FERRO filLECON, CALCEUM SILSCON, AND METALLIC CALCEUM A5 LADLE {NOCUL‘AN"?S FOR GRAY SRON Thesis for the Degree of M. S. MICHEGAN STATE COLLEGE Harman C. McCiure 19529 This is to certify that the thesis entitled FERRO SILICON, CALCIUI SILICON, AND M’E‘l‘ALLIc CALCIUII AS LADLE INOCULANTS FOR can IRON presented by Norman C. IcClure has been accepted towards fulfillment of the requirements for maegree mMr. Major professor Date “3y 26) 1949 - ._-—.—_-——- _- ‘- §\ FERRO SILICON, CALCIUM SILICON, AND METALLIC CALCIUM AS LADLE INOCULANTS FOR GRAY IRON By Norman C Egalure ' A THESIS Submitted to the School of Graduate Studies of Michigan State College of Agriculture and Applied Science in partial fulfillment of the requirements for the degree of MASTER OF SCIENCE Department of Mechanical Engineering 1949 .Iov— ACKNOLEDGEMENT The author wishes to eXpress his sincere thanks to Professor Howard L. Womochel for his technical advice and guidance which made it possible to conduct the research reported in this thesis. The author also wishes to take this Opportunity to thank the many people who willingly contributed their time and effort during the course of this investigation. 217858 Table of Contents I Introduction Page 1 II Survey of Published Literature A - Definition of Inoculation 3 B - Purpose of Inoculation 3 C - Value of Inoculation 4 D - Effects of Inoculation 5 E - Theories Concerning the Mechanism of Inoculation, 6 F - Data Concerning the Influence of Inoculation on Physical Preperties 15 G - General Statements About Inoculation Practice 23 III Crigional Research A - Scepe of Investigation 25 B - Melting Practice 1 - Preparation.of Holds 27 2 - Estimating the Charges 27 3 - Charging the Furnace 29 4 - Melting and Superheating 29 5 - Inoculation.Prcceedure 30 6 - Checks on Pouring Temperature 31 7 - Casting Proceedure 33 C - Investigation.Proceedure and Results 1 - Chemical analysis 55 2 - Transverse Strength.and Deflection 38 3 ' Chill Depth 41 tn -4 ox \n -> I l Brinell Hardness Tensile Strength Microscopic Examination Microphotography Condensation of Data, and Curves for Physical Properties D - Discussion 1.. 2- O\ U! -F \N I Chemical Composition Transverse Strength and Deflection Chill Characteristics Hardness Tensile Strength Graphite Distribution and Microstructure E - Summary and Conclusions 44 45 46 52 72 73 75 77 77 78 80 List of Figures ~From Survey of Published Literature-t Fig. 1 - Representation of Graphite formation, cooling rate, and transformation range on time temperature diagrams. (6) Page 8 Fig. 2 - Physical test data for 2850 F super- heat (10) 16 Fig 5 - Physical test data for 2950 F super- heat (10) 16 Fig 4 - 3.25% Ca - 1.22% 31 mottled base iron (25) 19 Fig.5 - 3.26% C - 1.81% Si base iron ( 25) 19 Fig.6 - 3.03% C - 2.21% 81 base iron, heat A (25) 29 Fig.7 - 3.03% C - 2.21% 81 base iron, heat B (25) 20 -From Origional Researchv Fig.8 - Approximate analysis of steel and pig irons 28 Fig.9 - Sample calculation for computing charge, heat 1 28 Fig.10 - Chemical analysis 56 Fig.1l - Description of Code Letters and Numbers 37 Fig.12 - Transverse load and deflection for heat 1 38 Fig.13 - Transverse load and deflection for heat 2 39 Fig.14 - Transverse load and deflection for heat 3 39 Fig.15 - Transverse load and deflection for heat 4 40 Fig. 16 - Transverse load and deflection for heat 7 40 Fig. 17 -Chill depth for heat 1 41 Fig.18 - Chill depth for heat 2 42 Fig.19 - Chill depth for heat 3 42 45 315.20 - Chill depth.for heat 7 Fig.21 - Brinell hardness numbers (heats 1 3.4 and 7) Fig 22 - Ultimate tensile strength (heats and 7) Fig.23 - Graphite distribution and micro- struoture for heat 1 Fig.24 - Graphite distribution and micro- structure for heat 2 Fig.25 - Graphite distribution and micro- structure for heat 5 Fig.26 - Graphite distribution and micro- structure for heat 4 Fig. 27 - Graphite distribution and micro- structure for heat 7 Fig.28 - Condensed data on physical prop- erties for Fe-Si vs Ca-Si Fig.29 - Condensed data on physical prop- erties for base irons vs Ca additions Fig.30 - Transverse test results: plot of bending load(#) 4. deflection(') - (re-$1, Ca-Si, and Ca) Fig.31 - Curves for actual %81 pick up vs transverse load and deflection, also vs tensile strength - ( Fe-Si and Ca-Si) Fig.32 - Curves for actual % 81 pick up vs total chill, also vs % carbon equivalent v (Fe-Si and Ca-Si) Fig.33 - Curves for % carbon equivalent vs transverse load and deflection, also vs ’2, 3,4 Page 44 45 47 4e 49 50 51 61 62 63 64 65 ‘W B". *m tensile strength (Fe-Si and Ca-Si) Fig.34 - Curves for % carbon equivalent vs clear and total chill - (Fe Si and Ca-Si) Fig.35 - Curves for % silicon vs transverse load and deflection, also vs tensile strength ( Fe-Si and Ca-Si) Fig.56 - Curves for % Silicon vs total chill, also vs % carbon equivalent (Ca-Si and Fe-Si) Fig.37 - Curves for % Ca added vs transverse and tensile prOperties, also vs chill (Ca and blank) Fig.38 - Curves for % Ca added vs transverse and tensile properties, also vs chill and % carbon equivalent (Ca and Ca-Si) 66 67 68 69 7O 71 List of Microphotographs m1 - Type A graphite distribution - 1001 Page 53 M2 - Mixed graphite distribution, type A and type D - 1001 54 M3 - Type D graphite distribution - 1001 55 M4 - Surface - .5% Si pick up from Ca-Si - 250: 56 IS - Surface - .5% Si pick up from Fe-Si - 2501: 56 M6 - Surface - Base iron, heat 7 - 2501 57 M7 - Surface - .22% Ca added, heat 7 - 250: 57 M8 - Core - Base iron, heat 7 - 250: 58 M9 - Core - .22% Ca added, heat 7 - 2501 58 I. IBERQDUCEION The subject of ladle inoculation of gray iron has been brought to the attention of the foundryman with increasing emphasis during the past twenty years. A series of papers on "High Test Cast Iron” by Smalley, Marbaker, and Coyle and Houston,published in the Transac tions of the American Foundryman's Association 4 e - 1929 (l) (2) (3), seemed to create an interest in the subject of inoculation. Since that time many papers, articles, and discussions have been presented on this subject and several theories to explain the inoculating action have been formulated. Although many of the papers have discussed ladle additions of calcium silicon and/or ferrosilicon, there has been a conspicuous lack of data comparing the relative effects of these inoculants on txe physical prOpe ties of gray cast iron. Any mention in the lit- erature of ladle additions of metallic calcium has been devoid of accompanying data on ph_sical prOperties. The purpose of this investigation is to pres nt some data on the relative effects of calcium silicon, ferro silicon and calcizm metal on the physical prop- erties of gray cast iron. Since it would be possible * Please refer to articles listed in the Siblioararhy. 2. to approach this subject from more than one angle,it was decided that a comparison would be made on comnercial gray irons with similar final chemical analysis. It was hoped that, from the resultiné data, some definite relations could be established as a comparison of these inoculants and some additional information could be presented on the theories of inoculation. II. SUBVsY OF BTSIISHnD lite..fflfifi A- Definition of Inoculation The 1948 "Metals Handbook” (4) states that in general, inoculation may be defined as ”the addition to molten metal of substinces designed to form nuclei for crystallization." The purpose, value, and/or effects of cast iron inoc ulation have been listed in the literature and attempts have been made to describe the mechanism accompanying successful inoculation practice. B- Purpose of Inoculation The 1944 "Cast letals Handbook” (5), remarks that ” in many cases, 3‘9? irons are inoculated with late I I. additions for the purpose of improving the structures and, consequently, the mechanical preperties. Eany d ifferent elements and combinations, such as ferrosilicon, calc ium silicide, graphite, and numerous comhegcially prepared inoculants, are used for this purpose. In fact, practically any late adCition appears capable of CD producing some effect." Lownie(6) divides the various inoculants into two groups: 1- (Have sole duty of producing inoculating effects sucn as changing graphite distribuiion and reducing chill); Ca; Ca-Si, Ca-Si-Ti, Fe Si, graphite, Si-C, Si-rn, Si-rn-Zr, Si-Ti, and Si-Zr. 2- (Also exert the effect that a change in chemical composition, alloying, has on the preperties in add- ition to inoculation effects); Cr-Si-Mn-Ti-Ca, Cr-Si- hn-Zr, No-Si, and Ki-Si. Burgess and Bish0p(7) consider those liSted in group 1, above, as straight graphitizing inoculants and those in group 2 as stabilizing inoculants. In as much as this thesis deals with group 1 (graphitizing) inoculants, and most of the literature surveyed covered the same group, some typical analysis for that tyoe as listed in the "alloy Cast Irons Handbook” (8) will be given: tea its: to ,*.n Others ( is) CaKetal lOO Ca-Si 50-55 60—65 Ca-Si-Ti 5-8 45-50 ---------------- 9-10 Ti Fe-Si ------- 50-90 Graphite -------------- lOO Si-C -------45-56 28-50 Si-hn ------- 47-54 -------- 20-25 Si-Mn-Zr ------- 60-65 --------- 5-7 5-7 Ar. 0- Value of Inoculation "The value or efficiency of a ladle inoculant depends on its ability to consistently p rform certain functions when added as a late addition to cast iron."(7). The authors of this statement (Burgess and BishOp) enlarge uron it by stating that successful inoculation requires: 1- reduction of tendency to chill and 2- improvement of physical properties a- development of high tensile and transverse strengths in irons possessing suitable base character- isitics b- deveIOpmcnt of optimum transverse properties, impfict resistance, and chill reductions at any given tensile strength level. D- Effects of Inoculation Effects of ladle inoculation, according to Lownie,(9) may be attributed primarily to the effect of the inocu- lant upon the size, shape, and distribution of the graph- itic carbon in the iron. "It has been established by many investigators that the effect of the inoculant is to change the graphite from a fine, dendritically orignted pattern to a coarse randomly oriented pattern." Lorig(lC) says that ladle inoculation seems to be the only reliable method in which the microstructure of superheated gray iron can be kept from becoming dendritic. His micrOphotographs show absence of dendritic graphite and ferrite for Fe-Si treated irons as compared with the dendritic structure of the same irons without ladle treat- ment. "Noticable effects of ladle additions on structure are reduction in size and influence on shape and dis- tribution of graphite flakes." Smalley(l) reported that cupola iron which would have been white and brittle in small section, or of uncertain properties in heavy section, produced normal graphite with a pearlite matrix after ladle treatment with Ca-Si. Loria and Shephard(ll) found that silicon- carbide inoculation controlled the chilling tendencies in alloy- ed and unalloyed irons. The chill depth was reduced and the structure of the chilled surface for wear resisting castings was refined. "Ladle additions prevent bad supercooled structures, such as cellular eutectic graphite coupled with ferrite", according to Lemoine(l2) who used both Fe-Si and Ca-Si in his eXperiments. Boyles(15) states that ladle additions seem to in— crease the number of crystallization Centers in the eutectic and therefore produce a small cell size. This has the effect of creating a normal grpahite structure in gray irons that would otherwise solidify in the dend- ritic graphite pattern. 3- Theories Concerning the hechanism of Inoculation Vanick(l4) is inclined to believe that "the mechan— ism of inoculation has been traced to a reaction which, when executed in its prOper time and place, forces the crystallization of graphite in the stable system. Commercial inoculants usually possess a high order of graphitizing power which is supplied through the presence of graphitiz- ing elements plus deoxidizers and degasifers. Their im- portant function of stabilizing the structure of the graphite leads to the production of a more dependable range of physical prOperties for metal of the same composition. " Lownie(9) believes that the random graphite pattern is formed,during inoculation, by the iron—graphite system without freezing directly into carbide is a temporary phase. "Therefore the effect of tie inoculant is to pro- du r1 V e graphitization at the eutectic in those cases where graphitization would occur normally below the eutectic." He lists tgree theor es for inoculating action: ) u' 1- Gas Theory; belief that inoculation is real deoxidation and reations between inoculant and dissolved gaSses cause the inoculant ffect. Lownie is skeptical about this action because chromium, a strong deoxidizcr, (D produces carbides and not graphit 2- Silicate Slime The ry; belief that submicrOSCOpic slime of ferrous silicate inclusions acts as coarse graph- ite nuclei. This theory is supported by the fact that a sl fl} ,2)‘ "‘27) silicon removing promotes abnormal, fine graphite. 3- Graphite Nuclei Theory; belief that graphite particles in the melt act as nuclei to accelerate the z This eXplanation for the act- graphiti tion phenomenon. ion of an inoculant has been the most widely accepted I ‘nvr A. .L 14 he origional st tenent, nade in 192o U ‘- althoug Piwowarsky(37) -"Eoltcn cast iron contains gr Thite nuclei in suspension, which as the me 31 freazed or Jsets", form the starting p int of co»rse graphite crystals." - has be h extended in more recent discussions. .0 V L Lownie(6) postulates that the process of nucleation appears in all of the theories as to the cause of inoc- ulating effects and presents tae foll wing graph to dem- onstrate his explaination of the inoculation phenomenon: Type A Graph ta. Type 0* E. Grip,“ t‘ White Iron Time. -——-- Fig.1. Representation of graphite form tion, cooling rate and transformation range on time-temperature diagrams. The purpose of this simplified diagram is to provide an aid to understandting, in a practical manner, how inoculants work. The temperature horizontal B indicates the division between the regions for normal and for inter- dendritic graphite formation, and C divides the regions where abnormal graphite is formed and where graphitiza - ion is suppressed. Cooling rates are represented by the arrows A F, A K, and A L for fast, medium, and slow cool- ing, respectively. The curved, dotted lines represent the start (subscript s) and finish (subscript f) of transfor- mation for any particular cast iron. The composition and castine proceedure of the gray iron in question will dete nine the position of the dotted lines. Us and Uf are typical transformation curves for uninoculated, medium carbon-medium silicon irons. Is and If indicate the relative position of the transformation curves for similar irons that have been completely inoculated. When using this chart the section size ( rate of cooling), composition, and inoculation effects deter- mine what éraphite formation will be obtained. The transformation curves for a low carbon-low silicon iron would be found to tie ridht of Us and Uf and for high carbon-high silicon irons to the right. The former would require extremely slow cooling to form ty-e A graphite and the later e tremely fast cooling to form white iron. Type A graphite flakes, formed during long transformation periods at low cooling rates are longs and coarser than type D flakes formed during shorter transformation periods at faster rates. If the time consumed during transformation to t’pe A in inoc- ulated iron is similar to the time required for the uninoculated iron to transform to type D, althoréa the rate may be faster in the later case, the graphite flake size should be siailar. The presence of an increased number of graphite nuclei in the melt due to inoculation makes it much more 1 kely that graphite will be formed in the normal graphite region. Bash (15) reached the conclusion that "inoculation produces graphite nuclei which cause the iron to solidify in the iron graphite system. Gray irons, having fine 10. graphite in a dendritic pattern, solidify in the meta- stable iron-carbide system as white iron. The eutectic carbide subsequently decomposes in the solid state to form iron and grapcite. Ladle inoculated g ay irons, having random flake graphite, solidify in the stable iron- graphite system. The flake graphite forms during the freezing of the graphite-austenite eutectic. The iron— graphite eutectic temperature :as found to be 56 to 70°F higher_ than the iron carbide eutectic in slowly cooled heavy section gray iron castings." Park, Crosby, and Herzig (16) also ascribe to the theory that graphite forms directly from the melt along with the austenite as the eutectic temperature is reached. They believe that the austenite-cementite eutectic and the austenite-graphite eutectic must be very nearly similar in temperature and therefore hardto distinguish. Boyles and Lorig (l7) plotted cooling curves on a 3.0% C - 2.5% Si iron as inoculated with Ca-Si and after holding at temperature for an additional thirty minutes after the inoculating treatment. The solidification of the eutectic started at approximatel; 10°F lower for t,e iron which had been held for thirty minutes and the effect of seeding had worn off. The comparative shape of the i curves after tne start of the eutectic formation is of special significance. In the inoculated, which showed a normal graphite distribution, the temperature rose immed- iately and as a result the size of the flakes formed was larger almost immediately. The curve for the iron in which the inoculating ef_ect -ad worn off remained at the /// ll. undercooling temps ature for a longer period of time and leped pward more gradually. This iron had a mofified graphite strucsure due to the iormation of most of tne graphite Us fine flakes before the temp— erature had leveled off. Although the;e was no difference between the size or distribution of tge primary dendrites there were many more crystallization centers for t-e formation of graphite in the normal iron. This resulted in an evolution of heat in a greater number of areas in the normal iron with a resulting rapid upswing of the curve whicn allowed for the form t on of type A graphite. The modified iron showed a mrch larger area in the center of each crystallization point due to the greater distance between centers and the smaller evolution of heat in- volvcd. D'Amico and Schneidewind (18) worked out '8' curves to c ompare the start and finish for the austenite and eutectic transform tions at constant temperature for several cast irons. Of Special inte est was the comp- arison between a eutectic iron ( 3.745 C - 2 .103 Si ) with and without ladle inoculation using Ca-Si (.4% Si added). The above researchers have developed the theory that the probable graphite pattern in a given iron is determ ned by: 1- tie rate of cooling during solidification, 12. 2 - the degree of undercooling experienced by virtue of its cooling rate during solidification and, 3 - the solidification characteristics of the iron. With rapid rates of cooling, appreciable undercooling accompanies solidification. In other word, the metal. remains molten until a temperature well below the theoretical greezing point is reached; the faster the cooling, the greater the degree of undercooling. The graphite pattern is directly related to the degree of undercooling for any particular iron; the flakes become smaller and more dendritic as undercooling is increased. An 8 curve may be drawn to represent the part: icular solidific tion characteristics of an iron and irons with different compositions will have different S curves. Each branch of the curve represents a part- icular graphite formation and corresponding points on different curves represent similar graphite patterns; therefore different irons will have different graphite patterns with si ilar amounts of undercooling. The upper branch of the curve (down to the knee) represents the normal graphite region. The middle portion(around the gradually curved part) is distinguished by a eutecti- forn graphite in dendritic distribution. In the lower part, mottled and white irons are found. D'Amioo and Schneidewind are convinced that the in- fluence of factors (such as inoculation) which affect the nature of the graphite pattern would only be attributed to their ability to change the solificiaation characteristics of an iron exposed to any given amount of undercooling. The effect of adding .4% silicon as Ca Si on the 3 curve is to move corresponding sections of it to the right and downward to an appreciable extent. "It is interesting to note that deoxidation with .4% 03-51 has made it nec- essary to undercool iron B {eutectic composition) to 1450 F and 1750 F respectively, in order that white and mottled structures are obtained. The corresponding temp- eratures in the untreated iron are 1675 F and 1925 F. In other words, the tendencey to form gray iron, or the urge to graphitize, is much stronger in a deoxidized iron than in a plain iron of the same composition; the addition of a deoxidizer demands much more drastic under- cooling for the formation of the metastable carbide phase." It was determined that the addition of a deoxidizer does not raise its temperature of solidification and therefore this effect could not influence the graphite formation. The increased silicon content could not cause the extreme effects shown by the inoculant, as evidenced by comparing the 8 curves of base irons with similar differences in chemical composition. Flinn and Reese (19) postulate that the preperties of any gray cast iron are dependent on graphite distribu- tion and the structure of the matrix. They believe that the fine, eutectiform, network type of graphite "fromed from eutectic cementite at emperatures below the eutectic, while the well distributed, larger, flake graphite 14. is formed at the eutectic." A normal distribution may be obtained if .556 Si is added to the ladle just before pouring. "The effect of this addition passes away with time (about 20 minutes) and therefore appears to be due to high silicon spots which promote the formation of an iron graphite eutectic." Inoculation seems to lessen the possibility of free ferrite formation by causing graphit- ization at solidification instead of at lower temperatrues by graphitization of cementite to graphite and lower carbon austenite. Crome (20) also ascribes to the theory that type A (normal) graphite flakes solidify form the liquid while type D (abnormal) graphite forms from the ausennite-cement- ite eutectic solid. He thinks that although the graphite nuclei may be actually dissolved in the melt there may still be enough points of carbon concentration due to inoculation that normal graphite will form. Bash (15) is convinced that inoculation with silicon alloys is successful because high concentration of silicon exist immediately adjacent to the added particle so that the solubility of carbon in that immediate area is exceeded and hypereutectic graphite precipitated. These particles act as nuclei although the silicon might become uniformly dissolved in the melt. He believes that the removal of oxygen caused by ladle additions does not aid in the form- ation of normal graphite but would be more likely to pro- mote iron carbide formation with resulting eutectiform graphite e \ 15. Mc Elwea (Egg) considers ladle additions such as Ti-Al-Si to be deoxidizing. He Speculates that or gen increases the chill and that ladle additions remove oxygen and theéfore remove chill. F- Data Concerning the Influence of Inoculation on Physical PrOpe ties. Smalley (1) listed the results of physical tests on a 2 .Bfi 0- 1.2 % Si cupola iron melted from 60» steel scrap and 40» high silicon pig with 120 ounces of Ca-Si added per ton of metal. hodulus of elasticity —28.6 x 166 p.s.i., tensile strength - 46,700 to 51,000 p.s.i., brinell hardness - 22 7 to 247, transverse pronerties (1 1/8" 1a. on 12 " centers) - load- 4,800 to 4,850# and deflection-.12 to .14". Another iron of 70% steel charge with 100 ounces of Ca Si ad ed per ton gave E-27.8 x 106, tensile-46,000 p.s.i., brinell-255, transverse load-4,BOO# and deflection -.16". These irons picked up prdctically no calcium and approximately .1m i. Lorig (10) compared the physical prOperties of two irons without ladle addttions, one with Ca~Si and one with Fe-Si ladle additions at various superheating temp- . O eratures from 2 BSOF to 3150 F. The data was taken on irons prepared in a high frequency induction furnace with 125 # charges and similar melting schedules. The irons were poured at 25500? unless superheated to a lower temp- erature. The ferro silicon used was the 75% Si grade but the grade of calcium silicide was not given. The ladle .4 -flu“‘zg '-"' lo. zaddjxtions were equivalent to .28 Si for 08-81 additions and .5523 Si for Fe-Si additcbons. Blank X Blank 0 Ca-Si Fe-Si Total c (25) 3.22 5.02 2 .99 2.88 % Si 1.87 2.21 2.21 2.12 Tensile p.s.i. 50,000 51,000 42,000 44,000 {Dransverse # 2,500 2,580 2,850 2,980 iDeflection " .245 .208 .595 .588 Brinell 179 188 195 199 'Fig. 2- Physical test data for 2850°F superheat Blank X Blank O CasSi Fe-Si Total C(%) 3.19 2.96 2.89 2.81 % Si 1.87 2.19 2.07 2 .11 Tensile p.s.i. 29,800 54,400 48,000 48,500 Transverse # 2,400 2,470 5,200 5,190 Deflections " .250 .207 .568 .400 Brinell 183 2 10 200 208 Fig.3'Physical test date for 2950°F superheat It was noted by Long thit the tensile and transverse strengths decreased to a minimum at approximately 27000F, for'the irons witnout ladle treatment, and then increased asiflm melting temuerature was r ised. Inoculation with fernrsilicon or calcium-silicnn eliminated tnis effect. Crosby and Herzig (21) added Fe-Si to a base iron of 5.0816 c - 2.17;; Si, melted in a '50 lb. induction furnace frmnlbw crrbon ingot iron and graphite. Corpariscn was mmkaon phisical tests with additions of 25,50,75, and 100 pa?cent silicon ad ed as Fe-Si five minutes before pouring. 17'. Edna superheat temperature was 2750.F and the irons we: Ixnxred at 2050.3. The uninoculeted iron had a deidritic grapfirtte distribution and the inoculated irons contained ruxnnal graphite, with a pearlite matrix in all cases. The tensile and transverse prOpesties increased up to 75H late additions of silicon and then decreased. The brinell hardness remained at approximately 215 in all inst noes. The tensile strength went from 57,500 to 46,000 and then dropped to 59,500 p.s.i.. The transverse load incre sed from 2,5 0 to 2,9220 and then decreased to 2,450#. The transverse deflection increased from .190 to .518 and then decreased to .245”. Type B-A.S. T. M. bars were used in t ese tests. Rotn (2 2) reported that ferro-silicor(80fiSi grade) was used as an inoculant on o4 consecutive prodution heats to produce sixty t ousand pound per square inch cunola iron. A11 heats exceeded the minimum requirement for tensile strength and many were without alloy additions. The amount of Fe-Si ad ed to the 1600 pound laules varied with the size of the castings to be poured. The average carbon content was 5.15 and the average siligon was 1.9g. The transverse strength exceeded 5,100# (1.2" dia. on 12” centers) in all cases and averaged 5,600fii lughes and Seenceley (25) pr sented data on 50 inch square, 5/q inch thick band saw tablesnrodiced from ord- inary ounola iron inocul tgd with ferro- silicon. The irmrovement in p.ysical preperties were listed as impact- 2/g transverse strength-50p, transverse deflection- 2b“, 1...: 0'.) tensile strength -52p, and bringll -23. A, Comstock and Stfl55”QTO@f (2i) compiled a consideeauie tle effect of inoculation with Ee-Si vim) L71 if. 0 F l “'5 -‘ 0’1 and Tine-Si (20; Ti-20fi Si). Late additions were made melt was complete in each case. two minutes before the O ~-\.‘-L— -,-. .v. :‘r- ”'3 1 - 'I 4“ . ° f ;~tuie was 20/0 l and tie pouring ~-,— \‘ J'— , '18 SL;-~.‘1‘-LJ I1 433”. c A J . .1- ure was 2550°F in all cases. Transvers test *aVen or 1.2" dianeter bars broken on 12" investigators showed no marked in rove- rent caused bv inoculation for base irons varying from 5 09 to 3.593 e and from 1.77 to 2.45% Si. Ladle add- itions up to .673 Fe-Si and up to 2% Fe-Ti were made. Base irons exhibited from 4,200 to 4,400# trans erse strength and inoculation practice showed up to 500% introveient as a maximum. Tensile strengths were improved only sligdtly in most cases with an improvement of 5,000 p.s.i. maximum-listed for Fe-Si additions and 8,000 p.s.i. for Fe-Ti. Base irons exhibited farm 51,000 to 39,000 p.s.i. tensile strength. Schnee and Barlow (25) develOped an alloy contain ng 6% Al-l2fi Si-BOQ Cu (alloy 5) which they found to be the most suitable conper base alloy for use as an inoculating agent in gray iron. This alloy was compared with .53 ladle additions of silicon as Fe-Si. They found that one of the advantages of alloy 5 over Fe Si was that it did not promote the formation of shrinks and blowholes. {The following tables list the comparison of physical 1.2" diameter bars and 18" preperties for several irons. centers were used 19. for the transverse test and the tensile diameters were .800". Addition .5» Fe-Si 1% A.5 Tensile p s 1 55,000 58,000 Brinell 185 210 TrenSterse loed # ,700 5,100 Deflection " .550 .500 Resilience " # 670 705 % Si 1.56 1.52. Fig 4 5.2 5» 0 - 1.2 2 g Addition Base Tensile p s i 50,000 Brinell 185 Transverse # 2,500 Deflection " .250 Resilience 555 5 Si. 1.81 5 C. 5.26 Si mottled base iron Fig 5 3.265 C - 1.812 Si base iron Addition Base Tensile p s i 52,000 Brinell 190 Transverse # 2,400 Deflection " .190 Resilience 290 2 Si. 2.21 ;L C. 5.05 Fig. 6 .52 Fe-Si 1% 1.5 52,000 57,000 170 195 2,500 2,700 .550 .520 510 580 2.52 1195 5.21 5.25 .5% Fe-Si lb A.3 55,000 59,000 170 2 20 2,600 3,100 .510 .280 550 600 2 .6 2.33 2 .97 2.98 5.055 0 -2.21 5 Si base iron (neat A) 20. Addition Base .59 Fe-Si lfi A.5 TEensile p s 1 52,000 54,000 42,000 Brinell 190 180 215 Transverse;¥ 2,400 2,700 5,100 Deflection " .190 .550 ‘ .510 ‘Resilience 290 040 000 A Si. 2.21 2.55 2.2 5 fi 0. 5.05 5.08 2.97 Fig. 7 5.05} C - 2 .21% Si base iron (Eeat B). Bash (15) compared the piysical probevties attained after inoculation, using 853 Fe-Si and Hi(.55%Si and .75 Ni ad ed), Jith the uninoculated iron. The transverse test bars were broken on 18" centers. The results were 59,700 psi tensile, 252 0 # transverse load, .255” trans- verse deflection, 22 ft # izod, .67" chill for the unin- oculated iron with 5.053 0-1.885 81- 1.0 fl Ki. Correspond- ing results after inoculation were 45,600 psi, 5,100#. .581”. 58 Ft # and .04" chill with 5-05fi 0-2.25m Si-l.7p Ni. Ladle inoculation provided normal graphite struct- ures instead of the dendritic grep ite structures found in the untreated iron. In another paper, Bash (26) has shown the i provement possible by means of ladle additions to low carbon, aust- enitic cast irons. Increases in trinsverse properties for 14y Ei-6fi Cu-2b Cr-2.25$ 0 base irons were listed with 1» Si added as Fe Si to the Ladle( the total silicon of the irons in all cases ranged form 1 to 2p). “xample: Trans- verse load 5,400 to 5,200#. transverse deflection .120 to —_—f———i 21. .680". Little or no effect from inoculation was noted 1f caflxmlwas up to 2 .755. The ladle addition of .75 Si as Fe Slim roved a 5.0p 0-2.0» Si-l.Ot Ni iron from 4,600 to 5¢Km# transverse load, from .120 to .150” deflection, andtflmm144,000 to 53,000 psi tensile strength. The brhnfll hardness remained at 241. sely to Late additions of graphite have been used laré reduce Chill according to Dierker(56), Schneidewind (27), and Hassari and Lindsay(28). ther pny ical prOperties did not change to any appreciable extent. McElwee and Schneidewind(58)(27)(29) worked with graphitizing inoculants of the Ti-Al-Si type. The most successful has been a commercial alloy called Graphidox No. 2 (7.5% Ti, 20% A1, 20% Si, Balance Fe.). Two pounds per ton of metal was added to the ladle for several heats and the physical properties compared with base irons of similar chemical composition. Typical results follow: final analysis 5.1% 0-2.123 Si, improvement from 2100 to 2700# and from .246 to .297" on transverse prOperties, tensile strength.changed from 55,000 to 40,000 psi and chill form .40 to .25 inches. eese(l9) disclosed th»t ladle additions of 11‘ Flinn.and n .7395 Si added as an alloy of 70;. Si—ioy-g Lin-1.04%; A1- .125 Ti- 'balarmxa Fe was the most successful out of several inoculat- irng agxnrts tnat were studied. Without ladle additions the Ifirvsirxil properties were 41,000 psi tensile and 2,500# and .lSCV"transverse load and deflection. After inoculation vwftd iflie above alloy this iron tested 60,000 psi, 3600f m M (\3 O and .250". The inoculated iron had a fine grained, light . gray fracture *nd random graphite distribution. They pointed out that all successful inoculants considered contained silicon although some of the uncessful inoculants also contained silicon. Burgess and Shrubsall(50) pr duced irons of a machinable grade, that would normilly have been white, by means of ladle additions of Si-kn—Sr type inoculating agents. For irons of 5.255 0-2.55 Si tee transverse properties were improved from 1,800 to 2,1005, and from .150 to .250" for tne most successful cases although there was but slight benefit in 33me inst noes. The tensile strength showed improvement of from 58,000 up to 45,000 psi and from 50,000 up to 46,000 psi. Chill depth wa reduced. Surgess and BishOp (7) presented tne following con- clusions concerning ladle additions of .255 Si as an alloy of Si-in-dr (60-65” Si, 5-75 Ln, 5-7NZr): l-Chill depth greatly reduced (2.75-5.03 0, 1-25 Si) 2-Tensile strength improved approximately 10,000 psi n with low 0 and low Si(improved Iron 40,000 to 50,000 psi) 5-Transverse strength improved ap roximately 500; with low 0 and low Si(improv d from 2,500 to 5,000#). 4-Transverse deflection improved apgroxinately .100” in all cases(in low C-low Si from .160" to.290”) Nit 5-“luidity improved slithly 6—Inocu1ant does not lose effect in less thin :5 W n. I ‘ r" IV '4 ._ . . . . w 1.) -', _ __1 .C‘ . .f 7...:‘f'31 _‘ ‘|"1_~“"."4’3L.I“ .1‘1""‘. x". 7 ”r ”f1 "-7‘ Q ,—) ‘7‘ pm ‘ '_ ,) ever} in low C-low Si irons. B-Refines cell size 9-Renders nigh 0-985 Si or low 0-1.22fi Si completely lO-Larger a cunts of inoculant(up to .85) sometimes are necessary in low carbon equivalent irons to improve transverse strength and deflection. Lownie(9) reported that ladle additions of Si-kn~dr were slightly superior to ladle additions of Ee-Si in the lower carbon equivalent irons. Physical tests on a 5.295 C, 2.05; Si iron showed tie same effect with .45» Si added as an inoculant with either alloy. The base iron test d 2,400? and .290” for a 1.2" dia. bar on 18” centers, tensile 57,500 psi and 255 brinell. Comparison with the inoc- ulated ircns showed 2,600# and .580” transverse pronerties, 44,000 psi tensile and 210 brinell. 0- General Statements Concerning Inoculation Practice Williams(5l) com ented that CavSi additions(about .46 Si added) did not show any advantage over Fe Si and was decidedly more expensive. The cupola process was mentioned as being an inoculation process. Soft iron from one cupola was added to white iron from the other cu 01a in tpe ladle and the e fact was the sa=e as inoculation. Francis(52) claimed that the Si may be as low as It in a 2.96 0 iron and inoculation with 100 to 120 ounces per ton of metal will still 0e successful. 5) \N Pearce(55), Eoyt(54), and Delbert and Eotaszkin( have reported that inoculation nit...; Ga-Si was found to be very satisfactory. 25 III ORIGINAL RESEARCH A'- Scope of Investigation .The first part of the present investigation was 'conducted in order to compare the physical properties of similar gray cast irons Which showed equivalent am- ounts of silicon pick-up after inoculation with ferro- silicon and calcium-silicon. Ladle additions were select- ed so that comparisons could be made after .1%, .3%, and .5% 81. pick up from either inoculating agent. Conditions were controled so that the obly variable would be the ladle inoculatn.used. For comparison between levels (1.e. -.1% .5% 31. pick up) the total silicon content would be variable. . The second part of the investigation involved inoc- ulation with metallic calcium. Comparisons were made on the physical prOperties of the base iron, .5% Si. pick up from both Fe-Si and Ca-Si, and ladle additions of Ca. The variables introduced were the type of inoculent and the total silicon content. An attempt was made to eliminate the later variable by pouring a base iron with carbon and silicon contents similar to earlier irons after .5% Si had been introduced as a ladle addition. This base iron was ladle treated, with approximately .04, .11, and .22% Ce added, to find the relative effects of Ca inoculation. The desired chemical compostion for the base iron was fixed at 2.85% c, 1.90% Si, .75% Mn,.lO% Phos, and .08% s. An indirect arc, rocking type electric furnace of 250 pound capacity was used for melting. 26. Three 1.2" diam by 21" (ASTM*. Type B) test bars, and two chill tests specimens were poured from each ladle to check results. Transverse tests were made on all bars. Representative samples from all heats were analyzed for the carbon and silicon content and were examined microscOpi- cally. The brinell hardness was taken on one bar from each ladle. Tensile specimens were broken on representative samples from.the heats that were the most satisfactory. The various properties were tabulated and important curves were drawn. The results were studied and conclusions were drawn regarding the inoculating effects of Fe*Si, Ca- Si, and Ca on the gray irons under investigation. 27. B - Melting Practice 1 - Preparation of Molds The arbitration bar and chill test molds were pre- pared from Lake Michigan sand with oil and cereal binder. The molds were baked overnight and then those to be used for transverse test bars were washed with a commercial "pink" core wash shortly before pouring each heat. This practice provided 1.2" diameter bars that were fairly smooth after wire brushing. 2 - Estimating the Charges Seven, 250 pound heats were poured during the course of this investigation. Computations were made to determine the required amounts of steel and various pig irons to produce the desired analysis for heat 1. After the carbon and silicon analyses for heat 1 were obtained, additional calculations were made to try to bring heat 2 closer to the desired analysis. This proceedure was carried out for each heat, although in some cases the deviations were slight. An adjustment was made in the ferrous sulpher additions for heats 4-7 to bring the sulpher content closer to .08%. Heats 1 through 6 contained similar amounts of each constituent in the charge but heat 7 required complete new calculations because the origional supply of low sili- con pig was exhausted. The approximate analyses for the steel and pig irons used as listed below were needed for the calculations. A*-Lot l pig B C D E -Old low Si Pig -Steel -Charcoal iron -New low Si pig 4.22 4.12 %Si 3.60 .70 .15 1.67 1.23 Q J .88 %Pho s .188 .109 .020 28 .055 Fig. B-Apnroximate Analysis of Steel and Pig Irons One to four pounds of silvery pig(code F), ferro-silicon containing 25% silicon, was includ d in the original charge for all he ts, Small amounts of Fe S (code G) were added to each heat a few minutes before tapping. The sulphur content of this alloy was 50%. The calculations necessary for figuring the charge for he t 1 will be used as an example of tne method used for all he ts (see Fig.9). *Code letter for lot 1 pig; code letters for other con- stituents of the charge are placed in a similar position in Fig. 8. Code Weight #0 #81 fun #P #3 A (1153) 4.25 4.14 1.52 .216 .020 B ( 70%) 3.00 .49 .32 .07o .027 C ( o4#) .03 .10 .26 .013 .013 F ( 1%) .25 G ( 1/4# .140 Total 7.28# 4.98# 1.90% .307# .200# Estimated fl 2.9l 1.99 .76 .12 .08 Fig. 9-Sample calculations for computing charge; heat 1. £3 3 - Charging the Furnace A.Detroit electric rocking furnace of the indirect arc type, silminite lining was used to melt the charge. The capacity of the furnace was 250#, therefore the charges had to be carefully placed to avoid contact with the fragile carbon electrodes. The steel was sheared into small pieces and placed on the furnace bottom along with the silvery pig. The large pigs, which had been wire brushed were then put in and the charge was ready for melting. 4 - Melting and Superheating The electrodes were brought into position and the current turned on. The average total K I during the heats was approximately 125. After a considerable portion.of the charge had melted the rocking mechanism was turned on. The Fee Additions were added, after the iron had become molten, at an average time of 75 minutes after the furnace had been started. All of the heats, with the exception of number 1, re- quired from 90 to 95 minutes to reach the desired super- heating temperature (2900 F). Heat 1 required 140 minutes before tapping when it becoame necessary to chip out and re- place an electrode which stuck and could not be preperly adjusted. Tapping temperatures, as measured with an Optical py- rometer, were 2900 P, with the exception of Heat 7 which reached 2950 P before tapping. An attempt was made, during a trial run between heats l and 2, to check the superheating temperature with an rayotube mounted in the furnace door. The experimental set up proved unsuccesful and was not 30 subsequently used during this investigation. 5- Inoculation Procedure Small, preheated ladles were used for pouring the test bars. Approximately 35# of metal was tapped into each ladle and then it was carried to a scale to check on the net weight. Ladle additions of dry Fe-Si or Ca-Si were made to the stream of iron as it was tapped and the ladles were rotated to provide for maximum inoculating action. With the exception of heat 7, for which the scale was not used, all ladles were brought up to 35# before pouring. Ladle additions of metallic calcium were made by rapidly 'dunking' the Ca under the metal surface and agitating it until the inoculating action had taken place. This was accomplished, without hazard, by wiring the desired amount of Ca to one end of a 10 ft pipe and mani- pulating the Opposite end. Heat 1 was used as a trial heat to determine the amount of silicon pick up form various amounts of Fe-Si and 03-81 added to the ladles. It was determined, by chemical analysis, that all of the Si added as Fe-Si was picked up by the iron. For Ca-Si, it was found, a computed .4% Si addition picked up .3% Si and a computed .7% Si addition picked up .5% 81. The calculations involved in determining involved indetermining the required amount of inoculant to add to each 35# ladle follow: .1% addition of Si .035# Si per ladle (For 90% Fe-Si) .1% Si 1 454 g/# 17.7 g Fe-Si/ ladle. (For Fe-Si) .1% Si 18 g, .3% Si 535, and .5% Si 88g. 31 (For 60% Ca-Si) .1% $1 I 454 26.4 g Ca-Si/ladle. (For Ca-Si) .1% Si 2 65, .4% Si 1055, and .7% Si 184 5. The ladle additions of calcium metal to heats 4 and 7 provided successful inoculating action. Evidently the calcium was not introduced quickly enough into the metal bath for heats S and 6, and was not sufficiently agitated for satisfactory inoculating action. The physical prOp- erties for heats 5 and 6 will therefore not be entered in this thesis. The pouring temperatures, as measured on the optical pyrometer for all aldles in heats 5 and 6, were checked with a chromel-alumel thermocuple so these heats provided some useful information which will be included under the next section titled 'Checks on Pouring Temperature.’ Heat 4 provided a comparison between the bass iron, Fe-Si, Ca-Si, and metallic Ca additions. The 10 gram Ca addition was equivalent to ICEVCa/ladlciiplgo : 06%Ca , 35#/ladle x 4546/# The scales were not used for heat 7 but an estimation of the weights for each ladle with Ca added was made and the estimated Ca additions were .04%, .ll%, and .22%. 6-Checks on Pouring Temperature Several checks were made during the investigation to determine whether the pouring temperature as measured with an Optical pyrometer could be relied upon to give an accurate and uniform pouring temperature for all test bars. A fine wire, platinumrplatinum rhodium couple was used on the trial heat, poured between 1 and 2, as a check. This produced the following results from four ladles: Optical Pyrometer Plat. - Plat.RhOd. Couple Apparent True Temp. M.V.Reading Temp. 2350 F 2570 F 14.8 2615 F 2500 2510 14.7 2600 2440 2665 15.25 2680 2565 2585 14.9 2630 A similar comparison on four ladles form heat 4 provided more accurate results. The IV reading On three ladles was equivalent to 2550 F and for the fourth 2595 F as compared with an apparnent 2400 or 2625 F true tempera- ture on the Optical. The millivoltmeter readings reached a constant value during this later test and it was indic- ated that the actual pouring temperature for all heats was 2550 F although they were checked at 2625 F on the optical. A larger chromel-alumel couple which had been pre- heated in a small resistance furnace was used in con- junction with heats 5 and 6. The results from all of the ladles are tabulated below for compariton: Optical Chromel-Alumel Difference 2675 F 2515 F 150 F 2675 2505 170 2665 2505 150 2625 2510 115 2625 2480 145 2525 2455 170 26 00 2440 185 2600 2390 210 2570 2390 '180 35 The Optical readings were always higher than the thermo- couple readings with differences ranging from 115 to 210° F(average 166, mean 170°F). The larger chromel-alumel couple evidently did not have sufficient time to reach equilibrium with the molten metal before the pyrometer indicated that the desired pouring temperature had been reached. ~ The bulk Of the data on temperature measurement points to a fairly reliable comparison between thermo- couple readings and Optical pyrometer readings. From this it was concluded that the apparent temperature read- ing of 2400’cn the Optical pyrometer, which had been used to determine the time for casting, was equivalent to approximately 2550°F in all cases and the possible varia- ble Of pouring temperature had been virtually eliminated for all heats. 7-Casting Proceedure Two wedge chill tests and three vertical arbitration bar molds were poured for each ladle throughlut this project. For heats 1,2, and 5, the ladles were tapped and cast in pairs with Fe-Si added to the first ladle and Ca-Si to the second in each instance. Increasing amounts Of the inoculating agents were added to success- ive pairs during the progress of the heats. The first four ladles for heat 4 were used to compare .5% Si pick up from.Fe-Si and Ca-Si; ladles 5 and 6 were tapped and cast as a pair with metallic calcium added to the last one. Increasing amounts of Ca were added to alternate ladles from heat 7 to compare with those without ladle 54 additions. The total time required to tap heats 1,2, and 5, varied from 20 to 15 minutes. Beats 4 and 7 required but 10 minutes for tapping. Each ladle was poured when it reached the desired pouring temperature. It should be noted that the larger additions of Ca-Si and metallic Ca required approximately two mdnutes additional time before casting. Apparently this was due to the exothermic action Of calcium when added to the molten iron. turf.“- ~59) 35 0- Investigation Proceedure and Resu11:s 1- Chemical Analysis Samples for chemical analysis were taken from the top half Of the 1.2" Dia. test bars after they had been broken. The surface was gound clean and drillings were Obtained from three different places along the bar. The total carbon content of a representative bar from the first and last of each heat, as well as addi- tional checks for heats l and 2, was determined. A carbon train was used to determine the carbon content: samples were burned in a combustion boat and the result- ing colwas collected in an absorbtion bottle and weighed. Standard samples were checked at the start and finish Of each set of carbon determinations. Silicon determinations were made on a representative bar from each ladle for heats 1 through 4, and from the first and last ladle Of heat 7. The method used for sili- con.analysis follows: dissolve with H N 03 and fume with perchloric acid, cool and dilute, filter through platinum cones using NO 42 filter paper, wash alternately with warm 5% H01 and distilled water, ignite in crucible and weigh as Si 01 . Standard samples were analyzed to check the determinations. . A check on the sulphur content, using titration; was made on bars from the first and last ladle of heat 5. A tabulation of the results from the chemical analysis has been recorded in Fig. 10. The code letters and numbers used in this table indicate the heat number, inoculating )0 agent, and percent of inoculant added. Fig 11 has been used to tabulate the code letters and their meanings. This code will apply to all data taken during the investi- gation Carbon (%) 1F]- - 3005 LFS ' 5.05 105 ' 2.99 201 - 2.65 2F5 - 2.65 204 - 2059 207 " 2053 3F1 - 2.88 307- 2076 4F5A - 3002 4C - 3003 731 - 2.95 705 - 2.87 Sulphur (%) 5Fl - .06 307 " 005 FigJD-Chemical analysis lFl 1F5 1F5 2F1 2F5 2F5 5Fl 5F5 5F5 56 Silicon (S) 1.95 101 - 1.98 2.15 105 - 2.02 2.56 105 - 2.2 0 107 - 2.54 2.24 201 - 2 .26 2.51 204 - 2.50 2.76 207 - 2.81 1.98 501 1.95 2016 304 - 2014 2033 307 "' 2036 4F5B - 2.42 407B 4B 2.44 - 1097 4C "' 2005 781 - 2023 703 - 202 4 57 Heat 1 lFl - .1% Fe-Si* 101 - .1% Ca-Si 1F5 - .5% Fe-Si 103 - .3% Ca-Si 1F5 - .5% Fe-Si 105 - .5% Ca-Si 107 - .7% Ca-Si Heat 2 2F1 - .1% Fe-Si 201 - .l% Ca-Si 2F5 - .5% Fe-Si 204 - .4% 03-81 -2F5 - .5% Fe-Si 207 - .7% 0a-Si Heat 5 5Fl - .1% Fe-Si 501 - .1% Ca-Si 3F3 - .3% Fe-Si 304 - .4% Ca-Si 5F5 - .5% Fe-Si 307 - .7% Ca-Si Heat 4 4F5A - .5% Fe-Si 407A - .7% Ca-Si 4353 - .5% Fe-Si 4C7B - .7% Ca-Si 43 -HO addition 40 - .06% Ca Heat 7 731 - NO addition 701 - .04% Ca 732 - NO addition 702 - .11% Ca 7B5 - NO addition 705 - .22% Ca * Indicated .1% Siaddcd as Fe Si (similar notations are used for other additions) Fig.1l - Description of Code Letters and Numbers 38 2- Transverse Strength and Deflection The 1.2" diameter by 21" long, arbitrahon bars were wire brushed to provide a clean surface. A hand Operated Olsen tester was used to break the bars. The load was applied midway between 18" centers. A dial gage was set up to measure the deflection at the midpoint. The results of the transverse tests for all bars was tabulated by heats in Figures 12 though 16. Code Load (#) Deflection Code Load(#) Deflection lFl 2060 .189" 101 2312 .230" 2238 .221 2342 .240 2171 .203 2386 .258 1F5 2133 .232" 103 2692 .370 2003 .202 2073 .228 2142 .222 2428 .294 135 2203 .255 105 2668 .347" 22 51 .262 2513 .302 2274 .275 2730 .365 107 2164 .242" 2801 .398‘ Fig. 12 - Transverse Load and Deflection for Heat L 59 Code Load(#) Deflection 0ode Load(#) Deflection 231 2123 .169" 201 2224 .174" 1774 .120 2155 .172 2014 .150 2273 .174 233 2124 .165" 204 3017 .337" 2335 .191 2926 .320 2369 .198 3142 .372 2F5 2322 .218" 207 5016 .355" 2224 .182 5016 .565 2197 .175 5115 .555 Fig 15 - Transverse Load and Deflection for Heat 2 Code Load(#) 5F1 5F5 5F5 2210 2226 2180 2156 2270 2248 2479 2562 2461 Deflection .175" .176 .175 .225" .218‘ .212 .252" .281‘ ..258 Code Load(#) Deflection 501 504 2577 2552 2187 2560 2457 2660 2826 5051 2920 .230" .227 .204 .303 .265 .329 .559" .385" .547 Fig. - Transverse Load and Deflection for Heat 3 Code 4F5A '4F5B 48 Fig. Code 7B1 7B2 7B5 Load(#) Deflection 2269 .284" 2265 .280 2256 .274 2078 .258" 2276 .280 2255 .281 1848 .185" 2110 .255 2168 .255. Code 407A 407B 40 40 Load(#) Deflection 2645 .374" 2557 .554 2280 .255 2505 .555" 2625 .554 2528 .265 2252 .242" 2516 .505 2507 .508 15 -Transverse Load and Deflection for Heat 4. Load(#) 2145 2224 2255 1845 2215 2160 2000 2105 Deflection .201" .210 .258 .150" .200 .185 .165" .187 Code Load(#) 701 702 705 2255 2207 2060 2580 2560 2565 2705 2610 2675 Deflection .2 20" .210 .195 .505" .299 .290 .521" .505 .521 Fig. 16 ~Transverse Load and Deflection for Heat 7. 41 3-0hi11 Depth The wedge shaped chill test specimens were broken and the measurements for total and clean chill depth were re- corded. The clear chill included the white iron fracture only while the total chill included the mottled gray and white fracture(tota1 distance from the sharp edge to the completely gray fracture). Some of the specimens exhibited fractures that did not have any section which was entirely gray. In that case the total chill measured 3 5/16", which was the maximum possible depth for the test pieces used in this investigation. The results,according to heat number, were recorded in Figs. 17 through 20. The chill tests for heat 4 were lost. Chill Depth in Sixteenths Of an Inch Code Total Clear Code Total Clear lFl 20 11 101 14 8 18 10 16 .8 lFS 15 8 105 7 5 11 6 7 6 1F5 10 7 105 6 5 10 6 6 5 107 6 5 Fig. 17 -Ch 111 Depth for Heat 1. 42 Chill Depth in Sixteenths of an Inch Code Total Clear Code Total Clear 2Fl 19 10 201 22 15 12 7 25 16 2F3 18 11 204 7 5 21 13 7 5 2F5 12 8 207 5 5 12 8 5 4 Fig. 18 - Chill Depth for Heat 2. Chill Depth in Sixteenths of an Inch Code Total Clear Code Total Clear 3Fl 55 19 501 15 10 33 20 14 10 3F3 14 9 304 7 5 l4 9 7 5 3F5 8 6 507 4 3 8 6 6 4 Fig. 19 - Chill Depth for Heat 3. Chill Depth in Sixteenths of an Inch Code Total 781 15 19 7B 2 ' 55 55 7B5 55 55 Clear 9 ll 16 17 16 15 Code 7C1 702 705 Fig. 20 - Chill Depth for Heat 7. Total 18 17 10 10 9 8 45 Clear 4s-a-cncnm 44 4 - Brinell Hardness The samples for hardness tests were cut 3/4" thick and were taken adjacent to the fracture of the 1.2" dla. test bars. After grinding, the brinell impressions were made using a 3,000 Kg load on a 10 mn steel ball. The data from these tests are recorded in Fig 21 for all heats. Code Brinell No Code Brinell No. 1Fl 207 101 197 1F3 201 105 197 1F5 207 105 197 107 207 2Fl 229 201 229 2F} 225 204 229 2F5 217 207 229 3F1 217 301 212 5F5 207 504 201 BPS 207 307 223 4F5A 192 407A .201 4F5B 197 407B 197 4B 201 4C 197 781 212 701 212 732 212 702 207 735 212 705 212 Fig. 21 - Brinell Hardness manners (Heats 1, 2, 3, 4 and 7). 45 5- Tensile Strength Specimens for tensile strength were taken from the lower half of the 1.2" dia. test bar adjacent to the fracture. The nominal diameter for the tensile test was .800". A'universal testing machine was used to break the tensile specimens and the breaking loads were recorded. The ultimate tensile strengths for heats 3,4, and 7 were listed in Fig. 22. Code Dia.(") Area(Sq.") Load(#) Tensile Steength 5F1 .799 .5014 18,420 36,800 psi 5F5 .799 .5014 19,030 38,000 5F5 .799 .5014 20,410 30,700 301 .799 .5014 19,220 38,400 304 .800 .5027 18,310 36,600 307 .798 .5001 21,500 43,000 4F5A .799 .5014 18,070 36,000 4F5B .799 .5014 18,750 37,400 4074 .801 .5039 21,380 42,400 4078 .800 .502 7 21,2 70 42,300 48 .799 .5014 18,760 37,400 40 ‘.799 .5014 22,230 44,400 731 .798 .5001 19,720 39,400 782 .799 '.5014 18,440 36,800 783 .799 .5014 18,980 37,800 701 .799 .5014 19,440 38,800 702 .798 .5011 22,710 45,400 703 .799 .5014 24,220 48,400 Fig. 22 - Ultimate Tensile Strength.(Heats 3,4, and 7). 46 6 - Microscopic Examination Samples for examination of core and surface micro- structures were cut from the brinell hardness specimens after the hardness readings had been recorded. Polishing and etching was accomplished in the following manner: 1- Rough grind on side of grinding wheel. 2- Polish on 120 grit paper disc. 5- Flat polish on #1 emery paper. 4- Wet polish on wax wheel using #520 abrasive. 5- Wet polish on wax wheel using #600 abrasive. 6- Alternate polish and etch on wet silk wheel using 'Met polish'. 7- Slight etch with 2% Picral-2% Nital solution for examination of graphite distribution at 100x. 8- Re-Etch for exaimination of matrix structure and graphite distribution at higher magnification. After the microstructures had been examined, and the results tabulated, representative samples were chosen for microphetographs to accompany this paper. The data on microscOpic examination has been assembled, by heats, in Figs. 23 through 27. ' No particular differ- ences in the structure of the pearlitic matrix was found, therefore only the amount of ferrite and its distribution was recorded. Graphite patterns were observed to fall in either type A, type D, or mixtures of these types. In general, the A graphite was approximately size 5 and the D graphite size 7, therefore only the type was recorded. These designations for graphite distribution and size were accord- ing to AFA specifications. .[1 .l I" Code 1Fl 1F5 101 107 Surface Similar amounts of D and A graphite Medium amt. of Fe. with.A graphite Small amt. of Fe. with D graphite Largely D grd.with fairly large amt. of Fe- Some A gr with medium amt of Fe. Largely A gr. with small amt. of Fe. - Some D gr. with small amt. of Fe. A graphite- Small amount of Fe. Core A Graphite- Small amts. of Fe. A graphite- A little more Fe. than lFl A graphite- Ne ferrite A graphite- Fc ferrite Fig. 23 - Graphite Distribtuion and Microstructures for Heat 1. Code Surface TF1 D gr.-FairLy large amt. of Fe. 2F5 D gr. - fairly large amt of Fe 201 D gr. - Fairly large amt. of Fe 207 A gr - Fairly large Amt. of Fe Core Largely D gr- medium amt. of Fe, some A gr. with no Fe Mixed: D gr-fairly large amt. of Fe, and A gr. with medium amt. of Fe Largely D gr - small amt of Fe A little more A gr than 2F1 with no Fe' A graphite - medium amt. of le Fig. 24 - Graphite Distribution and Microstructures for Heat 2. Code 5Fl 5F2 5F5 501 504 507 Fig. Surface D gr.- Fairly large amt. of Fe D gr. - Fairly large amt. of Fe Mixed: D gr.- medium amt. of Fe, and A gr,- medium amt. of Fe Mixed: D gr-medium amt of Fe, and A gr.- medium amt. of Fe Largely A gr.- small amt of Fe, some D gr.- medium amt. of Fe A gr.- very small amt. of Fe, some flakes at surface were smaller Core 49 Mixed: D gr. - medium amt. of Fe, and A gr. and no Fe A graphite Very small A graphite small We A graphite No ferrite A graphite No ferrite A graphite No ferrite amt. of Fe of Fe 25 - Graphite Distribution.and Microstructures ofr Heat 3 Code Surface 4F5A A gr.- small Amt. of Fe, smaller flakes at surface 4FSB A gr.- small amt. of Fe, small amt. of D gr. with medium amt. of Fe 407A A gr.- very small amt. of Fe, some flakes at surface were smaller 4073 A gr. - small amt. of Fe, smaller flakes at surface 43 Mixed: D gr- medium amt. of Fe, and A gr.- medium amt. of Fe 40 M.Mixed: D gr - medium amt. of Fe, and A gr.- small amt. of Fe 50 Core A graphite - Very small amt. of Fe A graphite - Small amt. of Fe A graphite - No ferrite A graphite - Small amt. of Fe A graphite - Medium amt. of Fe A graphite - Small amt. of Fe Fig. 26 - Graphite Distribution and Microstructures for Heat 4 Code Surface 7B1 D graphite - medium amt of Fe large amt of Fe 701 D graphite - medium amt. of Fe 702 D graphite - Medium Amt. of Fe 705 A gr. with'very Imall amt. of Fe, some smaller flakes with medium amt. of Fe 51 Core Largely A gr. with small amt. of Fe, some D Gr. with medium amt. of Fe Largely A gr. with medium amt. of Fe, some D gr. wdth farily large amt. of Fe Largely A gr. with very small amt. of Fe, lesser amt. of D gr. than 7B1 with small amt. of Fe A graphite - Small amt of Fe A graphite - Very small amt. of Fe Fig. 27 - Graphite Distribution and.Microstructures for Heat 7 52 7 - Micr0photography Micr0photographs at 1001 were taken.to represent the types of graphite distribution which were encountered in this investigation. A magnification of 250x was used to compare the microstructures and graphite distributuions for the maximum addition of Fe-Si with that of Ca-Si; also to compare the maximum addition of metallic calcium with the base metal. Although it would have been possible to present many more micrOphotographs in this paper, it was felt that the micros which were taken represented all of the significant comparisons that could have been made by this method. A B&L 'Research.Metallograph' with tungsten arc was used for taking the micro's. The following combinations were used: Objective Eypiece Bellows Setting Exposure Time 100x - 8X 51 65.5 cm. 4 sec. 250x - 21X 5x 61.5 cm. 10 sec. Eastman 1Wratten' metallographic plates with M Q develOper was used for the negatives and 'A20' paper with D72 develo; per for the prints. The etchant was a 2% nita1-2% picral solution. The specimens for Micros at 100x were etched for 1 sec. only to bring out the graphite distribution and those for 250x were etched for 4 sec. to bring out the matrix structure. 55 lee X [1 307- Light etch with 21 moral-2% nital solution. Type A (normal) graphite distribution. 35' n 2 381- Light etch with 25 pioral-zx nitel solution. Mixed graphite distribution- type A and D. 100 X 55 u 3 100 x 783- Light eteh with 2x moral-2:6 nital solution. Type D graphite distribution. 7 i i i 3 ! L—__.__§. I 4 2% Morel-2% Nital Etch 250 x 3F5- Surface- Mixed, D graphite with ferrite and A graphite with some ferrite (.51 Si pick up from Fe Si). 4! 5 2% moral-2% Nital Etch 507- Surface- A graphite with some ferrite (.57‘ Si pick up from Ca-Si). 250 X DI 1L 6 A “""amoaI-zx Hital Etch-5- 250 x 7B5-Surface-Type D graphite with fairley large amount of ferrite (Base iron, heat 7). -.—— .._--—-v———. M 7 27‘ Picral-2% Nital Etch 250 x 703-Surface-Type A graphite with small amount of ferrite (.22% Ca added, heat 7) M 8 ‘9 2% floral-2% Nital Etch 7B5-Core-Largely A graphite and some D graphite, some ferrite (Base iron, heat 7) M9 2% Floral-2% Dital Etch 250 x 705 Core-Type A graphite very small amount of ferrite (.22% Ca added, heat 7) 59 8 - Condensation of Data, and Curves for Physical Properties The accumulated data from the investigation was condensed and various curves were plotted in order that the physical properties resulting from different inocur lants might be more easily compared. The average carbon content for each heat was recorded. The transverse test data was plotted from the best bar out of each ladle. The average values would have been incorrect in many cases .beoause of possible surface defects on one or two of the bars from the ladle in question. The average value for chill was listed in the condensed data. The Brinell hard- ness showed no significant differences so it was recorded on the data sheets only. The condensed data concerning Fe-Si vs Ca-Si as inoculating agents was recorded in.Fig. 28. The condensed data for‘base irons vs metallic calcium ladle additions was listed in.Fig. 29. A plot of the bending load vs the deflection was made in.Fig. 30,to compare the effect of the inoculatns on the transverse properties. Curves for the physical properties is the actual % 81 pick up (Figs 31 and 32), vs the carbon equivalent (Figs. 35 and 34), and vs the % silicon (Figs. 35 and 36) were drawn from the first four heats to compare Fe-Si with.Ca-Si ladle additions. The results from heats 3 and 4 were included in this paper as representative curves. The physical prOperties were plotted against the % Ca added I Fig. 37), for heats 4 and 7, to determine the effects of -inoculation with.metallio calcium. Curves were plotted in Fig 38 to compare the effects of CaiSi additions on the ’0‘ ‘re’ 00 physical preperties with the effects of Calcium.metal Addi- tions. The pr0perties were plotted vs the % Ca added. The Ca-Si (30% 0a) additions of 26,105, and 184 grams were equi- valent to .05, .20, and .35% metallic calcium added, respecti- vely. The actual % Si pick up to use for plotting Figs. 31 and 32 was based on.the assumption that an addition of .1% Si as either Fe-Si or Ca-Si actually resulted in a .1% Si pick up. The actual % Si pick up was determined, for each ladle, from the chemical analysis for computed .3 and .5% Si pick up respectively. The actual % Si pick up for each ladle with Ca-Si additions was alos figured. Additions of .4 and .7% 81 as Ca-Si were calculated to result in .3 and .5% Si pick up to compare with Fe-Si inoculation. The actual % 81 pick up based on the chemical analysis follows: (Example for 1F3; $81 for lF3-% Si for lFl 2.15 - 1.97 .18 .18 .10 .28% 81 pick up, actual) Forcaloulated .3% 81 pick up .31 31 added as Fe-Si .43 added as Ca-Si IF3 - .28% 105 - .41%( .5% 31 added) 2F3 - .37 204 - .34 5F5 - .28 204 -.31 For calculated .5% Si pick up .53 Si added as Fe-Si .7% Si added as Ca-Si 1F5 - .493 107 - .553 2F5 - .62 207 - .65 3F5 - .45 307 - .55 4351- .45* 407A - .46 4F5B - .45 4073 - .47 * (1.97% Si base iron) Code 3 Si Te8e-# Tone-" T.0h. 0.0. Brinell 0 Eq-% Code 1 Si TeSe-# TOD.-. T.0h.. 0.0. Brinell 0 Eq-% Heat 1 Heat 2 2.60%0 Code Heat 3 331 TeSe-# ToDe-u T.Ch.- COCO Brinell 0 Eq-3 Tensile .1% 81 P0* Fe-Si Ca-Si lFl 1.97 2258 .221 19 11 207 5169 2Fl 201 ‘ 2.24 2.24 2125 .169 15 10 229 5.55 5F1 1.98 2226 .176 55 20 .217 5.48 36800 38400 Iote: * .13 Si pick up T. S. - Transverse Strength T. D. - Transverse deflection Total chill(16th of an")0 Eq-% T. Ch 101 1.89 2586 .258 15 8 197 5.68 2275 .174 24 16 229 5.55 501 1.93 2577 .250 15 10 212 3.46 0. Ch - Clear chill(16th of an')Tensile 36000 0 Eq - Carbon equivalent . (0 1/3 Si) Tensile (given in psi) .33 Si PU .53 Si PU Fe-Si Ca-Si Fe-Si Ca-Si lF3 105 1F5 107 2.15 2.20 2.36 2.34 2142 2730 2274 2801 .222 .565 .275 .394 15 6 10 6 7 5 7 5 201 197 207 207 5.77 3.79 3.84 3.85 2F3 204 2F5 207 2.51 2.50 2.76 2.81 2369 3142 2522 3115 .198 .372 .218 .355 20 7 12 5 12 5 8 4 223 229 217 229 5.44 5.42 5.51 5.54 5F5 504 5F5 507 2.16 2.14 2.33 2.36 2270 2660 2562 3031 .218 .329 .281 .385 14 7 8 5 9 5 6 4 207 201 207 223 5.54 5.55 5.60 3.61 38000 36600 40700 43000 COde 4F5A 407A 3 Si 2.42 2.45 To Do“. e284 0374’ Brinell 192 201 3.85 5.85 42400 Code 4FSB 4073 3 Si 2.42 2.44 T.S.-# 2276 2625 TeDe“u e288 e334 Brinell 197 197 0 Eq-% 5.85 5.85 Tensile 57400 42300 Fig 28 - Condensed Data on.Physical Properties for Fe-Si Vs Ca-Si Heat 4 3.0230 2.013 Si 0 1/3 Si -3.693 Heat 7 2.903 0 2.243 81 c 1/3 Si-3.65% Base Iron 43 2168# - .233" 201 Brinell 57,400 psi 781 2255# - .258“ Chill-17 and 10 212 Brinell 59,400 psi 782 2215# - .200" Chill-33 and 17 212 Brinell 56,800 psi 735 2105# - .187" Chill-35 and 16 212 Brinell 57,800 psi 62 0a.Added 40 - .06% Ca added 2 516# - .505 197 Brinell 44,400 psi 701 - .043 Ca added, 2255# ‘ .220 Chill - 18 and 9 212 Brinell 58,800 psi 702 - .113 Ca added 2580# ' .505" Chill-10 and 6 207 Brinell 45,400 705 - .22% Ca added 2705# - .321" Chill-9 and 4 212 Brinell 48,400 psi Fig 29 - Condensed Data on Physical Preperties for’Base Irons vs Ca Additions 151‘} . \r '74,. i i - 1 11.11:} Fig. ' O- TransverseTest Fosults 7E1 __ Plot of : ending Load _(#) vs Deflection .(") .. .__-. 267 - V .1: ...---_g,.4 i (Fe-82L. Ca-Si, and Ca) 1,. - ; 4 F £048.: 1 . ; 7 ._..__3,.”-__.ASF‘3I'$x 0 A5 (14:5, a . I 1 ,h I ‘3 (13-1- (7! '53:).r, S PL'Ck Ur; " ““"As Fe-Si To ‘AS Cad-Si D i i BA“ tron X . 04% Ca 214141845 X [C 7 18» -- . .Cé ‘. i l X a ellh‘rb~cs R4131 I X ”217/Rx, 11(4A = ‘ Eg/CF . z 7“ 3 .-__-.2m____-_ ., ‘T‘ - ‘ _.X femur f i 5 i , , Load f i E1307" C") [Dennis - , - l3 407:4 : : - ' 7‘71?- ‘ 32609“ ‘ 4 ~ 4 ; - t I 3F3 ”763‘ 7 ‘ G) t f . 7 x40 - 1500“” ~ ' ‘ — - -* -- * i i i ; 3 Trensluev-se Deflection - Inches ; .790 .130 .270 .370 .350 . 390 '. 1 l. . I 2m.____i-..-.,.,.-.- -. .' 3C, 13'“ woof—+4— - . g. ,- . [5120' ; @3783: (F59 0047555 r . 6’”. 4F51‘5 3H9 75, a _ _ - -___'7_.B_% TH. __. _ I a ('1’ / I' b I .'/”I _ .L~ .’.- ' x’) ' / H,;"/ .r" 5 ‘1‘ 7": if; : /C '\ .l /’ E ,2 ’ , 4g . \lfl _ a E 12 45‘ . E E ; § \‘ ‘5 ’ § 1 ; : -" E ‘ "3’ i i I .. Fig. [31- Curves! for Actual % Si Pick-up ' vs: Transverse Load and? Deflection, / also vs Tensile Strength } /" _ - __< Fe-Si-end-9e_-S:l)- . . : I E ! .. 4;,/’/ 7 E E \13/ E ‘ E 65 i - ----- ' - ‘ _ . 2.- . . ' .. ‘_ .47) I ’8 ; ___,..—-—-—" ‘ _..—-—-- W4 Feon ".1- nEBe Keane 'De£ .131" - >.1%l .th i 3 E . 3 4 -5' 7. S; PreLup (actual) \ -_— .vdp—u-_—-_.—__ — “7-4.- 1 H - 65 - -fo "“f J I I ' ' , I i | I Fig. 32- Curves for: Actual % $1 Pick-up [ Carbon ' 4 Equivalent 't Fussy-am: Cat-Sikh” . -- vs Total Chill, also vs 76 E S l ...... ..... 0 v ~. i —-- :.“.1 i 4. - - . *_ ,fi- ewxfl E 5.- k- a .N“ E I I ~ ' a i _ ; fii .4 H #H ., ._ l I ‘ n i l ‘ MN 0 l MWJ fl- ' . I "" > .--"'-' -| .__. - / 0 MI . [K . 4} ‘ "flirt“. . H ILOILSwv-v‘1 9 . ._' . E l | .13-. . 4‘ '( WWI-4"" - ' 0" . ' _ — -- — .- -¢ ~~~~~ .. s “ ..—I ------ ' — ‘ I I f.” I W”"" ’ -v--*’ [*‘M ,mwr—f', M'w‘gfi T ‘b‘: ) 'e Pr .0. ,0" -' H . :1 Lat-9:». '4 70 S). ’Ple '9‘? 00 'h‘\,// {k Ht + 9 / ’. ’ 4, re '54 “‘~ 4 - . T E .Ir‘ *-.4__‘._... - / . U / j E / i L 3 My _,, ... _ .aI-__1._.__..- 4 \ O 4.. . . ‘_ l 0“ — ~e-— v—-4-4 -- i n i Fig. 33- durves for % Carbon ~\ t E f 3 Equivalent vs TranSVIrse f/’ = Load and Deflectflon, also /’ l 78 Tensile Strength ~./;- --~-(Fe-Si and Germ).-— E ‘ ' t i i ; I" E --$ .. f h - -- n. n i E ,4" : i . )5 i 6 2 E ”I" F E / E. C ; . 5 “i duV'Fb-Sl - ”.180 ‘Ee’30 375 3+ 3.3 3.8 E ‘7. Cat-Egon Ejuilillcu‘t [C+ '/3 Si 3 'Trans. Def. ' In.» . ‘m-vz-r _' a?” .I' ‘ Ii... udE Nair: IE 30 [0' H |IE-L.EE . 67 E E Fig. 34- Qurves for 76 carbon -._.._... 4 = .___.equivalent .va clear and _ __ .. . ' l I II . _ Ia IIOEIII' Ill E E E Eh E E E l E E E E E total chill (Fe-Si aEnd 03-81) a E .- wig- - I . L Q 1,. I. f. H 1‘ \LIIlEIEtl .IElul-l‘l‘ .‘| .‘l-u. 4y. I I I. I." ' r! I 44'... -E|-_..1. EaEilE- Eh“ a ”E E E fl/ 1'. I - ’’’’’ E Tensile. ‘ /E / . E 5f V. I E \\\‘ ./ - E 5")/ I I E I J.»“‘-._\_ 1 - . I E r.“ i %,N\\-~_ Ht 4. Fe- S“ a | I" 'k. ‘ “—- ~ E " .3 . I WWI.“ _ .4 - E -. . ....... 4.33., .' , E E I I E E E E " E E a» E I E I ' I I E I ' ’ i 7 / .. i HA; L r7 7 . Cf -/’4+ ..f I E l ...... -..—i Ttlhslm I C- '4 _- ' 03f . - I' . E E E / “5 E f a ‘ I” i E I . "r" E r ' . E I I E / . it) A. . 4.2.30 I . E / 4 I E . " 4b “ a ’ / 14* E : , / / EFe-SI .. L 1 E E {E- g ‘4" I ‘ up H- - E i . 4,; EM..- _. f... "7". -4- _ -7; , --.-.. .7 7‘ . . . .4 ,..- Em-.- - - E E E 5 ' g 1’. I E - f/' E i ' .~ 1315. 35- Curves for f Silicon ’ j E .‘ / 3 I vs Transverse Lo aand ., E E , ~' 1’7; Deflection, also ‘ I , ‘ Tensile Strength _' '. E .'e ..,.-_ _ .4 ‘ 4.- 45,130 . 3"”‘E '3‘ 7 I (Fe-Si and (la-3i: i ; ' ' i £0ch " E E CE M £32,. E .. .2_._.. ..__._.._ i- + _ .... E - AH __ E r _ i .i‘h .7 “4.. _ -.- a - . 1 _ ., . -..-_... .. E 1.65. E j .. ‘i i ‘ ‘ EL 3 d' _2 ' E ‘ ' “ _E F i , I ; ' E ; ,../ 3 . . . E Ht- «kw '3. 0174 CE E. E I ‘1' ' y . ' “—T-‘J-m --- E . :.‘ . . 4 i- boo— -...“thm- E "l w’///;,_ . - E i E E / Ella . fe' ‘ “W | i I f I I / /3 I: I ._ ___ _ for- l - E .. i i - _ _ E /. -..;f, . : - t ., 4- ..44' ,4») I .... ~78 43/"!!! / ...- ?-SE ‘1’ / / t * 4 . W E E . 68 243 247 Silicom 2-5 70 \r T 1", . I //’f< ”*4 i .4949 I I T 111?? Pig. To M. -..—. .... 36' CurVeq tel Chill , (CaPSi for'% S also vs and F6951 “MR—“ii E AP———.-_——n—4b—_.———I- —.- _—l_—_ ———-—-_ —.—.——‘—— ———————————- —-Eb—.————————-—.-Ir————-- —-— l- -. Q . I "*4” Nit}? EFT" Es i I: A 9E; 3m; AI: ~7- E E E E I E j: 1 in I “' Tfi E I I E E I I E E l I A A L E 29 2. E zsy’lrtbh E ! I c-o—o-{p— — f E , , n H 4- 4 If i f ; a / l ' / y . I ' i 1/’ g .9 I m" I f— — — ,_ 1;» f .4‘" I : 70 i Fig. 37- Curves for % Ca Added v3 - ;»-Transverseiand Tensile~Properties, ‘ also vs Total Chill ; : i(Ca and Blank) 5 l f . ~~~- »;H% ‘ A i . 1+?: t// ‘I’ . i i I .. c—o—L -' 4». 310 H .x; 7 _, - , ,/ 'l I! j 4 , 3 I «- Transverse; 1' ; _ Def-Itch bu; Tntl‘C‘ 1 3 ++ 7 -+ - - .L,.n.u . i f 1 t 1 I L i . ; f ; r“ ' I1. ~ Hi 1 -4 F I ,1 M“. l ‘ - ,. 1 "‘ w 1 (Vote: ; 1, Hf‘I - {NW-7 F 3.51 I~ 0’ L I '. 1* f _ _. . , -... ; I /&,’- 74 ,x ,i' ‘ . , - C; I! 1 4 I ; {/3 I, 0,0 (I5. .16 w” . W§~ ‘ 4 Ca _ S" ...-g pa—C-O .....— ‘ ' 39§%- G 43.4.5.9; flMD # 3 Ca 5‘ ...» MM - L LHt,3,,-Ca”S£. _,--( '0 ”oft; . l . IW£3~2vgiL7oC [Hi “Way/fig . _ ‘ H3111 m‘: ‘ ‘ ’I 4__-_.l..--__._ H. g'éanfl" L- U____._.. HM} "38~~Gurvew~fbthfikarjfikhnrfinrk" fansv3rsa d Tensile Propertiet [80 vs Chi 1 ' ( a anion-$1) '. I . i , 72 D - Discussion 1 - Chemical Composition The desirability of having similar chemical compositions for the comparison of Fe-Si with Ca-Si has been previously expressed in this paper. In all cases the silicon contents for each pair of ladles using these inoculants were very close (average difference w.03%: maximum difference -.05%). During the tapping of the heats the carbon content decreased as much as .16% in one in- stance but the maximum difference noted for any pair of ladles was .06%. The sulphur content was analyzed at .05 to .06% for heat 3 but was adjusted to meet the desired .08% in the later heats. Other constituents were believ- ed to be similar to the desired analysis by calculation and were not determined analytically. Heats 1,3 and 4 were close to the desired carbon and silicon analysis for the base iron. Heat 7 was adjusted to provide a silicon content, for an iron with Ca additions, similar to the total silicon acquired after a silicon bearing inoculant (.3 to.7% 31 added) had been added. The carbon content for heat 7 was similar to that desired. The curves for carbon equivalent vs % $1, at the top of Fig. 36, show that heats l and 4 were very similar as to total silicon and carbon equivalent at all levels. Heat 3 was similar to these in regards to total silicon but the carbon equivalent was less. Heat 2 had a considerably higher silicon content but its low carbon 73 content placed it far below the other heats for carbon equivalent. The curves at the top of Fig 38 indicated that heat 7 was between heats 3 and 4 for carbon equiva- lent. The data on chemical analysis has shown that heats 3, 4, and 7 had comprable compositions and might be readilly compared; therefore these heats were chosen for represent- ative curves. The curves at the bottom of Fig 32 indica- ted that the % Si pick up from Fe-Si or Ca-Si was practic- ally the same at all levels throughout the heat, another factor to promote reliable comparisons. 2 - Transverse strength and Deflection The accumulated evidence on.the transverse properties of the irons under investigation conclusively pointed to the superiority of Ca-Si ever Fe-Si as an inoculating agent. The data on ladle additions of metallic calcium indicated that Ca (up to .22% Ca added) was at least as good as Ca- 81 and possibly even more effective. The only data in the published literature comparing Ga-Si with Fe—Si additions (see page 19) could not be con- sidered satisfactory beacause additions of .2% 81 as Ca- Si were compared with .5% Si as Fe—Si. The fact that the physical properties of these irons were comprable, although much less Ca-Si had been added, suggested the possibility that Ga-Si was more effective as an inovulant. One author (see page 23) stated that Ca-Si showed no advantage over Fe-Si with .4% Si added in either case. There was no data supporting this statement, therefore it could be given little if any consideration as a reliable source if inforv 74 nation. No data concerning Ca inoculation was discovered in the literature. The curves in Fig 31 showed that in heat 3 a .5% Si pic k'up from a Fe-Si ladle addition provided the same transverse properties as a .2% Si pick up form a Ca—Si addition. All of the other heats presented a less favor- able comparison for Fe-Si. At .576 31 pick up for both inoculants in.heat 3, the Fe-Si addition showed 2560# and .280” while the Ca-Si addition showed 2940# and .370“. Fig. 33 indicated that in general the lower carbon equivalent irons provided superior transverse strength and inferior transverse deflection when comparing Fe-Si or Oa—Si additions between heats. It was noted that the transverse load drOpped from ‘ 3035# for heat 3 to 2625# for heat 4 with an approximate .5% Si pick up from Ca-Si. With Fe-Si the drOp was from 2,560# to 2,275#. The ‘ transverse deflection stayed at about the same level for both high and low carbon equivalent irons at the maximum additions of Fe-Si and Ca-Si. For Ca-Si heat 3 was .385" and heat 4 was .350” average; for Fe-Si there was no change from .280". The plot of transverse preperties vs the total % silicon in.Fig. 35 also revealed the extent of the impove- ment gained in all cases by using Ca-Si as an inoculant. Fig. 37 indicated that increasing amounts of 0a im- proved the transverse properties. In Fig. 38, comparisons were made for the transverse properties between Ca and Oa-Si ladle additions. Even though the carbon equivalent 0F 75 of heat 7 was greater than heat 3 the transverse strength and deflection were similar for similar amounts of Ga add- ed- For heat 4 the small addition of Ca improved the trans- verse prOperties to a greater extent than similar amounts of Ca added as Oa-Si. It should be noted that although the lines for Ca-Si in heat 4 were dotted, indicating that no intermediate values were determined, the probability that the actual values would fall near to this curve would be great (compare with Oa-Si additions for heat 3, Fig 31). The transverse breaking load has been recorded vs the transverse deflection for the representative bar from each ladle poured during this investigation. With but one except- ion, all of the points representing base irons, .1% 81 pick up from Ca-Si and Fe-Si, .3% and .5% Si pick up from.Fe-Si, and a .04% addition of metallic calcium fall in the lower left hand corner (.150" to .2 85" deflection and 2,100# to 2,400# load). The lone point in the upper left hand corner was the .5% Si pick up from Fe-Si for heat 3. ‘In the upper right handtcorner (.300" to .395" deflection and 2,500# to 3,150#) the points representing .3% and .5% Si pick up from Ca-Si, and Ca additions from .05% to .22% were located. This would indicate the definite advantage to be gained by using Ca-Si or Ca as an inoculant, rather than Fe-Si, to improve the transverse properties. 3 - Chill Characteristics Ca-Si and Ca were found to be more effective chill reducers than Fe-Si, particularly when comparing the smaller additions of these inoculants. The curves for clear chill 76 followed the same pattern as those for total chill, so usually the characteristic curves for total chill were presented. Curves from he.ts l, 3 and 7 were used to compare the chill tendencies of the different inoculants. Fig. 32 shows that the greatest difference between Ca-Si and Fe-Si occurs at the .1% 81 pick up level for heat 3 (33/16” for Fe-Si and 15/16" for Ca-Si). On heat 1 the comparison was 1 9/16" for Fe-Si to 15/16" for Ca- Si. At the .5% Si pick up level the corresponding differ- ences were from 8 to 5 and from 10 to 6 sixteenths respect- ively for heats l and 3. An examination of the curves in- dicated that .3% Si addition as Ca-Si was as effective as a .5% Si addition from Fe-Si. The curves in.Fig. 34 included the clear chill and total chill vs the carbon equivalent. Ca-Si inoculation was as effective, in all proportions, for reducing chill on a low carbon equivalent iron (3.55% average) as for a higher carbon equivalent iron (3.75% average). This was not true for Fe-Si additions as evidenced by the curves for heats 3 and 1, respectively. The plot for chill depth vs .% silicon presented an excellent graphical picture of the superiority of Ca-Si, as a chill reducing inoculant, over Fe-Si. The curves for Ca-Si additions were observed to coincide while the Fe-Si curves were considerably higher for heats 1 and 3. The curves comparing the effect on chill from Ca addi- tions in heat 7 with Ca-Si additions in heat 3 followed approximately the same path with increasing Ca additions. It was concluded that 03 and Ca-Si were similar in their 77 action on chill and were both superior to Fe- Si inoculation. 4 - Hardness The data for Brinell hardness (Fig 21) did not show any significant trend for comparing the effects of the different inoculants on this preperty. The maximum spread during the entire course of this investigation was from 192 to 229, a difference of 37 points. Heat 3 was the most erratic of all with a variation from 201 to 223 BHN. The variation on ther heats was from 5 to 12 hardness numbers. 5 - Tensile Strength The available data for comparing the effects of Fe-Si with Ca-Si inoculation on the tensile strength in- dicated that Ca-Si was more effective than Fe-Si with the maximum additions of each. The curves for heat 3, Fig 31, showed a dip in the curve for Ca-Si with 3% Si pick up that could not be accounted for, although at the maxium additions Ca-Si was superior to Fe-Si. In.Fig. 33 the evidence pointed to the conclusion that Ca-Si provided similar improvement on the tensile strength for two irons which exhibited differnet carbon equivalents. In contrast, additions of .5% 81 as Fe to a higher carbon equivalent iron actually resulted in a slight- ly lower tensile strength while on a lower ca rbon equiva- lent iron the improvement was only slightly lower'than for silimar additions of Ca-Si. Evidently the lower carbon content of heat 3 as compared with heat 4 was the main 78 factor contributing to the above effects from.Fe-Si be- cause the total silicon contents were similar (see top of Fig 35.). Inoculation with metallic calcium was definitely superior to the other inoculants in regards to improve- ments of tensile properties. In Fig.38 the tensile curves for Ca additions exhibited a decided rise as comp- ared with similar amounts of Ca added as Ca-Si. Im- provement developed by .22% metallic Ca added to the ladle amounted to 10,000 psi. The maximum improvement noted from Ca-Si was approximately 5,000 psi and for Fe- 81 4,000 psi. 6 - Graphite Distribution and Microstructure Although there were no significant differences noted in the pearlite matrix structures throughout this invest- igation, variations were noted in the amounts of ferrite and in the graphite distributions. According to several papers presented in the discussion of the published lit- erature the physical prOperties were improved when the microstructure and graphite distribution were improved. This correlation was noted in the present investigation. When comparing Ca—Si with Fe-Si ladle additions for heats 1,3 and 4 it was found that the principle differ- ences in microstructure occured at or near the surface of the test bars. In all of these heats the core graphite distributions were normal (Micro l), and the Ca-Si addi- tions exhibited somewhat 1683 free ferrite in the core. 79 The maximum addition of Ca-Si (.5% Si pick up) produced' type A graphite wdth a very small amount of ferrite for these heats (Micro 5). The maximum.Fe-Si addition (.5% Si pick up) produced a mixed graphite structure (Micro2) at the surface with more ferrite in heats l and 3 (Micrc4), and was similar to Ca-Si for heat 4. Smaller additions of these inoculants for heats l and 3 indicated in the surface graphite distribution for Ca-Si at similar levels. Heat 2 showed that inoculation with Ca-Si provided improvements in the graphite distribution over Fe-Si for similar amounts added to the ladles. In fact, at the surface, the maximum addition of Fe-Si resulted in D type graphite formation (Micro3) and inoculation with a similar amount of Ca-Si produced a normal graphite structure. Improvements were noted as a result of metallic cal- cium additions to the base iron of heat 7. The base iron exhibited type D graphite at the surface (Micro6), which was mixed with type A graphite in the core (Micros). An addition of .22% Ca in the ladle produced a normal graphite pattern throughout with much less free ferrite. (Micros 7 and 9) Smaller additions of Ca caused lesser improvements in the distribution of the graphite and ferrite. It was noted that the maximum.additions of Ca-Si and Ca used during this investigation always produced irons with normal graphite distributions and little or no free ferrite. 80 E - Summary and Conclusions Fe-Si and Ca-Si have been widely used as inoculants since the paratioe of making late additions to gray cast iron was started. An investigation has been carried out to determine the relative effectiveness of these inoculants and also to make comparisons with ladle additions of metal- lic calcium. Although considerable attention has been given to inoculation theory and practice in the metallurgical lit- erature there has been little or no data given which pre- sented a valid comparison between the effects of Ca-Si and Fe-Si inoculation. The effect of the presence of Ca in the inoculating agent, or as an inoculant by itself has not been discussed in any of the literature. Data was presented in this intestigation on base irons of 2.6% o - 2.15% Si and 3.82 to 3.05% c - 1.93 t02.01% Si for Fe-Si and Ca-Si ladle additions. Data for inoculation with metallic calcium on base irons of 3.02% C -2.01% Si and 2.90%‘0- 2.24% Si was also presented. Comparisons were made of the effects of Fe-Si, Ca-Si and Ca on the transverse strength.and deflection, tensile strength, chill depth, hardness, graphite distribution and microstructure. It was concluded that, as inoculants, Ca and Ca-Si were definitely superior to Fe-Si in all instances, except that no significant differences were observed in the hardness data. The relative effects on the various properties are summarized below: 81 1- Transverse strength and deflection - At all levels of silicon pick up from Fe-Si and Ca-Si (.l% to .5%), Ca- Si was superior. The advantage of using Ca-Si over Fe-Si increased as the amounts of the additions increased. The effect 08 a .2% Si pick up from Ca-Si was equivalent to that of a .5% Si pick up from Fe-Si. Ladle additions of metallic calcium.were equal to or better than similar add- itions of Ca added as Ca—Si up to .22% Ca(maximum Ca metal addition). 2- Tensile strength - Additions of Ca were far super- ior to Ca-Si and Fe-Si in their effect on the tensile strength. Ca-Si was somewhat better than Fe-Si in this re- spect. 3- Chill depth - Ca and Ca-Si inoculation produced similar effects on the reduction of chill. In all instances Ca-Si, and therefore Ca, was superior to Fe-Si for chill reduction. An addition of .5% Si as Fe-Si was no more effective than a .3% silicon pick up from.Ca-Si. 4 - Graphite distribution and microstructure - Distinct- ive differences were noted in the graphite distribution and the amount of free ferrite at the surface of test bars from ladles inoculated with Fe-Si, Ca-Si and Ca. The effects of Ca and Ca-Si were similar in that normal graphite sturctures with very little free ferrite were produced throughout the samples with the larger additions of these inoculants. In some instances, type D graphite with a considerable amount of free ferrite were observed near the surface of irons inoculated with a .5% addition of 81 as Fe-Si. Comparisons of smaller additions, at similar levels, showed that Ca-Si 82 always produced less D graphite and less ferrite near the surface than Fe-Si. The presence of calcium in the most successful ino- culating agents suggested that the solidification charact- eristics of the base irons investigated were influenced by the resulting greater exothermic action from additions of this element. Bibliography 1 Smalley,0. "High Test Iron" Transactions A F At, vol.37,l929,pp 485-494. . 2 Marbaker,E.E. "High-Test Gray Cast Iron-Eur0pean Developements" Trans. A F A, vol.37,l929,pp. 405-416. 3 Coyle,F.B. and Houston,D.Mo l'High Test Cast Iron” Trans A F A vol37,l929,PP. 469-484. 4 "Metals Handbook" American Society for Metals,l948 Edition,p.7. 5 ”Cast Metals Handbook" A F A,1944 Edition pp.360-36l. 6 Lownie, H.W.Jr. "Theories of Gray Cast Iron", A F A, Pro-print Number 46-39. 7 Burgess,C.0; and Bishop "Alterations in Cast Irons and Properties Accompanying Use of a Strong Inocu- lant of the SiliconeManganese-Zirconium Type" Trans. A F A, vol.52,l944,pp.671-709. 8 TAlloy Cast Irons Handbook” A F A,1944 Edition,p 136. 9 Lownie,H.W.Jr. "Ladle Inoculation Improves Gray Iron Preperties and Structure" The Foundry,vol71, Nov.l943,pp.l26-127 and 188-202,Dec.l943,pp.ll6-ll7 and 194-197. 10 Lorig,C.H. "Gray Iron Metallurgical Practice" The Foundry, vol 67, March l939,pp.26-27,April 1939 pp. 74-76. ll Loria,E.A. and Shephard,H.D. “Silicon Carbide Ino- culation of Cast Iron” A F A,l947, Pro-print Number 47-49. 12 Lemoine,R.P. "Cupola High Test Iron" Trans. A F A, V01 0 42 , 1934 ,Pp 0745-761 0 * A F A (American Foundryman's Association) l3 ‘ 14 15 l6 l7 ".l8 19 20 21 22 23' to Cast Iron" The Foundry, vol 66, Jan.1938, pp.28-29,73. Boyles,A. "The Formation of Graphite in Gray Iron" Trans. A F A, vol 46,1938,pp.297-340. Vanick,J.S. "Cast Iron Foundry Practice" Trans. A F A, vol53,1945,p 51. Eash,J.T. "Effect of Ladle Inoculation on the Solidification of Gray Cast Iron" Trans. A F A, vol 49,194l,pp.887-910. Parke, R.M.,Crosby,V-A. and Harzig,A.J. "Graphiti- zation of Gray Cast Iron" Metals and Alloys, vol 9, Jan. 1938,pp.9-l4. ' Boyles,A. and Lorig,C.H. "Notes on the Undercooling of Gray Cast Iron" Trans. A F A, vol 49, 1949,pp.769 -788. ‘ D'AmiOO,C-D. and Schneidewind,Re "The Solidifioat- idn and Graphitization of Gray Iron" Trans. A F A, vol.48,1940,pp.775-803. Flinn, R.A. and Reese,D.J. "The Developement and Control of Engineering Gray Cast Irons" Trans. A P A, vol 49,1941,pp.559-607. '" Crome,L. ”Opinions on Graphitization" The Foundry, vol.73,Dec.l945,pp102-103 and 248-258. Crosby,V.A. and Herzig,A.J. ”Late SiliconAdditions Roth,E.L. "Sixty Thousand Pound per Square Inch “ Cupola Iron" Trans. A F A, vol 47,1939,pp.873-894. Hughes, H.P. andSpenceley,W. "The Developement and Production of Inoculated\Cast Iron" Abstract - The Journal of the Iron and Steel Institute, vol 151, 24 25 26 27 28 ‘29 38 31 32 1945, Number 1, p.5A. Comstock, GoF. and Starkweather, E-R. "Comparative Effects of Late Additions of Titanium and Silicon to Gray Cast Iron" Trans. A F A, vol 46,1938,pp353- 373. Schnee, V.H. and Barlow,T'E. "CoPper Aluminum, Silicon Alloys for Addition to Cast Iron" Trans. A F A, vol 47,1939, pp.725-741. Eash, J.T. "Effect of Ladle Inoculation on an Austenitic Cast Iron” .Trans. A F A, vol 50, 1942, pp. 815-829. Schneidewind, R. ”Ladle Additions to Cast Irons" Metal Progress, vol 38,1940,pp374-376. Massari, 8'0. and Lindsay, R.W. "Some Factors In- Fluencing the Graphitizing Behavior of Cast Iron" Trans. A F A, vol 49,1941, pp.953-991. Schneidewind,R. and McElwee, R.G. "The Influence fo Composition on the Physical Preperties of Gray Irons" Trans. A F A, vol 47,1939,pp491-512. BrugeSS, 0.0. and Shrubsall,A.E. “Machinable 1.5% and 2.0% Chromium Cast Irons to Resist Deterioration at High Temperatures" Trans. A F A, vol 50, 1942, pp. 405-445. Williams, J.H. ”Ladle Treatments for Iron" Foundry Trade Journal, vo168,Decl7,l942, pp345-349. Francis, J.L. "The Production.and Founding of Inoculated High DutyCast Iron" Abstract-The Journal of the Iron and Steel Institute, vol 151, 1945, mi- D . . 0 e . . - \ . . . e t . ' a ' e . . . . .._ . - I e . v . e ‘ ' , 4 . ' I e f . O . . , ‘ . a ‘ e . . 9 ‘ r . ' . mber l, p. 94A. 33' Pearce, J.C. "Acicular Cast Irons" .American Foundryman, vol 13, Feb. 1948, p. 41. 34 Hoyt, S.L; ”Metals and Alloys Data Book" Rhinhold Publishing Co, 1943, p.209. 35 Delbert, G. and Potaszkin,R. "Contribution to the Study of Special Cast Irons Having High Mechanical Resistance" Abstract - Metals Review, vol 20, Number 11, p.12, abstract 3-300. 36 Dierker, A-Ho ”Ladle Additions of Graphite to Gray Cast Iron” American Foundryman, vol 2, Nov.l940, pp.2-5. ' 37 Piwowarsky, E "Progress in the Production of High Test Cast Iron" Trans. A F~A, vol 34, 1926, pp 914- 985. 38' McElwee,R.G. "Deoxidation and Graphitization of Cast Iron" Trans. A.F A, vol 46,1938, pp.341-352. IIIIIHHIHIHIIHHIH 1293 03145 3453 3 ll. 3' I3 I I'll] 3 III II 3 l l .3! III III III II I I3 I II I]. III I II I l