THE BETA TRANSFORMATION IN BRASS AND THE , ' ~ . ‘ PHYSICAL STRUCTURE OF PEARLITE ThesiS fofth'e ADeg-ree of M s; T. A MICHIGAN STATE COLLEGE: ...... I- ' 0“ - .. v . ,. ' “ ~ .. 4 ~ . . .‘ . ' . ' v .- - u I . ‘ T 4. ..c A . I. ‘ ‘ .- | I o a ‘ ‘. QQQQQ aaaa ZZZZZ ''''' . . i . -. . u ..-l 1 1 c: . I . _. I. 11- . ‘. 0 1 .l . I .. .. . u . t ~ "0’ ‘_ '1’ q. . t. a r. 4.. - .L III. v “60! - o v . v v m.‘ Va n .. IA F o A . .1. J. . ;.§;§.§§§3§§§§c........_.a__.__....._.. _ .. . ... , .. , .. _...:............_.._,,,..,,:....._./..x.., . _ A . . . . I . a . . . ., A?! ~1flI-Illavllui run: u. .quflliu .4! a v c: T H E B E T A T R A N S F O R M A T I O N I N B R A S S and u: T H E P H Y S I C A L S T R U C T U T E O P E A R L I T B By Raymond E. Schwyn m Submitted in partial fulfilment of the requirements for the degree of Master of Science Graduate School Michigan State College Department of Chemical Engineering June, 1938 THE‘Q‘S The author wishes to acknowledge with thanks the kind assistance and timely suggestions of Professor H. E. Publow 1214.87 Part I THE ETA THAHSFOHMATION IN BASS Introduction Historical Experimental Summary of Results Suoeestions for Future Work {2 ii) Biblioaraphy Part II THE PHYSICAL STRUCTURE OF PEARLITE Introduction Experimental Summary of Results Sugnestions for Future Work CD 39 41 Part I THE ‘ETA TRANSFORMATION IN BRASS n: 7.01 UCTIDI‘? Much research has been done on the Beta transfor- mation in brass, but since the publication of the first paper on this subject in 1837 it has been a matter of controversy. It shall be the purpose of this paper to survey the literature to determine the points of disagreement, and attempt by experimental proof to clear up the dis— puted points. Beta brass is the copper-zinc alloy with concen— tration limits of 50 to 54.5 percent copper at room temperature. The concentration limits of the peritectic horizontal are from 45.5 to 59.5 percent copper and the r, temperature is 8500 C. At approximately 4700 C. the Beta brass shows a heat effect. This heat effect has been the cause of most of the disputes in the Beta transformation. The Beta brass solid solution is not malleable at ziny'temperature and is used commercially only in the czast form. The crushing strength is very much higher tllat Alpha brass but the tensile strength decreases with iricreasing amounts of the Beta constituent. The modulus Of' elasticity is low and the elongation is very low. $318 structure is as shown in the following photomicro- SITiph, a solid solution having very large grains. u . i2- .lrlul rat‘s}, ‘ -2- Beta Brass Annealed Slow Cooled 200 X w 1.... HISTORICAL The first work of importance concerning the copper-zinc equilibrium.was presented by Roberts-Austen (1) in 1897. In this work he determined the complete freezing point curve or liquidus and several horizontal lines which he termed "eutectics". Included with these so called eutectic lines was the line e"e' (figure 1) at about 4700 Centagrade and form 45 to 75 percent copper. Although he called this line a eutectic, he advanced no evidence to support his theory. The e"e' line has been the subject of con— troversy since the publication of Roberts-Austen's paper. Immediately after Roberts-Austen had presented his paper, Uharpy (3) contested the presence of a eutec— tic at this point. his metalIOQraphic studies showed that the ground mass which would correspond to the eu— tectic, although fine grained, failed to have the char— acteristic eutectic structure. Shepherd (5), in 1904, presented the first complete metallOgraphic study of the copper—zinc alloys. He also gave a complete thermal equilibrium diagram based on thermal and microscopic evidence. The bounding lines of Alpha and Beta were determined by annealing samples at various temperatures and quenching. Bis liquidus and peritectic lines, determined by cooling curves, agree w». 1 1. defl‘Jb .a1.| Hmt. u.” if .. Emir-2 .I... Figure 1 Co .HJ uper-Zinc Freezing Point Curve as Determined by Roberts-Lusten . A if--- Q Q\ 9N Quate a‘kkflukwfl a. a. 9». a Q K. .33wa w I _ b t 5. nwv Figure 2 Copper-.7: inc Thermal Equilibrium fl . hiagram by Shepherd 5.5.-.. ‘1 I. QNQQQQ NQVK§NQQNQ km. 9n. 3. 9% R g. an no «u l,» 3’ , r ll II : fir 6x §S QR. QQV Q90 QQQ 8k 8% 89 8% with the present diagrams (figure 8). Shepherd, however, was unable to find the line e"e' and states in his paper: "We are not able to verify his (Roberts—Austen’s) observations as to the existence of the line e"e'. We made repeated records of the cooling curves of alloys ranging in composition between 50 and 75 percent of copper without finding such a heat change in any case... we have also made tests of ingots above and below 4700 and quenching... It follows therefore that the curve e"e' is due to experimental error, possib- ly to the sticking of the pyrometer record." In 1906 Guillet verified the work of Shepherd, and was also unable to find the line e"e' (4). Tafel in 1908 had the same results (5). He too prescribed Roberts-Aus tin's findings (but more reluctantly) to experimental error. Tafel also located the peritectic points, using Tammann's method of thermal analysis, and these points are the same as used in the present diagram. Carpenter and Edwards (6), in 1911, found heat effects in alloys from 40 to 65 percent COpper at about 4700 by accurate pyrometrical work. In explaining this heat effect, their first hypothesis was that there was an allotropic change from.Beta to Beta' (figure 5). According to their second hypothesis, they assumed that the bounding lines of Beta met at 4700 and at this point the Beta transformed to Alpha plus Gamma. Figure 4 )I‘ W «Va? .1 4. o I v; ‘ Figure 3 carpenter and Edwards' First Hypothesis Transformation Beta to Beta‘ Sgt» NQVK§N9§NK & a. 8 DR 8 S3 50 m9. 1% eN$ w. To 1......) : _ p. b. _ 5.9» a s3 -- 69¢. be QQN QQQ Qfim QQQ . Q9 . Figure 4 Carpenter and Edwards' Second Hypothesis Transformation Beta to Alpha plus Gamir It. \ fixing! -I||.«r_+usi _ I. i m. 1-11 1... a 8. _ l _ f 3 5.3 .05» B 3/1 .. .i .. -- J. -. F/ I. /,1./ 7... r ' AIvILrlvo O LIL TL! n b -9- shows the diagram based on this assumption. After exam- ination of aged specimans of Beta brass, "the authors felt no doubt that in reality it consisted of two con— stituents." The next year Carpenter presented a paper which described his work in which he attempted to coalesce the Alpha plus Gamma in samples of Beta Brass (7). Samples he annealed below the conversion temperature for six weeks showed no change in structure. He then repeated the ex- periment with two alloys, one having a slight excess of Alpha and the other having a slight excess of Gamma. After annealing, two apparently new constituents appeared resembling Alpha plus Gamma. When these were reannealed at the conversion temperature, the Beta structure return- ed. From.this data he concluded Beta to be Alpha plus Gamma. Hudson in 1914 (8) disputed Carpenter's theory and contended that the heat effect was due to a polymorphic transformation. He annealed a heterogeneous sample con- taining Alpha plus Gamma at temperatures below the crit- ical point and produced a constituent that was neither .Alpha nor Gamma which he called Beta'. This method did not necessitate assuming a tardy coalescence of two sepa— :rate parts of Beta and proved that Beta' was a stable phase. X-Ray crystal analysis also fails to show the pre- sence of the Alpha constituent in Beta brass so it can— -10- not be a mixture of Alpha plus Gamma. Bain (9), in a aper presented in 1925 states: "The alloys in the nar— '0 row range aroung equal atomic proportions can be slowly cooled to room.temperature, giving what appears under the microscope as homogeneous grains of Beta brass. The X—Ray study shows that the large grains thus produced are truly of a single lattice.” He found the structure to be complimentary simple cubes of a 1:1 solid solution in the body-centered cubic arrangement. Owen and Preston (10) found by X-Ray analysis that quenched and annealed specimans show no difference in lattice structure. All other X—Ray work on Beta brass gave the same results and seems to prove that a trans- formation to Alpha plus Gamma could not be. iatsuda (11), in his work on brass, contended that the transformation was due to an internal energy change and not to a phase change. He also stated that the Beta transformation could not be suppressed by ordinary quenching. In 1926 Gayler (12) considered the Beta transform: ation. He "... inserted.. and arbitrary dotted curve, vfllich is rendered necessary as the lower boundary of a invo phase field for Beta plus Beta'." This is necessary ill order that the phase rule apply to the diagram. At the sunne time Houghton and Griffiths determined the critical jpoints by electrical resistance. They also inserted an -11- arbitrary line for the Beta plus Beta' region (15). Sykes and Wilkinson determined the boundaries ahd criticals by specific heats. They also made quantitative determinations on the specific heats at the Beta Trans- formation (14). Figures 5 and 6 give the copper-zinc thermal equi- 1ibrium.diagram.as it is accepted today. It incorporates the most recent of the theorys given above and the work of several others (15). As far as could be determined from.the literature considered, the double line of the Beta plus Beta' region has never been found experimentally, but merely inserted as an arbitrary line in order that the phase rule might apply. (Dhnnbers in parenthesis refer to bibliography.) -12- Figure 5 hodern Copper-Zinc Thermal Equilibrium Diagram 26-0) 30 o 2.09.0...— o~ o- 0 cu . 2- on 2. ow On 91 0» 8 On 0* on 3 or oo oo com-N. 00¢ a _ 3 mn- »..Q .K .34— “nu .nn¢—, , mm... 3:. Sn Tllr é? con “ \ n . _ mm a o -: m. . O L w u. A .Q a 95:00 W “by NON. N 02. . as d 9.32 - 539V W San AN uh?— % a 3:9.- . o- / ~3- ooa 30 // bSO-v/ . Co A :2 000. 36.5 65:3 / u}... co.- ~.o~ -13.. Figure 6 Beta portion of Copper-Zinc thermal equilibrium diagram as accepted to- day. N Percem - Bu Vcigm _—1‘ \ \ - _— _ 'p+L\quid fl 1 )5 Gamma 70 4O 3‘ 3O 4O ,0 60 1 -14- Preparation of Samples Many different methods were tried for the prepar- ation of the brass samples but the only satisfactory met- hods were the ones described below; In these methods the composition of the final product could be predicted to a close degree and a minimum of impurities were introduced. The flux used in the melting was a 50:50 mixture of sodium.chloride and potassium chloride. This flux prevents oxidation at the surface and dissolves oxygen already in the metals. About twenty—five grams of the flux was used in making a 200 gram sample. Graphite crucibles were used to keep the absorption of impurities from this source at a minimum, The melting was carried out in the gas furnace and the temperatures were measured with an optical pyrometer. The Zinc used was first melted and cast into minia- ture ingots of about 100 grams. They were then filed to the desired weights and attached to the end of a rod. Be- fore using they were covered with a small piece of paper. This paper burned before the zinc melted and formed a layer of carbon dioxide in the crucible and aided in pre- vention of oxidation. The zinc loss by oxidation and volitalization using this method of preparation was about two percent. From -15... this data, the amount of zinc to be added to a given amount of copper could be calculated to produce the de- sired alloy. One-hundred grams of copper shot was used in the preparation of each alloy. The shot was placed in a graphite crucible and covered with the flux. The cru- cible was then placed in the gas furnace using the max- imum.heating rate. When the temperature reached 81000 F. the zinc was quickly plunged below the surface of the copper and stirred for several seconds after the melting had been completed without lowering the flame. The crucible was then covered and the furnace was shut off. In the second method the zinc loss was held to less than one percent and the desired final composition could be accurately obtained. Several samples of high zinc brass were made first and their composition was deter- mined. The weight of a copper wire per foot was found and a piece of the correct weight to produce the desired composition was cut. The brass was then heated to just' above the melting point and the copper wire was stirred in. The same flux and crucibles were used for this method. In making the dilatometer specimans, the brass was cast into a preheated iron mold. The warm.mold produced solid castings with very little pipe. The approximate analysis of the brass samples was calculated from the known amount of materials used in the alloy, assuming all the loss to be zinc. -15- Two samples of Beta brass and one each of Alpha plus Beta and Beta plus Gamma were used in these tests. The approximate analysis (by weight percent) was as follows: Beta (a) 48 zinc 52 copper Beta (b) 46 zinc 54 copper Alpha plus Beta 44 zinc 56 copper Beta plus Gamma 58 zinc 48 copper Critical Point Determinations The furnace used in the critical point determination is shown in figure 7. A high external resistance was used so that the cooling rate could be varied. Plugs of insul- ating material were made for each end of the silica tube. Two thermocouples were used, one to record the temper- ature of the furnace and the other to record the speciman temperature. The speciman was made as nearly round as possible and was drilled so that the thermocouple could be placed inside. The furnace thermocouple was placed inside a piece of pure copper. Critical determinations were made on both of the samples of Beta brass. Several determinations were made Jith each sample varying the rate of cooling. In this manner it was possible to ascertain the effect of the cooling upon the critical point, both as to location and intensity. Thermocouples of very fine wire were used in order that they might be as sensitive as possible. -17- Figure 7 Critical Point Determination Apparatus I "7 “9-—- ‘ ‘ l3. ...!- “mt. wry-w," “ -18- Four determinations were made for each sample and readings were taken every minute. The results were as follows for Beta brass (a): Test_gne_ Test two Time Furnace Speciman Furnace Specimen Temp.OF Temp. OF Temn.OF Temp. OF 1 911 954 894 908 2 890 921 889 902 5 874 909 884 900 4 845 892 878 896 5 822 878 874 890 6 807 862 970 886 7 789 851 868 882 8' 772 858 861 878 9 759 820 859 874 10 758 809 856 870 11 851 866 12 848 861 15 844 860 14 842 859 15 859 858 16 857 855 17 855 851 18 852 849 19 850 847 go 828 844 81 88' 841 23 824 840 35 822 859 34 ’ 880 858 25 820 855 26 819 832 37 818 851 28 817 850 29 816 828 50 814 828 51 815 826 32 812 82 55 811 821 34 811 820 35 811 819 56 811 819 37 811 818 58 810 817 58 809 816 40 808 815 41 806 812 g 804 811 45 -19- Trial three Trial four Time Furnace Speciman Furnace Speciman Temp.OF Temp. OF Temp.OF Temp. OF 1 92 951 900 898 2 910 920 888 882 5 902 912 879 874 4 895 905 872 868 5 890 900 866 860 6 886 896 861 858 7 880 890 858 854 8 878 888 852 850 9 875 882 850 849 10 870 878 846 844 11 868 876 844 840 12 864 872 840 858 15 861 869 852 851 14 858 866 824 824 15 856 864 812 812 16 852 860 805 805 17 851 860 790 792 18 848 859 19 844 855 2 844 854 21 845 852 22 840 850 25 858 849 24 856 844 25 855 841 26 850 849 27 828 858 28 822 852 29 818 828 50 809 819 51 799 809 52 790 800 The results of these determinations have been plot- ted and are shown in fmgure 8. Since the furnace tempera- tures varied so much they were not used in plotting inverse rate curves as had been previously planned. Four tests were also made on the Beta brass (b) and the data has been tabulated below. The results of these determinations are shown in figure 9. Time Comx'JOiCflprfiZ‘Jl-J {OCDQOIUHP-Cflwl-J Test one Furnace Specimen Temn.OF_ Temp. OF 891 895 889 889 887 884 881 879 879 872 874 869 869 861 865 860 859 851 852 845 848 840 845 858 859 852 854 829 829 825 827 821 822 818 819 817 Test three 899 898 892 892 888 888 880 880 875 875 869 869 865 862 860 859 852 852 849 849 847 847 842 844 858 840 851 858 828 852 822 829 819 825 821 817 -20- mama Furnace Speciman Temp.0F Tfmgx (N? 895 889 891 882 888 880 885 872 879 866 875 860 868 852 862 848 859 841 852 858 848 850 842 828 858 820 852 816 828 811 Test four 891 883 878 874 871 857 862 858 855 852 849 846 842 841 857 835 831 829 825 821 898 890 885 879 877 871 858 865 859 857 852 851 850 848 842 840 858 834 852 829 From figure 8 it may be seen that the critical for the Beta brass (a) appears at about 8600 F. (4600 0.). The critical for the Beta brass (b) appears at about 8510 F. (4550 C.) as shown in figure 9. Increasing the rate of cooling tends to lower the temperature at which the critical appears in both samples, as is usually the case. however, there is not over 20 F. difference between the fastest and the slowest cooling rates used in these tests. By increasing the cooling rates, the intensity of the critical is greatly reduced. In spite of this, there is a definite deflection of the cooling curve using even the fastest cooling rate. The curve is similar to the general type to be ex- pected in the transformation of one solid solution to another. This is one point in favor of the theory that there is a transformation in brass from Beta to Beta'. Dilatometer Observations The instrument used for these tests was the Chevenard Dilatometer. One test was made with each of the two samples of Beta brass. Since the heat effect in the Beta trans— formation is so small, it was not shown in the heating curves, and these curves will not be included. However, with a very slow heating rate there should be a deforma- tion of the curve at this point. I.II|||| fill..- 14:53.1; Eb ltob.fiu‘| llll.l The data obtained from the dilatometer curves was as follows: (expansion in units per unit.) Heating Cooling Temperature Contractgon Temperature Contraction 0c. x 10-0 0c. x 10-5 Beta brass (a) 100 1.7 600 13.2 200 3.7 550 11.9 300 5.7 500 10.3 400 7.8 460 9.1 450 9.0 450 8.8 500 10.4 400 6.7 550 12.3 300 3.7 600 13.9 200 1.3 575 14.8 Beta brass (b) 20 0.0 500 15.2 100 1.8 550 13.7 200 3.8 520 12.8 300 6.4 500 12.3 400 9.3 480 11.6 420 9.9 460 10.8 440 10.4 440 10.0 460 10.9 420 9.0 480 11.5 400 8.6 500 12.0 300 5.4 520 13.6 200 3.1 550 14.4 100 1.0 600 15.8 20 -0.3 680 18.0 Photomicrographs of the Beta brass (b) are shown in figure 10. Samples were taken at both ends of the speci- man. The great difference in grain size between the tOp and the bottom of the piece was due to the different rates of solidification, since the piece was cast in a vertical position. All the photomicrographs of Beta brass in the next section of this report are of the Beta brass (a). Figure 10 Dilatometer Speciman Beta Brass (b) Annealed 900° F. 2 hours Slow Cooled (card 876 box B52)* Speciman (card 877 box B52) 50 X * Refer to speciman files ‘« --£' .1... 1?..- fir. : -26.- From the curve shown in figure 11, it may be seen that the transformation takes place over the range of 4400 to 4600 C. in the Beta brass (a). In the Beta brass (b), the temperature range is from 4550 to 4-500 C. as shown in figure 12. The curves obtained in this Beta transformation are also of the general type to be expected in a trans- formation from one solid solution to another. This is a very strong point in favor of the Beta to Beta' transformation. The coefficients of linear contraction have been calculated above, belowg and at the transformation tem- perature for both samples. This coefficient is accepted to mean the unit decrease in length per unit length for one degree drop in temperature. The results are as follows: Temperature units/ unit range 00;, Contractiog Beta Brass (a) gwjb) 550 - 600 0.000026 0.000050 450 - 460 0.000050 440 - 460 0.000040 200 - 500 0.000024 0.000085 From the data it may be seen that the coef- ficient of contraction is greater over the critical range in both cases. Also at the range above the critical, the coffecient is greater than over the range selected below the critical. ”11 J T‘) ..‘J Y ' V V ' Y ,V ‘ FY ‘—'— V .. .1 > p - 1 I ' - t ' f' ‘ ‘ ‘ ‘ ‘ ‘ ‘ _ . _1_ .. .. .. ,... . .... . . .. . 1.. . T . . .. .. . . . .--A . . . . . . " ‘._._,._ ,‘, , _ ..,,_ _, , , _ . , . .. , '.. . . .- . . . .... .. ,. .. . . . _:~o"-"““°"’.""92“-""""‘"1‘-"-.—.—--O—f'v"--‘.-‘0-- -----.... ,..._-_.-_._._.-. ->-.—o--—<-- --o-.-.—-.-4>--o—o—~¢.— - -.-_-..._--..- -‘--—¢—ov¢—‘1vo--ooo»-- ". . ... , , .: . | * . .... . . - . . . l . . c .. '.. . .. .. . ..,. ~$. .c p . ..o -. . - t ~ . ~ g.. . . , I ,. .. ‘ . ... . § . I ..--. .... .. ..1 .... ,. .. . o... o . ..:. . . . . . . - . - , . . . ~ - . —l A - H L g v - g v . I: ’ -‘o n . . . T "O. . . . . . I . . ' . , . _ .. ... . I. . . ‘- 27)» \J 4 bl } - .» t - v 'v v- ‘- 1 . , 1 . . . . . . . . . . . . ) . . g . I ‘ o—o-«oo-—-o*~—--1-~--ooo--- ---—--o—-o-- -~-- ---9-~--q:-~v- o .— 1L---.-....$-.-—‘-o-- ... ebe ......- -.—-AO—-Q‘-‘-‘yo-c—4O~O'4 ' - -e--- »- ( - - 7 - -..-...... ---v-0-*~- -.--9"—’"' "-"‘**" "'°‘°f‘"*‘—*r*""'""°‘ ... . , . Y. . . .. . .. ~.,; ,. t . . .1 i. - ~ I P Q _, ,, , . . .5 . .. . . .... 0 ~ A - - b .011 raCt On ' ; ' L 1 ; 3 f ‘ 1' ' " ' ‘ V V V . ""‘—?""‘ . g . . . . , . c -' u (_ - - ‘ v " ' U'n s--/‘unit ' , " r ' * t ' ' ' " ‘ 1‘ ' o .| .. . . . . . l . a ‘ . l . . . . . . . ‘. . , .. ‘3 ‘. u p u . y . t . 4 l I t y ----- >O~v--< —< - .*-¢-.H.-1p-....--.. .--...... .--...... --.-c--—. ~-a—oo---- ..----.. .-..-. t-v—avov V--- ~-Ov--v -—~-y~-’.— w———<-+-v-‘.— ---, -... --. --..,‘-- - . -...*---.-1y . -..... T . . .. . . . . ..-. . . . -. .. ...—r4. . ~—.* .- . - --.._...: o- - ~ r - _-- v. o ‘w- r - ‘ . - 9 - ‘ ‘ ‘—-—-—*" - ' ‘ ‘*)—‘ ' ' ' ' " ‘ ‘ '4 - . o .5 .I . p . s . ..u o . 0 ~ I 1- - .... ... . . . .. . i . . . -. ‘ . ~ - v ‘ v ... .... .. .. .. ... . ,. . . . O - -- . - f - ' -! " ' ' - . . . . . . . . . n . a in ‘ L5 1 7 - . . . 1 ... f .. . ; - n.. -- -- - .. . .. . .. .. . . A . . - .. .( .. l .. I . . - H"'QW0-o.——-o<-o- -_._._¢..-. ..--..._-, -.- 1 ‘ I O: . .. . . ' ~ . ' o . -.... . o... 3 l . --. _- 0-»“ n o 1 - .u . 94 ~- T .1qu . , . , , or- -1 - . .. 1 y . , .. 3 l . .., . . .‘~--u --.) :~ c I .. l . u c ‘ y- .. . . . . . . A .. . . F’*--f~'-o -o»~~——v—o~-oo-<-—-.-.-o-W-4rwv.-o~-o- ...-_...-.-.l.-.-.~—-- -»- .—~ -- u u .. . . A . . . . . . . 1 . . 1 1 . -A . ..... . . ... ..‘ . s g . - . - .. . u I . - 1 i ‘ .n . »- n o . ‘ . "~ - -.. o. . .. .. . . .. . . . --. .. . .. .. .. . . .. .. . . q . ‘— v . . .. ,. .. i . . . .. i H~~o+-o—o~-1-o----o-oau >QOO--+—OH<—‘va~-.OV‘- --..qo.or-4~-.vo‘.--»{~-o~-e 4 ' ' v - ix - -- 0. -~ - . t - O ~ -¢ . . . ;.. . .. .. , . . . ,. . . . . A Ll ' o O - . U , , ,, , - - v - . . . .... .. . . . . ,.. ... .. .....;... .... . ... ., . 4- - - . . ... . . . : v ""‘9°"‘-— .+>-Orv§-' o-.-—4—----—-4~-o . o~ .. 9 .. .... . .. . . . 0~ - .. ,r.. . .... ; §. . “a A A A - T+'_ r - I - .. . . p . .8 . . ... . »--.. . .. . .. .). .. . ,. . '--~~ o - ... . ... .. . . 6 ‘- «.... :.... . .. .. - c . .. ["O-vvv-o-o—o— -Ho...>---- ..--.-.._..-. ------.-- «..---...a. _-._.._..--‘ ..—.. V -9- .. . . , . , ..., . .. " 7- .- ? ... . .. . n .. V . A.. 1.- ' ‘v - -‘- I ... . ... . . 1 . . 1 .-l. .. .... .....fA . .. . ., . . A L A , 1 v v .. . . .. .. . .. . . . . ,. . . A . . .... ._ '-. . . ,. . .. . . ....... . -o... .. .-- .. ..... . .. .. , ... ~--evrco---—4)--—o~—o—o—-~‘>—so..—+.—~o-—(L-----¢-~o—o-4-~o--§.o-» ...—-+-.».—-}.-. '“4 - «'v- .... . . .. , . . . A ' ~«-‘- ~ .u »--q . . Q . o.‘ ..- .. .. .- ......... ..; . .... . .... . . .. . .. .. . .... ... . . a . . ' . ¢ . - . .. .. f ... .... .. ’l ...., . ' 0- .- . -.t . '. . .. .'... .T . . . -. ... .. .o -. . .... .. .-. . -4 .. . >v-..l i . . .1; . .5 . .l.. . .. ' >~§+oL-~-~-F-4 o—.—o—-¢--oo~<)—ooc—o;-o—.—..‘vo-~-»--‘-<-—~--o-J-ooa‘--— -«—-4 -.-- ‘; .4 ... A .. .- -. ... . "Y . ... . i. , . . - f »‘ . "v .- " ... ..I. . .. .. . . A.... . . o...- ' .. ‘ 9‘ ' " ' "‘| " 't " ‘ ' " - ... . .. . . . )... I. .A ,. 7' . .. 1 .. . . . . . g .. .. .. .. .-o -- a.‘ -- o . . .. . . ~ ‘ > ’ ‘ ' " ‘ ' ' r; - t - . . : .. . ....Ca. . .. .. ;. . . . ..-- . --.-. -- - ... ... . . .. . .. ... . .. . . . . . - -- -- v - ‘ ‘ " ' ~--«: . ...6--.- .. ' . 1.. -. .... .. ,. . ' f ‘ ' ' ‘ ‘ ' ‘ ' ’ ' ...... .. (... .... .. . . .. . a . ‘ 1 ' - 9 ’ ' ' >-—~-»o—<~.—- -... -... o—o—H—fu-O—- ~ 0 . 0—-—?—--o—o 4 . 87. o 4 o .-- ).« on- --. o --.—.J,_- .—< . . . . -— " ‘ ' ‘* ' “‘ ‘ ' ‘ ‘ ' "" "—‘ “ ‘—“t‘ ‘ ' ‘ ' ° “" "" ‘4 "' ‘ "' '—' ' 7 .. y.... .. . -. ....... , .. 1| .... . . . ' ‘ '~~ ' f ' ‘ .,. ..-l --. ... 9... .. , . .. _ ,. , . . - . .....«. .. .. o- - v '-.. -..q¢--~ ..i .. + ‘q . . . . ‘ ' ' ""+‘ ' ' h' ' ..... ...)... _.,- .... , _[ . . A “-1. t - . : i ‘T ‘ ‘ 7T7 Y 'L '4 '. . _ _'_ ..,,_ '_ _‘ . 3 . I , . I . .. ..... .. .. o . . . A. __A. _ . "l _ p ‘ - ... ___ . 4 ,. . i. ...4... . . o ...... . .__. _. | . ‘ I ‘ . . . .. 3 .. ,.: . . - t 7 ..... , ....o . - ’- . .. . . . . . ,, o ‘ - ---O- ‘ ‘ b-fi_.--+,. - - ‘WHQVQ-O.‘Q—o—-y—o—~—H+~ou-Oo p o-o-o—-- -—-~ 0-. ~ .—1 L - -.-. . o . - 4 .—.—.—--; -. . d \s‘ \‘L < - .—-. 4 . - v . ¢ . .-- L4 - >4 - - -- - r ---O ~ " "‘ Y " -‘ 0 "-‘"-.“‘ ‘ ‘ ‘ """“-‘ L“ ' " ‘ ' " ' "' r . . A..- .1 . p . . Q ~ 3 ‘ t . -- . . -o v f . . .. . . | . .... . p . - u -- q 0 > “ “ -- ' ‘ " " “ "" o --» . . . . .. . . . . .. .. .... ...... _ ‘.I ‘ ‘ _ . _ _ . . . . .. ... a. . . . . ... o "4" ‘ t "v (3 - ‘ 1k 1' f - ~ ~18LIP€THL UTE :égl‘éés h. 7‘ ‘ ' "‘ “ " ' . . .. . . | . . 5 . . . . . U - “ ‘ ' " ' ' ' " _ _ . '- . ..,.. , 7 , ., . . .. .. :. . .-. ~ I - .. .... .. .... .. , I . . . . -. -- 1‘ . ~ 9 ' > > " ' ' . .’ e’O-f--’~>--+-~r>‘"“*--e~~7--" _‘-',__-_{ ‘-_‘_-.‘ -~‘--.'""4*""“'*"“""“’ ._ .--.f-_...-.4;_.-_.-f-’-‘-r<--..-.-..--4 ,_.,. ....I-- o . . .. . 1 . ... r , .. . . . . . ..... A . . .. - . . ..o.-.. [..-- ‘ " "9 " -~; " ' a ‘ ' 'f" ' ‘A' I . 'h ' l 9". ” I . ... . ,,;..- . , . . .. . . 9 . . . . - 1 - fl . .. . . ..» A - ‘ ‘ . _. . -29.- metallographic Studies In the first series of experiments it was attempted to prove that the Beta transformation could be suppr ss— ed by ordinary quenching. An alloy containing 56 percent copper was prepared and the annealed, slow cooled structure is shown in fig— ure 15. This alloy was then heated to a temperature of 14000 F. for one hour and quenched in brine at 520 F. The speciman was cut in two pieces and the first piece was examined as quenched. The second piece was annealed below the critical temperature (8200 F.) for one hour and slow cooled. The structure of the quenched Alpha plus Beta brass is shown in figure 14. The photomicrograph shows that the Alpha constituent has been held in solution by the apid cool and the piece is now in solid solution. Figure H 15 shows the reannealed structure and the Alpha constit— uent has been precipitated. A sample of Beta plus Gamma brass (48 percent copper) imis then prepared and annealed at 15000 F. for one hour. 12fter quenching in brine at 520 F., the sample was divid— ed into two parts. The first part was examined as quench- emi and the second part was reannealed below the critical ixmnperature (8600 F.) for one hour as with the Alpha plus Beta samples. Figure 13 Alpha plus Beta Brass Cold Worked Annealed 820° F. 1 hour Slow Cool (card 373 box B49) 50 K -31- Figure 14 Alpha plus Peta Press Annealed 14000 F. 1 hour Quenched Ice Brine 50 X (card 874 box EbO) Figure lb Alpha plus Beta Press Above Sample Reannealefi 8200 F. 1 hour Slow Cool 50 X (card 875 box P51) “1k The structure of the quenched speciman shows that the Gamma constituent has been held in solution (figure 16). By reannealing, the Gamma constituent has been pre- cipitated as shown in figure 17. ‘ The above experiments prove by inference that the Beta transformation can be withheld by an ordinary quench. If it would be impossible to suppress this trans- formation, it would also be impossible to hold the Alpha and Gamma constituents in solution as has been done in the above tests. In passing through the criticals, if the transformation did take place, some of the Alpha or Gamma constituent would be precipitated. It follows, therefore, that the photomicrOgraphs (figures 14 and 16) represent the true Beta structure. The next work was done with samples containing no- thing but the Beta constituent. All the experiments in this series were carried out with the Beta brass (a). Figure 18 shows the Beta' structure after cold ‘work and annealing at 9500 F. for three hours. Hot work and annealing gave the identical structure with a pos— sible slight difference in grain size. From.this photo- Inicrograph and others of the annealed structure it may. be seen that the Beta brass shows no tendency to twin, has lost the characteristic copper structure, and the copper does not influence the Beta structure. A sample of the Beta brass heated above the inversion temperature (9000 F.) and quenched in brine is shown in Figure 16 Beta plus Gamma Brass Annealed 15000 F. 1 hour Quenched Ice Brine 50 X (card 864 box B42) Figure 17 Beta plus Gamma Brass Above Sample Reannealed 860° F. 1 hour Slow Cool 50 X (card 865 box B43) r- figure 19. Since it is possible to suppress the Beta transformation, this must be the true Bet structure. It may be seen, however, that there is very little dif- ference between the Beta and the Beta' structure. The brass was badly cracked by the quench along the grain boundaries as shown in the photomicrograph. Figure 20 shows a sample of the Beta brass that had been annealed slightly above the critical temperature (9200 F.) for 24 hours and then annealed slightly below the critical temperature (8200 F.) for 84 hours. This represents the true Beta' structure, but shows very little difference in structure. A sample of the Beta brass was Annealed at 14500 F. for six hours and slow cooled. In this case (figure 2 ) the only difference was in the grain size. Since there was very little apparent difference in grain structure, it was thought that there might be a difference in the texture of the grains. Photomicrographs of the annealed structure are shown in figure 93 and of the quenched structure in figure 22. A slight difference in texture is shown by these illustrations, but it is possible that this is due to the polish given the pieces and this data cannot be accepted as conclusive. -35- Figure 13 Beta Brass (a) Cold Worked Annealed 950° F. 1 hour Slow Cool 50 X (card 866 box B44) Figure 19 Beta Brass (3) Annealed 9500 F. 1 hour Quenched Ice Brine 50 X (card 867 box 545) Figure 21 Beta Brass (a) Annealed 14500 F. 6 hours Slow Cool 50 X ‘(Card 839 box 347) Figure 20 Beta Brass (a) Annealed 2200 F. 24 hours 8200 F. 24 hours Slow Cool 50 K (card 8‘3 box B4B) _- .. 37. Figure 22 {-1 Beta Brass (a) (Figure 19) 500 K Figure 23 Beta' Brass (a) (Figure 20) 500 X ‘5‘ (3 -58- SUMMARY OF RESULTS 1. From the cooling curves (both dilatometer and critical determination) it is apparent that there is a transformation in Beta brass from Beta to Beta'. 2. The Beta transformation can be suppressed by an ordinary quench. 3. The Beta critical temperature will show up at any reasonable rate of cooling, but it is best determined using a very slow 0001. Rapid cooling tends to lower the critical temperature slightly. 4. There is very little difference between the structures of the Beta and the Beta' brass. 5. Beta brass is not affected by copper in struc- ture and shows no tendency to twin, even when severely 'worked and annealed. SUGGESTIONS FOR FUTURE WORK 1. An interesting experiment would be to determine the effects of long ageing upon samples of Beta brass. It was noted by Carpenter (6 and 7) that Beta brasswire became very brittle after long ageing. He believed this to be due to a transformation to Alpha plus Gamma which was probably untrue. 2. It could be that there is a difference in grain texture in Beta and Beta' brass and this could be reveal— ed by high magnification photomicrographs. However, the polishing and etching would have to be very uniform for conclusive work of this type. 3. It would be possible to determine the properties of quenched samples of Alpha plus Beta and Beta plus Gamma brass with respect to the precipitation of the second constituent. Also determine whether or not the Alpha or Gamma constituent will precipitate at low tem- peratures on ageing. 4. Since the double line of the Beta to Beta' 'transformation has not been determined experimentally, it should be done. It would be possible to determine 'this double line by careful dilatometer work. It would .also be interesting to study the transformation care- :fully with the dilatometer. -40- 5. Very little is known about the physical proper- ties of quenched Beta brass. It may be possible that the properties of Beta and Beta' are very different and desireable properties might be obtained by quenching and drawing Beta brass. (l) (10) -41- BIBLIOGRAPHY Roberts—Austen, Transactiops’gf the Institute of Mechanical Engineers, parts 1-2, 1897, page 51. .-..— ...—..— .—o.->-_—. Charpy, Bulletin de la Soc. d{§ncouragement II, 1897, as quoted in Hoyt, Metallography, Part II, 1st edition, 1921. Shepherd, Journal of Physical Chemistry, volume 8, 1904, page 421. Guillet, Rev. de Hetallurgie, volume 5, 1906, page 245, as quoted in Hoyt, Metallography, part II, 1st edition, 1921. Tafel, Metallurgie, volume 5, 1908, page 542, as quoted in Hoyt, Hetallography, part II, 1921. Carpenter and Edwards, Journal of the Ipstitute of ygtals, volume 5, 1911. Carpenter, Journal of the Institute of MEtals, page 70, volume 7, 1912. Hudson, Journal of the Institute of Metals, page 89, number 2, 1914. Bain, Chemical andfiMetallurgical Engineering, page 21, volume 28, 1925. Owen and Preston, Procedings of_the Physics Sogiety (London), volume 56, page 1, 1925, as quoted in Houghton and Griffiths, Journal of the Institute of Metalsj volume 54, 1925. (ll) Matsuda, Sci. Rep. Tohoku Imp. Univ., page 4, volume 11, as quoted in Houghton and Griffiths, Journal of the Institute of metals, 1925. Gayler, Journal of the Institute of_§§tal§, page 245 volume 54, 1925. Houghton and Griffiths, Journal of the Institute of Metals , page 245, volume 34, 1925. Sykes and Wilkinsdn, Journal of the Institute of Hetgls, volume 61, 1957. Metals Handbook, American Society of Metals, 1936. Batchelor, The Transformation Point at 4700 C. ig Beta Bgassj Thesis for the degree of M.S., Michigan State College, 1932. *d nrl V OF P <13 j HudL LITE v e INTRODUCTION "With moderately slow cooling, the decomposition of austenite containing 0.80 percent carbon results in the formation of a constituent in which the ferrite and the cementite occur as alternate thin 1(mellae. These lamel- 1ae refract light in a similar manner to the lines of a diffrastion grating, giving the constituent a pearly ap- pearance, from which it has received the name pearlite."* From the surface structure thus described, it would be possible that the pearlite be either thin plates or elongated cylinders. is far as could be determined, no one has ever presented positive experimental evidence in favor of either one of the two possibilities. Sauveur** and most other workers in this field be- lieved that the pearlite lamellae consisted of the two constituents in thin plates. However, in their writings, they gave no proof to support this theory. The purpose of this paper shall be, if possible, to obtain experimental proof of the structure of pearlite. * Hetals Handbook, American Society for Metals, 1959 ** Sauveur, The Uetallography and Heat Treatment of Iron and Steel, page 65, 4th edition, 1955. 1,. .ma- .# _ y -44- Pearlite is a mechanical mixture of ferrite and cementite (Fe30). With carbon having a molecular weight of 12 and iron a molecular weight of 56, one part of carbon in steel will produce {7) 33456 «I- 12 12 or 15 parts of cementite. Assuming the percentage of carbon in pearlite to be 0.85 percent, then it will contain 0.85 X 15 or 12.75 percent cementite and 100 - 12.75 or 87.25 percent ferrite, or am... 12.75 or 6.8 parts of ferrite by weight for each part of cemen- tite. If the area of the ferrite and the cementite in a photomicrograph of pearlite would approach this ratio, it would prove that the pearlite was made up of plates of cementite and ferrite. A sample of 0.80 percent carbon steel was prepared, etched with picral, and photographed at 1000 X (figure 24). This plate was then enlarged to 4000 X (figure 25) and to 12000 X (figure 26). The area of the ferrite and the cemen- ~45.- Figure 24 0.80 Percent Carbon Steel Picral Etch 1000 X -, .‘y 7“.._""'9 “ta-5“?" a.“ .-'" -46- Figure 25 Enlargement Plate Figure 24 0.80 Percent Carbon Steel 4000X 47- Figure 26 Enlargement Flate Figure 24 0.80 Percent Uarbon Steel 12000 X tite was measured using cross section paper and the ratio was Calculated as follows: Figure 25 Area of ferrite— 4800 Area of Cementite- 1400 4800 1400 = 5.4 parts of ferrite Figure 26 Area of ferrite- 4800 Area of Cementite- 1511 4800 . 1511 = 5.2 parts of ferrite Since the above is calculated by area and the ratio had been calculated by weights, it may be that the density of the ferrite is not the same as the density of the ce— mentite and, therefore, the fact that the ration do not agree neither proves nor disproves the theory that pearl- ite is composed of plates of ferrite. The dark areas around the laminations in the pearl- ite are probably shadows between the ferrite and the cementite. Since we are unable to determine whether this part is cementite or ferrite, the mean area was used in the above calculations. While examining a pearlitic steel at a curved edge, it was noted that by changing the focus of the microscope it was possible to follow the pearlite laminations around the edge of the piece. This is very difficult to photo- graph but may be easily seen with the eye at about 1000 X. Figure 27 shows a series of three photomicrOgraphs taken -49.. Figure 27 Structure of Pearlite around Curved Edge ZOOO'X ° Azi (( -50.. around the edge of a speciman at about 2 00 X. Although it is possible that the cementite may have been smeared over the edge, it is improbable that this is the case since the laminations could be followed for quite a distance around the edge. In View of the above evidence, we may ass me for the present that pearlite is made up of alternate plates of ferrite in cementite, and not elongated cylinders of ferrite. -11. ,m -51.. SUHKARY OF RESULTS It is very probable that the structure of pearlite is alternate plates of ferrite in cementite. SUGGES”IONS FOR F1 UPE WORK; 1. Sauveur states that the ratio of the area of the cementite to the ferrite more closely approaches the ratio of l to 6.8 when the ferrite is etched with sodium pic- rate. By measuring the areas of a large number of samples, it may be possible to obtain this ratio. 2. By cooling a sample of pearlite for a long period through the critical range, it may be possible to obtain large enough laminations so that a good photograph may be taken of the curved edge. ' --A x r. -.. 16" 4* ,i ‘C'KIJI .-r‘. Iii-r" .. . . ... . . in? no: \ ... . s . .... < 0‘. HICHIGQN STRTE UNIV. LIBRQRIES I III III ||1l||||||l ll 312930059 2872