action speaks against it. At the moment of lighting, the 
 coil is cold, and the heater has no action on the mixture ; its 
 action is only progressive, following the warming up of the 
 coil, so that this disposition is only interesting in cases where 
 the blow-pipe must work without stopping for a long time. 
 This disposition is of no value in the works necessitating the 
 lighting of the blow-pipe for only short periods. 
 
 2d. Blow-pipes with interchangeable heads for various sizes 
 of flame. To facilitate the work, and to reduce the consump- 
 
 FIG. 9. 
 
 tion of gas to a minimum, it is necessary to use blow-pipes 
 giving a flow of gas in proportion to the work to be done. 
 In the case of a blow-pipe where one gas brings in the other, 
 the sections of the injector and of the pipes for the gas carried 
 in, the shape, the sections, and the length of the mixing and 
 egress chambers must be well determined for a given flow. The 
 result is the necessity of making a blow-pipe for each of 
 the necessary flows. Undoubtedly, certain manufacturers have 
 pretended that they obtained flames of different volumes, in 
 which the mixture of the gases was perfect, with the samfc 
 blow-pipe, the same injector, the same mixing parts, by simply 
 changing the mouth of the blow-pipe. This assertion has no 
 
 21 
 
foundation, and experience has shown the imperfect working 
 of these blow-pipes. 
 
 By a special adjustment of its blow-pipes the B. R. C. Com- 
 pany has realized the grouping in a head easily removable of 
 all the parts, the form and section of which is variable with 
 the required flow. The result is that with only one body of 
 blow-pipe and a series of these removable heads, it is possible 
 to obtain a great variety of flows, rendering possible the ex- 
 ecution of very different classes of work. 
 
 This disposition is of value for the shops where oxyacety- 
 lene welding is seldom made, and on pieces of very different 
 thicknesses. 
 
 3d. General comparison between the blow-pipes of the first 
 class (ga>s under pressure} and those of the second class (one 
 gas driving in the other.} 
 
 Certain shops, noticing marked differences between the work 
 obtained with the same gases (acetylene and oxygen), but in 
 one case with blow-pipes of the first class (gas under pres- 
 sure), and in the other with blow-pipes in which one gas 
 forces in the other, came to the conclusion that the former 
 were superior to the latter. 
 
 This superiority, although real, is not, however, as great as 
 one might be tempted to believe, because if certain blow-pipes 
 where one gas brings the other in are not carefully watched, 
 the few types generally used in the shops admit of a complete 
 mixture of the gases and of a perfect mixture, the flame of 
 which is in all respects similar to that of the other blow-pipes. 
 
 The inferiority, if it may be so called, of the blow-pipes of 
 the second class, arises from the following fact: 
 
 When the workman starts to work and regulates the flame 
 of his blow-pipe, its mouth is at the same temperature as the 
 surrounding air. In the course of the work the diameter of 
 this mouth increases in a certain proportion on account of the 
 heat; on the other hand, particles of melted metal or oxide 
 are always projected and may obstruct this mouth more or 
 less. The result is that the volume of the issuing mixture 
 is variable during the work. 
 
In the blow-pipes where the gases come in under the same 
 pressure this modification of the diameter of the mouth has 
 no other consequence than a variation of the flow; the pro- 
 portion of the gases in the mixture does not change, and it 
 follows that the nature of the flame is not modified. 
 
 On the contrary, in the blow-pipes where one gas carries in 
 the other, the quantity of oxygen passing under high pressure 
 through the injector remains nearly constant, notwithstanding 
 the variation of the orifice area, whereas the quantity of acety- 
 lene carried is subject to fluctuations. The consequence is a 
 certain irregularity of the flame, generally hardly noticeable, 
 requiring only a closer watch on the part of the workman. 
 
CHAPTER III. 
 
 APPLICATIONS OF THE HIGH TEMPERATURES FURNISHED BY THE 
 BLOW-PIPES. 
 
 The high temperatures that are so easily obtained by the use . 
 of blow-pipes have admitted of the execution of certain in- 
 dustrial works which were before impossible. We shall not 
 speak here of the cutting up of metals, which will be the 
 subject of a future article, but shall only indicate a few of 
 the applications of oxyacetylene welding, limiting of necessity 
 our description to the most typical current work. In fact, to 
 enumerate all the cases where oxyacetylene welding has been 
 the cause of any saving it would be necessary to review all the 
 branches of metallurgic activity. 
 
 BLOW-PIPE WELDING OF STEEL PLATE. 
 
 One of the most important applications of oxyacetylene 
 welding is its substitution for riveting and bolting in steel- 
 plate work. 
 
 To assemble steel plates, and in general in all cases of as- 
 semblage by oxacetylene welding, "marking" is the first opera- 
 tion; it consists in making in various places along the pieces 
 to be assembled several "drops" of welding, in order to main- 
 tain the two parts to be united in their respective positions. 
 Then welding is done in the following way: 
 
 AB, AB' are the edges of the two plates to be united, pre- 
 viously marked (Fig. 10). 
 
 We suppose that we start welding at a, moving forward 
 toward BB'. The flame of the blow-pipe is .allowed to act at 
 the point a until it causes a fusion of the metal throughout the 
 
 24 
 
plate at a; the metal in fusion affects the shape of a drop. The 
 flame is then brought forward from a to b; at b the same 
 operation is repeated, and so forth. 
 
 In fact, the successive fusion drops a, b, c, etc., partially cover 
 each other, and after cooling form an homogeneous mass. 
 
 The exterior face of the welding alone shows the way in 
 which the work was done by a series of ridges outlining the 
 
 B B 
 
 1 
 
 c 
 
 X 
 
 ) 
 
 
 
 
 X 
 
 
 
 I 
 
 } 
 
 
 
 * 
 
 / 
 
 
 
 V. 
 
 J: 
 
 / 
 
 A 
 
 FIG. 10. 
 
 successive fusion drops. These ridges are more or less 
 apparent, according to the ability of the workman and to the 
 volume of the fusing drops, the latter being in proportion 
 to the thickness of the plate. 
 
 In the assembling of two plates by oxyacetylene welding the 
 quantity of lost metal (by oxidation, for instance,) being ex- 
 ceedingly small, the thickness along the welding line is prac- 
 
tically the same as the adjoining parts, if there is no ma- 
 terial space between the two parts to be welded. -If, on 
 the other hand, there is a space between them, it is neces- 
 sary, after the edges of both plates have been brought to 
 the melting point, to melt a rod of the metal, which the work- 
 man holds in his hand; this supplementary metal falls by 
 drops on the melted edges and increases the melted mass in 
 the proportion desired. The nature of this supplementary 
 metal varies with the result to be obtained. Very often, in 
 order to have an invisible welding line and not to modify the 
 qualities of the metal along the latter, this line is "charged" by 
 means of metal taken from the piece itself, such as strips taken 
 from the plates; more frequently, in the case of iron or soft 
 steel, the "charge" is made by means of ordinary soft steel 
 wire, Y to % inch diameter, according to the thickness of the 
 piece to be soldered. The quantity of metal so added varies 
 with the space between both parts before welding, and with 
 the excess of thickness to .be given to the welded part over 
 the adjoining metal ; naturally, this excess of thickness may 
 be as great as the particular conditions may require, but in 
 general it is very small, and a little chipping or filing is enough 
 to take away the unevenness and bring the welded part to the 
 level of the adjoining surface. 
 
 Thin plates (less than 1/16 inch thickness) to be welded 
 with the blow-pipe require no special preparation of the edges 
 before assembling. The above sketches show the various cases 
 to be considered: 
 
 Fig. ii. Angle of the plates less than 90 degrees (concave 
 bottom welded in a shell). 
 
 Fig. 12. Angle of the plates about 90 degrees. The as- 
 sembling may be made, as desired, according to sketch A 
 or B. Assembling A requiring no addition of metal is more 
 rapid. 
 
 Fig. 13. Angle of the plates larger than 90 degrees (convex 
 bottom welded in a shell). 
 
 Fig. 14. Angle of the plates = 180 degrees. 
 
 The welding of plates of more than ^ m ch thickness re- 
 
 26 
 
quires a special chamfering of the edges. These sketches 
 show various examples. In general the preparation must be 
 such as to allow the flame of the blow-pipe to penetrate to the 
 very bottom of the part to be welded ; thus, there is no doubt 
 that the entire thickness is affected. If in some cases it is 
 impossible to make such a preparation it is then necessary to 
 proceed slowly, in order to thoroughly fuse the entire section 
 
 Pig. 11 
 
 Fig. 12 
 
 s 
 
 1 
 
 _J 
 
 Fig. 13 
 
 of the metal. In this case it will be better if the plates to be 
 assembled are not in contact but are 1/16 inch apart. 
 
 Fig. 15. Angle of the plates less than 90 degrees. Prepara- 
 tion A is better than B. 
 
 Fig. 16. Angle of the plates 90 degrees. Preparations A 
 and C are better than B. 
 
 Fig. 17. Angle of the plates larger than 90 degrees. 
 
 Fig. 18. Angle of the plates = 180 degrees. 
 
 The most delicate welding is that of pieces with re-entering 
 angles. This is, however, seldom the case; but if such is the 
 case it is better, if possible, to prepare the piece according to 
 sketch 20 or 21, or, if any advantage results from it, according 
 to sketch 22 or 23, a rigid rib being obtained which prevents 
 deformations of the piece. 
 
 27 
 
Before welding 
 
 Before welding After welding 
 
 _J 
 
 L_ 
 
 Fig. 16 
 Before welding After welding 
 
 Fig. 17 
 
In general, in preparing plates to be assembled by oxacet>- 
 lene welding care should be taken not to imitate shapes pre- 
 viously requiring bolting or riveting. 
 
 In the majority of cases oxacetylene welding does away with 
 a lot of preparatory work ; calking of edges, pulling apart of 
 rivets and other fastenings, operations always expensive and 
 which are always to be avoided if possible. 
 
 Let us consider, for instance, the case of a cylindrical tank 
 
 Fig. 18 
 
 Fig. 23 
 
 with riveted bottom and head. If this tank is not of a suf- 
 ficient diameter, and is not provided with a manhole, it will 
 be necessary 'to make it with at least a convex bottom or 
 head. Anyway, its making requires a riveted cylindrical shell 
 with two drawn heads at the ends to permit the riveting of 
 bottom and head and riveted bottom and head with calked 
 edges. 
 
 The same tank can be made by oxyacetylene process with 
 solid welded heads. 
 
 29 
 
We may remark, in passing, that oxacetylene welding has 
 rendered possible the making (volume and resistance being 
 equal) of tanks less cumbersome and lighter than those used 
 before its advent, in that it has made possible the building of 
 tanks with two convex bottoms without regard to the diameter 
 
 ---800-- 
 
 a>^ 
 
 700 
 
 r 
 
 FIG. 24. 
 
 and absolutely free of the double thickness of plates necessi- 
 tated by riveted coverings. 
 
 Nearly all the tanks built to contain gases under pressure 
 or very thin liquids, such as petroleum, are now welded by 
 the oxyacetylene process, because aside from the advantages 
 of weight, bulk and price which they have over the riveted 
 tanks, they do not leak, a quality which is difficult to obtain 
 by riveting, and even with subsequent tin soldering, particu- 
 larly when these tanks are supposed to travel and are, con- 
 sequently, subject to continual rough handling. 
 
Aside from the saving which may be realized by oxyacety- 
 lene welding over riveting by doing away, in a large measure, 
 with preparatory forge work, we must also consider the 
 economy of this process of assembling in itself. 
 
 Let us, for instance, consider the case of the ordinary rivet- 
 ing together of two plates of 54 mcn - 
 
 ist. Riveting (One Line of Rivets}. 
 
 Diameter of rivets, */ 2 inch; number of rivets per foot = 8. 
 
 Price paid to the workman per foot of joint : 
 
 Laying out the holes $ .006 
 
 Marking 0066 
 
 Drilling 0294 
 
 Chamfering 003 
 
 Riveting 0192 
 
 Calking plates 0048 
 
 Calking rivets 012 
 
 Total Jo. 0810 
 
 This cost of workmanship, obtained in a part of the country 
 where the salaries are not very high, does not include the 
 general expenses arising from the necessary power, keeping, 
 etc., of the machinery (punching tool, etc.) and heating of 
 the rivets. 
 
 It follows that riveting in one line of rivets costs per foot: 
 
 Eight -inch rivets, 1.23 X 4 cents $ .04 
 
 Workmanship (without general expenses) o .08 
 
 Total $0.12 
 
 Oxyacetylcne Soldering (Generator Acetylene}. 
 
 Chamfering of edges, per foot o .0108 
 
 C oacetylene $0.0186^ 
 
 Welding < oxygen -0312 > o .066 
 
 (^ workmanship .0162 ) 
 
 Total $0.0768 
 
 31 
 
This example shows conclusively that assemblage by oxy- 
 acetylene welding is more economical than by riveting. To 
 complete our comparison we shall consider the cases of the 
 building of a vertical tubular boiler shown in sketch No. 24 
 by oxyacetylene welding and by riveting. We shall not men- 
 tion the operations, which are similar in both processes of 
 manufacture: shearing and laying out of the plates, boring 
 holes, assembling and expanding of tubes, etc. : 
 
 ist. Oxyacetylene Welding (Generator}. 
 
 f Shell, 8.5 ft. X $.0054 $ .046 
 Chamfering of edges < Furnace, 2. 925 ft. X .0072 .021 
 
 (Uptake, 5.85 ft. X .066 .038 
 
 f Shell, 4.225^. X .066 $ .278 
 Welding < Furnace, 1.462 ft. X .21 .307 
 
 ( Uptake, 2.925ft. X .12 .351 
 
 Rounding and planing after welding $ .60 
 
 Forging of furnace (uptake and mouth) 2 .40 
 
 Turning of circular plates 40 
 
 Assembling of the boiler (mounter and help) 80 
 
 Welding 32. 5 ft. @ $0.27 8.78 
 
 Total $14.02 
 
 2d. Riveting. 
 
 Necessary plate: 
 
 For shell 5 . 28 pounds 
 
 For furnace 4 . 62 pounds 
 
 For furnaces flanges 9-9 pounds 
 
 For flanges of the outer circumferential plates, 5 1 . 48, total 
 
 71.28 pounds, @ $0.25 $ 1.78 
 
 44 ^-inch rivets, 5 pounds; 275 f-inch rivets, 112 pounds, 117 
 
 pounds @ $.04 4.68 
 
 Marking rivet holes i . 40 
 
 Flanging the uptake with forge heat i .00 
 
 Closing in on furnace boiler head flanges .80 
 
 Forging the furnace (uptake and mouth) 4 .00 
 
 f Inf., 105 pounds X $0.01 i.oc 
 Closing in the flanges of the plate { 
 
 [ Sup., 132 pounds X 0.009 1.19 
 
Turning of circular plates 
 Assembling the boiler 
 
 Chipping and calking heads i .60 
 
 Riveting 
 
 Total . '. 123 . 3 2 
 
 The above prices of riveting are established on the sup- 
 position that the chamfering and calking are executed by 
 compressed air (except for the heads, which require some 
 hand work). They do not include the general expenses (ma- 
 terial, coal and coke necessary for welding the charger and for 
 the various forge work). 
 
 These results show the considerable saving obtained by judi- 
 ciously using oxyacetylene welding in boiler making, and ex- 
 plain the development cf this process as soon as it was 
 known.* 
 
 The cost price may also be made much smaller by a pre- 
 liminary warming up of the parts to be welded by means of 
 the "Hauck" burner, using a less expensive combustible than 
 the oxyacetylenic or oxhydric mixture. It is evident that in 
 every instance where the method of manufacture, the shape of 
 the pieces, the place where the work is to be done, will admit 
 of such a warming up, a great advantage will result by such 
 bringing of the parts to be welded to the highest practicable 
 degree of heat; the more expensive combustible from the 
 blow-pipe is thus used only to cause the actual welding, which 
 the cheaper modes of heating cannot effect. 
 
 TENSILE STRENGTH AND ELASTICITY OF PIECES WELDED BY THE 
 OXYACETYLENE PROCESS. 
 
 The experiments conducted on plates welded with oxyacety- 
 lene, or oxyhydric blow-pipes, show that the results are about 
 the same if the proportions required to obtain a neutral flame 
 
 * These prices being those usual in French establishments must be proportion 
 ately increased for American plants. 
 
 33 
 
have been well kept. The unfavorable results that may have 
 been obtained with either one of the methods were due solely 
 to a defective regulating of the flame. 
 
 The tensile strength of pieces welded by the oxyacetylene 
 process is practically the same as that of the metal itself, and 
 in general is rather superior. 
 
 On the other hand, the elasticity is to some extent reduced, 
 which seems natural, the welded part having been melted and 
 then rapidly cooled, whereas the adjoining parts have been ob- 
 tained by fusion, followed by slow cooling or by laminating 
 or hammering, which operations increase the malleability. 
 
 If after welding care is taken to anneal the piece, the elas- 
 ticity is restored and becomes equal to that of the metal in 
 the primitive state. 
 
 In cases where the welded pieces have a tendency to stretch, 
 it will be well, whenever possible, to anneal them after weld- 
 ' ing;. 'it will be necessary to take this precaution in such works 
 as boilers, superheaters, etc. 
 
 REPAIRS TO STEEL BOILERS. 
 
 In repairing in general, and particularly in repairing plates, 
 the use of the blow-pipe is indispensable, because it very often 
 happens that by its use pieces may be saved which otherwise 
 would have to be replaced ; this fact alone results in con- 
 siderable economy. 
 
 One of the interesting applications of the oxyacetylene 
 process is the repairing of boilers. Nearly all the work of this 
 class accomplished up to the present time has been done with 
 dissolved acetylene, because portable tools are most con- 
 venient for this- class of work, as they avoid unnecessary 
 handling of heavy parts, and on the other hand the lower 
 temperature of the oxyhydric mixture renders its application 
 impossible in repairs where plates are above ^2 inch thick. 
 
 For an example we will mention some of the very interesting 
 work performed during the year 1906 by the use of dissolved 
 acetylene by the Societe 1' Acetylene dessous du Sud-Est in 
 the harbor of Marseilles. 
 
 34 
 
1st. Repairing Cracks Steamer "Eugene Pereire" of the 
 French Line, March, 1906. 
 
 The boiler furnaces of the mail steamer Eugene Pereire of 
 the French Line had numerous horizontal cracks above the 
 grate bars. There were about 100 of these, and in two 
 of the furnaces they extended from end to end of the corruga- 
 tions. 
 
 It had been attempted to stop the worst of these by plugging ; 
 but it would have been necessary to renew several furnaces, 
 which would have detained the steamer for two months and 
 caused great expense. All the cracks were wedged open with 
 chisels and welded; all repaired parts were annealed with 
 burners. In two spots where there were several adjoining 
 cracks, a part of the furnace was cut out and replaced by a 
 welded piece. No leak was observed at any of the 100 places 
 so repaired at the hydrostatic or steam tests. 
 
 Only the sweating of a few drops, caused by trifling lamina- 
 tions, were discovered, and a little calking restored the water- 
 tightness at such spots. The work lasted three weeks and cost 
 $300. From the month of March of that year the steamer 
 has been on the Algiers voyage, which is very trying for 
 boilers on account of its shortness, the fires being banked and 
 boiler temperatures changed so frequently. No trouble has 
 been experienced with any of the welded parts. 
 
 2d. Assembling of Welded Pieces; Work on the "Marsa," 
 June, 1906. 
 
 The unreliability of riveted patches on damaged boilers is 
 well known, particularly where the rivets are exposed to fire. 
 The use of oxyacetylene welding, by which two pieces may be 
 united end to end without butt straps, brings the plates to 
 their original condition and avoids all the inconveniences of 
 rivets. The work performed on the Marsa offers a remarkable 
 example of the results that may be thus obtained. 
 
 Of the four furnaces of this steamer, the steel plates A 
 and B riveted top and bottom to the fire-box, and the plates 
 composing the back end of the combustion chamber, C and D, 
 
 35 
 
were completely worn out. Portions of these plates 18 inches 
 to 36 inches long were cut out and replaced by welded pieces, 
 as indicated in dotted lines on the sketch (Fig. 25.) This work 
 was very successful, except on one of the sixteen pieces, which 
 was later replaced by a riveted patch. The welded part of this 
 
 FIG. 25. 
 
 piece broke several times, but observations made in the case 
 showed a defective quality in the plate to which the new piece 
 was joined. The other fifteen held good. In the course of the 
 work it was noted that the plates of the bottom of the boilers 
 were badly eaten away at E on a space of about 36 inches ; 
 oxyacetylene welding was used to restore these plates to 
 
 36 
 
FIG. 26. 
 
 37 
 
their original condition. In some spots they had been re- 
 duced by corrosion to a thickness varying between % and 1/16 
 inch. 
 
 3d. Repairing Corroded Parts on the "Cholon." 
 
 Oxyacetylene welding may be used to add metal directly to 
 the surfaces of plates, to repair corroded spots, such as are 
 frequently found in various parts of boilers. The flame of the 
 blow-pipe is directed upon the plate, and when the latter 
 begins to melt the workman presents to the flame a bar of soft 
 steel about 7 by 7, which melts and fixes itself in drops on the 
 corroded surface. 
 
 The repairs of the Marsa, already referred to, give a sample 
 of the value of the welding process, but the work performed 
 on the Cholon, of the Compagnie des Qiargeurs Reunis, from 
 Aug. 20 to Sept. 20, 1906, presents a still more striking case. 
 
 The eighteen corrugated furnaces of this steamer were 
 badly eaten away on the surface. There was corrosion on 
 each side and for some distance above the grate bars. 
 
 The work was difficult to perform, as the workmen were 
 compelled to be inside of the boilers ; and were inconvenienced 
 by the heat of the blow-pipe flame; and the places to be 
 welded were lower than the workmen's footing; 10,000 cubic 
 feet of dissolved acetylene and as much of oxygen were used; 
 about 200 pounds of steel were used to cover the corrosions 
 and restore the plates to their original thickness. This work, 
 at a total cost of $2,400, avoided the replacing of eighteen fur- 
 naces, as originally ordered by the government inspectors. 
 
 4th. Repairing Boiler Heads Worn by Corrosion or by 
 Repeated Calking. 
 
 A frequent fault in marine boilers is due to the grooving of 
 the flanged furnaces riveted to the combustion chamber. These 
 heads are under great stress on account of the expansion. 
 Leaks, which are in some cases frequent, require calking; but 
 each calking reduces the width of the collar or flange, and 
 after a series of calkings the parts are practically worn out. 
 
 By the use of oxyacetylene welding such defects can be 
 
very easily repaired. An addition of material restores the 
 plates to their original condition. The work is at times a 
 lengthy one, but presents no special difficulty. Advantage is 
 taken of a preliminary heating up to reshape, if necessary, the 
 
 piece to be repaired, to insure its close contact with the plate 
 to which it connects. 
 
 By making repairs as soon as a defect is noticed, boilers 
 may be kept in perfect condition and last indefinitely; such 
 repairs delaying the running schedule to an insignificant 
 extent. 
 
 Railroad companies are by this process enabled to repair 
 
 39 
 
their locomotive boilers at trifling delay, and consequently to 
 reduce considerably the capital otherwise tied up in repair 
 shops. 
 
 MANUFACTURE OF THIN STEEL TUBES. 
 
 The methods employed up to the present time for the manu- 
 facture of seamless drawn metal tubes are not economical 
 for very thin tubes, such as bicycle tubes or water or gas pipes 
 of large diameter. 
 
 Oxyacetylene welding solves this problem. The plates are 
 bent to form the circle, the edges being forced into close 
 contact. 
 
 The edges are welded by a blow-pipe, the tube is then drawn 
 through a form to rectify the thickness of the metal and give 
 
 to the tube the finished form, which cannot be exactly obtained 
 by bending. 
 
 Tube manufacture presents no peculiarity; but the large 
 number of similar pieces which are made at the same time has 
 brought about the adoption of a special welding process. 
 
 The rapidity of welding depends upon the power of the 
 blow-pipe used, the thickness of the parts, the total bulk of the 
 pieces, and on the nature of the metal. In the manufacture 
 in quantities the factors do not vary ; the result being that 
 the rapidity of welding is practically constant. It is, however, 
 evident that a workman, performing the work by ordinary pro- 
 cess, cannot, however great his attention, maintain an abso- 
 lutely steady pace; the ordinary welded metal along the joint 
 is therefore irregular, and this spoils the outward appearance 
 of the work. 
 
 40 
 
It is natural that in these manufactures in great quantities, 
 automatic movement of the blow-pipe over the joint must be 
 sought. The problem has been solved as follows : 
 
 The tube T, Fig. 28, secured with its two edges to be united 
 in proper contact, is placed on a carrier A moving on a bench 
 B. This carrier is automatically run at the speed found neces- 
 sary for the thickness of metal to be welded. On an extension 
 C of the bench B is placed the blow-pipe. 
 
 The work is generally performed by a woman or a boy. 
 The blow-pipe flame is regulated above the seam to be welded ; 
 
 FIG. 29. 
 
 the carrier A is set in motion. The tube runs under the flame 
 and is automatically welded. 
 
 The operator has only to watch and adjust the flame of 
 the blow-pipe so that it strikes the seam. 
 
 This simple method does not require skilled operators to 
 obtain sound welding, and if properly regulated the metal is 
 united at all points along the seam. 
 
 Whenever the work so warrants this automatic running of 
 the piece under the blow-pipe should be adhered to. 
 
 MANUFACTURE OF TUBING OF SPECIAL SHAPE. 
 
 We shall here mention several special articles of manufac- 
 ture which are rendered economically possible only by the 
 application of the oxyacetylene process : The manufacture of 
 partitioned tubes, the partial partitions of which guide the 
 fluids circulating inside ; manufacture of handle bars, etc., for 
 bicycles (Fig. 29) ; the forming of elbows impossible to 
 obtain by bending, whether on account of the small radius of 
 the bend or of the thinness compared with the diameter. 
 
 41 
 
In the case of large diameters, and when the regularity of 
 the shape is not absolutely essential, the elbow may be made 
 of cylinders welded together at the desired angle. 
 
 IRON, STEEL AND BRASS TUBES. 
 
 As soon as blow-pipes were found satisfactory, pipe-manu- 
 facturing shops were provided with installations which ren- 
 dered practicable, at small cost and with absolute guarantee, 
 work that had previously been impossible of performance. 
 
 At the present time pipe manufacturers in France possess 
 oxyacetylene welding apparatus by means of which they are 
 able to make in short order changes or repairs to their plants 
 which heretofore caused long delays and required taking 
 apart and reassembling of machinery and the attention of 
 special workmen. 
 
 End to End Welding. 
 
 When two tubes have to be united end to end, and they are 
 not more than 3/16 inch thick, it is only necessary to cut the 
 ends at right angles. If more than 3/16 inch thick the ends 
 must be chamfered, and the welding is then made as men- 
 tioned in the case of plates. 
 
 It is thus possible to obtain coils the developed lenghts of 
 which attain miles without any flange or other joints, which 
 completely avoids the leaks and the consequent hard and 
 costly work of examining and replacing the gaskets. 
 
 Pipe Branches. 
 
 Oxyacetylene welding avoids the making of collars in 
 branch work. It is only necessary to give to the end of tube 
 A the curve of tube B, with -which it is to be connected, and 
 to fit it snugly into an aperture in the latter, and bring the 
 parts under the flame. 
 
 Welding on Flanges. 
 
 Many accidents have occurred in consequence of the de- 
 struction of ordinary brazing, due to the voltaic couple formed 
 by the contact of copper and zinc. It is a well-known fact that 
 
 42 
 
Before welding After welding 
 
 Fig. 30 
 
 Fig. 31 
 
a hammer stroke on an iron flange brazed on an iron pipe 
 results, in many cases, in a failure of the brazing. Oxy- 
 acetylene welding absolutely avoids this cause of possible 
 accident, as there is no interposition of a chemically active 
 metal. 
 
 The work may be performed as in Fig. 31, by using the 
 ordinary flanges and making in A an addition of a proper 
 amount of metal to fill the joint up solidly. If the diameter of 
 the flange is large in comparison with its thickness, and trouble 
 is feared through the heating of its center and buckling, it is 
 preferable to provide the flange with a boss and to make the 
 weld in B (Fig. 32). 
 
 Flange work costs somewhat more than brazing, unless care 
 has been taken to bring the parts to red heat by some inex- 
 pensive fuel before the welding flame of hydrogen and acety- 
 lene is turned on. 
 
 But the safety is much greater if the pipes are to be kept 
 at high temperature and pressure, such as with superheaters, 
 digesters, etc. 
 
 REPAIRING OF CRACKS, FITTINGS, ETC., IN IRON AND FORGED STEEL 
 PIECES AND IN CAST-STEEL PIECES. 
 
 By application of the oxyacetylene flame, cracks and pittings 
 in any iron or steel pieces may be repaired rapidly and 
 economically. 
 
 In the case of cracks and blisters a groove must first be 
 made to enable the flame of the blow-pipe to reach the bottom 
 of the defect. The flame of the blow-pipe is then turned on 
 until the metal reaches a white welding heat. A piece of 
 (preferably the same) metal is then presented to the blast 
 until its melted drops solidly fill the crack. 
 
 Numerous steel manufacturing plants in France now use 
 the blow-pipe to repair blisters in castings, and manufacturers 
 in general find in it a most valuable instrument, which prevents 
 the necessity of discarding pieces in which defects may be 
 discovered, very often after large sums have been expended 
 for workmanship. Portable apparatus is in this case particu- 
 
 44 
 

 larly handy, as the piece may be repaired without taking it 
 from the machine-tool (lathe, planer, etc.,) on which it may 
 be adjusted. 
 
 The steamer Le Gaulois broke her stern post, the section 
 of the broken part being 4 inches by 9 inches. The edges of 
 the break were first chamfered with a special acetylene burner 
 in fifteen minutes, and the welding proper was then performed 
 in eight hours by the use of dissolved acetylene; 1,225 cubic 
 feet of this gas were consumed. To replace this stern post 
 would have cost from $3,000 to $4,000, and detained the 
 steamer a considerable time. 
 
 45 
 

 
 
 
 
OVERDUE. 
 
 A?8 27 
 
1 79283