The Iron Age 1915-07-15: Vol 96 Iss 3

THE IRON AGE New York, July 15, 1915 VOL. 96: No. 3 Designing High-Pressure Steel Valves Use of Cast-Iron Valve Patterns Con- demned—Proper Metal Distribution Essen- tial—Scientific Investigation Recommended BY E. G. There is an inclination on the part of valve anufacturers to use the patterns which were made or iron valves for pressures of from 175 to 250 bh. of steam for making steel valves for the same by somewhat higher pressures and the practice open to several objections from the users’ stand- mint and is probably preventing, at least to a cer- ain extent, the use of steel valves in places where hey could be used to good advantage. The primary reason for this pattern makeshift 3a desire to economize and as such is worthy up oa certain point, but unfortunately the economy benefits the manufacturer only, and it is a short- ighted economy at that. Some valve manufac- rers cast steel valves in their own foundry in or sar their cast-iron foun- 7, or buy the castings m their immediate vicin- y, while others must nd the patterns out to ome distance. There is ident economy, tempo- ily at least, in the first ase in using the same atterns even though frinkage and finish al- bwance may have to be ken care of by some Fig. 1 GREEN MAN in metal from the same master patterns and all sizes may be made from the same drawings and in case of emergency the patterns may be used inter- changeably for iron or steel. Consider the objections to using the same or a duplicate set of patterns for the two types of valves, no matter what is the apparent justification for such a course. In the first place there is the question of the cost of the valves to the customer. There is no denying that it will cost money to lay out new designs for steel valves and the manufac- turer who could get along with one set of patterns for the two valves will be put to the additional ex- pense of getting out a set of patterns for such sizes as he has a demand for in steel but possibly a saving of material suf- ficient to pay this expense within a reasonable length of time can be ef- fected with no reduction in the utility of the prod- uct: Is there not possibly an objection to the use of such a heavy pattern for steel valves, or if there is no actual objection other than would apply equally Fig. 2 to cast iron, is advantage hanges in the original sections to Indicate Unappreciated Features of Valve Design being taken of the addi- patterns which were made or cast-iron valves. In the second case the konomy in using the same or similar patterns is hot so evident as few manufacturers could spare me patterns from their cast-iron foundry long ough to get the necessary steel castings without lelaying cast-iron valve orders, and the difficulty is made worse from the fact that it is somewhat more lifficult to secure tight castings of steel and conse- fuently the patterns may need to be left in the eel foundry until the castings are partially fin- Shed so that there may be reasonable assurance at no serious flaws will be found. Comparatively €w steel valves are used at present and therefore k often does not pay to make up enough extra astings to take care of possible losses which can- hot be detected previous to finishing. But even this condition makes it necessary to have two ts of patterns for steel and cast-iron valves, the anutacturer may still consider it exnedient to ake a set for the steel valves which is a duplicate f his cast-iron valve patterns. His reason for such Course obvious as certain sizes may be made tional strength of the material to as great an extent as possible? The steel used: for -Valves will run from four to five times as’ Strong as cast iron in tension, and the material. is-so much more secure against failure from shocks such as are produced by water ham- mer, and from such piping strains as are caused by weight and by expansion and contraction, that a much smaHer factor of safety may be used and as to stresses due to transverse strains, although there is: not so much difference in the tltimate resistance of the two materials, the steel will de- flect just about one half as much under the same stress and as the most frequent failure in satis- factory service of valves is by leakage due to distortion, this fact may be taken advantage of in determining the proportions of the various parts. Such distortions are usually caused by one of two things; they are either due to strains caused by tightening the disc against the seat, or by strains caused by such external forces as expansion and contraction, or weight of adjacent parts. The lat- ter strains are closely related to the weight of the 125 a > 126 THE IRON AGE July 15, 1915 valves themselves and a saving in material by making the valve bodies thinner has the additional advantage of reducing the external forces tending to cause failure, for very often a pipe line is made up of a series of valves with fittings or short pieces of pipe between them. There is another very great advantage in this reduced weight which is that the stresses in the overhead framework of the building, due to the weight of the suspended pipe line, are very greatly reduced and as such framework is ordinarily not designed to carry any considerable weight this is a decided point in favor of lighter valves and fittings. Considering now some of the methods of de- signing steel valve bodies to secure the greatest advantage from the material employed, it may be stated that a cast-iron valve for heavy pressure is usually of about uniform metal thickness through- out, and it is a pretty clumsy affair. In designing a steel valve there is not the fear of failure occur- ring with the other material and rational principles may be applied. It is the writer’s purpose to men- tion only two points at which metal may be saved to advantage. One is in the necks of the valve. These are usually made heavy because they are adjacent to heavy end flanges whose thicknesses at present follow the extra heavy iron standard. It is to be hoped that a lighter standard of thickness may shortly be adopted by progressive manufac- turers, thus avoiding the necessity of making a2 valve heavier throughout than is necessary for strength or stiffness simply to accommodate it to a standard flange. But the flange is noc entirely to be blamed for heavy valve necks for there is an ap- parent opinion that they should be heavy to give the body stiffness. That idea the writer wishes to oppose for the neck is an unimportant part of the body and distortion here is unimportant, while in the case of steel valves where there is little danger of fracture, it may be a decided advantage to make an unimportant part light on purpose so that distortion may occur there and so protect the working parts. There is another section of a gate valve in particular where poor distribution of material is a disadvantage, especially in case the metal is not very much heavier than necessary to prevent leak- age due to deformation, and that is at the section at right angles to the pipe line and between the seating surfaces, see Fig. 2. In the case of a wedge disc gate valve it should be evident that as the distance across the top of the wedge is greater than that across the bottom, the thickness of the metal in the body which opposes the thrust of the wedge against the seats should be in the same ratio. If the minimum thickness necessary to hold the valve tight at the bottom is determined, the metal should be gradually increased in thickness up around the sides to a point opposite the top of the wedge, in proportion as the length of the portion of metal that resists the thrust increases. Since the stretch of the metal at any section is directly proportional to its length and inversely as its thickness, the distance across the wedge at the bottom divided by the thickness of the metal at the bottom should equal the distance at the top divided by the thick- ad a’ ness at the top, i.e. This would only be true t fs in case other conditions were uniform, but such is not the case, for at the top not only is the dis- tance greater but the thrust of the disc is not re- sisted by a direct or even a fairly direct pull in the material, and these forces must be transmitted by a flat surface which is subject to flexure. This flat surface, indicated by shading in Fig. ©. shoyy be one of the thickest points in the valve | od, and it should be stiffened by ribs on the out ide, ay due to its weakness the thickness ¢’ of net connecting the two opposite faces of the set at th large end of the wedge should be thicker than termined by the proportional lengths if possible Scientific principles have rarely been applied 4, valve design in the past, but with the high temper tures and pressures now in more genera! ys¢ time is ripe for engineers to enter this field they will find plenty of problems requiring exten and careful studq. A 6-In. Rotary Cutting Off Machine The Newton Machine Tool Works, Inc., Twenty. third and Vine Streets, Philadelphia, Pa., hg brought out a high duty 6-in. rotary cutting off mg. chine. It is designed especially for use in conne. \ 6-In. High Duty Rotary Cutting-Off Machine for Use | Connection with High Explosive Shells Which Have Parted to Provide for the Fracture Test tion with the manufacture of high explosive shell which have to be parted to provide for a fracture test. The hourly capacity of the machine is twenty pieces of 0.50 per cent carbon steel 6 in. in diameter. Two systems are used in making shells for the fracture test with this machine. In one the bar} fed forward by stock handling apparatus at the rear of the machine, which is not shown, the app* ratus being controlled from the operator’s position The other method is to cut the bars into double lengths on the builder’s cold saw cutting off m* chine and then place the double length bar in the machine illustrated and part it for the fracture test. This procedure does not require the use ® the stock handling apparatus and gives very rap! production on account of the speed with which the bars are turned out by the cold saw cutting off machine. The ends cut off on the saw form the to or open end of the shell, the fracture test being only required on the bottom. The double cutting off slide has an adjustable automatic stop for varying lengths of work and three changes of feed. It is arranged to take too measuring *°g x 25 in., which are staggered ' break the cut and arranged so that flat steel ca” be used, tool clearance being provided in the m* chine. A pump and attachment for lubrication 4 included in the equipment. The lubricant is °F Tuly 1 15 THE IRON AGE 127 tne chips and the latter are thrown Alloy tool steel for both the load rollers and xx through the opening in the front spacing rollers of the bearing, which has been de- scribed in these columns, is cut from bar lengths ine is driven by a 10-hp. adjustable- to the required sizes on automatic screw machines, iting current motor. Where a direct adjusted for the length of roller. Each roller is r is used, a unit of the accelerating inspected for defects before and after being hard- ved. ened and is then tested for hardness and toughness and finally sent to the grinding department for MA | finishing. The design of the bearing demands close = ING ROLLER BEARINGS observance of dimensions, and much of the grind- Meth lowed in American Roller Bearing ing -_ a aa ae ee pv s insure permanent accuracy. Works, Pittsburgh Before going to the assembling department, nerican Roller Bearing Company, Pitts- the inner sleeves are fitted with the roller retaining recently placed its new factory in opera- collars. These collars, made of cold drawn steel production of anti-friction non-lubri- tubing and heat treated, are put into position under bearings for shafting machinery and for pressure and keyed. The assembling operation is generally. The bearing operates with- simple. The load and spacing rollers are fitted un ition. It consists of heavy-wall cold- der the roller retaining collars on the inner sleeve, tubing for the sleeves and chrome tool making the rollers and the inner sleeve a single stet the rollers. unit, as indicated at the extreme left of the smaller d-drawn steel tubing for the inner and views. This method of construction has proved of eves of the bearing is received in mill value when the bearings are used as in heavy truck d cut into standard sizes by high-speed wheels, as it is often necessary to mount the roller pecial design. The rough tubing for both unit on the shaft or axle, and press the outer sleeve and outer sleeves is machined inside and’ in the wheel hub or bearing housing. rret lathes. The races or tracks on the Bearings for power transmission devices ar‘ eeves are shown among the accompanying furnished with outside sleeves of special designs to strations. After machining, the sleeves are conform to line shaft hanger and pillow block d red to the heat treating department for signs, but bearings for machinery are supplied with The heat treating department is. plain sleeves, permitting them to be installed in equipped with Tate-Jones furnaces, burning natural equipment designed for plain bearings. When the gas under pressure with the necessary drawing bearings ure so installed, no oil reservoirs or other inks. lubricating devices are provided in the housings, as sf After inspection for flaws, the sleeves are sent they operate, as stated, without the use of a lubri- to the grinding department for finishing both in- cant. This is possible, it is held, by the elimination side and out. The internal grinding is done on of sliding contact, the load-carrying rollers being grinders of special design. Two of the smaller illus- separated, at all times, by the spacing rollers. The trations show an inner sleeve before and after the load-carrying rollers lie within the race or track grinding operation. The sleeve shown to the ex- on the inner sleeve, while the spacing rollers are treme right is longer and are | assembled, [ i i en ee ee a ea, the two er-fre- supported by the race flanges and revolve in a dire tion opposite to that of the load rollers. The bear- ing provides that the inner sleeve surrounds the shaft or axle and takes the wear of the hardened | rollers. The bearing was described in THE IRON AGE of shed sleeve Dec. 12, 1912, and andre] ag 1 March 27, 1913. collars be- ressed into For out- ‘inding the eeves are set up special man- lrel tools shown arger view. ture is de- signed to allow set up, ninimum removing ght are the assembled rollers of the bearing: the outer shell or sleeve; the machined inner sleeve ; the inner sleeve with one of roller-retaining collars. Above is the rig for outside grinding Importance of Annealing Steel Casting German Tests Showing the Dependence of the Static and Dynamic Properties on Annealing Temperatures— Judging the Heat Treatment by the Fractured Test Piece In a recent article in Stahl und Eisen, Dr. P. the third, results of shock tests, Fig. 3. This |, Oberhoffer of Breslau, Germany, brought together diagram shows the great value of the shock test jj testing steel castings, especially those with the lowe carbon percentages, and all show the great influeys of annealing at the proper temperature. A cop, parison of the different curves shows the amount ¢ improvement, which varies from one material y another. Photomicrographs of No. 4 are given jp the paper to show the change in structure that »&. companies annealing. his results on the annealing of steel castings. He has previously published several papers on this subject, one of which was abstracted in THE IRON AGE, May 21, 1914, under the title, “Shock Tests and the Annealing of Steel.”” The composition of some of the steels worked on, arranged according to carbon percentage, is given in the table. Table of the Composition of the Steels sie c irbon, Manganese, Sulphur, Phosphorus, _ Silicon, A practical method often used for determin ng “0.11. 0.60 0025 pergent pereent the success of annealing is the observation of th . oo 0.98 er a ese 0.38 tensile test fracture. A fine structure shows with 4 0.40 1.11 0.039 0.027 0.21 certainty that the proper temperature has beg , ceeee ee 0°79 enna eee +4 reached. Unfortunately this method is not alway 7 0.69 1 + 0.022 0.021 25 of value, especially with the lower carbons wher 1 A 0.1 0.96 0.067 0.072 one the elongation and reduction of area are higher 3 0.13 0.4 0.029 0.03: 0.52 In such steels unannealed samples often give ver These steels were cast in blocks 250 x 250 x 600 mm. (9.84 x 9.84 x 23.62 in.). They were all free from visible blow holes or contraction cavities, nor t good elongation and reduction of area. There js however, another simple method to indicate whethe he proper temperature has been reached, which is i im 3 a i. 7. £ bo nefie No Sr 0462 C; 2 gc &é 60 - ~ e+ 4 ee . ® j ; “ =e I on 64+ o¥, ae -- No.4+ 0430 ee @-—+5—|—- mi pp a py} a @-+-—- B50 —=AP ThA wie VU L could any considerable amount of segregation be determined. Test pieces were cut out cold and the steel thoroughly examined, both as cast and after annealing for six hours at the desired tem- peratures and cooling slowly. The average values of the results of the various tests are given in tables in the original paper, and are also clearly shown in diagram form. Three of these diagrams are given here, the first showing the dependence of the tensile strength on the an- nealing temperature, Fig. 1; the second, the depend- ence of the elongation on the annealing tempera- ture, Fig. 2, measured in 200 mm. (7.88 in.), and 128 $—~a}- +—_-_—_+ ‘ and No|3-0.283C G |i 233 Annealing Temperature in Steel Castings illustrated in Figs. 4 and 5. These test pieces af from steel 1-B, with 0.13 per cent carbon, and bo! show very good ductility. Bar 1 was annealed # 800 deg. C., and bar 2 at 890 deg. C., the latter ing the correct temperature for this steel. Th rough surface of the first piece is clear evidente that it has been annealed below the proper temper ture, and only with the reaching of this tempel ture is the surface completely smooth. This appearance is less noticeable the lower elongation and reduction of area, but in these cas® the fracture is more crystalline, so that either fra ture or the appearance of the surface will give ® the rtain indication as to the annealing, especially ith soft steels, when a microscope is not available. In Fig. 6 the heavy firm line shows the best an- ng temperature as determined by these experi- and caleulated for steels with 0.8 per cent It should be remembered that exceeding these mperatures by a small amount, say 50 deg., is less ful than the bad results caused by not reach- g the proper temperature. The dotted line shows ie beginning of ferrite separation according to oerens and Meyer, for steels with 0.8 per cent mn. ery careful experiments were made with these THE IRON AGE 129 Fig. 2—Dependence of Elongation on Annealing Temperature Steel Castings I I steels and the thinner firm line shows the actual location of ferrite separation, the horizontal line, of course, being the pearlite. This is in much bet- ter agreement with the proper annealing tempera- tures than the dotted line. Fig. 7 shows the aver- age results of ultimate stress, elongation, etc., in relation to the carbon of the steels after proper an- nealing. Notwithstanding the varying manganese the curves are smooth. There was a short discussion of the paper, which mainly took up the difference in results found with steels 1 and 1-B. These steels were very similar in ; : ° 2 + i t ) a eee BOO —- ae. Ue ae 1000 Annealing Temperature in Degrees Cent i 4 ; ™ af @ Fig. 3—Results of Shock Tests on Steel Castings ie = f i tov 130 THE IRON AGE July 15, 195 Fig. 4 (Upper) Shows the Surface of a Broken Ter sile Test Zar of Under-Annealed Steel Castings Fig » (Lower) Shows the Smooth Surface of Properly Annealed Castings Fis " Comp: son of Ferrite Separati« wit \ Temperature composition, but gave very different physical tests, and while the explanation was not fully settled upon, it was thought that it might be due to a difference in casting temperature, or because the blocks cast were of different size. The diagram showing when factory operation be comes unprofitable, printed on page 25 of THE IRON AGE of July 1, was not credited as it should have been to James A. White, works manager, of the Worcester Pressed Steel Company, Worcester, Mass., who sub- mitted it in a discussion of the relation betwen pro- duction and costs at the recent meeting of the American Society of Mechanical Engineers at Buffalo, N. Y. Aluminum As a Check to Sulphide Segre. gation in Steel Ingots Forms in which sulphides may exist in st: was a subject of discussion by Prof. J. O. Arnolq G. R. Bolsover at the annual meeting of the [ro) Steel Institute in London in May. In a paper on 4 subject, published last year, the authors sai It is remarkable fact that the casts exhibit less we narked areas of the eutectic mixture a! aluminum while those presenting no evidence of containet o aluminum These facts suggest tl small mount of aluminum present plays some in forming the eutectic or in preserving the pat fallen-out manganese sulphide by checking its into relatively large masses. The paper presented this year contains the x of experiments made to confirm, if possible, the aut original suggestions. About 12 lb. of crucible cag No. 1499, weighing about 48 lb., was poured, withoy any addition of aluminum, into an iron mold 1% square. The remaining 36 lb., in a fluid state in th crucible, was put back into the furnace, and allowed; cool in a lump over night. Thus were obtained po tions of identical steel quickly and slowly cooled. Dri. ings were taken near the top of the ingot and near the top and bottom of the lump, and on analysis showed the following composition: iit Nor Ca Sili Man Sul- Pho bon, con, ganese, phur, phe ru Pel Per Per Per te? Cast No. 1499 Cent Cent Cent Cent Cent (near top) 0.26 0.075 1.11 0.517 0.02 0.26 0.084 1.07 0.53 0.021 1) 0.20 0.056 0.76 0.302 0.017 The main segregations noted in the lump are 0 per cent sulphur and 0.31 per cent manganese. Both these segregations are doubtless largely due to a con. mon cause, namely, the rising toward the top of th lump of MnS. An ingot, No. 1461, which had been cast 14 square, after the addition of about 0.03 per cent of pure metallic aluminum, was analyzed for segreg gation on drillings taken from both the top and the bottom of the ingot. The following compositio1 show} ¢ = nS =: 7 = =< =+ sc . = 325 835 85 ¢ — L é Th = < , q top 0.27 0.167 0.74 0.424 0.019 ¥% in. sq. bottom 0.28 0.158 0.78 0.478 6.018 The above figures show that the aluminum has pra tically prevented any segregation. This curious and unexplained action of aluminum was, at the meet ing of the Institute at Sheffield in 1905, proved B. Talbot in an exhaustive and most valuable paper “Segregation in Steel Ingots.” The amount of alum num employed by Mr. Talbot was very small, amount ing to only about 0.007 per cent per ingot. Of crucible cast No. 1540, weighing about 48 |, 9 lb. was poured into a 1%-in. square ingot, while the remaining 39 lb. was returned in the crucible to th furnace, and allowed to cool over night. To this cas about 0.05 per cent of metallic aluminum was added 5 min. before teeming the small ingot. The comp sition per cent of this ingot, and also of the top ané bottom of the lump, are set forth in the follow } analyses: Mat Sul Phos Carbon, Silicon, ganese, phur, phorus Per Per Pet Per Per er No. 154 Cent Cent Cent Cent Cent Cent Top « £ 0.5 0.163 1.02 0.493 0.02 Top ¢ im} » 30 0.163 1.02 0.493 0.021 Botte ‘ lump 0.26 0.187 1.02 0.480 0.021 The aluminum has viritually prevented any seg’ gation in the very slowly cooled lump, whereas sv segregation with reference to carbon, sulphur a0 manganese was well marked in the non-aluminum lump of cast 1499. The contrast in the segregation of the mangane sulphide, as shown by very fine photomicrographs in ™* original paper, is remarkable. The authors regard * as additional testimony to the power of aluminum ™ retarding and checking sulphide segregations A lt Four-Head Planing Machine Machine Company, Wilmington, Del., iilt a large planing machine with four achine, which will handle work 16 ft. eh and 30 ft. long, has been installed Island Navy Yard, Philadelphia, Pa. is employed and particular attention Leva the means for obtaining rapid manipu- the moving parts. i; has an over-all length of 55 ft. 4 in. nstruction was used, the various parts ae reed by box cross-girts and fastened ar th bolts and links. The top of the bed igh shaped to catch the dripping oil a ible rack and convey it to the central we One of the tracks on which the table it and the other V shaped, the angle of being 120 deg. e, which is driven by a cut steel rack, is and 11! ft. wide. To facilitate ship- THE IRON AGE 131 lower the cross-rail and power rapid traverse to the two tool heads mounted on the rail is provided by thessame motor. Both inside and outside gibs are used to clamp the cross-rail to the face of the up- rights. The tool heads are made right and left hand for close work and have vertical and angular adjust- ments of 48 in. and 36 in. respectively. The swiveling portion of the head is graduated and worm gearing enables it to be set over to any desired angle. The tool slides are of the square type and are held in place by adjustable gibs, taper shoes being provided for taking up wear. Both heads are provided with an automatic tool lifting device which is operated from the feed mechanism located at the right end of the cross-rail. The vertical or angular feed to the slides and the horizontal feed to the heads are controlled by the upper lever attached to each head. Auto- matic falling bearings are provided in the center of the cross-rail for supporting the weight of the saddle screws and splined shafts. The other two heads of the machine are located A Planing Machine with Four Heads for Handling Large Size Work member, like the bed, was made in pieces, sections being fastened together by slotted is of the builder’s standard double-plate de openings to remove the chips which through the stop holes in the top plate. prights are keyed and bolted to the sides | and are fastened together at the top by x section cross-girt. The uprights where lted to the cheeks of the bed are 10 ft. ne face is 24 in. wide. The cross-rail is a ‘ with a face width of 30 in. It is made ge n to permit either head to traverse the e between the uprights. An adjustable d to the top is provided which is relied fen the cross-rail and take up the sag due ned weight of the rail and the heads. erated through gearing from a 15-hp. ted on the top of the machine raise and one on each upright. They are of the builder’s ex- tension slide type, are counterweighted and have automatic feed up and down the uprights and a horizontal angular hand feed. Power for the auto- matic feeds is taken from two positive, quick acting feed boxes, one for each head, the weight being varied by graduated plates. Power rapid traverse is provided for all the heads and can be operated sepa- rately from each head by individual levers. When it is desired to engage the feed or the power rapid traverse the small handles at the end of the rail and on the side heads are turned to the right or left, it being thus impossible to engage both move- ments simultaneously. The machine is directly connected to a 75-hp. reversing motor, running at speeds ranging from 250 to 1000 r.p.m. This gives cutting speeds rang- ing from 18 to 36 ft. per min. and double these Ce — — rates for the return. The driving motor can be controlled from either side of the machine by levers connecting with the master switch and these are supplemented by a pendant switch carried on a swiv- eling arm. In this way control of the motor is secured: from any position that is convenient for the workman when setting tools or when the ma- chine is in operation. The driving gears are made of steel castings with teeth cut from the solid metal and the driving pinions are steel forgings. Provi- sion is made for taking up wear in all the driving shaft bearings. A special oil filter system operated by a 1-hp. motor is provided to supply all the driving gears, driving shaft bearings and table tracks with forced lubrication. TEST FOR SPELTER COATINGS* A New Method Offered for Use With and Wire Sheets It has been customary to express the weight of coating on wire in pounds per mile, while on sheet products the results are usually expressed in ounces per square foot. Obviously, the coating on wire expressed in pounds per mile would have a different meaning for each gage of wire. If the re- sults are expressed in ounces per square foot of surface on both wire and sheets, there will be a better understanding as to the thickness of coating on the respective products. In stating the weight of coating on galvanized sheets it is customary to express the weight based on one surface only, that is, a sheet containing 2 oz. of coating per square foot really contains 1 oz. on each side of the sheet. It is proposed to express the weight of coat- ing on wire in ounces per square foot, and also to use such lengths of wire that the number of grams of coating found will be equivalent to ounces per square foot, without calculation. These lengths must be such that the surface coated is equal to 5.079 sq. in. It is likewise proposed that the sam- ples for determining the weight of coating on gal- vanized sheets shall be 2%, x 214 in. (area = 5.079 sq. in.). The number of grams of coating on a section of this size will also express the weight of coating in ounces per square foot without calcula- tion. The method for determining the weight of spelter coating consists of using a small amount of antimony chloride in hydrochloric acid (sp. gr. 1.20). Antimony chloride appears to hasten the solution of the coating, and after the coating has dissolved a thin film of antimony plates on the sur- face of the base metal and retards the solution of iron or steel. Experiments have shown that sheet steel 214 x 214 in. which loses 50 mg. in five min- utes in cold hydrochloric acid (sp. gr. 1.20), will lose in that time only 1 mg. in the same acid con- taining 80 mg. of antimony per 105 c.c. of acid. DETERMINING SPELTER COATING OF SHEETS In the proposed method the metal is immersed in the acid only one minute, which is long enough to dissolve several grams of coating, yet the amount 32 THE IRON AGE July 1915 familiar, and a determination can be mad time than is occupied in making the Preece toy For determining the weight of coating oy gal vanized sheets cut several samples 2% x 21, j, from various parts of the sheet. These imples about five in number, should be weighed togethe and immersed singly for 1 min. in 100 ce ¢ hydrochloric acid (sp. gr. 1.20), to which has boo added 5 c.c. of antimony chloride prepared }y dis solving 20 g. of antimony trioxide in 1000 cc x hydrochloric acid (sp. gr. 1.20). The same 100 ¢¢ of hydrochloric acid can be used for at least fixe samples. Five cubic centimeters of the antimony chloride, however, should be added for each samp on account of the antimony being removed from the solution by the iron. The samples are washed and scrubbed unde running water, dried with a towel, and laid jp , warm place for a few seconds. The samples ar again weighed together and the number of gram: lost is divided by the number of samples takey Each gram corresponds to 1 oz. of square foot. JESS coating per DETERMINING SPELTER COATING OF WIRI A small section of the galvanized wire should le stripped in hydrochloric acid containing antimon chloride. The diameter of the black wire shoul then be carefully measured in order to determin the length of wire, such that the number of grams of coating will represent the number of ounces per square foot of surface. [These lengths are given in a table in the paper. | The method of making the test is very similar to that outlined for galvanized sheets, except that the wire is first cleaned with carbon tetrachloride or gasoline, and after being carefully weighed is placed in a tall glass cylinder containing hydrochloric acid (sp. gr. 1.20), to which has been added from 2 to 3 c.c. of antimony-chloride solution of the same strength as used on galvanized sheets. The reason for using one-half the amount of antimony chloride in the case of wire is on account of taking one-half the area. As previously stated, the coating on gal- vanized sheets is expressed in ounces per square foot, considering one side only, when in reality this amount of coating represents 2 sq. ft. of surface. After immersing the entire length of wire for | min. it will be found convenient to pour the acid solution into another tall cylinder in order to fe cilitate removing the wire. The wire is then scrubbed under running water, wiped, thoroughly dried in a warm place for a few seconds and again weighed. Each gram lost corresponds to 1 oz. of coating per square foot. For direct compariso with the weight of coating as expressed on gal: vanized sheets, this figure should be doubled. A copy has been received of a pamphlet reproducing a lecture by Sir Robert A. Hadfield entitled “History of the Metallurgy of Iron and Steel,” delivered at 4 meeting of graduates of the Institution of Mechanica! Engineers, London, February 8, 1915. The lecture covers early metallurgy, French metallurgy in the eigh- teenth and nineteenth centuries, metallurgy in the mi¢ dle of the last century, the conservation of new Tf : of iron or steel dissolved is negligible. The small amount of antimony that plates on the surface of the sample can easily. be removed by scrubbing sources, and a detailed description of the author's tT search work, with special reference to manganese steel, low hysteresis steel, steel alloys, low temperature ¢x mem a 1a? under ‘running water. This method is one of the periments, etc. 7 7 ° . . . . fe : most rapid and accurate with which the writer is - — a —_—_— . it *From ‘a paper by J. A. Aupperle, metallurgical engineer. The Baldwin Locomotive Works are now employing St American Rolling- Mill Company, Middletown, Ohio, read 299 " . TY a Qc shich was i} Salons the ‘American Séclety for esting Materials, Atlante On mens compared with 3900 last January, which a City; N. J., June’ 22, ' about the low point in the company’s operations. lee fe ; H ; 4 Ew { | ; — »ve With 3-In. Checker Walls mpanying drawings show a hot blast ed and patented by P. J. Brown, presi- P. J. Brown Construction Company, Ohio. Particular emphasis is placed on of the piers in the bottom of the stove ree access for cleaning and also to the of the arches supporting the checker ch aims to give strength of construction to eliminating displacement of arches yration. ter wall constructed of 9-in. brick com- rcles the stove, inclosing the checker CTETHTGT Ge Ua apr A SLA NOAS SE S Fy ry > rottey eS D a Stes | ° SS Hes. ss | f 4 ¥ Hy HH) ih RR Ay N RS S>> . 4 SS i y RN 3 ih NH BH N 3 : Section ©-D Sections of Brown Hot Stove 15 THE IRON AGE Li ~~ -~ work and combustion chamber. This wall also sup- ports the dome. The checker work is not bonded into this wall, which permits an independent ex- pansion of checker work from that of the dome and back wall. In repairing this stove checkers may be taken out and replaced, it is explained, without disturb- ing the back wall or the bridge wall of the stove. The checker walls are constructed of 6 x 9 in. standard straight brick, laid on the 3-in. face with alternate courses running at right angles to each other, giving a 9-in. square checker with a 3-in. checker wall. The 3-in. brick is calculated to give the structural strength necessary, combined with the thin wall to make it the ideal checker construc- tion. With the exception of shapes in arches and dome, the stove is constructed entirely of standard size brick 3 in. thick. Industrial Plant Automobile Trailer A trailer that can be coupled to any type of automobile can probably be put to considerable use around a manufacturing plant in making local de- A Trailer Designed for Coupling to an Automobile and Har dling Loads Which Are too Heavy for a Small Ca ind not Heavy Enough for a Large Motor Truck liveries, as well as for carrying loads to and from freight and express offices. Such a car can be used in place of a heavy truck for making deliveries of small loads, or by small, light, manufacturing plants that do not have motor trucks. A trailer designed to be attached to an auto- mobile for these and various other purposes has recently been brought out by the Erie Trailer Mfg. Company, Erie, Pa. It is made in various sizes and capacities and in numerous models to suit requirements. The one shown has a strongly con- structed wood box, 8 ft. long, 46% in. wide and 16% in. high, weighs 500 Ib. and has a capacity of 2000 Ib. It has 30-in. wheels with 1%-in. solid rubber tires. Ball, roller or box bearings are fur- nished as desired, and some of the models have four wheels. Couplings are furnished to fit all makes of automobiles and after the coupling is affixed to the car body it requires but a moment to couple or uncouple the trailer. The consumption of aluminum and the production of bauxite in the United States in 1914 were the larg- est ever recorded. The United States Geological Sur- vey report shows production in 1914 of 219,318 gross tons, valued at $1,069,194, compared with 210,241 tons in 1913, valued at $997,698. Aluminum consumed in this country in 1914 was 79,129,000 lb., against 72,- 879,000 Ib. in 1913 and 65,607,000 Ib. in 1912. The pro- duction was 150 Ib. in 1884, 550,000 Ib. in 1894, and 8,600,000 Ib. in 1904. Universal Strainometer of Simple Design’ A Single Instrument That Can Do the Work of an Extensometer and Com- pressometer on Various Gage Lengths BY S. H. GRAF In the municipal, commercial, or college labora- Briefly stated, the instrument consists tw tory for testing materials a wide range of work is’ simple adjustable frames, each carrying two scr; usually handled, and specimens of sizes other than’ bearing on the gage marks on the specime) the ordinary standard ones must often be tested. of the frames carries an Ames dial and pivots Although most of the ordinary routine tests are a rod held rigidly in the other frame, the dial j; made without deformation measurements within the cating twice the actual deformation. A togg! elastic limit, these data are sometimes required, clamp prevents the frame from separating fron the pivot, and a slender steel rod actuates the staf of the gage head. The error due to tilting of the dial frame within the range of any test is of y consequence, and within the elastic limit it is not a readable quantity. Figs. 2 and 3 show the instrument applied t various specimens and indicate its range; this range includes specimens either in tension or compression up to 8 in. in diameter or square, and of any gag length from 2 in. up. To adapt the instrument t different gage lengths it is only necessary to pivot and dial rods of different lengths; for the ordinary lengths of specimens, rods of drill steel 1 We One e able, while for special tests where the length ma be considerable, light wooden strips with steel serts in the ends are perhaps most satisfactory. In addition to its universal applicability this strainometer also possesses the following advan- tages: Fig. 1 Universal Strainometer in Case making it necessary in the past to have at hand at least the following: 1. An extensometer for 2 and 8-in. gage lengths. ; 5 ara ; : 1. Ease of use; no special skill being required t 2. A compressometer for small specimens, with spe- St : ae als ; read the dial. Dial is zero setting so that there is! ‘ial collars. a s , , nA troublesome zero reading to subtract; the reading di 3. A compressometer for 6 or 8-in. concrete cylinders. vided by two gives the net deformation. ) It is believed that a simple, direct-indicating, - Readings may be taken “on the run” with 7 even by an inexperienced observer. Fig. 4 shows load deformation diagrams plotted from readings actuall) taken “on the run” by students who had never before read an extensometer or compressomoter of any kind universal instrument for the measurement of both extension and compression would serve all ordinary purposes of the commercial and city laboratories. The writer and his associates, C. L. Knopf and ; 3. A change in rate of deformation from any caus “ > > ) « : ava , r) > > * . 51WePYs ros . . ° R. B. Boals, have experimented for several years jg at once plainly apparent. with various arrangements to accomplish the pur- 1. The instrument may be easily and quickly reaé- pose indicated and have finally arrived at the device justed and applied to any specimen within its range shown in Fig. 1. It is simple, and with ordinary care does not get out of order. *From a paper presented at the Eighteenth Annual Meet- - ne of the American Society for Testing Materials at Atlan- 5. It may be constructed by any good mechani tic City, NJ. June 22 to 26, 1915 The author is assistant The only delicate part is the indicating mechanism, con- professor of experimental engineering, State grricultura ol . . eo A ‘ - : jege, Corvallis, Ore sisting of the ordinary form of 0.3 by 0.001-in. Ames et) i j tit 1%) Fig. 2—The Strainometer as an Extensometer on a Tensile Fig. 3—The Strainometer Used as a Compressometer in : Testing Machine for Steel Testing Cast Iron ie \ 134 ie 4 ie is ar Shae 4) 5 ain back and celluloid crystal, which may complete and ready to attach. st of the instrument complete is low. uracy and reliability of the Ames dial ed to strain measurements have been well _and it is not necessary here to cite cali- ta for the dial. Repeated calibrations of ete strainometer as just described, both mierometer and against test bars of known have shown the instrument to be fully as is others designed to read to 0.0001 in. eed the extensometers and compressometers on ) et, while fundamentally of very precise de- sig so complicated and cumbersome, as well Elongation in Bin., in 0 04 08 12 1.6 20 _—————— [ 1000 ; / , 60 j / j f 4 } tl L dive bil = * = : : E at Bin rT. al Lompr 1 lest Tension Test 16t of A 1 S+ 1-Deformation Diagrams Readings Taken “on the Run” by Inexperienced Operators as slow and difficult to read, that their apparent ac- iracy as shown by calibration cannot be obtained nder operating conditions. Some objection has been made to instruments having only two (instead of three) points of attach- ! ut numerous studies made by means of the Berry strain gage on the distribution of stress in is specimens under test have convinced the uuthor that under proper conditions of gripping tension specimens and of bedding compression specimens, the two-point instrument will, with equal n centering, give the average deformation as hfully as the other. If provided with three dials crometers the three-point instrument is useful showing roughly the distribution of stress, but makes the instrument too complicated for all ordinary purposes. Three observers would be re- quired if readings were to be taken on the run, and even when load is applied by increments, it is diffi- iit for a single observer to read three dials or micrometers accurately. The strainometer as described has now been in ise for nearly a year in the laboratory at the Ore- gon State Agricultural College, and has proved very satisiactory, not only for class work, but also for mmercial and investigational work. The author Wishes to say that he and his associates claim no great originality for this strainometer, for it really en es a combination of principles previously ap- various other instruments. c Studebaker Corporation has taken out life in- ipon each of its employees in the operating de- ts, office men not being included. It is stated amount placed for each individual is sufficient ' for sickness and funeral exvenses and leave for immediate needs of the beneficiaries. 1915 THE IRON AGE 135 Book Review The Electric Furnace. By Alfred Stansfield. Pages, XIII + 395, 6% x 9% in.; illustrations, 155. Published by the McGraw-Hill Book Company, New York. Price, $4. This is a second edition, revised and enlarged, the first appearing in 1907. In its 400 pages it covers the whole field of electric furnaces, describing those for aluminum, calcium-carbide, brass, carborundum, zine, tin, copper, nickel, alundum, ferrosilicon, ferrochrome, quartz, phosphorus, carbon bi-sulphide, silicon, gra phite, nitrogen, caustic soda, sodium, magnesium, etc., besides most of those for the iron and steel industry. In trying to cover all industries using electric fur naces it does not give the reader a comprehensive idea of the possibilities of these furnaces in any one indus try. The volume is a good example of the printer's art and has many excellent illustrations, some show- ing great detail. About one-quarter of the book is devoted to de scriptions of electric furnaces and processes for mak ing iron and steel. In the limited space given to this important branch of the subject Professor Stansfield not only brings out the good points of nearly all of the electric furnaces used in foundries and steel mills, but gives space also to many furnaces existing only on paper. His work shows the hand of the careful re viewer rather than of the experienced metallurgist and practical operator. The central thought of the book is the great prog- ress made in all electrochemical industries using an electrothermal or an electrolytic and electrothermal! furnace. In referring to the use of direct current in the electric furnace, the author would limit its em ployment to cases where it is needed for electrolysis—a position which calls for emphasis in view of the fre quent attempts to use direct current where only the thermal action is wanted, the electrolytic action prov- ing to be a positive detriment. It would have been more correct, in saying that “Bessemer and open-hearth steels of practically pure iron only differ from wrought iron in having been fused,” to add that their physical properties are vastly different. Moreover, it is hardly correct to say that “in making steel by the Bessemer process the carbon is completely removed and the necessary amount is added in the ladle.” The average amount of carbon in blown metal before recarburization is well known and making steel in the ladle has gone out of fashion with the better steel makers long ago. The Lyon furnace of 1909 for making electric pig in California is said to be “substantially the same as the Domnarvet furnace.” There is a similarity in their construction; the great difference is, however, that the Swedish furnace operates as an are furnace, whereas the 1909 Lyon furnace was operated as a resistance furnace. Much less space might well have been given to paper furnaces, laboratory furnaces, college experiments and to furnaces of which only one has even passed the ex- perimental stage, while more should have been devoted to metallurgical discussions; for example, on the ad- vantages of electrically melted ferromanganese for the steel maker. The treatment is perhaps as thorough as could be expected from a book covering so larwe a subject, em- bracing electric furnaces for all possible uses, in com- paratively so few pages. It should be welcome as a preliminary review for any interested, prior to study- ing specifically the furnaces belonging to one of the many industries represented. Cc. H. VOM RB. The Swedish iron and steel market has decidedly improved from the depression of 1914. The exports for the first three months of 1915 were 90,800 metric tons, compared with 74,200 tons to April 1, 1914, and 87,900 tons to April 1, 1913. The pig-iron exports were 43.300 tons, against only 19,600 tons for the first quarter of 1914. Of the country’s 125 blast furnaces 97 were in operation on March 31. d THE IRON AGE July 15 High-Speed Blowers and Compressors The characteristics of one of these blov shown in the accompanying drawing. It ex. The De Laval Steam Turbine Company, Trenton, N. J., has developed a line of centrifugal blowers and compressors that can be operated at peripheral velocities of 450 to 600 ft. per second. It i 1S pos- plained that the form of the head-delivery teristic can be modified according to the wor done. Where forced draft is supplied t furnaces, the head-delivery characteristic sible in practically all cases to connect the blower or compressor directly to the De Laval steam turbine and a pressure of from 3 to 4 lb. per square inch can be generated in a single-stage blower. Among the fields for which this blower can be em- ployed are the supplying of forced draft to boiler furnaces and for coal gas plants and coke ovens. The housings of the double-suction type are in one piece, and are usually made of cast iron. The bearing bracket is cast solid with a cir- cular inlet ring or bell fitting into the eye or open ing of the casing, which is of sufficient diameter to enable the impeller to be removed. The housing is formed so as to provide a diffusor and volute, the discharge opening of which is adapted for the attachment of piping. The impeller is built up on heat-treated chrome-nickel-steel disk, thickened at the center to form a hub and to turn the air which enters axially in a radial direction. One side of each blade or vane is riveted directly to the disk, while the other is fastened to a steel side plate of the same material as the hub. The entrance edges of the blades are set at an angle, which is relied upon to receive the air without shock, and are formed to give the desired characteristics. A laby- rinth packing or tightening ring attached to the side plate meshes with a corresponding groove in the circular entrance nozzle. This arrangement, it is emphasized, minimizes leakage from