SOME. FACTORS AFFECTING RETAINED AUSTENITE 1N ALLOY STEELS Thesis for the Deg?“ m‘ M; S‘ MECHIGAN STATE COLLEGE William fames Buckley 3:947 T515515 —. ‘ ' e . .\\:‘-. .9 - ‘ ' . n n\%.x¢:“\4. Enffl‘n‘mvyn mm“ Agoh L61} ~1R33~m:k\-s‘b 2‘3"?” ;“M.t;‘ I wag“. JfTIFuii #199, Oh to This 1. to certify that the thesis entitled Some Factors Affecting Retained Austenite in Alloy Steele. presented by Willium J. Buehler has been accepted towards fulfillment of the requirements for Chemical and M.S. degree in Retellurgi cal Engineering. Major professor— a’”. ' SOME FACTORS AFFECTING RETAINED AUSTENITE IN ALLOY STEELS By William James Buehler \. Submitted to the School of Graduate Studies of Michigan State College of Agriculture and Applied Science in partial fulfillment of the requirements for the degree of MASTER OF SCIENCE Department of Chemical and Metallurgical Engineering 1947 THESiS Table 9§_Content§ Introduction ‘ Experimental Procedure Part I Part II ‘Part III Diagrams Carbon analysis apparatus End-quenching apparatus Experimental Results Outline Carbon gradient data Graphs of carbon gradients Photomicrographs for Part I End-quenching data End-quenching graphs Microstructure of end-quenched bars Photomicrographs for Part II Photomicrographs for Part III Qiscussion Conclusions gossible Euture Work Selected References kn; r a J an I a K.“ ’wv‘ QC 7 1r“ . Page 3-11 12—16 16—17 18 19 20 21 22-26 27-51 32-57 58-59 60 61-65 64-69 70-72 75-81 82 85 84-86 Introduction Solid solutions in which gamma iron is the solvent and carbon is the solute are called austenite (1). The solid solution austenite is formed by heating a steel containing between 0.03 and 1.7% carbon to the austen- itizing temperature. The austenitizing temperature is a function of the amount of carbon and alloying elements present in the steel. Gamma iron is quite important because practically all heat treating is done from this solid solution range which occurs above about ISSOOF. Upon rapid quenching it has been discovered that certain quantities of austenite remained untransformed in the martensitic matrix. Retained austenite is particularly undesirable from the standpoint of practically all physical properties due to the fact that it is considerably softer and more plastic than martensite. The subject of the decomposition of austenite in steels, because of its immense practical importance, has been studied very carefully in recent years (2) (5). The decomposition products have been classified in three general groupings - pearlite, bainite, and martensite. 0f the three decomposition products, bainite has received only scant attention. On occasions it was found that not all the austenite was transformed (4) especially in higher carbon steels containing high percentages of manganese, nickel, and chromium. This untransformed or retained austenite has caused some investigation and several methods of quantitative analysis have been tried. Some factors affecting the retention of austenite are as follows: (1) Carbon content (2) Quenching rate (quenching media) (3) Austenitizing Temperature (4) Sub-atmospheric cooling (5) Cold working (6) Tampering temperature and time The fact that some austenite was retained on quenching higher carbon steels was known since the very early days of metallurgy.(5) It was originally felt that more aus- tenite was retained when more drastic cuenching from a higher temperature was employed. This idea was later discovered to be incorrect, when certain alloy steels retained more austenite after 011 quenching than they did after water quenching (7). The suggested explanation for this phenomenon was based primarily on the effect of stresses and stress distribution during quenching. When the problem was considered from a stress stand- point, care had to be taken to distinguish between compressional and tensional stresses within the steel being considered. Since austenite has a greater density -4- than.martensite, compressive stresses will promote the retention of austenite while tensional stresses will promote its decomposition. In connection with this idea, experimental evidence showed that more austenite was retained in the exterior, on water quenching, while upon oil quenching, more austenite was retained in core area. The earlier work caused some disagreement on the lowest possible carbon content at which austenite was retained regardless of the rapidity of quenching rate. “over and Engel (11), by the use of X-ray analysis found that a minimum of 0.60% carbon was necessary in quenched steels for the retention of austenite. Davenport and Bain (12), claimed that they found traces of retained austenite in 0.54% carbon steel that was quenched. Esser and Cornelius (13) found that the maximum amount of retained austenite results at cooling rates Just below that of the critical cooling velocity. With either increasingly higher or increasingly lower cooling rates, the amount of retained austenite was found to steadily decrease. The idea of cooling rates was ex- tended as the possible reason why more retained austenite was found in small sections quenched in oil than in corresponding sections quenched in water. The cooling rate of the oil was more nearly that of the critical cooling velocity and thus retained more austenite. The maximum amount of austenite was retained when the steel was quenched from just above the A03 or Acm line. High-temperature quenching was found to give less retained austenite (5). Several ideas have been put forward on the retained austenite to martensite transformation that occurred during cooling to very low temperatures. Early exper- imenters (14) found that the transformation occurred by a step-wise cooling to sub-atmospheric temperatures, transformation began at the -20°C. step. 0n holding at this temperature the transformation ceased after a period of time and no further transformation occurred until the temperature was again dropped. Complete, or nearly complete transformation occurred if enough increments of temperature drop were applied. From the step-wise action of the transformation it was believed that the real cause of the austenite to martensite change was brought about by the deformations, stresses, and strains which had occurred during cooling. Fletcher and Cohen (6), stated that aging at room temperature between the hardening and sub-cooling treatments lowered the temperature at which the retained austenite started to transform on sub—cooling and reduced the amount of transformation achieved by any given cooling treatment. Virtually complete decomposition of the retained austenite could be accomplished by sub-cooling to-250°F. to-260°F., if the prior time at room temperature was kept within several minutes. Transformation of retained austenite due to relief of stresses and strains was definitely established when the effect of cold working on the retention of austenite was studied. Hardened steels which contained 0.55% carbon were found by Bain (7) to be susceptible to retainaiaustenite after quenching. Those alloy steels in which more than the normal quantities of nickel, manganese, and chromium were present can have a carbon content even lower than 0.55% and still retain unstable austenite. Tamaru and Sekito (15), found by X-ray studies evidence of retained austenite in steel containing as little as 0.40% carbon. detained austenite was transformed by heating to a suitable temperature for a definite length of time. Partial trans- formation has occurred at very low temperatures, with the occasional formation of cracks. A proportion of retained austenite amounting to 10% to 25% has been found in commercial steels. In many steels, the presence of a surprisingly large pro- portion of austenite had scarcely any effect on the hardness of the quenched steel. To obtain a reduction of 10 points Rockwell required the retention of fairly large amounts of austenite. The explanation for this was based on the uniform distribution and orientation of the austenite in the martensitic structure. Bain (7) states that some high chromium steels quenched from an austenitizing temperature have been found to contain as high as 80% to 90% retained austenite. The austenite in plain carbon steels has been substantially unchanged during heating for brief periods at 450°F. The complete transformation of austenite has been verified by X-ray diffraction and dimension change with tempering. Careful studies (7) made on the isothermal trans- formation of retained austenite reveal that the product of transformation at the tempering temperature was not hanimartensite, but rather one of the slightly softer structures of the bainite group. The hardness of this bainite was Just a little greater than that of the tempered martensite. The behavior of some steels to get slightly harder at room temperature as a large quantity of time has passed indicates that possibly the retained austenite, which is unstable after quenching, has slowly transformed into martensite. Bain (7), has referred to this phen- omenon as a lingering austenite transformation. Tampering at low temperatures, such as 200°F., has greatly speeded the transformation or "aging period" of freshly quenched high carbon steel. French (8) stated that in a steel of around eutectoid carbon, it was found that the water quenched steel with less than 5% austenite had poor fatigue resistance, while the same steel quenched in oil contained 5% retained austenite and showed an appreciable improvement in the fatigue limit. The effect of the 5% austenite was thought to exert a cushioning action rather uniformly throughout the cross-section of the fatigue sample and thus cause increased fatigue resistance. When wear resistance is desired, the retention of large quantities of austenite is found to be a disad- vantage, since it is much softer and more plastic than martensite. French(8) also stated that if sufficient quantities of manganese and nickel were added to a steel, it could remain completely austentic. Hardened steel, containing tetragonal martensite and retained austenite, passes through three structural changes on tempering (9). During the first stage (200°F. to 350°F.) the tetragonal martensite undergoes a decom- position which causes a precipitation in the higher carbon concentration regions and this transition pre- cipitation accounts for the darkening of the martensite plates. During the second stage of tempering (450°F. to 550°F.) it is quite certain that the retained austenite is transformed. The transformation product has the general appearance of an acicular bainite. The main reason for believing that the transformation product is bainite is that the second stage tempering temperatures are well above the martensitic range of transformation. The third stage of tempering (550°F. to 7500?.) is characterised by the decomposition of the transition precipitate formed during the first two stages. The decomposition forms cementite particles which gradually coalesce into a spheroidized structure with a ferrite matrix. Liedholm (10), supports Bain's statement on the transformation of retained austenite in plain carbon steels at a range of temperatures in the vicinity of 455°F. The transformation reaction causes an increase in magnetization of carbon steel by about 3%. It was found also that the relationship between the amount of austenite present and the magnetic properties defied the attempt of mathematical formulation. The investigation 0n cobalt high speed steels indicated that considerable austenite was retained after tempering at temperatures -10... of 900°F. or less. The retained austenite did, however, decompose rapidly upon tempering at 1000°F. and higher temperatures. The first changes in the austenite occurred at temperatures between 700°F. and 800°F. which indicated that the austenite transformation occurred over a range of temperatures rather than a sharp change at a definite temperature. No evidence was found in the literature to support the transformation of retained austenite in plain carbon steels over a range of temper- atures. In view of the fact that the transformation of retained austenite in alloy steels occurred over a wide range it may be assumed that a similar system of trans- formation occurred in the plain carbon steels. -11- Egpggimental Procedugg Part I Relation of Carbon Content to the Amount of Retained Austenite Five steels were chosen for the experimental work, one plain carbon steel SAE 1010 and four alloy steels SAE 2015, SAE 2540, SAE 5145, and SAE 4640. The analysis of the five steels were given in table 1. 221.319.; 9. n.9, .r; a a: a. a9. SAE 1010 .15 .55 .016 .045 SAE 2015 .55 .54 .017 .024 .55 .75 SAE 2540 .297 .71 .011 .017 .22 5.42 SAE 5145 .40 .72 .016 .020 .69 .159 SAE 4640 .40 .65 1.82 .25 (Aver.Spec) The bars were first cut to a convenient length (6 inches)-and then placed in a lathe chuck to be faced down on both ends and center-drilled. Upon center-drilling each end of the five bars, they.then were turned on centers to the largest possible diameter giving a taper of no more than .001 of an inch from one end to the other. After all the preliminary machining was completed, the bars were then ready to be case carburized. The carburizing was done in a 2.0" steel pipe carburizing bomb, using a commercial solid carburizing mixture of the following analysis: -12.. Table _2_ 3:003 10-12% Na2C0§ 2-s% Ca003 2-5% Coke 25-50% Charcoal (Type F.S.R) Balance Two of the smaller diameter bars were placed in a bomb together, while the larger bars were carburized individually. When two bars were placed in the same bomb together care was exercised to keep the bars equidistant from the walls of the bomb and from one another. The commercial carburiser was packed very tightly about the samples. The sealed bombs were placed in a muffle type furnace controlled at 1700°F., and left there for a period of 15 hours. At the end of the 15 hours, the bombs were removed and allowed to cool in still air. The carburised bars were then placed on lathe centers again and checked first for possible warping by the use of a dial indicator. If it was found that the bar had warped appreciably it was aligned by a few well placed hammer blows, while on the lathe centers. Maximum.warping occurred, as would be assumed, in the smaller diameter bars. The large bars (SAE 5145 and SAE 4640) exhibited practically no deformation during carburizing. After the proper alignment had been secured -13- the bars were ready to be machined. The machining operation consisted of removing, by the use of a lathe, layers of the carburized case of very definite thickness and catching the chips from each individual layer in a clean, oil free, enve10pe. The enve10pe was carefully marked to designate the steel and the exact layer it contained. The layers were removed in a systematic ‘ way, the first layer was .002 of an inch thick (a diameter decrease of .004) and each following layer was .005 of an inch thick, to a depth of approximately .062 inches or the core whichever came first. It was possible to tell when the core had been reached by the ease of machining and the type chip produced. The entire length of the test bar was not machined, thus leaving a stud about .75 inches long to be used later in the metallo- graphic analysis of the carburized cross-section of the respective bar. The steel chips were then analyzed for their carbon content in a standard carbon train (see diagram). The carbon contents of the various layers of the five steels were determined and plotted against the distance from the surface of the respective bars. The graph of carbon content vs. the distance from the edge of the bar indicated the carbon gradient in the case. Uare was \ -14- exercised to repeat exactly each step of the procedure involved in the carbon analysis so thatvreproducible results were obtained. Bureau of Standards samples were used to check the accuracy of the "carbon train" at the beginning of each run. The 35 inch studs which were not machined were then heated in used carburizing compound to a temper- ature of l700°F throughout and quenched. The quenching media was stirred water for the SAE 1010 and stirred oil (100 deg. cent.) for the four alloy steels. The five samples were then tempered at 400°F. for a period of 1.5 hours. The tempering was done to transform the light etching tetragonal martensite to a body centered cubic martensite which was dark etching when 5% nital was used as an etchant, thus making a sharp differentia- tion between it and the light etching retained austenite. ‘11 samples were then mounted in bakelite with a steel band mounted around their periphery to help maintain a flat edge on the polished samples. Polishing was done in two steps; first the sample was lapped on a lead lap usingmmedium lapping compound, then it was finished on a felt wheel using levigated alumina as the polishing compound. The polished samples were then etched with 5% nital. After etching the samples were ready to be photomicrographed. The procedure used in taking the Photomicrographs was to start at the edge of each sample -15.. and proceed to the core taking photographs at 500 magnifications of each .0064 of an inch until the retained austenite was no longer visible. This pro— cedure gave slight overlapping in each photomicrograph, allowing later matching of prints and thus forming a continuous photomicrograph from case to core. Part II Effect of Quenching Rates on Retained austenite The same five steels were used for the second phase of the experiment as were used in Part I. The bars were turned on a lathe to 0.500 inches diameter by 4.00 inches long. Flats were milled on each side of the bars (180 degrees apart) to a depth of .010 of an inch. One end of each bar was faced off while the other end was drilled and tapped to accomodate the 10-52 thread on the standard Jominy test bar holder. A sixth bar was machined from SAE 4640 steel to a diameter of 1.000 inch and a length of 4.000 inches and milled and tapped similarly to the other five bars. The six bars were then case carburized for a period of 15 hours at 1700°F. The carburizing procedure was the same. as that used in Part I. All six bars were then cleaned carefully and copper plated in a cyanide bath for 0.50 hours at a current density of 15 amperes per square foot. The six plated bars were then placed in a furnace -16- containing burning used carburizer which produced a slightly reducing rather than oxidizing atmosphere. The bars were heated completely to a temperature of 1700°F. All the bars were then quenched according to the standard Jominy end-quenching procedure (see detailed diagram), the 0.50 inch bars being quenched in a 0.25 inch stream of water while the 1.00 inch bar was quenched in a 0.50 inch stream of water. The quenched bars were then polished on the flats etched (5% nital) and examined microscopically to determine the critical transition points on the bars. Representative photomicrographs were then taken at 500 magnifications at these points. Rockwell-C hardness readings were then taken on the polished copper-free flats of the six end-quenched bars. The readings were taken every sixteenth of an inch for 1.50 inches from the quenched and, and then every eighth of an inch between 1.50 inches and 2.50 inches from the quenched end. These Rockwell-C values were then plotted as hardness (ordinate) vs. distance from.the quenched and of the bar (absdssa). A 1.00 inch bar as well as a 0.50 inch bar of SAE 4640 steel was end- quenched in hopes of obtaining some correlation of the cooling rates in the 0.500 inch bar as compared to the known cooling rates in the 1.00 inch bar. -17- Part III Transformation Range of Retained Austenite in SAE 1010 (carburized) Five samples of SAE 1010 (carburized, 1700°F., 15 hours, solid commercial carburizer) were quenched in stirred water. Each of the five samples were then tempered carefully at a different temperature, temperatures being 350, 400, 425, 440, 450°F. and the time of tempering 1.5 hours. The samples were then mounted as in part I, in bakelite. Polishing was done on a lead lap and a felt wheel. The polished samples were etched with 5% nital by swabbing the polished surface with saturated cotton swabs. The same approximate distance was chosen from the edge on each sample and a photomicrograph was taken at 500 magnifications. Effort was made to choose a rep- resentative spot characteristic of the amount of retained austenite in the chosen distance from the edge. One photomicrograph was taken of SAE 2015 (carburized, 1700°F, 13 hours, solid commercial carburizer) quenched in oil (100 deg. cent.) and tempered at 430°F. This was done to note any transformation in retained austenite over that of the 400° tempered sample. -18- Figure‘l The apparatus used for the carbon determinations Oxygen supply tank Gas pressure regulator Combustion furnace Oxygen washing bottle (cone. H2804) Ascarite tube (C02 removal) Combustion tube Zinc pellets C02 and 0 washing bottle (cone. 32304 and Cr03) 2 C02 and 02 washing bottle (cone. H2804) Ascarite weighing bottle Bench -19- F ure g The apparatus used for the end quenching KEY A- Carburized test bar (%" dia. x 4“) 8- Water supply nozzle (stream é" dia., 2%" head) C- Test bar adapter and centering attachment D- Test bar holder -20- Experimental Results The experimental results were divided into three major divisions and several sub-divisions, as follows: £9.22}. (a) Carbon gradient data for carburized SAE 1010, 2015, 2340, 3145, and 4640. (b) Graphs of the carbon gradient data for the five steels listed in Part I (a). (c) Photomicrographs of retained austenite in cases of the five steels listed in Part I (a). '6 F d‘ It: (a) End-quenched hardness data from the five 0.5 .inch and one 1.0 inch, end-quenched bars. (b) Graphs of the end-quenched data of Part II (a). (c) Microstructure of end-quenched bars (observing the polished flats). (d) Photomicrographs of the critical points on the polished flats of the five 0.5 inch end-quenched bars. '2’ fl cf AH e]... VH -Photomicrographs showing the transformation of retained austenite in SAE 1010 and SAE 2015 with increased tempering temperatures. -21- Cut No. (OCDQCfiUIiFCflNl-J F' F4 Id as +4 <3 Table 7 Data 0f Carbon Gradient In Carburized SAE 4640 Steel Distance From.Surface .000 - .002" .002 - .007" .007 - .012" .012 - .017" .017 - .022" .022 - .027" .027 - .032" .032 - .037" .037 - .0426 .042 - .047" .047 - .052" .052 - .057" 1.000 1.000 1.000 1.000 1.000 1.000 1.000 1.000 1.000 1.000 1.000 1.000 Sample Wt. gr gr gr gr gr gr gr gr gr gr gr gr .0381 .0407 .0355 .0322 .0307 .0303 .0275 .0242 .6882 .0242 .0207 .0162 Wt. 002 gr gr gr gr gr gr gr gr gr gr gr gr % Carbon 1.040 1.110 .968 .878 .837 .826 .750 .660 .714 .660 .565 .440 Part I Photomicrographs The photomicrographs in thds section represent steels that were all subjected to the following treat- ment and specifications. 1. Carburized for 13 hours at 1700 deg. Fahr. 2. Quenched from 1700 deg. Fahr. in «103014503 Water- SAE 1010 Oil- SAE 2015 SAE 2340 SAE 3145 SAE 4640 All steels were tempered at 400 deg. Fahr.,1.5 hrs. Etchant- 3% nital Transverse section Magnifications 500 X The number of the photomicrograph, steel, and distance from the surface in inches will be listed, in that order, on the page Just pro- ceeding each set of pictures. -52- }-1 i -‘4 ("v V ‘ Lj-LLJ... ‘— ’) l ‘12:; at“. . Q»... R}. Q . k a I A r- -4; .2. .‘I . J "’ _ I t u..i.V—'Q\ l at )lllli. V. ‘ w. rprbifitil x -54.. '5! u r- r I i ‘-' 3. ..‘ r a r. I «. r‘ 1 -. l '3 u, ‘ ,1 f‘ l e-a'? -55.. 8 SAE 2015 ~ 0.0000-0.0064 9 SAE 2015 0 . 0064-0 . 0128 10 SAE 2015 0.0128—0.0192 43 .-. ) v K" . .‘i ' ~- -57.. 11 _ SAE 2015 0.0192-0.0256 12 _ SAE 2015 0.0256-0.0320 -38.. 13. SA! 2015 0 .0320-0 . 0384 aha ..1 A | |||.l.¥l..lLll.o . \ vilifiiAAP. .. . . . p i ~39- r... r..- 5... o .lJlII.ll.1 9...... .413 r‘ on; J .. . .- r/VanaZ-l. - ..J -40- SH‘IU‘DVV " \ O '»./ O \./ rl ..- 7" .I .\ .n. .. C -41- ; . .1. . 3.: I'!’ ' ‘ I- " ‘ £14 0.0654-c, ' '43 ' " .1 . 7' L n ‘ . . - '0 _ k 0 0’ -'K_ '. ""5"“ , . 5‘ I . p ' . BB .2349.“ gas ' ‘ -L 8 .b, ._N- ... \ 9. ' . o . a . .: . I u A - I1 ‘ l . ,.. .. . -,r I." ’4 ‘1. in 0 L. ‘3‘ . . A .‘ g. , , "5 ' .15" ' o' . _ - ‘. Q j“ t! 4 .., ’ “IT. " ' .‘:.\ -42- I . . ‘ . l ’- . v ’ -1. ' l .r- ' ‘ I g! I l . , l , . ' l . n n: . .\ ‘1. L 'h .7 - :1 ' 9 ‘ I. {I ' 1. — . . _ . r " ‘0." ' v ' I ‘ ' " a. I 'I‘. I l . ~ ' 4, - . ' e l h n ‘ . . - ‘ . . - , s ‘ ‘ ' 3: A “‘.' I - : . 1' ' “ ' . ‘5 ' Q ' . ‘ I ’ . ‘ ' . “ I I ' . _ - ' . ,x-Jf . ' ‘ ‘ l I)» . '~ "‘3‘ J L." .7 ‘V - < . . - s11“ ’ : ‘ . . . 7 I .’\ 1 I“; .\. 7.14 w.- J 41.4 . v . . . . ._ i a . .- . . , .e . . . - I . .A , .t. 3 O .W- . 3 x v . v . O I a .3 \ w. . . . .J .- , .1 q a r 2 IV .. .. o . 4 y . . I “1.!ilflufilc. .c A lllluueLJ .. .. . .- ‘ _ _.._—.l'-" -43- . .. «Qu.-..t..||l.PI|.fr 33...... Iv # -44- 40.0448. 'ogoas4 ' . I B. N aria-tr— .— _'_ I \ -, | . 5 \ 'I Am.- I -7— .A I o - .\ l n 2' I y. I V.‘ ‘i l . , '~. t 7 ‘ 3 I . . _ “r - “HA J , i1 "-r—a ‘ . ‘ ."K\' I I. :;-‘l‘| rr‘ .\4 . .V_ r" ’ ~45- . Nut“: .rfvu.. -46- 1.4 2... \';Ja" -47- -48- n ‘I .535, viii 0’ :~\A. I... .1 b 1 ' K 1 w w i. x . ll; It?) .M . . ’JIJ ' V a (‘11) f\ r\ . 0215;, u". ‘\ (I ”J l K ,1._'J A—‘J '12-}. l (;‘-§5’3".‘;5.”), O -50- ‘ “in"?!m. .3 v - » 4 v . J -' a ‘ \ t"... a, .3 r :1“ ‘ a n , .,._~ \J.’ g ‘1; -51- 39 SAE 4640 0.0000-0.0064 40 SAE 4640 0.0064-0.0128 -52- 41 SAE 4640 0 .0128-0 . 0192 42 SAEE464’0 H 0 .o.0192-o.0256 ‘ % ‘ "O ‘3; \‘ .' i 0 fl,. \ ' “ . o “ ‘ I, ~ .17 49V” " ‘ a ‘ ' ‘ ‘1 ”lgi'é J;‘\'}£x. ."-‘ -.' 4’ ‘w‘. "it , 1. b): ((3 '-. w' 0‘ ‘ we. 390%“? \“a .g.“ - 0 «- .Hgfi-z-.v " v.03 ‘b ‘.-' t v x [3%) ,-. 'I I '31! ‘~ .~ fl 9 a» .’~.' .1‘. . ‘ . Vick" -54- by} '0' . I‘T/r'i.’;’\\ ‘2 ’aag“"'§‘ ,L";'¢1é“‘ . ~.:1!‘ .1 2:":‘6‘1—1 ' ‘r’ 11) a .b— ’ 1.. ,‘ :‘;»,, :5 Lfii'fi‘o’ \‘S , '. Logo Rw“!44, it. - ‘3 n “V ‘ M. I _.‘ "‘ ’ ' '. 1"”. :o.‘ 5.33:! ‘fl .." ,..“ J‘:"“ ”‘59,“, f - g"; 4:." T“ ’1 x02, “‘4', 1. ‘\ 0 '. \ .\ -’ (~12 gafi-.' . V‘ “ I1 9". - I“ .I~:o ' ' ‘ \I.. a. {’1' ~ \ -54- 45 SAE 4640 0 . 0384-0 . 0448 46’ .sas‘464c;_ 0.044e-0.0512 L’! I. K; is") >7 .9 v .- "‘5 «11",...3; \‘vc 0.,' (V... ‘- ’ I 'h ."‘ o.‘ A:"._' ‘\ a»; ’. drew-3‘ . . ' . 7‘ ' . u’ e -55- O :1 \ a I. c ’l y x. 1. f. f 1‘ I. V ‘ h _ I \ VA" 'w .5 I‘) .‘J I .81} . 1 *2 ct V, '1 ,\ -56- \ \ u o ’\ he. ‘ J B -57- All Bars 5" Except Those Designated Otherwise Distance From.Bar End 1(16ths) 2 {OCDQCDUll-PCR 11 12 13 14 15 16 17 18 19 20 21 N SAE 1010 58 59 55 47 46 45 45 44 43.5 42.5 42.5 43.5 43.5 43.5 43 42 42 42 41 41 41 Table 8 dominy Bar Data (Hardness-RC) SAE 2015 55.5 58 59 58 58.5 57 55 52 49 4e 44 45 45 42.5 44 45 44 45 45 44 45 SAE 2340 46 48 50 50 49.5 50 49.5 49.5 51 51 51 51 51 52 51.5 51 51 SAE 3145 53 56 55 56 56 56.5 57 57 57 55.5 55.5 56.5 56 54.5 55.5 56.5 54.5 55 54.5 55 58 Continued On Next Page -58- SAE 4640 55 56 57 58 58.5 59 59 59 59.5 60 60 60 60 60.5 60 60 59.5 59.5 58.5 58 SAE 4640(1") 63 62 62 62 62 62 61.5 61.5 61 61 60 59.5 59 59 58 57 56 55.5 54 52 51 Jominy Bar Data (Continued) Distance SAE SAE SAE SAE SAE SAE From Bar End 1.93.2 _2_g_1_§ fl 3145 _4_649_ 4640‘ 1") 22 41.5 43 50.5 58 56.5 51 23 41 42 51 54 55 50 24 40.5 42 51 55 52.5 49.5 26 59.5 42 51,5 55.5 50 48 28 39 41.5 51,5 55 49 ' 47.5 30 38 41 52 55 48 46 32 37.5 40 52 54.5 44.5 45.5 34 36.5 39 52 54.5 43.5 45 36 36.5 39 52 55.5 43 45 38 35.5 39 53 54.5 40.5 44.5 40 36 38 53.5 54.5 40 43 -59- Inl.31‘. . . Part H (g) Microstructure of End-quenched Bars SAE 1010 (1) The highest percentage of retained austenite was observed at the quenched end of the bar (photo - 59). (2) The austenite concentration gradually decreased .until at 0.150 inches from the quenched and of the bar patches of austenite and martensite were observed in a matrix of fine pearlite (photo - 58) (photo - 60). (3) The austenite and martensite patches disappeared at 0.375 inches from the quenched end of the bar. (4) The remainder of the bar was pearlite in varying degrees of coarseness. SAE 2015 (1) Very large retained austenite areas (50%) were found from the quenched end of the bar to 0.090 inches from the and (photo - 50) (2) Then there was a band of lesser austenite concen- tration between 0.090 and 0.150 inches (photo - 51). (3) Between 0.150 and 0.210 inches large austenite patches (50%) were observed again (photo e 52). (4) At 0.700 inches from the quenched end patches of combined austenite and martensite surrounded by fine pearlite were observed. The range of this formation was from 0.50 to 0.75 inches from bar end (photo- 53). —61- (5) The area between 0.210 and 0.500 inches contained scatterings of small austenite areas. (6) At 0.940 inches from the quenched end distinct carbide networks were observed and these networks con- tinued, in varying degrees, to the end of the bar. The cementite network began when the austenite-martensite patches disappeared. SAE 2340 (1) Retained austenite was quite constant, up the bar from the quenched end, with regard to quantity and size of areas to 0.50 inches (photo -454). (2) At 0.50 from the quenched end there was a slight indication of carbide network formation (photo - 55). (3) The retained austenite disappeared almost com- pletely when the carbide networks became quite pronounced at 1.50 inches (photo - 56). SAE 3145 A trace of retained austenite was observed at the extreme quenched end of the bar only. This was the only position at which any austenite was observed (photo - 8). SAE 4640 (1) There was a gradual decrease in the size of the austenite areas from the quenched end to the 1.310 inch position, at which point the carbide networks appeared (photo - 61) (photo - 62). -62.. (2) Between 1.310 and 2.625 inches from the quenched end the carbide network persisted. (3) At distances greater than 2.625 inches, there was pearlite in varying degrees of coarseness. -63- Part _;_§hotomicrographs The photomicrographs in this section represent steels that were all subjected to the following treat- ment and specifications. 1. Carburized for 13 hours at 1700 deg. Fahr. End quenched in a 0.25 inch stream of water All steels were then tempered at 400 deg. Fahr. for a period of 1.5 hours Etchant- 3% nital Longitudinal section Magnifications 500 X The number of the photomicrograph, steel, and distance from the quenched end will be listed, in that order, on the page Just proceeding each set of pictures. -64- “I l I" .Z . ’ 2““ ‘3" JL; (11'23 ($491.1 ”:9; r ~ '\) 50¢..J )3 éu Alr"..-‘ ‘2'“) x‘l.&\ _. \1. l -65- HOT}. 1 1-)! 1 d .qu u: 1) ‘1'} {)1 .1 ‘3 ,. 1. r3 1 -66.. kg. ..|4¢..H . amt ..r #. x: ‘1'-» . mil 5 .. .II.I.V4I‘L"P..I .. 4f) [(11). a I“ O , , .... r; .1 \l. . . . .. . ..« . f. ...3.uf".. a. v. t .. y. 0.4 an - . u l.- . . c . 44...». t... r$¢< a . . a ._ . T. v.,- 3%.”... 1.“.m‘. . -67- .3 a}. J It 7‘ a r... I. J '4?! -68- ‘ . )(llu‘fil heriison “ w -68- r ‘1‘. 1" '1 '1‘) d #3.:‘13 -69.. Part III Photomicrographs The photomicrographs in this section.represent steels that were all subjected to the following treat- ment and specifications. 1. 2. Carourized for 13 hours at 1700 deg. Fahr. Quenched from 1700 deg. Fanr. in Water- SAE 1010 Oil- SAE 2015 Etchant- 3% nital Transverse section Magnifications 500 X The number of the photomicrograph, steel, and the tempering temperature (deg, Fahr.) will be listed, in that order, on the page just pro- ceeding each set of pictures. -70- .... 15L {x )ll)..|. ills-1.-.. J lul‘ ‘. fl .. '1.— " .. -71- 'I ~72- Qiscussion During the experimental work some doubt had arisen as to whether or not the light constituent, in the photomicrographs was actually retained austenite. It was thought that this light etching material might have contained untempered martensite. The best results obtained as far as proving that the light etching material was actually austenite, was obtained from work done on varying the tempering temperatures of sections of SAE 1010 car- burized steel, that contained appreciable quantities of light material in the microstructure. Another indication was obtained by correlating the microstructure of the polished flats on the 0.5 inch end quenched bars to the Rockwell-C values obtained on the same surface. The work on tempering of carburized SAE 1010 steel was done by obtaining five pieces of the steel all containing approximately the same quantity of the light etching constituent. These samples were then tempered at varying temperatures between 350°F. and 450°F. and it was observed, as was shown in the photomicrographs of Part III, that little change occurred in the light areas until 440°F. was reached, where there was appreciably less white constituent than at 425°F. At 450°F. only very slight traces remained. This transformation was quite in agreement with Bain's work (7), as he claimed -73- a transformation temperature of austenite in plain carbon steels at 455°F. As was mentioned before an indication of retained austenite was obtained in the 0.50 inch bars of and quenched SAE 1010, SAE 2015, and SAE 4640 steels. The curves of these steels (Figure 8) showed a definite correlation between the areas of greatest retained austenite and the Rockwell hardness. 0n the other hand, SAE 2340 and SAE 3145 gave results that were definitely not expected in view of their microstructures. As an example, the SAE 4640 gave a Rockwell-C reading of 6 points lower at the quenched end where the cooling rate was the highest, than at a position .810 (approximately) inches from the quenched end. This was definitely abnormal for the end-quenched curve and thus indicated that the light etching material Imust have been somewhat softer (austenite) and caused this abnormality. Rockwell-C and file hardness tests were also taken on the quenched rounds of Part I. These results were discounted because Rockwell-C values were taken instead of Rockwell superfical, and it was thought that the Rockwell-C "brale' might have penetrated the case instead of revealing the surface hardness.' The file hardness test showed that the surface of the steel in every case was -74- softer than the file. This could mean that the case contained retained austenite or that the tempering treat- ment at 400°F. made the steel softer than the file. The photomicrographs were observed in Part I and it may be said that the higher the carbon contents, up to the maximum.of the surface, the more the tendency for the retention of austenite. It cannot, however, be said directly that the microstructures of Part I give a direct indication of the amount of austenite at a definite carbon content because other factors were involved. One of the factors was the difference in quenching rate from the surface to the center of the round quenched. The quenching media had a definite effect in the amount of austenite obtained. All in all, in quenching a carburized round, two variables will always be involved and therefore both must be considered in evaluating the data obtained by this method. If the photomicrographs in Part I are observed, it will be noted that the retained austenite seemed to decrease on the very edge. This was caused by a slight decarburization which either occurred during the carbur- ization or later when the samples were heated in used carburizer to be quenched. Two of the stools showed decarburization in their carbon gradients (SAE 3145 and SAE 4640). -75- The carbon gradient curves had to be weighted in order that they would present somewhat of a smooth curve for comparison with the corresponding photomicrographs. Several points strayed from the general trend of the carbon curves and upon repeating the run, similar inconsistencies were encountered. The inconsistencies were probably due to inhomogeneities in the carburized layers involved and also due to the extreme difficulty in collecting the "oil free“ chips. The values obtained should have been quite close inasmuch as extreme care was exercised in repeating the procedure very exactly each time a run was made. The retained austenite in any one layer or carbon content in the carburized sections of Part I varied considerably as the piece was revolved. When the photo- micrographs were made for Part I, this was taken into consideration and an average or representative point was chosen in every case. The same inconsistency arose on the flats of the end-quenched bars but in this case, observations were made down the center of the polished flats. These discrepancies probably occurred due to differences in the carburized cases and also they may have been due to an inhomogeneous austenite before quenching. The latter explanation was very possible due -76— to the fact that an austenitizing temperature of only 1700°F. was used and the thme at temperature was as short as possible to prevent undue decarburization. In Part III, it was found if several pieces of the same carburized steel were subjected to varying tempering temperatures that the inherent inhomogeneous structure caused differences in the results obtained. This was not true of the 450°F. temperature because practically all of the retained austenite was transformed throughout the complete sample. However, at lower tempering temper- atures, errors could have very easily crept into the work by using poor judgement in choosing the area to photograph. The tempering effect could have been observed more accurately by using one piece of steel and one area, and this area could have been photographed at small temperature incre- ments starting at 350°F. and continuing until 450°F. had been reached. Tampering time had a definite effect upon the amount of transformation especially in the temperatures around 450°F. and for this reason it was imperative that constant tempering times were used to insure any comparison whatsoever. The microstructures of the end-quenched bars in PartII seemed to indicate that the higher the quenching rate, the greater the quantity of retained austenite. Previous work -77- has shown, on cooling rates, that there is a region near the critical cooling rate, which gives rise to more retained austenite, and this has been used to explain, in some degree, the greater retained austenite in oil quenched parts than in the same parts water quenched. Only the SAE 2015 steel gave an indication of the fact that there may be certain cooling rates less susceptible to the retention of austenite. A transverse banding of retained austenite in the bar was observed along its length. From the quenched end to 0.090 inches from the quenched and existed a dense mass of austenite, then there was a band of lighter austenite concentration between 0.090 and 0.150 inches, past which large austenite par- ticles were observed and these gradually tapered off at a distance of 0.70 inches (approximately) from the quenched end of the bar. The area seemed too well defined to be considered an inhomogeneous segregation although it was not an impossibility. Some error was undoubtedly intro- duced when the copper was removed from the flats by hand honing and for this very reason, no definite statements can be made other than the close relation of the hardness curve to the microstructure. The error, although probably only a few thousands of an inch, may have meant a difference in carbon content of 5 to 20 points. The microscopic data obtained in Part II had to be correlated to relative cooling rates within the bars because of a lack of high speed temperature recording equipment. The 1.00 inch SAE 4640 bar was run in an effort to try and link up to its known cooling rate that in a corresponding 0.50 inch bar of the same steel. The only comparison that was possible was that between the points on each bar where the carbide network just began to form. After a very careful micro- scopic examination of each bar, it was decided to throw out this means of getting at the cooling rate, due to the wide discrepancy in the starting point of the carbide network. Another method was used to try and correlate the quenching rates of the 0.50 inch and 1.00 inch bars by the use of previous data obtained on SAE 4063. By graphing cooling rate (ordinate) vs. distance from.the quenched end (abissa) for the SAE 4063 steel, it was found that the curves were extremely hard to interpret as such, and that before a comparison could be made it would be necessary to have more data particularly on the steels used in this experbment. Hetallographic methods (16) are not the only means of getting at the amount of retained austenite in a steel sample. As a matter of fact, there are several methods much.more satisfactory but again, a lack of time -79.. and equipment formed a barrier to a possible check on the results obtained. The determination of retained austenite has been done by magnetic methods, X-ray methods, specific volume changes, electrical resistance methods, dilatometric methods, and micro-hardness determinations. An effort was made to link retained austenite with the Mr temperatures of the five steels used in the exper- iment. hfter several calculations were made, it was found that the “f temperature in the high carbon high alloy steels fell below that of room temperature. The If temperatures for plain carbon steels of a similar carbon content were found to come quite close to that of room temperature. Lower carbon contents caused the Mr temperature to fall above room.temperature. The data used in the calculations (17) showed quite a wide discrepancy and this had to be considered in the calculation of Hf temperature. It was observed that the plain carbon steels contained the least retained austenite while the higher alloy steels exhibited the most retained austenite in s similar carbon content layer of Part I. This phenomenon might have had some bearing on the fact that the higher carbon contents showed more retained austenite, and that it was necessary for the carbon content to be quite high in order to have the Mr.temperature fall below room.temper- ature. Also correspondingly lower carbon contents exhibited the same retained austenite percentage in alloy steels ~80— as compared to the plain carbon steel, this being due to the effect of the alloying elements on the “f temperature. —81- Conclusions Higher carbon contents promoted the retention of greater quantities of austenite, all other factors remaining constant. The retention of austenite is not always the great- est at the critical cooling rate. This was indicated by the alternate bands of retained austenite in the SAE 2015 (end-quenched), with a band of very little aus- tenite between them. The light etching constituent austenite was complete- ly transformed into one of its transformation pro- ducts, in a plain carbon steel, at about 4500 F. if sufficient time was used during tempering. Retained austenite appeared to transform, upon temp- ering, over a range of temperatures rather than at any single temperature. The higher alloy steels retained more austenite upon quenching than did the plain carbon steels. -82.. Possible Futur§_fl9rk More work should be done in the future on the end— quenched bars. It is felt that if a means of determin- ing the exact cooling rates within the bars during quench- ing were made available some interesting data could be obtained in connectionnwith.the work on retained austenite. A means of determining more correctly the effect of increased carbon content could be obtained by quenching, in the same media, several thin strips of the same steel, carburized to different carbon contents. By using thin strips the variable caused by the quenching rate could be completely eliminated. In this connection, strips of the same carbon content could be quenched in different medias to determine the effect of varying quenching rates upon the same section of a similar steel. Also some work should be done in the future regard- ing the effect on the retention of austenite of quench- ing from higher temperatures in the austenite range. -85... 1. 2. 3. 4. 5. 6. 7. 8. 9. Selected References Metals Handbook, 1939 Edition, American Society for Metals, Cleveland, Ohio G. V. Smith and R. F. Mehl: "Lattice Relationships in Decomposition of Austenite to Pearlite, Bainite, and Martensite," Metals Technology, April, 1942 A. B. Greninger and A. R. Troiano: “Crystallography of Austenite Decomposition,? Metals Technology, August, 1940 E. C. Bain: "Alloying Elements in Steel," American Society for Metals, Cleveland, Ohio, 1939 Samuel Epstein: "The Alloys of Iron and Carbon," Vol. I Constitution, McGraw-Hill Book Co., 1936 3. G. Fletcher and M. Cohen: "Subatmospheric Trans- formation of Retained Austenite," American Society for Metals Transactions, Volume 34, 1945 E. C. Bain: "Alloying Elements in Steel," American Society for Metals, Cleveland, Ohio, 1939 H. J. French: "Alloy Constructional Steels," American Society for Metals, Cleveland, Ohio, 1942 D. P. Antia, S. G. Fletcher, and M. Cohen: "Structural Changes During the Tampering of H1gh Carbon Steel," American Society for Metals Transactions, Vol. 32, 1944 -84- 10. ll. 12. 13. 14. 15. C. A. Liedholm: "Retained Austenite and Its Decompo- sition Range in a Quenched Cobalt High Speed Steel," American Society for Metals Transactions, 1935 F. Wever and N. Engel: "Uber den Einfluss der Abkuh- lungsgeschwindigkeit auf die Temperatur der Umwan- dlungen, das Gefuge and don Feinbauder Eisen-kohl- enstaff-Legierungen"(Effect at Cooling Velocity on the Temperature of Transformations and the Structure of Iron-Carbon Alloys), Mitt. K - w. Inst. Eisenforschung, v. 12, 1950 pp. 93-114 ' E. 3. Davenport and E. C. Bain: "Transformatien'of Austenite at Constant Subcritical Temperatures," Trans. Am. Inst. Min. Met. Eng., vol. 90, 1930, pp. 117-154 H. Esser and H. Cornelius: "Die Vorgange beim Anlassen abgeschreckter Stahle"(0ccurrences during the Tampering of Quenched Steels), Archiv f. d. Eisenbuttenwesen, v. 7, 1934 G. Tammann and E. Scheil: "Die Umwandlugen des Austenite und Martensits in geharteten Stahlen" (The Transformations of Austenite and Martensite in Hardened Steel), 2. anorg. allgem. Chem., v. 157, 1926 K. Tamaru and S. Sekito: "On the Quantitative Deter- mination of Retained Austenite dn Quenched Steels," Sci. Rep., Sendai, ser. 1, v. 20, 1930 -85- 16. P. Gordon, M. Cohen, R. 3. Rose: "The Kinetics of Austenite Decomposition in High Speed Steel," American Society for Metals Transactions, Volume 31, 1943 17. R. A. Grange and H. M. Stewart: “The Temperature Range of martensite Formation,“ Metals Technology, June, 1946 -86-