The Iron Age 1926-08-19: Vol 118 Iss 8

THE IRON AGE New York, August 19, 1926 ESTABLISHED 1855 i. VOL. 118, No. 8 Motor Plant Steps Up Production Changes in Equipment Reduce Time for Machining Fly- wheels, Clutch Disks and Other Parts—Heat-Treating Furnaces Equipped with Loading Device BY L. Plainfield, N. J., plant of the International Motor Co. were described in THE IRON AGE, Aug. 16 and 23, 1923, pages 389 and 481, respectively. At this plant motors only are manufactured. Since that time the company has had to resort to still further improve- ments to keep in step with the increasing demand for Mack trucks and the new development of Mack buses. The latter field has called for two new motors, a four- cylinder and a six-cylinder, each cast en bloc, whereas the motors for trucks are built with cylinders cast in pairs. While the manufacture of these motor blocks has called for no new methods of production of out- standing interest, per se, it has put a load demand on some of the parts departments which had to be met with increased output. How the demand for stepping- up the output of some of the new motor parts was met is outlined in this article. An explanation of the method of setting piece rates at this plant will assist in understanding the produc- GP of the production methods in use at the Fig. 1 (below)—A Ma- chine for Each Multi-Cut Operation of Roughing and Finishing Both Sides of Flywheels Is Saving Approximately % Hr. per Wheel. Time between cuts is saved by air- chucking S . . LOVE tion time given in connection with some of the pieces shown in the accompanying illustrations. The time studies are made on the decimal basis, and production time taken in hundredths of an hour. When a new method is installed a first-class operator is timed in producing ten pieces. To the actual average time re- quired, 30 per cent is added to establish the allowed time as a piece rate for the job. Any reduction of this time by the regular operator earns for him a bonus. Comparisons given below of old time and new are of allowed time. . Output of Flywheels Increased An outstanding factor in the new time reductions was the introduction to the shop of Simplimatic lathes. These machines, while they perform the service of tur- ret lathes in the production of several operations through multiple tooling, have no turrets. The two tool slides are provided with a group of tools to do desired work and are also equipped with special barrel Accuracy Ie Held by Hardened Register Ping for Finishing Cuts. Clutch fit and rim are machined im one setting. Fig. 2 (above) shows method of roughing crankshaft flange fit and the seat of other side of wheel 7 : : : MR RRsat on NE NO Been ett te 474 THE " cams incorporated in the feed mechanism, one for each slide, to control the feed and rapid advance or with- drawal of the slide by power. Fast and slow turning of the cams and fast and slow rotation of the lathe spindl may also be controlled by gear changes incor- head end of the machine. Aside from machine in a group is 1ipped to do only one sequence of operations, and is * chi ft nce set up. Yet these are not strict- ngle-purpose machines, for by changing the cam y may, on a change in design of the porated in the tnese possibile changes, each » machine the changed piece, or to handle an entirely different piece. To facilitate chuck- ng and reduce time between cuts, each machine is pro- { r-operated chuck. y of four of these machines is engaged flywheels, such as shown in Figs. 1 to 4. f these flywheels, 19 in. and 22 in. in diam- (at Right)—Too f Multi Cuts Are irranged on Two Slide: } Kith Roughing ” hing Operations 18 are shou i eter, are machined. Change from a lot of one size to a lot of the other size is made simply by adjusting the chuck, changing the tool slides with tools and the tool slide cams, to control proper tool advance, feed or dwell for one slide or the other. The four machines are used for roughing and finishing the two sides of the wheel, one machine being employed on each operation. The roughing operation, Fig. 1, on the first side turns the wheel, faces and bores the flange, faces the clutch fit, the hub, and bores. The wheel is next re- moved to the second lathe in the group for the rough- ing operation on the other side, Fig. 2. In this opera- tion the wheel is gripped by outside jaws on the periphery and the machining consists of facing the Pa second side of the flange, boring the flange and also ey boring the crankshaft flange fit. The arrangement of tools on two slides for multiple cuts is shown in Fig. 3. The chucks for the finishing operations, instead of depending upon the accuracy of chuck jaw faces for close setting, are provided with three hardened stops each, with ground faces, as shown in Fig. 4, against which the turned face of the flange abuts. While in the first finishing operation chucking registers from a surface which is only rough turned, the second which calls for one dimension of extreme accuracy registers from a completely finished surface. In the first finish- ing set-up the outside diameter is turned, the flange is faced and bored and has a radius turned on the outside edge. Also the clutch fit is faced as well as the hub, being faced and bored. The final operation, and the one calling for the most IRON AGE August 19, 1924 precision in the turning of flywheels, is the finishing of the second side. In this set up, as in that for finishin the first side, the casting, with one side completely fi: ished, is registered with that side against the thre. stops with hardened and ground faces. The machinin; consists of facing and boring the flange, with a radiv turned on the edge, and the boring of the crankshaf: fit. This last operation is checked with a “go” and “no-go” gage and is held to a tolerance of plus « minus 0.0005 in., as it comes from the lathe, withou: any subsequent reaming operation. Without increasing the number of men on the fly wheel job there has been a marked increase in the pro- duction, with a saving of 43.8 min. per wheel. The old method employed two men on two turret lathes and re- quired 0.93 hr. per wheel. The new method employs two men on four machines and turns out a finished wheel in 0.20 hr., a saving of 0.73 hr. per wheel. With Peme Da es Fig. 4 (Left)—Hardened and Ground Stop Pins Register the First Fin- ished Side for Proper Setting for Finishing the Other Side of the Fly- wheel the complete battery a wheel entirely finished is turned out every six to seven minutes. Time Saved in Machining Clutch Disks Further use of the Simplimatic to reduce produc- tion time has been made in the manufacture of clutch disks, shown in Fig. 5. On this operation one man cares for two machines and equals the former output of three men. The disks are 21 in. in diameter and must be faced on the clutch side and turned to suit the clutch fit in the flywheel. The old time for this side was 0.26 hr. In the second machine the back of the disk is faced and the bore machined in 0.65 hr., making a total time for the finished disk of 0.91 hr. By the present method disks are turned out at the rate of one every 0.105 hr. or 6.30 min., a saving of about 48 min. per disk. Automatic Crankshaft Lathes Facilitate Production Another recent innovation at this plant is the in- stallation of a battery of LeBlond automatic crankshaft lathes. These machines are used for turning crank- pins which are late: finished to exact size on grinders. The former time for turning these pine was 0.75 hr. per shaft and the work was done in a non-automatic lathe, one pin at a time, but all four pins being done in the same machine. In the battery of automatic lathes, part of one of which is shown in Fig. 6, each machine turns two pins at once with tools mounted on double carriages at front and rear. The two slides on the rear carriage carry \ugust 19, 1926 g. 6 (at ght) Mul- ¢ Cutting d Machining f Two Pins at Time Make a Saving f 86 Mén. Each on Four- Crank- shafts two tools each. The slides on the front carriage carry one tool each placed midway between the rear tools opposite, but with cutting edges slightly overlapping. In this manner the pin journals are turned on the ends by the rear tools and the middle sections by the front tools. Two machines are used for a four-throw crank- shaft and turn the pins to within 0.003 in. of finished size, which amount is later removed by the grinder. The turning time is 0.15 hr., a direct saving of 0.60 hr. or 36 min. per shaft. Other recent improvements in methods have to do with grinding operations. The first of these is on a camshaft machine, Fig. 7, and consists of a flexible Fig. 5 (at right)—A Saving of 45.6 Min. Is Made on Each Clutch Disk. This is regarded as no mean accomplishment in this day of split-second machine production THE IRON AGE 47 Fig. 7—Avoidance of Chatter, Thereby Eliminating the Hand Stoning of Cam Noses, Has Re- sulted in a Saving of 4 Hr. per Four- Cam Camshaft. Tolerance also re- duced coupling drive in place of the older geared drive method for controlling movement from the master cam to the cam being ground. It is stated that the change has saved 4 hr. per shaft when it is considered that due to lack of chatter or gear marks, no stoning of the nose of the cam is needed. The actual time on the grinder has also been reduced from 0.65 hr. to 0.55 hr. The machine equipped with this flexible coupling driven :on- trol has also enabled the company to cut tolerance limits 40 per cent. The other grinding operation is a development from finish turning pistons, to grinding by the “plunge” or straight-in feed method with a wide wheel, then to the Fig. 8 (at left)— Traverse of Work Past Grinding Wheel or Vice Verea Hae Been Found Desir- able, to Maintain Accuracy on Pistons Gv peering ep taeah nates dace see neee einem EI Us AMINE: oe RR A Es ac ah heat Rehan der ha a Po hil ii 8 ae em 1 Nee tl eB ti os 8 e- = —_ 476 THE IRON AGE 3 \ - <> acai use of a standard grinder with narrow wheel traversed over the face of the piston, Fig. 8. It was found that, due to spring in the casting, as between the skirt end of the piston and the rigid closed end, tolerances could not be maintained by the straight-in feed method. These tolerances are plus 0.000 in. for roundness and minus 0.001 in. for taper. Also the time has been cut two-thirds over the previous method of finishing pis- tons. Improved methods have also been installed in the heat-treating department where a number of furnaces have been provided with an unusual type of loading New Company to Rust-Proof Steel A new company has been incorporated, entitled the Cromilite Process Corporation, by interests identified with the New Haven Sherardizing Co., which has a plant in Hartford, Conn., and with other interests identified with the Metal Protection Corporation of Indianapolis. The officers of the new company are: President, August F. Schoen; vice-president, James E. Patten, Indianapolis, and treasurer, Wesley I. Charter, Hartford. The new company will engage in the rust- proofing, metal finishing and chrome-nickeling of steel under a process invented by Mr. Patten. The first unit of the new company’s plant is to be established at Akron, Ohio, and its operation is expected to commence some time this month. The company is also acquiring property adjoining the Hartford plant at Windsor Street and it is stated that a third plant will be estab- lished in Springfield, Mass. New German “Wafios’” Barbed Wire A German barbed wire machinery manufacturer has patented a machine for barbed wire of one strand and two German manufacturers have been licensed for the sole production of this sort of barbed wire. The sales are said to be satisfactory. The per ton prices are higher than the usual two-strand wire, but the linear prices are much lower. At present the B. W. G. No. 12 wire is sold at 17s. 10d. ($4.25); the B. W. G. No. 13 wire at 18s. 744d. ($4.50), and the B. W. G. No. 14 at 19s. 6d. ($4.75) per cwt., f.o.b., which means about 6s. to 7s. per cwt. more than the usual barbed wire. The difference is the weight, as 1000 kg. of wafios wire has a length of 15,530 meters (B. W. G. No. 13) and of the usual barbed wire only 8600 meters, so that 1000 meters of wafios wire weighs only 644 kg., August 19, 1926 Fig. 9—A Mo- tor - Operated Loading De- vice Helps to Increase Furnace Out- put in Heat- Treating De- partment device, shown in Fig. 9. This consists of a pusher bar extending across an extension of the furnace bed. The pusher arm is operated by crank levers actuated by motor drive and gear reduction mounted below the ex- tended furnace bed. Work is placed to fill the surface of this extension, when the motor is started and the pusher bar is called on to push that load through the furnace door and return to its outer position, when an- other batch is placed for insertion into the furnace. This process is continued until the full depth of the furnace bed is filled with parts to be annealed, car- burized or otherwise heat-treated. whereas the usual wire weighs 113.3 kg. The buyer of wafios wire is getting, therefore, if buying one ton, about 75 per cent more wire and is paying 50 to 60 per cent more for it. The two plants making this wire are selling monthly increasing quantities in Germany and other European countries and have had success overseas. Spanish Ore Shipments Low WASHINGTON, Aug. 16.—With only 31,000 gross tons of iron ore exported from Bilbao during July, as against 89,000 tons in June, the situation in the Spanish iron ore industry is becoming more unfavorable, ac- cording to a cable from Commercial Attaché Charles H. Cunningham, Madrid. Stocks on hand amount to 800,000 tons, and several important mines of long standing have had to shut down. Canadian Automobile Output Gains The total automobile production in Canada during the six months’ period ended June 30, 1926, amounted to 124,878 cars, as compared with 98,647 in the cor- responding period of 1925, according to a report to the Department of Commerce from Lynn A. Meekins, trade commissioner at Ottawa. Automobile production in Canada during June totaled 21,571 units, of which 17,785 were passenger cars, 2586 were trucks, 1354 chassis and 26 taxicabs and buses. More than three miles of enameled metal shelving are contained in the new office and hardware wholesale warehouse of Hibbard, Spencer, Bartlett & Co., Chi- The new building is 13 stories in height and “aspaed all of this shelving has been placed on three oors. cago. How Metal Tubing Should Be Tested Tensile Properties as Affected by Shape of Test Specimens—Tube Diameter the Influential Factor BY N. S. [ is often desirable in conducting tensile tests for | determining the properties of material in struc- tural shapes to load the entire cross section, thus minating the expense of machining the conventional und or flat specimens. Perhaps the best example of his is found in testing tubing which is extensively ised in aircraft and other construction as an echro- ymie section. Heretofore, it has been generally assumed that such full section tests indicated properties of material! di- rectly comparable with those obtained when testing the conventional round and flat specimens. This, how- ever, is not the case since the characteristic of the section loaded has a very prominent influence on the ndicated properties of the material. This assumption has led to errors in test results with consequent fail- ires in service. Ordinarily, tubing that has a tensile strength within the capacity of the largest testing machine convenient- ly available is tested in full section while larger tubes are flattened and specimens prepared similar to those used for sheet metals. Furthermore, it is current prac- tice to specify that tubing show given tensile proper- ties regardless of whether full section or flat specimens are used. Under these conditions test results may be in error as much as 100 per cent and, unfortunately, these errors are on the unsafe side leading to the acceptance of materials having low ductility. Some recent tests made at the Naval Aircraft Fac- tory demonstrated that the percentage elongation over a given gage length of a tube, tested in full section, was very different from that of a flat specimen pre- pared from the same tube. For example, three dif- ferent types of tensile specimens were machined from a single bar of mild carbon steel to dimensions shown in foot notes of Table I. They were all of conventional design and the round and flat types were so cut that their longitudinal axis coincided with a line midway between the inside and outside periphery of the tube specimen, thus insuring comparable material. These specimens were tested with results shown in Table I and Fig. 1. With these and other preliminary tests as a basis, a more extensive investigation was made covering three classes of materials: S. A. E. No. 1025 steel, S. A. E. No. 2330 steel and an aluminum alloy, commonly known as duralumin or 17S. Aircraft Factory, Philadelphia *Engineer of tests, Naval 80} — 10}— 6 s — S i 50}— io = 40/— | & | Mud pe r | _.|__ filet Specumet : OTEY* The tube diameters and gage thicknesses of ma- terials used in this investigation are shown in the first two columns of Tables II, III and IV. In order to ob- tain materials of reasonable uniformity, a process of selection from physical tests was adopted in which three tubes of each size and gage were tested and one of the three selected which had a tensile strength nearest the specified value, as follows: Specified Material Tensile Strength S.A.E. No. 1025 Steel ( Normalized) 000 Ib. per sq. in S.A.E. No. 2330 Steel (Heat- Treated) 125,000 Ib. per aq. in Duralumin (Heat-Treated) 5,000 Ib. per aq. In, The specified chemical analyses under which these materials were purchased and inspected were: Steel Carbon Mang Phos., S.A.E. No Per Cent Per Cent Per Cent 1025 0.20 to 0.30 0.50 to 0.80 0.045 Max. 0.040 max, 330 125 to 0.30 0.50 to 0.80 Sulph Ni., Per Cent Per Cent 0.050 max. 0.045 max 3.25 to 3.76 Duralumin Al., Mang., Mag., Copper, Per Cent Per Cent Per Cent Per Cent 92.0 min 0.4 to 1.0 0.2 to 0.75 3.5 to 4.5 From each test tube, as above selected, three full section or tube specimens and four flat specimens were taken, Fig. 2, and were machined to dimensions shown in Figs. 4, 5 and 6. The flattening of sections for flat specimens indicated in Fig. 5 was done by pressing be- tween steel plates in a testing machine. This operation does not seriously affect the tensile properties of the material since cold working occurs principally in a direction normal to the longitudinal axis of the speci- mens. Furthermore, these effects of cold working are practically eliminated by heat treatments applied sub- sequently to flattening and machining. In all cases the flat specimens were placed inside of their corresponding tube specimens during heat treatment operations to aid in maintaining uniform temperature control of each lot. The following heat Fig. 1—Reaults of Tensile Tests on a Single Bar of ; S.A.E. No. 1025 Steel, Using Three Types of Specimens, Shown eat in Table I ne (rl eran ia thi ante sh lta: Staite ets itt an pha ni gs ite PR ROE Ip RAO BRE NN = : f) j : : : : ee fj met Hl rhe $74 itt fi] Lif i aa 4 4 thi <a : 4} 5 re all machining eT? held for 20 I ‘ 2 I ue ; I t ed) dium and pe Ss 201 deg. Fa quen d / , oun Two Flat \One PF 4 _ —_ 4 =— j ( f Specime fron Tubes 4 sg - ne e " c Fig a G+/0D a May 6+ tD+ 9 ~ A ? bus - ‘rf 6+2D+ = 7 + - ay f , - "D vr D D 4 6 A “4 Fig. 4 , Sec? + trom Tube rot to he < > > , eate 7 gry , = > a+ re viatte Ng Ira aay” ared ing 40 FIG.6 I ig 5 ec - . pap ee --— FY a G ; 7 Fig. 6 in water at room temperature After this treatment the material was allowed to age in still air at room temperature for 25 days before testing Specimens having a computed breaking load below 50,000 Ib. were tested in a 50,000-lb. capacity uni- versal testing machine and those of greater computed load in a 200,000-lb. capacity machine. Auxiliary self- aligning test grips on flat specimens and the set up shown in Fig. 4 on tube specimens were used. Much care was exercised in conducting these tests to insure the elimination of any variables in testing methods. Complete results of these tests are shown in col- umns (C), (D), (G) and (H) of Tables II, III and IV. August 19, 1926 on - ee ee a > Table I—Comparison of Elongation of Round, Flat and Tube Specimens Cut from Single Bar of 1025 Steel Gage Percentage Elongation Lengt SN nin Tube Round Flat \, 70 68 80 ay 53 38.8 40 12 41 26.9 285 a an 9 21 9 @o « é 35.2 21.2 22.2 e 32 19.1 20.2 2u, 29.5 17.6 18.7 2% 27.5 16.5 17.2 24 ly 21.5 4 19.3 Lit. tensile strength, lb. per sq. in 70,600 71,500 70,500 Dimensions : Tube—1.372 in. outside diameter and 0.878 in inside diameter. tound—0.500 in. gage diameter and 3 in. parallel length Flat—Same as shown in Fig. 2. 1°) |; VED SOESERES OROLEBEMOBBDET TAP THEIL CTT The values shown are the average of three full section and four flat specimens. From these data it will be observed that no well defined difference in tensile strength exists between full section specimens and flat specimens of the same material. The elongation of full section specimens, however, is consistently greater than that of flat specimens and further examination may well be confined to these differences. As a probable working. basis for analyzing these results, let us assume that the differences in elongation may be due to variations in: (a) Tensile strength. (b) Geometric characteristics of the section loaded. Variations in tensile strength for a given materia! can be practically eliminated by correcting the elonga- tion values to correspond to specified tensile values, which for these materials and heat treatments, are as follows (on next page): O+:c01)/4c10C0NMERON ORE :ONEPDELEDS LED OY FTO TERESCORATRONOEANSRESEE YOOERESEEDURPEDEES OREEDOTTTREGDES “EDEL WEOTTNERTUENESNECERSS ELSIE Fe Table I11—Comparison of Tube and Flat Specimens of 1025 Steel, Normalized S = , pee . , & E = 5 $ . £2 ee Pe ee eS > © & 2 S38 %E bem ee ez Ors si g5 G8: G52 Se 82 GE BE Sce o> Oa =o 256 A> mS 3 3 ou =o = “zn 2 ZnS #2 w§% B32 2 23% + -¢& SS eeae s be ca ea ¢< sa o2#5 Ee S§§ go” gsm 82 Sa 32 3a £ oe cE ge ‘Sag Fa, Og Of Ee Eu See S= S& «ep F868 SP S28 ss SS sec ZQ Ze Se Sik | A ik O88 Oh Aa % 0.028 56,400 55,900 32.2 25.0 33.2 26.5 7.7 % 0.028 55,300 55,300 37.0 30.1 37.2 30.3 6.9 % 0.035 61,200 56,200 35.5 31.0 41.1 32.0 9.1 % 0.042 55,300 64,800 36.7 29.5 36.9 29.3 7.6 % 0.058 57,000 655,500 41.0 382.0 42.8 32.8 10.5 % 0.035 64,560 63,100 39.5 30.3 48.5 36.5 12.0 % 0.049 52,000 50,700 46.8 35.5 43.5 31.9 11.6 % 0.058 55,600 54,600 41.3 80.6 41.5 30.2 11.3 1 0.035 65,760 66,200 36.2 27.0 45.6 34.2 11.4 1 0.042 63,300 57,900 36.3 27.8 43.5 29.5 14.0 1 0.058 52,970 50,500 46.7 353 44.5 31.8 12.7 1% 0.058 58,600 55,600 44.8 33.1 47.5 83.4 14.1 1% 0.035 51,600 48,600 44.2 32.3 40.8 27.5 13.3 1% 0.058 60,400 57,160 43.2 30.8 48.8 32.2 16.6 1% 0.049 63,700 62,370 38.2 27.0 45.1 81.8 13.3 1% 0.042 60,400 59,100 41.0 27.4 46.0 30.1 15.9 ‘a 0.035 63,360 61,900 41.5 28.5 49.8 34.0 15.8 Ye 0.058 64,960 60,200 43.8 28.5 51.2 32.0 19.2 “a 0.049 63,100 65.500 36.0 20.4 43.0 25.5 17.5 2 0.049 71,500 72,200 35.5 22.7 49.5 $2.2 17.8 “s 0.058 57,450 56,800 47.5 30.8 50.5 31.9 18.6 Fe (0058 59,050 57,900 47.8 30.0 52.2 32.0 20.2 1% 0.049 63,500 59,400 47.7 32.4 57.6 35.8 21.7 1% 0.058 61,300 58,600 43.8 28.9 60.0 31.3 18.7 1% 0.058 52,700 49,000 49.5 31.6 46.8 27.3 19.5 1% 0.065 57,700 56,600 52.2 32.8 655.5 34.0 21.5 2 6.058 58,300 54,100 53.8 36.8 658.5 35.0 23.5 2 0.065 56,700 54,300 52.0 $1.6 654.0 $1.8 22.7 a 2 Cc D @ HH Be Be SHCTUTTUEN ) ARTY EEN CHEN VEDEOTEDOGETNGDNERERECREENERERY CXTEETESET ETE Circular No. 101, “Physical Properti April 23, 1924. August 19, 1926 THE IRON AGE * Ch ee | TTL INeSlepeste |__| (ae Fig. 7T—Relation Between Elongation and Tensile Strength for 72 Different Steels, Bureau of Standards Circular No. 101 (at Left) Fig. 8—Material Is S.A.E. No. 1025 Steel (Normalized) ; Tensile Strength 55, 000 Lb. per Sq. In. Symbols: E per- centage elongation in 2 in. for tube speci - " 4 mens; e = percentage elongation in 2 | Carbon Stee/ in. for flat opesimens ; D = nominal out- ALI casteaee eonnoned side diameter of tube in inches; M modulus of elasticity (30,000,000 lb. per sq. in.); S tensile strength (55,000 Ib. 4 } per sq. mm.) 5 © Tt 20 s0 Tensile Strength - P/.1 x 10* S.A.E. No. 1025 Steel (Normalized), 55,000 Ib. per sq. in. tensile strength S.A.E. No, 2330 Steel (Heat-Treated), 125,000 Ib. per sq. in. tensile strength Steel, Heat-Treated Table Il1I—Comparison of Tube anf Flat Specimens of 2330 Duralumin (Heat-Treated), d = 2 55,000 Ib. per sq. in. tensile strength x a a E g e2 : , t te e 3 = ia It was assumed that the elongation varied as some : @ Ss. &. 4. 4. oe . ° . = > 2 . . 2 inverse function of the tensile strength, but their exact = ; 5:2 S:2 ME BE SE 88 Ons relation was not known. This relation was determined s§ an “She S52 S& 22 GE BE a=% approximately by plotting per cent elongation in two ©; Og Geo Gg 2 =e 32 33 $s% inches against tensile strength for 72 different steels’ 48 92 Sue Eta sa SP ta 32 5 as shown in Fig. 7. It was found that this could be 5—E go ag He 3 . Fe fe g turned to a mathematical expression representing the °- 5S& E55 556 G2 ge 66 Sh Aag inverse slope 1.2 of a straight line curve. a 1 0.028 160,000 155,000 65 35 89 6.1 38 > j ; ion values 1! 0.058 152,000 147,000 11.1 7.5 14.1 9.2 4. re lation (slope 1.2) thus obtained, the elongati : 1% Soak 148300 188000 86 643 106 «688. SR shown in columns (C) and (H) of Tables II, III and ! ii, 0.035 155,000 150,000 80 45 105 6.7 44 were corrected to correspond to specified tensile values 1% 0.208 142,000 142,000 21.3 15.5 25.0 181 6.9 given above. The numerical difference between the 1% 008 e000 oes 3) OY ise eH OG corrected elongation values for tube specimens (column 14 9.942 188.700 ishae 84 48 188 63 88 : own iy 0.0 ’ 5 4. . E) — tons am ee — e) are sh 1% 0.058 177.000 170.000 9.0 48 140 70 70 in columns E-e of the same tables. 1% 0.049 165,000 173,000 102 61 145 7.7 68 : far have been rea- i% 0.042 156,000 152,000 93 65.1 122 66 6.8 If the assumptions made thus he quantitative 1% 9.058 179,300 165,600 123 80 193 114 1.9 sonably accurate, it would appear that the quantit 1% 0.042 151,000 193,000 120 45 153 77 78 influence the geometric shape of full section specimens 1% 0.187 144,000 140,000 19.0 12.0 22.7 139 688 has : son is represented by the values 1% 6.068 161,000 158,000 128 61 17.7 81 9.6 on their elongation is rep from flat specimens, 1% 0.049 144,000 141,000 116 56.5 189 66 1.4 (E-e) since the values (e), taken from fla oo f the 1% 0-120 162,000 156,000 17.0 9.1 236 120 116 may be considered a fair index to the ductility of © 1% 0.049 163,000 156,000 132 71 183 95 88 material as ordinarily measured. , it hin f 0.058 143,000 136,000 156 82 18.7 91 986 In Figs. ese different elongation : 2% 0.058 159,000 157,000 13.7 7.5 186 99 6.7 (E n Figs. 8, 9 and 10 th + diameter of tube for each 2% 0.120 171,000 156,000 168 10.3 25.0 13.6 11.4 -¢) are plotted agains 2% 0.187 145,000 142,000 21.7 121 261 142 119 s Bureau of Standards’ A B Cc D G H E o E-e ‘ Values taken from United ee of Materials,” dated 19 ee Nominal Outside Diameter of Tubing, inches ° ; S.A.B. No. 2330 Steel (Heat-Treated), Tensile Strength 125,000 Lb. per Sq. In. Fig. 9—Material Is e elongation in 2 in. for tube specimens; ¢ = percentage elongation in 2 Symbols: E = pon. nominal outside diameter of tube in inches; M = Sena of elasticity in. for flat peoit0, 000 Ib. per sq. in.); S = tensile strength (125,000 lb. per aq. in.) 290 SA ROC ORBEA ARIA RIN Om er 480 THE IRON AGE August 19, 1926 tice are decidedly in error which permit a direct com- parison between tensile properties obtained from fu section tube specimens and those obtained from fiat specimens cut from the same tubes. Indeed som specifications require greater elongation for small di ameter tubes than for larger ones. This again reminds us of the very great need for 4 , k t Specimens of Tat 7} ( ‘ I d mMmeAtior in Correcte S z we FE 2 a standard test specimens and methods designed as scien- 62 se ae =¢ tifically accurate as knowledge affords which could b« ; ~ RS 5S sa 25" calibrated for conditions and materials they are t eB B ES = 5 =i 58 represent. Such accuracy is not required in all test- - HS2 Ue OF Bo pe Cee ing work; however, reliable standards are a very neces- 6 23.0 21.5 26.0 24. 0 a SS SEEN 7 29 200 258 22.1 7 7,900 2 20.2 25.1 22.0 1 f Rf 26. 2 23.9 25.1 5 29 t i 24 4 20 vail 25 é 28.1 290.9 é 00 25 2 28.1] 20.5 } 2 2 24.2 t 7 2 28.8 22.8 ‘ f 4 { 1S 2 20.9 4.f i ~ 23 5.4 ( G H E ‘ E-« Typical Failure of Duralumin Tube terial examined. It was found that these data could Fig. 11—Typical Failures of Steel and closely approximated by an empirical formula. The Duralumin Tubes rves drawn through these data clearly indicate that, r a given material, the most influential factor is the sary thing in setting tolerances and allowances for } ended to represent all three materials thus: ibe diameter. It was also found that, in working youtine tests. Our weights and measure systems are from one material to another, the empirical formula very good examples of the utility of fixed standards. The writer is indebted to the management of the V2D? X M X 0.015 Naval Aircraft Factory for encouraging this work and the Naval Intelligence Bureau fur permission of pub- s lication. n pecime!l (per = a ens (per cent Actual Cast Iron Pipe Increase in 1925 in PEL ae PO Was 4.5 Per Cent RNC LOD. POS -OG. SB.) An editorial in THE IRON AGE of July 29 on “High a ee ae es Output of Pipe and Tubes” stated that “from 1913 to "he curves shown were plotted from this formula. 1925 production of cast iron pipe fittings increased The quantity 2D° in the above formula was left 4 per cent, to 2,324,047 net tons or 2,075,042 gross this form to show its approximation to the radius tons.” The statement was based on the statistics for last year, published by the American Iron and Steel Institute. However, the writer of the editorial had overlooked the fact that revised figures recently issued by the institute showed that the output of cast iron pipe in 1925 was 1,924,735 tons, a reduction of about 400,000 tons from that originally published. The error was in the production of gas and water pipe, including culvert pipe. The institute’s revised total for such pipe was 1,414,252 tons, whereas the original figure was 1,813,- 564 tons. Adding to this the figure for soil and plum- bers’ pipe, which remained unchanged at 510,483 tons, gave the total of 1,924,735 tons. We regret the repe- tition of the incorrect figure in THE IRON AGE, since the case of full section specimens) and ratio of gage jt has led to some exaggerated statements in financial thickness to gage length are of minor importance when reviews concerning the expansion of the cast iron pipe compared to the influence of the tube diameter. This industry in the past year. The actual increase over orne out by the reasonable agreement of these data the 1924 output (1.841.350 net tons) was 4.5 per cent; with the empirical curves shown. : ae but 1924 had made an unusual record, being 20 per cent It is evident that current specifications and prac above the best previous year gyration of the tube. How the modulus of elasticity, ng an elastic property of the material, affects the ngation which is a plastic property, cannot be readily explained. Fig. 11 may shed some light on this point which it will be noted that, in the failure of a steel the principal reduction in diameter occurs at the point of fracture whereas for duralumin tubes a majority of the reduction in diameter occurs at the lug. It may thus be that, in the early stages of the test, the elastic properties, or stiffness of the material, control the point of greatest deformation. The influence of such variables as gage thickness, ratio of tube diameter to distance between plugs (in is i : TT V>De eT T ; ; ; . , J s iieasijihieietilonlil = F-e at D<x M x 0.0/5 ] & 5} 6 , | | | ? Wy —— I aa le L Dee | Q 05 10 1.5 2.0 Nominal Outside Diameter of Tubing, inches Fig. 10—Material Is Duralumin (Heat-Treated) , Tensile S ? trength 55,000 Lb. per Sq. In. Symbols: E = percentage elongation in 2 in. for tube i 5 : _ pe ) . specimens; e = percentage elongation in 2 in. for flat specimens; D = nominal outside diameter of tube in inches; M = modulus of elasticity (9,750,000 lb. per sq. in.): S — ’ per sq. in.) tensile strength (55,000 Ib. Silicon as an Alloy in Steel German “Freund” Steel Compared With Other Steels of Similar Composition—Effect of Silicon, Manganese and Nickel—Use for Structural Purposes BY DR. H. W. GILLETT’ the Deutsche Reichsbahn Gesellschaft, are show- ing a good deal of interest in a “new” low-carbon high-silicon structural steel known as “Freund,” or “F,” steel which the Aktien Gesellschaft Freund first produced in 1925 by melting in the Bosshardt furnace. Extensive study of the steel was carried on by the Verein Deutscher Eisenhiittenleute, and reported in Stahl und Eisen® and in the Zeitscrift des Vereins Deutscher Ingenieure*. Their work was aimed not only to study the proper- ties of the steel and its behavior in rolling, but also to show whether or not it could be successfully made in furnaces other than the Bosshardt. Their conclu- sion was that it could be, and their data show quite definitely that it can be made in the open-hearth or the electric, and probably even in the converter. The Bosshardt furnace is a special, patented form of open-hearth furnace, so far only built in a 3-ton size, which is said to be operated at an extraordinarily high temperature. The claim has been made that the high furnace temperature is responsible for the properties of the steel and for its alleged freedom from oxygen. Discussion still rages in Germany as to the need for the Bosshardt furnace and while the Deutsche Reichsbahn Gesellschaft is having a 10-ton Bosshardt furnace installed at the Linke-Hofmann-Lauchhammer A. G. for a more extended comparative test, the general opinion seems to be that the analysis is the essential thing and not the production of F-steel in a Bosshardt furnace. This is the more probable because work done in Italy upon steel] whose composition and properties prac- tically coincide with those of the “new” steel was de- scribed by Paglianti® 14 years ago, and steels varying but little in composition and properties were used in England, in the steamships Mauretania and Lusitania, about 20 years ago. Similar American steels made for the Navy were reported on by Bisset in 1910°. All these were of course made in ordinary types of fur- naces, The term “new” applied to the “Freund” steel ap- pears to be a misnomer as regards metallurgical knowledge of steel of high-silicon content, although quantity of production of high-silicon steel of quite as low a carbon content as the Freund composition does not yet appear to have been undertaken outside Ger- many. The effect of silicon in structural steel has been on record as a result of the work of Tetmajer, published in 1884, and of that of Guillet in France, Sir Robert Hadfield and Thomas Turner in England, before 1900, as well as of many other workers in more recent times. Especially in view of Paglianti’s work, it is difficult to see how any claim to “newness” of the Freund steel could be substantiated, nor have any patents been cited that deal with the composition. Since it would thus appear that no patent on the composition and no requirement of a special furnace ( ' ERMAN structural engineers, especially those of 1 Published by permission of the director of the United States Bureau of Standards. 2 Chief division of metallurgy, Bureau of Standards. * Vol. 46, 1926, page 493. * Vol. 70, 1926, page 861. 5 Metallurgie, Vol. 9, 1912, page 217. ‘ A Aug. 25, 1910, page 442. ee ee ae jones American Society of Civil * Bur G. K.—Transact Engineers, Vol. 86, 1923, page 1292. 481 would lie in the road to commercial manufacture of the high-silicon steel, attention may be concentrated upon the properties and usefulness of steel of that type Tests of Freund Steel Specimens of five heats of Freund steel were sub- mitted to the Bureau of Standards for test and, while they were insufficient in number to establish by them- selves the properties of such steel, the results of the tests are in general agreement with several extensive German tests. A full account of the tests and a com- parison with the German data are to be issued as a technologic paper of the Bureau of Standards, in which published work on high-silicon steels is reviewed. Only a very brief abstract of the information available on this type of steel can be given in the space here allotted. The outstanding feature of the steel is its composi- tion, about 1 per cent silicon and not over 0.15 per cent carbon, the combination giving a high yield point and high ductility. The properties of the steel are com- pared with those of other structural steels in the table, from which it will be seen that high yield point and high ductility may, with sufficiently low carbon, be ob- tained by raising either manganese or silicon or by the use of nickel. Nickel in American Practice American practice has been to raise the yield point by the use of nickel, or, as in the Delaware River Bridge (see M in the table) to use a little silicon and much more manganese, without materially reducing the carbon from the percentage used in unalloyed struc- tural steel. Considering only the cost of the alloying elements, silicon is sometimes more expensive and sometimes cheaper than manganese. If large scale manufactur- ing tests, needed to establish the cost of production, should show that the cost of finished low-carbon silicon steel would be less than that of structural manganese steel at periods of high prices for manganese and low prices for silicon, it would appear that American manu- facturers might well consider being prepared to make the silicon steel. German engineers appear to be satisfied with the properties of the steel but they require further i:for- mation on the cost of the steel when made in regular large-scale production. Since German steel makers are actively working to obtain that information, it seems probable that the steel will soon be in commércial use in Germany, and its export in competition with Amer- ican steels may ultimately be expected. From the German data and the Bureau of Stand- ards tests, the properties of the steel make it seem suitable for ordinary structural use. The claim that oxygen is absent was not substantiated either in the German or the bureau tests. The steel was evidently treated with aluminum, and carried too many non- metallic inclusions to make it promising for exacting use under repeated streas. Notwithstanding the “dirtiness” of the German steel, it gave 58 to 86 ft.-lb. in the Izod impact test, and the specimens did not break completely apart. The proportional limit of the specimens tested at the Bureau of Standards, (using a Ewing extensometer on an 8-in. test length) ran from 49,000 to 57,500 Ib. per sq. in., and averaged 53,000 Ib. The steels contained up to about 0.20 per cent copper and up to 0.15 per cent a Cie hii iss aS ghee: > AA A RRO Ms Ee mR A Rm ML Mt eeee Sich eas MANNY ia Ea a eee ew a ee 422 THE IRON AGE oensnannunenevoveensaenstec mmette qmmmnasnennccsesencensenenecnsannsnnen ravens irene nennenennensemnansuanas tistneenaneaeel August 19, 1926 sunenanopee 04a ose e¢8000A088SAN000585800SNN0NSANNSDOGRONORROEEDEDOSSSEOUSESOOND-SEEAUNOOONSHSHUSOUCERSNPUNGASUNENUEDOSENSEUESSIOTISDLESNOOOSSISISOUUSSS ESSENSE ONSEN RRR Table of Properties of “Freund” and Other Steels Elonga- , Tensile tion, Reduction Yield Point, Strength, Per Cent of Area, Mr S Ni Cr Lb. per Sq. In. Lb. per Sq.In._ in 8 In. Per Cent — 6.500 76,000 25 63] : . a = 48.500 68,500 22 45 | B 59.500 75.500 27% 68% | a 4 55.500 72,500 27 61 Low-carbon silicon steels of “Freund ; 54.500 70.500 22% 48 composition. hs ry , 53.500 69,0006 28 55 E 60,000 78,500 28 45 Fe 52,500 82,500 26 29 High-carbon silicon steels. . 62.500 92,000 25 45 : 18 19.000 76,000 23%* 67 Annealed nickel. I : 60.000 100,000 18 33% #$=Nickel. : > a 64 56.000 90.000 13 Nickel-chromium, annealed and i " normalized, 52.500 90,500 25 48 as < 16,500 79,000 25 52 Plain carbon, with some silicon. 60,000 82.500 26 54% High manganese, low carbon. 36,000 50,000 44* 72 Plain low-carbon rivet steel. 30,000 55,000 251 os Specification, structural. 45.000 80,000 18% 30 Specification “silicon’’—structural, 5 36,000 55,000 4a German “No, 37.” 40,500 72,500 ad e German “No. 48.” 44.500 76,000 25 _ German “No. 58.” Ay Freund” steels made at Bureau of Standards. Mi ‘Freund” steel from German tests : ric si ns ; I irt licon s (yer! en-hearth silicon ste¢ }? on ste ) Bi Rg I te | ) ‘ footns é Ste M ‘ m figures given by Hadfield, R. A., “Metallurgy and Its Influence on Modern Progress, Al see Kreb KF. W Mech. Eng. 48, 1926, page 448. K. Nici Louis Bridg ‘ see Burgess, G. K., footnote 7 Mayari) Memphis Bridge | M High-manganese structural steel, Delaware River Bridge—Bureau of Standards test. P at tural steel, Delaware River Bridge— Bureau of Standards test. High ma inese low carbon steel—see Lang, G.—Stahl und Eisen 31, 1911, page 181. R t iverage hosphorus and sulphur committee tests—Proc. A.S.T.M. 22, (1) 1922, page 108. Q St tura teel for buildings—A.S.T.M. Specifications A-9-24. R Struct il ste tentative A.S.T.M. specification, A-9-25T for “silicon” steel. , H 5 Minir figures for three grades of German structural steels reported by Graf and by Gehler—see Bauingenieur e 2 nd 5, 1924, page 630 + CUECTE TER URODLUOOEE OHO ORERESOOELORORO OREN ORATERERODS POERE NT EE ceEROREESETEREORO CET RONEN renreremmRDCrIOT”@ chromium, neither one intentionally added. Some Ger- man analyses show a little nickel, also adventitious, from the scrap used. The steel is a little lighter than ordinary structural steel, the specific gravity averag- ing 7.78 against 7.84 for the ordinary material. If, as seems probable, engineers in this country come to designing statically stressed structures on the basis of the yield point of the steel, as is being done in Germany, a demand for material having a higher yield point will grow up here just as it has in Germany. With such a demand, the problem of the steel maker will become how to produce such material at the lowest possible cost. The Cheaper Alloying Elements The possibilities of the cheaper alloying elements, in steel to be used as rolled, as well as in the heat- treated condition, deserve more attention than has heretofore been given them. Manganese is already gaining recognition as the alloying element it really is. The amounts of silicon ordinarily used have generally Apprenticeship Study of Chamber of Commerce of the United States A relation between employer and minor wherein mutual obligations and duties are agreed upon (usually in writing—the indenture) by both parties concerned, is one definition of apprenticeship set down in a 50- page pamphlet issued by E. W. McCullough, manager of the department of manufacture of the Chamber of Commerce of the United States, and prepared by Rob- ert H. Spahr. The publication does not attempt an exhaustive study, but offers valuable information selected from the experiences of such States as Wis- consin and Massachusetts, where State supervision has made great strides in the training of future skilled workers in the trades. The impression prevails, but the pamphlet points out that it is a mistaken one, that the training of apprentices is a monopoly of the large company. With the aid of community or State cooperation the small been considered as having relatively little effect upon strength. Yet, by raising the silicon content to around 1.00 per cent, one can raise the yield point of a 0.13 per cent carbon steel from 36,000 to 56,000 lb. per sq. in., retaining good ductility (elongation of 25 per cent in 8 in. and a reduction in area of over 60 per cent). Hence its influence as a potent alloying element can scarcely be denied. Reaching this yield point by raising the carbon in- volves a sacrifice in toughness and ductility. Reach- ing it by adding nickel increases the expense. If the choice of alloying element lies between manganese and silicon, and if the carbon content must be held low, the lower carbon content of ferrosilicon as compared with ferromanganese adds another advantage on the side of silicon. Burgess’ has called attention to the possibility of using both silicon and manganese in adjusted com- position, according to the prevailing prices of the ele- ments, but recent American practice seems to have neg- lected the possibilities of silicon. company as well as the large may do its share without facing the prospect of losing the youth it has trained to a larger concern. A conclusion emphasized by the study is that, even though the employer and the labor union are willing to devote time to apprenticeship, the intervention of a third party, usually the State, has been necessary to formulate a definite working plan. Nor can the prospective apprentice be neglected. He must first be won to the idea and then rewon by the employing company, if he is to lay hold of the oppor- tunities which are his and translate them into a future of high achievement. Among interesting statistical statements is the one that carpenters may expect about 70 per cent employ- ment, also that bricklayers average about 180 days per year, and a third one—that 3300 carpenters came to the United States from Europe in 1923. Of 25 well known industrial companies named as maintaining apprentice schools, 23 are in the steel and metal-work- ing trades. Large Ferroalloy Output in South Producer of 50 Per Cent Ferrosilicon Is Heavy Con- sumer of Hydroelectric Power—Plant Operated Continuously on Three 8-Hr. Shifts nooga is the Southern Ferro Alloys Co., manu- facturer of 50 per cent ferrosilicon. Out of 147,- 22,398 kwhr. furnished by the Tennessee Electric Power Co. to Chattanooga industries for power pur- poses last year, this one company used over 20 per cent. Tn largest consumer of electric power in Chatta- Load Factor Averages 96 Per Cent Having an annual capacity of 7500 tons of 50 per cent ferrosilicon, the plant runs continuously in three 8-hr. shifts with a load of 5000 to 5200 kw., or 120,000 to 124,800 kwhr. per 24 hr. The load factor is very high, averaging 96 per cent. The plant, which occupies a tract 175 x 330 ft., is equipped with three open-top electric furnaces, two taking 3-phase current and rated at 2400 kw. and 1800 kw. respectively, and a third on single-phase current rated at 850 kw. The two larger furnaces have three electrodes each, and the 850-kw. furnace has two elec- trodes. The 2400-kw. furnace uses electrodes 24 in. in diameter and 8 ft. in length, weighing over 2000 |b. apiece. The two other furnaces are equipped with 20-in. electrodes. The electrodes are manufactured from a carbon mixture by the National Carbon Co., Niagara Falls, N. Y. Suspended above the furnace, the electrodes extend into the raw materials with which the furnace is filled. As their lower ends are slowly consumed, the electrodes are fed into the furnace. About 10 in. of each elec- trode is used up daily. The top of each electrode is threaded and fitted with a plug over which a new electrode is serewed when the electrode in use has been largely consumed. This plan saves the butt ends of the electrodes. The electrode holders are heavy copper pipes, which convey both current to the elec- trodes and water for cooling the electrode clamps. Silica Is Reduced at 6000 Deg. The load on the furnaces is automatically con- trolled by Thury regulators. The current is received at 6600 volts from the public utility and is stepped down to 85 volts. For the single-phase furnace there is an 833-kva. water cooler transformer, furnished by the Pittsburg Transformer Co.; for the 2400-kw. fur- nace there are three 833-kva. transformers supplied by the same company. The Westinghouse Electric & Mfg. Co. furnished three 650-kva. air-cooled transfor- mers for the 1800-kw. furnace. The raw materials used are high grade silica rock (averaging about 98% per cent silica), coke breeze, for reducing the silica to silicon, and heavy steel turnings, bought largely from the railroads, for the iron which combines with the silicon. The reduction of the silica takes place at a temperature of about 6000 deg. Farh. The furnace shell is of steel, with a first lining of fire brick and a second of carbon. Shipped Largely in Balk Tapping is at intervals of 1 hr. or 2 hr., according to the size of the furnace. The yield per tap ranges from 1300 lb. of 50 per cent ferrosilicon from the 2400-kw. furnace to 650 Ib. from the 850-kw. furnace. The tap runs out on silica sand in pancake-like form. When it is cold it is lifted by crane hooks and trans- ferred to an industrial car. It is then rolled along- side a railroad spur, which runs into the plant, where sand is removed from the ferrosilicon with a wire brush. The ferrosilicon, when stone cold, is very brittle and is easily broken with a hammer into brick- The Open-Top Furnaces Are Charged Con- tinuously With Silica, Coke Breezeand Heavy Steel Turnings. The ferrosilicon ia tapped at in- tervale of 1 hr. to 2 hr., ae- cording to the size of the fur- nace ashes 2 pak a Ee ER GNI MK. AT bens + reheat Nb giant a a Bets & os es 484 THE a Silica Rock and Coke Are Discharged From Railroad Cars Through a Track Hoppe r toa Belt Conveyor, Which Transfers Them to an Ele- ator. The elevator raises them to a two-wau pout, which is set to discharge the silica and coke into separate storage bins imps. Most of the ferrosilicon is shipped in bulk