E: ,_ t :5; ¢ ,u-a ?! 0 s it ,‘ ' b ”v . ~- a?“ '0 .3! I 0 5"; h. . - -'. n V. i". '4' as ‘ I... 1" .. u a...” “HESS LIBRARY Michigan State University ll!|l!l'lilll‘|ll1l|l This is to certify that the thesis entitled THE EFWCT OF LATE ADDITION OF GRAPEIITE ON THE PROPERTIES AND I‘TICROSTRUCTURES OF INDUCTION FURNACE, HIGH STRENGTH CAST IRON presented by Carl L. Langenberg has been accepted towards fulfillment of the requirements for degree in MAS TEE OF SCIENCE ,. ,1 I,:'4 /. t. /I Ad/ , ' ..-/ - , _ r / 4 ‘ " , - ': /;'3°*'i-'~"2’I"84/ \ Liz/g; 7. .g 5.. r 7 Major professor Date (I: 5” ,/(/, ./;/\5/ 0-169 THE EFFECT CF LATE ADDITION OF GRAPHITE ON THE PROPERTIES AND LICRCSTRUCTURE CF INDUCTION FURNACE, HIGH STREQGTH CAST IRON By CARL LOUIS LANGENBERG A THESIS Submitted to the College of Engineering Michigan State University of Agriculture and Applied Science in partial fulfillment of the requirements for the degree of MASTER OF SCIENCE Department of Metallurgical Engineering 1957 ABSTRACT A study was made of the effectiveness of graphite as an inoculant for hypoeutectic gray cast iron. Iron was melted in an induction furnace. Some heats were in— oculated with 0.47 per cent graphite. These were matched in chemical composition by untreated heats. Transverse bars, chill test bars, quenched wedges, slow cooled wedges and bars containing thermocouples were poured. Comparative data on chemical composition, transverse strength, deflection, graphite distribution, cell size, chilling tendency, centers of eutectic crystallization, and cooling curves are included for untreated and graphite inoculated heats. The following conclusions are made: (1) A carbon increase ranging from 0.22 to 0.36 per cent can be achieved by ladle addition of 0.47 per cent graphite during tapping. C2) Inoculating with graphite causes a marked reduction in chill. (5) Graphite inoculation causes the formation of a large percentage of Type A graphite and materially increases transverse strength and deflection. (4) A slight decrease in cell size is achieved by treating with graphite and graphite treatment nucleates many small centers of eutectic solidification. (5) The initial temperature of the eutectic reaction is raised and the duration of the reaction is increased by graphite inoculation. ,/ 67 Approved, Major Professor ii ACKNCfiLEDGEMENT The author wishes to gratefully acknowledge the encouragement and aid extended by Dr. R. L. Homochel and Dr. D. D. McGrady during the course cf this study. The counsel of Dr. A. J. Smith is also appreciated. Assistance contributed by hr. B. D. Curtis is also acknowledged. iii II. III. IV. VI. INTRODUCTION.... FXPERILIEITAL PROCEDURE. ..... . . . . . . . . . TABLE OF CONTENTS DATA AND RESULTS.................... A. Chemical Composition. ............... . Mechanical Preperties....... B C. Structure..... D Chilling Tendency.......... E‘J‘ DISCUSSION..... COI‘ICL-USIOIJSOO0.0.000....OOOOOOOOOOOOOO BIBLIOGRAPHY... Cooling Curves...... ....... iv Page 12 12 15 17 28 59 47 48 LIST OF TABLES Chemical Compositions of Experimental Heats" .. Mechanical Troperties......................... Structure of Transverse Bars.................. Structure of Quenched Wedges.. ...... ........... Chilling Tendencies. ...... ... ...... .... ..... ... Cooling Curve Data ..... .... ..... ... ...... ...... Comparison of Effects of Gra)hite and Calcium. _ l 24 29 55 45 Figure Page 1. American Foundrymen's Society Chart for Graphite Distribution in Cast Iron ............ 2A 2. Graphite Distribution at Centers of Transverse Bars..................... ..... ..... 18 5. Graphite Distribution Near Surface of Trans- verse Bars.. ...... .......... ......... .. ...... . 19 4. Eutectic Cell Size Chart..... ............... .. 21 5. Photomicrographs Showing Cell Sizes of Untreated and Graphite Inoculated Irons. ..... . 22 6. Centers of Eutectic Crystallization in Untreated and Graphite Inoculated Wedges Quenched During Solidification ..... ........... 25 7. Graphite Flakes in Centers of Crystallization of Quenched fledges..... .......... . ...... ...... 28 8. Graphite Distribution in Slow Cooled Wedges... 27 9. Untreated Chill Test Wedges.... ............. .. 50 10. Graphite Inoculated Chill Test Wedges. ..... ... 51 11. Untreated Chill Test Blocks.... ............ ... 52 12. Graphite Inoculated Chill Test Blocks. ....... . 55 15. Cooling Curves for Untreated Irons............ 56 14. Cooling Curves for Graphite Inoculated Irons.. 57 15. Comparison of Cooling Curves....... ......... .. 38 16. Ledeburite Adjacent to Very Fine Graphite Adjacent to Normal Graphite ............ ....... 45 vi I. INTRODUCTION Cast iron is basically an alloy of iron, carbon, and silicon. Because it is relatively inexpensive, easy to melt and cast, readily machineable, quite wear resistant, relatively resistant to corrosion, relatively insensitive to notches, and has high damping capacity, cast iron occu- pies a position of wide application and great usefulness. The microstructure of cast iron determines its mechani- cal properties and hence its usefulness in various appli- cations. Therefore, accurate control of microstructure is essential in the production of high quality cast iron. There are many useful types of cast iron: white, gray, chilled, mottled, malleable, nodular, and alloy. This work will be devoted mainly to gray cast iron which is a relatively complicated alloy. Flinn and Reese (11) state: "To prepare a series of irons of all important com~ binations of carbon, silicon, manganese, phosphorus, and sulfur would require over a half million heats and literally millions of tests would be needed to investigate the en- gineering properties of these irons. Even with this colos- sal amount of work, the effects of alloying elements, and of variations in metal processing and handling, would still remain to be studied." Gray cast iron has been described simply as graphite 1 2 in steel. Actually the microstructure of cast iron can be divided into two principal parts, graphite and matrix. The matrix can consist of any of the following constituents or combinations of them: ferrite, pearlite, cementite, stead- ite, martensite, and austenite. The character of the matrix plays a part in determining mechanical prOperties. However, this is overshadowed by the greater influence exerted on mechanical properties by the character of the graphite flakes. Distribution, orientation, size, and amount of graphite flakes are all factors in the determination of mechanical properties. This effect is also amplified by the fact that inferior matrices usually accompany undesir— able (abnormal) graphite, while matrices exhibiting super- ior properties generally accompany desirable (normal) graphite. Five types of graphite flakes have been classified jointly by the American Foundrymen's Society and the American Society for Testing Materials. These are shown in Figure 1. For most applications high quality cast iron is regarded as consisting of type A graphite in a pearlite matrix. Many factors play a part in the determination of cast iron microstructure. These are: chemical composition, degree of super heat, time at super heat temperature, pouring temperature, cooling rate, and inoculation. Ladle treatment with any of a number of inoculants tends to (bl /d/ o-’ a. I o v I. .‘ggi “ ‘l r"; .\ " . . . f t “3.... .-.11 {N c .4 ads? R'l- ’g.)-I'. I ’ .9 t '..r. W . 'r‘f‘ . "" ' ' " .1- ID. . ‘ umb‘s ‘. e . ‘ \,“~. " I mdxl\ ...5.‘ "3’ i’ {1" ‘ J, R“: )5" b . 5 - ”If? 9,," ~.' t 9 ‘ 4 .-l .' \.’ .. ‘fi- 3 .0" (e) FIG. 8-9. A.F.A.-A.S.'l'.M. Graphite Flake Types in Gray (last lron. 100/. (a) Type A, L'niform distribution. random orientation. (1)) 'l'ypc 3. Rosette groupings, random orientation. (r) Type (I. Superimposed flake sizes, random orientation. ((1) Type 1). lnterdendritic segregation, random orientation. (e) Type 1‘), lnterdendritic segregation, preferred orientation. (All specimens unett'hed.) Figure 1. American r.undryuen's Sveiet f C art for Grapl‘tite Dis tributi tn in Cast iron. 5 bring about a more desirable structure than can otherwise be achieved with iron of a given chemical composition. Many inoculants have been used in the past and are being used in attempts to improve the structure of cast iron. Some of these have proven to be effective while others have not. This work was done in an effort to evaluate the ef— fectiveness of graphite as an inoculant. Graphite is be— ing used to some extent commercially for the purpose of reducing chilling tendency. Some previous work has been done regarding the use of graphite as an inoculant. Dierker (6) added approximately 0.11 per cent carbon in the form of graphite to cupola irons and electric furnace irons. He found that tensile properties of the cupola irons were not materially affected. In the case of electric furnace irons, the transverse breaking load was increased by 100 to 400 pounds and the deflection was increased somewhat. Finer cell size and less chill were also observed. Chemical analyses were not reported so observed effects could have been at least par- tially due merely to an increase in carbon content because of graphite addition and not necessarily to an inoculating effect. Womochel (4) observed that an addition of 0.44 per cent graphite to an arc—furnace hypoeutectic iron at the furnace Spout had a negligible effect on the transverse properties. Some decrease in chill was noted however. 4 Dahlberg (5) reported a reduction in chill achieved by inoculating with approximately .10 per cent Mexican graphite. An increase in transverse strength was also ob— served. Various graphite particle sizes were used. How- ever, the determination of their relative effects was inconclusive in the opinion of the present writer. Also, due to difficulties in temperature measurement, inoculation was done at various temperatures. This may have affected the results. It was pointed out by Norbury and Morgan (8) that the addition of 0.05 per cent flake graphite caused the forma— tion of coarse graphite flakes in an iron that would other— wise contain fine "supercooled" graphite. This lends support to the graphite nuclei theory suggested by Piwowarsky (7). This theory states that the late addition of a graphi- tizer causes minute graphite precipitates to form which in turn nucleate flake formation. This idea is also sub- stantiated somewhat by the fact that superheated irons tend toward the formation of type D graphite. This would lead one to believe that superheating causes these minute graphite particles to go into solution. It is also well known that inoculants lose their effect when iron is held in the ladle for some time between inoculation and pouring. This could also be caused by the graphite particles becom- ing disolved in the iron. Eash (9) reports that an addition of 0.05 per cent 5 graphite to the ladle had a very slight effect on the struc- ture and strength of 2.50 per cent carbon, 2.65 per cent silicon iron. However, the addition of 0.05 to 0.10 per cent graphite to the mold during pouring increased the tensile strength approximately 6000 psi. by improving the structure. He concludes that irons which contain type D Graphite solidify in the metastable iron-cementite system as white iron, with the eutectic carbide subsequently de- composing to form iron and graphite. Eash also points out that in the case of inoculated irons, type A graphite forms during the freezing of the graphite—austenite eutectic according to the stable iron-graphite system. This is thought to be accomplished by graphite nuclei provided by inoculation. Boyles (2) concludes the following regarding the freez- ing of cast iron from his work with quenched specimens: "Imimary austenite freezes out in the form of dendrites which continue to grow down to the eutectic temperature. "Crystallization of the eutectic begins at centers which grow equally in all directions, forming a cell—like structure. "Segregation takes place in two stages — (a) between the primary dendrites and the liquid — (b) from the crys— tallization centers of the eutectic outward into the boundaries of the cells. "Constituents formed during the freezing of the 6 eutectic occupy the interstices of the primary dendrites. The graphite flakes and the phosphide eutectic are thus re- stricted by the size and distribution of the dendrites." Boyles also states that three types of nucleation occur during the freezing of cast iron. The first of these is the nucleation which originates the formation of dendrites of primary austenite. The next type of nucleation which occurs is that from which centers of eutectic crystallization form and grow to become colonies of graphite flakes. The third type is nucleation of individual graphite flakes forming simultaneously with the growth of eutectic colonies. Smith (15) states that the principal function of an inoculant is probably to nucleate the formation of primary austenite crystals rather than graphite, thus giving a fine uniform grain size in the solid crystals around which the graphite flakes will form. However, McGrady(5) states, "The spacing of the primary dendrites, although it varies somewhat in normal and ab- normal irons, cannot be taken as the principal cause of variations in graphite distribution." Work was done by Spangler and Schneidewind (10) inject— ing graphite under the surface of molten metal by means of nitrogen. Carbon content of the melt was increased by as much as 1.00 per cent in some cases by this method. This work was not done primarily for the purpose of studying the ’7 effect of graphite as an inoculant, but rather to investi- gate the possibilities of producing significant changes in the carbon content by injection. However, it was found that chilling tendency was decreased, and strength equiva— lent to other inoculated irons was obtained. Structure was mainly type A graphite in pearlite. The improvement in properties observed may have been due in part to the addition of nitrogen to the iron. Dawson, Smith and Each (12) reported that an increase in nitrogen content results in an improvement of mechanical properties by causing the graphite flakes to become thicker and shorter. Since the literature is rather inconclusive regard- ing the effect of graphite inoculation on the properties of cast iron, the present work was undertaken in an attempt to more definitely determine this effect and if possible to shed a little more light on the mechanics of inoculation. II. EXPERIMENTAL PROCEDURE Experimental work was done in the following manner: All heats were melted in a 55 pound capacity induction furnace with a power rating of 20 kilowatts. A basic furnace lining was used. Molten iron was brought to 28750 F. as measured by an optical pyrometer before being tapped into ladles. This tapping temperature is within the range of tapping tem— perature used in commercial foundries. Also, a tapping temperature of 28750 F. provides sufficient time for moving the iron to the molds, skimming the ladle, reading the tem- perature with either an optical or an immersion pyrometer, and pouring at 26500 F. Experimental work was divided into two parts. The first nine heats were studies of structure, strength, chilling tendency, and cell size. The final five heats were studies of nucleation and chilling tendency. Pig iron, ingot iron, and ferro-silicon were charged in the cold furnace. When the charge reached a temperature of 27000 F. the ferro—manganese, ferro-phosphorus, and ferrous sulfide were added. The graphite was added con- tinuously to the ladle during tapping of the inoculated heats. During tapping the furnace was moved in a circular path above the ladle in order to carry the graphite under 8 the surface of the molten iron. The following is a typical charge used for an inocu— lated heat. Analysis of Pig Iron 0 Si Mn P S 11.6 lbs. of L1 pig 2.72 2.21 1.01 0.14 0.054 14.7 lbs. of L2 pig 5.58 2.00 0.87 0.16 0.044 5.7 lbs. of ingot iron punchings 0.902 lbs. of 25% ferro-silicon 0.045 lbs. of 80% ferro-manganese 0.060 lbs. of 25% ferro-phosphorus 0.056 lbs. of ferrous-sulfide 0.155 lbs. of graphite More high carbon pig iron and less‘low carbon pig was used in the heats to which graphite was not added. Transverse bars were used to determine the physical properties of the various heats. Inoculation generally has a greater effect on the structure near the surface of a casting. Therefore, the transverse test of unmachined bars is more revealing in this case than the tensile test of machined bars. Dry sand molds made from a mixture of lake sand, cereal, water, and linseed oil were used for standard 1.2 inch transverse test bars. The molds were baked for four hours at a temperature of 4750 F. They were then washed with a non—carbonaceous silica slurry, and the baking cycle was repeated. The transverse bars were poured in 10 the vertical position. Three bars were poured from each heat. The bars were tested on 18 inch centers: the break- ing load and deflection at the instant of breaking being measured. The diameters were measured and the corrected breaking load and triangular resilience (corrected breaking load x deflection x l/2) were determined. Photomicrographs were made of typical transverse test bars. Micro samples were also etched deeply and examined at 25 power to deter- mine cell size. Two types of chill test bars were used. One type is a wedge approximately 4 x 2 x 1/2 inches and the other is a rectangular chill block approximately 5/4 x 2 1/4 x 5 7/8 inches. Four chill test bars were poured from each heat. Chill tests were poured in molds made from the same sand mixture as the transverse bar molds. However, the chill test molds were not washed. Chill test bars from compared heats were broken at comparable positions and clear (white) and total (white plus mottled) chill was measured. Chill measurements obtained from different bars of the same heat agreed very well. Typical chill test bars were photographed. A one inch diameter cylinder five inches long, con- taining two chromel-alumel thermocouples was poured from each of the first nine heats. Cooling curves were obtained by using these thermocouples in congunction with a high Speed electronic recorder. The final five heats were poured into chill test molds 11 and wedges approximately 1 x 5 x 6 inches for nucleation study. These wedges were cooled slowly until the start of eutectic formation as evidenced by the cooling curves on the recorder chart, then quenched in water. These wedges were sectioned, polished, etched, and Visually examined. Typical nucleation study wedges were photographed. Two quenched wedges were poured from each heat along with one wedge of the same dimensions which was allowed to cool slowly in the mold. When inoculated heats were being poured, a small bar was poured from the furnace before graphite addition. This bar was used for the determination of carbon content of the heat before the addition of graphite. Final carbon, silicon, manganese, phosphorus, and sulfur contents were also determined for each heat. The greatest difference in carbon equivalent (%C + %Si) 9 between any two heats whose properties are compared is 0.04 per The compared The compared The compared cent. greatest heats is greatest heats is greatest heats is difference in manganese content between 0.06 per cent. difference in phosphorus content between 0.042 per cent. difference in sulfur content between 0.012 per cent. III. DATA AND RESULTS Part A. Chemical Composition In studying the effect of graphite as an inoculant, several 35 Pound heats of hypoeutectic iron were inocu— lated with 0.155 pounds of RECARB-X Mexican graphite. These heats were then closely matched in composition by untreated heats. The various heats are arranged in sets for comparison of compositions in Table 1. Relatively low carbon and silicon contents were used as this grade of iron is more likely to contain abnormal graphite. The first heat poured, L5T, was inoculated with 0.47 per cent of No. 8 Mexica graphite. Since only about 0.18 per cent carbon increase resulted, it was decided to use RECARB—X graphite for all but one of the subsequent inoculated heats. No. 8 graphite is sized all through a six mesh screen with a maximum of 5 per cent passing a 50 mesh screen. RECARB-X is sized all through a 20 mesh with a minimum of 55 per cent passing a 100 mesh screen. Therefore, HECARB-X affords much more surface area in contact with the molten iron. Carbon increases ranging from 0.22 per cent to 0.55 per cent resulted from inoculating with 0.47 per cent RECARB—X in pouring heats L4T, L5T, L9T, L12T, L15T, and L14T. 12 13 0H. mm. mm. 00. 00. mm. mm. mmmmnqu connmo & pama0>azum 0 $ 00.0 00.0 000.0 0ma.0 00.0 00.0 H0.0 00.0 000.0 :0a.0 00.0 00.0 00.0 000.0 00H.0 00.0 0H.0 00.0 00.0 000.0 000.0 H0.0 00.0 00.0 00.0 000.0 0ma.0 00.0 00.0 H0.0 000.0 00H.0 00.0 00.0 00.0 00.0 000.0 Hma.o 00.0 40.0 00.0 00.0 000.0 msa.o 00.0 00.0 00.0 000.0 00H.0 H0.0 00.0 00.0 00.0 000.0 s0a.0 00.0 0H.0 00.0 000.0 00H.0 00.0 00.0 00.0 00.0 000.0 mma.0 H0.0 00.0 00.0 000.0 msa.0 00.0 00.0 00.0 00.0 0 m :2 00 0 sagas 0 HmapfiqH pcoEmHm pamonmm 00000 A00202Hmmmxm mo onaHmomzoo q<0H2000 H mqmda opagmmpo w .02 osoz Numm¢omm Xnmmdomm wcoz Nnmmdomm mpfiemmpo 0-00 mqoz Numm¢omm mooz Xnmm 0 2 O O O 3 4 5 O O 6 7 I'm, 'J (‘HI Sm. Snxmm. htl: l'w \T \ Mumwn \I'IU\ ..y '3’- gufe 4. Eutectic Cell Size Chart. After Adams (I -\ J 25 A comparison of the number and size of eutectic cell centers was accomplished by water quenching wedges at the beginning of eutectic solidification as indicated by the cooling curves obtained by imbedded thermocouples. A wedge from each heat was also cooled slowly. All wedges were sectioned, polished, etched and visually examined. As is shown in Table 4, all inoculated quenched wedges con— tained many small centers, while untreated quenched wedges contained a small number of larger centers as was reported by McGrady (5) in the cases of other inoculants. Figure 6 shows the appearance of typical inoculated and untreated wedges which were quenched at the beginning of eutectic solidification. Figure 7 compares microstructures of the same quenched wedges as were shown in figure 6. Figure 8 compares micro structures of slow cooled wedges from the same heats as quenched wedges shown in figures 6 and 7. It is apparent that the slow cooled wedge from inoculated heat L14T contains Type A graphite while the wedge from untreated heat L16B contains Type D graphite. 24 TABLE 4 Structure of Wedges Quenched During Solidification Iron No. L12T L15B L14T L16B Treatment RECARB-X None RECARB-X None Structure of Qpenched Wedges Many small centers of eutectic crystallization Few large centers of eutectic crystallization Many small centers of eutectic crystallization Few large centers of eutectic crystallization :2 dc. in '1 [ 3.111 ‘3'?! C l ('3 f- . .H’ Crvs ii *c ._ J, u A. c . . I .4’{ "[3 o 1 I t \ b .LJ f l C =4 1i thwat Llff z" - .J( ,‘ y 11‘. (A r ._ A. .74 6.0 . 3 ~ r: m n . dft C n Ln. H0 C H. 7‘. .L. 14 i.) . ('T‘l‘v’lr3 r‘ . 31001., L 7. EJT'Cf'E‘L'.‘ ALI-”3 1:11:1'L’ a >_ r‘ H v . ......) in V9’ mf‘ nn .2? "‘ LPK‘A‘ Ulj‘rk f 11‘ 1711* ' - k" “V _)\,g;}(\ t 1 I - ”. 11(1, 7 V1 ,‘ .. v 11 “ L ‘ ““109” WWW“ a") L 1. - UK ’\ ’ 'i" _ '-:., _ Kl “Ly qr 1 1 1,.} no A 71,37. ... n'. . Q Figure 0. l"! Chxuflyibe DistflbutlcmljflldnlvV Cooled Wedges. X100. (8) Graphite Incculated Heat L14T. (b) Ifiitreated Inuit L168. Part D. Chilling Tendency As was to be expected from reports of previous work, graphite inoculation was found to reduce chilling tendency considerably. Table 5 points this fact out in comparing chilling tendencies of irons having practically the same compositions. All results shown are averages of at least two samples. Graphite inoculation was found to decrease clear chill an average of 15/52 of an inch, and total chill an average of 26/52 of an inch in the wedge chill bars. Clear chill was decreased an average of 11/52 of an inch, and total chill was decreased an average of 56/52 of an inch in the rectangular chill bars. Figure 9 shows wedge chill tests from untreated heat L8B, while figure 10 shows wedges from inoculated heat L5T. Figure 11 is a photograph of rectangular chill tests from untreated heat L8B, while blocks from inoculated heat L5T are shown in figure 12. Iron No. L4T L7B L5T L8B L9T LlOB LllT L12T L15B L14T L163 TABLE 5 Chilling Tendencies of Experimental Irons Treatment Chill depth in 52nds of an inch Wedge Sample ECARB-X None RECARB—X None RECARB-X None BB-6 RECARB-X None RECARB—X None Clear Total 6 15 18 42 4 10 18 56 4 lO 17 57 5 9 5 11 15 5O 29 Rectangular Sample Clear 4 l5 4 16 15 10 15 Total 8 56 10 $4.31.). ,"> I h 1" V 1 .21 t, e <1 ire l7). Lj ’"1 ‘L “-Ql 'rx-} 1 v A". rm. ~ 17. ...L 2 r11" 2 ~-. ”M ' rnvnre 11. atroatoi - 1 Test niocwo. n;;u qu- .] Part E. Cooling Curves As previously stated cooling curves were made of several heats by means of a high speed electronic recorder in conjunction with chromel-alumel thermocouples placed in the molds. Cooling curves for three untreated heats, L78, L8B, and LlOB are shown in figure 15. These show an initial eutectic temperature of about 19900 F. with an undercool- ing of 5 to 100 F. The length of time required for the eutectic solidification was approximately 50 seconds. Cooling curves for four graphite inoculated heats, L4T,.L5T, L6T, and L9T, are shown in figure 14. It can be seen that in the case of graphite treated irons the eutectic formation began in the range of 2020-20500 F. No undercooling is evident. The eutectic reaction re- quired approximately 70 seconds. A graphical comparison of typical cooling curves for graphite inoculated and untreated irons is given in figure 15. Table 6 gives a summary of the initial temperature of eutectic formation and the duration of the eutectic reaction in the untreated and the graphite inoculated heats. 55 TABLE 6 Summary of Data Obtained from Cooling Curves O F. at Start of Time Required Iron No. Treatment Eutectic Reaction in Seconds L4T RECARB-X 2020 60 2020 60 L5T RECARB—X 2040 72 2050 78 L6T RECARB—X 2055 78 2055 66 L9T RECARB-X 2057 78 L78 None 1966 50 1986 ‘ 48 L83 None 1995 40 1990 60 L10B None 1985 50 1970 55 Average for graphite heats 2040 70 Average for untreated heats 1982 51 \1 49 't‘tl'ul‘. ..." li."’x’l.‘..l‘ll’.l .~ n.’ .I.."\r.~ lbl.I-III|[I-u ..- .Io-I.‘ .Il.. ..ll 0 '11.?! . ‘Ir 11" til I III‘I.|.|‘II411."IIIIIII¢O| I] 57 A _ l . 1'. I- ,lI'l i4 .lll.|.|ll||‘0n-|l4ll|.l-4"» -c- )I . n .5 [£950.4’“. I. DJ 7A., “v'l¢l‘|l“‘ \IL! r: w 0,._._-- IV. DISCUSSION reater carbon pick up was achieved in this study than has been reported previously by other workers using ordinary methods of addition. This is probably due to the fact that the furnace was moved about over the ladle during tapping in an attempt to push the g aphite beneath the surface of the iron. Also, a relatively high tapping temperature was used. Regarding carbon pickup, Spangler and Schneidewind (10) conclude that negligible results are usually obtained by means of ladle treataerts with granular graphite. Angus (14) states, "Carbon in the form of graphite is quite frequently used as a mild inoculant for the control of chill, but it is very difficult to obtain a pickup in the ladle of greater than about 0.10 to 0.15 per cent. Additions of fine graphite to the ladle are likely to produce a considerable amount of dirt, both in the foundry itself by air flotation and also in the ladle as scum." Carbon pickup was not extremely uniform in the present study, ranging from 0.22 to 0.56 per cent. These differ- ences were apparent during pouring, with some heats show— ing more graphite floating on the iron than others. This situation could probably be improved by a metering device at the furnace spout for controlling graphite flow in 59 40 conjunction with a means for insuring uniform iron flow in tapping. Uniformity of pickup could also be improved by use of a gas to carry graphite beneath the surface of the iron as reported by Spangler and Schneidewind (10). This study showed a greater increase in transverse strength and deflection due to graphite inoculation by ordinary methods than has been reported previously. This may be caused by a greater graphite recovery than was previously achieved. Microstructure of graphite inoculated heats was found to be mainly Type A graphite in pearlite, a definite im- provement over the predominantly Type D graphite in pearlite of the untreated irons. Again, probably because of rela— tively high graphite recovery, these data show a somewhat greater effect on structure than reported by previous investigators. Cell size of inoculated irons was found to be slightly smaller than that of untreated irons. Smaller cell size has been shown by Adams (1) to be a characteristic of irons having better mechanical properties. Also, thrady (5) states, "Successful inoculation is accompanied by a de- crease in cell size." Cooling curves showed that graphite inoculation brought about a considerable raising of the initial eutec- tic temperature. This would be expected to accompany an improvement in structure and properties. McGrady (5) 41 states, "Successful inoculation is accompanied by an ele— vation of the range of eutectic temperature." Undercool- ing was eliminated by graphite inoculation. However, as pointed out by hoGrady (5), "The elimination of undercool— ing does not necessarily result in an improvement in graphite distribution." Wedges quenched during eutectic solidification show a small number of large centers of eutectic crystalliza— tion in the untreated irons and a relatively large number of small centers in the inoculated irons. This is evi- denCe that graphite inoculation serves to nucleate centers of eutectic crystallization. Micro examination showed the eutectic centers in the untreated quenched wedges to contain Type D graphite ex- tending to the edges of the cells. The slowly cooled wedge from the same heats also contained mainly Type D graphite. The eutectic centers in the inoculated quenched wedge contained mainly Type A graphite. Slowly cooled wedges from inoculated heats also contained mainly Type A graphite. Since eutectic solidification occurs at higher tempera- tures in inoculated iron, possibly the rate of growth of austenite in the eutectic is slow enough to allow the formation of Type A graphite. In the case of untreated irons in which the eutectic solidification occurs at lower temperatures, perhaps the growth of austenite in the eutectic is too rapid to allow the formation of Type A 42 graphite. Larger primary dendrites of austenite in un— treated irons, which have had more time to grow before eutectic solidification begins, may also contribute to the formation of abnormal graphite. Fine Type D graphite flakes were found in the interior of the inoculated wedges. This graphite was apparently formed during quenching. Chilling tendency was very greatly reduced by graphite treatment in the ladle. This was to be expected from re- sults of previous work. However, as pointed out by McClure, Khan, McGrady, and Womochel (l5), chill reduction is not necessarily associated with an improvement in graphite distribution. Their work showed aluminum to be a powerful chill reducer but not a promoter of Type A graphite formation. Table 7 compares the average effect of graphite in— oculation to the average effect of calcium inoculation of similar irons as reported by McGrady (5). Calcium evi- dently has a greater beneficial effect on mechanical properties than does graphite. This is probably the result of a slightly better structure in the calcium treated iron brought about by eutectic formation at a somewhat higher temperature and over a longer period of time. Graphite appears to be slightly more effective in reducing chill than is calcium. No Type D graphite was evident in the interior of quenched wedges treated with calcium. 45 TABLE 7 Comparison of Effect of Ladle Additions of Graphite and Calcium on Physical Properties and Eutectic Formation of High Strength Cast Iron. Temp. at Start Duration Treatment of Eutectic of Eutectic Undercooling 0.47% O RECARB-X 2040 F. 70 sec. None *l.0% 0 Calcium 2048 F. 76 sec. None Transverse Triangular Breaking Load Deflection Resilience 0.47% RECARB-X 5055 lbs. 0.515 in. 477 in. lbs. *l.0% Calcium 5550 lbs. 0.581 in. 656 in. lbs. Block Chill 1/52 in. Carbon Eguivalent Clear Total 0.47% RECARB-X 5.58 4 9 *l.0% ’ Calcium 5.56 6 10 5 After MoGrady (5). 44 Figure 16 is a photomicrograph taken approximately 1/8 inch from the surface and the thermocouple of a quenched wedge inoculated with graphite. Type A graphite flakes and ledeburite are separated by areas containing very fine graphite flakes. Areas of very fine graphite are similar in shape to areas of ledeburite. This may indicate the arrest by quenching of the following sequence of events: 1. Formation of ledeburite or cementite from liquid. 2. Breakdown of ledeburite or cementite to form very fine graphite flakes. 5. Agglomeration of very fine graphite flakes to form larger Type A flakes. Possibly_figure 16 can better be explained by apply— ing the three following structural conditions for eutectic freezing suggested by Boyles (2): "(a) Conditions where there is no breakdown of the FeBC during freezing. This results in the formation of a cementite-austenite eutectic and produces the ordinary structure of white iron. "(b) Conditions where the eutectic freezes as austenite + cementite, the cementite breaking down immediately to form graphite before all the eutectic is frozen. This is the type of reaction which, in the opinion of many metallur- gists, occurs adgacent to the chilled part of a wedge casting, giving rise to the Type D graphite structure. "(0) Conditions where all the Fe5C breaks down as it Figure 16. Quenched Wedge Inoculated with Graphite. X500. Very Fine Graphite Between Areas of Ledeburite and Formal Graphite Flakes. 46 comes out of solution in the liquid without forming any solid particles of cementite. This reaction produces a graphite-austenite eutectic structure such as that seen in Fig. 52, the flakes growing outward directly into the remaining liquid." The ledeburite in figure 16 may be the result of con- ditions in paragraph (a). The very fine graphite may be the result of Conditions in paragraph (b). The larger graphite flakes may have been caused by conditions in paragraph (c). If this is the case, the liquid—solid interfaces at the instant of quenching were probably be- tween the large graphite flakes and the areas now contain- ing very fine graphite flakes. It is apparent from this investigation that the mechanism of graphite flake formation Cannot be defin— itely determined by means of quenching experiments because an infinite cooling rate cannot be attained. This investigation also indicates the need for an investigation of the duration of the beneficial effect of inoculation with graphite. V. CONCLUSION The following can be concluded from this study: 1. A carbon increase ranging from 0.22 to 0.56 per cent can be achieved by ladle addition of 0.47 per cent graphite during tapping. 2. Inoculating with graphite reduces chilling ten— dency to a considerable degree. 5. Graphite inoculation results in a definite in— crease in transverse strength and deflection. 4. A graphite inoculated iron contains predominantly Type A graphite. 5. Cell size is decreased slightly by treating with graphite. 6. Graphite inoculation causes the nucleation of many small centers of eutectic crystallization as opposed to a small number of large centers in untreated iron. This is evidenced by wedges quenched during eutectic solidification. 7. Graphite inoculation raises the initial tenpera- ture of the eutectic reaction approximately 580 F. and increases the length of time required for eutectic solidi- fication in the test bars by approximately 19 seconds as shown by cooling curves. 8. The results of this investigation would seem to lend support to the idea of minute graphite particles in the melt serving as nuclei for centers of eutecuic crystallization. 47 VI. BIBLIOGRAPE R. R. Adams, "Cast Iron Strength Verses Structure," Transactions, American Foundrymen's Association, Vol. 50 , 1945 1 A. Boyles, "The Structure of Cast Iron," Published ov U the American Society for Metals, 194? D. D. thrady, "The Effect of Ladle Additions of Aluminum, Calcium, Silicon, and Graphite on the Eutectic Solidification, microstructure, and Pnysfical PrOperties Thesis, hichigan of Hypoeutectic Gray Cast Irons. State university, 1956 H. L. Eomochel, "Efifect of Ladle Additions of Some Alloys and Active hetals on the Properties and hicro— structure of Gray Cast Iron," Thesis, hichigan State College, 1954 H. R. Dahlberg, "Effect of Graphite Particle Size in the Inoculation of Gray Cast Iron," Foundry, Vol. 71, August, 1945 A. H. Dierker, "Ladle Additions of Graphite to Gray Cast Iron," American Foundryman, Vol. 2, November, 1340 n n. Piwowarsky, "Progress in the Production of High Test 48 10. ll. 12. 15. 14. 49 Iron," Transactions, American Foundrymen's Association, Vol. 54, 1926 A. L. Norbury and E. Morgan, "Effect of Melting Con- ditions on the Microstructure and Mechanical Strengths of Gray Cast Irons of Various Amounts of Carbon and Silicon," Foundry Trade Journal, May 15, 1950 J. T. Eash, "Effect of Ladle Inoculation on Cast Iron," Transactions, American Foundrynen's Association, Vol. 49, June, 1942 G. E. Spangler and Schneidewind, "Increasing Carbon Content of Cast Irons by Ladle InJection," Transactions of the American Foundrymen's Society, Vol. 65, 1955 R. A. Flinn and D. J. Reese, "The Development and Control of Engineering Gray Cast Irons," Transactions, American Foundrymen's Association, Vol. 49, 1941 J. W. Dawson, L. V. Smith, and B. B. Each, "Some Effects of Nitrogen in Cast Iron," American Foundryman, Vol. 26, July, 1954 M. C. Smith, "Alloy Series in Plysical Metallurgy" Published by Harper and Brothers, 1956 H. T. Angus, "Carbon Pickup," The British Cast Iron Research Association, Journal of Research and Develop- ment, Vol. 4, No. 6, June, 1952 15. 50 N. C. chlure, A. U. Khan, D. D. XcGrady, and H. L. fiomochel, "Inoculation of Gray Cast Iron ~ Relative Effectiveness of Some Silicon Alloys and Active ll Metals as Ladle Additions, American Foundrymen's Society, Preprint No. 57-2 Date Due Demco-293 WWII” ' l i l I I, II II I I II I