llfllfilllll I t I l I IIIHIIHI ’HTHS X STUDY OF THE GRAIN GROWTH HF A LUW ITARBON STEEL THES‘S FOR DEGREE. OF M. 8. JAMES WARE: PERCY 1926 .J—.—.HL ‘- v‘vID-A ~w‘ ~3- - :- 11-1555: ....m...:...:....! d1! tin. 21. fl F... 1.1.3.; Elk»! _ . gr...n,.,...9.........l is T. ....... . ‘lnullv i P--. 1‘ I ‘ .fi A STUDY OF THE} GRAIN GROWTH OF A LOW CARBON STEM r‘ k. .. O .. 1*; A STUDY OF THE. GRAIN GIIDW’IH OF A LOW CARBON STEEL Thesis Submitted to the Faculty of Michigan State College of Agriculture and Applied. Science In Partial Fulfillment of tb Requirement: for a Degree of Master of Science James Ward Percy June, 19%. THEE-:1“)? JEEEEEZEZEEEE With the continued use of the microscoPe and. exhaustive chemical and thermal analysis, much has been learned concerning the internal struc- ture of pure metals and alloys. Many hundreds of observations of metals and alloys have been made by Which equilibrium diagrams have been cal- Oculated and produced. Since this is a study of low-carbon steel, a brief review of the crystallizing process of a low-carbon steel might well be made. This is quite simply done by referring to the diagram (1) given on page 2. Beginning at the t0p of the diagram on the X oPdinate, representing a low-carbon steel such as was used hereinafter, the crystallizing process is as follows: On reaching a temperature of 27320 F. (15000 O.) on AB, austenite or the solid ,solution of F830, cemen- tie, in gamna iron begins to‘ crystallize. At the line AB, this precipi- tation has been canpleted until the mass has completely solidified. As the cooling continues ani the temperature of 16l6° F. (880° c.) on GS is reached, ferrite or pure irefi'begine to precipitate. At a temperature of 13370 F. (725° 0.) on P8 the austenite has completely broken down, leaving a matrix of ferrite with crystal of eutectoid or pearlite, lamellae of ferrite and cementite. It has been found, however, that the size of these final grains of ferrite and pearlite were by no means constant or the same. They seemed to vary because of many factors. In fabrication, the pouring tanpsrature, size of ingot, rate of cooling, and nature and quantity of impurities were found to be potent factors. It was also found that, with 103334 assumes} .AW. eiuflffiflu \thuUtJTm NMNN, Nkb .n w h; u. n. N. \ NEG .wfikafisxmu u .flerth «aux- W+ i _ d \ a Ne? w 4st§etet t 12”er W. uses»: “SN b 5* i. .wbkzwxewu \MA «5333!. o Q \b§\ V / °\\ww\ V NunN \\ III! \ Q \33 // a . a “i famqpaummfiaez fi § § ‘§ 19. the advent of heat treatment, this gram-size varied with various treat- ment. Mathewson and Phillips (2) suggest, that for each tetnperature of anneal, a mean size of new grain is fomed. It was found that continued annealing at this same tauperature produced little or no increase in grain size. This condition has been called a state of "grain-size equilibrium" (3). This grain-size equilibrium or maximum grain size is best produced at a critical temperature which Howe has pleased to call a "germinative temperature" (4). £113 temperature, it has been found, is different for different metals am alloys and may vary with different degrees of cold-working in the same metal. This might more prOperly be stated as a short range of temperature, below and above which little growth is observed. In general terms, grain growth seems to be favored by grain fineness, grain-size contrast, and by prior plastic deformation. How- ever, tha latter may be, actually, a combination of the first two causes. It is logical to believe that the higher the temperature, the greater mobility and, hence, the greater grain-growth - provided the alloy does not materially suffer from the extreme heat. Obstructions, such as sonims, slag, oxides, etc., tend to Oppose grain-growth. Decrease of grain-size contrast likewise hinders growth. The rate of growth scans to increase with the temperature. While all of the above causes of grain-growth have been formu- lated upon observations, no specific results of the grain-growth of a carbon steel were available. In this report all factors except that of tauperature were reduced, or eliminated, as far as possible. The re— port becomes a study of the effect of heat on the grain-growth of a carbon steel. 4- The steel used in this report was one which corresponded to an S.A.R. 1020 steel. It was fabricated in an American mill by the electric furnace process. The material was obtained as a bar 1" in di- ameter (1" cold rolled). A preliminary examination, chemical and micro- scepic, showed it to be almost entirely free of sonims and other occlud- ed matter. The specimens used for heat-treatment were cross-sections of the bar about 3/8" thick. All specimens were out from one bar ~— 30 that variations in composition, prior plaster deformation, and prior heat-treatment might be reduced to a minimum. All heat-treatments were carried out in s Hoskins Furnace, Type Edi-104, 4" in diameter. Since this is standard equipnent, no further description is necessary. All tenxeerature measurements were made by means of a Leeds 6. Northrup Hand-compensated Potentiometer Indicator and a Hoskins ILA. thermo- couple. The themocouple was introduced through the cover of the furnace. All photomicrographs were made with a Banach 5: Lamb Metalloscope, Type nan-9. Unless otherwise stated, all photomicrographs were of longitudinal sections and were taken at 100 diameters with the following settings on the metalloscOpe. Objective................................16 Inn Eyepiece................................. 7.5 x Bellows..................................53 cm. Aperture.................................3/S" stOPOOOOOOOOOOOOOOOOOOOOO0....0.0.0.0...01 (larfi’ harposure..................................90 sec. Source of light........................... 6 volt, 108 watt Tungsten Bandllasda. Developer.................................Ii-Q Time of deveIOping........................72 secs. For those micrographs taken at 500:, the following were used: Objective“.............-.................. 4 an Eyepiece.................................. 7.51 Bellows...................................4l.5 cm. iperture..................................1/4.n Step......................................l Exposure..................................6oo sec. Time of deve10ping........................72 sec. For those taken at 10001, the following settings were used: Objective....._............................l.9 (oil imnersion) Eyepiece..................................7.5 Bellows...................................39 Aperture..................................l/8" Stop......................................l upcsure..................................l hr. Time of deveIOping........................80 secs- All photographs were taken on plates manufactured by the Eastman Kodak 00., Eastman Gamercial brand. The specimens photographed for this report were etched with 2% solution of nitric acid in ethyl alcohol, the time varying from 4 - 20 seconds, depending on the condition of the sample. ' 5‘... O I U 9 . s s . s s . o . o s - g \ I s I a ‘ I O h P ‘ O U I l sljga I O 9 D s . I O Q \ . l O ) O i I h ) . . t I O l 3 s I i O 1 O O I b I s o . n v 0 1 I O 0-] Of a series of samples, heat-treatments at various temperatures, with varying durations were made. In the body of the experiment, the rate of cooling was the same, i. e., 100° F. (56° 0.) per hour to 10000 F. (558° c.) and from thence within the mrnace with its natural cooling period. As a secondary experiment, the effect of the rate of cooling thru the critical temperature range was determined. One samme of the 1" bar, 12" long was drilled on its central axis to receive a thema- ccuple. This saInple at each step was held at 19000 F. (10380 C.) for 1/2 hour, cooled to 1500° F. (816° 0.) in about 5 minutes and then cooled from 1500° F. (816° 0.) to 1200° F. (649° c.) in two hours and in one- half hour, in the furnace, in air, in oil, and in water. After each stage of treatmnt,ta section of the piece was rancved, examined, and photographed. As a final check on the effect of the 1900° F. (loas° 0.) treatment, the piece was held at 19000 F. for 1/2 hour and cooled thru the critical range within the furnace. In this experiment, grain—size determinations were made by a specially devised method. This consisted in measuring the average grain diameter of the pearlite in the photographs taken at 1001. This was taken only as a simple method of comparison in this report. QECU§SION When this report was begun, the effect of temperature on the grain-size and growth of the steel was its ultimate purpose. However, before many determinations or heat-treatments had been.made, another factor was introduced. It was noticed in the examination of the heat- treated specimens, that the pearlitic areas seemed to follow a definite orientation. This orientation seemed to be closely allied with.the roll- ing grain, so-called, and seemed to produce a laminated or striated effects on the longitudinal sections of the specimens.‘ As successive heats were made, this effect became more pronounced. Thus it was incorporated in the body of the report. Since this is a study of a series of'temperatures of varying duration on a single steel, it seems well to examine photomicrographs of spechmens of the heats. As has been said, all photOgraphs were taken of longitudinal sections of the specimens, which presented a section parallel to the direction of rolling. Figure l is a photomicrograph of the steel, as received. The average grain diameter is about 60 units. a careful examination shows no evidence of a regular striaticn of the pearlite areas. While evidence of some occluded matter can be found, the steel, in.general, is fairly clean. Figure 2 is a section of the above specimen presented at 500x. Figure 3 is another section of the same specimen taken at 1000:. Pearlite in its lamallar state is found. Figure 4 is a.micrograph of a Specimen.heated to 17000 F. (927° 0.) and cooled to 1000° F. (538° 0.) at the rate of 100° F. per hour. The average grain diameter at 100: is about 75 units. Here the first signs of the lemmated or striated effects are found. The pearlite is arranged in definite rows the length of the photograph. Figure 5. This specimen was held at 1700° F. for 1 hour and cooled at the usual rate of 1000 F. per hour to 10000 F. The grain diameter has increased to about 90 units. The striated effect is still more evident. Figure 6. This specimen was held at 17000 F. for 2 hours and cooled. The grain diameter has increased to 110 units. The striated effect is still evident. Figure 7. This specimen was held at 17000 F. for 5 hours and cooled. The grain diamter has increased to 135 units. The striated effect is still evident, though not quite so pronounced. Figure 8. This specimen was held at 1700° F. for 10 hours and cooled. The grain diamter is now about 170 units. The striated effect is still evident. Figure 9. This specimenuwas held at 1700° F. for 36 hours. The diameter of the pearlitic area has increased to 225 units. The striated effect is evident, though not sharply pronounced. The decrease in total pearlitic area is very probably due to dicarburization. Figure 10. This final specimen of this heat was held at 17000 F. for 72 hours and cooled as usual. While a great amount of decarburiza- tion has taken place, the grain diameter has increased to 240 units. The striated effect is still present. Throughout this 17000 Ft heat, it will be noticed that the grain diamter slowly increased from.75 to 240 units. The striated ef- fect persisted. The duration of the heat seams to be responsible for the decarburizat ion. Figure 11. This is the first specimen of the next heat. It was heated to 18000 F. (983° 0.) and cooled. The average grain diamter 'was found to be about 140 units. The abnormally large grain shown in.the photograph seems to be canposited of four grains. The striated effect again is evident, though less pronounced than in the 17000 F. heat. Figure 12 shows a specimen held at 18000 F. for 1 hour and cooled as usual. The average grain diameter is about 185 units. The striated‘ effect is still evident. Figure 13 shows a specimen held for 2 hours at 18000 F. and cooled. The average diameter is but slightly larger, 200 units. The pearlitic areas seem to coalesce to form long ”stringers." Figure 14. This specimen was held at 18000 F. for 5 hours and cooled. The grain diameter has increased to 225 units. The striated effect is evident but slightly. Figure 15 shows a specimen.Which.has been.held at 18000 F. for 10 hours. The grain diameter has increased to 250 units. There is now but little evidence of striation. Signs of occluded matter, very pro- bably ferric oxides,are seen. A small amount of decarburization.has taken place. In this 1800° F. heat, the grain-size has increased from 140 to 250 units. Practically all evidence of striated orientation has been obliterated. Figure 16 is the first section of the 18500 F. (10100 0.) heat. It was brought to the heat and cooled as usual, 100° F. per hour. The diameter of the pearlitic area is about 200 units, or the size of the grain of the specimen held at 18000 F. for 2 hours. Again, the striated orientation is but slightly evident. Figure 17 shows a sample held at 18500 F. for 1 hour. The grain- size is but a little larger, 210 units. The striated effect is practic- ally obliterated. Figure 18 shows a specimen held for 2 hours at 18500 F. The grain diameter is now about 255 units. The striated effect can only be discerned with careful examination. Figure 19. This specimen was held for 5 hours at 18500 F. The grain-size has reached a size of 250 uni ts, almcst a maximum. The pearlitic areas are well broken up. Figure 20 shows the longest anneal at the 18500 a. heat, 10 hours, followed by the cooling. The grain-size has now reached itslnmxnnmn of 265 units. Very little evidence of striation is perceptible. At this point, a section of the sample shown in Figure 10 and a sample of the steel as received were heated together for 4 hours at 18500 F. and cooled 1000 F. per hour. Figure 21 is the specimen.Which had been held fer 72 hours at 17000 F.,'While Figure 22 is a sample of the original material. The size of the average pearlitic area is about the same, the diameter being about 250 units. Figure 23 is the first member of the 1950° F. (lO65° c.) heat, being brought to heat and cooled. The average grainpsize is about 190 units. There is little evidence of striation. Figure 24 is of a sample held at 19500 F. for 1 hour. The di- ameter of the average area is about 210 units. Figure 25 shows a specimen which was held at 1950° F. for 2 hours. The grain diameter is about .220 units. Figure 26 shows a specimen which was held at 19500 F. for 5 hours. A large amount of decarburizaticn has taken place, Figure 27 is of a specimen held at 1950° F. for 10 hours. A greater degree of decarburizaticn has taken place. It scans evident in the 19500 F. heat that the rate of grain growth is rapidly decreasing. Long duration of the heat only serves to decarburize the steel, thus reducing the amount of pearlite present. At this temperature all effect of previous heat treatment or mechanical work- ing has been removed. Figure 28 is the first specimen in the section of the experiment devoted to the effect of the rate of cooling through the critical range on the grain growths. The specimen shown in Figure 28 was held at 1900° F. for 30 minutes, cooled in the furnace to 1500° F. in about 5 minutes, then cooled from 1500° F. to 12000 F. in 127 minutes. The grain-size is roughly that of Figure 19 (q.v.). Figure 29 shows a specimen held at 19000 F. for 50 minutes, cooled to 1500° F. in 5 mimites, and then to 1200" F. in 33 minutes. ihe grain-size is little changed. Figure 30 shows a specimen held at 19000 F. 30 minutes, cooled to 15000 F. in 5 minutes, and then cooled to 12000 F. in 1.95 minutes in air. A roughly formed pearlite is found with evidence of the susten- ite grain boundaries. Figure 31 shows a specimen.held at 19000 F. fer 30 minutes, cooled to 15000 F. in 5 minutes, and then quenched in oil to 12000 F. in 4 minutes. ' Figure 32 shows Figure 51 at 500 1. Here it is shown that a little evidence of the austenitic boundaries remain. All ferrite and cementite is in very finely divided lamaller form. Figure 33 shows a specimen held at 1900° F. for so minutes, cooled to 1500° F. in 5 minutes, and then.quenched in water to 1200° F. in .12 minutes. Figure 34 shows Figure 33 at 500 x. Somthat more of the austenitic boundary remains due to the sudden change in.mobility. The structure, however, is of a very fine nature. Preper heat-treatment, probably a quench and.anneal, might bring the specimen back to its orig- inal grainpsize, as received, though the structure might vary a bit from the original. Figure 35 shows a specimen held at 19000 F. for 30 minutes, and cooled in the furnace after the last series of cooling treatment had been made. All evidence of these treatments are shown to be obliterated in this anneal. It will be remembered that, in this later section of this report, dealing with the effect of cooling, specimens were cut from one bar. Since a section.was out off after eadh treatment and examined, all specimens were subjected to the preceding treatments. mhus, the sample shown in Figure 55 was subject to all these later experiments. CONCLU§I ON In a consideration of the results brought out by these exper- iments, the effect of teznperature on grain-growth is of primary impor- tance. A portion of the data has been represented graphically on the follOWing page. It seems, first, as was expected, that the rate of grain- grOWth increases With an increase in temperature. Secondly, there seems to be an equilibrium grain-size for each temperature. This equilibrium size seems to increase with an increase in tanperature. At present, it seems a bit logical to believe there may be some maximum equilibrium size Which all temperatures tend to approach, though perhaps never reach. At any temperature, the rate of grain-growth decreases with the duration of the heat. If grain-size alone is concerned, an 18500 F. heat, with a dura- tion of 2 - 5 hours, seems to furnish Optimum conditions. Beyond this temperature, no great increase in size is effected and decarburizaticn becomes serious. Because of this last fact, the 19500 F. heat was not represented on the graph. The matter of striation effect is also of great importance. It was brought out in the 1700° F. heat, very probably by a release of in~ ternal stresses due to the increase of mobility at 17000 F. This effect was made less pronounced as the temperature was increased. Again, a 2 - 5 hour heat at 18500 F. scans to be effective in obliterating this striated effect. While no observations have been made to substantiate the belief, 0\ m. m k 0 hi V h: § n Vi x . vex Ax?“ hUMw‘ flex Cog \ \c \ a} \ § 313,19“; 0/6 inf/0J9 V’o SAM/fl it seems possible that sections or "stringers" of nonpmetallic inclusions omight be broken up at 1850° F. as were the "stringers" of pearlite. Should this later belief prove true, it might be of very great importance from an industrial viewpoint.. While the section of the experiment devoted to the effect on the grain-size of the rate of cooling through.the critical temperature was of secondary importance, several interesting points were noted. First, the grain-size is dependent upon the rate of cooling. As is to be expected, after a quench there is evidence of the retentidn of the austenite grain boundaries. The most important point seems to be the effect of the final 18500 F. heat. All traces of all rapid cools and quenches were entirely removed. The data obtained from the experiment was entirely insufficient to detennine the exact rate of cooling necessary for a small pearlite grain, similar to the original. It seems to be between a 2 minute to 8 30 minute cool thru the critical range. This question of the effect of the cooling rate on grain-growth is obviously one for further study. In concluding, we wish to acknowledge the aid of Mr. H. T. Publow, Professor of metallurgy, upon Whose advice and instruction, the success of these experiments largely depended. 1. 2. 3. 4. 5. W American Society of Steel Treating Handbook, page A-2D. Recrystallization of Gold-worked Alpha Brass Re-annealing, Am. Inst. Min. 8. Met. Eng., Vol. 54, pages 608-658. Discussion of (2), Jeffries, ilib. 660. Grain Growth Phenomena in Metals, Jeffries, Am. Inst. Min. & Met., Vol. 56, pages 571-582. Detennination of Grain-Size in Metals, Jeffries, Axner. Soc. Min. Met. Bugs, V01. 54, pages 594-507. -15- As rec'd 500: Iig.2. As rec'd 1001 113.1. 1700‘! To heat. rig.4. As rec‘d lOOOx 113.3. or .. 3.9.0.113... l f . J. J (r . ., so I . i . e... 433.....1fi. ., .....y «X. a: $4842.». .. .. .1. a. f. . s. M It- v ergtfif. 75 t . 'c %z‘ . p} . o n,‘ .’:~).I [fl r. . . .3ch wg < saw - .',3-Dkri germ-1. View.» (1:. . . I. . . r . v H O 7*. I ..\v . ease... . . . . {newts .. e . F I ... I... . . .. u e s . ’ ‘(‘.h(o.l. s . n. . . s.......| r .‘ . .‘v......... 1......A... . 1700 3.92 hours. 113.6. 1 hour. I 1700‘ Iig.5. . .4 . . . . .. . . .. . , . . .c , ,. . . .. ...v . . . s. , . I 2. . I. . ..' v . . ,. . . a. . . n v 0% s :: . r . - vi . 2. . ... .a a . .. .. . . .. ..I.. L .. . 1. : K . v.2. . .u .v e .e.‘ I . a . w. .. I A... a .a J. :..W. .. :Y. y. I I 3.. :0 IL. «\v f... ,_ .7 4’ ...... .. . ,. __ u}. .1 11...: a. and... a bra : s z .w .."r\.. .u Jesse“... 10 hours. 1700°F. iig.8. 5 hours. F. 1700 Fig.7. MS... 1‘. < )1. - Haiti?! 1... (t. 1 hour. Iig.6. 1700 F.°2 hours. 1700°F. rig.5. aumsewmw. 10 hours. 1700°F. Iig.8. 5 hours. 1700°F. Fig.7. _....r .. P. s One... ”v V- .. .4. s .w yaw! . .. i \s. h.» 72 hours. 1700°F. Fig.10. 1800°F. 1 hour. 113.12. 56 hours. 1700°F. ,..:.. e w .... ... .flr. . LLJ .. .. .. a .\ m 4 it, w... .. ._.... . .a/ sea” .5... fl 1 , . L. , . ...u~7w . .skw ; .w. . . . . _ a. quantifies ow. . .. m , I p y . .. . ./. .1.Wfikaba2. , ,...............Huw .53.. Fig.9. To heat. 1800°F. 513.11. . 3 E'::. 7,3: 1_ ltgv; 'm: 1 fix . . ‘7‘". .- M'.'1’ ,JJ, lei 1‘ ‘5‘ .1. Fig.15. 1800°F. 2 hours. Fig.14. 1800°F. 5 hours. 313.15. lsoo°r. 10 hours. Fig.16. 1850°F. To heat. 113.17. laco°1. 1 hour. 113.18. lsso°1. 2 hours. ’ I 'g .‘7 fl . I, ‘ 9 \ .' ' ~21. ‘. \‘n :‘n ’. .' . . I - _ . ‘_ . .' ' "t' 7 _ : 17" . 113.19. leso°r. 5 hours. 113.20. lsso°1. 10 hours. Fig.21. 1700°F. 72 hours. Fig.22. As received. Heated together - 1850°F. 4 hours. Fig.23. 1950ol'. To heat. Fig.24. 1950°F. 1 hour. 113.25. 1950°1. 2 hours. 113.26. 1950°1. 5 hours. 113.27. 1950°F. 10 hours. 313.28. Cooled 2 hours thru critical temperature. , 7 1: saw. is: 1x3“ “ N531 : ‘5 “gig? ‘1 EQQAVM 37* - .‘xq ‘ _l:\‘$‘:§_ p {‘7‘} . ,/ F {.1 ,. s . Fig.29. Cooled 55 minutes Fig.50. Cooled 1.9 minutes thru critical temperature. thru critical temperature. ’ 192.1; " 9‘15""? . . . v. . I I ‘3'. A ~ ~ . * '. ”fir, ' 'V-J‘M " . . . s ' ' ’2' ‘d‘kfi‘g’l -- s ’frvf, .7 .. . I 7 s : ‘1 h ': s Fig.51. Cooled .4 minute thru critical temperature. o1 - r a a . y"... _ . xx».- . w. . . I 1’ . . .... . .. an? . .. ‘3 3.. t. .1}..- .bFup...Cr.|,-Pu . .- 3... v . ., . .&>&NVH~..~.MWO . . 1 0.1 L _. _ 500x. Same .34. Fig minutes 12 thru critical temperature. 313.35. Cooled Furnace cooled 1900°F. Fig.55. 071.811...) - Prlaflnl' Elli-Ilia}- —.-.r—‘. 1 r .fiiww MICHIG I III III 3 1293 STATE UNIVERSITY LIBRARIES | IIILIIIIIIHI 3 42 8786