The Iron Age 1933-09-28: Vol 132 Iss 13

THE IRON AGE September 28, 1933 J. H. VAN DEVENTER G. L. LACHER Ww. W. MACON T. H. GERKEN R. E. MILLER Editor Managing Editor Consulting Editor News Editor Machinery Editor F. J. Winters G. EMRNSTROM, JR. BURNHAM FINNEY GERARD FRAZAR Pittsburgh Detroit Boston Cleveland Chicago Editor Emeritus Washington Cincinnati CONTENTS Uno Multos Structure Alloy Steel Welding Facilitates Bridge Work Heat Treatment Cast Bright Annealing Controlled Atmospheres Life Turning Tools Influenced Shape Insulating the Open-Hearth Furnace Exhibitors National Metal Congress New Equipment News Automotive Industry Washington News NUMBER Construction and Equipment Buying Products Advertised Index Advertisers THE IRON AGE PUBLISHING COMPANY I’. J. FRANK, President G. H. GRIFFITHS, Secretary Cc. 8S. BAUR, General Advertising Manager PUBLICATION OFFICE: Corner Chestnut and 56th Sts., Philadelphia, Pa. EXECUTIVE OFFICES: 239 West 39th St., New York, Y., Member, Audit Bureau of Circulations ADVERTISING STAFF Member, Associated Business Papers Emerson Findley, 311 Union Bldg., Cleveland B. L. Herman, 675 Delaware Ave., Buffalo, N. Y. H. K. Hottenstein, 802 Otis Bidg., Chicago Published every “ Thursday. Subscription Price: Peirce Lewis, 7338 Woodward Ave., Detroit United States and Possessions, Mexico, Cuba, $6.00; Charles Lundberg, 45 Kent Rd., Upper Darby, Canada, $8.50, including duty; Foreign, $12.00 Del. Co., year. Single Copy 25 Cents Cc. H. Ober, 239 West 39th St., New York Robinson, 428 Park Pittsburgh W. C. Sweetser, 239 West 39th St., New York Cable Address, ‘‘Ironage, N. Y."’ D. C. Warren, P. 0. Box 81, Hartford, Conn. SEVENTY-NINTH YEAR SERVICE THE METAL WORKING INDUSTRY der Act “be re | MEMBER | 2 PART | 45 7 2 THE IRON 28, 1933 Page use Steel Plates has ushered new era the design housings for heavy machinery. Plates the designer has avail- able homogeneous material the strength which known within very close limits, enabling him Figure PLATES for Heavy-Machinery provide the proper factor safety with minimum weight. Manufacturing costs are lower when housings are made from Steel Plates. The low cost per pound and high tensile strength Plates, to- gether with the facility with which Housings they can fabricated, make pos sible the design machines which are lighter, better able stand rough usage, and more economical Company, General Offices: Bethlehem, Pa. STEEL PLATES q q ~ ¥ @. THE IRON ESTABLISHED 1855 = SEPTEMBER 28, 1933 Vol. 132, No. Uno ROM one many. From Tubal Cain, the world’s first legendary worker metals, the engineers, metallurgists and technicians who will throng National Metal Con- week, What dramatic story fire the imaginations and ambitions the thousands who are following and who fol- low the pioneer footsteps the man who first harnessed the miraculous power fire and who thus opened the door concealed treasure house technical accom- plishment. these followers Tubal Cain, who have carried the torch progress that lighted, this issue The lron Age dedicated. wie 1—Structure normal sample cooled still air from 750 deg. (upper left). Fig. 2—Same but cooled from 950 deg. (upper right). Fig. 3—Same cooled from 1000 deg. (lower left). Fig. 4—Same cooled from 1100 deg. (lower right). etched per cent Nital for sec., 250 dia. J 42 5—Structure sample heated 1107 deg. for min. and cooled 626 sec. (upper left). Fig. 6—Sample heated 858 deg. for min. and cooled 221 sec. (upper right). Fig. 7—Sample heated 1115 deg. for min., cooled 482 sec. (center left). Fig. heated 945 deg. for min., cooled 337 sec. (center right). Fig. 9—Sample heated 1097 deg. for min. and cooled 380 sec. (lower left). Fig. 10—Sample heated 991 deg. for min., cooled 437 sec. (lower right). All samples cooled 850 350 deg. Photomicrographs etched per cent Nital for sec., 250 diameters. 14—The Iron Age, September 28, 1933 tructure that when series steel samples were heated and then normalized, cooled still air about deg. C., previous work' was discovered and then examined microscopically there was interesting microstruc- ture evidence the specimen’s original temperature was excess 980 The following work inquiry into the condition which had resulted this formation. This microstructure will referred Widmanstatten. Justification for the use this term tions photomicrographs samples one these steels, which had been heated 1000 and 1100 deg. C., re- spectively, prior cooling. More usual structures are shown Figs. and samples this steel, which had been normalized 750 and 950 deg. C., respectively. The steel was the following analysis: Carbon 0.33, Manganese 0.54, Nickel 1.27, Chromium 0.70. Since, after the forging experiments had been completed, insufficient this steel was left for further investigation, bar the following analysis, in. diameter, was obtained for the tests described below: Carbon 0.33, Manganese 0.69, Nickel 1.30, Chro- mium 0.73. The samples were all machined from the same bar but varied from diameter diameter %-in. steps. Two holes were drilled each sample, one axially depth equal half the length the sample, and the other slightly off center and parallel the first. These holes were used accommodate thermocouple and handling bar respectively. The heat- ing furnace was made globars and could minutely regulated. The samples were protected means silica tube, and were prevented from nitrogen evacuating the furnace. The sequence operations during the treatment individual samples were follows: The furnace was raised and held predetermined temperature (which will referred for about hr. The heating the sample was done gradually, and accurately 10. Ellis, “Further Experiments the Forgeability Steel,” Transactions, American Society for Steel Treating, 21, 1933, 673. the 3 che hal ang hol hel ute: gro Thi sity OWEN ELLIS Director Metallurgical Research, Ontario Research Foundation, Toronto, Canada loy Steel measured calibrated thermo- solution carbon, etc. body-cent- iron carbide, i.e., the products the couple. After holding the maxi- ered iron and iron carbide, but dissolution saturated solid solution mum temperature for 130 min. which these constituents vary their carbon, etc., face-centered iron. the sample was permitted cool mode distribution. “Conglomerate” This interstitial matter corresponds regulated rate. careful chart might well referred constitution, though not necessity the cooling temperatures was made. “species” structure, which pearl- chemical composition, the con- Most the work was done samples ite, sorbite, troostite, and mixtures glomerate. these are “varieties.” the samples The structures obtairied subsequent 850 and 1100 deg. C., the samples With rare exceptions, the cooling author elaborates his previ- having been heated and held curves covered the range 850-150 deg. ous work, and determines the during minutes, are filed Table After each sample had cooled conditions under which Widman- and arranged graphically Fig. 16. 150 deg. the thermocouple, which statten structure appears steel Table the samples have been ar- noved anc calibration was re- 350 deg. C.). checked. The sample was then cut 0.69 per cent manganese, 1.30 per cent nickel and 0.73 per cent chro- From Table and Fig. will half, with the cut being made right that when falls bel bout hole. The solid half was then care- greatly ennanced 350 deg. C.) requisite for ground, polished, etched for tion the Widmanstatte sec, (usually fresh solution raised beyond 850 deg. decreases somewhat, and, further, that cel parts concentrated nitric acid 100 later date will shown that when falls below about 915 deg. parts ethyl and finally under suitable cooling conditions (1680 deg. F.), this time cooling Pho- steel can exhibit four critical falls about one-third that when critical examination, and these pho- tere solid ution interstitia and held temperatures below about tographs 750 diameters were taken. form the Widman- 915 deg. the time cooling be- statten structure. Mr. Ellis will came important its effect upon the Structure Related Cooling present the complete paper the amount solid solution carbon, in. their objec discovery where iron carbide precipitate the conditions requisite for the produc- the eutectoid temperature). The tion the Widmanstatten structure proportion conglomerate these this steel. The structures observed samples became distinctly less their lel the samples subsequent cooling from time cooling (850-350 deg. C.) in- after having been raised and creased. This effect cooling rate at- have been divided into four main been heated temperatures above groups, which are classified Table about 915 deg. C., simply because the ‘he The author has found difficult maximum times cooling these ex- fit these structures into the classifica- periments only just prevented the for- tion adopted though mation Widmanstatten structures. ce. both classifications. Typical Grain Size Stee examined the conglomerate was gen- samples consisting entirely erally troostitic. conglomerate, the sizes the grains term “conglomerate” has been varied accordance with the tempera- The term “Widmanstatten” refers for the structures classified Table These areas comprise (1) blades and the average number grains ely between struc- body-centered iron (light portions) shown Table are constituted solid and (2) interstitial matter (dark por- ig. 17. Journal, Iron and Steel In- tions), composed solid solution will seen: (1) that marked Stitute, Vol. 114, 1929, 391. carbon, etc., body-centered iron and grain size resulted from The Iron Age, September 28, | : — 4 why Ate: ale h, ag the r y Table Classification Structures Observed the Steel Subsequent Various Rates Cooling Group Grains conglomerate enclosed ferritic network............. Figs. Group Grains conglomerate, together with grains conglomerate plus Widmanstatten, both enclosed ferritic network...... Figs. Group conglomerate and Widmanstatten approximately Ur ed = i ¢ pr ¥ sample heated 1111 deg. for min., cooled 290 sec. (upper left). Fig. 12—Sample heated 853 deg. for min., cooled sec. (upper right). Fig. 13—Sample heated 1002 deg. for min., cooled 197 sec. (center left). Fig. 14—Sample heated 891 deg. for min., cooled sec. (center right). Fig. 15—Sample heated 1117 deg. for min., cooled sec. (bottom). All samples cooled 850 350 deg. Photomicrographs etched per cent Nital for sec., 250 diameters. Table Time gmax. Time Qmar ofCooling (15 Min. Cooling (15 Min. Deg.C.) ing) Structure Deg. C.) ing) Structure 734 1014 234 936 684 1036 221 858 671 1083 210 978 626 1107 209 1069 630 1006 207 995 614 1108 205 893 585 993 198 1002 940 194 891 492 1037 189 900 482 1115 185 865 478 919 182 1017 471 893 175 1049 449 847 170 874 449 918 + 169 854 + 437 991 165 1043 413 1099 150 893 409 1041 149 900 380 1097 139 930 337 945 123 936 323 993 109 892 323 947 + 98 903 4 320 1001 892 299 1051 875 298 932 874 290 1111 853 263 928 1048 251 938 1124 251 1037 891 247 978 1121 240 874 16—The Iron Age, September 28, 1933 4 +. © heating samples temperatures above 1000 deg. C., and (2) that grain growth proceeded most rapidly within the temperature range 900-1000 deg. Samples heated (15 min.) tem- peratures within the range 900-950 deg. contained both large and small grains conglomerate. structure typical such material shown Fig. 18, reproduction photo- micrograph sample that had been heated and held 940 deg. during min. and had cooled from 850 350 deg. 532 sec. may noted that, whatever the type structure within the grains, their size appeared the same for any given value When rates cooling resulting Widmanstatten structure (Group Group were employed, this opinion found support, not upon observation grain size, but rather upon observations the size the blades and plates the structure. Time Affects Structure Experiments were made deter- mine the effect time heating the structure the steel. Table Sample Grains per Number (Deg. C.) Millimeter N.S. 115 1108 110 212 1083 113 211 1936 115 209 1006 119 207 993 239 168 940 530 198 919 964 170 893 1204 183 874 1331 178 854 2152 these the time heating and hold- ing was extended from 130 min. one series Was approximately 1100 deg. (2010 deg. F.). The struc- tures produced result cooling different rates from this tempera- ture after 130 min. heating were prac- tically identical with those shown Figs. 11, and 15—photo- micrographs samples which were heated about 1100 deg. for min. before cooling about the same rates the samples this series. Fig. shows the structure sample N.S.126, which was heated and held 1100 deg. during 130 min. and then cooled from 850 350 deg. 351 sec. The likeness be- tween this structure and that shown deg. C., time heating min., time cooling 380 sec.) manifest. The structural similarity between other samples heated about the same tem- perature for different times (15 and 130 min.), though cooled about the same rates, was many cases even more striking. What was true sam- ples heated 1100 deg. for the two different times and cooled different rates was true the samples heated about 1050 and 1100 deg. ~ on °C. | tur sal he: ? an hi; in sl ain 7100 hin 2000 em- 1050 1500 ure 7000 een ‘ing 950 to 500 900 950 7000 1050 ere Fig. (at left)—Effect time cooling structure ort, samples heated different temperatures. but 100 200 400 500 600 700 Fig. (above)—Relationship between maximum temperature re. was only when was below use the word “below” here needs 975 deg. and upwards proceeds 1000 deg. that the time heating since the Acs point for rapidly that equilibrium reached ter- became important its effect struc- this steel was the neighborhood within few minutes the introduc- ture. Grain growth then became evi- 790 deg. C.—only deg. below tion the sample into the heating dent after prolonged heating. For when was 850 deg. Grain zone the furnace; hence, little example, sample N.S.170 was heated growth quickens rises from change structure results from pro- and held 893 deg. during 850 950 deg. temperatures (Concluded Page 82) min., then cooled (850-350 deg. C.) 205 sec. The structure the sample was group Table (Fig. 6). the other hand, sample N.S.174, which was heated 900 deg. during 130 min. and cooled (850-350 deg. C.) 189 sec., had the structure shown Fig. 21. few well-marked centers throughout the section were large grains Widmanstatten structure. The smaller grains were about the same size those sample N.S.170. Fig. shows similar structure old- sample (N.S.188) which had been heated 875 deg. during 130 min. and subsequently cooled from 850 tely 350 deg. 192 sec. The structure within all the larger grains this ling sample was like that Fig. 22. was era- Group 2b. rac- Samples heated for the longer period about 850 deg. did not show the same grain growth did the samples heated for the same time higher temperatures. Fig. shows ame the largest grain (only seven others tire cross section sample in. diameter (N.S.180), which had been 130 heated and held 857 deg. dur- 350 ing 130 min. and had been cooled (850- 350 deg. C.) 225 This structure own may compared with Fig. that 1097 sample (N.S.179) heated and time held 858 deg. during min. and The cooled deg. C.) 221 sec. ther The cooling curves the two samples tem- agreed closely except the eutectoid and temperature; the carbon change be- the ginning 640 deg. the latter even sample, and ending 585 deg. the former. The carbon change be- gan the former about 600 deg. will seen that grain growth slow 850 deg. and below. The 2500 7 ‘oo a 18—Structure sample heated 940 deg. for min. and cooled 532 sec. (upper left). Fig. heated 1100 deg. for 130 min. and cooled 351 sec. (upper right). Fig. heated 900 deg. for 130 min. and cooled 189 sec. (center left). Fig. heated 876 deg. for 130 min. and cooled 192 sec. (center right). Fig. 22—Sample heated 876 deg. for 130 min. and cooled 192 sec. (lower left). Fig. 23—Sample heated 857 deg. for 130 min. and cooled 225 sec. (lower right). All samples cooled 850 350 photomicrographs etched per cent Nital for 250 diameters, except Fig. 21, which 100 diameters. The Iron Age, September 28, 1933—17 | | | Ae Beam £ YS : ~ Welding Progress ELD metal, equal physical properties the structural steel used for bridges, can now deposited the metallic arc. Such progress notable, for indi- definite transition the de- sign and technique welding bridges for railroads and highways. The economic situation our country such that many the bridges for highways and railroads strengthened rather There known method for the strengthening bridges equal welding. Reconstruction highway bridges welding has been practiced quite extent even without the advantages the newer electrodes. Now that the newer electrodes pro- vide weld metal with properties equal structural steel, there nothing strengthening railroad bridges ex- cept perfecting the nique the handling the elec- trodes developing the details design. Standard specifications for struc- tural steel for bridges (A.S.T.M. Designation: 7-29, paragraph call for certain physical properties and tests. Tests have been made which show that the newer electrodes deposit weld metal which generally (below)—All-weld-metal tension and specific gravity test speci- mens. Specimens from left right were welded flat, vertical and overhead positions, using G.E. type W-20 electrode. (at Boiler Construction Code reduced section tension-test specimens taken from 0.20 carbon steel pipe 5/16, and in. wall thickness. Specimens from left right were welded flat, vertical, and overhead positions using the type W-20 electrode. 18—The Iron Age, September 28, 1933 ROGRESS the development heavily-coated electrodes has made possible the effective use welding for strengthening, recon- ditioning and fabrication high- way and railroad bridges, many which because present economic conditions will have strength- ened rather than replaced. Results tension, bend, impact, shear and other tests that show the suitabil- ity welding for this work are cited Mr. Vogel meets the requirements for tensile strength (55,000-65,000 Ib. per sq. in.); yield point (0.5 tensile strength, not less than 30,000 per sq. in); and elongation in. (22 per cent). Tests have also been practically meet the requirements for bend tests. They have not been made for elongation in. that would involve the deposit large amount weld metal which would not duplicated any practical design. addition these tests, Charpy impact tests, and also Charpy tension im- pact tests, have been made which prove that the weld metal equal superior impact resistance rolled steel. examination the “Rules for Construction Unfired Pressure Vessels, Section A.S.M.E. Boiler Construction Code,” will show that Class welds should meet all the requirements usually specified for the structural steel for buildings and bridges. examination this specification reveals many interesting facts, and particularly the fact that weld metal has now become de- pendable that may used for pressure vessels practically all de- Boilers” the same Boiler Con- struction Code calls for the same welding requirements Class Section VIII, except all seams are re- quired X-rayed. The develop- ments the welding power boilers and other pressure vessels have indi- cated possibilities the use weld- ing for bridges and buildings far be- yond present practice. The welding bridges Ger- many, Czechoslovakia, Poland and Australia has been developed far greater extent than the United States and Canada. Possibly this due the fact that foreign coun- tries heavily-coated electrodes have been popular for some years. The development trodes this country cal tou sol fe’ pa ri Ww caused principally the require- ments the boiler construction code mentioned above. Existing Structures Reconditioned The economic situation touched briefly before. rea- sonable expect that for the next few years railroads and highway de- partments must use structures long possible. will absolutely necessary recondition strengthen many bridges. has been the experience many engineers that bridges are usually weak only few minor points and that the strengthening drilling for and riveting additional parts ex- tremely costly. The newer method burning off otherwise removing paint and rust, and then welding additional parts. Fortunately, gas-cutting equipment now available which will cut out any segments holes necessary, that parts can fitted over the rivet heads. this way, possible ship bars and structural shapes direct from the mill the site and fit them the structures required, with- out shop fabrication. drawings which give the details and dimensions the various parts are available, such gas cutting preparation additional members can done, course, the shop accordance with the usual routine. If, however, alter- ations are required drawings are missing, then possible the Bridge Reconstruction and Repair ANDREW VOGEL General Electric Co., Schenectady, work the site. this event, gas- cutting equipment becomes indis- pensable tool. The tests described below follow closely the specifications the Ameri- A.S.M.E. Boiler Con- struction Code free bend test from 0.20 carbon steel pipe 5/16, and in. wall thicknesses. Specimens from left right were welded flat, vertical, and over- head positions using the type W-20 electrode. weld-metal Charpy im- pact test specimens welded flat position using W-20 elec- trode. can Society for Testing Materials. The tension specimens parallel de- sign the specimens described the A.S.T.M., and the tension test values are uniformly more than 65,000 per sq. in. Yield point usually not taken for such tests, yield point more measure the steel bars TABLE TESTS W-20 ELECTRODES Position Specific Reduction Welding Gravity Area Flat 7.795 38.5 Vertical 7.799 35.8 Vertical 48.5 Overhead 7.795 42. Average 7.7965 41.275 Yield Tensile Elongation Point Strength in.. sq. in.) sq. in.) 21.5 54,600 20.5 50,300 68,500 21.5 53,000 70,000 24.0 49,500 65,600 21.875 51,850 69,900 The Age, September 28, 1933—19 welded together than the weld it- self. Where yield point taken, usually over 50,000 per sq. in. Elongation in. not taken, elongation over such long dimension would merely register the elongation the steel bars welded together and would not show the elongation the weld metal itself. Elongation in. taken typical all-weld- metal tension specimen made ac- cordance with the specifications Section VIII the Rules for Con- struction Unfired Pressure Vessels the A.S.M.E. Boiler Construction Code. Class code requires mini- mum elongation per cent in. The tabulation following [Table gives the results group tests all-weld-metal tension specimens made accordance with the A.S.M.E. Boiler Construction Code. The column headed “Position Welding” indi- cates whether the weld metal was de- posited the V-shaped grooves be- tween the plates, with the groove flat (horizontal), vertical over- head position. The A.S.T.M. code calls for mini- mum elongation per cent in., while the A.S.M.E. code calls for minimum elongation per cent in. These specimens met the requirements the A.S.M.E. code and practically met those the A.S.T.M. code with respect elonga- tion in. the other items, the all-weld-metal tests, using W-20 elec- trodes, more than fulfilled the require- ments either code. rt) or Fes | E all ng at le- ait or le- e- | \ in ; =), TABLE STANDARD REDUCED SECTION TENSION TESTS (Values per sq. in.) POSITION WELDING SPECIMEN Fiat Vertical Overhead PLATE CGP 71,500 63,600 57,800 CKP 51,400 64,950 48,300 BIP 67,000 46,550 52,900 BDP 50,400 57,100 50,900 144-IN. PLATE MKP 65,300 63,500 63,200 CGP 65,200 65,000 65,300 CFP 63,350 55,900 62,400 CHP 64,400 71,700 64,450 1-IN. PLATE BDP 53,700 54,300 55,400 XTP 62,750 62,550* 59,050 ABP 63,550 63,750 61,900 BIP 64,600 63,700 64,100* 63,400 57,300* 63,400 All specimens except those marked the asterisk broke in the plate, well out- side the weld. (Required tensile strength plate 45,000 per sq. in.) *broke weld and plate. The A.S.M.E. Boiler Construction Code requires tension tests welded joints. The specimen called re- specimen. The 5/16-in. specimens were taken from 8-in. pipe located with its longitudinal axis horizontal. The in. specimens were taken from 14- in. outside diameter pipe with its longitudinal axis horizontal. The in. specimens were taken from plates rolled 14-in. outside diam- eter simulate pipe joints normal position with longitudinal axis hori- zontal. Strips were cut from the pipe joints. Those designated “flat” were taken from the top the pipe, those designated “vertical” from the sides—that is, the horizontal diameter—and those marked “over- head” from the bottom the pipe. Specimens taken are more truly indicative the quality welds which can produced the welder using W-20 electrodes the awk- ward positions involved welding around pipe. The tabulation [Table shows that all welds broke the plate stock well outside the welds, with the exception three specimens. These specimens failed the weld 57,300, 62,550, and 64,100 per sq. in. re- spectively. They therefore met the requirements the A.S.T.M. code and the A.S.M.E. Boiler Construction Code. will noted that some the test values are lower than 55,000 per sq. in. This was due fail- ure the plate the lap-welded steel pipe, which only required have tensile strength 45,000 per sq. in. Meet Severe Bend Tests Bend tests are shown for steel plates 5/16, and 1-in. thick welded together with the new heavily-coated electrodes. The plates are shown 20—The Iron Age, September 28, 1933 bent practically flat themselves, with the weld metal sustaining maxi- mum elongation the outer surface the bend. These bend tests follow paragraph the A.S.T.M. Specifi- cations for Structural Steel for Bridges nearly practicable, but were made comply with the A.S.M.E. Boiler Construction Code. Table III, Stand- ard Free Bend Tests,” was prepared from tests made specimens taken from pipe, mentioned. The 5/16-in. specimens were taken from 8-in. pipe, the specimens were taken from 14-in. outside diameter pipe, and the specimens were taken from plate rolled 14-in. outside diameter. The flat, vertical and overhead desig- nations indicate specimens taken from top, side and bottom pipe with its longitudinal axis horizontal. The A.S.M.E. Boiler Construction Code requires that Class specimen shall bent cold under free bending conditions until the least elongation measured within across approxi- mately the entire weld the outside fibres the bend test specimens per cent. This severe require- ment, the specimen tends bend outside the weld, thus making ex- ceedingly difficult bend the weld it- self. The tabulation shows that even this drastic test can met many specimens. Charpy Tension Impact Tests The next series tests was made the standard Charpy impact test specimens, and show the ability the weld metal resist impact. The aver- age values Charpy impact tests vary between and ft.-lb. for struc- tural steel. Tests weld metal de- TABLE STANDARD FREE BEND TESTS (Values per cent elongation outer fibre) POSITION WELDING SPECI- MEN Flat Vertical head age 5/16-IN. PLATE CGP 26.7 28.8 29.9 CKP 27.0 44.0 BIP 36.0 30.0 38.0 BDP 25.0 27.0 31.0 Average 29.2 28.2 35.7 31.03 PLATE MKP 41.0 31.3 32.0 CGP 35.5 35.5 34.2 CFP 36.0 36.0 38.3 CHP 30.0 26.0 35.5 Average 35.6 32.2 35.0 34.27 1-IN. PLATE BDX 35.0 30.9 30.7 29.5 30.0 ABP 25.8 29.6 24.4 BIP 23.9 21.5 20.6 29.1 26.5 28.9 Average 28.1 28.4 27.0 27.83 General Average 31.04 posited the W-20 electrode varied from 24.4 27.9 ft.-lb. This is, therefore, immediate and direct comparison between weld metal and rolled steel with respect im- pact, and the values obtained from the weld metal compare favorably with rolled steel. addition the standard Charpy impact test specimens, Charpy tension impact tests [Table IV] were made specimens using three types elec- trodes. The first type standard sul-coated rod extensively used steel frames for buildings and identified G-E type the other two, known G-E types W-20 and W-21, are heavily coated. The tests, wherever failure occurred the plate stock, illustrate the fact that the weld metal generally superior impact the plates that are welded together. These test results were made avail- able through the courtesy Col. Jenks, Watertown Arsenal, Water- town, Mass., and the photographs and data are well worth consideration demonstrating new form com- parative impact test. The superiority this type impact test over the standard notch bar test lies the fact that the entire welded joint tested, rather than small specimen from the weld metal itself. Nick Break Test Shows Weld Quality Another form test that known the nick break test. This can simply and inexpensively performed. shows the quality weld metal laid down the welder simple and direct manner possible. any defects occur, due errors the technique the welder, this particu- lar specimen will expose the defects immediately and give the welder opportunity improve his methods. This test can made the site. only involves welding together two plates, shown the illustration, cutting the specimen into strips with gas torch, making nick with the gas torch, supporting the specimen, and then striking the specimen heavy blow. The fracture easily inter- preted terms satisfactory un- satisfactory weld metal porosity with respect defects resulting from failure remove slag other foreign material. Specimens can also made ac- cordance with Fig. U-13 the A.S.M.E. Boiler Construction Code, Section VIII; but, while the external appearance more satisfactory than | \ 1 e 4 ‘ | < q | aA test specimens using low- carbon steel plates. They were welded flat posi- tion using G.E. Types W-21, and W-20 elec- ted trodes. tal im- the IG. (at left) Boiler Steel Welding Committee the steel pipe 5/16 report was issued September, 1931, thicknesses. Specimens from left the American Welding her right 5/16 and in. groups New and were welded flat, vertical and will observed that the tabula- sts, overhead positions; and specimens tion indicates three groups speci- late welded mens. The first three specimens were veld welded such manner produce 3/8-in. fillet welds two passes. The her. first pass consisted 5/32-in. W-20 electrodes, current 170 amp. being used. This pass was applied rapidly, ter- already made indicate that the securing deep penetration the root and shear value, when heavily- coat- the weld, but building only electrodes are used, about small fillet. The second pass was made om- the same as, slightly higher 3/16-in. W-20 electrodes, 220 amp. rity than, for bare lightly-coated being used. This pass completed the the electrodes. weld 3/8-in. fillet, and also re- fact when heavily-coated electrodes fined the first pass. root the weld. were with single pass using specimens prepared the gas torch, tudinal Shear Fillet Weld Tests” Those the third group were the results are the same. [Table was based group welded with two passes, each case tests made accordance with the 3/16-in. W-20 electrodes 220 amp. Shear Tests Welds Made With specimen designated Fig. 54, page being used. This third group speci- laid Heavily-Coated Electrodes 109, the report the Structural mens was rotated that the surface and Shear tests are, course, very im- tests this type weld the metal. Practically all structures are designed that the weld metal TABLE TENSION IMPACT TESTS fects shear. Many hundreds tests have (Tested Watertown Arsenal) been made determine the shear Specimen Electrode Ft.-Lb. Failure hods. fillet welds, using ordinary WG-2 Weld, fusion zone bare lightly-coated electrodes. One 393.5 Center weld hundred seventy-three tests were re- 459.6 Center weld men, using heavily-coated electrodes. Tests 934.0 Plate nter- *Reinforcement left on. un- The above specimens were butt welds plate. The width the specimen its narrowest section was in. other WELDING CONDITIONS NO. Specimen Plate Bevel Space Electrode Amps. Are Volts Passes the in. Std. 1/4 in. W21 260 35-38 ernal than The Iron Age, September 28, ae? Sq 4 Pp Ke, | 2 ¥ Ante 2 the pool molten metal was hor- izontal. The first two groups specimens were welded with the specimen flat position, thus simulating normal position most frequently encountered structures and bridges. The test values show that they failed almost uniformly 12,000 lb. per linear inch more. The one exception failed 11,950 per linear inch. The uniformity the results im- pressive. brings out the fact that heavily-coated electrodes can produce uniform strength bare lightly- coated electrodes. The test results the latter were uniform particularly empha- sized page 340 the article “Tests Metal Arc Welds” the June, 1929, General Electric Review. Table should studied con- junction with Figs. and From these data, will observed that two groups tests were made with two- pass welds and one group with single-pass weld. Fig. shows speci- mens W-12-1, -2, and -3, made with two-pass welds, after being tested, and the fractures indicate the high quality the weld metal. While specimens W-12-4, -5, and -6, made with single-pass welds, showed shear test values equal the other specimens, tests made bending bars perpendicular the weld, trated Fig. indicated that greater ductility obtained two-pass welds, virtue the refinement the metal the first pass. The specimens illustrated Fig. were welded one side only. The specimen the left was made with two-pass weld and the strips were bent downward hammer blows shown. The specimen the right was made with single-pass weld and the strips were bent down only part way when the weld fractured. These bend tests are confirmed and explained photo- micrographs, which show the refine- ment the first pass, resulting from the application the second pass. appears, therefore, that these tests should lead the conclusion that two- pass welding preferable single- pass welding. The steel work railroad bridge, and highway bridge crossing Iron Age, September 28, 1933 railroad, subjected severe corro- sion due the smoke and fumes from locomotives. Various types acceler- ated corrosion tests, such spray, weatherometer, and immersion equal parts hydrochloric acid and water, show that welds made with bare corrode faster than welds made with heavily-coated electrodes, such the type W-20. Immersion acid solutions and salt spray tests are well known, but the weatherometer test not generally known. The weatherometer device which the specimen subjected ultra-violet rays continuously and spray water alternately and off for 15-min. periods. All these tests indicate that, while weld deposits made with bare electrodes corrode slightly more rapid rate than struc- tural steel, welds made with heavily- coated electrodes corrode either the same slower rate than structural steel. Based accelerated corrosion tests appears that welds bridges, subjected severe corrosion produc- ing conditions, should made with heavily-coated electrodes. Inspection Methods Improved With tests various types indicat- ing that the weld metal produced heavily-coated electrodes superior all respects that other types electrodes, and also equal that structural steel, are now the shear fillet weld test specimens made according Fig. 54, report A.B.W. Structural Steel Welding Committee. Welding was flat position, using G.E. W-20 electrode. point where possible use these electrodes for the strengthening, re- conditioning and fabrication rail- road and highway bridges. Inspection methods have improved with the introduction coated electrodes. Visual inspection has been found satisfactory guide all forms fillet welds, and even very deep butt welds. Devel- opment the technique welding, called, has resulted making every weld good weld. Progress tech- nique has kept pace with progress quality electrodes. The result that welding engineers may now approach the welding bridges with the high confidence developed experience pressure vessels and power boilers. New Era Design and Fabrication Undoubtedly are now the point where new era type develop- ment will occur the design and con- struction bridges. This seems have commenced European and other countries far greater extent than the United States, probably because European engineers have al- ways been more economical the use metal. Material always been cheap, and labor expensive, Amer- ica, while exactly the reverse has been true Eurpoe. Conditions the rela- tive cost metal labor are now changing the United States. Eco- TABLE V—LONGITUDINAL SHEAR WELD TESTS See Figure 54, Page 109, “Report Structural Steel Welding Committee the American Bureau Welding,” issued Sept., 1931 Inches Welding Weld Specimen Procedure Failure passes 6.500 W-12-1 5/32 in. W-20- 0-A 6.500 2nd 3/16 in. W-20- 220-A 6.562 W-12-4 pass 250 3/16 in. W-20 6.250 220-A 6.125 W-12-7 passes 6.312 3/16 in. W-20 220-A 6.062 deg. angle 6.312 Total Failure Failure inch 79,400 12,200 12,400 80,250 12,200 75,800 12,100 78,650 12,600 75,350 12,300 75,650 11,950 13,350 13,150 — sit pr lay dis il- the 8—Specimens showing difference left shows bend tests two-pass weld; the other fracture from bend test single pass weld. nomic factors have also made neces- sary extend the life bridges and structures much possible because the financial condition the rail- roads and all governmental divi- sions. These tests should considered Relation Temper Colors following may throw some light the debatable question the relation tempering colors temperature. That is, whether the temper colors represent the actual condition (not the temperature) the steel itself. Some tool makers maintain that the efficiency the tool, both hardness and other prop- erties, the same whether the color has been obtained short heating heating lower temperature, and that the ultimate results are indicated the temper colors, independent the method obtaining them. Others aver that this not the case. dis- agree with the former and contend that temperature, regardless color, the only true indication good tempering and believe the following will prove that temper colors are best but surface indication. heated strongly one end, the other end remaining cold, there Produced over the heated portion layer oxide scale, the thickness which diminishes gradually with the distance from the hot end. Some- where the center the strip the thickness the oxidized layer will ductility between in. fillet welds made with one and two passes. One with the thought that they start new era design and fabrication, and that, therefore, structural and bridge engineers should investigate the pos- sibilities strengthening recon- ditioning structures and bridges HOLDEN New York Chapter, Society for Steel Treating Secretary, American become comparable the wave length light and over this section the films give rise series inter- ference tints commonly known “temper colors.” The colors are the result the fact that two trains light waves are reflected respectively from the outer and inner surfaces the oxide film. there interference be- tween the crests some the waves with the troughs some the others, there produced variance colors, ranging from light yellow the thinnest coating the oxide blue the thickest portion. Be- low certain thickness oxide, colors are produced since the oxide film thin produce wave lengths short not affect the eye. easy comprehend, from the foregoing, the necessity for having polished bright surface order obtain the colors, essential that there some reflecting medium; and also that the various tints pro- duced are not the color the oxide, which more less dark gray black, but are caused the steel re- flecting different colors the thick- ness the oxide increases from the the use heavily-coated electrodes. The writer wishes acknowledge the valuable assistance Waugh the General Electric Co., who fur- nished the test data and photographs for this article. Temperatures color the steel itself blue and, when very thick, black. The longer the strip allowed exposed heat, the thicker the oxide will become, the non-visible oxide becoming yellow; the brown, blue; and on, and finally the heated portion might coated with black oxide depending the temperature which the strip exposed. An- other case point that, with correct amount oxygen present, the colors would form quickly and that, with too little oxygen present, the color might not form all take longer time than usual. From the foregoing one can better understand how time and temperature have marked effect the temper colors and one can readily see why that tempering certain tempera- ture for given length time would much better than just draw straw brown blue the case might be. While the colors are guide tempering there are certain conditions that may tend vary the colors regardless the temperature employed such time and atmos- phere. Also temper colors are only surface indications and not criteria the whole body the tool die. The Iron Age, September 28, 1933—23 4 4 ry BS : int ont er- | en j la- ow =) aa RIGINALLY, more extend- use cast iron was only brought about because the low price and ease with which con- formed pattern when cast. Re- cently, the metallurgist turned cast iron material which was worthy his research, and the re- sult was greatly improved metal which conformed the rigid speci- fications design engineers. This manuscript resume and exam- ination heat treatments which result particularly high strength cast irons unusual ductility, shock resistance and fatigue strength. Future applications are indicated. Mr. Morken will pre- sent this paper the Detroit meeting the A.S.S.T. vvyv ECAUSE its humble origin and homely methods production, cast iron has for generations been considered inferior material and was held quite unsuitable where strength, ductility, and shock resistance were necessary. occu- pied obscure position during pe- riod tremendous metallurgical de- velopment other fields. During only the last few years, cast iron has begun receive attention alloy capable refinement and worthy intensive study. Improved from the electric melting furnace awakened the metallurgist new interest cast iron engineer- ing material, and the metallurgist became able produce better irons, the engineer began adopting them materials worthy his use. Use the electric furnace resulted cast iron which possessed more than four times the strength the conventional “old” cast iron, and addition iron greatly improved shock-resistant properties. During the last two years number startling and revolutionary applica- tions heat treated cast iron have been made and others, still more startling, are the process de- velopment. Heat Treatment Gray Iron considering the heat treatments applicable cast iron the first type examined will ordinary gray iron. The gray appearance the fracture caused, course, free graphite, and the matrix may com- posed pearlite, pearlite and ferrite, pearlite, ferrite and cementite, with other minor constituents present. The gray iron foundryman has learned that the ideal iron, similar that illustrated Fig. which graphite individually fine par- Iron Age, September 28, 1933 The Heat Treatment Cast lron CARL MORKEN Foundry Engineer, Detroit Electric Furnace Co. ticles, uniformly distributed through the matrix. This iron contained 2.70 per cent total carbon, 1.70 per cent silicon and had unalloyed tensile strength 52,210 pounds per square inch. His attempts produce such iron are frequently frustrated un- foreseen uncontrollable foundry conditions, and free ferrite results, the detriment hardness and wear resistance, free cementite ob- tained, much the detriment ma- many instances, due the design castings, stresses are set the casting the metal freezes. Since the ultimate value any casting equal the theoretical strength less the internal stresses, (upper left)—Pearlitic gray cast iron produced rocking indirect arc elec- tric furnace. and equally large average grain size. Etched, 100 diameters. able iron made standard long cycle anneal, 140 hr. Etched, 100 diameters. White iron melted indirect arc rocking electric furnace. Fig. (upper right)—Air furnace malle- Note large graphite particles Note relatively large pro- portion pearlite cementite and the manner which the free cementite has been broken and distributed. primary graphite present. eters. Etched, 100 diam- Fig. (lower right)—Same iron Fig. but annealed with 20-hr. cycle, having fine graphite nodules, uniformly distributed and fine grain size. Note absence pearlite. Etched, 100 diameters. 3h | 7 desirable remove those internal stresses created during the freezing. The stress removal accomplished simple heat treatment frequently referred “normalizing” “ag- ing.” The castings are heated point ranging from 800 1100 deg. (430 595 deg. C.), and main- tained maximum temperature until equilibrium reached. ordinary treatment for relief internal stresses consists holding for four hours 900 deg. F., and, when prop- erly conducted, such treatment has apparent effect upon the physical properties the castings. Because the graphitizing tendency silicon, advisable use low tempera- tures for relieving stresses high silicon irons, and, per contra, higher temperatures are permissible for low silicon content. second heat treatment for gray cast iron the anneal, used break down the free cementite that rough castings may more readily ma- chined. The treatment consists mainly heating above the critical tempera- ture and cooling slowly tempera- ture well below the critical, followed dumping the charge into the air for convenience handling. The general treatment consists holding point from 1400 1650 deg. for Ballay, Transactions, American Association, Vol. XL, 1932, . two six hours and cooling the furnace black heat. The brief soak above the critical temperature serves decompose the free cementite with- out seriously affecting the eutectoid cementite. Slow cooling through the critical range facilitates this decom- position. The annealing, outlined, decreases the hardness and causes reduction the physical strength the iron. The reductions hardness and strength are variable, depending upon temperature, time, and the com- position the iron. The ideal treat- ment that which the hardness decreased only enough produce the desired machinability. third treatment hardening, which frequently called “marten- sitic quench,” and may may not followed draw. The ob- ject the quenching produce extreme hardness, increase hard- néss while retaining machinability, increase the strength the iron. JA (upper similar Fig. annealed with 24-hr. cycle. pearlite remains. Etched, 125 diameters. Fig. (upper right)—Same iron Fig. but heat treated produce high strength with total cycle Etched, 125 diameters. Fig. (lower left)—Same iron Figs. and but given heat treatment for high strength. Properties illustrated Fig. Etched, 100 diameters. Fig. (lower retention pearlite. The white iron, similar Fig. was elec- trically melted and the total annealing cycle was about hr. Etched, 100 diameters. The casting quenched from above the critical temperature. One limita- tion that immediately suggests itself the formation quenching cracks, but these are obviated by: (1) selec- tion quenching medium, (2) cor- rect design heating and quenching cycle, (3) redesign the casting, and (4) use alloys increase the hot strength the iron lower its critical temperature, whereupon lower quenching temperatures are used. Since the critical temperature cast iron varies with the carbon and silicon content and with alloys, common practice quench from 1450 1550 deg. F., which well above the critical. The quenching medium determined largely the design the casting and the hard- ness desired, and usually either oil air, although water some- times used. means the quenching treat- ment, unalloyed cast iron with Brinell hardness about 400 readily obtained, while the hardness alloyed metal runs 500 more. These irons are used for with- standing severe wear and abrasion, and are not commercially machinable. has described quenching treatment which good hardness obtained with quenches oil from about 1500 deg. F., and draws from 400 deg. and 600 deg. relieve quenching stresses. Nickel used the iron improve susceptibility heat treatment and lower the critical temperature. The iron machinable and has martensitic structure. Brinell hardness about 350. with steel, cast iron fre- quently machined the “as-cast” condition, following which quenched, drawn and the machining completed slow speed grind- ing. The strength gray cast iron increased quenching and draw- ing, and strength values well over 85,- 000 Ib. per sq. in. are readily obtained with fair degree machinability. White Iron its name indicates, white iron iron such composition pro- duce white fracture. All the carbon combined and graphite exists the iron cast. White iron has been produced for years for abrasion resistant parts and for making malleable iron, but has limited commercial application. Since white iron brittle, some- times found necessary anneal slightly order preclude break- age the castings handling. This anneal consists, usually, holding 1500 1600 deg. for three four hours, which precipitates small amount graphite and breaks down some the massive cementite, there- increasing the toughness the castings. Most verted white iron poured con- into malleable iron some The Iron Age, September 28, 1933—25 2s = > wear ob- due 23 form similar malleable iron. Rapid strides have been made during the last year this field, with attention directed primarily toward shortening the annealing cycle and improving the physical properties the iron. This activity has developed series irons with distinctly new properties. Malleable Iron Malleable iron produced from white cast iron suitable anneal- ing process. has made this type iron the subject most comprehensive treatment. White iron primarily alloy iron, carbon, and silicon, containing small amounts other elements. The carbon en- tirely combined and graphite exists, and the structure consists mainly massive cementite and pearlite. producing malleable iron from this material, the combined car- bon completely broken down, and the carbon precipi