‘73?“ ‘. .v. ‘ THE EFFECT or LADLE ADDITIONS OF ALUMINUM, CALCIUM, smcon, AND GRAPHITE ON THE aurscnc sduomunou, MGROSTRUCTURE, AND. PHYSICAL paopmnss OF HYPOEUTECTIC GRAY CAST mons M fath- M a! Ph. I). ma smuummsm Donton WMcGrady 1956 IHESIS This is to certify that the thesis entitled THE EFFECT OF LADLE ADDITIONS OF ALUMINUM, CALCIUM, SILICON, AND GRAPHITE ON THE EUTECTIC SOLIDIFICATION, MI CROSTRUCTURE, AND PHYSICAL PROPERTIES OF HYPOEUTECTIC GRAY CAST IRONS 0 presented by BENTON DELBERT MCGRADY has been accepted towards fulfillment of the requirements for i _P_h._1_)_._ degree in _Me ta]. lgri cal Engi ne er ing DOCTOR OF PHIIDSOPHY Majoé prozessor J 2 Date W 0-169 THE EFFECT OF LADLE ADDITIONS OF ALUMINUM, CALCIUM, SILICON, AND GRAPHITE ON THE EUTECTIC SOLIDIFICATICN, MICROSTRUCTURE, AND PHYSICAL PROPERTIES OF HYPOEUTECTIC GRAY CAST IRONS BY BENTON DELBERT MCGRADY A THESIS Submitted to the School for Advance Graduate Studies of Michigan State University of Agriculture and Applied Science in partial fulfillment of the requirementsfbr the degree of DOCTOR OF PHILOSOPHY Department of Metallurgical Engineering 3 5 v- 1956 _ 1’ L é/J.‘ (/57 gA/Slléa ABSTRACT A study was made of the relative effectiveness of pure aluminum, calcium, silicon, and graphite as inoculants for gray cast irons of a nominal composition of 2.85% carbon, 2.25% silicon, 0.90% manganese, 0.17% phosphorus, and 0.07% sulfur. Cast irons melted in an indirect- arc rocking furnace were inoculated with aluminum and calcium and compared to untreated irons of identical analysis. Comparative data are included on graphite distribution, chemical analysis, cell size, chill depth, transverse strength, deflection, and triangular resilience. Cast iron melted in a high frequency induction furnace were poured into dry sand.molds and time-temperature solidification curves auto- matically recorded for untreated iron and for irons inoculated with .0.6% aluminum, 1J3% calcium, 0.55% silicon, and 0.5% fine graphite. Comparative cooling curves are presented for the induction furnace irons along with associated data on physical properties, microstruc- ture, and chill depth. Small wedge-shaped castings from a blank iron and from an iron inoculated with calcium.were quenched into cold water at the start of the eutectic formation and the resulting microstructures studied for difference in primary dendrite size and distribution. The following general observations and conclusions are made. (1) The use of aluminum as a ladle addition caused a marked reduction in chill, but had very slight or negligible effect on graphite distribution and no effect on physical properties. Cell size was not changed by the late addition of aluminum and the amount of carbon and sulfur in the cast iron remained the same. Aluminum raised the temper- ature of the initial formation of the eutectic by about 50°F. on the average. The microstructure showed largely type D graphite. (2) Cast irons inoculated with 0.55% Pure silicon were not significantly differ- ent from the corresponding blank irons. (3) The use of 0.5%lof fine graphite as an inoculant raised the temperature of initial eutectic formation by about éO'F. on the average. (h) The late addition of 1.0% of pure calcium metal to the ladle resulted in a cast iron that contained a large proportion of type A graphite, that showed a marked reduction in chill and in cell size, and that exhibited a definite increase in physical properties. (5) The use of ladle additions of calcium raised the temperature of initial eutectic formation by 60-80'F. as compared to a corresponding blank iron. The calcium inoculated irons showed eutectic cells forming from.a large number of nuclei as compared to blank irons which contained relatively few centers of eutectic cell formation. The addition of calcium decarburized and desulfurized the molten cast iron and evidence was Obtained to indicate that the decar- burization was caused, at least in part, by the formation of calcium carbide. The size and distribution of the primary dendrites of austenite was not appreciably changed by inoculation with calcium. (69 Strong evidence was obtained that inoculation is a nucleating process and this study points to either a carbide or a sulfide as the nucleating agent. (7) The elimination of undercooling does not necessarily result in an improvement in graphite distribution. (8) The rate of growth of the eutectic cells is an important factor in graphite shape and distribution. (9) Successful inoculation is accompanied by an elevation of the range of eutectic temperature and by a decrease in cell size. (10) The spacing of the primary dendrites, although it varies somewhat in normal and abnormal irons, cannot be taken as the principal cause of variation in graphite distribution. (11) A decrease in carbide stability as measured by chilling tendency is not necessarily related to graphite distribution. (12) The poSsibility exists that a group of elements related to calcium may also be effective as inoculant for gray cast iron. MM Approved, Major Professor PREFACE This study is a continuation and extension of the general prOblem initiated by WOmochel at al with regard to the relative effectiveness of various active metals and other elements as inoculants for gray cast iron. Because of the complex nature of the alloy and the consequent large number of variables, the literature contains very little funda- mental information about this topic. This research was undertaken in order to better determine the nature of the time-temperature solidification curves, particularly in the region of the eutectic transformation. Parallel observations of microstructure and physical properties were also made. ii ACKNOWLEDGEMENT The author would like to acknowledge the guidance and counsel of Dr. H. Im‘WOmochel in this study and express his appreciation for valuable discussion to Dr. A. J. Smith and Dr. R. L. Sweet. The en- couragement of Mr. J. W. Hoffman and the financial assistance of the Engineering Experiment Station is also gratefully acknowledged. Mr. F. Landstrom, Mr. B. D. Curtis, and Professor D. G. Triponi contributed important services and assistance to the study. iii TABLE OF CONTENTS I. II‘JTRODUCTIONO0QOO0.0.0.0.000...oOOOOOOOOOOQOOOOOOOOCOOO II. GEIJEIML HISTOI)~ICAL SURVEYOOOOOOOOOOOOOOOO0.00.00.00.00. III. DISCUSSION OF FACTORS AFFECTING GRAPHITIZATICN......... A. Chanical FactorSOOOOOOOOOOOOOOOOOOOO0.00....0. B. Thermal Factors............................... IV. HPFMIWIGTAL PRWEDUF‘EOOOO0.000000000000000000000...... A. Rocking FUrnace Irons......................... B. IndUCtion Furnace IronSoooo00000000000000.0000 C. PhySical PropertieSOOOOOOOOOOOO0.0...000...... V. DISCUSSION OF EXPERIMENTAL DATA AND RESULTS............ A. Rocking Furnace Irons......................... 1. Aluminum additions...................... 2. Calcium additions....................... B. Induction Furnace Irons...................... l. Time-temperature solidification curves.. 2. Microstructure and physical properties.. 30 Tlater-quenChed ironSOOOOOOOOOOOOOOOOOOOO VIC SUIVDTARY AND CONCLUSIONSQOQQ000000000000.00000000000000. VII. BIBI‘ImWYOOOOOOOOO0.0...COO...OOCOOOOOOOOOOOOOOOOOO. iv Page 13 18 22 22 2h 26 29 29 29 31 39 39 57 85 89 Table l. 10. ll. 12. 13. LIST OF TABLES Data on Aluminum Inoculation of Rocking Furnace Iron T3h............ Data on Calcium Inoculation of Rocking Furnace Iron T35............. Data on Calcium Inocudation of Rocking Furnace Iron T39............. Data on Calcium.Inoculation of Rocking Furnace Iron Th0............. Comparative Data on Calcium Inoculation of Rocking Furnace Irons.... Comparison of Inoculation Effect of Aluminum and Calcium on ROCking mrnace Irons............................................. Effect of Calcium Inoculation on Carbon and Sulfur Content or caSt IronOOOOOO0.00.00.00.00...I.0.0.0.0....OOOOOOOCOOOOOOOOOOOO. Comparison of Carbon Equivalents of Rocking Furnace Irons........... Comparison of Effect of Ladle Additions of Aluminum, Silicon, Calcium, and Graphite on the Initial Temperature of Eutectic Formation Of Induction Fm'nace IronSOOOOCOCOOOOOOCCOOOOOOOOOOO0.0... Comparative Data on Calcium, Aluminum, and Silicon Inoculation Of Induction Furnace caSt IronOCOOOOCOCOOOOOOOOOOOOOOOOOOOOOOOOOOOO. Comparison of Inoculating Effect of Calcium, Aluminum, and Silicon.. on Induction Furnace Irons.......................................... Summary of Average Results of Calcium, Aluminum, and Silicon Additions to Induction Furnace Irons................................ Comparison of Carbon Equivalents of Induction Furnace Irons......... Page 30 35 36 38 L9 55 56 So Figure l. 3. h. 16. 17. LIST OF FIGURES American Foundrymen's Society Chart for Graphite Distributi-on in 08.81.; IronOOOOOOOC0.00IIOOOOOCOOCOO0.000...... Section Through the Ternary Iron-CarbonFSilicon Diagram at 2.0% SiliconOOOOOO0.000000000000000000000000...... Microstructures Obtained by Quenching Partly Solidified Hypoeutectic Gray Cast Iron. (Boyles)............. Time-Temperature Curves for Solidification of Hypoeutectic Cast Iron. (Boyles)............................ Time-Temperature Curves for Solidification of Hypoeutectic Cast Irons. (Eash)............................. Cooling Curves for Uninoculated Cast Iron.................... Cooling Curves for Cast Iron Inoculated.with Aluminum........ Cooling Curves for Cast Iron Inoculated with Calcium......... Cooling Curves for Uninoculated Cast Iron.................... Cooling Curves for Cast Iron Inoculated with Fine Graphite... Comparison of Typical Cooling Curves......................... Cooling Curves for'Water Quenchedeedge Samples.............. Typical Microstructure of Type A Graphite Distribution....... Typical Microstructure of Type D Graphite Distribution....... Dendrite Size and Distribution in Iron T53 Inoculated With 1.0% Galoiunlmoooo.0000000000000oooooooocoooooooooooooooo Dendrite Size and Distribution.in Iron T53 Inoculated With 100% 08101111“...00000000000000.000000000000000000000cocoo Dendrite Size and Distribution in Uninoculated Blank Heat T9400...0.0.0.0000...OOOOOOOOOOOOOOOIOOOOOOOOOOOOOOOOOOO Page 10 11 DO no L2 L3 m 146 h? be 51 52 53 59 Figure 18. 19. 20. 21. 22. 23. 2h. 25. 26. 27. 28. 29. Dendrite Size and Distribution in Uninoculated Blank Heat T94000000000.......OOCOIOOCOOCOOOCOOOOOOOOOOOOOOCOOOOO Effect of Inoculation on Number of Eutectic Cells.......... Graphite Flakes Formed in Eutectic Cells................... Graphite Flakes Formed in.Eutectic Cells................... Ledeburite Adjacent to Fine Graphite Constituent........... Microstructure of Iron T53 Near Top Surface of‘Wedge Quenched at Start of Eutectic Formation.................... Microstructure of Iron.T53 Very Close to Top Surface of wedge Quenched at Start of Eutectic Formation.............. Microstructure of Iron T53 Near Center of wedge Quenched at Start Of EnteCtic SOlidification........................ Same Spot as Figure 25. Polarized Light................... Decomposition Time for Cementite. (Hanemann).............. Time for Decomposition of Cementite at'Temperature Below the EUteCtiCo (Berman and Kondic)......................... Free-energy/Composition Diagram of Iron-Carbon System...... vii Page 61 63 6h 65 67 68 69 71 7b 7b 75 I. INTRODUCTION cast iron has served the needs of mankind since the Chinese produced cast-inn tools for the farmer and woodworker some 2,200 years ago. Howe (2h) writes that the period of cast-iron metallurgy begenin the 114th century with the production of ornamental castings in Sussex. Vannocdo Biringuccio, (6) writing in the first edition of his 'Pirotechnia" published in 15140, reports a process of steelnaking in which mess of wrought iron were imersed in a bath of molten cast iron. Biringucdo writes..."bloons weighing 30-140 lbs. each of iron... are put while hot into this bath of molten iron...then keep it in this nelted material with a hot fire for four to six hours..." By the end of the 16th century cannon weighing as much as three tons were cast. In 1735 Darby showed how to make cast iron by mean of coke, thus replacing the previously used wood charcoal. Thus for six centuries foundry-en have made iron castings to supply man's requirements for douestic and agricultural use - and more recently for engineering purposes. With the exception of steel ingots, cast iron articles constitute the largest tonnage of coatings made of any metal or alloy. The majority of these articles are made of grey cast iron. The casting method is admirably suited to making products of intricate nature, and the alloy itself is readily uchinable. The products of the iron foundry vary from pipe fittings to auto-otive cyliuier blocks, and their weight varies from a fractional pound to more than 50 tons. Following the early works (25) of such pioneers as Sorby, Martens, Tschernoff, Brinell, Osmond, and Hadfield before the turn of the 20th century, and the studies of Howe, Sauveur, Stead, Campbell and others in the early 1900's, cast iron was subjected to increasingly intensive study as to nicrostructure and associated physical properties. By 1910 Moldenke (35) had written a book aboutmlleable castings. Patents (29) on the late addition of small amounts of certain metals and alloys were granted in the United States about 1920. These late-addition ‘ materials, or inoculants, were usually added in relatively small amounts to the molten metal in the ladle Just before pouring the castings. The improvements in graphite form and physical proper- ties were much yeater than would logically be expected on the basis of simple alloy effects or slight change of chemical composition alone. Perhaps the high point in the study of the effect of inoculants on graphite distribution in gray cast iron came in 191:8 with the simultaneous discovery in England and the United States of 'ductile" cast iron. This new engineering material, which has graphite present in the form of nodules or spherulites rather than as the usual flake tour, has been described as the outstanding foundry development of the century. Certainly the pasthenty years, and particularly this decade, has been a period of fruitful. work in explaining some of the basic transformations in cast iron. Hultgren (26) states, "The question of whether the gaphite which is observed in cast iron after solidification is formed directly from the melt or as a decomposition product of an 3 initially precipitated solid carbide constituent has been the subject of experiment and discussion for a considerable time. For the primary graphite in hypereutectic irons, and also for the coarse eutectic graphite, formation directly from the melt now seems to be the generally accepted mechanism. This is not so, however, for the fine graphite, nor for the spherulitic graphite produced during cooling under special condi- tions.” Along the same line of thought Morrogh (33) has written, "The graphite phase in gray cast irons can be obtained in the flake form, in the undercooled form, and as spherulites. The conditions necessary for the formation of these structures are well known, but the funda- mental reasons for these modifications are not established. The view that the flake or lamellar form of graphite in gray cast iron forms as a hypereutectic or eutectic constituent directly from the melt seals to be widely accepted. It is, however, not easy to see why the graphite originating from a eutectic transformation should be so unlike a typical eutectic constituent in appearance. The undercooled form of graphite presents even more problems." In discussing the decomposition of cementite during the solidi- fication of cast iron, Berman and Kondic (5) state, "Many methods of producing commercially different forms of graphite in cast iron are known, but no generally acceptable explanation has been offered for the mechanism of graphite formation, nor are the reasons for obtaining different cast-iron structures under certain conditions fully under- stood.. . . There is still disagreement on whether graphite forms directly from the eutectic liquid or is a product of the decomposition of cementite.... Any general theory of graphite formation in cast iron must be based on the solution of this problem." It thus appears that the methods and conditions for altering the form.of the graphite in cast iron are reasonably well known, but that considerable work yet remains before establishing a satisfactory under- standing of the basic reasons for these modifications of graphite based on fundamental principles and concepts. A.comprehensive survey and critical.review of various theories explaining the mechanism of inocu- lation of gray cast iron has been presented 'by“womochel (h8) and need not be repeated in detail here. The purpose of this investigation is to make a fundamental study of the effect of late additions of aluminum, calcium, silicon, and graphite on the eutectic solidification, microstructure, and physical properties of gray cast irons. II. GENERAL HISTORICAL SURVEY Since the earliest work on the iron-carbon diagram (1890-1900) the general mechanism of freezing of cast iron has been known. The phenonemon of graphitisation was the point in question and the exis- tence of cementite and graphite together in the same alloy was the condition demanding explanation. Piwowarsky (37) reports that as early as 1906 Benedicts (14) had suggested the possibility of the formation of flake graphite from the solidified cast iron. Howe concurred with this concept during the years 1909-1911. In his book published in 1916 Howe (23) used the general idea of the stable and metastable equilibria and described quite clearly the sequence of events occuriug during the freezing of iron-carbon alloys without, however, giving details as to the manner in which graphite flakes were fomed. In 1906 Wust and Petersen (117) had shown that the carbon content of the eutectic is lowered by the presence of silicon and in 1913 Sauveur (39) plotted a diagram based on this work which related the silicon percent to the percent of car- bon in the eutectic. (A 2% Silicon iron is shown as having a eutectic at 3.81 of carbon). Howe (25) states that he considers the work of Tim (152) as being the first to clearly identify; .: primary cementite in white (I hypereutectic cast iron. Between 1915 and 1931 the work of Ruer and Goerens (38), Goerene (15), and Hanemann (18, 19), based on a study of the thermodynamics of both the iron-carbon and iron-iron carbide system, showed that the possibility existed that under certain conditions graphite in gray cast iron could form directly from the melt. Honda and Murakami (22) and Honda and Endo (21) wrote during the period 1920-1929 that graphite in gray cast iron is a result of decom- position of the carbide component of the eutectic. In 1921; Heyn and Bauer (20) studied the variation in graphite content of three cast irons containing different amounts of silicon and concluded that the graphite form was dependent on cooling rate and that the greatest part of the graphite had formed during the eutectic halt, not only for the low- silicon, hypoeutectic iron but for the eutectic and the hypereutectic cast irons as well. is research on the subject of graphitisation was continued two schools of thought gradually evolved - one maintaining that a graphite-austenite eutectic formed directly from the melt, and the other that cementite-austenite eutectic always formed first and that the formation of graphite was a secondary reaction. Hurst (27) writing in 1926 presented the arguments on both sides and pointed out the unsatisfactory state of knowledge at that time and the need for additional research on the tapic. Hanemann, (18) as a result of quench- ing experiments published in 1931, concluded that graphite formed directly from the melt and he published two photomicrographs one of which showed nests of graphite flakes surrounded by ledeburite and the other a single graphite flake extending at each end into ledeburitic areas. Boyles (8) in 1937 showed by a series of quenching experiments that graphite is not present before the eutectic solidification begins. From studies made in 192).; by pouring silicon containing iron in a thin 7 stream into cold water, Northcott (36) concluded that the absence of elemental carbon in the couples indicated that for hypoeutectic cast irons the graphite forms from the carbide phase, not disproving, however, the possibility of direct gramite formation. Hanemann (18) in 1931 and later studied the time required for decomposition of cementite, and concluded that relatively long times (hours) were required even at temperatures above lOOO'C. More recent work (5) has indicated that as little as four minutes may be needed at temperatures close to the eutectic. The discussion as to the mechanics of formation of the various types of graphite possible in grey cast iron has continued until the present day. Recently (19514) Morrogh (33) published work supporting the carbide decomposition hypothesis to explain type D graphite and in the same issue of the journal Hultgren, Lindhlm, and Rudberg (26) presented data to support the idea of direct formation of fine graphite from the melt. Ladle additions to improve the properties of cast iron are reported in the literature as far back as 1900 in the work of Wust (146) who studied lead, sine, tin, aluminum, sodium, and magnesium additions as related to degassing, purifying, and improvements in density. In 1908 Holdenke (3h) investigated the use of ferrosilicon, calcium, and vanadum ani recomnended these materials for deoxidising and increasing the density of cast iron. It has been known since about 1920 that late additions of small amounts of ferrosilicon, calcium-silicon, and other "inoculants" to the ladle just previous to casting can, under proper conditions, greatly 8 improve the graphite structure of cast iron (bl) . The original U.S.A. patent (29) to A. F. Meehan, which was granted in 1922, along with later patents and additions, covered in general the late additions to white and grey cast iron of calcium-silicon, calcium, and other alkaline earth materials either separately or in combination. An extensive literature about the mechanics of graphite formation and hypotheses to explain the effects of inoculation has developed in the past 30 years. Outstanding work based on a fundamental approach to understanding the formation of graphite in grey cast iron and the effect of inoculants has been published during the past decade. A series of papers by Alfred Boyles was brought up to date and expanded into a book (10) published by the American Society for Metals in 19117. The research and discussion of Rash, Schneidewind, D'AmicO, MacKenzie, Lownie, Lorie, and many others has been of great value. The research of the British Cast Iron Research Association (BCIBA) as presented in papers by Morrogh, Norbury, Williams, et a1 is or excellent quality. The Royal Institute of Technology and the Swedish Institute of Metals Research in Stockholm have contributed important data by authors such as Hultgren, Lindblom, and Rudberg. The work of Withnoser, Ward, Pelsel, and Piwowarsky continues in Germarw, and that of de Sy in France to mention only a few. The discovery in 191:8 that late additions or ladle inoculation of low-sulfur gray cast iron with cerium or magnesium could produce a nodular graphite form has not only created a new engineering material but has served to focus attention and intensify interest in recent years on fundamental studies of the inoculation phenomenon. I. III. DISCUSSION OF FACTORS AFFECTING GRAPI-UITIZATION The physical properties of gray cast iron are largely determined by the character or the matrix proper and by the size and distribution of the flakes of graphite throughout this matrix. The approach to the problem of graphitisation in cast iron has customarily been on the basis of binary or ternary equilibrium diagrams involving iron, carbon, and silicon. Cast irons are known, however, to behave much differently from pure alloys. Indeed it may well be that there is need for additional careful study of the iron-carbon-silicon alloy system. Figure 1 shows the classification chart for graphite distribution in gray cast iron as standarized by the American Foundrymen's Society and the American Society for Testing Materials. Because of superior physical properties, a uniform distribution and random orientation, as illustrated in type A, is usually preferred. However gray cast iron may also have graphite present in other forms as illustrated by types B, D, and E. A cast iron having a type D graphite pattern has inferior properties to an iron of the same chemical analysis, but with type A graphite. Figure 2 shows a section of the iron-carbon-silicon thermal equilibrium diagram at 2.0% silicon. Cast irons display the phenomenon of metastability, by which the solidification may proceed in either the stable iron-graphite system or in the metastable iron-iron carbide system. The effect is no doubt caused by the presence of several elements with rather diverse influences on solidification or freezing. 9 was mammfmbliw Figure 1. American Foundry-en's Society alert for Graphite Matribution in Cast Iron. cent. Temperature oc+Ca Carbon, per cani- Flc. 4—IRON-CAIBON Duonu m Plumes or 2.00 Pas Can-r qucow (Gasman, MARSH mu Smocnrow). Figure 2. Section Through the Ternary Iron-Carbon-sSilicon magram at 2.0% Silicon. 12 The mere presence of a stable graphite phase in gray cast iron is not sufficient to determine or control its quality. It is, rather, the size, shape, distribution, and orientation of the graphite masses which most concerns the metallurgist, for these are factors basic to the quality and strength of his product. Knowledge of the factors that influence graphitization is inade- quate at present. Among some topics which might well merit additional study and discussion Altekar (fl) has included: 1. Mode of occurence of carbon in molten cast iron. 2. Possibility of bringing about molecular groupings in the melt. 3. Mechanism of formation of carbides. h. Mechanism of stabilization of carbides by the presence of carbide-stabilizing elements. 5. Mechanism of decomposition of carbides, with and without the presence of graphitizers. 6. Control of the freezing of the eutectic. 7. Phenomenon of undercooling in cast irons. The achievement of nodular graphite in as-cast iron as the result of inoculation or late additions to the ladle, has had a most helpful effect in attracting the attention of research workers and has resulted in a large number of diverse explanations for the microstructures obtained. No overall comprehensive explanation has been offered to explain the several variations of graphite form in gray cast iron. It would see: logical that the various types of graphite bear a definite relation to each other. The conditions that change one graphite type to the other can perhaps be classified into two main parts— thermal factors and chemical factors. Manipulation of the 13 several alternatives under each of the main factors could give any size, distribution, and orientation of graphite flakes. A. CHEMICAL FACTORS AFFECTING GRAPHITIZATION Each element present in cast iron is effective, to a greater or lesser extent, in influencing the formation of graphite in cast irons. However, the effect of a particular element may be increased or minimized by the presence of some other element or group of elements. Thus no classification of these elements based upon their individual effects is entirely rigorous. The work of Boyles (10) and Piwowarsky (37) is prominent in the considerable literature that now exists on this subject. Boyles, by means of a series of quenches from successively advancing stages of solidification, was able to follow closely the initiation and progress of graphitization during eutectic freezing. The photomicrographs in Figure 3, from the work of Boyles, well illustrate the mechanism of solidification of hypoeutectic cast irons and may be summarised as follows: 1. Primary dendrites of Austenite solidify from the liquidus temperature down to the range of tenperature of the eutectic. 2. Crystallization of the eutectic begins at centers which grow outward spherically in all directions and form a cell- like structure. 3. Constituents formed during the formation of the eutectic occupy the interstices of the dendrites. Thus graphite flakes Fig. 1‘) 75ample No. .10. annched from 2010’ F. Heat-tinted. X 20. 0. Fig. ZO—Dclail nf Strucmrc in Fig. 1‘). Heal-tinted. X Figure 3. Microstructures (btained by Quenching Partly lidified Hypoeutectic Gray Cast Iron. Heat Tinted. Upper Picture :20. Lower Picture x100. After Boyles (10). 15 are restricted by the amount, size, and orientation of the primary dendrites of austenite. h. Segregation takes place in two ways--~first, between the primary dendrites and the liquid, and second, in the solidi- fying eutectic between the core and the radially growing boundary. 5. Graphite flakes start to appear with the gradual freezing of the eutectic and the flakes increase in size and number down to the point of complete solidification. In high purity Fe-C-Si alloys the amount of both carbon and silicon profoundly influences graphitization. Larger amounts of carbon not only increase the amount of eutectic at the expense of primary austenite, but also increases the readiness and amount of graphitization. Silicon is generally considered to replace one-third of its own weight in carbon and to lower the solubility of carbon in austenite. The presence of silicon increases the eutectic temperature and extends the range of temperature during which the eutectic freezes. Although the presence of silicon facilitates the process of graphi- tization, it is probably not indispensable. It is readily seen that carbon and silicon together determine the relative amounts of primary austenite and eutectic and that the influence of those elements that segregate into the eutectic is reduced for the higher compositions. The presence of more primary austenite leads to a very restricted interdendritic spacing for the formation of graphite flakes. For cast irons containing higher amounts of carbon and silicon, and consequently fewer primary dendrites, the graphite flakes are characterized by random orientation and comparatively 16 unhindered growth. Boyles found that the graphitization of pure Fe-C—Si alloys always resulted in type D graphite because of the strong tendency of the eutectic to graphitize, and he showed that it was the presence of sulfur, manganese, and hydrogen that made the commercial cast irons solidify so differently from the pure ternary alloys. Sulfur and hydrogen are carbide stabilizers and the introduction of either one of them into pure alloys resulted in an increase in size of graphite flakes. There was an optimal carbide stabilizing action which resulted in a maximum.f1ake size, and additional concentrations of sulfur or hydro- gen thereafter decreased flake size and eventually resulted in white cast iron. Supposedly the carbide stabilizers acted to increase flake size by their ability to delay graphitization by making the carbide more persistent. This prevented a rapid graphitization. Too great a car- bide stability'however sharply reduces the availability of carbon for the formation of flake graphite. Both sulfur and hydrogen are reported to segregate in the eutectic where they exert all their influence. Manganese considerably modifies the effect of sulfur. A complex series of FeS and.MnS solid solutions are fonmed. The influence of sulfur on graphitisation.when thus excluded from.the freezing eutectic because of the presence of manganese, is not definitely'known. Thus three elements---sulfur, manganese, and hydrogen---changed the high purity'ternary alloys into cast irons. Boyles investigated this trio rather thoroughly and found that their manipulation could produce any type or combination of types of graphite flakes. 17 Another class of elements are used to affect graphitization. These elasnts are generally added late as ladle additions and, although added in very small amounts, have the ability to bring about a remark- able change in graphitization. These elements are either graphitizers or carbide stabilizers. Why certain elements stabilize the carbide and others promote its decomposition is still an unresolved question. Bol- ton (7) has suggested the following possible ways in which a ladle addition may act: 1. It may form a mixed carbide more stable then cementite. 2. It may form a mixed carbide that is very unstable and which may accelerate the decomposition of all carbides. 3. It may affect eutectic carbon concentration. 1;. It may affect the eutectic temperature. 5. It may obstruct- atomic movement in the matrix. The effect of some elements on the persistence of carbide is not consistent and some elanents which are generally regarded as a carbide stabilizer under one set of conditions may act as a graphitizer under another set of conditions. The influence of boron, bismuth, tellurium and many other elements remains ambiguous. Boyles observed that the immediate factors that influenced graphite flake size were the rate at which graphite became available and the rate of solidification. By maintaining a constant rate of slow cooling, and varying carbide stability by chemistry, Boyles was able to change the rate ofcarbon availability and thus obtain a wide range of flake sizes. Thus it is contended by Boyles, Morrogh, and others that the 18 conditions for the formation of fine graphite flakes can occur in two ‘ways. First, when a very high nucleation rate and a high carbon avail- ability are present, a simultaneous formation of numerous centers of graphitization throughout the melt results in closely spaced, small, underdeveloped flakes of graphite. This contention is justified by the observation that high purity'alloys of Fe-Si-C, Ni-C, and 00-0 always show finely dispersed graphite, and that this condition persists until a carbide stabilizer in suitable amount is added. The second condition for the formation of small flakes of graphite occurs if an excessive persistence of carbide causes the growth of the flakes to lag behind the advancing front of solidification, whereupon the flakes are pinched off becausezthe rate of solidification exceeds the rate of availability of carbon or of decomposition of canentite. In the case of extreme stability of the carbide it may be decomposed very abruptly at lower eutectic freezing temperatures, or may even persist indefinitely to produce white cast iron. The stability of carbide is sensitive to thermal treatment as well to chemical variations. B. Thermal Factors Affecting Graphitization Thermal factors that operate during the freezing of the melt also affect graphitization during the solidification of the eutectic. Some of these closely interrelated factors are: l. Superheating temperature. 2. Time at maximum temperature. l9 3. Pouring temperature. 1:. Rate of cooling in the mold. 5. Temperature of start of graphitization. 6. Duration of graphitizing period. 7. Tanperature at end of graphitization. 8. Amount oftundercooling of the melt. Superheating of the melt results in a fine type D graphite. Piwowarsky has suggested that superheating causes a breakdown of the molecular carbides in the melt thus greatly increasing the urge to graphitize. Studies by Zapffe (159) and Morrogh (31) have shown that the fine graphite which results from superheating is caused by an increased absorption of hydrogen by the melt. Perhaps the rate of cooling of the melt and particularly of the eutectic is the most important thermal factor, for it in turn seems to control other conditions such as the temperature at which graphite starts to form, the duration of graphitization, and the extent of undercooling. Generally speaking the greater the rate of cooling, the lower the initial temperature of graphite formation, the shorter the duration and the more severe the undercooling. A compilation of conclusions about the effect of rate of cooling on graphitization based on the work of Schneidewind (hO) , Bolton (7), D'Atlico (12), Piwowarsky (37), and Morrogh (fl) is given below: 1. The rate of cooling affects the temperature of the eutectic arrest. Very slow cooling may result in a eutectic temperature as high as 1200'0 (2190'F) whereas with fast cooling it may occur at 1000.0 (1832'F). 20 2. The greater the rate of cooling and the lower the arrest temperature, the shorter the duration of graphitization, and the finer the graphite pattern. 3. The higher rates of cooling result in more severe undercooling. b. 'With extensive undercooling the growth of graphite flakes is hindered, possibly by the viscous nature of the melt. 5. Very high cooling rates produce a metastable product. o. ‘Hith gradual variation in cooling rate, a gradual and ordered change in the graphite pattern takes place. 7. For a particular iron every type of graphite is formed at a definite temperature. Raising the temperature of graphitization yields longer and coarser flakes whereas lowering the temperap ture of graphite formation results in finer and.more numerous flakes. 8. The eutectic freezing of cast iron can'be made to occur over a wide period of time and over an extended range of temperature. The above conclusions, although compiled from.various sources, are in agreement and emphasize the fact that graphitization depends upon time-temperature conditions. From these conclusions it is seen that carbide stability brought about by control of cooling rate progressively decreases flake size. The graphite flakes continue to get smaller until undercooling is sufficient to allow only finely dispersed type D graphite. The limit of undercooling is the complete suppression of graphitization in any fonm and the formation of white iron. Bash (13) and Merrogh (31, 33) have shown that type D graphite results from.decomposition of carbides immediately after solidification 21 of a white eutectic. 0n the other hand Hultgren (26) supports the possibility of direct formation of graphite from the melt by eutectic reaction. Little information is available about the carbide phase in cast iron. It is evident however, that the persistence of the carbide is a factor of major importance in the process of graphitization. A study of the literature indicates the possibility that during the freezing of the eutectic either: 1. Direct graphitization may proceed without the formation of a carbide phase. This may be true with high-carbon, high— silicon irons containing only small amounts of stabilizers, or even in the presence of usual amounts of stabilizers when these are diluted by a large amount of eutectic as in near- eutectic alloys. (Hanemann (18); Boyles (9); Piwowarsky (37); Hultgren (26) .) 2. The formation of graphite may be preceded by the formation of carbides. (Eash (13); Morrogh (31.. 33) .) In summary it may be said that graphitization is greatly affected by the relative persistence of the carbide phase and that this persistence can be altered by chemical variations, by variations in cooling rate or by simultaneous manipulation of both factors. Available evidence indicates that increasing the cooling rate, lowering the eutectic temperature, constricting the range of temperature of eutectic freezing, reducing the duration of graphitization and increasing the persistence of carbides by chanical control, all tend to promote finer graphite flakes in an orderly manner starting with type A and proceeding to type B, E, andDe 00. Mei . .3.) \ IV. WTAL PROCEDURE Part A. - Rocking Furnace Heats The experiments discussed in Part I of this report were made in an indirect-arc rocking furnace of 250 lb. capacity. The cold.furnace ‘was charged with pig iron, low-carbon steel scrap, and 27% ferrosilicon. The composition was adjusted inmediately after the meltdown by an appropriate addition of 80% ferromanganese, 2 5% ferrophosphorus, and SDZIiron sulfide. The analysis of the pig iron used was: 1 c Si Mn P : 5 Lot 1 14.15 1.18 0.72 0.128 0.039 Lot 2 14.12 1.23 0.88 0.21. 0.033 A typical charge consisted of: 110 lbs. of pig iron, Lot 1 143 lbs. of pig iron, Lot 2 61 lbs. of steel plate 12.7 lbs. of 27% ferrosilicon 0.7 lbs. of ferromanganese 0.6 lbs. of ferrophosphorus 0.2 lbs. of iron sulfide All of the melts were heated to a temperature of 2893'1". An optical pyrometer reading was made through the gout of the furnace. After tapping the heat into 3) lb. pre-heated ladles, the metal was 22 23 allowed to cool to the pouring temperature of 2650-2675‘F. The ladle additions of metallic aluminum and calcium were made after securely wiring the active metal to a l/2 inch diameter steel rod about 6 ft. long which was bent at 90' some 1 ft. from the end. The active metal was then plunged quickly under the surface of the ladle and agitated until all. reaction ceased. Blank ladles were treated with a 5 inch length from a similar rod. Some two or three minutes elapsed between tapping the furnace and pouring the molds. Metal from each ladle was cast into four or five standard 1.2 inch vertical test bar molds. Chill specimens were poured in pairs, either as small wedges or as chill blocks. The data on physical properties was secured from the transverse breaking load on 18 inch centers and from the deflection. The greatest effect of inoculation is noticed at the surface of castings, the break- ing of the unmachined transverse bar is thus a sensitive way of revealing effects of inoculation. Carbon equivalents ($0 + 1/3% Si) of cast irons compared in this paper will usually not vary more than 0.1%. This is a measure of con- trol that compares favorably with that of similar experiments reported in the literature. The amount of carbon and silicon was determined for each ladle, and the analysis for manganese, phosphorus, and sulfur was made for each heat or, frequently, for each ladle if subject to variation as a result of inoculation. No attanpt was made to determine the mounts of active metal retained in the cast iron as a result of inoculation. Work reported in the literature has given residual magnesium and calcium 2b as low as 0.008%. (3) Retention of any considerable amount of active metal is apparently not necessary in order for the effect of an inocu- lant to be quite pronounced. Part B. - Induction Furnace Heats In order to proceed with a second series of heats made in a small 20 kw. induction furnace, it was first necessary to produce some melting stock. To this end the rocking furnace was charged as follows: 90 lbs. of Pig Iron, Lot 1 93 lbs. of Pig Iron, Lot 2 15 lbs. of low carbon steel strip 2.2 lbs. ferrosilicon (27% Si) 1.0 lbs. ferromanganese (80% Mn) 0.22 lbs. iron sulfide (50% s) The analysis of the pig used was: Analysis Pig Iron % c Si Mn P 5 P18 Imn, LOt l hels 1e18 0072 Del-28 0e039 Lot 2 b.22 1.67 0.1.8 0.19 0.018 The 200 lbs. of metal in the rocking furnace was poured into four 50 lb. cylinders about 14 inches in diameter and 20 inches long. The analysis of the slugs thus made was found to be: i C Si Mn P 5 $1118 3-TOP 3e67 1e56 -- .- 0e059 Slug h-Bottom 3.66 1.51 0.8145 0.152 -- Slug S‘BOtton 3.9 1e78 "" -" .- Slug 6-Bottom 3 . 57 l. 72 - - -.. 25 Sections were cut from.the large cylinders of melting stock in order to charge the induction furnace. The cold charge to the induction furnace consisted of all of the ingot iron and ferrosilicon, and part of the melting stock. A typical induction furnace charge of 31 lbs. consisted of: 12.06 lbs. slug 3 12.06 lbs. slug h 1.30 lbs. ferrosilicon (27% Si) 5.60 lbs. ingot iron punchings 0.09 lbs. ferromanganese (80% Mn) 0.06 lbs. ferrophosphorus (25% P) 0.03 lbs. iron sulfide (33% S) is soon as the original charge had melted sufficiently, the remainp ing melting stock was added. Additions of the ferromanganese, ferro- phosphorus, and iron sulfide was made as soon as complete melting of all the chargehad occurred. All heats were brought to a temperature of 2850'F. and were then poured into a small hand ladle. The blank heats each had five inches of iron rod stirred into them until.melted, and.the inoculated heats had the active metal added as described previously. Pouring temperature was 2650’F. by optical pyrometer measurement. There was poured from each induction heat two wedge chills, two block chills, and three standard transverse bars as in Part A. In addition a small cylinder one inch in diameter and five inches long was poured. The cylinder contained a pair of chromel-alumel thermocouples and a dual record was made on the strip-chart of a high speed electronic recorder of the progress of solidification as indicated by latent heat evolution 26 at both thermocouples. Chemical analysis of the various irons was made Part C. - Physical Properties 1. Transverse Strength The dry sand molds used in this work were made from a well-mixed aggregate of lake sand, cereal, water, and linseed oil. The core sand molds were baked at h7S'F. for four hours. The transverse bars were washed with a non-carbonaceous silica slurry and again baked as before. The other molds used were not washed. Standard transverse bars 1.2 inches in diameter were poured from each heat. The bars were cast in vertical core-sand molds and were wire-brushed before testing on 18 inch centers. The diameter of the bar and the deflection at the breaking point was measured. From these data the corrected breaking load was calculated and the triangular resilience determined as 1/2 (corrected breaking load) (deflection). 2. Chill Tests (a) Wedge chills Two wedge chill specimens per heat, each about )4 x 2 x 1/2 inches, were poured into core sand molds. The pair was broken at a similar section and chill data based on the average depth of clear chill (white) and total chill (white and mottled) measured in 1/32 inch units. (b) Block chills Two block chills each about L. x 2 1/8 x 3/14 inches, were poured for each heat into a core sand mold containing a heavy metal chill face. Pairs of block chills were broken at comparable positions and chill 27 data taken as the average of clear and total chill in 1/32 inch units. 3. Photomicrography (a) General Small specimens were taken.near the fracture of representative transverse bars so that the microstructure could be studied from the surface into the center of the bar. The samples were processed on coarse and fine grit dry abrasive paper followed by'a fine abrasive used.in water suspension on a'wax*wheel. Final polish was on a silk cloth using AB Metpolish.#1 for hard metals. The etchant was either 2% nital or 11% picral. The etched samples were dipped into boiling water and immediately dried. (b) Cell size Cell size was measured at 25x by comparison of a heavily etched sample with a chart offered.by'Adams (1). This procedure requires a heavily etched sample in which the cell size is outlined by the lighter areas of steadite. (c) Heat tinting Certain samples were given a final polish and a very light etch and were then placed in an electrically heated muffle furnace at too- SDO°F. for l to 8 hours. Time and temperature of heat tinting were necessarily varied for each different sample. h. TemperatureIMeasurement Cooling curves were recorded on the strip chart of a high speed electronic recorder. The recorder had two scales which were calibrated in millivolts and which operated independently each with.its own.ampli- fier system at a response rate equivalent to some 1000'F. per second. 28 Two chromel-alumel thermocouples of 22-gauge wire were attached to the recorder by means of extension lead wires. The thermowuples were insulated by a fine-bore porcelain tube and were enclosed in a somalhat larger tube of fused silica (Vitreosil) . The welded hot junction of the thermocouple was protected by a thin layer of alundum refractory cement which was applied wet and thenbaked at a bright red heat. ' The recorder and thermocouple were calibrated by comparison with the freezing point temperature of high purity zinc, aluminum, and copper. The recorder was also compared at intervals to a portable millivoltmeter potentiometer wl'dch had been checked with the melting point of pure electrolytic copper using a noble metal thermocouple of platinum - platinum 10% rhodium for which a bureau of standards cer- tificate of calibration was available. Reproducible accuracy of the different chromel-alumel thermocouples, based on the noble metal thermocouple, the portable potentiometer, and the automatic high- speed recorder, was taken as -_r S'F. 5. Chemical Analysis Chemical analysis of the various irons for carbon, silicon, manganese, phosphorus, and sulfur was made by standard methods. V. DISCUSSION OF EXPERIMENTAL DATA AND RESULTS Part A Rocking Furnace Irons 1. Aluminum Additions. In order to determine the effect of aluminum as a ladle addition, two ladles of molten cast iron were inoculated.with 0.07% and.0.5% aluminum respectively and two additional ladles were untreated to serve as blank irons. 1 tabulation of results of he at T-Bh is given in Table 1. The chemical analysis of'the iron was not altered by the addition of aluminum. A comparison of the two blank irons and the two irons inocu- lated with aluminum indicated that no significant changes occurred in the transverse strength, deflection, or triangular resilience. However, for the iron inoculated with 0.5% aluminum, therewas a marked drop in clear and total chill. A cell size of 2-3 was found for both the aluminum.treated irons and the blank irons. A study of the microstructure of the transverse bar near the point of fracture revealed that the iron inoculated with 0.07%ialuminum had a similar'microstructure to the blank irons. Both irons showed type D and E graphite at the surface, with type A increasing in amount toward the center of the bar. The ladle addition of 0.5% aluminum resulted in a more abnormal.microstructure than found in the corresponding 29 mum mum mum mum seam sees mm\a Hamo mmuqa Hmum omaaa 0-~ fiasco seeam 0mm 00m new man .Heem NHN.0 mmgm «Hm.0 mega HHN.0 megs ama.0 omnm .sa .moH .Heeo .nesna H00.0 u- a0.0 an.“ «0.0 400.0 ma.0 00.0 mm.m s0.~ m00.0 .. a0.0 0N.N 40.“ 400.0 NH.0 00.0 am.m N0.m nlhw m. u mmmwams< «m o ansam aeeam dd so.o H4 m.o u. mafipducd ans soaH somehow woflxoom Ho moavudsoocH Edcflsaad so same A mqmde mwnama mausma «NIJMB «dizma e02 soaH 31 blank iron, and the inoculated iron was definitely type D and E graphite at positions on the diameter where the blank iron was partly type A mixed with type D and E. Both samples tended toward type A graphite at the center. A study of the microstructure at 5001: showed no apparent difference in matrix structure. Both samples showed little or no ferrite associated with the type D and E graphite. 2. Calcium Additions. A study of the effect of calcium as an inoculant was made in heats T35, T39, and Th0. A 200 lb. heat of cast iron was melted in the rocking furnace using the same charge as for heat T311. Results for the ladle addition of 0.8% calcium as compared to the blank iron are shown in Table 2. A loss of carbon resulted from the addition of calcium metal to the molten cast iron. A marked decrease in chill was also obtained. An additional study of calcium inoculation was made in heat T39. There was added 0.75% and 1.0% calcium respectively to a pair of ladles for comparison with a pair of untreated ladles from the same heat. Results are shown in Table 3. Again the addition of calcium served to decarburize the iron and produced a marked decrease in chill. Heat Th0 was run to obtain information about the effect of a small addition of 0.1% and to secure some blank irons of comparable analysis to the calcium inoculated irons of heat T39. The addition of the smaller amount of 0.1% calcium resulted in only a small decrease of carbon in the iron as shown in Table 1:. A definite drop in chill was obtained. Table 5 presents data for comparison of three cast irons inoculated with different amounts of calcium with two blank irons of similar analysis. Iron T39-lC, which was inoculated with 1.0% calcium, showed a microstructure Iron Nbe TBS-c T35-B Iron NOe T39-lc T39-20 T39-IB T39-2B Iron No. Tho-10 ThO-ls Tho-23 ThO-BB 32 TABLE 2 Data on Calcium Inoculation of Rocking Furnace Iron T35 Addition 0.8% Ca Blank SC 2.68 2.78 Block 7681 %S Chill 1/32 inch 2.h0 0.063 h-8 2.38 0.060 18-33 TABLE 3 Data on Calcium Inoculation of Rocking Furnace Iron T39 Block Addition Analysis % Chill % 0 Si Mh P 5 1/32 inch 1&0 Ca 2072 zehé 0087 0e19 OeOhZ 2-3 0.75 Ga 2.7h 2.h8 0.88 -— 0.0h0 h-6 Blank 2.91 2.h5 0.86 -- 0.053 10-22 Blank 2.85 2.51 -- -- 0.050 13-27 TABLE 14 Data on Calcium Inoculation of Rocking Furnace Iron Th0 Block Addition % % % % Chill 0 Si Mn 5 1/32 inch 0.1% Ga 2.70 2.h3 0.83 0.053 5-12 Blank 207h 20h? 0e86 OeOSh 12-28 Blank 2e72 2.h8 Oe86 .. 1h.2h Blank -- 2 ehé - -— 9-1h 33 m m mu4 4 mn4 onsm eons ~m\a Hare Ham mmw, 4N4 mum mam .Heem umH.o HmH.o mmm.0 H4M.0 mam.0 9G.“ .ooaeoa mm4m 00mm m4mm 00mm mamm .noa .mcaae .. n. 00.0 04.~ ma.~ 4m0.0 .. 00.0 a4.~ 4a.~ mm0.0 nu no.0 M4.~ 0~.m 040.0 .. .00.0 04.m me.~ «40.0 0H.0 50.0 04.N «5.0 m a. u esmwwona mm 0 seaam asaam so 0H.0 so m~.0 so oo.H m essences meoaH oomcnsm mcfixoom Ho soapaasoocH sawoamo do «use o>dpsnddsoo m mqm4a 00-040. mHu04a 0H-04a 00-0me oatmma eoz soaH 3b of type A graphite from the center to the very edge of the transverse bar. Iron TLO-lC, to which a ladle addition of 0.1% calcium was made, had type A graphite predominating and had almost as good a microstruc- ture as iron T39-lC to which 1.0% calcium was added. The corresponding blank irons, Th0-1B and Th0-2B, showed considerable type D and E graphite particularly near the surface. Cell size of the calcium treated irons was h-5 in all cases as compared to 3 for the blank irons. As shown in Table 5 all of the cast irons inoculated with calcium showed a marked reduction in; chill and superior results for the transverse breaking load, the deflection, and triangular resilience. Data on the relative effectiveness of aluminum and calcium as inoculants is shown in Table 6. The ladle addition of 0.5% aluminum did not alter the chemical analysis of the iron and had no beneficial effect on microstructure or physical properties except to sharply reduce the amount of chill. The use of calcium as an inoculant resulted in a marked improvement in microstructure and an increase in physical properties. Cell size was reduced and chill depth was lowered in calcium inoculated cast irons as compared to the corresponding blank iron. Calcium additions to molten cast iron resulted in some decarburization of the cast iron and in a reduction of sulfur. Table 7 compiles the available information on the effect of ladle additions of calcium on the carbon and sulfur content of cast irons. Generally speaking an increase in the amount of calcium used in the ladle addition resulted in a 13' ger loss of carbon and sulfur. The use of 1.0% of calcium as an inoculant, for example, resulted in a drop in carbon from 3.0h% to 2.814% and decrease in sulfur from 0.065% to 35 ;:01 43W\ .36“ \I U 0:0 0-0 .000 noes 00\H :20 sooam Have 40-40 0:0 00-00 0-4 40-40 00-0 00-40 0-0 H40 000 000 000 040 404 000 000 .Hmom 000.0 mam.0 HmH.0 040.0 000.0 000.0 000.0 400.0 03H .Hmon 0040 0000 0000 0000 0040 0400 0040 0040 .ena .eoaaa 00.0 04.0 00.0 040.0 00.0 00.0 04.0 00.0 400.0 00.0 04.0 40.0 040.0 00.0 04.0 00.0 00.0 04.0 00.0 000.0 00.0 04.0 00.0 000.0 00.0 00.0 00.0 400.0 0H.0 00.0 00.0 00.0 0 0 oz. Hm-l, 0 u mammwac< meoaH oossnsm mcwxoom do 2500000 new assassad mo poemmm wswpsasoosH mo somfiasmsoo 0 mum4a xdmam mo 0.0 0500 00 m0.0 Madam mu H.o xnsam H4 m.0 a. nowpfivcd mwloae OHImma 00-040 00-000 00-040 00-040 001400 mmo 023.000 Ito .0; awhizz‘ Z. WET». mmm emo.Il mmo.I m¢__r eom_I .w , w,._ wv;; , M .I 00mm 00 _ r _ _— __ _ mo 143 Nwmu Nm0_ no? may: ¢0N_ 0. $32.23.; 1.22 324302. zom. emqo mo... mm>mno ozjooo I e. .2... WMPDZIA Z. .32.... l j _ _ _ _ . . Iéikl .wIimii _ , ; loom. 1009 I OOON I 00_N .-Ioomm M .220440 IFIS owbdjzooz. 20m: Fm<0 mom mm>m30 023000 l.m .0; mwh32_2 Z_ MIC. Nmm I. 1009 hMO. I I00m_ mac. l l OOON m¢: l IOO_N VON. l l 00mm M The use of pure silicon as an inoculant for cast iron in heat T55 produced a cooling curve very similar to the blank irons except that no undercooling was recorded. The temperature at the start of the eutectic reaction was 1980°F. and the curve held horizontal at constant temperature for some to seconds. In connection with a study of graphite as an inoculant, cooling curves on a second pair of blank irons, L7 and L8, were determined. These curves are shown in Figure 9. An initial eutectic temperature of about 1990'F. was recorded along with an undercooling of 5'-IO°F. The duration of the formation of the eutectic in the blank irons was again about 90 seconds. Figure 10 shows the cooling curves obtained from a series of four heats, Lb, L5, L6, and L9, in.which fine graphite was used as an inocu- lant. The initial temperature of eutectic formation was increased to the range 2020’-2050°F. The duration of the eutectic reaction was about 70 seconds. A summary of the results of cooling curves obtained is shown in Figure 11. These are typical or average curves‘based on the studies Just discussed. A tabulated summary of the effect of each inoculant on the initial temperature of eutectic formation and on the duration of the entire eutectic formation is given in Table 9. These data show that calcium was the most effective inoculant in raising the range of temperature of eutectic solidification. Graphite ‘was also effective in this way but not to quite such a marked degree as calcium. Aluminum.was less effective than either the calcium or graphite, although better than silicon which showed.littls or no effect on the temperature of eutectic fermation. lI6 ammEmw 02083 .22: emqo ou...<.5002_z: mod 3393 92.600 lime; mwFDZ.2 2.. m5; a i _ _ _ _ \Im0_ l mmO. I l mu: l ¢0N_ o. r r a p e a J 00m_ 099 000m 00_N OONN no lI7 .m..r_ra<7.0 m2; It; awkd.43002_ 20m: Hmdomou mm>m30 02.4000 l0. .0: wmwnzi 2. m5; 1:8 .2835 $3.0 5.3 Beanaoé .mtxddmo m2: some :5; 3.5309: .220an $0.. :23 82.382. 532334 wood It; $2.382. -nnEneooE eozdemm ozoommwdm: xzfim Tm .8538; 52 €qu xzqnm Tn .353 03-58 32;: no zomiquoo I: o mmknzrz 7: mete 4000) l _ l L L a _ I... Comparison of Effect of Ladle Additions of Aluminum, TABLE 9 Silicon, Calcium, and Graphite on the Initial Temperature Iron No. T-hS T-h6 T-h6 T-h? T-SS T-Sl T-h8 T-h9 L 7 L 8 L h L 5 L 6 L 9 of Eutectic Formation of Induction Furnace Irons. Addition % Blank Blank Blank Blank 0.55 Si 0.6 A1 0.6 A1 1.0 Ca 1.0 Ca Blank Blank 0.5 Cr 0.5 Gr 0.5 Cr 0.5 Gr 1970 1970 1970 1968 1980 2007 2015 20h? 2033 2057 20 57 2038 2038 1966 1986 1995 1990 2020 2020 zoho 2050 2055 2035 2037 Duration of °F at Start Eutectic of Eutectic (Seconds) 60 50 90 55 L2 Ave. °F 1970 1980 2025 20h8 198k 2039 Ave. Seconds b2 76 b9 SO 2. Microstructure and Physical Properties. The blank cast irons of heats Th5, Th6, and Th7 were very abnormal in graphite distribution. Figure 1b is typical of this type of micro- structure. Very little type A graphite was observed, even at the center of the transverse bars. The physical properties of these blank irons are listed at the top of Table 10 and a summary based on average results is shown.in Table 12. The blank irons as a group showed.a large amount of chill and were relatively low in transverse strength, deflection, and triangular resilience. Heats T50 and T51 to which.0.6% aluminum was added as an inoculant showed a microstructure very similar to the blank irons. No improve- ment in graphite distribution resulted from the ladle addition of aluminum. The physical properties, as tabulated in Table 11 for com- parison.with corresponding blank irons, were not significantly changed. The addition of 1JDZ calcium as an inoculant to heats Th8 and Th9 produced a cast iron having type A graphite predominating throughout the cross section of the bar. Figure 13 is typical of this type of microstructure. Table 11 shows that the physical properties of the calcium treated irons are definitely superior to those of the correspond- ing blank iron of similar chemical analysis. The physical properties of an iron to which a late.addition of 0.55% silicon was made were not appreciably different from the proper- ties of a blank iron of similar analysis. Table 12 is a summary of average results for the blank irons and for the irons to which aluminum, calcium, and silicon were added as inoculants. Table 13 is a comparison of carbon equivalents of the several inoculated irons and the corresponding blank irons. 51 mwm ~m0_ mm0_ 0v: #0.? m .JQPMZ $20440 o$0.. It)». QMHQJDQCEI 0 .omeqnzoozi aoz 2m: urchin waiwzcrm modem; QmIOZMDO mowed} mom mu>m30 OZSOOQIN. .0; mMHDZ¥¢ Z. MIC. T TI n a q a _ a 00 0. 00m; 000m 005 Figure 13. Typical Microstructure of Type A Graphite Distribution. Iron T5314 Inoculated with 1.0% Calcium and Cooled in a Dry Sand Mold. 1100 53 Figure 1h. Typical Microstructure of Type D Graphite Distribution. Uninoculated blank heat TSlIM of same analysis as iron in Figure 13. Cooled in a Dry Sand Mold. x100 00-00 00-00 man» mum mane 00-0 0010 00.00 00:00 00.00 noon 00\4 adage .3on 000 NAN one 000 mmm 000 mom 000 000 000 .Hmom Hom.o 00H.0 00m.o NNN.O mma.o 004.0 000.0 000.0 000.0 000.0 0““ .HHon 00:0 00mm 0000 0000 0000 0000 0000 0000 0000 00:0 .mpa .mcwne 000.0 00.0 00.0 000.0 00.0 00.0 000.0 00.0 00.0 000.0 00.0 00.0 00.0 000.0 00.0 00.0 000.0 00.0 00.0 00.0 000.0 ma.0 40.0 000.0 00.0 00.0 00.0 000.0 00.0 00.0 00.0 000.0 00.0 00.0 00.0 000.0 00.0 00.0 00.0 000.0 00.0 00.0 00.0 m. .0 m nammmane am .0 00 0R0 Anooosov madam 300003 00 09H dd 0.0 H4 0.0 .0 0.0 .0 0.0 acdam anaam madam n. 830.2 mm0 ame 000 Hma moms nme 0000 may 000 000 000 000 .02 cosH soaH Ammo oomcnsm nowposocH mo cowpmasoosH soofiafim 000 «ESQHESH4 .Efiwoamo no spam 0>prnmeoo n: mum