AMERICAN SOCIETY FOR METALS Metals Park, Ohio 44073 Metals/Materials Technology Series ROLE OF REFRACTORY ELEMENTS IN COBALT-BASE ALLOYS M. F. Rothman, R. D. Zordan Cabot Corporation Kokomo, Indiana 46901 D. R. Muzyka Cabot Corporation Boston, Massachusetts 1984 ASM Conference on Refractory Alloying Elements in Superalloys — Effects and Av y Rio de Janeiro, Brazil 6-13 April 1984 8402-009No part of this paper may be reproduced, stored in a retrieval system, or transmitted, in any form or by any means, electronic, mechanical, photocopying, recording, or otherwise, without the prior written permission of the publisher. Nothing contained in this paper is to be construed as a grant of any right of manufacture, sale, or use in connection with | any method, process, apparatus, product, or composition, whether or not covered by letters patent or registered trademark, nor as a defense egainst liability for the infringement of letters patent or registered trademark. SAN: 204-7586 Copyright 1984 American Society for Metals All rights reserved8402-009 ROLE OF REFRACTORY ELEMENTS IN COBALT-BASE ALLOYS M. F. Rothman, R. D. Zordan ‘Cabot Corporation Kokomo, Iriana 46901 D. R. Muzyka Cabot corporation Boston, Massachusetts ABSTRACT Refractory metal elements play important roles in the design and application of cobalt~base su- peralloys. As in the case of nickel~base super Alloys, alloying elements such as molybdenum and tungsten are traditionally added to the cobalt base for the purpose of providing solid solution strengthening; hovever, the addition of various refractory metals can profoundly influence the basse physical metallurgy, mechanical behavior, and environmental properties of cobalt-base ma~ terials, From this perspective, the major re~ search pertinent to a description of the effects of refractory elements has been reviewed and sup~ plenented with limited results of recent previ- ously unpublished work. WHEN RECOUNTING THE HISTORY of the development of modern-day cobalt-base “superalloys,” nost. authors are apt to begin with a reference to the use of STELLITE® alloy No. 21 (or “Vitallium”) investment-cast turbosuperchar~ ger buckets in the early 1940s. However, the origins of today's wrought and cast cobait-base high-temperature alloys can be traced back to the work of Elvood Haynes in the early 1900s. Haynes was the first to exploit the cobalt~ chromium-tungeten and cobalt-chromiua-molybde~ nua systens for their excellent hardness and tool-waterial characteristics in the form of the STELLITE alloy fanily, initially patented in 1913. The key aspect of this work was the addi~ tion of tungsten and/or molybdenum to the co- balt-chromfun alloys Waynes had patented ear- Ler in 1907. The true extent of the improve~ fuent in elevated-temperature strength and vear~ resistance properties achieved with these re~ fractory metal additions was certainly not fully appreciated until many years later. Yet, it is clear that the roots of the present-day cobalt~ base superalloy family tree reach back to nearly 101 the turn of the century. The chemistries of some important cobalt-base alloys are given in Table 1. ‘The role of the refractory metal elements molybdenum, tungsten, tantalum, columbiua, and vanadium in cobalt-base superalloys is at once straightforward and complex. One hears the alloys of thie family commonly referred to as jolid-solution-strengthened* materials, with the primary role of refractory elements in the alloys often described as being to provide such strengthening. It 12 well known, however, that these sane refractory elenents participate in nunerous phase reactions in cobalt-base alloys, any of which can be used to provide various forns of alloy property enhancement. Table 2 is a description after Sullivan, et. al.(2) of the roles ascribed to those elements commonly found in cobalt-base superalloys. ‘Anong the effects associated with refrac~ tory metal elenents in addition to solid-solu~ tlon-strengthening are the strengthening effects related to carbide precipitation and intermetal~ Lie compound precipitation. ‘The purpose of this paper is to examine in sone detail the more complex roles played by re~ fractory metal elements in cobalt-base superal~ loys. This will be done from the viewpoint of both the physical metallurgy of the materials and their behavior in aggressive high-temperature or wear environments. Anong the many roles of refractory elements in cobalt-base superalloys which will be covered are: (1) Influence upon the nature of carbides, (2) Stacking fault ener~ sy considerations, (3)'Structure/aechanical pro~ perty effects, (4) Influence upon wear proper~ tfes, and (5)' Influence upon high-temperature envitonental properties PHYSICAL METALLURGY ‘A complete description of the physical me~ tallurgy of cobalt~base superalloys 1s beyondesttion of Cobale-tateSupees} Lye sett aatoy wo. 1-40) Suge oly’ a eat) Wake? 202 ‘iszsrer® 33 SRG 0 Cant ety Ho. 130) eessees the scope of this paper, and excellent conpre~ hensive reviews of the subject are to be found tn'Wagner and Hall,(3) Moral, et al.,(0) Sine,(5) ‘and Sullivan, et ai.(2). tare ve will present only a brief overview of the key consideration relevant to the physical metallurgy of the al~ Joya, so as to better introduce the sections which fotlow. Cobalt-base superalloys are normally char- acterized by a facercentered cuble (FCC) aatrix containing second phases. Although the second phases are usually various different carbides, precipitation of other intermetallic compounds ‘An example of the profusion ie also observed. fxn ety ni fri etn MMs of phases sometimes encountered is given in Table 3 after the work of Takeda and Yukawa(6) ‘on WAYNES alloy No. 25 (ALloy 1-605). Ae MES ey Ho. 256) wo Sot et sets rato te ne tart eit olor msm eat ee Ane ofa Tek Mo Lee than ni eitfcent pina ware ideti- food tosating tae Soe ea ad the other @ close-packed hexagonal (HC) phase. ‘Transformation of the FCC matrix ro an HCP seructure {s known to occur in a number of com baltshase superalloys, For example, the materi als of the MULTIPHASE® alloy fantly derive their considerable strength in part froa'a etrain~ Induced ¥CC to HCP transformation. (/-10).Sub- mitial atrengthening can be achieved when guch 102alloys are cold worked and aged, as illustrated by the data in Table 4 for HAYNES alloy Wo. 25 bar. TAME 4 Yensile Properties for HAYNES alloy No. 25 Bar cold Worked 30 Percent ond ‘Aged Six Hours at Vartous Tenperatures Aging Temperature U.T.S.Y.8. Elongation CP) est ket) AeRolled 236 198 10 700 240 196 n 900 250 a 10 as ae a 9 Transformation of the FCC watrix to an HOP Structure ir closely related to the common occurrence of stacking faults in many cobalt— base superalloys, Stacking faults and stacking fault energy considerations are dealt with later in this paper. ‘The majority of commercial cobalt-base al- Joys derive their strength at high teuperatures fron @ conbination of solid solution strengthen ing, from precipitation of carbides, and, to a Lesser extent, from precipitation of intermetal- Tie compounds, ‘The solution strengthening of cobait-chromium alloys through ternary additions of molybdenun, tungsten, columbiua, ané tantalum has been studied by Rausch, et. al-@2) and Drapier, et al.(3) Rausch established chat all of these elenents provide substantial in- Greases in 925°C (1700°F) yield strength vhen added. to cobalt-base alloys containing 15% to 25x chrosiua, This ts shown clearly in Figure 1, Solubility Limtes at 1200°C (2190°F) for these same elenente in the cobalt-chroalua syoten vere deterained by Drapier, ct al. The Cobalt-rich corner of a superinposition of the {200% ternary isothermal sections is show in Figure 2. Various different carbides have been iden~ tified in current comercial cobalt-base alloy ‘A suamary of the various types observed for rep~ Fesentative alloys hae been derived from a va Hety of sources and 1s presented in Table 5. In discussing the role of carbides in cobale~ base superalloys, it {8 necessary to draw the distinction between primary carbides present in the cast or wrought materials in the as-annealed or as-produced condition and those carbides ‘which precipitate upon subsequent. purposeful heat treatment, cooling from the heat treatnent, or service thermal exposure. From the work of 103 TERNARY ADDITION (w=) Fig. 1: The effect of molybdenum, tantalum, and tungsten additions on the 925°C (1700°F} short-tine yLeld strength of Co-Gr base alloys. After Rausch, et al-(12) LSELS. LITNINAN 1200°¢ (2190°F) isothermal sections of (X= Mo, Wy Cb, Ta) showing cobelebase solid-solution Held boundaries.‘asblde Pace taprtedly Feesent ta Varloue Cobale-taen Soperattoys rae es slg Me Suite ley He. 38 0) ey ae te AE Mah she Clause ond Weeton, (14) Yorrow, et al.,(15) Teel, C185" and Zordan, (179 Se de appazene thar the type and morphology of primary carbides can substantially influence the elevated temperature Propertiae and wear resistance of cobsltbase Soperaldoya, Many investigators have also examined the effects of secondary carbide preetpitation upon Cobalt alloy properties, The association of Strengthening of NAR-M 509 with precipitation Of 30g, carbides during aging bas been well ocutented, C1820)" York on less complex cobalt base systens has also been extensive, clearly cetablishing a strong relationship between secondary carbide precipitation and property changes. 21-23) Figure 3 illustrates. the har~ dening response associated with Npa0g carbide precipitation during aging of Co-23ce-0.26C and Coclonis250r-04250 alleyes co-25er-.266 700% en 450 TIME (HOURS) Fig. Co=25cr-0.26C and Co-25Cr-0.25¢ alloys. Beaven, et al.(23) Wardness versus aging time data for After 104 Intermetallic compound precipitation in cobalt~base superalloys is a subject which has | been as well investigated as carbide reactions, though perhaps less productively in terms of commercial exploitation. The wealth of possi~ ble intermetallic compounds available for po- tential use in strengthening cobalt-base super- alloys 1s evident from an exgeination of Table 6, adapted from Drapier, et al.(13) and others. rae 6 Intorstalite Conpound Types 15 Cobale-Base Super compound serveture os(Hoy Wy Tay Gb) HP, Ordered cesthh cyte reo, ontred cello, tn CO) tenets Lave Pi ate ret | cont, Wy tian Pane an cutie, ornred (em, ma, 1), ota Despite this, generation of an intermetallic precipitate Combining the high-temperature ota~ bility and versatility of the analagous N13(A1,T1) y" phase in nickel-base alloys hi proven'to be a largely elusive goal. Alloys en ploying such phase precipitation as the prinei~ pal aode of high-temperature strengthening have Jet t0 gain much comercial acceptance, though Substagetal research has been performed in this area. (26-32) ‘CARBIDES. The type and morphology of prinary car- bides can substantially affect. the propertion cf cobait-bace superalloys. ‘The influence of Tefractory metal additions to cast cobalt-base Alloys upon carbide uorphology and alloy prgpes~ ties hae beon thoroughly explored. (4-11, 33°39) Morroy, et al-C3) evaluated the effects of substituting aolybdenum for. tungsten in cast SLloye FSiei1¢ and HALA 509. They found. that, for HAR-M 509, substitution on an equivalent tonic percent basie produced nore rounded, digerete, and less continuous He script. carbide structure. ‘This vas accompanied by a slight Inerease in both tensile. and strees rupture 4ve~ tility, which vas also observed for the molyb- denunvoubstituted FSX 414 despite no apparent Tefinenent of the prinary carbide etructure. These results are in basic agreement vith the } earlier work of Griffiths and Phelps(®9) on | substitution of molybdenum for tungeen inSTELLTTE alloys No. 4 and No. 7, yielding STELLTTE alloys Wo. 5 and No. 8, respectively. ‘Taken together, the studies indicate that, on ‘an atopie-percent basis, molybdenun and tung- Sten are similar in their effect upon carbide structure. ‘The influence of tungsten on the mlcro~ structure and wear properties of alloys in the STELLITE alloy No. 12 family (31.5%Cr-1.52C~ (0.5ihin-0.8%81-22Fe-0% to 16% WeBalance Co) was examined by de Brouwer, et a1.(34) although the effect of inereasing tungsten levels upon car~ bide morphology wae not reported, x-ray dif Fraction analysis of extracted residues vas performed. The resulta, which are given in Table 7, show that M7C3 {6 the only carbide foraed upon solidification of the alloys con~ taining up to 8% tungsten. oe "3 ss 8 SBT M05 as ST W365 + Mg pity se ee 507] 4, At the 12% tungsten level a mixture of M7C3 and 'gC carbides is observed. When the tungsten Level is raised to 16 percent, only MgC-type carbides are found in the as-cast material. ‘Thie change in the carbide type with in~ creasing tungsten found by de Brouwer, et al. was reportedly accompanied by an appreciable lowering in the alloy liquidus temperature vith Little change in the solidus. This is shown in Higure 4, and it implies a finer dendrite structure for the high-tungsten compositions. More recent vork by Durand-Charre, (35) however, reports little change in the liquidus tenpera~ ture for tungeten varlations butween 8.4 and 14.9 percent in STELLITE alloy F-type materials (1S4-25¢r-22Ni-1.5C-W-Co Balance). Wot hardness was also evaluated by de Brouwer, et al. for the 0% to 16% tungsten variations on STELLITE alloy No. 12 and the results related to performance in a 600°C (1110°F) metal-to-metal wear test. Hardness at 600°C (1110°F) as a fraction of room tempera: ture hardness is plotted versus tungsten content in Figure 5b, and veight loss in the wear test ie plotted versus tungsten content in Figure 5a, 105 O Liquidus © settee 1400 S 1300 — © 200 z 1100 b r sa a6 Fig. 4: Bffoct of tungsten content on the solidification temperatures of STELLIT: 12-type alloys (1.3% G);. After de Brouwer and Coutsouradis. 34) wanowess ar s00te/ Pig. 5: (a) 600°C metal-to-netal wear behavior of alloys as a function of tungsten content. Data taken from de Brouver and Coutsouradis.(39) (b) Fraction of room temperture hardness retained ‘at 600°C as a function of tungsten content in STELLITE alloy No. 12 materials. A combined plot of vear weight loss versus hardness at 600°C (1110°F), including addi~ tional data reported by de Brouwer, is shown in Figure 6. The authors note that tungsten is {ndispensibie to ensure that the alloys retainim T T T T T e ‘ e Sob 4 oe 4 e 5 1 ,@ | e7——~s- 200 250~=«400=S«80==«SO Hardness at 600°C, BV 6: Wear of STELLITE 12-type alloy deposits as a function of their hardness at 600°C. After de Brouwer and Coutsouradis. (2! 9 high hardness at clevated temperatures. The relationship of wear resistance to hardness, though not absolute, also indicates the need for high tungsten 1s That the hardness i not an absolute guide to the wear resistance of cobalt-base super alloys ig illustrated by the recent work of Zordan.(17) Figure 7 contains photomicrographs which depict the as-cast microstructures of four different cobalt~base alloys vith different car- bon and tungsten contents. Data relating the hardness and abrasive wear resistance of these same four alloys is presented in Table 8. Clearly, hardness 1s no guide to either ox- pected wear performance or microstructural condition in this case. The finer particle size and interparticle spacing conbined with the increased amount of MyC3 and MgC carbides Bi Awa v TAME 8 low sexs ASTM 6-658 Procedure (7) Vouume Tage tardness ———Alley__ Gam)_GRookwet_ @) STELLITE alloy Yo. 20. 23.3 53 SELLLAME alloy No. 151.7 3 STELLITE alloy Wo. 696 68.2 50 STELLITE alloy No. 663.8 rt {in STELLITE alloy No. 20 probably contrtbute to {te auperior weet perforasnee. ‘the influence of the addition of elenents such as columbiun, tantalum, zireontun, and Cteantun upon the primary carbide strecture of cobalt-bage:alloye’ hao also been thorsushl explored. (36-39) youdelis and Kwon(28) found that addittons of up to 2.32 colunblur oF 3-252 tantalun dramatically refine the primary carbide size in cast STELLTTE alloy No. 21 by promoting the formation of fine NC carbides tn’ place of coarser Mz4Cg carbides. Drapier, et al.(39) ob= served sintler behavior for 1-52’ colusbiun adit— tions to tron-base alloye containing 23% chrom{— tm, 52% cobale/Sk nickel, 42% cobait/9% nickel, 32k cobale/12% nickel, all vith 0-252 carbon, The carbide refinement was accompanied by a aud- stantial increase in stress/rupture.steesgeh, 40 shown in the Larson-Miller plot tn Pigere 8. TRESS SO l moan (ase i oe L Fig. 7: Comparison of the microstructures of four alloys with varying tungeten and carbon content; STELLITE alloys No. 20 (a), No. 1 (b), No. 694 (c), ad No. 6 (d). Light etching particles are M703. Dark etching particles are MgC, After Zordan.Pa Tae 20 + teen P = (K)(20 + log t) “ we a ww wo ey Py Pe re at a sauce foe of PT ee als 300 wool, 8 szcessnicaser-0.25¢-1.scb Seo 40 Lg sncontanicasrco.a5e-1.5¢8 ls val tn 00} ™ 3 a zoo a) = 100 & EL ~\ qe & 7 2 10 F sakssnesspurne ‘ fe ‘o ol. 8 7 ab 2 6 40 4; 5 oe 42 ao 46 40 50 52 x 1073 po P= TCR) (20 + top © Pm TEKN20 + Ke * Fig. 9: Effect of 0.4% Cb and 0.83% Ta additions upon the siress/cupture properties of Alloy X-40, Adapted from Davin, et al.( Fig. 8: Streso rupture properties of Fe-Co-Ni- Gr-€ alloys with added columbiua. Data from Drapter, et al. Barlier work by the sane authors(38) for 0.83% tantalua and 0.4% columbium additions to Alloy K-40 with 0.5% carbon had yielded comparable improvements in rupture properties, shown in Figure 9. Fukui, et al.(3/) performed the same type of studies involving small titanium, colusbiun, fand zirconium additions to Cast FSX~414 alloy. ‘They found that much smaller eutectic carbides were formed when, for example, 0.15% titaniua, 0.19% zirconium, and 0.24% colunbiun were added to the base alloy. Optimization of 980°C (1800°F) rupture strength, however, was found to be possible only by controlling the ratio of carbon content to the amount of NC carbide sta~ bilizers added. This 1s illustrated by the plot in Figure 10 of 1000-hour rupture strength at 980°C (1800°F) versus the atomic weight ratio of (T+ Cb + 2r)/C. This shows that optimun Tupture 1ife 12 obtained at an atomic weight ratio of MC stabilizer element to carbon con- tained of about 0.3. Fukul concludes that the decreage in rupture strength at higher ratios 4e due to excessive primary MC stabilization, depletion of matrix carbon content, and reduc tion of in-service strengthening by ¥93Cg car- bide precipitation. Goneyary results wore reported by Davidson, (24) who examined the aging charac~ teristics of HAYNES alloy No. 25 and molyb= denun-substituted alloys of the same fanily 107 oarze casoors ad sess to proouce nurrune Tn 1000 HOURS Gur eye lente eee AEA NG, sant Fig. 10: Relationship between 982°C (1800°F) str rupture strength of FSX~414 alloys odified with various NC formers and the ratio of such elenents to the carbon content (atomle percent basis). After Fukui, et al, (37 with additions of titantum and columbium. In hie studies involving 16-hour exposures at tem peratures between 650°C (1200°F) and 1100°C (2015°F) Davidson observed that 1~percent addi~ tions of colusbium or titanium to the nolyb- denun-substituted heats shifted the peak in thehardening response upward from 700°C (1290°F) to 800°C (1470°F). Ue related this shift to the observed precipitation of secondary colum— bium-rich, MC carbides, which appeared in a 200 thick platelet morphology accompanying the more profuse globular Mp3Cs precipitation. The Cbc platelets apparently precipitated pre~ ferentially upon dislocations, and being finer in size and more thermally stable, could logi- cally be expected to improve the strength of the material more efficiently than the larger Mg30g precipitates. ‘This secondary precipitation of MC car- bides and associated strengthening was also observed by Ranaswany, et al.(21) for Co-20¢r =L0Ni-5Ta alloys with’ from 0.13% to 0.29% car- bon. They report that TaC precipitation occurs upon aging in the range from 650°C to 850°C (1200°=1560°F), wih repeated precipitation upon dislocations giving rise to the formation of UCP stacking Faults in the FCC matrix, TaC pre~ cipitates also form preferentially upon stacking faults induced by prior cold work, with the aged wrdness increasing appreciably as a result, much as in the case of cold-worked and aged HAYNES alloy No 25. STACKING FAULTS ‘The role of stacking fault energy in nickel-hase superalloys 16 perceived to be straightforward, with the addition of those elenente which pronote lover fault energy lead ing to high strength. In cobalt-base alloys the situation 42 complicated in that low fault ener- gy can also lead to the FCC* HCP allotropic transformation which weakens cobalt~base alloys. As Will be discussed, a balance is required. The occurrence of HCP stacking faults in otherwise FCC matrix cobalt~base superalloys is widely known, and the relevance of stacking fault energy considerations to the occurrence of the FoC + HCP allotropic transformation has been discussed by several authors.(2+4,5,41) The role of various elements in the occurrence of the allotropic transformation bas been re viewed by Sime(>) and in detail by Glanet, et a1,(1) In general, it 1s proposed that refrac~ tory metals such as'molybdenum and tungsten tend to atabilize the HCP structure while others such tantalua and coluabium appear to be FCC sta~ bilizers. This is {1lustrated by the diagram presented in Figure 11, after Sims.(5) ‘The relationship between stacking fault energy and the allotropic transformation is that the fault energy approaches zero when the FCC matrix transforms allotropically to HOP. Sullivan, et al.(2) point out an important con~ sideration in the fact that, even in the suppos~ edly stable FCC temperature regime, the stacking fault energy {e sufficiently suall to allow for fatacking fault formation to appreciably influ- fence creep properties. Sims proposes that the diagram shown in Figure 11 relating the influ~ fence of various elements upon the allotropic transformation can also be used as at least a general qualitative guide to the effects of 108 ree hanse0 1 Fig. 11: Effect of alloying additions (in atom de percent) on the HCP+ FCC transformation in cobalt as a function of solubility in FCC cobalt. After Sine. (3) alloying additions on stacking fault energy: further cautions that, for cobalt-base alloys, stacking faults varn of proximity to the HGP structure, and can be regarded as a precursor to more massive HCP formation. ‘The relationship between the chemistry of cobalt~base alloys and stacking fault formation fo otill a relatively complex issue, hovever. For example, Drapier and Coutsouradis({2) exan~ ined Co-Ce-ta alloys aged at 700°C (1290°F). AS wentioned previously, tantalum night be expected to ralee the stacking fault energy of Co-Cr al- Joys; however, in thie case, precipitation of Costa was observed, and as a consequence, the density of stacking faults in the tantalum depleted satrix increased substantially during aging. A similar phenomenon was noted by Ranaswany, et al.(21) for the Co-20Cr-10Ni~Sta~ 0.13 to 0.296 alloys they studted, although tn this case TaC precipitation on dislocations led to the fault formation. Such precipitatfon-induced fault foraation in cobalt-bage alloys 19 well documented in the Licerature. (21-25,82543) “Beaven, et al, (23) describe an interesting case for’ Co-25%Cr- 0.2626 alloy. Material predeforaed 10 percent displayed sone initial transfornation to the HGP phase. Aging was performed at 650°C (1200°F) for up to 500 hours, and, in the early stages of aging, Mz3Cg precipitation was ob- ferved to occur upon the Prank partial disloca- tons bounding the individual stacking fault With longer tines, recrystalltzation and carbide coarsening were observed to occur simultaneous ly, eventually resulting in the complete trane- Foruation of the matrix toa fault-free UCP structure after 500 hour The mportance of stacking fault energy considerations in the wear properties of cobalt: ‘base alloys has also been recognized. (4) Low stacking fault energy appears closely related Heto supgrtér galling performance. Dennis and Stiiite5) evaluated the wear resistance of pute Co-to\and Con clectrodeposited. coatings Sn forging ates, and they found a clesr rela” Clonahip teevees the weer volome measured oa the forging dies and the percent of tungsten or olybietun in the cobalt alley coating. Ths {fe show in Figure 12. Fig. 12: Relationship between the alloying ele~ ment content of Co-Mo and Co-W die coatings and their degree of wear and hardness after, forging 1000 billets. After Dennis and stii1.(45) Surprisingly, while hardness increased continu- ally with increasing tungsten or molybdenun content to very high levels, vear resistance appeared to stabilize at an’ optimum low level after a content of about 4%-6% molybdenum or 10z-15% tungeten was reached. Since both of these alloys were pure binary, single-phase materiale, it might be suggested that the ex- fotence of an optimum level of molybdenum or tungsten addition for achieving improved wear resistance could be related to a critical low level of stacking fault energy. INTERMETALLIC COMPOUNDS ‘The atrengthening of cobalt-base super- alloys by precipitation of intermetallic con pounds has remained very much secondary in 1 portance compared to carbide strengthening. Nevertheless many intermetallic compounds have been identified in cobalt~base alloys, and se- veral investigations have been undertaken in the past to establish which of these phases 109 hold the most promise for comercial exploite— tion. Sone of the intermetallic compounds which show some value in this respect are | CogW and Co3Mo (both ordered hexagonal), CoyTa (either metastable ordered cubic or rhombohedral forms), CozTi (ordered cubic), and CopCb (Laves hexagonal) ‘An excellent summary of the precipitation hardening characteriatice of intermetallic com pounds in Co-Cr-W, Co-Cr-Mo, Co-Cr-Ta, and Co~ Gr-Cb alloys has been presented by Drapier, et a1,(13) The 100-hour aging response of selec ted alloys at various temperatures is shown in Figure 13. bd (21909F1/2h«/WO 4 a) it ete me ets ternary Co-Gr alloys. After Drapier, et al. Substantial strengthening ts evidenced in all of these systems, yet the strength is readily lost at temperatures above the range from 750° to 850°C (1380" to 1560°F) as a result of overaging. ‘A great deal of the effort expended in the ech for a stable, high-temperature strength= ening precipitate in cobalt-base alloys has beentrected atthe Cost analog of the.) precipi tate in alckel-base superalloys. ( se ptobtce'teeke ete staptey of thie Initially coherent, ordered FCG precipitate at tenperatures above 750°C (1380°F). A number of Investigators have exaained the Influence of re~ Eeactory metal additions to Co-Cr-T or Co-Ni~ CeT alloys, with mixed results. Drapter, et a1, (0) reporeed that transforaation of the or dered FCC Coytt y' precipitate in Go-Cr-Tt al- loys to an HGP structure 1s promoted by the Addition of uolybdenua in excess of 3% to 5%, with the same effects expected for the other refractory clenents. Sintlar results are re~ ported by Viatour, et al.(29) who observed Aiscontinuous, lamellar HCP phase precipitation 4m Co-Ti-Wo and Go-Ti- alloys aged at 700°C (1290°F).Kineeics of the precipitation were observed to increase with increasing tungsten content in those alloys ‘A different behavior was noted by Viatour, et al. for ternary additions of tantalua to Com 1 alloys. Here discontinuous lanellar precip- tation was not observed at 700°C (1290°F), but ‘only at 800° and 900°C (1470° and 1650*”).” At 000°C (1830°F) platelets of an intermetallic UCP precipitate vere found, Likely a version of the 8 Cosfa-type structure: This difference for tantalun was not found to carry over to Co-Cr- TUX alloys investigated, hovever, where addi- tions of over 1 atoaie percent of ‘molybdenum, tungaten, oF tantalua vere all found to promote formation of the lamellar phase upon aging at 700°C (1290"F). Walder and Martyt31) {avestiga~ ted the effects of minor columbium additions to a Co-16XGr-4ZT~31Mo alloy, and although no micro~ structural results vere reported, it 1s clear from the hardness data that columbium does not Increase the precipitate stability tenperature range beyond the 750°C (1390°F) of the base alloy. Results obtained relative to more complex systems have been fever and perhaps lees 11~ Tuminating as far as understanding structure/ couposition/property relationships in these alloys. Further investigation by Drapler and Coutsouradie(42) into the Co-Cr-Ta systens modi~ fied with nickel, vanadiva, aluminum, and titan ium denonstrated that an effective strengthening ordered FCC Co3Ta-type precipitate could be obtained with aging at least up to 800°C (170°). Ae aging texperatures approaching $00°C (1650"F), however, the rhombohedral Costa was always observed. Perhaps the most systenatic studies per~ formed upon complex syateas were done by Acunetu Taking the approach that anala~ gous nickel-base superalloys ake simultaneous Use of y'-forming elements, solid solution strengtheners, and carbide forsers, Acunclus evaluated complex multicosponent cobalt-base alloys containing as many as 11 elenents, The fain thrust of this vork vas still directed at Increasing the stability of the y'-type precipl~ tate, In this respect he concludes that the greatest tuproveuent in the stability of the y phase (in a base alloy containing nominally 110 20%Ce-152NI~32T4~3KAL-Balance Co) is achieved Ehrough the addition of columbium and/or tanta~ lun. ‘This {s readily apparent from an examina: tion of Figure 14, which depicts the stress rup- ture properties of alloys containing various re~ fractory wetal combinations. be 10NG0 + 190 9 sraessnurrute Se ate ten Sa 2 Se a 5 sa s lect 3 sree rnc, Ba, qok eels aon QOS a0 * mS Ra hs 10.0 XN iS 20 sists P= feR)(20* Joa 9 1m come Effect of various refractory metal elenent combinations upon the stress rupture properties of a cobalt-base alloy. Data from ‘Acunetua. (28) Herchenroder(46) observed a basically simi~ lar trend in alloys containing 21%¢r-242N1 and varying amounts of molybdenum, tungsten, and tantalum, but allowed to age harden during test- ing (test performed upon as-solution-treated ma~ terial). He noted a synergism between the tung- sten and tantalua content, claiming a ratio of tantalum to tungsten between 1.5 and 2.5 as being optimum. This ie illustrated by the data in Figure 15. The disadvantageous effect of nolybdenum addition was also noted. One novel use for intermetallic compound precipitation in cobalt-ba been the development of the TRIBALOY® cial wear-resistant materiale. These alloys such as TRIBALOY alloy No. 1-800 (28%Mo-17%Cr~ 328i-Balance Co), develop massive Laves phase in fan HP matrix. ‘The hardness of the Laves phase is about 1100 on the Vickers scale, which con pares quite favorably with that of typical tool steels (950 Vickers). Such alloys are idealpen (20 + tes 9 srasss nurture 11758¢-1203%¢ (2130°F-2200°F) Fe 10000 + toe 9 Fig. 15: Stress-rupture properties of Co-Cr-Ni alloy containing various refractory metal additions. Data taken from Herchenroeder (46) choices for wear applications in poorly or pon lubricated, high-temperature environments (47) Currently,’ such materials are being considered for cylinder sleeve materials in advanced diesel engines. ENVIRONMENTAL RESISTANCE The role of various elenents in the pro~ ceases of oxidation and hot corrosion of com balt-base superalloys has been generally sunmar- ized by Sins,() and more specifically addressed by Beltran.(48) table 9 is a summary of Sins's catagorization of elenental effects in oxida~ tion, It is evident from this summary that he considers all refractory elenents haraful to Some degree, with the exception of tantalua. This 18 in baste agreement with the data of Davin, et al.,(49) who found molybdenum and co~ lunbitm £0 be particularly deleterious in Co- 30%cr-X alloys tested in still air for five hours at various temperatures. Beltran(48) as~ eribes the inferiority in oxidation behavior of Alloy WI-52 compared to Alloy MAR-M 509 to the colunbiun content of the former versus the tan- talum content of the latter. le also points to the 9 tantalum tn MAR-M 302 as the probable reason for that alloy's excellent relative re~ sistance to oxidation, citing a comparison of all of these alloys nade by Wheaton in earlier work.(50) ‘This comparison is presented in Figure 16. Interestingly enough, Boone and m TABLE 9 Role of Alloying Additions to Cobalt-Rase Materials in Oxtdatton(5) Probable Effect of Addition ALloying on the Oxidation Behavior of Element. ‘4 6o(20-30)6r Base ‘Mdtantun _Innocuovs at Low Level: Hirconlun —Innocuovs at Low Lavele Carbon Slightly Deleterious; Tes Up Chromtun| Vanadium ‘araful, Even at 0.52 Coluabiua — araful Even at 0.5% Tantalum Beneficial to Hoderate (< 5%) levela Molybdenum Harmful; Forms Volatile Oxides Tungsten _Innocvotis below ~ 1800°F (lo00"C), Kaearer > 1800°F (200°C); Foras Volatile Oxides veers Benefietal; Improves Seale Adherence Nickel May Bo Slightly Deleterious Manganese Beneficial; Induces the Fornation of Spinels Iron Tends To Induce Spinel Formation Sullivan(52) equate the inferiority in thermal fatigue resistance of cast Alloy WI-52 in com parison to MAR-X 302 to the poorer oxidation resistance of the colunbtun-bearing WI~52, though differences in carbide morphology were also cited as being related to the degree of ‘oxidation penetration. The beneficial effects of tantalum as op- posed to columbiua are algo deaonstrated in the Work of Irving, et al.(1) for Co-20%cr-K alloys containing nominally 4%, 7% and 10% columbium or tantalum. Tests were performed in oxygen and air at temperatures between 900° and 1100°C (1650° and 2010°F) for up to 350 hours. The tesults illustrate that columbium accelerated the oxidation rate while tantalus proved benefi~ cdal at all tenperatures. This is shown by the curves in Figure 17 for oxidation studies con— ducted at 1000°C (1830°F). The authors relate the ability of the tantalum-containing alloys to form a complex oxide, Cra0y, to the observed better oxidation resistance. An analogous. coluabiun-bearing oxide was not observed. Beltran(48) gimilarly warns against large additions of vanadium.WEIGHT GAIN (mo/en?) Time (HOURS) Fig. 16: Static oxidation of selected cobalt=base superstioys at 1095°C (2000°r). After Beltran. (48) In the case of classical “hot corrosion” (iay804 Eluxing of protective oxide scales with consequent internal attack), the role of Tetory wetgle 1s somevhat less well defined, Beltrant48) generalizes that, to a degree, Te~ fractory elenente affect hot corrosion reste tance in a fashion quite similar to the way in which they affect oxidation resistance. Ue cites columbiun as being quite haraful in Alloy WE-52, but Less so in Alloy $816. Tung- ten way be deleterious in UAYNES alloy No. 25, but tantalum io cited as being innocuous in MAR-H'302. Performance of these alloys 12 com pared in Pigure 18, Nagarajan, et a1.(53) found that the pres- ence of wolybdents in Co-10% to 25%¢r alloys caused accelerated and sonetines catastrophic hot corrosion for eamples tested at 900°C (U650°F) An NagS0g and NaCl + Nag804 environ~ ments. In Nag80q, addition of 10x molybdenum to 10% to 15% chromium alloys results in catas- trophic corrosion caused by classical NayS0, acidic fluxing. In the higher chromtua-content alloys, addition of 10% molybdenun results in catastrophic corrosion as a consequence of the formation of low melting point, volatile oxides of molybdenum which cause extensive scale break avay. In concurrent work, Nagarajan, et a.04) Feport that tantalum additions of up to 102 ap- pear innocuous by contrast. ‘They caution, how ever, that when oufficient tantalum 1s present to eduse a continuous second-phase network to 2 x co-00r-7. ‘a co-20C4-9 & WEIGHT GAIN (eo/en” THAME (wOURS) WEIGHT GAIN (ne/ee®) ea 30s SES a7 uite (hours) Fig. 17: Comparison of the 1000°C (1830°F) oxidation characteristics of various Co-20% alloys with additions of coluabiun and tan- talum. Teste conducted in flowing oxygen. (2) columbian series (by tantalun series. After Irving, et al. develop, extensive penetration attack along, phase boundaries will occur; Richards and Stringer(35) have studied the influence of tungsten, tantalum, and columbium additions to Co-25%Cr alloys upon resistance to corrosion in molten NaS0, at 900°C (1650°F). They found that the base alloy, Co- 25%Cr, exhibited very Little attack after a 100-hour immersion in the molten sulfate. In contrast, the addition of 7.5% tungsten caused extremely rapid corrosion to occur in molten Na7S04 at 900°C (1650°F). Replacing the tungsten with 2.5% colunbiun caused the rate of attack upon the material to be even more rapid, hovever, replacing tungsten vith tantalum at levels up to 8%, however, substantially reduced the rate of attack.SUYMARY AND CONCLUSIONS Considerable literature existe which char~ acterizes the role and influence of refractory metals in cobalt-base superalloys. It has been observed that elenents such as molybdenum, tung~ sten, colunbiua, tantalum, and vanadium play in portant roles in strengthening of these alloys. They do so by providing solid-solution strength- ening, by influencing carbide morphology and precipitation characterietics, and to a lesser %0 60 40 oo mana 509. MAR-M 302 20 x40 MAXIMUM PENETRATION PER SIDE (MILS) MAXIMUM PENETRATION PER SIDE (mm) 870 755°C Tawe Fig. 18: Relative hot-corrosion reststance of cobslt-base superalloys. After Beltran. ( extent by providing for precipitation of inter- metallic conpounds. Refractory elements exert & strong influence upon the FCC to HCP allo- tropic phase tranaformation in cobalt~base al— Joys and are important in considering stacking- fault related phenomena in these materials. ‘the observable effects upon various cobalt base alloy properties, which are at least par~ tially attributable to refractory elenent con~ tent, include taproved hardness, high-tempera- ture strength, phase stability, and resistance to various forms of wear. Refractory elenent additions to cobalt-base alloys are generally Yess favored for thelr effect upon environmen tal properties such as oxidation and hot cor~ rosion., Some are quite deleterious, while other, such as tantalum can be beneficial. REFERENCES 1, STELLITE ~ A Wletory of the Haynes Stellite Company 1912 to 1972, Ralph D. Gray, Pub- Tished by High Technology Materials Divi~ ‘sion of Cabot Corporation, Kokono, Indiana (981). 113 Lo. a. a. B. 1a. 45. 16. aw. Cobalt-Hase Superalloys ~ 1970, @. P. Sullivan, #. J. Donachfe, Jr, fand F. R, Morral. Published by the | Cobalt Inforwation Center, Brussels, Belgium (1970). | ‘The Physical Metallurgy of Cobalt~Base Superalloys,” H. J. Wagner and A. M. Wall. DMIG Report 171, published by the Battelle Menorial Institute, Columbus, Ohto (1962). “Mterostructure of Cobalt~Base High ‘enperature Alloys,” F. R. Morral, L, Habraken, D. Coutsouradis, 3. M. Draper, and M. Urbain. Technical Report No. D8=21.1, American Society for Metals, Metals Park, Ohio (1968). c. T. Sims, Journal of Metals, Vol. 21, No. 12 (1969), pp. 27-42. 8. Takeda and N. Yukava, Nippon Kinzoku BLL, 12 C1367 )5 pps T8480 | ‘A. H. Graham and J. L. Youngblood, Met. ‘eang., Vol. 1 (1970), pp. 423-430. J. M. Drapier, P, Viatour, D. Coutsouradis, and L, Habraken, Cobalt, No. 49 (1970), pp. 171-186. M. Raghavan and B. J. Berkowitz, Scripta Met., Vol. 14 (1980), pp. 1009-1012. J. 8. Slaney and R. A. Niebsolo, Metailogeaphy, Vol. 16 (1983), pp. 137-160. V. R, Ishwar, Cabot Corporation Internal, Report No. 11198, Kokomo Technology Department (1983). J.J, Rausch, J. B. MeAndrew, and 6. R. Simcoe, High Tenperature Materials IL: Proceedings of Conference, April 26-27, 1961, Cleveland, ‘Ohio. Gordon & Breach (1962), pp. 259-279. J. M. Drapler, J. L, de Brouwer, and D. Goutsouradie, Cobalt, No. 27'(1965), pp. 59-72. F. J. Claus and J, Ws Weeton, NACA ‘Technical Note No. 3108 (1954). H. Morrow IIL, W. P. Danesi and D. 1, Sponseller, Cobalt, 1973-No. 4 (1973), pp. 93-102, K. Lebl, Cobalt, No. 38 (1968), pp. 29-35. R. D. Zordan, to be published.18. 19. 20. a. 22. 23. 24. 25. 26. 27. 28. 29. 30. 31. 32. 33. J. M, Drapier, V. Leroy, C. Dupont, D. Coutsouradis, and L. Habraken, Cot No. 41 (1968), pp. 199-213, M, J. Woulds and T. R. Cass, Cobale, Now 42 (1969), pp. 3-13. A.M. Beltran, C. 1. Sims, and MN. 7. Wagenheim, Journal of Metals, Vol. 21, No. 96 (1969), pp. 39-47. V. Ramaswany, P. R. Swann, and DIR.P. West, Microstructure & Design of Alloys. ‘Proceedings of Conference, London, 1973. Institute df Metals (1973), pp. 340-344. P. A, Beaven, P. R. Swann, and D.R.F. West, Journal of Materials Seience, Vol. 13 (978), pp- 691-701. P. A. Beaven, P. R. Swann, and D.R.F. West, Journal of Materials Seience, Vol. 14 (979), pp: 354-364. J. H. Davidson, Superalloys, Metallurgy. ‘E Manufacture. Proceedings of Conference, Seven Springs, Pennsylvania, Septenber 12-i5, 1976. Claitor's Publishing Division (1976), pp. 275-286. B. E. Jacobson, Met, Trans., Vol. 11A (1980), pp. 1167-1186. ©. Rogister, D. Coutsouradis, and L: Habraken, Cobalt, No. 34 (1967), pp. 3-9. B, Bollenrath and W. Rhode, Cobalt, No. 34 (1967), pp. 18-36. D. 8. Acuncius, Stellite Division of Cabot Corporation Internal Report No. 8173 (972). P, Viatour, J. M. rapier, and D. Coutsouradis, Cobalt, 1973-No. 3 (1973), pp 67-76 J. M. Drapier, B. Viatour, D. Coutsouradis, and L, Wabraken, Superalloys, Metallurgy & Manufacture. Proceedings of Conference, Seven Springs, Pennsylvania, Septeaber 12-15, 1976. claitor's Publishing Division (1976), pp. 395-404. A. Walder and M. Marty, Modern Developments in Powder Metallurgy, Vol. 14 (1981), pp. 115-130, N. Gs Ingesten, R, Warren, and M. Dahlen, Metal Science, Vol. 17 (1983), pp. 159-166. W, Griffiths and R. B. Phelps, Metallurgia, Vol. 69 (1964), pp. 2-4. 34. 35. 36. 37. 38. 39. 40. a 42. 43. 4b, 43. 46. a7. 48. 49. 50. sa. 52. J. Ls de Brouwer and D. Coutaouradie, Cobalt, No. 32 (1966), pp. 141-147. ©, Durand-Charre, $. Haar-Thibaule, and B. Andries, Metallurgie, Vol. XXII, No. 1 (1982), pp. 33-38. W. V. Youdelis and 0. Kwon, Metal Sctence, Vol. 17 (1983), pp. 379-384.” Y. Fokut, R. Sasaki, F. Hataya, and T. Kashimura, Met. trans., Vol. 124 (4981), pp. 1033-1039, A. Davin, J. M. Drapier, and D. Coutsouradis, Cobalt, 1974-No. 2 (1974), pp. 32-33, J. M. Drapier, A. Davin, and D. Coutsouradis, ‘cobalt, 1975-No. 2 (1975), pp, 32-42. S. Wamar-Thibault, M. Durand-Charre, and B. Andries, Met. Trans., Vol. 13A (1982), pp- 545-550. A, Glamei, J. Burman, and J. Friese, Cobalt, No. 39 (1968), pp. 88-96. J. M. Drapier and D. Coutsouradis, Cobalt, Wo. 39 (1968), pp. 63-74. K. Rajan, Met. Trans., Vol. 13A (1982), pp. 116171166. K. J. Bhansali and A. B. Miller, Wear of Materials. Proceedings of Conference C981), pp. 179-185. J. K. Dennia and F. A. Stil1, Cobale, 1975-No. 1 (1975), pp. 17-28. R. B. Herchenroeder and C. M. Augustine, U.S. Patent No. 3, 802, 934 (April 9, 1974). ©. B. Cameron and D. P. Ferris 1974-No. 3 (1974), pp. 49-53. » Cobalt, A.M. Beltran, Cobalt, No. 46 (1970), pp. SLi. A. Davin, D. Coutsouradis, and L. Wabraken, Cobalt, No. 35 (1967), pp: 69-77. H, L. Wheaton, Cobalt, No. 29 (1965), pp. 163-170, G. W. Irving, J. Stringer, and D. P. Whittle, Corrosion Science, Vol. 15 (1975), pp. 33% D. W. Boone and C. P. Sullivan, Fatigue at Elevated Temperatures, ASTM Publication STP-520 (1973), pp. 401-414.53. V. Nagarajan, J. Stringer, and D. P. Whittle, Corrosion Science, Vol. 22, No. 5 (1982), pp. 41-453. 54. Ibid, pp. 407-427. 55. M,N. Richards and J. Stringer, British Corrosion Journal, Vol. 8 (1983), pp. 167-172. ACKNOWLEDGEMENT ‘The authors would like £0 express their sincere appreciation of the invaluable assis~ tance rendered by Julia Carter in the prepara~ tion of this paper- ‘TRADEMARKS, MULTEPHASE 1s a registered trademark of SPS Technologies, Inc. STELLITE, HAYNES, and TRIBALOY are regis~ tered trademarks of Cabot Corporation. MAR-M Le a registered trademark of Martin Matala Corporation. AIRESIST is a registered trademark of Garret Turbine Engines.This paper is subject to revision. Statements and opinions advanced in papers or discussion are the author's and are his responsibility, not ASM's; however, the paper has been edited by ASM for uniform styling and format. Printed and Bound by Publishers Choice Book Mfg. Co. Mars, Pennsylvania 16046