\ 3 P& '* ^ " ^^ *~ ^ NBSIR 83-2679-2 Technical Aspects of Critical Materials Use by the Steel Industry Volume II B II Proceedings of a Public Workshop; "Trends in Critical Materials Requirements for Steels of the Future; Conservation and Substitution Technology for Chromium". 4-7 October 1982 Vanderbilt University Nashville, TN Public Workshop Sponsored by: U.S. Department of Commerce, National Bureau of Standards U.S. Department of the Interior, Bureau of Mines U.S. Department of Defense, Army Research Office June 1983 Center for Materials Science U.S. Department of Commerce National Bureau of Standards Digitized by the Internet Archive in 2013 http://archive.org/details/technicalaspectsOOmehr NBSIR 83-2679-2 Technical Aspects of Critical Materials Use by the Steel Industry Volume II B: Proceedings of a Public Workshop; 'Trends in Critical Materials Requirement for Steels of the Future; Conservation and Substitution Technology for Chromiun." Sponsored by: U.S. Department of Commerce, National Bureau of Standards U.S. Department of the Interior, Bureau of Mines U.S. Department of Defense, Army Research Office 4-7 October 1982 Vanderbilt University Nashville, TN Chairman Allen G. Gray American Society for Metals Adjunct Professor of Metallurgy Vanderbilt University June 1983 U.S. Department of Commerce, Malcolm Baldrige, Secretary National Bureau of Standards, Ernest Ambler, Director U. S. Depository Copy Contents VOLUME I: SUMMARY REPORT EXECUTIVE SUMMARY ix INTRODUCTION 1 PROCESSING 6 Current Economic Status of the Industry 6 Technical Status of the Three Major Sectors ... 7 Technical Developments in Processing 9 Continuous Casting 9 Primary Melting and Refining Processes ... 10 Real-Time Control 13 Near Net Shape Technology 15 Surface Modification 16 Steel Plant Refractories 18 Import Implications of Advanced Technical Developments 19 Recycling 20 SUBSTITUTION 23 Chromium Functions in Steel 24 Substitution Options for Chromium in Corrosion and Oxidation Resistant Stainless Steels ... 24 Substitution Options for Chromium in Structural Alloy Steels 29 INSTITUTIONAL FACTORS 32 Specifications 32 Qualification 32 Composition of the Natural Defense Stockpile . . 33 Research 34 CRITICAL MATERIALS AND PRODUCTIVITY 37 REFERENCES 39 ii VOLUME II A: Proceedings of a Public Workshop; "Trends in Critical Materials Requirements for Steels of the Future; Conservation and Substitution Technology for Chromium". Significance of the Workshop CRITICAL MATERIALS NEEDS FOR FUTURE STEELS AND THE CHALLENGE OF THE CHROMIUM SITUATION, Allen G. Gray, Workshop Organizer and Chairman; Vanderbilt University and American Society for Metals. PI Workshop Keynote Session INDUSTRIAL LIFE WITHOUT CHROMIUM-TECHNOLOGICAL CHALLENGES, Arden L. Bement, Jr. , TRW Inc. P2 POTENTIAL FOR CRITICAL MATERIALS CONSERVATION IN THE INTEGRATED STEEL INDUSTRY, Gordon H. Geiger, Chase Manhattan Bank. P3 POTENTIAL AREAS FOR CHROMIUM CONSERVATION IN STAINLESS STEELS, R. A. Lula, Consultant. P4 THE VALUE OF GENERIC TECHNOLOGY: SUBSTITUTION BASED ON HARDENABILITY, Dale H. Breen, Gear Research Institute. P5 Users' Views on Steels Needed For The Future Technological Trends on Critical Materials Required For These Steels TRENDS AND NEEDS FOR FUTURE STEELS IN BUILDINGS AND BRIDGES, Lewis Brunner, American Institute of Steel Construction. P6 THE MATERIALS USE AND RESEARCH OUTLOOK IN THE RAILWAY INDUSTRY, William J. Harris, Association of American Railroads. P7 TRENDS AND NEEDS FOR FUTURE AUTOMOTIVE STEELS, George H. Robinson, General Motors Corporation. P8 CATERPILLAR TRACTOR CO. VIEWS ON STEELS NEEDED FOR FUTURE APPLICATIONS, Dennis B. O'Neil, Caterpillar Tractor Co. P9 CHEMICAL EQUIPMENT - USER'S VIEWS ON STEELS NEEDED FOR THE FUTURE, Edward A. Kachik, Materials Technology Institute. P10 ELECTRIC UTILITY VIEW OF THE USE OF CRITICAL METALS IN STEEL, Robert I. Jaffee, Electric Power Research Institute. Pll THE USE OF CHROMIUM IN STEELS FOR AEROSPACE, Rod Simenz, Lockheed-California Co. P12 USER'S VIEWS - OIL COUNTRY USAGE TRENDS IN CRITICAL MATERIALS FOR STEELS OF THE FUTURE - CHROMIUM, John W. Kochera, Shell Development Co. P13 in Impact of Developments in Manufacturing and Process Controls on Conservation and Recovery of Critical Materials STAINLESS AND SPECIALTY STEELS AOD, EBR, LR, VAR, VIM, VOD, James T. Cordy, Universal Cyclops Specialty Steels Co. P14 MINI MILLS-TECHNOLOGICAL INNOVATIONS AND FUTURE ALTERNATIVES, Peter H. Wright, Chaparral Steel Co. P15 FUTURE RAW-MATERIAL REQUIREMENTS FOR STEEL PLANT REFRACTORIES, David H. Hubble and K. K. Kappmeyer, U.S. Steel Corp. P16 RECYCLING-PRESENT AND FUTURE: "POTENTIAL FOR CONSERVATION" Herschel Cutler, Institute of Scrap Iron and Steel. P17 Conservation and Substitution For Chromium in Stainless Steels for Chemical Use and For Corrosion Resistant Applications AN OVERVIEW OF THE POTENTIAL FOR CHROMIUM CONSERVATION IN STAINLESS STEELS FOR CORROSION APPLICATIONS, Gerald L. Houze Jr., Allegheny Ludlum Steel Corporation. P18 SUMMARY OF STUDIES OF METAL PROPERTIES COUNCIL ON CHROMIUM CONSERVATION IN STAINLESS STEELS FOR CORROSION APPLICATIONS, A. 0. Schaefer, Metal Properties Council. P19 CHROMIUM CONSERVATIONS IN STAINLESS STEELS: STATUS OF MPC ACTIVITY, Jim Heger, Consultant, Metal Properties Council. P20 CHROMIUM SUBSTITUTION PLANNING AT THE AMERICAN STERILIZER COMPANY, Roy S. Klein, The American Sterilizer Co. P21 ALTERNATES FOR STAINLESS STEELS IN THE CHEMICAL PROCESS INDUSTRIES, Edward A. Kachik, Materials Technology Institute of Chemical Processing Industries. P22 OPPORTUNITIES FOR CONSERVATION OF CHROMIUM IN CHEMICAL PROCESS EQUIPMENT, Robert A. Gaugh, ARMCO Inc. P23 APPROACHES TO CHROMIUM CONSERVATION IN MATERIALS FOR CHEMICAL PROCESSING INDUSTRIES, Aziz I. Ashpahani , Cabot Corporation. P24 POTENTIODYNAMIC CORROSION BEHAVIOR FOR SEVERAL FE-MN-AL AUSTENITIC STEELS, Rosie Wang and R. A. Rapp, Ohio State University. P25 IV VOLUME II B: Proceedings of a Public Workshop; "Trends in Critical Materials Requirements for Steels of the Future; Conservation and Substitution Technology for Chromium". Conservation and Substitution For Chromium in Stainless Steels and Alloys For Heat Resistant Application OUTLOOK FOR CONSERVATION OF CHROMIUM IN SUPERALLOYS, John K. Tien, Juan M. Sanchez, and Robert N. Jarrett, Center for Strategic Materials, Columbia University. P26 DEVELOPMENT OF 9 CR-1 MO STEEL, Vinod K. Sikka, Oak Ridge National Laboratory. P27 SILICON-MOLYBDENIUM DUCTILE IRON FOR ELEVATED TEMPERATURE SERVICE TO CONSERVE CHROMIUM, Jan Janowak, Climax Molybdenum Co. P28 THE 1982 STATUS REPORT OF MN-AL-FE STEELS AS REPLACEMENT FOR STAINLESS IN HEAT RESISTING AND CRYOGENIC APPLICATIONS, Samir K. Banerji, Foote Minerals Co. P29 Conservation and Substitution For Chromium in Carburizing, Heat Treatable Steels and Bearing Steels THE DEVELOPMENT OF NEW ALLOYS TO REPLACE CHROMIUM IN CARBURIZING STEELS FOR GEARS AND SHAFTS, Carl J. Keith and V. K. Sharma, International Harvester Co. P30 CHROMIUM-FREE STEELS FOR CARBURIZING, George T. Eldis, D. E. Diesburg, and H. N. Lander Climax Molybdenum Co. P31 RARE EARTH BORON STEEL (25-Mn-Ti-B) GEARS REPORT FROM NANCHANG GEAR PLANT, BEIJING, CHINA, Presented by Dale H. Breen and Allen G. Gray. P32 POTENTIAL FOR SELECTIVE HARDENING BY INDUCTION IN CHROMIUM-FREE STEELS, Peter A. Hassell, Ajax Magnethermic Corp. P33 BEARINGsSTEELS OF THE 52100 TYPE WITH REDUCED CHROMIUM, Chester F. Jatczak, The Timken Co. P34 Conservation and Substitution For Chromium in Structural Alloy, High Strength, and High Strength Low Alloy Steels AN OVERVIEW Oh CONSERVATION AND SUBSTITUTION FOR CHROMIUM IN STRUCTURAL ALLOY, HSLA AND ULTRA HIGH STRENGTH STEELS, Robert T. Ault, Republic Steel Corporation. P35 ALTERNATIVE COMPOSITIONS FOR FUTURE HSLA STEELS, Brian L. Jones, Niobium Products Co. P36 Potential For Advanced Technologies in Chromium Conservation-Coating Systems And Surface Modification Technology; Ceramics, Composites And Intermetallics OPPORTUNITIES FOR SURFACE MODIFICATION TECHNOLOGY IN CONSERVATION OF CHROMIUM, Peter G. Moore, Naval Research Laboratory. P37 SALT BATH TREATING AS AN ALTERNATIVE FOR CHROMIUM PLATING, William G. Wood, Kolene Co. P38 CLAD METALS: MATERIAL CONSERVATION THROUGH DESIGN FOR CORROSION CONTROL AND HIGH PERFORMANCE, James T. Skelly, Texas Instruments, Inc. P39 POTENTIAL FOR POLYMER CONCRETE TO CONSERVE ALLOYS IN ENGINEERING APPLICATIONS, Jack J. Fontana, Brookhaven National Laboratory. P40 ELECTROLESS NICKEL AS A SUBSTITUTE FOR CHROMIUM PLATING IN INDUSTRIAL APPLICATIONS, Ronald N. Duncan, Elnic Inc. P41 DEVELOPMENT OF DUCTILE POLYCRYSTALLINE NIJ\L FOR HIGH TEMPERATURE APPLICATIONS, C.T. Liu and C.C. Koch, Oak J Ridge National Laboratory. P42 INJECTION MOLDING CERAMIC PARTS FOR HIGH TEMPERATURE APPLICATIONS, Beebhas C. Mutsuddy and Dinesh K. Shetty, Battel le-Columbus Laboratories. P43 Information Stockpile - Summary of Comments DEVELOPING AN INFORMATION STOCKPILE TO AID IN SUBSTITUTION PREPAREDNESS, Robert T. Nash, Vanderbilt University. P44 VI OUTLOOK FOR CONSERVATION OF CHROMIUM IN SUPERALLOYS John K. Tien, Juan M. Sanchez and Robert N. Jarrett Center for Strategic Materials Henry Krumb School of Mines Columbia University New York, New York 10027 I. INTRODUCTION AND BACKGROUND Chromium is widely recognized as an essential alloying element in superalloys. This point is illustrated in Table I, which shows chromium contents for typical superalloys. Wrought nickel-base and iron/nickel-base alloys constitute the majority of superalloys and contain roughly 15-20% Cr. Cast nickel-base alloys contain typically 10-15% Cr and cobalt-base alloys 2o-3o% Cr. Since the total tonnage of superalloy production is a mere fraction of the steel production, the amount of chromium used in superalloys does not constitute a very sizable portion of the 12 3 total end use chromium in the U.S. ' ' (see Fig. 1). However, the readily available ferrochrome used in stainless steels cannot be used as the source of chromium in the production of superalloys, since steel-grade ferrochrome usually contains too many deleterious tramp elements and impurities. Moreover, iron certainly is a major contaminant in the nickel-base and cobalt-base superalloys. Thus, the more expensive vacuum grade, P26-1 19, ,5 18. ,0 15, .0 15. .0 TABLE I. Chromium Content of Typical Superalloys Wrought Ni-Base Cr Content, % V \ Waspaloy Udimet 720 Udimet 700 Nimonic 115 Cast Ni-Base IN 738 16.0 Mar M200 9.0 IN 100 10.0 Alloy 713C 12.5 Iron-Base Inconel 718 18.5 Alloy 901 13.5 Incoloy 903 0.0 Cobalt-Base X 40 25.5 Mar M509 21.5 Mar M322 21.5 Waspaloy, Udimet, Mar M, and Nimonic, IN, Inconel and Incoloy are trademarks of United Technologies, Inc., Special Metals Corp., Martin-Marietta Inc. , and International Nickel Company, Inc. , respectively. P26-2 4 5 spectral grade, or electrolytic* chromium metal is required. ' Figure 2 shows the superalloy use of refined chromium and since over 50% of this chromium metal is refined overseas, this data clearly indicates the critical role played by chromium in superalloy production. Although superalloys use only a small fraction of the total chromium imported, they are the largest user of high quality chromium metal for which the D.S. has a limited production capacity. In order to establish the feasibility of a Cr conservation program in superalloys, we will next briefly assess the role played by chromium in the structural (or bulk) properties and in the corrosion (or surface) properties of these alloys. As a bulk alloying element, Cr is found to be the major component of f^oCg and M^C^ carbides, a minor component of MgC, and not present in MC carbides '' (Table II). In cobalt-base alloys carbides play a very important role since they are the strengthening phases. Table III shows the basic types of cobalt-base alloys. Note that the total amount of carbide formers in these alloys are about constant, but as the tungsten to chromium ratio increases, M g C carbides replace the Cr 23 C 6* Furthermore, with the addition of Ti-Ta type * Difficulties also arise in using electrolytic chromium due to salt impurities resulting from that process (4). P26-3 TABLE II. Carbides in Cobalt-Base Superalloys Alloy Cr W+Mo Ti+Cb+Ta+Zr Total Carbide Carbide Formers Type X 40 0.5 25.5 7.5 Mar M509 0.6 21.5 7.0 Mar M322 1.0 21.5 9.0 4.2 7.5 33.0 32.5 38.0 Cr 23 C 6 (W?2rLc (Ta,TiJC w*c Tic TABLE III. Carbide Types in Superalloys and Their Metal Content Carbide Metal Atoms M 7 C 3 Cr M 23 C 6 Cr + (Mo,W) M 6 C (Mo,W) + Cr MC (Hf,Ta,Nb,Ti,V) P26-4 refractory elements, the Ci^Cg carbides are replaced with MC's which do not contain Cr. In fact, refractory metals are purposely added to Co-base superalloys in order to replace chromium carbides, which in general play a detrimental role in the alloy's mechanical properties. When Cr^oCg forms, it precipitates heterogeneously at the grain boundaries and at stacking faults. This stacking fault precipitation, in turn, acts as a crack initiation site resulting in low temperature embrittlement. Chromium, however, is always present in Co-based superalloys at a concentration of about 20% Cr, in order to form a protective oxide layer. The latter is necessary since cobalt oxide is volatile, offering no protection of its own. The same carbide decomposition reaction seen in Co-base superalloys is used in the heat treatment of nickel-base superalloys. However, in the latter system, the reaction is used to produce the typical grain boundary morphology shown in Fig. 3, with both forms of carbide (MC and M^oCg) present after heat treatment. In a study* on the role of cobalt in superalloys, we found that cobalt affected the type of carbide found in nickel-base 6 7 superalloys ' (see Fig. 4). Removing cobalt from the standard * "The Role of Cobalt on Nickel-Base Superalloys", NASA COSAM (Conservation of Strategic Aerospace Materials) Grant NAG 3-57 conducted at NASA-Lewis Research Center, Special Metals Corporation, Purdue University and Columbia University. P26-5 alloy destabilizes the MC carbide and results in more M^^Cg. Despite this effect the mechanical properties were essentially unaffected (Fig. 5) , indicating again that the chromium carbides are replaceable. As part of the same investigation, it was found that the precipitation of Cr^oCg carbides resulted in a drastic reduction in the hot workability of Nimonic 115. The effect is illustrated in Fig. 6, which shows the failure during the first reduction of the chromium carbide containing alloy, while the alloy with the standard was rolled to the final 3/4" bar. The carbide morphologies are shown in Fig. 7. The spheroidized MC carbides aid workability, while the chromium carbides at the grain boundaries embrittle the alloy. Thus, it was concluded that in Ni-base superalloys the Cr^oCg carbides are not essential for mechanical properties and, indeed, sometimes such carbides are detrimental. Beside carbide effects, chromium also affects the properties of the matrix since, like cobalt, it partitions to that phase. Similarly, chromium is a relatively weak solid solution strengthener compared to the refractory elements. The chromium effect on gamma prime fraction is also expected to be small, like cobalt's effect, since the Ni-Al-Cr and the Ni-Al-Co ternary phase diagrams are quite similar. Certainly a chromium free alloy would be less prone to sigma formation since chromium is a major component of that phase. The effects of chromium on stacking fault P26-6 energy and gamma prime anti-phase boundary energy should certainly be considered, but no intrinsic mechanical problems would be expected if chromium were removed from a superalloy as was the case of 903, a chromium free version of alloy 901. The main reason for the presence of chromium in superalloys is, of course, the marked improvement in oxidation/hot corrosion resistance it imparts to the alloys. For hot corrosion resistance o an alloy should simply have as much chromium as possible (Fig. 8). This amount is limited, however, by the 20-25% chromium solubility in the matrix ' , beyond which sigma will form. Since the solubility of chromium in gamma prime is small (about 3-4%), high gamma prime volume fraction alloys are limited in their total chromium content. Thus, such alloys have intrinsically poor hot corrosion resistance. In applications, the superalloys are usually divided into two groups according to their operating temperatures: the intermediate temperature range (above 800 C) for turbine blades. At the lower temperatures fine grain, wrought alloys with up to about 40% gamma prime are used. At high temperatures, the strength requirement precludes all but the high gamma prime fraction, high strength cast alloys. Because of the large gamma prime fraction, high temperature alloys have less chromium and must be coated for adequate corrosion protection. The mechanism of corrosion protection in the complex P26-7 superalloys is not simple. In the case of iron-nickel and cobalt base alloys Cr^O., forms as the predominant oxide. If the alloy contains sufficient chromium (Fig. 9), this Cr 2 3 layer is continuous and it has a low enough permeability for oxygen to prevent oxidation of the other metallic species. Such an oxide is protective by itself. However, if an alloy does not contain enough chromium (see Fig. 10) , a continuous layer does not form and other oxides occur instead, such as volatile CoO or TiO^ or P orous N ^0 or Fe 2°3 r wn i° n are permeable to oxygen. This oxide situation allows continuous oxidation of the exposed f 10 surface. In the case of nickel-base superalloys, the thermodynamically preferred oxide is Al 2 0.j, which has the lowest permeability of any of the oxides. However, the aluminum content in these alloys is too low to form a continuous, protective Al^O-j layer. As the oxide starts to form aluminum is drawn to the surface, dissolving gamma prime and forming a denuded zone marked 'Y' in Fig. 11. Due to the depletion of aluminum in the denuded zone, other oxides begin to form as fast as oxygen diffuses through the oxide. In the nickel-base superalloys, the chromium content is high enough that CrJ^^ also forms in the initial oxidation reactions. This oxide essentially behaves like a diffusion barrier for the oxygen and the stable configuration in Fig. 12 is reached. The oxygen activity below the Cr 2 °3 layer is P26-8 sufficiently low so that Al^CK alone forms and Al diffusion across the denuded zone is the controlling step. This oxide combination is the source of the excellent oxidation resistance of the superalloys. Even though an aluminum oxide layer would provide the best oxidation protection, superalloys require chromium to form a continuous layer. The C^O, layer is irreplaceable for resistance to hot corrosion in the presence of molten salts. Without this layer (due to a low chromium content), the liquid salts and oxygen preferentially attack grain boundaries (Fig. 13) causing premature intergranular failure (Fig. 14). The rupture life of this specimen was considerably shortened by the presence of NaS0 4 (see Fig. 15). For this reason low chromium alloys are protected by coating with corrosion resistant alloys (NiCrAlY, beta, etc.) or plasma spraying (chromizing or aluminizing) . The compositional differences between the coating and substrate (such as a beta phase alloy shown in Fig. 16) form a diffusion couple that can result in a multitude of intermetallic phases at the interface. Differences in mechanical and expansion characteristics between these two alloys are unavoidable. Accordingly thermal fatigue failure at the interface are likely to occur so the substrate always contains a considerable amount of chromium as a safety net. P26-9 10 In summary, chromium can be removed from superalloys without a severe effect on mechanical properties, but it is essential at the surface for oxidation resistance especially in the presence of salt contaminants. II. Possible Chromium Conservation Options in Superalloys The current solution for resisting hot corrosion in superalloy design is the addition of as much chromium as the matrix can hold without forming sigma phase and then coating the alloy if that amount is not sufficient (see Fig. 17). Aside from the conservation consideration that over 90% of the chromium is not needed, the presence of this chromium, in fact, can deleteriously affect the alloys. As mentioned, these negative aspects include sigma and mu phase instabilities, excessive M 23 C 6 P rec ipi tat i° n ' anc3 relatively high thermal expansion coefficients. The chromium requirement of superalloys can be redefined by three criteria: 1. The near surface of the alloy (a minimum of 10 microns) must contain sufficient chromium for corrosion resistance (^20%). 2. This layer must be integrally bonded to the substrate with no continuous intermetallic layers or physical interfaces. 3. The configuration must be sufficiently stable over pro- longed periods ( " 1000 hours). We performed a simulation of the P26-10 11 diffusion process (Fig. 18) and determined that the initial 10 micron profile would be stable at 800 C, but a deeper profile of 50-100 microns would be required for 1000 hours of operation at 1000°C (Fig. 19). The available techniques for producing surface chemical changes are: 1. Claddings and coatings — Using current diffusion techno- logy, chromium-free base alloys can be coated or clad by conventional techniques. However, this approach has been rejected, to date, since a crack forming in the coating would result in disastrous substrate degradation in hot corrosion environments. Currently all base alloys that are coated contain at least 8% chromium for insurance. 2. Plasma spraying and overlays — Overlays also have a dis- tinct interface and the same failure consequences as claddings or coatings. 12 3. Dual composition powders — By using two different powder compositions for the interior and surface of the disk components, a sizable amount of the chromium could theoretically be saved. However, to our knowledge, this technique has still not been successfully used to produce a reliable crack-free component. 13 14 4. Pack chromizing ' — This technique has been success- fully used to reproduce a surface alloy on stainless steels. The steel component, for example, is usually packed in a mixture ferro-chrome or chromium metal and alumina powders with a small amount of ammonium chloride and annealed in an inert or reducing P26-11 12 environment until the required diffusion profile is achieved. The gaseous chromium chloride is the carrier exchange gas of the chromium atoms rather than the metal/metal interface as is the case for the true diffusion couple. Currently this technique has been used experimentally to apply coatings to superalloys, but most of the research is unpublished company proprietary or published in the Soviet Union. This technique might be applied to achieve a smooth concentration profile on the surface of a chromium-free alloy by adjusting chromium and nickel activities in the pack or the exchange gas. However, we do not know of any success, to date, of producing a significant surface concen- tration of chromium without forming an alpha chromium layer. 5. Laser glazing ' or electron beam surface welding — Laser glazing is a rapid solidification technique that has been used to construct disk shapes from layers of nickel- molybdenum alloys roughly 100 microns thick. Conceivably this technique could be used to apply a "layer of a chromium containing alloy to a chromium-free superalloy substrate. However, as yet, no conventional alloy has been laser glazed without forming surface cracks. 17 18 19 6. Chromium ion implantation r ' — Ion implantation is readily used as a surface alloying technique for low temperature wear or for aqueous corrosion resistance. However, the depth of the alloy layer of chromium is on the order of .1 micron, which is well below the 10 micron minimum required. Additional difficul- ties include sputtering loss and a low beam current (measured in microamps). Hence, straight ion implantation can be eliminated P26-12 13 at this time as a viable alternative. 19 7. Ion recoil implantation — As an alternative to using chromium ions as the source, recoil implantation can conceivably use a vapor or electrolytically deposited layer (of chromium) as the target for a high current (milliamps) beam of argon ions. The atom localized impact can in theory drive the chromium into the surface to a depth of approximately 1 micron. Multiple anneals and reapplication (Fig. 20) could result in a substantial profile to a depth of 10 microns or more with no sharp chromium alloy interface. Isothermal annealing of the implanted layer would most likely produce a smooth concentration profile following Fick's law. Since no diffusion barrier exists at the 10 micron depth, repeat implant- ation/anneal cycles will distribute previous implant layers over an ever-increasing range before a substantial concentration is developed at the 10 micron depth. However, we propose that the diffusion range can be restricted by the application of a steep thermal gradient. Controlled laser annealing ' is such a technique. Laser energy can be concentrated to within a few microns of the surface and thus constrain chromium diffusion in the near surface layer. With this method a high concentration near-surface layer can be built with a minimum number of implan- tation cycles. The same recoil mechanism may be used to bond a thicker chromium layer to the alloy surface. The argon ion beam disrupts P26-13 14 the interface yielding a smooth transition layer for the diffusion couple without the problem of decohesion. In this alternative scenario, laser anneal can again serve as the spatially contained heat source for driving the chromium rapidly through the interface resulting in a chromium bearing surface alloy. In summary, we assert that the chromium metal in superalloys is a critical strategic material which is quite vulnerable to sup- ply side fluctuations. We have shown that chromium is essential for the elevated temperature oxidation and hot corrosion resis- tance of these alloys. However, the high chromium content required for the surface protection is not needed for the mechanical properties of the superalloys and, indeed, is often considered more detrimental than beneficial for these properties. Accordingly, we have reviewed the various methods of achieving the srface requirement of chromium while eliminating the bulk of the chromium from the alloys. While current methods of protecting low chromium alloys have drawbacks when applied to chromium-free alloys, several alternate methods of surface alloying, including ion implantation and laser annealing, have recently shown possi- ble applications in the area of conservation of chromium in superalloys. P26-14 15 III. References 1. J.L. Morning, N.A. Matthews and E.C. Petersen, in Mineral Facts and Problems 1980 , U.S. Bureau of Mines Bulletin 671, 1980, pp. 167-182. 2. U.S. Bureau of Mines Commodity Data Summary, "Chromium," 1980-1981. 3. L.R. Curwick, W. A. Petersen and H.V. Makar, U.S. Bureau of Mines IC #8821, 1980. 4. Private communication, W.J. Boesch, Special Metals Corporation, New Hartford, New York. 5. Private communication, A. FitzGibbon, Elkem Metals, Pittsburgh. 6. R.N. Jarrett and J.K. Tien, Met. Trans. A, Vol. 13A, June 1982, pp. 1021-1032. 7. J.K. Tien and R.N. Jarrett, in High Temperature alloys for Gas Turbines 1982, eds. R. Brunetaud et al., D. Reidel Publishing Co., Dordrecht, The Netherlands, pp. 423-446. 8. The Superalloys . eds. C.T. Sims and W.C. Hagel, John Wiley and Sons, Inc., New York, 1972. 9. O.H. Kriege and J.M. Baris, Trans. ASM, Vol. 62, 1969, pp. 195-200. 10. f.s. Pettit and j.k. Tien, in Corrosion Fatigue, 1972, pp. 576- 589. 11. J.K. Tien and J.M. Davidson, Advances in Corrosion Science and Technology, Vol. 7, 1980, p. 1-51. 12. Private communication, J. Stulga, Crucible Inc. Research Center, Pittsburgh. 13. R. Sivakumar, Trans. Indian Inst, of Metals, Vol. 33, No. 5, 1980, pp. 398-403. 14. R. Sivakumar and L.L. Seigle, Met. Trans. A, Vol. 7A, 1976, p. 1073. 15. E.M. Breinan, B.H. Kear, CM. Banas and L.E. Greenwald, in Superallovs: Metallurgy and Manufacture. Claitor's Publishing Div. , Baton Rouge, 1976, pp. 435-450. 16. D.B. Snow, E.M. Breinan and B.H. Kear, in Superalloys 1980 . ASM, Metals Park, 1980, pp. 189-203. 17. A.B. Campbell III, B.D. Sartwell and P.B. Needham, Jr., U.S. bureau of Mines Report RI #8387, 1979. P26-15 16 18. C.W. Draper, Journal of Metals, June 1982, pp. 24-32. 19. Private Communication, D. Moon and W. Nahemow, Westinghouse Research and Development, Pittsburgh. 20. G. Dearnaley, Journal of Metals, Sept. 1982, pp. 18-28. 21. Private communication, S. Copley, University of Southern California, Los Angeles. P26-16 17 Other Other |[ SUPERALLOYS Fig. 1 1980 Total U.S. chromium consumption (849 million pounds) Other Nuclear^ x 11 % 6 % Fig. 2 Superalloy use of refined chromium metal P26-17 18 Fig. 3 Microstructure of a typical nickel-base superalloy with carbides at the grain boundaries 1.0 0.8 < 0.6 UJ *r* 0.4 UJ 0.2k cr _j j_ 0.0E i i i i * ' i i t i i M B C B j L J i u 5 10 15 WEIGHT % COBALT Fig. 4 Effect of cobalt on carbides in Udimet 700 P26-18 19 1500 fD Q. c 1000 0) m Hi cc H 500 t — i — i — i — i — i — i — i — i — i — i — i — i — i — i — i — r Tensile Strength 0.2% Yield Strength DISK 25°C J I I I I I I I I I I I I I I I L 5 10 15 WEIGHT PERCENT COBALT Fig. 5 Room temperature tensile properties of cobalt modified Udimet 700 P26-19 20 _!" "_.IL. Nimon\c115 0%C©bolt Fig. 6 Failure of low cobalt Nimonic 115 during rolling due to chromium carbide content Fig. 7 Carbide microstructure of standard N115 (right) and low cobalt N115 (left) . Note the nearly continuous chromium carbides in the cobalt-free alloy. P26-20 21 40 30 < E t- 20 10 U- 100 • U-700 » U-500 yRENE-41 • HASTELLOY X \ \ _ M3I3 I I 10 _L J_ 50 20 30 40 WEIGHT %Cr Idealization of the effect of chromium on the hot-corrosion resistance of nickel-base alloys in burner-rig tests. ALLOY Cr Content y' fraction CrOn y) Cr(in y') Wrought Waspaloy Udimet 700 NiMONIC 115 19.5 15.0 15.0 20. 45. 50. 23. 22. 24. 2. 3. 4. Cast Inconel 713c IN 100 MarM 200 12.5 10.0 9.0 50. 65. 55. 23. 22. 17. 3. 3. 3. Fig. 8 Phase compositions (9, 6) and the effect of chromium on hot corrosion resistance (8) . Note the chromium solubility in the the gamma and gamma prime phases of all the alloys as nearly constant. This limit is determined by the sigma phase boundary in the multicomponent phase diagram. P26-21 22 CR 2 3 ALLOY + IRON- CHROMIUM CARBIDES V*i S ■ 30M . Fig. 9 iron- chromium alloy with protective chromium oxide layer zm **& MRH9 9Jm .-■'<■■<■ CrS - .% * -10M. Fig. 10 Oxidation of a low chromium alloy P26-22 23 Fig. 11 Formation of a gamma prime denuded zone below the alumium oxide in a nickel-base superalloy Cn 2 3 A1 2 3 Metal SUPERALLOY OXIDE SCALE Fig. 12 Superalloy oxide protective scales P26-23 24 Fiq 13 Preferential grain boundary attack in the hot 9 ' corrosion of a nickel-base superalloy P26-24 25 Fig. 14 Intergranular failure of a hot corrosion/ stress rupture specimen o < Q: h- m 12 10 8 GRAIN SIZE: 300/i.m 982°C 107 MN/m 2 VAC ^ -5.16x10 /sec -a 6- = 1.12x10 /sec ^00 200 TIME (HOURS) AIR 300 Fig. 15 Corrosion induced premature failure of the above specimen P26-25 26 ZONE 1 P(NiAI) matrix + bcc a Cr, Mo ♦ substrate phases 1SWJL single phase 0, Cr, Mo, Ti, Co in solution ZONE 3, P(NIAI) matrix, TIC, M C , and o 6 o phase in coating affected substrate Fig. 16 Diffusion boundaries of a beta phase coated superalloy P26-26 27 CONVENTIONAL SOLUTIONS r+r* 15 % Cr MCrAlY r+r' 10 % Cr CONSERVATION OPTIONS Cr-rich Alloy 0% Cr MCrAlY Cr-rich Allov 0% Cr 10 urn Implanted With 15% Cr 0% Cr MCrAlY Cr Implanted Layer , 0% Cr Fig. 17 Methods of providing superalloys with adequate corrosion resistance P26-27 28 30_ _As Implanted Temp = 800 C OAR DEPTH (in urn) 8 10 c(x,t) = 2(2Tr) i 6(l+x) 3/2 f.XP[^]ERF CC^ilfli] ♦ EXP[^f]ERF fl#$& i ] WHERE: x = 2Dt 52 x-x. n = a = $ = total ion dose AND THE INITIAL PROFILE IS: c(x,o) = -£r exp[- -L°l ] Fig. 18 Diffusion profile of a surface alloy in a nickel-base alloy P26-28 29 30 DIFFUSION PROFILES Initial Depth lOum Exposure for 1000 hrs 4 8 12 IB 20 DEPTH (in urn) 30,. LU a 20 CJ a 10 DC X CJ Initial Depth 50um 1000 hrs at 1000 C J I L. 4 8 12 DEPTH (in urn) 16 20 Fig. 19 Required profiles for 800°C and lOOO°C exposure P26-29 30 30_ Initial Implantation 2 4 8 DEPTH (in urn) 10 Cr+ 30,. Anneal 2 4 6 a DEPTH (in urn) 10 i; m pp.-implant 2 4 B % a DEPTH (in urn) 10 Cr+- Repeat 2 4 B DEPTH (in urn) Cr+ Specimen Surface Fig. 20 Multiple implantation approach P26-30 By acceptance of this article, the publisher or recipient acknowledges the U.S. Government's right to retain a nonexclusive, royalty-free license in and to any copyright covering the article. DEVELOPMENT OF MODIFIED 9 Cr-1 Mo STEEL 1 Vinod K. Sikka Metals and Ceramics Division Oak Ridge National Laboratory Oak Ridge, TN 37830 The status of development and commercialization of a modified 9 Cr-1 Mo alloy is presented. The alloy is modified by the addition of 0.06 to 0.10% Nb and 0.18 to 0.25% V. The alloy is recommended for use in the normalized and tempered condition (1038°C for 1 h, air cooled to room temperature; 760°C for 1 h, air cooled to room temperature). Heat treat- ment, Charpy impact, tensile, and creep properties of the alloy are described in detail along with brief description of other properties. The modified alloy has creep strength that exceeds that of standard 9 Cr-1 Mo and 2 1/4 Cr-1 Mo steels for the tem- perature range from 427 to 704°C. The total elongation and reduction of area values for all test temperatures and rupture times up to 22,500 h exceed 15 and 70%, respectively. The estimated design allowable stresses for this alloy are higher than those for standard 9 Cr-1 Mo and 2 1/4 Cr-1 Mo steel. At 550°C and above, these values are twice those of the other alloys. Operating experience on this alloy is being obtained by installing tubes in various steam power plants. 1. Introduction A 9 Cr-1 Mo steel with properties improved over the 2 1/4 Cr-1 Mo steel and that of other ferritics in the ranges from 9 to 12% Cr and 1 to 2% Mo has been developed recently [1, 2]. The development of this alloy was funded jointly by the U.S. Department of Energy fossil energy Research sponsored by the Office of Fossil Energy and the Office of Breeder Reactor Technology Project, U.S. Department of Energy, under contract W-7405-eng-26 with Union Carbide Corporation. P27-1 and breeder reactor development programs. The purpose of this paper is to describe the status of this development, with special emphasis on the status of approval of this material in the ASTM Specifications and the A SUE Boiler and Pressure Vessel Code, 2. Chemical Specifications The composition specifications of modified 9 Cr-1 Mo steel are listed in Table 1 and compared with those of standard 9 Cr-1 Mo steel. The main difference with the modified alloy compared with the standard alloy include 1. addition of niobium and vanadium, 2. a specified range for each element, and 3. a specification for nitrogen, which is not listed for the standard 9 Cr-1 Mo. Table 1. Chemical analysis of modified 9 Cr-1 Mo steel and its comparison with standard 9 Cr-1 Mo steel Content range, wt % Element Modified 9 Cr-1 Mo (Grade 91) Standard 9 Cr-1 Mo (Grade 9) Carbon Manganese Phosphorus Sulfur Silicon Chromium Molybdenum Nickel Vanadium Niobium Nitrogen Aluminum 0.08-0.12 0.30-0.60 0.020 max 0.010 max 0.20-0.50 8.00-9.50 0.85-1.05 0. 40 max 0. 18-0. 25 0.06-0.10 0.030-0.070 0. 04 max 0. 15 max 0.30-0.60 0.030 max 0.030 max 1.00 max 8.00-10.00 0.90-1.10 P27-2 Both niobium and vanadium are added to the alloy to improve its elevated-temperature strength properties. Microstructural work has indi- cated that the improved strength of the modified alloy comes from two factors. First, fine M23C 6 precipitate particles nucleate on Nb(C,N), which comes out first during the heat treatment. Second, the vanadium enters M 2 3C 6 and retards its growth at the service temperature. The finer distribution of M 2 3C 6 adds to strength, and its retarded growth holds the strength for long periods of time at the service temperature. 3. Heat Treatment The alloy is recommended for use in the normalized and tempered condition. The normalizing treatment consists of heating the alloy to 1040°C, holding for 1 h for material up to 25 mm thick, and then air cooling to room temperature. This treatment produces a fully martensitic structure. The typical hardness in this condition is Rockwell C40. The tempering treatment consists of heating the normalized steel to 760°C, holding for 1 h up to 25-mm thickness, and then air cooling to room temperature. The typical hardness in this condition is Rockwell B95. Optical and transmission electron micrographs of tempered martensite are shown in Figs. 1 and 2. Figure 1 shows that the alloy is single phase Figure 1. Typical microstructure of modified 9 Cr-1 Mo steel after the nominal normalizing and tempering treatment (1040°C for 1 h, 760°C for 1 h). Note that the grain size is very fine. P27-3 Figure 2. Transmission electron micrograph of a specimen normalized ~at 1040°C for 1 h and tempered at 760°C for 1 h. The micrograph shows that tempered martensite consists of dislocation substructure and carbides (M 2 3C 6 and MC) on both the grain and subgrain boundaries. (free from 6-ferrite) and has a fine grain size (ASTM 8—9). The trans- mission electron micrograph shows that the tempered microstructure has high-dislocation-density subboundaries in the matrix. The subboundaries are stabilized by the precipitation of carbides on them. Carbides also precipitate at the prior austenite grain boundaries. The tempering response of the alloy can be described by the Hollomon-Jaf fe (HJ) tempering parameter. This parameter is given by HJ = T(C + log t) , (1) P27-4 where T - temperature in kelvins, t = time in hours, C = HJ constant. The optimized value of C for two commercial heats (30176 and 30394) characterized to date was determined to be 22.3. The correlation of tempering parameter with hardness, Charpy V-notch energy, 0.2% yield strength, and total elongation at room tem- perature is shown in Fig. 3. The 0.2% yield and ultimate tensile strengths are well correlated with room-temperature hardness (Fig. 4). Such correlations are useful when the material has to meet both the tensile property and hardness criteria. The grain coarsening response of the alloy was examined as a func- tion of the normalizing temperature and time (Fig. 5). Grain coarsening of the modified alloy occurred only when the normalizing temperature was increased about 100°C or more above the specified normalizing tempera- ture of 1040°C. Even in the coarsened condition, the grain size for the modified alloy remained near ASTM No. 5. 4. Commercial Melting and Fabrication Eight heats of this alloy have been melted by commercial vendors (Table 2). Many vendors have fabricated this alloy (Table 3). The overall consensus is that melting and fabrication of this alloy presents no technical problems, although costs will obviously vary depending upon the melting and fabrication process selected. 5. Mechanical Properties The main emphasis on the mechanical properties of modified alloy was limited to Charpy-impact, tensile, and creep testing. Each of these properties is described here briefly. 5. 1 Charpy-impact Properties The curves of Charpy V-notch impact energy versus test temperature for two commercial heats of modified 9 Cr-1 Mo steel are compared with a heat of standard 9 Cr-1 Mo steel in Fig. 6. All tests were conducted on P27-5 3 — O uj O K — => «. r-. ^ tf> i- ■» — « £ I I I I I I " u ooo v> O -J A <= 3 mj O *- — * e o Ui Vs. A " ■ s A/ — I- u, r si * / / P / /■ J> 41 - - 2 o 4-1 T3 fi cd ^ 43 4-) 60 c a) v< 4-1 CO -o r-l a) •i-l 5^ B^S CN • o m> >, 60 J-i CD c a) >■> ex $-i cd & o a CO CO a> C T3 M cd X <4-l o c o •H • 4-1 a; O t-i c 3 fl 4-> 14-1 CO J-i CO <1) a, CO s CO 0) 4-1 M o S u CO u 4J CO CO a a 60 O c •r-l •H 4-1 J-I CO a) 60 a C e o a> rH H i ♦ zzz) i • ro P27-6 HARDNESS (ROCKWELL C) 130 120 — 110 x 100 o Z UJ or t- UJ CM 6 (a) 90 80 — 70 60 50 6 8 10 12 14 16 18 20 22 1 1 1 MODIFIED 9Cr-1Mo 1 1 1 STEEL I I 850 O- HEAT 30176 A - HEAT 30394 NORMALIZED: 1040 °C. 1h 800 TEMPERED: 732-816 °C. 0.25-16.0 h O 750 ~ 0. 2 — °y 700 ~ »- o 2 ^ CiS^ 650 uj or S^ A 600 Q UJ &xf A O 550 v CM £&A?° — 500 O A Jgv /§> 450 "7 I l I I I i 400 86 88 90 92 94 96 HARDNESS (ROCKWELL B) 98 100 HARDNESS (ROCKWELL C) 130 6 8 10 12 14 16 18 20 22 1 1 I I I I I 1 1 A_ 120 r?^y^ 110 A ^X" I »- O z UJ o: K 1/) 100 A .^^ v "a ^5|T ° — UJ -J V) 2 UJ t- 90 80 S < ? u — UJ < 2 »- 70 MODIFIED 9Cr-1Mo STEEL 3 O- HEAT 30176 A- HEAT 30394 — 60 NORMALIZED: 1040 *C . 1h TEMPERED: 732-816 °C. 0.25- 1 1 1 1 -16.0h _ I 850 — 800 750 | — 700 — 650 600 =! in z UJ 550 H 500 2 — 450 400 (b) 86 88 90 92 94 96 HARDNESS (ROCKWELL 8) 350 98 100 Figure 4. Room-temperature strength versus hardness for two commercial heats of modified 9 Cr-1 Mo steel. (a) Yield. (b) Ultimate, P27-7 t>u 1 1 1 1 1 1 1 1 (18501 J 1 (1900) 1 (1950) 1 1 (20001 i i (20501 1010 1038 1065 1093 1121 5b i i i i 1 TEMPERATURE "CCFI FOR 8h MOLD 50 6 _ 15 NORMALIZING TEMPERATURE XI'FI / 6 E i. o- •927(17001 V- 1038 (19001 6 i 40 f— A- 954 (17501 d- 1065 (1950) - £ 35 L_ a- 982 (18001 O-'093(2O0O) o- ■1010(1850) 6-1121 (2050) UJ •- 7! ' JO | i» — 1 < a 3 25 Ul 03 a- _ 3 o _ 15 j- 1 A^---^* d v O — (0 rfO — (1850) (19001 (1950) (20001 (2050) 1010 1038 1065 1093 1121 5 1 1 1 1 i I I i 1 1 1 1 1 TEMPERATURE *C CF) FOR 1h i I I I I I HOLD 1 - 5.5 - 60 - 65 tti 70 80 8.5 9.0 95 100 266 278 274 278 282 286 290 29.4 298 3Q2 306 310 314 318 32.2 32.6 (»10 5 ) HJ i TI22.3* log I ) Figure 5. Grain coarsening behavior of modified 9 Cr-1 Mo steel. Table 2. List of commercial heats melted on modified 9 Cr-1 Mo steel Heat Melter Heat size (tons) Melting practice* 2 F5349 Quaker 4 30383 CarTech 15 3039 h h CarTech 15 30176 e CarTech 15 30182 c CarTech 15 10148^ Electralloy 15 XA3602 Combustion Engineering 0.5 91887 CarTech 2.5 YYC982C Sumitomo, Japan 2.0 59020 NKK, Japan 5.0 AOD AOD AOD/ESR AOD/ESR AOD/ESR AOD and AOD/ESR Air induction Electric/ESR Vacuum induction Vacuum induction a A0D = argon-oxygen decarburization. ESR = electroslag remelting. ^Heat 30394 is half of heat 30383 and was electroslag remelted. ^Heats 30176 and 30182 are the same heat. Their numbers are dif- ferent because CarTech identifies ingots separately. "Heat 10148 was melted by AOD process. Half this heat was sub- sequently electroslag remelted. P27-8 o u CO I Cfl CO OJ iH XI CO H u o 4-1 CO CJ T-l U X cfl fa T3 O X 4-1 a) a e o cfl O •H CO fa o 3 "3 O fa 05 CO ID c CJ •H H P O u 3 -3 O fa 2 cfl 0) 33 5h .3 X X fa CJ fa o c o C fa M o CD o 01 OJ -< 4-1 X 01 o 33 CL) O T3 u HI fa 0) 4-» T3 0) « iH •3 CJ O 33 UO m in ^H CM CM o in CO vX> in v^ CM -* CM CM v£> co O r^. CM -H CM CM vO CO O l-» CM — i cfl cfl OJ 4J OJ 01 4-1 CD 0) u X CO M X J-l cfl M X X cfl 3 .H cfl 3 3 iH cfl 3 3 CO H fa m H H fa 03 H H cfl cfl fa fa T3- CD 3 >-i 4J X CD O 33 CD X X CD •i-l r-\ c X CD 4-1 O CD J-3 « fa CD O m 4-1 o 33 o ■3 CD U CD 00 O 33 in CM m -a- CM CD u u fa cfl CO •H CQ « fa 00 CD 05 OS oi 05 os ai ai 04 cm C cj cfl Cfl Cfl Cfl Cfl Cfl Cfl Cfl Cfl •H T-( w W w w w W Id w w 4-> 4-1 I 1 1 1 1 1 1 1 1 .H CJ Q Q Q Q Q Q Q a Q Q O o o Q a O a CD cfl O O O o O q O o O O o o o o o o o X i-> < >< < < < «* < . >-, >% f^ o o O O 1-1 iH iH iH iH iH .H rH X J= X J3 J= J3 rf3 X. X x: X X cfl CO cfl CO i-i o o CJ CJ CJ CJ CJ o CJ CJ CJ o U u M l-i (D CD CD CD CD CD CD CD CD CD CD CD CD 4J 4-1 4-1 4_) ^! H H H H H H H H H H H H CJ o o CJ cfl U M M )-i Vj l-i m Vj Vj l-i ^ l-i CD CD CD CD 3 CO cfl cfl cfl cfl CO cfl cfl cfl cfl cfl cfl ■H ■H iH <-4 C U U c_> U c_> U o U CJ cj CJ O w Cxi W w ON CO CO CO o oo 00 00 00 r^ -3- -* -3- -4t CO CO co CO CO CO i-H r— 1 •—1 < — i 1— t 1 — 1 ^H 1— 1 m o o o o o o o o o o o o o o o o fa co co co co co CO co CO CO co CO CO f— 1 1— 1 1—1 1— * P27-9 co (4 CO o CU 4-1 iH CO rH CO r-t co r-t CO rH 3 co a, co a CO o. co a. e o o CJ CO o CO o CO o CO O •H e •H Vj rH U rH r4 rH VJ rH CO •~\ o o M a> o eu a CU CJ CU CJ 4J ^ 4J 4-1 .£> > >s > >o > So > 5o C/> 33 •r-t •H cfl •H cj •H CJ •H o •H CJ N-^ B a Ui M B fe c 3 c s rH W 3 3 2 3 33 33 33 H CJ "3 CO 3 3 cfl CO cfl S5 T3 52 3 S 3 S3 3 3 3 00 3 iH 4-1 rH ■3 3 14 p CU P Vj OJ O cfl •3 T3 rH •3 •3 e x: rH u TJ 4J •3 "3 O *3 "3 3 O o o o o CU o O O o o 33 33 33 33 33 CJ 33 33 S3 53 33 II S CO > Q 09 a) I/O CO M n c -3- o >a- -3- 60 /~N CU J«! • • • • 3 T3 N O o> 00 co vD UO CO V43- •H OJ tH •iH T— ( CM r— 1 4-1 3 CO r3 rH G H CU 4J e 4-1 CJ 3 OJ 14 -a o o CO -H o CM - UO r^ UO 3 rH co o • CO 4-1 CJ s o U 4J CJ oj 3 CU CU 00 X TJ CU OJ "3 4-1 4J cO O c cu CU OJ cu 4J CU cu H iH O CO CO 4J £> 3 £> •O J3 J3 Cfl ^o J3 CU rO U ^H r-t CJ o 3 3 3 3 rH 3 3 « On Cu a, o oi H H H H Oh H H II H 00 H o •H 4J 4-1 1 1 1 1 1 •H 4J r-t o Q o Q Q Q Ul 4J S S X S a 3 a> co O o O O O •H M H M M N S k <: < > > > > •r-t a CO a o rH CJ >s CO o r-l CJ >s CJ CO Q. O rH CJ >. CJ CO a o CJ >s cj CO a. o rH CJ :*, o 00 U 3 rO U 3 o OJ •3 3 3 m -^ -^. 1 — ^■^ •^^ 3 60 0) >, M !*.rH >s rH >.rH >% rH •H >> 4J O cfl O CO O CO O CO o CO c u X rH rH CO iH CO rH CO r-t CO r-\ CO o CU o u > Vj > u > U > CO •H o o o 4J iH 4-1 tH 4J •H 4J iH 4-1 •H 3 00 4J 4J 60 CJ C CJ 3 CJ C O C CJ c J3 3 •H iH U CD 33 OJ 3D CU 33 CU 33 CU 33 6 u a B **4 t^ "^ 3 iH ■H iH rH rH o 3 3 3 V! ^ W W W W W o CM CO u CM co o Csl 25 5S II Q 4J 00 oo 00 00 oo O oo oo o O o CO -* sr -tf -* -d- vD CTi o> CM CM CM CU ^H i-H r-t rH rH CO CJ CJ O O O X o o o o o 3 >H >4 o> ON OS r- >4 U0 U0 UO P27-10 250 200 -a 150 o cr 100 50 0.2% Si ESR -100 0.5% Si ESR Figure 6, 0.5% Si AOD 0.6% Si STANDARD 9 Cr-1 Mo WR ORIENTATION 200 160 120 -o > o a<> d - A A 1 i i i 1 1 1 1 1 1 1 1 1 1 1 1 1 1 1 1 1 1 1 1 UJ or < 2 O o Q UJ or uu 90 - i 1 ' ' 1 1 | 1 ' ' > | 1 1 1 1 | I I 1 1 80 O A 0A° D 70 — 60 - HEATS F5349 30182 - 50 40 TEST TEMPERATURES CO 0-482 0-649 A- 538 6-677 D-593 V-704 i i i i 1 i i 1 1 30176 30383 30394 10148 91887 XA3602 1 1 1 1 1 1 1 1 1 1 1 1 i i i 10^ 10' 10' io- 10* TIME TO RUPTURE (h) io- Figure 11. Plots of (top) total elongation and (bottom) reduction of area as functions of rupture time for commercial heats of modified 9 Cr-1 Mo steel tested in the normalized and tempered condition. Data are for test temperatures in the range 482 to 704°C and for eight commercial heats. P27-19 «50 140 (30 120 | MO too 90 80 70 O 60 50 40 — 30 20 — «0 — ■ 2 1/4 Cf - 1 Mo (NORMALIZED AND TEMPERED) MODIFIED 9Cr-1Mo (NORMALIZED AND TEMPERED) 2 «/4 O - 1 Mo (ANNEALED) BRITISH COMMERCIAL 9Cr-tMo (NORMALIZED ANO TEMPERED)- 350 400 450 500 550 600 650 TEMPERATURE (*C ) 700 750 800 Figure 12, Estimated design allowable stresses as a function of tem- perature for modified 9 Cr-1 Mo steel. Design allowable stress values for standard 9 Cr-1 Mo and 2 1/4 Cr-1 Mo steel are also included for comparison. Results of this study have shown that the thermal expansion coefficient of modified 9 Cr-1 Mo steel is the same as that observed in this study and reported in the literature for the standard 9 Cr-1 Mo steel [4]. However, the thermal conductivity was found to be related to the silicon content. Higher silicon content tended to be associated with lower thermal conductivity. We have tested modified 9 Cr-1 Mo heats containing 0.2 and 0.5% Si and recommend that the average values for these two heats be used as typical values. P27-20 6.2 Fatigue and Creep-Fatigue Tests These tests are in progress at Oak Ridge National Laboratory, Sandia National Laboratory, Northwestern University, and the University of Connecticut. Data available have shown that the modified alloy has about the same fatigue strength (total strain range versus number of cycles to failure) as type 316 stainless steel in the range from 525 to 59 3°C. The alloy also has superior fatigue life beyond 105 cycles to 2 1/4 Cr-1 Mo, standard 9 Cr-1 Mo, and type 304 stainless steel. Microstructural observations on fatigue-tested specimens are reported by Jones [5] . 6.3 Fatigue Crack Growth and Fracture Toughness These tests are currently in progress at the Hanford Engineering Development Laboratory. Results available showed that the crack growth rate of the modified 9 Cr-1 Mo steel is about the same as that observed for 2 1/4 Cr-1 Mo and 12 Cr-1 Mo steel in the range from room tempera- ture to 538°C [6]. 6.4 Steam and Air Oxidation The behavior of modified 9 Cr-1 Mo steel and other alloys (2 1/4 Cr-1 Mo, 9 Cr-2 Mo, 12 Cr-1 Mo, and type 304 stainless steel) in superheated steam at 482 and 538°C are available for a period of 28,339 h. Results of this study have shown that silicon is very potent in reducing the oxidation rate in steam at 538°C of chromium-molybdenum steels [7]. Because of its lower silicon content, the modified alloy showed slightly higher weight gain in steam at 538°C than that observed for standard 9 Cr-1 Mo alloy. Air oxidation data at 593°C were obtained on both the modified 9 Cr-1 Mo and 2 1/4 Cr-1 Mo steel for a period of 20,000 h. These data showed that the weight gain for 2 1/4 Cr-1 Mo steel was about eight times that observed for the modified 9 Cr-1 Mo alloy. Tests are now continuing to measure the weight gain for longer periods. P27-21 6.5 Welding The weldability of modified 9 Cr-1 Mo steel is being investigated with the Varestraint hot cracking test, Battelle under- bead cracking test, Tekken Y-groove test, hydrogen sensitivity test, and stress-relief cracking test. Results available thus far have shown that the material is free from hot cracking susceptibility, a preheat temperature of 200°C can prevent hydrogen sensitivity, and there are no stress-relief cracking problems for a postweld heat-treatment temperature of 7 32°C. The filler wire composition for gas tungsten arc welding, electrod com- position for shielded metal arc welding, and the combination of filler wire and flux for submerged arc welding are currently being optimized to satisfy the strength and ductility criteria set for the alloy. 7. Operating Experience on Modified 9 Cr-1 Mo Steel Tubes Industrial operating experience, which aids earlier approval of the data package for the ASME Code, is being obtained by installing tubes of this alloy in various conventional power plants. The status of instal- lation of modified 9 Cr-1 Mo tubes in various utility power plants is summarized in Table 5. The range of utilities involved is international: United States, United Kingdom, and Canada. In most cases the tubes being replaced are stainless steel. The longest operating time has been reached for tubes installed at the Kingston Steam Plant. The tubes installed in Kingston plant are shown in Fig. 13. The modified 9 Cr-1 Mo steel tubes were supplied with type 347 stainless steel safe ends. The safe ends were welded on with Inconel 82 (ERNiCr3) filler wire. The final welds between type 347 and type 321 (existing tubes) were made on location. Tubes in the Agecroft power station went into operation in April 1982 and those in Lambton and Nanticoke are expected to go into operation in October 1982. This operating experience will be very use- ful for obtaining the approval of this alloy in the ASME Code. 8. Status of Commercialization The use of a modified or new alloy in actual application requires that it be included in ASTM specifications and various sections of the P27-22 60 60 60 60 e 3 c C CO 1-1 •r-( 1-1 i-l •3 •3 3 4-1 4-1 4-) 4-1 01 a> 4-) CO (0 cfl CO 3 3 3 I-l I-l I-l U 3 3 4-) 0> 0) CU 01 cfl 3 CO a o o P. o a o ■H 0- ■H O-i XI CM CM cu 1—1 00 00 i-i oo as ON i-l ON i—i i—l 3 ~H 1— 1 CN 4J 00 00 u I-l •H CO o ON >-. on cu cu s 00 i— 1 I-l •—I X X 1-4 ON cfl B s f— 1 i-H 3 i-i 0) cu CO cu •H l-i •H 4-1 4-1 0) 4-1 >n »-l X) I-l Ou a XI ct) CO a 01 a 0> 0) 3 n S <£ [K <: CO CO 4-) H u CD 0) CO 0) X H-i 0> oo o CM vO ON ON ~H j-J s o x •— i ~H co 3 3 Z 4J o S O o o —H CO 60 X 2 S 1 u 4-1 c c •H "3 dl CU X O i—i o o O o> a CO r-4 dl D. co CO CO >* CO ON -* cu •H V4 CD 3 XI 01 01 01 01 — , 0> *^^. 3 I-i o. P. a. i—i CXT3 ■—1 H >n >N >> >N 4J H H H CM H CO CM •4-1 o •H ex T3 O S S CD 0) 60 VJ 3 3 •H 4-1 o CM <4-i 4-1 4J CO /-N co co oo ^> 00 8 oo O CO 3 u u as on CO I CO CO M 0) o m in m o m m m 60 60 •H at a^-' as a s O 0) 4-1 m 4J a; CO S-l cu o 4-1 y-i c o i-l Vj (J o> 4-1 i-i u cu 4-1 C4-I T3 4-1 0) cfl cu 3 o c CO 4-1 >, 0) 4-» >-, cu rt o CO u X i-i cfl I-l i-i U X CO o 0) cfl (-1 0) 01 cu cu 3 I-i 3 4J • tH X •3 0> 4-1 X 4-1 4-1 ■3 0) w (j 3 a cfl I-l 3 3 3 Ou CO 0) 0) O 3 cu cu CU 0) O 3 • XI a. O CO X a. X X CJ CO 4-1 13 3 3 o> 01 3 0) 0) 3 CO H en co <* CO BO on CO 4J c m cu s-l S CO 4-1 J«i I-l a) CO 5-i 0) •H 0> 3 o 3 4-1 c 01 o CO H O 4-1 c CO » I-l c_> I-l a, c H 3 CO e A CO •H CM 4-1 o & ■H o 4-> CO Cfl M-l 1-1 3 o • Cu 4-1 3 u 4J i-t 4-1 O 4-) O o CO cfl 0) •H C_> i-l I-l cfl 4-1 1-1 60 tH s c c o 4-4 X 4-1 (D c P- s 3 • 3 0) CO s 3 i— i •H <0 4-1 60 CO 3 .£> « H CO < ,-J Z cO H >n 0) i-H O •H U 4-1 CJ 3 O o •H I-l 4-1 •a i-i cfl O O o >~. CO OJ CO CJ 60 I-i I-l 4-1 > >% i-H •H cu 3 •3 •3 i-l 4-) w ■3 iH •H >N *-\ >, ^N i-l 0) •i-l W w 4-1 S 3 PC 3 •H 01 u c Cfl /"> •3 -3 4J CO o CO (-1 4J i-H V-i • o 3 O 3 I=> CO X o CU 1-1 cfl CU ^ 1-1 3 •w 3 o> 4-1 •H 3 O I-l 3 • I-l 3 l-i 3 c 3 u o Vi 4-J Ph 4-1 3 o ^^ 4-1 ^-^ 4J ^-^ 01 1 01 0) 3 3 H o u O O P27-23 Modified 9 Cr-1 Mo tubes Figure 13. The modified 9 Cr-1 Mo steel tubes before going into operation in the TVA Kingston Steam power plant. P27-24 ASME Boiler and Pressure Vessel Code. The ASTM specifications are in fact prerequisites for the consideration of this alloy by the ASME Code Committees. The schedule for the ASTM and ASME Code approval is as follows: An application for an ASTM specification for the plate and tube product was submitted for approval in May 1981. The specification application for the forgings, piping, and fittings was submitted in May 1982. The tube subcommittee has approved the specifications, which now must be approved by the main committee. The plate specifications are near approval, but the other specifications are still under con- sideration by the appropriate subcommittees. The data package for the inclusion of modified 9 Cr-1 Mo steel in Sect. I and VIII of the ASME Boiler and Pressure Vessel Code was sub- mitted in June 1982. At the request of various users and producers, the appropriate subcommittees of ASME are expected to start reviewing the data package during September 1982. 8 . Summary The status of development and commercialization of a modified 9 Cr-1 Mo alloy is presented. The alloy is modified by the additions of niobium (0.06-0.10%) and vanadium (0.18—0.25%). The alloy is recommended for use in the noramlized and tempered condition (1038°C for 1 h, air cooled to room temperature; 760°C for 1 h, air cooled to room temperature). Heat-treatment, Charpy impact, tensile, and creep properties of the alloy are discussed in detail and other properties are described briefly. Some of the key facts about modified 9 Cr-1 Mo alloy include the following. 1. The ductile-to-brittle transition temperature (68-J) of modified 9 Cr-1 Mo is lower and the upper-shelf energy higher than those of standard 9 Cr-1 Mo alloy. These properties were much better in the electroslag-remelted material. 2. The creep rupture strength of the modified alloy is higher than that of the standard 9 Cr-1 Mo and 2 1/4 Cr-1 Mo steels for the entire creep temperature range. The improvement in lO 5 -!* creep-rupture strength P27-25 is very significant at temperatures above 500°C. Higher strength is also accompanied by total elongation and reduction of area values higher than 15 and 70%, respectively. 3. Higher creep strength of the modified 9 Cr-1 Mo steel provides design allowable stresses that exceed those of 2 1/4 Cr-1 Mo and standard 9 Cr-1 Mo steel for the entire temperature range. At tempera- tures of at least 500°C, the design allowable stresses for the modified alloy are twice those observed for 2 1/4 Cr-1 Mo and standard 9 Cr-1 Mo steel. 4. The specifications for various products of this alloy have been submitted to ASTM for approval. A data package has also been submitted to the ASME Boiler and Pressure Vessel Code for approval of this material in Sects. I and VIII. 5. Operating experience on this alloy is being obtained by installing its tubes in various steam power plants. The author thanks the following for their contributions: R. E. McDonald (ORNL) and G. C. Bodine (CE, Chattanooga) for contribu- tions to melting, fabrication, and heat treatment and helping to prepare tubes for installation in various power plants; W. J. Stelzman (ORNL) for Charpy V-notch studies; R. H. Baldwin (ORNL) for tensile and creep testing; M. C. Cowgill (Westinghouse-Advanced Reactors Division) for creep testing; J. F. King (ORNL) and C. D. Lundin (University of Tennessee) for welding studies; R. K. Williams (ORNL) for physical properties; and M. K. Booker (ORNL) for the data analysis. Our special thanks are also due to the research staff of Climax Molybdenum Company of Michigan for its contributions to the success of this program. Continued encouragement and support of E. E. Hoffman (DOE, Oak Ridge Operations) and P. Patriarca (ORNL) to the success of this program also are greatly appreciated. P27-26 9. References [1] Sikka, V. K. , Ward, C. T., and Thomas, K. C. "Modified 9 Cr-1 Mo Steel — An Improved Alloy for Steam Generator Application," presented at the ASM International Conference on Production, Fabrication, Properties and Application of Ferritic Steels for High-Temperature Applications, Warren, Pa., Oct. 6—8, 1981, to be published. [2] Bodine, G. C., and McDonald, R. E. "Laboratory and Pilot Commercial Process/Product Development of Modified 9 Cr-1 Mo Ferritic Alloy," presented at the ASM International Conference on Production, Fabrication, Properties, and Application of Ferritic Steels for High-Temperature Applications, Warren, Pa., Oct. 6—8, 1981, to be published. [3] Orr, J., et al. "The Physical Metallurgy of Chromium-Molybdenum Steels for Fast Reactor Boilers," pp. 91—109 in Int. Conf. Ferritic Steele for Fast Reactor Steam Generators s 30 May— 2 June 1977 , British Nuclear Energy Society, London. [4] Williams, R. K. , et al. "The Physical Properties of 9 Cr-1 Mo Steel from 300 to 1000 K," paper presented at the 17th Int. Thermal Conductivity Conference, Gaithersburg, Md., June 15-19, 1981, to be published in the proceedings. [5] Jones, W. B. "Effects of Mechanical Cycling on the Substructure of Modified 9 Cr-1 Mo Ferritic Steel," presented at the ASM International Conference on Production, Fabrication, Properties and Application of Ferritic Steels for High Temperature Applications, Warren, Pa., Oct. 6—8, 1981, to be published. [6] James, L. A. Fatigue-Crack Growth Behavior in Ferritic Alloys for Potential GCFR Structural Applications, HEDL-TME 80-71, Hanford Engineering Development Laboratory, Richland, Wash., December 1980. [7] Griess, J. C, and Maxwell, W. A. The Long-Term Oxidation of Selected Alloys in Superheated Steam at 482 and 538° C, ORNL-5771, Oak Ridge National Laboratory, Oak Ridge, Tenn. , July 1981. P27-27 Si Mo Ductile Iron for Elevated Temperature Service to Conserve Chromium Jay Janowak" INTRODUCTION High silicon molybdenum ductile cast iron is a relatively low cost ferritic heat resistant material. In the trade it is called high silicon moly iron or just plain SiMo. It is already in direct competition with high chromium steels in some applications on a purely economic basis. It is a popular material for turbo charger housings and exhaust manifolds. It has been found to be the superior material for rabble teeth in a moly-sulfide ore roaster. And finally, SiMo is a candidate for furnace grates. All of these either are, or could become, high chrome steel applications. Should there be a "critical" need to reduce our use of chromium or to decrease our dependency on chromium, we believe the high silicon moly ductile cast irons could be considered as an option to high chromium steel. It is probable, but somewhat speculative, that relatively minor design changes could allow SiMo ductile iron to function in place of high chromium steels in numerous situations. In the next few minutes four applications will be reviewed to provide a feeling for the characteristic properties of this material. SiMo ductile is basically a standard ferritic nodular (ASTM 60-40-18) to which 1.5°o silicon and \% molybdenum are added to create the h silicon 1 moly heat resistant ductile cast iron. Jay Janowak is Manager of Foundry Development, Climax Molybdenum Company P. 0. Box 397, Arlington Heights, Illinois 60006. This is the oral text of a workshop presentation "Trends in Critical Materials Require- ments for Steels of the Future Conservation and Substitution Technol- oqy for Chromium" at Vanderbilt University, October 5, 1982 P28-1 -2- TURBOCHARGERS The Si-Mo nodular iron concept was developed at the Climax research laboratory in the sixties. Among early anticipated applications were relatively low cost tubing for use in hydrogen gas generators and forming dies for titanium. These did not materialize, but Climax engineers brought the concept to the attention of a prominent turbocharger manufacturer, and, after an extensive development program, k% Si - 0.6% Mo ductile iron emerged as an improved housing material. Figure 1 shows an automotive turbocharger. To give an idea of size, a small car unit would fit inside a 1 gallon milk container. A cut away view is shown in Figure 2. Turbochargers are like miniature air compressors. Hot exhaust gases spin the turbine which through a common shaft drives the blower. On demand, fresh compressed air can be forced into the combustion chamber of the engine to provide added power in trucks and off-road vehicles. Using a different approach, this concept is applied to improve auto fuel economy through utilization of smaller, lighter weight engines. Exhaust gases are typically in the 1300 to 15oo°F range. Housings can become red hot as shown in Figure 3. Dimensional stability , oxidation resistance, and high temperature strength and toughness are important properties provided by SiMo ductile iron in this application. Let me emphasize the critical criteria. Containment - Housings must be able to contain a "wheel burst" for obvious safety considerations. On vehicles, heavy walls are prohibitive because of weight and space restrictions. Therefore, the relatively P28-2 weaker gray irons are not practical. Consequently, the alternatives are narrowed to ferritic or austenitic ductile irons or stainless steel. However component complexity favors the castability of a ductile iron over stainless steel options. While austenitic irons are stronger at temperature, the ferritic irons have better ductility. Since containment is related to a combination of strength and ductility, either iron will serve the purpose. Figure 4 shows strength and toughness at 1300°F of a h% silicon ductile iron with increasing molybdenum contents. Notice that molybdenum substantially increases strength at 1300°F and also note the elongation of over 50% at a level of 1% molybdenum. Ductile behavior is evidenced in Figure 5 which shows a housing that contained a wheel burst at temp- erature during a scheduled "run-away" test on a dynamometer. D i mens i onaj S t a b i \ i t y - Dimensional stability is essential in general, but, specifically, the turbine blade/housing "air-gap" must be maintained at a minimum level. Divider wall stability is also very important. Thermal expansion, phase changes, oxidation growth and distortion can influence the "air-gap" and other contours visible in the cutaway view (Figure 2). Silicon and molybdenum both work in the direction of improv- ing the dimensional stability of nodular irons. Ferritic ductile iron has a very low coefficient of thermal expansion which is d imens ional 1 y beneficial compared to an austenitic material. The addition of silicon further improves dimensional stability by raising the temperature of the phase change to austenite as shown in Figure 6. Note that, at k% silicon, ferrite is stable up to over 1500°F. P28-3 -it- Molybdenum provides dimensional stability by increasing the strength at temperature and by improving creep resistance. In addition, molybdenum adds resistance to stress relaxation, thereby reducing warping and distortion. Figure 7 shows constrained thermal fatigue specimens. Both are high silicon ferritic iron. One sample accumulated a bulge - it had no molybdenum. The other sample contained .6% molyb- denum. It showed less bulge and survived 3 times the number of cycles to failure. This benefit of molybdenum applies to both ferritic and austenitic irons. Oxidat ion - Ductile iron is inherently more oxidation resistant than gray iron because the graphite is in discreet spheres as will be seen later. Ihis is a mechanical influence related to the transmission of oxygen. Chemically, silicon and aluminum can add greatly to oxidation resistance in cast irons. The silicon effect is shown in Figure 8. Here, weight gain is plotted versus time at 1200°F for various silicon levels. Notice the initial weight gain in all cases. This is explained by the mechanism of oxidation resistance apparent in Figure S. Silicon preferentially oxidizes at the surface to provide a silicon oxide layer which prevents further penetration of oxygen. Notice also the discreet spheres of graphite characteristic of ductile iron. This is a different graphite shape than that of gray cast iron which has interconnected graphite flakes that create a virtual "pipe line" for oxygen. This silicon scale is particularly tenacious. Without adequate silicon, this scale would exfoliate and rapidly abrade the turbine blades. Silicon provides this benefit to both ferritic and austenitic irons. By limit- ing oxidation and thereby growth, high silicon irons have additional dimensional stability. P28-4 -5- To bring this section to an end let me point out that turbocharger housing materials can be classified according to service temperature. Gray irons can serve up to 1000°F. Unalloyed ferritic nodular irons can serve up to 1300°F, and with additions of silicon and molybdenum up to 1500°F. Austenitic irons with silicon and molybdenum may serve up to 1700°F while temperatures beyond that point may require high chromium steels or superalloys. High chromium stainless steels are alternatives to the irons at all these nominal temperature levels, particularly when special thermal or environmental situations challenge the cast irons. In such cases, silicon and molybdenum additions to ductile iron make it a suitable alternative to high chromium steels. EXHAUST MANIFOLDS Exhaust manifolds are another hot engine item (Figure 10). Traditionally, automotive exhaust manifolds were gray iron. However, higher operating temperatures, demand for weight reduction, and the desirable aspect of low thermal conductivity (for faster catalytic converter light off) has resulted in a change to predominantly unalloyed ferritic nodular iron. Most recently, a substantial interest has been shown in fabricated stainless steel exhaust manifolds and there are indications that some of the new cars may have these as standard. The principle advantage of fabricated stainless steel is weight reduction - since the remaining properties should be adequately provided by the SiMo ductile irons. In fact, it is believed that SiMo ferritic ductile irons are superior to the fabricated stainless steel in design flexibility, noise damping and distortion resistance. Thermal distortion can make it impossible to replace an exhaust manifold once removed from the car. P28-5 -6- In extreme cases, during operation, thermal distortion can lead to exhaust gas leaks which can be a noise problem and a potential fire hazard. Building a further case for the iron, it is important to recognize that it is now possible, with new casting technology to cast ductile iron exhaust manifolds having sections as thin as one-sixteenth of an inch. The jury is still out on this application but SiMo ductile irons are clearly in competition with stainless steel having more than M% chromium. In a specific case involving very large diesel engines with pistons over 15" in diameter, SiMo ductile iron replaced tubular wrought 2iCr-lMo steel exhaust manifolds. A design change was desired by the manufacturer to reduce cost and improve performance. The new design provided a common tuned exhaust system instead of multiple individual exhaust tubes. Initially, unalloyed ferritic nodular iron was considered and had the necessary or improved corrosion resistance to condensate. However, concern over oxidation and creep led to the selection of SiMo ductile iron which provided the necessary confidence level for the new design. In this case, SiMo ductile iron replaced 2iCr-1Mo steel for a cost and performance benefit. GRATES Furnace grates (Figure 11) are another application where SiMo ductile iron is being considered. An industrial incinerator operator was looking for a lower cost alternative to HC steel. HC steel is 26-30%Cr, k%U'\ (max), .5%C(max), 2%Si(max), and UMn(max). An enter- prising foundry looking for new business decided to pursue the SiMo iron with our guidance. A review of the application revealed certain P28-6 -7- needs. Metal to metal wear resistance was needed to protect against a sliding action inherent in the design. Resistance to spike heating up to 1500°F was desired, but normal grate operating temperatures were said to be on the order of ^00°F because the grates were effectively insulated by the burning rubbish. In addition, the grates were cooled by cold air forced up through the grates from below to provide oxygen for burning. Toughness was important because motor blocks sometimes were dropped from a height of 20 feet onto the grates. There was no feeling for the corrosion resistance needed. A comparison of available properties was made and shown in Figure 12. Based on this information a decision was made "to go" and trial castings were put into a controlled pilot experiment side by side with HC steel. After 6 months the incinerator was shut down and inspected. SiMo ductile grates showed less wear than HC steel. Then, sometime after the 6 month check, the grates were found to have signi- ficantly deteriorated (Figures 13 and 14). A preliminary investigation revealed that molten aluminum had apparently fluxed the iron causing a reduction in the melting point. The iron thus appeared melted in an environment where temperatures only rarely reached 1500°F. The investigation has begun only recently, and as of this moment, the relative resistance of SiMo iron and HC steel to aluminum fluxing is not known. In refuse systems where metal separation occurs after burning, relative resistance to such fluxing action may end up being a critical material selection criteria. P28-7 -8- Specific tests are planned to determine the relative resistance of SiMo ductile and HC steel to aluminum fluxing. It is however clear from the results to date that, except for this fluxing by aluminum, the SiMo grates performed equally well or better than those of HC steel. ROASTER TEETH - The last example is a case where SiMo ductile iron has been found the superior material in an ore roasting furnace. Five years ago, engineers at our moly sulfide roasting plants began a program to improve the life of the roaster rabble teeth. A number of rabble teeth standing upside down and waiting to be put into service are shown in Figure 15- They hang from a beam that rotates inside a roaster as shown in Figure 16. This view shows the roaster in the empty and cold condition. The main shaft is also visible, The tooth closest to the shaft is in location number one. In Figure 17 is a view showing the red hot rabble teeth doing their job which is to roll over the molybdenum sulfide in an oxidizing atmosphere at up to 1500°F. The rabble teeth are basically plows which must stand up to an abrasive and especially hostile environment of hot molybdenum oxide, oxygen, and high sulfur gases. At times, during operation, it is necessary to maintain a free flow condition. In Figure 18 is an operator poking to free teeth from the substance being plowed when it becomes gummy creating excessive drag on the shaft. You might say some impact resistance is needed. The standard material was ferritic nodular iron with a silicon carbide insert at the plow location for abrasion resistance. A number P28-8 -9- of materials were involved in the controlled experiment including SiMo ductile iron, high chromium white irons and stainless steels. The basis for selection of these materials for test was fairly straightforward. The SiMo ductile iron was included on the premise that it might last sufficiently longer than standard ferritic ductile iron to meet the objective time and still be low cost. High chromium white iron is well known for abrasion resistance but high temperature endurance in this environment was unknown. Stainless steel was assumed to be the best material to withstand the corrosive environment. It was felt that significantly extended life could justify the added cost of this high chromium-nickel option. Results after one year of continuous service are apparent from pictures of the teeth in Figures 19-22. Standard teeth in Figure 19 are ferritic ductile iron with about 2% silicon and no alloying elements. The smallest tooth suffered the greatest deterioration and was closest to the shaft where the environment is most hostile. The other teeth were relatively further away from the shaft and in environments of declining hostility. This same pattern follows with all of the materials tested. CF8 (cast 30M stainless steel (l9%Cr-9%Ni) is shown in Figure 20, 28% chromium white iron in Figure 21, and k silicon - 1 molybdenum ductile iron in Figure 22. Wear data are numerically summarized in Figure 23 in terms of "percent of original weight remaining" on the vertical axis and roaster arm location along the horizontal axis. Position 1 is hotter and more hostile. Position 1^4 is cooler and less hostile. P28-9 -10- The roaster teeth can be classified by performance into two groups; the high silicon alloyed nodular iron (Alloys 1 and 8) - and all the rest } as clearly evident. There is a distinct difference in the way the nodular irons eroded and the way the stainless steels and high Cr white irons eroded. The nodular irons developed a thick oxide scale which appeared to protect the metal surface and reduce the rate of oxidation. The stainless steels, Alloys 3 and 9, did not develop a thick protective scale, and thus, eroded very rapidly. Part of the rapid deterioration of the stainless steel may have been related to the nickel content. However a specific study would be necessary to verify that possibility. The high Cr irons, Alloys 6 and 7, did develop an oxide scale but it apparently had inadequate stability in the roaster envi ronment . It is clear that the high chromium alloyed materials did not perform any better than the standard alloy in the roaster environment. The oxidation protection normally accompanying high chromium contents in ferrous alloys is apparently just not effective in this environment. On the other hand, it was certainly a pleasant surprise to find the SiMo ductile iron as the most servicing material in this aggressive envi ronment. In conclusion I would simply like to summarize these case histories and restate our premise. P28-10 -11- 1. Four silicon - one moly ductile cast iron is currently in competition with high chromium steels for elevated temperature service in exhaust system components such as turbocharger housings and exhaust manifolds. SiMo is a common turbocharger housing material. It has replaced 24 Cr - 1 Mo in an exhaust manifold. And, it is being challenged by fabricated stair'ess steel for automotive exhaust systems. 2. SiMo ductile iron is being considered as a replacement for 27% chromium HC steel in industrial incinerator grates. 3. SiMo ductile iron has been found to be the superior rabble tooth material in the highly corrosive - erosive environment of a moly sulfide roaster over standard ferritic ductile iron, 21% chromium white iron, and \%% chromium - 3% nickel stainless steels. h. It is probable, but speculative, with relatively minor changes i. design or performance requirements, that SiMo ductile cast iron could replace high chromium steels in numerous high temperature appl icat ions. should a critical need to do so develop. P28-11 c* Figure 1 Automotive Turbocharger P28-12 Figure 2 Automotive Turbocharger -- Cutaway View P28-13 Figure 3 Turbochareer in Service on Test Stand P28-14 8,000 Q. X 1- o z bJ tr 16,000 14,000 Q 12,000 UJ >- Q 7^ 10,000 < UJ _l 8,000< UJ 6,000 FERRITIC N0DULAR-4%Si 788C0450F) ANNEAL 704C (1300F)TEST TENSILE STRENGTH 0.5 1.0 15 2.0 MOLYBDENUM ,% Figure 4 Tensile Properties of 4% Si Nodular Cast Iron at 704 C (1300 F) P28-15 Figure 5 Turbocharger Housing Damage After Containment of Wheel Burst P28-16 TRANSFORMATION TEMPERATURE (A,) £ 700h 2 3 4 5 6 SILICON CONTENT, % 1300 Figure 6 Effect of Silicon Content on Ac^ of Nodular Irons at Various Carbon Contents. Curves Are From the Literature, Data Points Represent Climax Data (Dilatometry) for 3.1% 4-1 CO 0) H 09 a> | CO CO a) CO 4J CO Q C o 0) o M 1-1 o o •d c CO e o o CO o •H e 0) CO a> fH .O CO H C CN • 0) MS t-i m mo m u o^ O 3 0< O U M 3. CO •H 4J • C M MUM M 4J 0/-s O O o> O Cm W w 9 U o s 8>8 c X o cu cu •O 6 O CO o z o z m CO CM ON CO © en fH oo O O O rt • • • • o o o o r>. CO en oo o xo* o en rx. m ?h O O iH o o o o o on en m m m en en cm rx. in O iH vo «* «a- .•a- oo cn »o* cn m o on o o iH o rx. x» cn cn CN CM CO N (O 00 CN CN CM CN CN CM CM fH m o 00 CO m CM CM o SO CM rx. CM en CM CO en CM «a- oo en CM m m o oo o 5% Cr. Table 7. Weight Gain and Composition of Some Fe-Mn-Cr-Al Alloys (ref. 15). Alloy Composition Weight gain at 800 °C (mg cm" 2 ) e-%Mn- -%Al-%Cr 24 Hours 100 hours 5 4 5 .292 (18 hrs) __ 5 5 5 .068 .327 7.5 4 5 .196 .420 7.5 5 3 .122 .493 7.5 5 5 .063 .215 10 4 5 .115 — 10 4 12 .061 — 15 5 5 2.663 The results on two alloys tested for 100 hours at 600°, 800°, and 1000°C under 200 torr oxygen pressure are shown in Table 8. Convoluted scales with almost identical morphologies and composed of an outer layer rich in Mn and Al, and an inner layer of AI2O3, formed on both alloys oxidized at 1000°C. Table 8. 100 Hour Oxidation Tests (ref. 15) Alloy Composition Weight Gain (mg cm" 2) Fe-%Mn-%Al 600 °C 800 °C 1000°C 10 6 .171 .117 .450 20 8 .101 .242* .253* *Weight gains unreliable due to oxide spallation. P29-16 17 Based on this work the authors concluded that suitable alloys from the Fe-Mn-Al system can be produced with oxidation resistance at 800°-1000°C equivalent to that of heat resisting stainless steels. The final formula- tion of these alloys by possible addition of Cr, C, and rare earths is now being undertaken in order to develop better room temperature properties and high temperature strength. OTHER UNPUBLISHED AND ONGOING WORK 1. University of Leeds, U.K. - G. E. Hale (ref. 19) This work performed by G. E. Hale constitutes a part of his soon to be submitted Ph.D. thesis at the University of Leeds. It concentrated mainly on the austenitic alloy with about 30Mn, 8A1, 0.8-1.0C, bal. Fe. Some work was also done on a duplex austenitic/ferritic alloy and a fully ferritic material. The composition of various alloys studied are shown in Table 9. Table 9. Alloy Compositions (ref. 19) ALLOY NO. ROOM TEMPERATURE MICROSTRUCTURE COMPOSITION Mn Al C Fe Others 2 Austenitic 30.4 7.6 0.81 BAL 0.014N,0.36Si 0.009S, 0.017P 3 Austenitic/ Ferritic 17.1 5.0 0.07 BAL 0.06Si 4 Ferritic 8.9 6.9 0.07 BAL 0.007P 5 Austenitic 30.9 8.0 0.97 BAL 0.008N, 0.17Si 0.011S, 0.007P P29-17 18 The Ferritic Alloy ; The ferritic alloy No. 4 shows a distinct "script" phase together with intergranular grain boundary precipitation in the as-cast state (Figure 16) , but this can be removed by hot rolling and/or solution treat- ment at temperatures of 1100°C and above, followed by water quenching. The exact nature of this "script" phase has not yet been determined. Unfortu- nately, this alloy has proved extremely difficult to machine and very brittle at room temperature with a Charpy impact energy of only about 4 Joules. Aging after a high temperature solution treatment followed by quenching results in a fine precipitation in both the grain boundaries and grain interiors. No distinct increase in hardness was detected with increasing aging time, however. The Duplex Alloy ; The duplex alloy No. 3 shows some interesting effects and may well be worth more detailed study. After hot rolling, the alloy contained 73% austenite and 27% ferrite (the relative proportions of the two phases were measured with an X-ray dif fractometer) , but on solution treating at 1200°C and higher followed by rapid quenching (preferably into iced brine) , the alloy becomes single phase ferrite in fairly thin sections (1-2 mm) . Larger sections and/or slower cooling rates lead to austenite precipitation both at ferrite grain boundaries and also within the grains. Reheating the quenched material allows austenite reversion to occur and the amount of austenite formed can be controlled by varying the temperature and time of the reversion treatment as shown in Table 10. This experiment was performed in samples measuring less than 3 mm in thick- ness. In larger sections the proportion of austenite formed is generally higher due to slower cooling rate. P29-18 Table 10. 19 Influence of Reheating on the Percent Austenlte Formed in the Duplex Alloy No. 3 (ref. 19) Reheat Temperature (°C) % Austenite 1250°C/30 mins/IBQ* 16.5** 150° 10.5 200° 14 400° 18 600° 15.5 700° 51 800° 62 1000° 68.5 All spec ime ns were solution treated at 1250°C for 30 minutes in silica tube (under a partial atmosphere of argon) , quenched into iced brine and then reheated at the temperatures shown for one hour followed by water quenching. * IBQ: Ice Brine Quench ** Quenching out of silica tube gives a slower cooling rate than a direct quench from air and this has caused some austenite precipitation during cooling. Tensile and impact data have been measured for this alloy in the hot rolled duplex state and are given in Tables 11 and 12, respectively. Table 11. Tensile Data at Room Temperature for Alloy No. 3 in the Hot Rolled Condition (ref. 19) DTS(MPa) Q.2Z PROOF STRESS (MP a) Z REDUCTION IN AREA Z ELONGATION (15 mm gauge length) 625 485 80 321 P29-19 20 Table 12.. Energy Absorbed in Charpy V-Notch Test for Alloy No. 3 in the Hot Rolled Condition (ref. 19) • O — V IMPACT ENERGY TEMPESAXUitti ( C) (JOULES) 20 193 193 -40 148 -70 102 -85 98 -196 32 ' These figures show acceptable strength levels coupled with fairly good toughness down to at least -85°C (98 Joules). At -85°C, the predominant failure mode was trans granular ductile by microvoid coalescence. Some small areas of cleavage were observed and presumably these are associated with the ductile— to-brittle transition within the ferrite. It is clear, however, that the predominantly austenitic matrix: restricts the propaga- tion of any cracks which originate in the ferritic regions. The absence of a rapid ductile- to-brittle transition in this alloy unlike that observed in certain ferritic steels suggests that alloys of this type could- be useful in certain cryogenic applications. Some limited impact energy data have been measured in samples con- taining about 25% auatenite obtained by a reversion treatment mentioned above. The Impact energy of this specimen, at -40 °C was only about. 27 Joules compared to 148 Joules for the hot rolled sample (Table 12) with 73Z auatenite. This is consistent with previous observations of decreasing ductility with decreasing auatenite content of these alloys. The fracture mode of the sample with 25% austenite was predominantly brittle P29-20 21 transgranular cleavage with small areas of ductile fracture around reformed austenite forming a continuous network along the prior ferrite grain boundaries. This amount of austenite is apparently not sufficient to arrest brittle cracks originating from the ferritic matrix. This result again emphasizes the inherently brittle nature of the ferritic Fe-Mn-Al alloys. The Austenitic Alloys ; A large proportion of this work has been centered on austenitic alloys with compositions close to Fe-30Mn-8Al-lC . Transmission electron microscopy was used to help characterize the aging behavior of these alloys. Hardness versus Aging Time curves were produced for Alloys 2 and 5 at temperatures between 450° and 650°C. In the case of Alloy No. 2, Figure 17, the classic type of age-hardening response was observed with lower tempera- tures eventually achieving higher hardness values but at greatly increased times. For Alloy No. 5, Figure 18, aging between 450 °C and 550 °C showed the standard type of hardness-time curves, while above 600°C, the maximum hardness was lower and occurred after a longer time due to over aging. In Alloy No. 5, the hardness values were always somewhat lower than those measured for Alloy No. 2 at the same temperature. The solution treatment temperature in both cases was 1050 °C for one hour followed by water quenching. It is not yet clear why there is a difference in the aging behavior of the two alloys; the only compositional difference being the carbon content. The transmission electron micrographs in Figures 19-23 show the fine structure which develops in this material on aging. Selected area diffraction analysis has indicated that these precipitates form along <100> directions in the austenite matrix and exhibit a cube-cube orientation with P29-21 22 the austenite, Figure 20. Selected area diffraction analysis also shows the presence of super— lattice spots and that the precipitate has a simple cubic structure and is likely to be ordered. This has been confirmed by x-ray analysis of the extracted precipitate particles, from which a lattice parameter of 379.6 pm was measured and this is in good agreement with the published data for Iron Aluminum Carbide Fe3AlC, although the interplanar spacings for Manganese Aluminum Carbide Mn^AlC are also similar. It would appear that the precipitate is a Fe3AlC type in which there has been some partial substitution of manganese for iron. This has been further confirmed by energy dispersive x-ray analysis (EDX) of the precipitate powder in the scanning electron microscope where iron, manganese and aluminum peaks were detected. (EDX analysis does not detect elements below sodium and therefore carbon cannot be observed by this method. ) The tensile and impact properties of the two alloys both in the solution treated (ST) and quenched condition and after aging at close to the peak hardness values have been measured, Table 13, and are similar to those determined by other workers. The effect of aging time at 550°C (Alloy No. 5 after solution treat- ment: 1050°C/lh/WQ) on the tensile properties is shown in Figure 24. With short aging times, it is possible to further improve the strength without too great a loss in ductility. However, the low temperature impact strength drops off rather drastically on aging. For example, the room temperature impact energy after 1 hour aging at 550°C is still a respectable 119 Joules (Table 13) at -196°C it drops to 5 Joules. The impact energy of the solution treated material shows a small drop from 206 Joules at room temperature to 192 Joules at -90°C and then falls to 106 Joules at -196°C — still a very high value and a low P29-22 a cm o Z •d e ctf m i >> O L. o OS Q •M O OS a e e 0) s 0) Eh CO .Q in i O u o 4J O O CM 10 E (0 £ to U 0) O 3 Z O a fe U u iQ a) £ c u u u ~ 0) £ > 4J o o» w c V c >-* ■h a> (0 3 tT <0 C C7> -h E W E m dP iH c o 4J (0 o cu 3 D TJ < « c CO «-. & 10 W (0 Eh CU 4J c cu E R3 a> 1H Eh 4J «0 CD z 6 o O 3= J-> L So" 1 u 23 vo O CM 9\ * S •H 00 S z m CM in CD CM o CM CM CO CM co 0) CO CO in a» co V CM CM m in CD m cm o O O O in O vo m o> in CM en VO 00 CO CO CO CO o CD o CM 0) O VO CO in CO co O m O O o o o I 2 2 \ o O ~ m m Eh m O co Eh CO Eh CO O m m Eh CO a O O © ^ * * ^ vo CO vo vo iH HT t-H iH \ \ "^ \ o m O m 8 8 m m in vo + + + + Eh Eh H Eh CO CO CO CO CM o z o cy o m o Eh CO CM O o CD Eh CO •P e CD E +> aJ O U P to •H P Li O Sh •d CD Li 3 CO a CD E p O z s z P29-23 24 ductile-brittle transition temperature. Therefore, the increases in strength which occur with aging are offset by a substantial decrease in toughness together with a measurable loss of ductility as well. The aged material failed in a predominantly intergranular brittle mode and was caused by the grain boundary precipitates which are present in the peak-hardened state. 2. University of Calgary, Canada - W. J. D. Shaw et. al. (ref . 20) The work has just started this summer focusing on the creep behavior of an Fe-Mn-Al steel. The high temperature environmental (in steam) creep and electrochemical corrosion behavior will also be studied. Preliminary results of the short time creep tests are shown in Table 14 for an alloy (#SB-8-24-2) containing 30.1Mn-8.3Al-1.5Si-1.07C-Fe bal. in the hot rolled condition. These early results indicate the creep characteristics of this alloy to be equivalent to 9Cr-lMo steels but infoerior to 304 SS. The micro- structural characterization and the corrosion tests will be starting shortly. 3. Virginia Polytechnic Institute and State University, Blacksburg, VA - J. H. Wilson, T. Sudarshan et. al. (ref. 21) This work is also just getting underway. The major emphasis here will be to study the effects of humidity and hydrogen on the tensile strength and torsional fatigue life of Al-Mn stainless steels. In addition, the effects of oxygen and hydrogen on the crack growth rates will also be determined. The initial alloys being tested have the following compositions (Table 15). Table 15. Composition of Fe-Mn-Al Alloys Being Studied at VPI&SU Alloy %Mn %A1 %Si %C %Fe SB-8-24-1 30.1 8.3 0.12 1.07 bal. SB-8-24-2 30.1 8.3 1.54 1.07 bal. P29-24 25 bfi 60 C w >» £ u « • ea JZ .C 60 CO O ^ cm d rt CO «So in • iu-) O O 4J 4J > 01 li CJ C .O E-t a 4J rH 3 CO 0) ftS ^■1 0) r\l D. o 1 CU J-l sj CU to (SI 1 H 00 < 1 >s 1 pa h c CO 2 g •h B •H rH 0) H c oen «0 o 0) 0) iJ o 0) 4J ea u u .cen O. — I cu e o u -^ x o c • -H 60 09 u 0) rH •H C3 OJ e 0) J-l 3 H CU a CO PL, CO CO a> n CO 1-1 0) 1 z o cu Cu C/3 CM CN O m o r» O O VO r^ «3- 00 «* rH CN rH CO CT» c vO o CO en m en en CM m cm cm vo CM CM o m sr rH © iH odd CM o oo cr» m m CM 00 CX» 0\ m <■ rH NO en o 00 O en vO en O T-l r^. CT. o o o o o o o O o o m o o o m o o O m in iH o en en rH CM o O iH rH tH rH rH rH iH iH rH rH iH rH O O o o o o o o o O O o o o o O O O o o o O O O O o o O O O A M A * •* m •fc #s •t #» o *3" o sr O CM o CM o CM o CM o CM O CM o rH m m + * en a> cm r * r* co 60 c •H H t-l 3 V V o cu r-l CO 3 CO rH cu •H m CO 00 U-l o u o Ou 4J c ^ •H o 1-1 rH t-l rH o- •H 4J •o CO cu a 4J a. CO o cu 4J H CO P29-25 26 The first samples cathodically charged with hydrogen and also charged during tensile straining showed approximately the same degree of suscepti- bility to hydrogen embrittlement as most common austenitic stainless steels. Intergranular fracture, typical of hydrogen- induced cracking was observed, Figure 25. Utility of these steels in food processing and fertilizer industries is envisioned depending on the results of this work. 4. EG&G Idaho, Inc., Idaho Falls, ID - G. Korth (ref. 22) A major program on Mn-Al steels has also started recently at EG&G Idaho. This will be a rather comprehensive study involving tensile, impact, creep, fatigue, corrosion, stress-corrosion and weldability properties of these steels. Two alloys with about 30Mn/8Al/lC/Fe, one without Si and one with 1.5% Si have been produced for the initial tests. Early indica- tion of weldability is encouraging. Some corrosion test coupons in geo- thermal wells are being tested. Other experimental work is just beginning. SUMMARY In summary, it is gratifying to note that considerable progress has been made over the past year and a half on the evaluation and characteriza- tion of Fe-Mn-Al alloys. However, most of the work done and the information presented in this report are somewhat preliminary in nature. Much work still remains to be done for complete evaluation and understanding of various properties of this alloy system. It appears that in certain applications such as heat resistant materials at moderate temperatures, corrosion resis- tance in some environments, higher strength-to-weight ratio materials, and cryogenic applications, etc., these alloys offer good substitution potentials for chromium-bearing steels. P29-26 27 It is hoped that the interested agencies will offer assistance and encouragement for continued and more thorough investigations of this new family of steels, of f ering _one option in the search for conservation and substitution technology for chromium. ACKNOWLEDGEMENTS The continued interest of Dr. Allen Gray in the Fe-Mn-Al alloys as potential substitutes for Cr-bearing steels is greatly appreciated. The author is grateful to the ASM and the AIME for providing the preprints of the manuscripts of the two papers by R. Wang et. al. and the paper by K. Narasimha Rao et. al., respectively. The efforts of the individual contributors who have privately communicated with the author and provided the information and results of their work presented in this report are gratefully recognized. I also wish to express my special thanks to F. Boratto and P. Tomas for providing the reprints of the papers published in Brazil and Australia. Finally, I wish to acknowledge the continued support of Foote Mineral Company and its permission to present this report. REFERENCES 1. S. K. Banerji; "An Update on Fe-Mn-Al Steels," Proc. of the Workshop on "Conservation and Substitution Technology for Critical Materials," Vanderbilt University, Nashville, TN, USA, held June 15-17 (1981). 2. S. K. Banerji; "An Austenitic Stainless Steel Without Nickel and Chromium," Metal Progress, p. 59, April (1978). 3. S. K. Banerji; "An Austenitic Stainless Steel Without Nickel and Chromium," Trans. Indian Inst, of Metals, Vol. 30, No. 3, p. 186, June (1977). 4. J. C. Garcia, N. Rosas and R. J. Rioja; "Development of Oxidation Resistant Fe-Mn-Al Alloys," Metal Progress, p. 47, August (1982). 5. R. Wang and F. H. Beck; "An Austenitic Mn-Al-Si Steel for Propellers of Sea-Going Vessels," to be published in Metal Progress. P29-27 28 6. R. Wang and R. A. Rapp; "Marine Corrosion Behavior of Several Fe-Mn-Al Austenitic Steels," to be submitted for publication. 7. K. Narasimha Rao, R. Sivakumar and M. L. Bhatia; "An Evaluation of Al-Mn-C Austenitic Steel," submitted to Journal of Metals. 8. L. C. Casteletti and D. Spinelli; "The Resistance to Oxidation and the Electrical Resistivity of an Fe-Mn-Al System Alloy," Dept. of Mat Is. Sci., EESC, Univ. of Sao Paulo, Sao Carlos, Brazil; Proc. XXXVI, Brazilian Soc. for Metals (ABM) Annual Congress, p. 249, held in Recife, Pernambuco, Brazil, July 5-10 (1981). 9. F. C. R. Assuncao and M. F. da S. Lopes; "Study of Oxidation of Fe-Al-Mn-C Alloy at High Temperatures," Dept. Matls. Sci. & Met., Catholic Univ., Rio de Janeiro, Brazil, ibid, p. 329. 10. A. J. A. Bushinelli, J. C. Dutra and W. May; "Study of the Micro- structural Transformations in TIG Welding of an Fe-Al-Mn Alloy," Fed. Univ. of Santa Catarina, Florianapolis, Brazil, ibid, p. 521. 11. L. C. Casteletti and D. Spinelli; "Mechanical Properties of an Austenitic Steel of the System Fe-Mn-Al," Dept. Matls. Sci., EESC, Univ. Sao Paulo, Sao Carlos, Brazil, ibid, p. 299. 12. A. P. Tschiptschin and H. Goldenstein; "Precipitation Reactions in Low-Carbon Fe-Mn-Al Alloys," Dept. Met. Eng. , Univ. of Sao Paulo, Brazil, ibid, p. 117. 13. J. R. T. Branco and F. J. M. Boratto; "Domain of the Austenitic Phase in the Fe-Mn-Al System at 1000 °C," Met. Tech. Section, Tech. Center Found, of Minas Gerais-CETEC , Belo Horizonte, Brazil, ibid, p. 175. 14. P. Tomas; "Elements Essential to High Temperature Sulphidation Resistant Iron-Based Alloys," Univ. of New South Wales, Kensington, NSW, Australia, Proc. 35th Annual Conf . of the Australasian Inst, of Metals, p. 90, held in Sydney, Australia, May (1982). 15. P. R. S. Jackson and G. R. Wallwork: "The Development of Fe-Al Alloys for High Temperature Oxidation Resistance," Univ. of New South Wales, Kensington, NSW, Australia, ibid, p. 78. 16. D. J. Chakrabarti; "Phase Stability in Ternary Systems of Transition Elements with Aluminum," Met. Trans. B, Vol. 8B, p. 121, March (1977). 17. P. Tomas and S. K. Banerji, private communication (1982). 18. R. A. Perkins; "Sulphidation-Resistant Alloy for Coal Gasification Service," Qtrly. Report #FE-2200-6, Lockheed, Palo Alto, CA (1976). 19. G. E. Hale, Dept. of Metallurgy, University of Leeds, Leeds, U.K.; Ph.D. thesis to be submitted, private communication (1982). P29-28 29 20. W. J. D. Shaw, Dept. of Mech. Eng., University of Calgary, Calgary, Alberta, Canada; private communication (1982). 21. T. Sudarshan and J. H. Wilson, Depts. of Materials and Agricultural Eng., VPI&SU, Blacksburg, VA, private communication (1982). 22. G. E. Korth, EG&G Idaho, Inc., Idaho Falls, ID, private communication (1982) . SKBanerji:erd 10/82 P29-29 30 Weight gain, g/m» 1010 carbon steel •■>_.;..- -rM-- 300 feW Time, t* Fig. 1. Plots of weight gained versus time at 700°C (1290°F) for the six experimental alloys of Table 1, type 304 stainless steel, and a 1010 carbon steel. Parabolic regions of the data curves are presented here. Note that the slope for alloy B is similar to that for type 304. Pressure: 1 atm (100 kPa) of flowing oxygen, (ref. 4) P29-30 31 Weight gain, g/m« .V^ 1010 carbon steel 0? Tltne,s^ 10O &£>£*'. &*&***+!£*& 200 Vj > ; 300 VJV;.«^^^.ii»i»*.i.:i^.'.»-jSri*».^A^ Fig. 2, Plot of weight gained versus time at 500 C C (930°F) for the six experimental alloys of Table 1_, type 304 stainless steel, and a 1010 carbon steel. Parabolic regions of the data curves are presented here. Note excellent behavior of alloy F. Pressure: 1 atm (100 kPa) of flowing oxygen, (ref. A) P29-31 32 4 AISI 3044 • pouring 13 (31Mn, 7.5A1.1.3S1, .93C,Fe) without silicon 12 4 6 8 12 24 ae Zg — — — ~ : 70 *° time (hours) ,w Fig. 3. Weight gain as a function of time at 700°C in air. (ref. 8) Q00Q2 024 Fig. A. Weight gain as a function of time at 800°C in air. (ref. 8) P29-32 33 1603 - 1790 s 177.0 1 • /• S|750 >> u X # ascent t 1730 M» decline 4J (0 •H 00 01 * 1710 ■• 1690 - •/ 1 1 1 1 II 1 1 100 200 300 400 500. „„J3PjQ„„i.7C •teraperature^C) 700 612 Fig. 5. Temperature dependence of resistivity for an Fe-31Mn, 7.5A1, 1.3Si, 0.93C alloy, (ref. 8) P29-33 34 pouring \0 © AIS! 304 (REF 8 ) A AISI 316 (REF. 8) L AISI 202 (REF. 9 ) 24 100 200 300. 400 500 600 700 800 Test temperature (°C) Fig. 6. Variation in tensile strength with temperature, (ref. 11) P29-34 35 • Pouring 13 e AISI 304 (REE 8) A AISI 316 CREF 8 > A AISI 202 (REE 9 ) 24 100 200 300 400 500 600 700 *-B00 Test temperature CO Fig. 7. Variation in yield strength with temperature, (ref. 11) P29-35 36 "Pouring 13 AISI 304 (REF 10) o A AISI I AISI 316 (REF. 10) 202 (REF 10) 20 40 60 Reduction (%) 80 Fig. 8. Variation in tensile strength with the degree of reduction, (ref. 11) P29-36 37 60 Reduction (%£° Fig. 9. Variation in yield strength with the degree of reduction, (ref. 11) P29-37 38 • Pouring 13 A AISI 202 (REE 10) AISI 304 (REE 10) * AISI 316 (REFIO) Reduction Fig. 10. Variation in elongation with the degree of reduction, (ref. 11) P29-38 39 « v Pouring 13 o AtSI 201 (REE t>) A AtSI 301 (REF10) I 20 40 60 Reduction' (%9° Fig. 11. Variation in hardness with the degree of reduction, (ref . 11) P29-39 40 £50 £ 180 « (Pouring 13 A Fe - 16 Mn - 13 Cr - 2 Mo « 02 N (REF. II ). A Fe-l6Mn- l3Gr-l5Si-CX2N o o o °o °o °o °o °o LO O LO O l/> *»» U"> IT) ID 10 LU LU P LU O < 03AH SS3N0UVH P29-46 47 Wfcr : v: I o-^S/w j Fig. 19. Alloy No. 2. Hot rolled bar - ST 950°C/30 mins/water quench (WQ) Cold rolled 90% ST 950°C/1 hr/WQ Age 600°C/3 mins/WQ Note the presence of very fine precipitation after aging for only three minutes, (ref . 19) P29-47 Fig. 20. Alloy No. 2. Treatment as in Fig. 19 but aged at 600°C for 30 minutes and water quenched. Dark field micrograph from (010) precipitate spot in Fig. 21. (ref. 19) • • • * v teoO L 00 Vl*c.c Fig. 21. Alloy No. 2. Selected area diffraction pattern from area seen in Fig. 20 showing superlattice spots intermediate between the austenite matrix spots, (ref. 19) P29-48 49 Fig. 22. Alloy No. 2. Treatment as in Fig. 19 but then aged at 600 °C for one hour and water quenched. Bright Field Micrograph, (ref. 19) Fig. 23. Alloy No. 2. Dark Field Micrograph of area seen in Fig. 22, Zone axis just of [001]y. Operative reflection is a (100) precipitate spot. (ref. 19) P29-49 50 80 60 40 60 c o 1-1 w "O c to CO 01 u < c 20 -h 4J u 9 0> OS 3*8 ST+Q 10 20 30 40 Aging Time (Hours) 50 Fig. 24. The effect of aging time (at 550°C) on the tensile properties of Alloy No. 5 after ST 1050°C/lHr/WQ. (ref. 19) P29-50 51 l/Pl&SU Fig. 25. Hydrogen- induced inter granular cracking in an Fe-Mn-Al alloy, (ref. 21) P29-51 THE DEVELOPMENT OF NEW ALLOYS TO REPLACE CHROMIUM IN CARBURIZING STEELS FOR GEARS AND SHAFTS CARL J. KEITH AND V. K. SHARMA Materials Engineering and Technology International Harvester Company Hinsdale, Illinois ABSTRACT A status report on the Bureau of Mines sponsored research to develop a new series of cost-effective chromium-free steels is provided. These new steels will be capable of adoption by a broad range of industry as a substitute for the chromium-containing 8600 and 4100 standard grades of heat treated constructional alloy steels. Based upon several socio- economic scenarios and their projected impact on alloying element cost and availability in 1985, International Harvester's computerized metallurgical design system has been used to design two substitute steels. Both steels, a manganese-molybdenum substitute and a manganese-nickel-molybdenum substitute, are expected to provide microstructure, heat treat response, and mechanical properties equivalent to the 8600 and 4100 steels. Small experimental heats of both chromium-free replacement steels were prepared and their hardenability response and mechanical properties were evaluated and compared to the base line 8620 steel. P30-1 INTRODUCTION Chromium is both a critical and strategic alloy element. With the exception of approximately 8 percent chromium recycled from stainless steel scrap, the U.S. relies totally on imports. Chromium resources in the Western Hemisphere are limited to about three-tenth of one percent of the world resources of chromite ore. Identified world supplies of chromium are sufficient to meet conceivable demand for many centuries, however 97 percent of the world's chromite reserves and 98 percent of its identified resources are located in Rhodesia (Zimbabwe) and South Africa^). This very high concentration of world reserves in a politically unstable area, coupled with the relatively rapid rise in world demand and limited substitution potential, creates a potentially dangerous situation. The National Materials Advisory Board in their report on contingency plans for chromium utilization points out that "at some future time the major source of chromite could become inaccessible to the United States as a result of political action. "(2) Depending on the level of economic activity, sixty-to-seventy percent of chromium consumed annually in the United States is used for metallurgical purposes. The bulk of this usage is in the production of heat resistant steels, superalloys, and corrosion- resistance stainless steels. At the present time there exists no viable substitute for chromium as used in these heat and corrosion resistant applications. However, approximately 17 percent of the chromium consumed annually is used in the production of the constructional alloy steels. In these applications chromium is used primarily for its effect on hardenability and, while its use is both highly efficient and cost-effective, it can be substituted for. Two particular grades of constructional alloy steel, chromium-molybdenum (4100) and nickel-chromium-molybdenum (8600), account for 60 percent of the chromium used in the constructional alloy steels and thus approximately 10 percent of the total U.S. demand. Conservation by substitution of this chromium could reduce demand by 10 percent or alternatively expand existing supply thereby extending the supply for those heat- and corrosion-resistant applications for which there is currently no viable substitute for chromium. The purpose of the Bureau of Mines sponsored research described in this paper is to develop a new series of chromium -free constructional alloy steels capable of replacing the two chromium-containing AISI 8600 and 4100 steels. The goal is to substitute for the chromium in these two steels other cost-effective alloying elements in which the U.S. is self-sufficient or for which a more secure and stable source of supply exists. In the event of a chrome supply disruption or significant price escalation in the time period 1985-1995, the new chrome-free compositions can be adopted by a broad range of U.S. industry without significant testing and development effort. To be readily acceptable to domestic steel producers, the new chrome-free steels have been designed to avoid the necessity for special production practices which are not currently required for the production of the standard grades. For example, although it is more cost-effective to design chrome-free steels with manganese level exceeding 1.30 percent, the manganese range was limited to 1.00 - 1.30 percent because a broad range of domestic steel producers are not accustomed to producing higher manganese steel on a commercial basis. P30-2 SUBSTITUTE STEELS The acceptance and adoption of new steels as standard grades is a relatively lengthy process requiring many years. Most of the currently used constructional alloy steel grades were developed in the late 1930s and adopted in the early 1940s as National Emergency Steels during World War II. Engineering design of parts and components, using these steels is well advanced, and manufacturing methods rely on process control parameters developed only after long and painful experience. It would be very difficult to replace parts made from these standard steels with parts made from new alloy compositions having different hardenability, heat treat response, and mechanical properties without redesigning the parts. Redesign of the parts would involve extensive engineering and manufacturing test programs. Engineering test programs would be required to ensure achievement of required engineering performance, and reliability and manufacturing test programs would be required to establish new process control parameters to enable consistent production of parts within required tolerance levels. However, experience has shown that standard steels can be replaced by new steels of different alloy composition provided that the new steels exhibit the same hardenability, heat-treat response and mechanical properties. By careful control and balance of the composition, new steels can be developed to replace the standard steels in current parts without the need for redesign of the parts and with only nominal changes in process control parameters. Basic assumption in the development of such substitute steel is that the engineering performance of a heat treated component is controlled by the carbon content, microstructure, and residual stresses. The microstructure and residual stresses are governed by the carbon content, hardenability, and martensitic start-and- finish transformation temperatures. Alloy composition controls hardenability and transformation temperatures. Opportunities, therefore, exist to develop substitute steels composed of alternate cost-effective and/or readily available non-strategic alloy elements provided the new steels have the same carbon content, hardenability, and martensitic transformation temperatures. The use of substitute steels as replacements for the standard steel grades is not a new concept. The SAE specification 31081 was developed to provide a uniform means of designating new steels during a period of usage prior to their acceptance as standard steel grades. Beginning in the early 1960s, new replacement steel grades developed by industry have been listed in this specification as EX steel grades. At International Harvester a computerized metallurgical design system has been used successfully over the past ten years to develop a number of proprietary, minimal cost, alloy compositions which are presently widely used in the production of IH products. None of these new steels is a chromium-free steel. At the present time, the use of chromium to enhance hardenability is highly cost-effective and the wide spread implementation of chromium substitution would not be in the economic best interest of the nation. As noted earlier, in the design of new chromium-free steels alloy costs have been minimized within the constraints of self-sufficiency and/or availability considerations. Computer-harmonizing or the CH portion of International Harvester's metallurgical alloy steel design system CHAT (Computer Harmonized Application Tailored) was used to design new chromium-free steels. The IH CHAT system has been fully described elsewhere w^). The CH portion of the system is a computer program which uses linear and separable programming to optimize a primary or objective function while at the same time satisfying a series of additional functions. When the CH program is used to design replacement steel grades, the primary or objective function is described as the total cost of the alloy additives required to make the steel. This primary function is P30-3 minimized, while at the same time a series of restrictive functions embodying minimum and/or maximum base and case hardenability requirements, transformation temperature requirements, and additional limitations aimed at controlling intergranular oxide penetration during heat treatment and carbide formation during carburizing are all simultaneously satisfied. ALLOY STEEL COST FACTORS One of the key pieces of information required in designing a cost-effective replacement steel is the alloy steel cost factors. The cost factors used in the CH portion of the CHAT system are the raw material costs to the steel mill to add one percent of the alloy element to a full heat of steel. These cost factors are modified for the size of the heat and, the yield and recovery factors which account for oxidation, metal slag reactions, etc. and affect the efficiency of the alloy element additions. The CHAT system selects the optimized combination of alloy elements (minimized cost) meeting the input requirements and limitations. The resulting CHAT composition is then modified to provide an optimum price in terms of its "chemistry grade extra" as determined by commercial steel product price books. This difference between steel cost and steel price must be clearly appreciated. Although there is a close correlation between steel mill cost and steel product price, there is not always a one-to-one correspondence. In designing the chromium-free steels, the 1981 alloy cost factors were projected to 1985 and then modified to reflect the following four future chromium supply scenarios considered probable in 1985: Scenario A; Chromium supply will remain stable with prices rising in a steady manner as world economic and political conditions continue to evolve in a relatively orderly fashion. Based upon world conditions existing at this time, this scenario appears to be the most likely and is estimated to have a relatively high 75 to 80 percent probability. Scenario B: Chromium remains available but supply and price fluctuate as the principal African producers seek to form an OPEC -like cartel and attempt to control pricing. This scenario is considered to be eminently possible due to the concentration of resources in southern Africa. However, this scenario is assigned a relatively low probability of 15% because of the high dependency of that area on export of minerals for foreign trade; the relatively low economic development of the area; and the reasonable availability of lower grade ores from other sources (including industrial and governmental stock piles) for both alloy element and ferro alloys. Scenario C: Chromium supply from African sources is cut off as a result of a regional war or other occurrences causing a complete breakdown of export capability from the South African sources. As a result of political unrest, the potential for insurrection and/or war in this area is very real. The relative isolation of South Africa in the United Nations and world community and the active support of Soviet Bloc nations for the South African opposition groups are destabilizing factors. However, the interest of the major western powers in the mineral resources of the area and the relative economic and military strength of the nation of South Africa in relation to its own neighbors indicate that a war could probably be contained without seriously affecting export. The probability of a complete breakdown of South African production and export potential is therefore considered to be relatively low, i.e. 5 percent. P30-4 Scenario D: This scenario assumes the breakdown of African chromium supply as in Scenario C above, but then further assumes that at the same time Soviet and Albanian sources of chrome supply would also be cut off. This fourth scenario D is assigned a very low probability of 1 percent. The costs for the alloy elements developed from various scenarios are shown in Table I. The cost forecasts, developed by the Resources and Technology Forecasting Group of IHC Corporate Technology as a part of an ongoing study, are intended to represent long-term trends rather than short-range fluctuations. TABLE I - STEEL ALLOY COST FORECAST FOR 1985 Unit = $/ton (2,000 lbs) Ni* Mo* Cr* Mn* Ferroalloy s Scenario do NiO CIO M0O2 do Cr20 3 do MnO FeCr FeMn FeSi i Scenario A 9,760 31,240 489 97.2 1,431 750 1 , 240 , Scenario B 10,736 34,364 562 107.0 1,646 862 1,302 Scenario C 11,712 37,488 3,600 116.6 3,856 937 1,364 Scenario D 11,712 37,488 3,960 116.6 4,242 937 1,364 1981 Prices (Nov-Dec 1981) 7,160 21,340 360 : 71.5 1,067 538 900 * Price per ton of Contained Metal: Scenario A: Stable world conditions, 75-80% probability Scenario B: Producers' Cartel, 15% probability Scenario C: South African Breakdown, 5% probability Scenario D: Scenario C plus Soviet and Albanian sources cut off, probability 1% The alloy element cost forecasts were reviewed with a number of experts in various areas of alloy availability. Of particular interest were comments from Climax Molybdenum Corporation regarding the projected costs for molybdenum oxide. The forecasts, conducted at a time when molybdenum was still considered to be in short supply, projected a 1985 molybdenum price tending toward $16/lb under stable world conditions of Scenario A. The Climax Molybdenum representatives suggested that the existing 1981 producer price of $6.85/lb would prevail throughout 1982 and would not exceed $10/lb by 1985. It was pointed out that the continuing economic recession, depressed steel market demand, and increased world production had affected producer inventory levels to a point where they could not be adjusted in the near-term future. Figure 1 and Table II reported in American Metal Market, July 1982,(5) tend to confirm this outlook. P30-5 10. m 8.00 6.00- 4.00 2.00 22.0 - 17.0 -12.0; -7.0 0. 00 I 1 1 ! " ! "!" ! " ! " ! " ! " ! " ! " ! " !" 1 "!" 1 "!" 1 2 - 76 78 80 82 84 86 88 90 MOLY OXIDE. REAL PRICE 1980$ n U.S. INVENTORY. MONTHS CONSUMPTION CHASE ECONOMETRICS FERROALLOY SERVICE MAY 1982 AMM— MOLYBDENUM SUPPLEMENT Figure 1 - Molybdenum Price and U.S. Inventory (5) TABLE II - THE OUTLOOK FOR MOLYBDENUM^) (in thousands of metric tons) Consumption United States Other Countries TOTAL 1982 1985 1990 25.4 43.6 69.0 34.7 60.0 94.7 36.5 80.8 117.3 Supply (Mine) Capacity Operating Rate Production 117.5 .69 81.7 140.3 .74 103.7 156.2 .84 131.6 Price of Oxide ($/lb) 1980$ $ 5.60 $ 6.80 $ 7.93 The cost factors used in the CH computer program significantly affect the selection of the optimum composition. To more fully evaluate the problems of cost projections, a cost sensitivity analysis for two competitive alloying elements, nickel and molybdenum, was performed. This cost sensitivity study included the constraints imposed in matching the metallurgical characteristic base and case hardenability of the 8600 and P30-6 4100 series steels, and is not applicable to other steel grades having different metallurgical characteristics. The results indicate that for the 8600 and 4100 series steels a Mo/Ni cost ratio of 2.4 is a critical point. Above this ratio (i.e., Mo cost greater than 2.4 times Ni cost), nickel substitutes to reduce Mo content; below this ratio, nickel does not substitute for Mo. A review of the history of alloy costs over the past 20 years indicates that the Mo/Ni cost ratio has existed on both sides of the 2.4 critical ratio. As it will be seen later, it is precisely for this reason that two different chrome-free compositions, a manganese-nickel-molybdenum and a manganese-molyb- denum steel, have been developed in this program. COMPOSITION OF CHROMIUM-FREE STEELS Manganese, nickel and molybdenum, are three possible elements which can compensate for the base and case hardenabilities lost due to the removal of chromium in the chrome-free steels. Increasing the base carbon can increase the base hardenability, but in designing substitute steel only a minimal change in the base carbon is permitted. Obviously a change in the base carbon will have no effect on the case hardenability. Silicon contributes to case hardenability significantly but has no significant effect on base hardenability unless increased beyond approximately 0.60%. Silicon range in standard AISI steels is 0.15 to 0.35%. Producers and users of carburizing steels do not have experience with higher silicon carburizing grades of steels. Therefore, three key elements which can be effectively used to replace chromium are manganese, nickel, and molybdenum. Although manganese is a relatively inexpensive alloying element, consideration was given to the use of both high and low manganese in the computer design of the chromium-free replacement steels. South Africa currently supplies approximately 35% of the Western demand for manganese^', and the formation of a producer cartel for chromium which included the South African sources for chromium would in all probability affect manganese also. A breakdown of the African export capability for chromium would also involve export capability for manganese. These factors, and their effects on manganese costs, are included in Table I, Steel Alloy Cost Factors, under Scanarios B, C, and D. However, it is assumed that short-term manganese shortages could be offset by increased production from other western suppliers (primarily Australia and Brazil). Four different manganese ranges varying from 0.40 to 1.50 percent Mn were considered. These four manganese ranges, along with additional restrictions of fixed silicon range, no chromium, and metallurgical base and case hardenability requirements for 8622 and 4118 steels, were included as input to the CH program. Using the 1985 alloy steel cost factors for various scenarios, the computerized alloy steel design system provided optimal chromium-free compositions for each manganese range. Chromium-free replacement steel compositions developed for the AISI 8620 steel are shown in Table III. The CH program first provided only the -1/4 chemistry. The +1/4 chemistry was established after a due consideration to the chemical range limits (tolerances) for the various alloying elements established by the American Iron and Steel Institute (AISI). This was accomplished within the CHAT system by varying the input requirements. Input requirements representing the minimum, maximum, and/or median values of the metallurgical characteristics describing the steel were varied systematically until the resulting new composition met AISI range tolerances and the primary objective function was optimized. Notice that the chromium-free chemistries in Table III are applicable to all four scenarios. Changes in the alloy cost factors resulting from various scenarios did not alter the composition. The computer selected the same cost-effective chemistry irrespective of alloy cost factors in Scenarios A to D, which indicates that P30-7 TABLE III COMPOSITIONS OF CHROME-FREE REPLACEMENTS FOR AISI 8620 STEEL ITEMS AISI 8620 CH 1 VERSION SELECTION ■ - STANDARD RANGES .40-.6096 Mn .70-.9096 Mn 1.00-1.30% Mn 1.20-1.50% Mn -1/4 +1/4 -1/4 + 1/4 -1/4 + 1/4 -1/4 +1/4 -1/4 +1/4 C Mn Si S Cr P ; Ni Mo .18 .23 .75 .85 .20 .30 .03 .05 .45 .55 .01 .03 .47 .63 .17 .23 .18 .45 .20 .02 .01 .72 .23 .55 .30 .08 .05 .83 .18 .75 .20 .02 .35 .42 .23 .85 .30 .08 .55 .48 .18 .23 1.07 1.23 .20 .30 .02 .08 .35 .55 .27 .33 .18 .23 1.27 1.43 .20 .30 .02 .08 .25 .35 .27 .33 D Ib Min 1.42 1.30 1.41 1.53 1.71 Dj, Max 2.33 2.06 2.19 2.45 2.64 D Ic Min 4.31 5.03 4.15 4.09 4.39 D, max 6.45 7.08 5.83 6.21 6.23 M s Base 7970 7450 830O 7820 807O 755° 7910 7360 7840 7310 M 10 7790 7270 8120 7640 7890 7370 7730 7180 7660 7130 M 50 712° 661° 7450 698° 7 220 670° 707° 6510 6990 6470 M 90 612° 560O 6450 5970 621o 5690 606O 5510 598Q 5460 M f 410° 358° 443° 3950 4 20O 3680 404O 349° 3970 3440 M s Case 305° 2870 343° 330O 3210 304O 305O 2840 2960 2780 A cl 1329° 1330O 13360 13390 13190 13180 1313° 13110 13120 13130 A c3 1528° 1515° 157 20 1565° 1545° 15310 1537° 1523° 1540O 15280 A e3 1493° 1477° 1516° 1504° 1498° 14810 1490° 1471° 1488° 1473° Cost Factor $- 1981 $9.50/cwt $19.50/cwt $14.85/cwt $10.85/cwt $10.05/cwt $- 1985A $12.10/cwt - - $13.20/cwt $12.60/cwt $- 1985B $14.90/cwt - - $15.25/cwt $13.85/cwt $- 1985C $17.80/cwt - - $15.90/cwt $15.30/cwt $ - 1985D . $22.40/cwt - - $15.90/cwt $15.30/cwt Note: Dj values calculated from +1/4 chemistry. P30-8 under the constraints imposed relative per dollar contributions of various alloying elements to hardenability does not change with the increase in the alloy cost factor from Scenario A to D. Table III shows +1/4 chemistries, characteristic hardenabilities, and various transforma- tion temperature for the 8620 and the computer harmonized replacement versions for the four manganese ranges. Present and 1985 estimated prices under the four scenarios are given in the last column. The .40 - .60 Mn range results in the development of a Mn-Mo steel with a .70 -.85 percent molybdenum range. It would be a costly steel, and its use would require more extensive testing. The .70 - .90 Mn range results in the development of a Mn-Ni-Mo steel, which again is relatively costly. The 1.20 - 1.50 Mn range results in the development of the most cost-effective compositions. However, in current steel production practice, the use of manganese above 1.20 to 1.30 percent involves additional problems in ingot casting practice due to a tendency of the high manganese steel to cause cracking if allowed to cool in the ingot. This problem can be avoided by vacuum degassing and/or by controlled processing in the steel mill. Since the chromium-free steel to be developed in this project is intended for use by a broad range of industry, without a significant change in current production practices, the 1.20 - 1.50 percent Mn range is rejected, as are the .40 - .60 percent Mn and .70 - .90 percent Mn ranges, because they are too expensive. Based on these reasons, the 1.00 -1.30 percent Mn range which results in a replacement steel slightly more expensive than the 1.20 - 1.50 percent Mn steel, is selected as the optimal manganese range. The 1.00 - 1.30 percent Mn steel CH composition was further refined to provide an optimum price in terms of the chemistry grade extras and to reproduce all metallurgi- cal requirements of hardenability and transformation temperatures of the parent 8620 steel as closely as possible. Table IV shows the revised Mn-Ni-Mo composition. In Table III, for the 1.00 -1.30 percent manganese steel, the computer indicates an optimal cost with a steel containing nickel at a range of .35 - .55 percent. However, commercial practice does not distinguish (provides no price break) a difference on nickel content below a value of 0.7 percent. Therefore the nickel content in Table IV was increased to a range of .55 - .65 percent without affecting cost. This increase enhances case hardenability without significant increase in the base hardenability. Another modification in the chemistry is a decrease in the carbon range from 0.18 - 0.23 percent to 0.16 - .21 percent. Examination of the output of the CH program, Table III, indicates that the high carbon, case hardenability (Dj c ) is the primary factor influencing the alloy composition. In order to meet the minimum Di c requirement, the base hardenability, D15, is being exceeded. Experience has shown that although the increased base hardenability does not affect the engineering performance of the new steel, manufacturing process problems are sometimes encountered as a result of increased core hardness in comparison to the previously used standard steel. The problem of increased core hardness or base hardenability can be ameliorated by a slight reduction in the carbon content. This reduction in the carbon content reduces the base hardenability without affecting the case hardenability. The resulting calculated Jominy hardenability band for the chrome-free steel matches the standard 8620 very closely, as shown in Figure 2. P30-9 TABLE IV COMPOSITIONS OF CHROME-FREE REPLACEMENTS FOR AISI 8620 STEEL Items AISI 8620 CH" - revisions' 1.00-1.30% Mn Mn-Ni-Mo 1.00-1.30% Mn Mn-Mo -1/4 +1/4 -1/4 +1/4 -1/4 +1/4 C Mn Si S Cr P Ni Mo .18 .23 1 .75 .85 .20 .30 .03 .05 .45 .55 .01 .03 .47 .63 .17 .23 .16 .21 1.07 1.23 .20 .30 .02 .08 .55 .65 .27 .33 .16 .21 1.07 1.23 .20 .30 .02 .08 .01 .05 .37 .43 D15 Min 1.42 1.55 1.52 Dj5 Max 2.33 2.28 2.19 Di c Min 4.31 4.42 4.43 Dj c Max 6.45 6.41 6.11 M s Base 7970 7450 800° 748° 8160 766O Mio 7790 7270 782° 730O 7980 7480 M50 7120 6 6lo 7160 6630 7310 68IO M 90 612° 560° 6150 563° 6310 580O M f 410 o 358° 413° 361° 429° 379° M s Case 305° 287° 2990 299° 3130 2970 A cl 1329° 1330° 1307° 1307° 1324° 1326° A c3 1528° 1510° 1540° 1540° 1561° 1550° A e3 1493° 1477° 1490° 1490° 1507° 1494° Cost Factor $- 1981 $9.50/cwt $10.85/cwt $9.60/cwt $ - 1985A $12.10/cwt $13.20/cwt $11.85/cwt $ - 1985B ! $14.90/cwt $15.25/cwt $12.95/cwt $ - 1985C $17.80/cwt $15.90/cwt $14.30/cwt $- $1985D $22.40/cwt $15.90/cwt $14.30/cwt NOTE: Di values calculated from +1/4 chemistry. P30-10 C/3 co LU z Q cr < X o 6 5 5 5 3 5 5 25 o □C 1 5 I I I I CALCULATED H-B I I \ ■X. V •N •♦>. •^ ft 1 2 1 6 2 2 4 2 8 3 2 JOMINY DISTANCE 1/16 INCH Figure 2 - Standard 3ominy hardenability band for AISI 8620 steel compared to the calculated hardenability band for the chrome-free replacement steels. It should be pointed out that the Mo/Ni price ratio used in this portion of the CH development program is approximately 3.2, as shown in Table I. Consideration of a Mo/Ni price ratio below 2.4, which actually exists at the time of writing this report, results in the development of the manganese-molybdenum composition shown in Table IV as the most cost-effective replacement steel. Since the Mo/Ni price ratio has historically varied on both sides of the critical ratio of 2.4, both the Mn-Ni-Mo and the Mn-Mo chromium-free steels are considered as future replacement compositions. The selection of the most cost-effective replacement will depend upon the relative costs of nickel and molybdenum at that future time when the need for the chromium-free steel occurs. The calculated Jominy hardenability band for the Mn-Ni-Mo and for the Mn-Mo steel is represented by the same curves in Figure 2. Similar to the 8620 steel, CH development of chromium-free replacements for a representative 4100 series steel (4118) is given in Tables V and VI. The ladle chemical analysis ranges for the chromium-free replacement steels are summarized in Table VII. The calculated hardenability band for the Mn-Ni-Mo and Mn-Mo chromium-free replacement steels are compared to the standard AISI 4118 steel hardenability band in Figure 3. In the development of the replacement steels, it is assumed that the chromium-free steels will be produced using essentially BOF practice and that the steels will be ingot cast rather than strand cast. If electric steel practice is used, the melt composition can be modified to account for the high residual element content normally present in the scrap charge. For the steel users, the chromium-free steels will have to be provided in a fully-killed fine grain condition with a standard silicon range, since the 8600 and 4100 steels used domestically in carburizing applications are generally specified to be fully-killed fine grain steels. P30-11 TABLE V COMPOSITIONS OF CHROME-FREE REPLACEMENTS FOR AISI 4118 STEEL ITEMS AISI 4118 CH' VERSION SELECTION - ■ STANDARD RANGES .40-.6096 Mn .70-.9096 Mn • 1.00-1.30% Mn 1.20-1.50% Mn -1/4 +1/4 -1/4 + 1/4 -1/4 + 1/4 -1/4 +1/4 -1/4 +1/4 C Mn Si S Cr P Ni Mo .18 .23 .75 .85 .20 .30 .03 .05 .45 .55 .01 .03 .01 .05 .10 .13 .18 .45 .20 .02 .01 .52 .23 .55 .30 .08 .05 .63 .18 .75 .20 .02 .35 .27 .23 .85 .30 .08 .55 .33 .18 .23 1.07 1.23 .20 .30 .02 .08 .01 .05 .17 .23 .18 .23 1.27 1.43 .20 .30 .02 .08 .01 .05 .15 .18 D Ib Min D,, Max DJ Min DJ^Max 1.15 1.76 3.17 4.29 1.05 3.75 1.68 5.25 1.17 3.18 1.83 4.58 1.22 1.87 3.02 4.27 1.37 1.99 3.30 4.19 M s Base 813° 765° 8330 785° 809O 757° 804° 7530 793° 743° M 10 795° 747° 815° 7670 7910 7390 7860 7350 7750 7250 M 50 728° 681° 748° 701° 724° 672° 7190 6690 708° 658° M 90 627° 580O 647° 600O 6 230 5710 6180 5680 607O 5580 M f 426° 378° 4460 3980 4220 370O 417° 366° 406O 3560 M s Case 321° 307O 3470 3340 3240 307O 3170 301O 306O 290O A C 1 1343° 1348° 1336° 13390 13190 13180 13240 1326° 1320O 13220 A c3 15360 15250 1560° 15530 15370 1523° J 1540O 15310 15390 15280 A e 3 1507° 1495° 1516° 1504° 1498° 14810 1501° 1487° 1496° 148 20 Cost Factor $- 1981 $4.90/cwt $12.10/cwt $10.75/cwt $6.10/cwt $5.65/cwt $ - 1985A $5.95/cwt - - $8.00/cwt $7.00/cwt $ - 1985B $7.65/cwt - - $8.95/cwt $7.70/cwt $- 1985C $10.50/cwt - - $9.90/cwt $8.45/cwt $- 1985D $15.10/cwt - - $9.90/cwt . i $8.45/cwt NOTE: D\ values calculated from +1/4 chemistry. P30-12 TABLE VI COMPOSITIONS OF CHROME-FREE REPLACEMENTS FOR AISI 4118 STEEL ITEMS AISI 4118 CH" - REVISIONS 1.00-1.30% Mn Mn-Ni-Mo 1.00-1.30% Mn Mn-Mo -1/4 +1/4 -1/4 +1/4 -1/4 +1/4 C Mn Si S Cr P Ni Mo .18 .23 .75 .85 .20 .30 .03 .05 .45 .55 .01 .03 .01 .05 .10 .13 .16 .21 1.07 1.23 .20 .30 .02 .08 .25 .35 .17 .23 .16 .21 1.07 1.23 .20 .30 .02 .08 .01 .05 .22 .28 Dib Min 1.15 1.26 1.25 Di D Max 1.76 1.85 1.81 Dj c Min 3.17 3.17 3.33 Di c Max 4.29 4.74 4.70 M s Base 8130 7650 8110 7590 8180 768© M 10 795° 747° 7930 7410 800° 750° M 50 728° 6810 727° 6740 7340 6830 M90 627° 580O 626° 574° 633° 5820 M f 4260 378O 4240 3720 431° 381o M s Case 321° 307° 310O 2920 316° 300O A cl I3430 1348° 1310O 13170 13240 13260 A c3 15360 15250 15430 1530O 15520 15410 A e3 1507° 1495° 1500° 14840 1507° 14940 Cost Factor $- 1981 $4.90/cwt $7.50/cwt $7.10/cwt $ - 1985A $5.95/cwt $9.30/cwt $8.85/cwt $ - 1985B $7.65/cwt $10.25/cwt $9.80/cwt $ - 1985C $10.50/cwt $11.30/cwt $10.65/cwt $ - 1985D $15.10/cwt $11.30/cwt $10.65/cwt NOTE: Di values calculated from +1/4 chemistry. P30-13 TABLE VII CHROME-FREE REPLACEMENT COMPOSITIONS FOR STANDARD 4118 AND 8620 STEELS LADLE ANALYSIS RANGES Chemistry Ladle Range, Percent 4100 TYPE STEEL* 8600 TYPE STEEL* AISI-4118 Steel Mn-Ni-Mo Replacement Mn-Mo Replacement AISI-8620 Steel Mn-Ni-Mo Replacement Mn-Mo Replacement Carbon .18 -.23 .16 -.21 .16 -.21 .18 -.23 .16 -.21 .16 -.21 Manganese .70 - .90 1.00 - 1.30 1.00 - 1.30 .70 - .90 1.00 - 1.30 1.00 - 1.30 Chromium .40 - .60 r r A0 - .60 r r Nickel r .20 - A0 r A0 - .70 .40 - .70 r Molybdenum .08 - .15 .15 - .25 .25 - .35 .15 - .25 .25 - .35 .35 - .45 Dib>**,min,in 1.15 1.25 1.25 1.40 1.55 1.50 Di D **,max,in 1.75 1.85 1.80 2.35 2.30 2.20 Di c **,min,in 3.15 3.15 3.30 4.30 4.40 4.45 * The carbon content of the Mn-Ni-Mo and Mn-Mo replacements for the 8620 and 4118 steels is 0.02% lower than the standard grades. It is anticipated that this 0.02 % reduction in carbon content would be maintained throughout the entire replacement series. ** Dj values calculated from chemistry +1/4 range. r = residual level; Si range = .15-.3596; S = .05% max; P = .04% max. CO CO LL1 z Q a < x o 5 o o DC 6 5 5 5 4 5 3 5 2 5 1 5 v 4 118 H I C A L C U 1 ■BAND . A T E D H - B / — i 1 "^ k \ v \ "-•? 1 2 1 6 20 2 4 2 8 3 2 JOMINY DISTANCE 1/16 INCH Figure 3 - Standard Jominy hardenability band for AISI 4118 steel compared to the calculated hardenability band for the chrome-free steels. P30-14 EXPERIMENTAL RESULTS Small, 100 lbs, experimental heats of the chromium-free replacement steels were produced in order to experimentally confirm that hardenability, microstructure, and properties predicted for the chrome-free steels correlate with measured values. The experimental heats were produced at Climax Molybdenum and processed into forged bars or plate samples. Jominy hardenability tests were performed on the base composition, and carburized Jominy hardenability tests were performed to evaluate the high carbon hardenability of the chromium-free steels. Fracture toughness tests on specially prepared high carbon heats of the chromium-free replacement steels were conducted in accordance with procedures outlined in ASTM E-399. Six experimental heats were produced. The chemical analysis of the six heats is shown in Table VIII. TABLE VIII - CHEMICAL ANALYSIS OF EXPERIMENTAL HEATS OF REPLACEMENT STEEL FOR THE AISI 8600 SERIES Chemical SAMPLE IDENTIFICATION 8620 Mn-Ni-Mo Mn-Ni-Mo Mn-Ni-Mo Mn-Mo Mn-Mo Analysis Base Line Lo-Side Hi-Side Hi-Carbon Mid- Hi-Carbon % by Weight Heat Heat /Ml Heat #38 Heat #39 Heat #47 Heat #70 Carbon .21 .19 .23 .91 .21 .93 Manganese .85 .86 .97 1.21 1.12 1.14 Silicon .23 .18 .35 .32 .21 .31 Sulfur .03 .03 .03 .03 .03 .02 Phosphorus .02 .02 .03 .02 .02 .01 Nickel .47 .55 .68 .58 .04 .04 Chrome .46 .03 .11 .05 .05 .05 Molybdenum .20 .21 .26 .34 .30 .40 Aluminum .053 .071 .085 .09 .082 Nitrogen - .009 .015 .012 .019 .011 D Ib, in ch 1.75 1.35 3.05 _ 1.75 _ D Ic , inch 5.15 3.95 7.25 6.35 5.10 6.20 Jominy hardenability test results conducted on the low side heat of the Mn-Ni-Mo steel (heat #41) are shown in Figure 4. The measured curve falls below the standard curve and exhibits lower than the expected results. Investigation revealed that the grain size in this heat (ASTM #9) was finer than anticipated. In general, in hardenability calcula- tions an ASTM grain size of No. 7 is assumed. The measured curve is compared to calculated Jominy curves assuming grain sizes of ASTM #7 and ASTM #9 in Figure 5. Note that the calculated hardenability curve assuming grain size of ASTM #9 falls closer to the experimental curve. P30-15 » CO CO UJ z Q cr < i o 6 5 5 5 3 5 1 r\ _ «a i 1 1 1 1 1 1 1 1 i i 1 1 1 1 <* 1 M C \ » ' ^ § 25 o o a. 1 5 4 8 12 16 20 24 28 32 JOMINY DISTANCE , 1/16 INCH Figure k - Hardenability curve measured for the Mn-Ni-Mo low side heat /Ml compared to the calculated 8620 low side hardenability curve. CO CO UJ z Q cr < X o UJ o o cr 6 5 5 5 45 3 5 2 5 1 5 c a i nn 1 A T ST 1 -inc. 1 1 I I I 1 1 I 1 1 CALCULATED, G3- 1 1 1 HEATI41 MEASili 1 ■ ■ ■■ 1 < < V \ \ 4 8 12 16 20 24 28 32 JOMINY DISTANCE . 1/16 INCH Figure 5 - Comparison of the measured low side hardenability to calculated hardness curves for different grain sizes. P30-16 The Jominy curve for the high side heat of the Mn-Ni-Mo steel is compared to the high side curve for the standard 8620 steel in Figure 6. The chemical analysis of the high side heat exceeds the desired analysis resulting in a higher hardenability and, hence, a higher measured curve. In producing experimental heats, it is very difficult to obtain exactly the analysis desired. Jominy hardenability tests conducted on the mid-band heat of the Mn-Mo chromium- free steel are shown in Figures 7 and 8. This steel also exhibited a finer than expected grain size, and Figure 8 demonstrates that when the calculation is adjusted to include the finer grain size the measured and calculated curves correlate very well. The occurrence of finer grain size in the experimental heats is related to a high aluminum content (.07 - .09 percent) and a high nitrogen content, which provides a larger number of A1N particles to affect the nucleation and growth of austenite. < X o LU o o DC 6 5 5 5 4 5 3 5 2 5 1 5 V 1 8620 HI-SIDE CAL 1 1 1 1 CULATED — 1 1 " \ \ "*•'•.» ~ - . ,, 1 2 1 6 2 2 4 2 8 3 2 JOMINY DISTANCE 1/16 INCH Figure 6 - Comparison of the high side Mn-Ni-Mo heat #38 to the calculated 8620 high side hardenability curve. Carburized Jominy hardenability tests were performed on the three sample heats and results are shown in Figures 9 through 12. Figure 9 shows the 8620 standard steel base line test. All tests were pack carburized in new carburizing compound and heated at 1700° F for eight hours, and direct quenched using a standard Jominy bar quench fixture. The surface carbon content in the chromium-free steels was in general reduced, as compared to the 8620 standard sample. This reduction indicates a reduced tendency to form carbides as a result of the elimination of chromium, and implies that these steels would tolerate a greater variation in carbon potential in the furnace than "the standard chromium containing steels. P30-17 CO CO UJ z Q cr < x o o o DC 6 5 5 5 45 3 5 2 5 1 5 8 6 2 MID- 1 1 1 BAND CALCULATED-- — ■ v 1 1 I H E A T # 4 ' M E A SURE \ t \ 5^ 1 2 1 6 2 2 4 2 8 3 2 JOMINY DISTANCE 1/16 INCH Figure 7 - Comparison of the mid-band Mn-Mo heat #47 to the calculated 8620 mid-band hardenability curve. CO CO LU Q cr < i o LU o o cc 6 5 5 5 45 3 5 2 5 1 5 ^ CALCULATED, GS- 7---- ■ I I I CALCULATED, GS-9-- \ H E A T # 4 7 MEASURED — vV NW X »*: . 1 2 16 2 2 4 2 8 3 2 JOMINY DISTANCE , 1/16 INCH Figure 8 - Comparison of the measured hardenability were for heat #47 to calculated hardness curves for different grain sizes. P30-18 CO CO UJ z a a. < x o LU o o JOMINY DISTANCE 1/16 INCH Figure 9 - High carbon (carburized) hardenability curves developed in the 8620 base line steel. CO CO UJ z o a. < i o UJ O o 6 5 5 5 45 3 5 2 5 1 5 0.95 CARBON-' — i 1 1 0.90 CARBON 0.80 CARBON 1 2 1 6 20 2 4 28 3 2 JOMINY DISTANCE 1/16 INCH Figure 10 - High carbon (carburized) hardenability curves developed in the Mn-Ni-Mo high side heat #38. P30-19 CO to UJ z Q tr < T. O UJ 5 *: o o tr 6 5 5 5 4 5 3 5 2 5 1 5 ^rT^^F 0.80 A.'ISON' —I 1 1 1 h 0.70 CARBON - — ——— — • I I I I I 0.60 CARSON 1 2 1 6 20 2 4 2 8 3 2 JOMINY DISTANCE , 1/16 INCH Figure 1 1 - High carbon (carburized) hardenability curves developed in the Mn-Mo mid-band heat /M7. 6 5 CO CO UJ z Q tr < i o UJ 5 o o tr JOMINY DISTANCE , 1/16 INCH Figure 12 - High carbon (carburized) hardenability curves developed in the low side Mn-Ni-Mo heat /Ml. P30-20 Results of the fracture toughness studies performed on high carbon steels to simulate case microstructure are given in Table IX. High carbon chrome-free replacement steels were poured especially for this program. The data for 8695 steel, included for comparison, was obtained at IH under another in-house program. Ki c values are averages of at least three measurements. X-ray retained austenite (RA) measurements given in the table were made using a rotating and tilting staged) to compensate for the texture in the specimens. TABLE IX PLANE STRAIN FRACTURE TOUGHNESS VALUES FOR HIGH CARBON STEELS Sample %C ———————— Re RA Ki c , Ksi Yin. Mn-Mo Mn-Ni-Mo 8697 0.93 0.91 0.97 60.5 60.5 60.0 22% 23% 23% 16.2 16.1 15.2 Fracture toughness values for all three steels are essentially the same. A slightly lower value for 8697 steel may be attributed to a slightly higher carbon content of 8697 steel. COST BENEFIT OF THE CHROMIUM-FREE STEELS The cost benefit to be derived from the development of chromium-free replacement steels for the 8600 and 4100 standard steels may be estimated as the difference in the costs required to purchase the chromium-containing standard steels vs. the costs required to purchase the chromium-free replacement steels, assuming that a chromium supply interruption has occurred in 1985. o Ml Q < cr o 12.00 10.00 — w 8.00 O o 6.00 < cr •"" 4.00 X Ul 2.00 I I I I I I * - 8 6 2 EXTRA / / • / * - /;,/" - ^^ • «^ „"* / — >*4 1 1 8 EXTRA I I I I I I 19 6 4 19 7 2 19 8 19 8 8 TIME , YEARS Figure 13 - Chemistrys grade extras for AISI 8620 and 4118 steels: 1960 to 1982 P30-21 5 o w o o < DC y- x LU LU Q < tr o 2 4.00 2 0.00 - 16.00 - 12.00 - 8.0 - 4.0 - t r r SCENARIO D \" SCENARIO C !" SCENARIO B SCENARIO A 4. 19 8 1 19 8 3 19 8 5 TIME . YEARS Figure 14 - Predicted effect of the various 1985 chrome supply scenarios on the 8620 steel grade extra. The cost of the two AISI standard grades of steels in terms of their grade extras is shown for the time period of 1960 to 1982 in Figure 13. The effect of the various chromium supply scenarios on the 8620 steel grade is projected to 1985 in Figure 14. The 8620 steel grade extra increases from $12.10/cwt in Scenario A to $22.40 in Scenario D. The steel grade extras for the corresponding Mn-Ni-Mo and Mn-Mo chromium -free replacement steels are shown in Figures 15 and 16 respectively. Both of the replacement steels at a 1981 grade extra are more expensive than the AISI 8620 steel. In 1985 the Mn-Ni-Mo replacement steel becomes cost-effective only under the 2 q 2 4.00 2 0.00 I- O 16.00 o "* 12.00 cc 111 LU Q < a. o 8.0 - 4.00 - SCENARIOS C&D SCENARIO B SCENARIO A X 19 8 1 19 8 3 19 8 5 TIME , YEARS Figure 15 - Predicted effect of the various 1985 chrome supply scenarios on the Mn-Ni-Mo (replacement for 8620) grade extra. P30-22 chromium supply Scenarios C and D. The Mn-Mo replacement for the AISI 8620 at a 1981 grade extra is slightly more expensive ($9.50/cwt vs. $9.60/cwt), but it becomes cost-effective under all scenarios in 1985. Based on the 1978 domestic production of 2,042,176 tons of 8600 type steel, the Mn-Ni-Mo and Mn-Mo steels provide an annual cost advantage of 257.3 and 322.6 million dollars, respectively, in Scenario D. 5 O 2 4.00 I- C/3 o o < cr i- x Ul w Q < cc o 2 0.00 16.00 12.00 8.0 4.0 SCENARIOS C&D SCENARIO 3 SCENARIO A 19 8 1 19 83 19 85 TIME , YEARS Figure 16 - Predicted effect of the various 1985 chrome supply scenarios on the Mn-Mo (replacement for 8620) grade extra. A similar cost analysis for the AISI 4188 and its two chromium-free replacement steels is shown in Figures 17 to 19. Projected 1985 steel grade extra cost for 4118 steel increases from $5.95/cwt, Scenario A, to $15.00/cwt under the conditions of Scenario D. These chrome-free steels, however, are not cost-effective in Scenario A, h- o o < a: t- x in tu Q < cr o 2 4.00 2 0.00 16.00 12.00 8.00 4.00 SCENARIOS C&D SCENARIO 3 SCENARIO A 19 8 1 19 8 3 19 8 5 TIME , YEARS Figure 18 - Predicted effect of the various 1985 chrome supply scenarios on the Mn-Ni-Mo (replacement for 4118) grade extra. 5 o CO o o < cr H X LU ai Q < cr 2 4.00 2 0.00 16.00 - 12.00 - 8.0 4.00 - SCENARIO A J_ 198 1 19 83 19 85 TIME . YEARS Figure 19 - Predicted effect of the various 1985 chrome supply scenarios on the Mn-Mo (replacement for 4118) grade extra. P30-24 The economic advantages of using chrome-free steels under various scenarios is summarized in Table X. The replacement of both 8600 and 4100 types steels with the Mn-Ni-Mo and Mn-Mo steels, under conditions of Scenario D, will result in a combined cost advantage of $455.8 and $556.4 million, respectively. TABLE X - ECONOMIC ADVANTAGE OF CHROME-FREE STEELS Steel Type Grade Extra Cost for Each Scenario, $/cwt 1981 1985A 1985B 1985C 1985D AISI 8620 Steel 9.50 12.10 14.90 17.80 22.20 Mn-Ni-Mo 8600 Replacement 10.85 13.20 15.25 15.90 15.90 Mn-Mo 8600 Replacement 9.60 11.85 13.15 14.30 14.30 AISI 4118 Steel 4.90 5.95 7.65 10.50 15.00 Mn-Ni-Mo 4100 Replacement 7.50 9.20 10.25 11.30 11.30 Mn-Mo 4100 Replacement 7.10 8.80 9.80 10.65 10.65 Annuali zed Cost Adv£ tntage (Pena Ity)* - Millions of Dollars Mn-Ni-Mo 8600 Replacement (55.2) (44.9) (14.3) 159.3 257.3 Mn-Mo 8600 Replacement (4.1) 10.2 71.4 224.6 322.6 Mn-Ni-Mo 4100 Replacement (139.8) (174.7) (139.7) (42.9) 198.5 Mn-Mo 4100 Replacement (118.3) (153.2) (115.6) (8.0) 233.8 Combined Cc >st Advantage (8600 and 41 00 Steels) - Millions of C )ollars Mn-Ni-Mo Replacement (195.0) (219.6) (154.0) 116.4 455.8 Mn-Mo Replacement (122.4) (163.4) (44.2) 216.6 556.4 * Annualized cost advantages based upon 1978 domestic steel production of 2,042,176 tons of 8600 type steels and 2,687,512 tons of 4100 type steels. P30-25 The Mn-Mo replacement steels provide approximately a 22 percent additional cost advantage over the Mn-Ni-Mo steels. This is correct only under the assumption that the molybdenum market will be capable of absorbing the additional increased demand without increasing the molybdenum price above the critical ratio of 2.4- Based upon the 1978 production volume of 4,727,000 tons (combined) of 8600 and 4100 series steels, 17.5 million lbs of molybdenum were required to produce the standard steels. In eliminating chromium to develop the new chromium-free replacement grades, an increase in molybdenum content is required. The manganese-nickel-molybdenum steel grade would require 27 million lbs of molybdenum to produce 4,727,000 tons of steel. This is an increase of approximately 10 million lbs of new molybdenum demand. The manganese-molybdenum steel grade would require 35 million lbs of molybdenum, an increase of 18 million lbs of new demand. The selection of either the manganese- nickel-molybdenum grade or the manganese-molybdenum grade as the most cost- effective replacement for the 8600 and 4100 standard steels is dependent upon the ability of the molybdenum market to accept and accommodate a rapid increase in new domestic demand at the time the chromium supply interruption occurs. If the molybdenum market is capable of accepting 18 million lbs of new demand without raising the price of molybdenum above the critical ratio of 2.4, then the manganese- molybdenum replacement is the most cost-effective. Based upon current evaluation of the molybdenum market, it seems that the molybdenum market will be able to absorb the additional demand and the manganese-molybdenum replacement steel would be the most cost-effective chromium-free replacement steel. CONCLUDING REMARKS Based upon the several socio-economic scenarios and their projected impacts on alloying elements cost and availability in 1985, two chromium-free substitute steels have been designed. Both steels, a manganese-molybdenum substitute and a manganese nickel-molybdenum substitute, are expected to provide microstructure, heat treat response, and mechanical properties equivalent to the 8600 and 4100 steels. The use of chromium at the present time is highly cost-effective and the implementation of the chromium-free steels under the present conditions would not be in the best economic interest of the nation. However, the development of the two chromium-free steels provides materials information which can be saved or stored in a "National Substitution Data Bank" against a future time when an interruption in the chromium supply occurs. Under the worst case scenario - breakdown in South African chromium supply and at the same time, cut-off of Soviet and Albanian sources - the use of chromium-free steels is expected to provide savings of over half a billion dollars annually. It is also possible that the existence of chromium-free substitute steels could moderate future chromium producer supply and pricing policies. Thus, the development of these substitute steels significantly enhances the National "substitution preparedness" with regard to chromium as used in the constructional alloy steels. Obviously, additional testing is required to develop engineering design data and to confirm equivalency of at least one of these chromium-free steels under production conditions prevalent in U.S. industry. Availability of this data will further encourage the ready adoption and widespread use by industry of the new chromium-free steel in the event of a shortage in the supply of chromium. P30-26 ACKNOWLEDGMENT This project was sponsored by the United States Department of the Interior, Bureau of Mines under Contract No. JO 113104. The support of Messrs. 3. T. Dunham and K. W. Mlynarski of the Bureau of Mines, Washington Office, in developing the project is sincerely appreciated. The contract was administered through the Materials Section of the Bureau of Mines, Albany Research Center, under the direction of Mr. H. W. Leavenworth with Mr. Max L. Glenn serving as technical project officer. The assistance of Mr. Glenn in providing technical and administrative assistance is gratefully acknowledged. REFERENCES 1. Morning, 3. L., Mineral Commodity Profiles, Chromium - 1977, Bureau of Mines, U.S. Department of Interior, May 1977. 2. Contingency Plans for Chromium Utilization, NMAB Commision on Sociotechnical Systems, National Research Council, National Academy of Sciences, Washington, D.C., 1978 -NMAB - 335. 3. Breen, D. H., Walter, G. H., Keith, C. 3., and Sponzilli, 3. T., "Computer Based System Selects Optimum Cost Steels," Metal Progress, November 1973. 4. Keith, C. 3., Sponzilli, J. T., Sharma, V. K., and Walter, G. H., "International Harvester's CHAT System for Selecting Optimum Compositions for Heat Treated Steels," Hardenability Concepts with Applications to Steel, edited by Doane, D. V. and Kirkaldy, 3. S., published by AIME, 1978. 5. Ralph, W. L., "Lessons from Nickel," Molybdenum Supplement, American Metal Market, 3uly 1982. 6. U.S. Bureau of Mines, Mineral Commodity Summaries 1980, Washington, D.C., Government Printing Office, 1980. 7. Shin, S. W., and Sharma, V. K., "Application of A Tilting and Rotating Specimen Stage to X-Ray Retained Austenite Measurement in Textured and Coarse Grained Steels," SAE Trans., Paper 800428. P30-27 CHROMIUM-FREE STEELS FOR CARBURIZING D. E. Diesburg, G. T. Eldis and H. N. Lander* INTRODUCTION Despite the fact that most commonly used carburizing steels contain on the order of 0.5 to 1% chromium, the prospect of a restriction or total cut-off of chromium supply is not terribly troubling to the carburizing steel metal- lurgist. There already exist several standard chromium-free carburizing grades. The general methodology of developing new carburizing steels to replace existing grades, substituting more abundant and less expensive alloying elements to reduce overall steel cost while maintaining desired properties, is well developed, thanks to past materials shortages of one kind or another. And especially over the last five years or so, we have gained a great deal of knowledge concerning the effects of all the commonly used elements, indi- vidually and in combination, on the critical properties of carburized com- ponents, so that designing chromium-free steels to meet the requirements of typical carburizing steel applications should pose very little difficulty. Indeed, the only area in the realm of carburizing steels where the prospect of a chromium shortage gives us pause is the relatively specialized application of elevated temperature service. At present, all elevated temperature carbu- rizing steels contain chromium, and the steel properties are heavily dependent on a dispersion of alloy carbide precipitates. However, the authors are con- fident that, if work is begun now, we can be ready to meet any future chromium shortage with a suitable low-chromium or chromium-free steel for this applica- tion, too. *D. E. Diesburg, G. T. Eldis and H. N. Lander are Research Supervisor, Research Manager, and Sr. Vice President, Research and Development, respectively, Climax Molybdenum Company of Michigan, Division of AMAX of Michigan, Inc., Ann Arbor, Michigan. P31-1 In the discussion which follows, we will first look at existing standard carburizing steel grades to get an indication of the economic impact of a chromium shortage. We will then consider the effects of various alloying elements on hardenability and on the mechanical properties critical to carburizing steel performance, to show the directions we are likely to follow in our pursuit of new chromium-free carburizing steels and the properties these steels are likely to possess. Lastly, we will look briefly at the most critical problem for chromium-free carburizing steels, namely, elevated temperature service. EXISTING STANDARD GRADES In any discussion of carburizing steels, we usually find it convenient to classify the steels into one of three core hardenability ranges, low, medium or high. Figure 1 illustrates these three ranges as schematic hardenability bands showing the depth of hardening in an end-quench test to be expected for each range. Table I catalogues various domestic (SAE standard) and European grades in the respective groupings. The low-hardenability grades consist of those steels used for rela- tively small carburized components, specified by the U.S. automotive industry as SAE 4000, 4100, 4400, 4600, 5100, 8600 and 8700. The SAE 8600 and 8700 grades have the highest alloy hardenability and represent the hardenability most often required for such components. SAE 4028 has a hardenability similar to that of SAE 8617; however, a substantial amount of the hardenability comes from the carbon content. This is an important point. Carbon is a very potent element in increasing hardenability, and as such is about the most cost-effective P31-2 "alloying" element. But higher core carbon contents generally lead 1 2 to relatively inferior bend strength properties. ' Thus, the SAE 4028 grade would be used primarily in those applications where specific bending stresses are relatively low. Of the low-hardenability grades, SAE 4000, 4400 and 4600 are chromium-free steels. The other grades contain nominal chromium contents ranging from 0.5 to about 1.0%. In general, the alloy cost of the chromium-free grades is higher. The medium-hardenability steels are represented by grades such as SAE 4300 and 8800, both containing 0.5% Cr. These grades are used when the section sizes being quenched are too large for the low-hardenability steels to maintain an adequate core hardness. At present, there are no standard chromium-free steels that fall into the medium-hardenability classification. The high-hardenability grades include such steels as SAE 4815, 3310 and 9310. The SAE 4800 grade is a chromium-free Ni-Mo steel while the SAE 9310 and 3310 steels are Ni-Cr steels containing 1.2 and 1.6% Cr, respectively. These steels are used to produce large components with tough martensitic cores and also for smaller components expected to encounter service too severe for the low- or medium-hardenability steels. In this respect, these grades are often considered to be high performance grades. Within the various core hardenability ranges, there are several new grades which have been proposed and made commercially available which represent more economical alloy compositions than existing standard grades. These proposed alternates have been assigned SAE EX (or PS) numbers. Table II shows a few examples of EX steels designed by the Climax Molybdenum Co. and compares them with the SAE grades they were designed to replace. All of these alternate P31-3 steels contain higher levels of chromium, manganese and molybdenum than the corresponding SAE grades, illustrating a principal reason for the frequent use of chromium in carburizing steels: At today's availability and price, chromium provides an economical source of alloy hardenability. But what of the technical merits of chromium? Cost factors aside, can we achieve desired properties without this alloying element? Let us consider various attributes of carburizing steels and what we know about the effects of alloy additions. HARDENABILITY We normally subdivide hardenability into two categories. Core hardenability is a measure of the ability to achieve hardness at depth in the low carbon core of the carburized component. It is the hardenability of which we've been speaking in the discussion thus far of hardenability ranges. Case hardenability is a measure of the ability to achieve martensitic microstructures , free of softer pearlitic and bainitic transformation products, in the high carbon case of the carburized component. Figure 2 illustrates the effects of the common alloying elements on core 3 hardenability. Molybdenum and manganese are, individually, of comparable ef- fectiveness, although molybdenum is substantially more effective if the steel *There is no standard grade with a hardenability range similar to EX55. Its hardenability exceeds that of SAE 9310 and SAE 3310, but with a nominal carbon content of 0.17% and a rather low chromium level of 0.55%. P31-4 contains a minimum of about 0.75% Ni. Chromium is somewhat less effective than manganese or molybdenum, and nickel is relatively ineffective by itself. Silicon has no effect on core hardenability at the levels typically found in carburizing steels. If we included a cost factor into these data, turning the abscissa into "alloy cost," the order of the elements would be reversed with manganese being by far the most effective alloying addition, followed by chro- mium, molybdenum and nickel. That is at current prices and availability, of course. From the core hardenability standpoint, elimination of chromium poses no major difficulties. The hardenability decrease resulting from elimination of, say, 1% Cr from the steel could be readily recovered by an increase of 0.6% in the combined Mn + Mo content. The ratio of manganese to molybdenum selected for such a substitution would depend upon several factors, among them the overall level of those elements already present in the steel, since ex- cessive use of any one element can aggravate the problems of alloy segregation during solidification. Other factors to be considered will be discussed later. Figure 3 shows the influence of molybdenum, manganese, chromium and nickel on case hardenability, in the carburized areas of the component where carbon 4 content is usually on the order of 0.8 to 0.9%. The relative effectiveness of the elements is quite different than for core hardenability. In addition, molybdenum and nickel share a common attribute which makes them, relative to manganese and chromium, even more effective than Figure 3 implies. These elements are immune to depletion from the matrix by the mechanism of surface oxidation. During normal carburizing in endothermic atmospheres, the oxygen potential is high enough to induce oxidation of chromium, manganese and silicon. These elements are depleted from the matrix at the surface of the P31-5 specimen, as illustrated in Figure A which shows the results of electron probe 6 microanalysis on a 1. l%Mn - l.l%Cr steel carburized in an endothermic atmosphere. Figure 5 shows the microstructural effects of this surface oxidation, a de- crease in hardenability at the surface and transformation to pearlite and bainite rather than martensite. Depending on depth of oxidation/depletion, size and quenching rate of the component, subsequent surface treatment (lapping, shot peening) and in-service loading experienced by the component, these non- martensitic transformation products at the surface could ultimately prove detri- mental to performance. Turning back to Figure 3 and our example of removing 1% Cr from the steel composition, we could maintain the same level of case hardenability by simply adding 1% Mn. (Note that this is substantially more than the 0.6% Mn required to maintain a constant level of core hardenability, as discussed before.) However, given that the steel will most probably already contain at least 0.5% Mn, a full manganese-for-chromium substitution would not be a good choice. It would result in an "excessive" level of manganese (1.5% or more), and would only aggravate the surface oxidation problem (Figure 4). A more suitable sub- stitution might be, say, 0.4%Mn + 0.2%Mo, which would maintain both the case and core hardenability lost by the removal of 1% Cr. Regardless of the exact combination of elements chosen to substitute for the chromium, it is quite clear that the substitution can be made without experiencing any reduction of steel performance from the hardenability standpoint. P31-6 MECHANICAL PROPERTIES There are two rather independent mechanical properties of carburizing steels that determine their overall performance: (1) the ability to resist pitting and (2) the ability to resist bending. In most applications, car- burized components are overdesigned with respect to bending fatigue simply because complete fracture is a less tolerable situation than is surface degradation such as pitting, although severe pitting also can eventually lead to complete fracture. Pitting Fatigue Factors influencing pitting fatigue are more difficult to evaluate experimentally than are the factors influencing bending fatigue. Although it is generally believed that pitting behavior is dependent on hardness and microstructure, at times it has been difficult to confirm this experimentally. ' 9 Recent research has shown pitting to be associated with inclusions. Pitting resistance has also been shown to be dependent on the lubricant. A proper lubricant can maintain a film between the contact surfaces and significantly improve the pitting fatigue lives of carburized components. Soft microstruc- tures are generally accepted as having poor resistance to pitting, although hard microstructures can also exhibit pitting. Surface decarburization, the presence of a large amount of retained austenite, and the transformation of austenite to pearlite or bainite during quenching can cause soft regions to occur in a carburized case and thus are believed to lead to an eventual pitting failure. Figure 6 shows pitting damage as revealed in the scanning electron microscope. P31-7 All common alloying elements used in carburizing steels, namely chromium, manganese, nickel and molybdenum, are added to obtain the required hardenability . In this respect, the elements are interchangeable as already discussed. However, there are independent limiting factors that determine the maximum amount of each element that should be added, in addition to the segregation and surface oxidation problems already mentioned, which could affect pitting resistance. Nickel and manganese are both austenite stabilizers and therefore must be used in moderation if retained austenite in the carburized case is expected to be a problem. Chromium also has a carbide formation limitation. High chromium contents are believed to promote the formation of large carbides in the case during carburizing, thus steels such as SAE 3310 and SAE 9310 generally require extra care during carburizing and heat treatment. Reducing or eliminating the chromium contents in carburizing steels is not expected to have a strong effect on the pitting resistance, provided the corresponding loss in hardenability is regained by adding manganese, nickel or molybdenum and retained austenite content remains under control. In support of the contention that chromium-free steels will offer adequate pitting resis- tance compared to chromium-containing steels, one can look at the current successful use of existing grades of chromium-free steels such as SAE 4000, 4400, 4600 and 4800. There are no reported indications that these grades are any more prone to pitting than the chromium-containing steels. Bending Fatigue and Impact Bending fatigue properties of carburized steels are highly dependent on the compressive residual stress that develops at the carburized surface during P31-8 11 12 quenching. This is illustrated nicely by the improvement in high cycle 13 fatigue limit resulting from shot-peening, shown in Figure 7. Shot-peening prior to carburizing does not provide a similar improvement in fatigue behavior 14 (Figure 8) . Altering the residual stress at the surface so as to make it less compressive, as by improper heat treatment or other processing, can have a dramatic negative effect on fatigue behavior. This is illustrated in Fig- 13 ure 9 by the data for EX55 subjected to a carbonitriding treatment. The particular cycle employed resulted in a very high (70%) retained austenite content at the specimen surface. Despite this high austenite content, the bulk residual stress at the surface was compressive and the fatigue proper- ties quite good. A refrigeration treatment reduced the austenite content to 40%, but also resulted in a net bulk tensile residual stress at the surface. The effect on fatigue properties was disastrous. The only expected effect of chromium per se on constant load amplitude high-cycle fatigue behavior would be the result of chromium depletion from the surface, either by severe oxidation or by massive carbide formation. This could allow pearlite or bainite to form at the surface, thus causing the surface to have less compressive residual stress, perhaps even a tensile stress. Figure 10 illustrates the theoretically calculated effect of compressive re- sidual stress on fatigue limit for various assumed degrees of surface oxida- 16 tion. Here, the oxidation product is assumed to act as an internal defect (stress concentrator) which intrinsically reduces fatigue limit. The im- portance of maintaining a substantial compressive residual stress at the sur- face is quite clear. Reducing or eliminating chromium in carburizing steels is thus expected to have no detrimental effect on high-cycle fatigue behavior, P31-9 provided the lost case hardenability is replaced with other alloying elements to ensure the surface microstructure remains martensitic. Indeed, to the extent that chromium-free alternate steels might be less susceptible to surface oxi- dation effects, an improvement in properties might be expected. The apparent sole dependence of bending fatigue performance (fatigue limit) on the surface residual stress is true when the applied loads are rather low, with maximum loading in the vicinity of the laboratory-measured fatigue limit. However, most carburized parts in service are subjected to a wide spectrum of loading, and the infrequent occurrence of loads well in excess of the fatigue limit must be anticipated. Such overloading can result in material damage, so that the effective fatigue limit after the overload is reduced. The magnitude of overload that can be tolerated before signifi- cant damage occurs can simply be referred to as "toughness," and is_ highly dependent on alloy content as well as the fracture toughness and residual 14 stress further away from the carburized surface. The relative ability of the steels to resist overloading can be evaluated by measuring the impact fracture strength of the carburized component, that is, the load required to fracture the component in a single blow. Climax has been evaluating the effect of various alloy combinations on the impact fracture strength for several years and has published some generalizations that can be used to assess the contribu- tion of chromium to bending fatigue life when random high and low loadings are i 18 encountered. For low-hardenability steels, Figure 11 indicates that chromium generally has a negative effect on impact fracture strength over the range of 0.2 to 1.2%. Thus, no detrimental effect on impact fracture strength or "toughness" is P31-10 expected if chromium is eliminated from carburizing steels, provided adequate alloy substitution is made to maintain hardenability . In the medium-hardenability range, the steel compositions generally in- clude various nickel-chromium-molybdenum combinations. The influence of chromium on impact fracture strength is dependent on the nickel content of the steel. Combined with nickel contents of 1% or less, chromium shows a negative effect on impact fracture strength similar to that experienced in the low-hardenability steels. In combination with nickel contents exceeding 1.0%, chromium has no influence. So again, using chromium-free steels for medium-hardenability should have no detrimental effect on "toughness." All high-hardenability steels contain more than 1.5% Ni, and as might be expected from the trends in low- and medium-hardenability steels of various chromium and nickel contents, chromium content has no effect on impact fracture strength in the high hardenability steels. The excellent performance of SAE 4800 steels provides ample evidence of the toughness of chromium-free high-hardenability steels. EX55 contains 0.55% Cr, and it has exhibited even better performance than SAE 4817 (other factors, such as the higher hardenability of EX55, are partly responsible for this). A chromium-free version of the EX55 composition, containing 2% Ni and 1% Mo, has been evaluated in the impact frac- ture test and has exhibited similar behavior to that of EX55. Therefore, in all three hardenability ranges, chromium-free steels are expected to have a toughness equal to or better than currently used chromium-containing steels. P31-11 ELEVATED TEMPERATURE SERVICE Special steel compositions are available for carburized components to be used at temperatures up to 315 C (600 F) . Table III gives some examples. Such steels are required to retain adequate case hardness and toughness at these ele- vated service temperatures . Typical requirements might include a minimum case hardness of 58 HRC after 1000 hours exposure at 315 C (600 F) , along with a minimum fracture toughness specification. To a large extent, these properties are dependent upon precipitation of various alloy carbides during tempering at temperatures above the intended service temperature, to form a strengthening dispersion stable at the operating temperature with a morphology not overly detrimental to fracture toughness. A variety of alloy carbides have been identified after heat treatment of such materials, including M5C in the lower carbon core regions of the samples, and M23C5, M2C and MC in the carburized case, depending on the exact 19 steel composition and heat treatment temperatures used. The extent to which chromium enters into these various carbides is not well defined, and its neces- sity from the standpoint of precipitation kinetics and precipitate stability is even less well defined. Thus, at present, we can't be certain how success- ful we would be in meeting the requirements of elevated temperature service with a chromium-free composition. The authors are optimistic, however. From 20 the standpoint of "carburizability , " the lower the chromium content the better. Carbide precipitation and secondary hardening certainly do occur in vanadium, 21 molybdenum and tungsten alloyed steels without chromium additions, and such precipitation is the major prerequisite for good elevated temperature proper- ties. The comparably good performance of X-53 and X-2M, with 1 and 5% chromium, P31-12 respectively, also bodes well for the development, if eventually necessary, of chromium-free compositions. CBS 1000, with only 1% Cr, has performed even better in Climax laboratory tests than X-53 and X-2M, and X-ray diffraction studies have shown it contains no M23C6, the chromium-rich carbide. Indeed, the main difficulty in developing chromium-free steels for elevated temperature service may have nothing to do with service performance at all, but rather be related to slight atmospheric corrosion of components in inventory. Rather remarkably, it was found in developing the CBS 1000 steel that nominally 1% Cr was the minimum amount necessary to avoid problems of component degradation 19 in storage. SUMMARY Although chromium is a common alloy addition to carburizing steels, all the evidence indicates it is not necessary for performance in typical (ambient temperature) carburizing steel applications. Chromium is currently added to low-, medium- and high-hardenability steels for economic reasons. Chromium- free steels are expected to perform equally well and, in many situations, may even provide better performance than currently used chromium-alloyed grades. Additional work is necessary to determine the full potential of chromium-f ree carburizing steels to meet the requirements of elevated temperature service, but the prospects appear promising. P31-13 REFERENCES 1. V. S. Sagaradze, "Effect of Carbon Content on the Strength of Carburized Steel," Metal Science and Heat Treatment, March 1970, No. 3, pp. 198-200. 2. C. Kim and D. E. Diesburg, "Fracture of Case-Hardened Steel in Bending," submitted for publication in the Journal of Engineering Fracture Mechanics. 3. C. A. Siebert, D. V. Doane and D. H. Breen, The Hardenability of Steels , American Society for Metals, Metals Park, Ohio (1977), p. 101. 4. C. F. Jatczak, "Hardenability in High-Carbon Steels," Met. Trans., 4 (1973), pp. 2267-2277. 5. I. Kirman, G. Mayer and F. W. Strassburg, "Einfluss von Nickel and Randge fiige auf die Brucheigenschaf ten einsatzgeharteter Stahle," Harterei-Techn. Mitt. 29 (1974), pp. 88-94. 6. Y. E. Smith and G. T. Eldis, "New Developments in Carburizing Steels," Met. Engr. Quarterly, 16, No. 2 (1976), pp. 13-20. 7. J. S. Learman and G. T. Eldis, "Effect of Bainite in the Outer Carburized Case on Rolling Contact Fatigue Life," SAE Paper No. 760665. 8. S. L. Rice, "Pitting Resistance of Some High Temperature Carburized Cases," SAE Paper No. 780773. 9. D. H. Breen, formerly with International Harvester, currently with the Gear Research Institute, unpublished research. 10. J. P. Sheehan and M. A. H. Howes, "The Role of Surface Finish in Pitting Fatigue of Carburized Steel," SAE Paper No. 730580. 11. D. A. Sveshnikov, I. V. Kudryartsev, N. A. Gulyaeva and L. D. Golubovskaya, Chemicothermal Treatment of Gears," Metal Science and Heat Treatment, No. 7, July 1966, pp. 527-532. 12. Chongmin Kim, D. E. Diesburg and G. T. Eldis, "Effect of Residual Stress on Fatigue Fracture of Case-Hardened Steels," ASTM STP 776 (1981) pp. 224-234. 13. Chongmin Kim, D. E. Diesburg and R. M. Buck, "Influence of Sub-Zero and Shot-Peening Treatments on Impact and Fatigue Fracture Properties of Case-Hardened Steels," Journal of Heat Treating (3), 1_ (1980), pp. 3-13. P31-14 REFERENCES (Continued) 14. R. W. Buenneke, C. R. Dunham, M. P. Semenek, M. M. Shea, M. B. Slane and J. E. Tripp, "Gear Single Tooth Bending Fatigue Test," SAE Paper No. 821042. 15. B. Hildenwall and T. Ericsson, "Residual Stresses in the Soft Pearlitic Layer of Carburized Steel," Journal of Heat Treating (3), 1_ (1980), PP. 3-13. 16. D. E. Diesburg, C. Kim and W. Bulla, "Impact and Fatigue Fracture of Carburized Cases Related to Fracture Toughness and Residual Stress," Presented at Heat Treaters Conference, Wiesbaden, West Germany, 1981, proceedings to be published. 17. T. B. Cameron and D. E. Diesburg, "The Significance of Impact Fracture Strength of a Carburized Steel," to be presented at the Spring Meeting of TMS-AIME in Atlanta, March 6, 1983 (to be published). 18. D. E. Diesburg and Y. E. Smith, "Fracture Resistance in Carburizing Steels Part II: Impact Fracture," Metal Progress, 115 (6), June 1979, pp. 35-39. 19. C. F. Jatczak, The Timken Co., private communication. 20. C. F. Jatczak, "Specialty Carburizing Steels for Elevated Temperature Service," Metal Progress 113_, No. 4 (1978), pp. 70-78. 21. G. A. Roberts, J. C. Hamaker and A. R. Johnson, Tool Steels , American Society for Metals, Metals Park, Ohio (1962), pp. 215-218. P31-15 Table I COMPOSITION, HARDENABILITY AND COST OF COMMERCIALLY AVAILABLE STANDARD CARBURIZING STEELS Nominal Composition, Wt-% '•■•I »» ■ r« Hardenability," Steel' C Mn Cr Ni Mo D|, mm $/ton c Low Hardenability 4028 0.28 0.80 — — 0.25 40 157 4118 0.20 0.80 0.50 — 0.10 36 98 4422 0.22 0.80 — — 0.40 39 184 4617 0.17 0.55 — 1.80 0.25 37 323 5120 0.20 0.80 0.80 — — 41 55 8620 0.20 0.80 0.50 0.55 0.20 48 189 8720 0.20 0.80 0.50 0.55 0.25 50 208 *18CrMo4 0.18 0.65 1.05 — 0.25 53 163 *20NiCrMo2 0.20 0.75 0.50 0.55 0.20 45 189 *20MoCr4 0.20 0.75 0.40 — 0.45 48 253 *16MnCr5 0.16 1.15 0.95 — — 52 97 Medium Hardenability 4320 0.20 0.55 0.50 1.80 0.25 60 327 8822 0.22 0.85 0.50 0.55 0.35 62 252 High Hardenability 4815 0.15 0.50 — 3.50 0.25 66 512 3310 0.10 0.55 1.60 3.50 - >127 499 d 9310 0.10 0.55 1.20 3.25 0.10 89 530" *14NiCrMo13 0.14 0.45 0.95 3.25 0.25 99 559 M7CrNiMo6 0.17 0.50 1.65 1.50 0.30 >127 431 *20NiMoCr6 0.20 0.75 0.40 1.60 0.45 87 436 a Asterisk denotes European standard. Others are SAE grades. "Calculated from composition assuming ASTM grain size No. 7. c Grade extra charge based on U.S. industry pricing. d Does not include $25/ton electric furnace extra, which is required. P31-16 Table II COMPOSITION, HARDENABILITY AND COST OF SEVERAL SAE GRADES AND THEIR EX ALTERNATES Nominal Composition, Wt-% Hardenability," Steel C Mn Ni Cr Mo D|, mm $/ton SAE 8620 0.20 0.80 0.55 0.50 0.20 48 189 EX 24 0.20 0.87 — 0.55 0.25 48 154 SAE 4320 0.20 0.55 1.85 0.50 0.25 60 327 EX 29 0.20 0.87 0.55 0.55 0.35 63 256 SAE 4815 0.15 0.50 3.50 0.25 66 512 EX 30 0.15 0.80 0.85 0.55 0.52 80 357 EX 55 0.17 0.87 1.80 0.55 0.75 >127 528 Calculated from composition assuming ASTM grain size No. 7. b Grade extra charge based on U.S. industry pricing. P31-17 Table III CARBURIZING STEELS FOR ELEVATED TEMPERATURE SERVICE Nominal Composition, Wt-% Steel C Mn Cr Ni Mo W V Cu 0.3 2.0 1.4 0.4 3S 1000 0.14 0.5 1.15 2.9 4.7 X-53 0.18 0.4 1.05 2.1 3.3 X-2M 0.15 0.3 4.90 — 1.4 P31-18 CM <2 31V0S SS3NQHVH m3M>IOOd (U u o CO C o CU ,-J X) J-l u-l cO O EC o c 3 to I c TJ -H C w U CO B a C/) (U i-i 3 00 •H P31-19 Mo(>0.75Ni) 0.4 0.6 0.8 1.0 ALLOYING ELEMENT (%) Figure 2 Effect of Various Elements on Hardenability at 0.2% Carbon Content (Reference 3) P31-20 0.25 0.50 0.75 1.00 1.25 1.50 PERCENT ELEMENT 1.75 2.00 Figure 3 Effect of Various Elements on Hardenability at 0.85% Carbon Content (Reference 4) P31-21 1.2 CASE DEPTH, mm 0.02 0.04 0.06 1.0- o CL H 0.8 h z UJ h- o 0.6- o >- o -I 0.4h 0.2 — i — i — i — r NOMINAL Cr.M( 0.001 0.002 0.003 CASE DEPTH, in. Figure 4 Chromium and Manganese Concentration Profiles Developed in the Specimen Case During Carburizing of a 1.1 %Mn - l.l%Cr Steel (Reference 6) P31-22 .CARBURIZED SURFACE Figure 5 Scanning Electron Micrograph of Carburized Case in 1.1% Mil- 1.1% Cr Carburizing Steel P31-23 TOOTH ROOT LINE OF INITIAL ^MATING CONTACT PITTING ZONE Figure 6 Scanning Electron Micrographs of the Heavily Pitted Region on the Compression Face of a Driven Gear P31-24 OISTANCE FROM SURFACE , in. 0.01 0.02 0.03 0.04 0.05 o a. CO co lj a: CO -J < o CO u AC -200 -400 i 1 t 1 r CARBURIZED SAE 4028 -0 io^T UNPEENED -600 PEENEO 1 1 1 10 -20 co* CO Ul a: -40 H to -60 < CO uj -80 * 0.2 0.4 0.6 0.8 1.0 1.2 DISTANCE FROM SURFACE, mm 2000- o CL CO 2 iooo oc co I I I I I I t 1 i i I 1—— i — r-r] 1 i ■ i | CARBURIZED SAE 4028 PEENED 10 -300 200 ■ ■ ' I I I I I I L J-J-L 10' 10* CYCLES 10* '@*1 1 1 CO CO 111 a: 100 co 10 Figure 7 Residual Stress Profiles and Stress-Life Curves for Carburized SAE 4028, Before and After Shot-Peening (Reference 13) P31-25 qi c oi 'avcn CJ 00 CD I I 1 1 Cl& ™ ^^ ■ - z - _ UJ _ - UJ - - Q CD m LJ ■ Z a - UJ UJ •k UJ z 7 CL UJ ZJ — - Z UJ - 3 CL MZ O m o • y^^fc ^^ - o y» ° - - O «K o - - ^J* _ - • - - 6/ ^ - "my ' • ~ o - I _j ' • • <0 O lO O CO UJ -J o > o (0 m ic0i 'avcn CM 2 00 c •H a ^-s O CI O -H CNI N 00 u 3 a) »-t n co •H CJ u 3 o U cO u o •H O P-. cfl C •U CO CO £ Q 4J cO 0) CO <4-l (-1 •H H I CO O 00 0) 00 P31-26 DISTANCE FROM SURFACE, in. O.Ol 0.02 0.03 0.04 0.05 200 I 1 r i i ■ n CARBONITRIDED SAE EX55 20 0. - 5> 2 ^-TREATED AT j« CO CO -85C(-120F) P * (0 (0 UJ or 1- CO y* -20 UJ o: H CO -1 -200 - \ ^^^^^ J < < o -NO SUBZERO -40 o CO TREATMENT CO UJ UJ or -400 _ i I 1 I 1 1 -60 or 0.2 0.4 0.6 0.8 1.0 1.2 DISTANCE FROM SURFACE, mm 2000- o 0. CO CO uj 1000 or H CO i i I i 1 \ i i 1 1 i i i 1 r CARBONITRIDED SAE EX55 T TREATED AT -85C(-120FV 1111 I ■ I ' I L NO SUBZERO .-TREATMENT 10' 10 10 CYCLES 10 6 V8/_ -»> J_Ll 300 200 r CO CO UJ or 100 co 10 Figure 9 Effect of Refrigeration Treatment on Residual Stress and Stress-Life Behavior of Carbonitrided EX55 (Reference 13) P31-27 DEPTH OF SURFACE OXIDATION, 10" 4 in. M O en" -(0.5% Ni) i i~-«* 600 500 400 CO •» 0) 0) o 0) 4J e> w x: o c <0 '0 ** *•»» E e E u -««: E >J 5 to E E C 0»H fj CO CM CM Wl VH > r-» ov V© 00 vo r» r* n3-H (C • • • • • • • u a J o o o o o o o u m in in o • • • o -u O CM i o\ »H CM en i VC 1 Sm c ^r ^ en **• V v I ^ 1 o Vj u V5 a C 0) 'S o VI U (S m m cr «3»M • • -H 9= J-i r-» O I CM ^r IN o I t*» 1 cc 3 to VO NO vo VO VO VO I in 1 4J to 1 ^ S Cm !S mm Cm 2 b 0) M «•* E-« U O c c: o n c c O c o •^ 4J 5 J ^ o u 5 c 'S E 5 O E •M j- c O.Q c^o OX1 o 0£ O I 0£ O 1 o a c a ►J u J J-i CO »J.H (0 ►J VJ CD iJ »-i to c v- u n rj •«c ••^ 10 13 ir. H 3 u U + u U 4 u + +J C^4-» .H U 3 rH 3 O + + + 4J f) r-i M O E CJ IB4J fl n:< Srf mm «Mi M S < o: 4J CO «M a Q) 0) 0> t-i O Q l a) a ti E E i a o < 1 £ 3 G £ c e <0 IC IS C Cn CO H •H o (0 to to •H M u (0 b (0 *^ *^ «^ Cu fl • •J a> O M 4-» 3 4J VO VO in VO U) o ^J" o m . OV vV U -* •h r- CM (N o o VO CM r»- c: E ^ U vo r- 00 vO r» o 00 «-l i—» m •ri U SI 1-« t-l tH «H H tM'"' 1 E~ M H ox: O \w- JJ lO f» m m. m M> m ON CTV m a: tu co m vo en en »-l m r» CM CM ■^r ^r e a-H •H M JJ 6" U «u 1 •H o J-l E-» CQ s o o r> o o 6 O n o o o o 6 o a o id « ri 65 — —■ $i s \ 60 vv — ^G \ III _| 65 < o U) 50 j? N i -C r-Mo 50 co 1 1 ' a '6/7-H 4-0.20C ! ' ^ en 5" 1 O.I ! ! 40 \ CO 1 1 1 1 r V«-i i i i ! N:I700 30 v f^- | A.I700 >j>J . *?. CNJ_ ~\ 1 ! 1 1 M PO _ 1 |^r<^U> Mill C-Mn-Cr li-Cr-Mo 00 111 CD S_ r3 CD u o 60 50 40 30 20 L5/40-H J. 0.37- 0.44 C , 0.60-/. 00 Mn, 0.60-1.00 Cr_ - C-Mn-Cr 60 50 40 30 20 U_„ — ' — «-— I' "' —•■ — , >J i !?" N-) *~> ' J M CN, M i ! ! 1 1 1 1 ! T~~ C-Mn-B 50 40 30 20 *i 1 , 1 i i 1 I^R77 H i 1 I 30-0.39 C 1 00- 1.50 Mn ir^^ i ; i |s!>? i N j i i l ! ! i i \L' 1 V 5 " i i 1 M 1600 3 1550 1 ! ! V ! i i\ I i ! j i iV i I s Vcvi f-. i : i 1 IN? t. 1 o 4 n ! t i i i ivji | 1 i ^ CO ' i i ! !..! M C\J ^ K -CO i i j | i N:I6 00 ' A. 1550 CV'_ c\j _c\i -t\j.C\j 1 C-Mn-Mo i ! 42-H t— • — : — r- 0.60-1.00Mn, _ 0,20-0 .30Mo 1600 — i — I — i — r S S ^ k ^ <\1 (\J c\i cv °\J C-Mn-Cr 20 24 28 32 2 4 8 12 16 20 24 28 32 Distance From Quenched End of Specimen, Sixteenths These few examples should suffice to illustrate how existing standard AISI grades of steel can be substituted for other existing grades in the event of a raw materials shortage. P35-3 - 4 - Alloy Substitution Capabilities Via the SAE EX Steels As mentioned previously, the SAE series of EX steels represent the SAE numbering system for new grades of alloy steels. As a function of time the EX steel either is dropped because of lack of interest in the alloy or it becomes a standard SAE grade if the alloy becomes popular and widely used. The currently available EX steels and the standard SAE grade for which they are normally substituted are shown in the list. THE EX STEELS AND EQUIVALENT STANDARD GRADES Composition {%) EX Equivalent No. C Mn Cr Mo Other SAE Grade 10 0.19-0.24 0.95-1.25 0.25-0.40 0.05-0.10 .20-0.40 Ni 8620 15 0.18-0.23 0.90-1.20- 0.40-0.60 0.13-0.20 - 8620 16 0.20-0.25 0.90-1.20 0.40-0.60 0.13-0.20 - 8622 17 0.23-0.28 0.90-1.20 0.40-0.60 0.13-0.20 - 8625 18 0.25-0.30 0.90-1.20 0.40-0.60 0.13-0.20 - 8627 19 0.18-0.23 0.90-1.20 0.40-0.60 0.08-0.15 .0005 B min 94B17 20 0.13-0.18 0.90-1.20 0.40-0.60 0.13-0.20 _ 8615 21 0.15-0.20 0.90-1.20 0.40-0.60 0.13-0.20 - 8617 24 0.18-0.23 0.75-1.00 0.45-0.65 0.20-0.30 - 8620 30 0.13-0.18 0.70-0.90 0.45-0.65 0.45-0.60 .70-1.00 Ni 4815 31 0.15-0.20 0.70-0.90 0.45-0.65 0.45-0.60 .70-1.00 Ni 4817 32 0.18-0.23 0.70-0.90 0.45-0.65 0.45-0.60 .70-1.00 Ni 4820 33 0.17-0.24 0.85-1.25 0.20 min 0.05 min .20 Ni min 4027 34 0.28-0.33 0.90-1.20 0.40-0.60 0.13-0.20 - 8630 36 0.38-0.43 0.90-1.20 0.45-0.65 0.13-0.20 - 8640 38 0.43-0.48 0.90-1.20 0.45-0.65 0.13-0.20 - 8645 39 0.48-0.53 0.90-1.20 0.45-0.65 0.13-0.20 - 8650 40 0.51-0.59 0.90-1.20 0.45-0.65 0.13-0.20 - 8655 54 0.19-0.25 0.70-1.05 0.40-0.70 0.05 min _ 4118 55 0.15-0.20 0.70-1.00 0.45-0.65 0.65-0.80 1 .65-2.00 Ni 4817 56 0.08-0.13 0.70-1.00 0.45-0.65 0.65-0.80 1 .65-2.00 Ni 9310 57 0.08 max 1 .25 max 17-19 1.75-2.25 - 30303 58 0.16-0.21 1.00-1.30 0.45-0.65 _ _ 4118 59 0.18-0.23 1.00-1.30 0.70-0.90 - - 8620 60 0.20-0.25 1.00-1.30 0.70-0.90 _ _ 8622 61 0.23-0.28 1.00-1.30 0.70-0.90 - - 8625 62 0.25-0.30 1.00-1.30 0.70-0.90 - - 8627 63 0.31-0.38 0.75-1.10 0.45-0.65 _ .0005-0.003 B _ 64 0.16-0.21 1.00-1.30 0.70-0.90 - _ _ 65 0.21-0.26 1.00-1.30 0.70-0.90 - - _ P35-4 - 5 - Currently there are about 30 different EX steels which can be substituted for a large variety of standard SAE grades. Generally the use of an EX steel results in a reduction in either the Ni content or the Mo content compared to the standard SAE grade. As we have experienced both Ni and Mo shortages in the past 10 to 12 years the utilization of the SAE EX steels has been both industrially and technologically significant. Alloys "On the Shelf" as a Result of Recent Research Recent Republic Research efforts to develop bar products with reduced levels of critical alloying elements while maintaining the required heat treatment response and mechanical properties have resulted in a number of "on the shelf" alloys ready to go as demand arises. The table below sum- marizes some of these developments. NEWLY DEVELOPED PRODUCTS WHICH ARE EXAMPLES OF MATERIALS SUBSTITUTION Product Area Old Grade 3310 New Grade 3120 AT loy Savings Al Toy Bar Reduce Ni from 3.50 to 1.75% Modified Reduce Cr from 1.50 to 0.60% Alloy Bar 4815 3120 Reduce Ni from 3.50 to 1.75% Modified Reduce Mo from 0.25 to 0% Al loy Bar 4620 Modified 3120 Modified Reduce Mo from 0.50 to 0% Alloy Bar 4815 "2Ni-lMo" Reduce Ni from 3.50 to 1.75% Alloy Bar EX-55 "2Ni-lMo" Reduce Cr from 0.55 to 0% Alloy Bar 4118 10B18 + V Reduce Mo from 0.11 to 0% Reduce Cr from 0.50 to 0% Alloy Bar 4027 C-Mn-Cr Reduce Mo from 0.25 to 0% Ultra-High- HP 9Ni-4Co "lONi-lMo" Reduce Co from 4.5 to ( 3% Strength Alloy Steel The first alloy shown (3120 Modified) is actually now off the shelf and in use in automotive steering gear systems. A second alloy "2 Ni - 2 Mo" is currently being evaluated for use in rock-drilling bits and would reduce Cr content from 0.55% to 0% compared to the currently used EX-55 steel. In the vacuum-melted specialty steel area "10 Ni - 1 Mo" steel is under development as a replacement for the HP 9 Ni - 4 Co steels in the event of a cobalt shortage. P35-5 - 6 - Computer Prediction Capability as an Alloy Development Tool In the past several years various hardenability prediction models have been developed. The most accurate computer prediction model is the Mini- tech hardenability and mechanical property predictor. This prediction system provides a rapid and highly accurate means of predicting harden- ability, heat treatment response, and mechanical properties as a function of steel composition. This therefore provides a rapid and reliable means of designing alternate bar product alloys based on current alloy costs and availability without going through a lengthy program of melting and eval- uating experimental steel compositions. Examples of the type of information that the Minitech predictor can pro- vide and the accuracy of the calculations are shown below. The first and most important information desired is the core hardenability of the steel. Shown in Figure 1 are comparisons of the experimental and Minitech pre- dicted hardenability curves for a given composition of 8620 steel. The excellent agreement between experimental and predicted results in apparent, For carburizing grade steels such as 8620 the next piece of information an alloy developer needs is the case hardenability or carburizing response of the alloy. Shown below in Figure 2 is an example of the ability of the Minitech system to calculate or predict the case hardenability for a given set of heat treatment conditions and steel composition. Again, excellent agreement is observed between experimental and predicted results. P35-6 to O) c ra - 7 - Grade: 8620 Mn Si Ni Cr Mo Cu AT 0.21 0.83 0.008 0.020 0.31 0.52 0.51 0.16 0.17 0.038 70 \ 60 50 | 40 O A Experimental Predicted (Minitech) e 30 20 10 e L 6 a s 8 12 16 20 Distance From Quenched End (1/16") 24 28 Figure 1. Comparisons of the Experimental and Minitech Predicted Hardenability Behavior for 8620 Steel. P35-7 - 8 - 70 60 o 50 on 00 00 O) c: -a s- 40 O Experimental A Predicted (Minitech) a s o A O e__A-~0 A 2 6 A O A ' O A KJ Q g Grade: 8620 30 20 10 C Mn Si Ni Cr Mo Cu Al 0.18 0.82 0.023 0.021 0.31 0.44 0.57 0.17 0.09 0.03 Carbon Potential = 0.90% Carburized 1700 F - 2.5 Hrs, Cooled in 1 Hr to 1575 F, Held at 1575 F - 0.5 Hr, Oil Quench, Temper 300 F - 1.5 Hrs 0.010 0.020 0.030 0.040 Case Depth (inch) 0.050 0.060 0.070 Figure 2. Comparisons of Experimental and Predicted Case Harden- ability Response for 8620 Steel. P35-8 - 9 - The next information that is desired, for a new alloy, are the mechan- ical properties of the heat-treated steel for various section sizes of bar products. Shown below are the calculated (Minitech) and experi- mental mechanical properties of pseudo-carburized (core properties) 8620 steel. MECHANICAL PROPERTIES (EXPERIMENTAL VERSUS PREDICTED) OF PSEUDO-I :arburized i 3620 STEEL (1700 -»• 1550 F, Oil Quench , 300 F Temp ier) Bar Diameter (inch) BHN 388 399 UTS (ksi) 199,500 199,000 YS (ksi) 157,000 181,000 % Elong : 13 14 t RA 49 47 Method 1/2 Experimental Predicted 1 255 268 126,700 134,000 83,750 112,000 21 21 53 60 Experimental Predicted 2 235 235 117,200 118,000 73,000 96,000 23 24 58 64 Experimental Predicted Considering the normal heat-to-heat variability this agreement be- tween calculated and experimental mechanical properties is considered to be good. For through hardening steels such as 4140, the quenched and tempered mechanical properties as a function of tempering temperature are of interest to the design engineer and the metallurgist. A comparison of the quenched and tempered mechanical properties (experimental ver- sus predicted) are shown in the table below for two different grades of alloy steel. QUENCHED AND TEMPERED MECHANICAL PROPERTIES (EXPERIMENTAL VERSUS PREDICTED) FOR 4140 AND 8650 STEELS Temper Temperature YS (ksi) UTS (ksi) % 1 Elong % RA Grade (F) Exp 238 Pred 235 Exp 257 Pred 247 Exp 8 Pred 11 Exp 38 Pred 4140 400 38 600 208 199 225 215 9 13 43 44 800 165 169 181 188 13 15 49 49 1000 121 128 138 149 18 19 58 57 8650 400 243 268 281 275 10 10 38 32 600 225 221 250 235 10 12 40 41 800 192 196 210 213 12 13 45 45 1000 153 146 170 166 15 17 51 54 1200 120 113 140 135 20 21 58 60 P35-9 - 10 - These data, generated for 1-inch round bar product, demonstrate good agreement between experimental and predicted values. In addition to the value of this powerful prediction tool to the alloy developer, the produc- tion metallurgist can use this computer program very successfully as a quality control tool and in heat disposition. In summary, there exists today a very large body of knowledge concerning the metallurgy and application of carbon and alloy bar products. This knowledge bank includes, in part, the development, application and over 35 years of experience with the WWII National Emergency Steels, the SAE EX steels most of which originated from the 1969-1970 time period when there was a nickel shortage due to a strike, and for the last several years in- cludes computerized hardenability prediction systems which are highly ac- curate and extremely fast and result in a significant shortening of the alloy development cycle time. With this knowledge base the steel industry is positioned to respond very rapidly to any possible raw materials short- age. In terms of alloy bar steels there would not be a serious problem in the event of a chromium shortage. The technology is in hand to readily find alloy substitutes for Cr containing alloy steels through the judicious use of Mo, B, and V. Whether it is a carburizing grade such as 8620 or a through hardening grade such as 4140 there is not a problem in finding an adequate non-chromium containing substitute. B. Structural Alloy Plate Under the category of structural alloy plate materials we will briefly discuss the possibilities for chromium conservation and alloy substitution in the following classifications of plate materials. 1) Constructional Alloys 2) Abrasion-Resistant Alloys 3) Weathering Steels 4) Pressure Vessel Steels 1) Constructional Alloys As shown in the table below, many of the high strength 80 to 100 ksi Y.S. structural, weldable plate steels contain Cr and Mo. These quenched and tempered plate materials (covered under ASTM Specifica- tions No. A514 and A517) are produced in a variety of thicknesses up to 6" and are intended for use in welded bridges and other structures. P35-10 - 11 - Constructional Alloys (High-Strength Plate) Compos i tion Y.S. ASTM Name C 0.16 Mn 0.55 Si 0.27 Ni Cr 1.70 Mo 0.50 Other Cu,V,B ksi 100 Spec. SSS-100 A514 RQ100 A 0.16 0.60 0.27 - - 0.55 B 100 A514 N-A-XTRA-100 0.21 0.85 0.65 - 0.65 0.28 Zr,B 100 A514 T-l 0.15 0.80 0.20 0.85 0.55 0.55 Cu,B,V 100 A514 HY80 0.18 0.25 0.27 2.75 1.40 0.35 - 80 - HY100 0.20 0.25 0.27 3.00 1.40 0.35 - 100 - HY130 0.12 0.75 0.27 5.00 0.55 0.45 V 130 In ASTM Specification A514 there are fifteen different compositions or Grades listed (Grades A through Q) some of which are shown in the table above. Of these fifteen different grades there are four grades that do not contain chromium. One of these grades is shown in the table above and it can be seen that higher Mo levels and the addition of boron are utilized in order to meet the hardenability require- ments. The different compositions indicate that the balance of alloy- ing elements is primarily based on hardenability considerations and experience has indicated that the non-chromium containing grades meet the specification and customer requirements equally as well as the Cr- containing grades. Similar to the situation with heat treated bar products it is preimarily a matter of achieving hardenability most economically. Therefore, if there were a chromium shortage there would not be a technical problem in the area of quenched and tempered constructional plate steels. The situation with HY80, HY100, and HY130 steels for Naval ship appli- cations is not as straight forward and adequate substitutes for these grades are not readily apparent at this time. Later in this session Dr. Brian Jones will have information to share concerning the possi- bility of substituting microalloyed HSLA plate materials for HY80 steel. 2) Abrasion-Resistant Plate Steels A similar analysis of abarasion resistant plate steels shown by the compositions listed in the table below reveals that there is an ample choice of quenched and tempered (Q & T) abrasion resistant plate mate- rials that are Cr free. The users can select from a variety of C-Mn, C-Mn-Mo-B, and C-Mn-Mo-Ni-B Q & T alloys. Therefore, chromium avail- ability would not be a problem for abrasion resistant plate steels. P35-11 - 12 - TABU M I • ABRASION-RESISTANT ALLOYS Compos it .ion, % Hardness, Name C Mn 0.04 S 0.05 Sj 0.20-0.35 Cu J> 0.85-2.0 Mo 0.15-0.60 Other (.) _ Bhn_ SSS-AR-321 0.25 0.40-0.70 0. ,20-0, .40 Ti or V, B 321 SSS-AR-360 0.25 0.40-0.70 0.04 0.05 0.20-0.35 0. 20-0, ,40 0.85-2.0 0.15-0.60 Ti or V, B 360 SSS-AR-400 0.25 0.40-0.70 0.04 0.05 0.20-0.35 0. ,20-0, .40 0.85-2.0 0.15-0.60 Ti or V, B 400 RQAR-321 0.25-0.32 0.40-0.65 0.035 0.040 0.20-0.35 _ 0.80-1.15 0.15-0.25 _ 321 RQAR-340 0.25-0.32 0.40-0.65 0.035 0.040 0.20-0.35 - 0.80-1/15 0.15-0.25 - 340 RQAR-360 0.25-0.32 0.40-0.65 0.035 0.040 0.20-0.35 - 0.80-1.15 0.15-0.25 - 360 RQAR-400 0.25-0.32 0.40-0.65 0.035 0.040 0.20-0.35 - 0.80-1.15 0.15-0.25 - 400 RQC-321 0.28 1.50 0.040 0.050 0.20-0.60 - - - B 321 RQC-340 0.28 1.50 0.040 0.050 0.20-0.60 - - - B 340 RQ321A 0.12-0.21 0.45-0.70 0.035 0.04 0.20-0.35 - - 0.50-0.65 B 321 RQ340A 0.12-0.21 0.45-0.70 0.035 0.04 0.20-0.35 - - 0.50-0.65 B 340 RQ360A 0.12-0.21 0.45-0.70 0.035 0.04 0.20-0.35 - - 0.50-0.65 B 360 RQ321B 0.12-0.21 0.45-0.70 0.035 0.04 0.20-0.35 - - 0.45-0.60 1.20- •1.50 Ni, B 321 RQ340B 0.12-0.21 0.45-0.70 0.035 0.04 0.20-0.35 - - 0.45-0.60 1.20- -1.50 Ni, B 340 RQ360B 0.12-0.21 0.45-0.70 0.035 0.04 0.20-0.35 - - 0.45-0.60 1.20- -1.50 Ni, B 360 3) Weathering Steels Weathering steels are HSLA 50 ksi min Y.S. plates and shapes intended primarily for use in welded bridges and building were weight savings and durability are important. The ASTM A-588 standard specification covers ten different grades of weathering steels (Grades A through K) and three of these grades do not contain chromium. While most of these steels contain C-Mn-Si-Ni-Cr-Cu and V they can be made to have adequate atmospheric corrosion resistance, in the absence of Cr, by increasing the Ni and Cu levels. A second option would of course be to do without weathering steels and get out the bucket of paint and paint brush. The bottom line with weathering steels is that an absence or shortage of Cr would not be a problem. 4) Pressure Vessel Plate Steels ASTM Standard Specification A387 for pressure vessel plates, covers the Cr-Mo alloy steel plates intended primarily for boilers and pres- sure vessels designed for elevated temperature service. This specifi- cation covers the well known and widely used 2-1/4 Cr - 1 Mo and 1-1/4 Cr - 1/2 Mo steels. P35-12 - 13 - The market for the ASTM A-387 grades (there are 8 different grades in this specification) is some 8,000 to 9,000 tons/year domestically and 15,000 to 20,000 tons/year worldwide. Concerning plate products there is only one major domestic producer (Lukens Steel) the remainder of the steel being foreign, primarily from Japan. The most popular grade is Grade 11 (1-1/4 Cr - 1/2 Mo) which is used extensively in process piping and heat exchanger tubing. The next most popular grade is Grade 22 (2-1/4 Cr - 1 Mo) which is primarily used as a plate product for pressure vessels and high temperature, high pressure hydrogen ser- vice for petrochemical applications. In these grades if you lower Cr content you loose high temperature strength. As the writer understands the state of the art of technol- ogy in this class of steels there are no known substitutes for these grades and furthermore there is apparently no research work being done to find substitutes or reduce Cr levels in these pressure vessel steels. The current thinking is that there is no known substitute for Cr when it comes to resistance to hydrogen at elevated temperatures. The Climax Molybdenum Company is doing research to develop an alloy which will have greater high temperature strength, greater harden- ability and greater resistance to hydrogen. This has lead to the development of a 3 Cr - 1-1/2 Mo - .10 V steel. Unless the section sizes can be reduced significantly, this development, while technolog- ically significant, will not be beneficial from a chromium conserva- tion viewpoint. In this product area the bottom line would appear to be that Cr sub- stitution or conservation does not look favorable at this time and there is a real need for research in this area. As there are a limit- ed number of domestic suppliers of A-387 plate steels this is an area where the Federal Government should step in and sponsor some R&D work. II. HIGH STRENGTH LOW ALLOY STEELS The high strength low alloy (HSLA) steels and their possible vulnerability to a chromium shortage will be reviewed in depth by the next two speakers on the program; Mr. Michael Korchynsky of Union Carbide Corporation and Dr. Brian Jones of Niobium Products Company. Their presentations will cover plate steels, skelp for line pipe applications, dual-phase steels HSLA microalloyed steels, weathering steels and HSLA steels in general. Dr. Koo from Exxon Re- search will discuss the optimization of alloying elements in low carbon dual phase steels and medium carbon structural steels. III. ULTRAHIGH STRENGTH STEELS The major use of ultrahigh strength specialty steels is in aircraft forging applications. These critical load bearing forgings are used in such applica- tions as engine mounts, tail section forgings, wing mounts, flap tracks and landing gears to mention just a few. The landing gear typically comprises about 14% of the weight of the aircraft, therefore, high strengths and good fatigue resistance are mandatory. P35-13 - 14 - The primary ultrahigh strength steels which have been used for landing gears over the past 30 to 40 years are shown in the table below. What is readily apparent from this list of steel compositions is that all of the alloys con- tain chromium. What is not nearly as readily apparent is whether a chromium- free steel could be developed for this very demanding high hardenability application. Aircraft Landing Gear Steels Composition Alloy C Mn Si Ni Cr Mo V Other 98 BV 40 0.43 0.85 0.65 0.75 0.90 0.50 0.04 4330 M 0.30 0.95 0.27 1.80 0.85 0.40 0.08 4340 0.40 0.75 0.27 1.80 0.80 0.25 - 300M 0.40 0.75 1.65 1.80 0.80 0.40 0.80 HP 9-4-30 0.30 0.25 0.10 7.50 1.00 1.00 0.10 4.5 Co HP 310 0.40 0.75 2.50 1.80 0.90 0.40 0.20 There is not currently any alloy development work in progress which has as its goal to develop a Cr-free steel for ultrahigh strength steel aircraft landing gear applications. As both military and commercial aircraft utilize the same alloys for landing gears (300M being the current work horse alloy for landing gears) this would appear to be another very appropriate and vital area for the Federal Government to sponsor alloy substitution research. SUMMARY In summary, the state-of-the-art of substitution for chromium in structural alloy, HSLA, and ultrahigh strength steels has been reviewed. With regard to heat treatable structural alloy bar products there are ample chromium-free substitutes for most grades and a high technology base for rapidly developing alternate chromium-free structural alloy bar steels. Concerning both struc- tural alloy plate steels and abrasion-resistant plate steels there has been available, for a considerable period of time, adequate Cr-free steels in both of these categories of Q & T plate materials. Likewise for ASTM A-588 weath- ering steels, there are a number of existing Cr-free weathering plate steels. The as-hot rolled microalloyed HSLA plate, strip and skelp steels generally do not contain chromium and these steels would not appear to be a problem area in the event of a chromium shortage. The two major alloy steel areas which would be severely crippled in the event of a chromium shortage, in the near future, are the ASTM A-387 pressure vessel plate steels and aircraft forging and landing gear steels. For both of these classes of steels there are no known Cr-free substitute alloys and there is no research in progress to develop such substitutes for either the pressure ves- sel steels or landing gear steels. Both types of steel are utilized in criti- cally important applications and it is felt that these two areas should re- ceive serious consideration by the Federal Government to sponsor research work to develop adequate chromium-free substitute materials. P35-14 Alternative Compositions for Future HSLA Steels The technology of high strength low-alloy (HSLA) steels has made tremendous advances over the past decade. This has come about because of the development of controlled-rolling techniques and research into the benefits of microalloys. There now exists a family of micro-alloyed steels capable of providing high yield strength combined with low tempera- ture toughness and weldability. They have been utilized primarily in the pipeline and automobile industries. The purpose of this paper is to illustrate the properties available from this class of steel and to suggest that their field of application be extended to include areas traditionally reserved for heat-treatable chromium-containing steels. The Development of HY 80 In the mid 1950' s, design requirements called for a steel with 80 ksi yield strength for use as plates and frames in submarine pressure vessels. This necessitated the steel to provide adequate strength, ductility and notch toughness together with good weldability under shipyard conditions. Strength and notch toughness were, of course, also required from the weld deposit and HAZ. In the USA, following a fairly exhaustive investigation of the alloy approaches then available, this led to the development of the heat-treatable Ni-Cr-Mo steels known as HY 80. These are fully-killed aluminum-grain-refined steels which develop the desired strength proper- ties following water-quenching from 1650°F and tempering in the range 1200-1250°F. A specification for chemistry and mechanical properties for HY 80 is given in Table 1. P36-1 The difficulty in alloy selection was in obtaining an approach which provided the high strength levels required together with acceptable low temperature toughness levels. To achieve this 25 years ago it was felt necessary to use heat-treatable nickel-containing steels which have a shallow-sloping brittle-ductile transition curve and therefore, retain respectable toughness at low temperatures. This however, in the steels concerned is achieved only at the expense of weldability. Due to what is now considered to be a relatively high carbon level and the chromium and molybdenum additions, HY 80 is a steel which can be very prone to hydrogen- cracking. Its carbon equivalent using the International Institute of Welding formula (see Table 2) is around 0.70. This means that successful welding requires the use of low-hydrogen techniques usually in conjunction with pre-heat to allow hydrogen escape during post-weld cooling. Three pre-heat calculation analyses are indicated in Table 2 all recommending levels of 150°C or 300°F minimum. In the 1980' s, however, we no longer need to use heat-treatable nickel steels to achieve good low temperature properties. Controlled rolling and microalloying have now produced materials sufficiently fine-grained to give the required strength levels and to push transition temperatures to extremely low temperatures. Since the pipeline industry for which these steels were primarily designed views field weldability as a primary requisite, there has been a great incentive in this direction and these properties are achieved with carbon equivalents below 0.35 indicating excellent material tolerance of hydrogen and resistance to HAZ cracking. Low Transition-Temperature Micro-alloyed Steels Table 3 shows typical compositions of several low-carbon microalloyed P36-2 steels which have been developed in recent years. Selection of microalloy approach is made on the basis of final property requirements, controlled- 0) rolling schedules and mill capabilities. The versatility of microalloy approach can be best illustrated by considering the development of X70 grade (70 ksi yield strength) linepipe steels. A few years ago it was considered necessary in producing these materials at thicknesses of about 0.6 inches to use a Nb-Mo approach with about 0.04% Nb and 0.30% Mo. This led in Italy for instance to the 0) development of the Molytar Steels. In 1979, however, due to a temporary sharp rise in the molybdenum price, alternatives to the Mo addition were sought. In Italy, it was found expedient at that time to use Nb-Cr steels with 0.2% Cr. While this does not have the hardenability effect of Mo, suitable controlled rolling schedules were easily able to achieve X70 grade requirements. Further work, however, established that even the chromium addition could be replaced and that, with the correct heat- treatment and rolling methods, a dual micro-alloy addition of Nb and V could do the job. The Northern Border Pipeline, therefore, which represented an order of more than 500,000 tons of mainline 0.6 inch wall X70 grade pipe, used no molybdenum and more than 90% of it from manufacturers across the world, used 0.04% Nb and 0.05-0.07% V as the microalloys. Some heats did utilize the Nb-Cr approach, however, and this allowed a direct comparison to be made of the tensile properties achieved by replacing 0.20% Cr by a much smaller vanadium addition. It can be seen from the distribution of values shown in Fig. 1 that no variation in yield strength attained was observed and only minor changes in ultimate tensile strength. P36-3 For plate of this gage and strength, therefore, there is no requirement to use chromium and a selection of microalloy approaches are available. For thicker grades, it is common to use molybdenum additions once again and for high toughness at severely low temperatures - the most stringent requirement of HY 80 specif icaiton - several novel microalloy developments are being promoted. Tables 4-6 illustrate a development by Nippon Steel known as ultra low carbon bainitic steel or ULCB. Here a very small addition of boron (.001%) leads to a transformation from ferrite-pearlite to a bainitic structure. In order to utilize the effects of boron, nitrogen has to be fixed by a titanium addition and carbon reduced to very low levels (< .03%). In ultra-low carbon steels the role of niobium is also important since it can remain in solution additionally promoting the achievement of a super- fine bainite structure. The properties achieved in this way are most impressive. It can be seen in Table 5 that yield strengths in the vicinity of 80 ksi can be attained in plates up to 22 mm thick with very low drop weight tear test transition temperatures. The more commonly specified 50% Charpy crystal- Unity temperature is in all cases less than -80°C. For the alloy designated UB-5 it can be seen that a yield strength of 95 ksi is achieved due to the additions of Mo and Ni, in 20 mm plate. This excessive yield strength can in fact be traded off to allow X80 properties in plate greater than 1 inch in thickness. Table 6 shows the obtained submerged-arc weld metal and HAZ proper- ties. The high toughness is explained by the grain growth suppression effect of titanium nitride particles and the elimination of martensite formation due to te low levels of carbon. P36-4 The requirement for very low-temperature finish-rolling (controlled- rolling) can also be relaxed under certain circumstances. Recent research (S) investigations on an HSLA steel containing 0.5% Ni and 0.13% Nb has indicated properties in excess of X70 grade together with very good low temperature toughness in both base plate and submerged-arc weld. These were obtained in plate rolled to 0.6 inch thickness at a finishing tempera- ture of 1400°F. Composition and properties of this steel are shown in Table "7 . The low temperature toughness characteristics are illustrated in Fig. 2. . Conclusions The microalloyed controlled-rolled steels developed mainly for line- pipe applications are able to develop 70 ksi and in certain instances, 80 ksi yield strength together with excellent low temperature toughness properties, both in the plate material and in submerged-arc weld metal and HAZ. They can be used in thicknesses certainly up to 1 inch and, with additions of molybdenum, to greater thicknesses. They need contain no chromium and offer very significant advantages in terms of weldability over the heat-treatable Ni Cr Mo HY 80 steel traditionally used for shipplate applications. The microalloys involved are Nb, V, Mo and Ti. While niobium is currently obtained largely from Brazil and Canada, there are ample deposits in the USA, currently not regarded as commercially exploitable. The other microalloy elements are widely distributed thoughout the world including sources in the USA. P36-5 References 1. "The Metallurgy and Welding of QT 35 and HY 80 Steels", The Welding Institute. Research Report. 1974. 2. B. L. Jones, "Metallurgical Considerations in Linepipe Production", American Welding Society Conference on Pipeline Welding and Inspection, Houston, September 1982. 3. M. A. Civallero, C. Parrini and N. Pizzimenti, "Production of Large- Diameter High Strength Low Alloy Pipe in Itlay", Microalloying 75 Conference, Union Carbide, Washington, D.C. 1975. 4. H. Nakasugi , H. Matsuda and H. Tamehiro, "Ultra-Low Carbon Bainitic Steel for Linepipe", Conference on "Steels for Linepipe and Pipeline Fittings", Metals Society, London, 1981. 5. F. Heisterkamp and K. Hulka, Niobium Products Company Ltd., internal report, 1982. P36-6 Chemical composition and mechanical properties specification for HY80 steel plates (from MIL-S-16216G ships) Chemical composition Element Per cent Not less than Not more than Carbon Manganese Phosphorus Sulphur Silicon Nickel Chromium Molybdenum 0.10 0.15 2.00 1.00 0.20 0. 18 0.40 0.025 0.025 0.35 3.25 1.80 0. 60 Residual elements Titanium Vanadium Copper S + P < 0. 045% 0.02 0.02 0.25 Tensile properties Thickness range mm (in. ) 0.2 per cent proof stress, (to) Elongation on 50mm (2in. ) Reduction in gauge length, % area, % Minimum Maximum Minimum Minimum Less than 16 (0. 625) So - loo 19 - 16 (0. 625) and over lo - 95 20 55 longitudinal 50 transverse Impact properties i. Minimum energy absorption in Charpy V notch test Plate 11. 5mm (0. 5in. ) to 50mm (2. Oin. ) incl. Plate over 50mm (2. Oin. ) 68J (50ft lb) at -84°C No single specimen < 61J (45ft lb) 41J (30ft lb) at -84°C No single specimen < 34J (25ft lb) P36-7 Table I. Carbon equivalent, welctaDincy ana pre-neac calculations tor HY 80 and similar steels (Ref. 1). Preheat temperatures to avoid HAZ cold cracking (NCRE approach) Carbon % Thermal severity number (TSN) equivalent, 4 8 12 16 20 24 0.4 _ N 40°C 70°C 85°C 95°C 100°C 0.5 40°C 65°C 90°C 105°C 115°C 120°C 0.6 65°C 90°C 110°C 125°C 135°C 140°C 0.7 90°C 110°C 130°C 140°C 145°C 150°C 0.8 110°C 125°C 140°C 155°C 155°C 160°C Comparison of hardenability of HY80, Ql (N) and QT35 steels using carbon equivalent \jCjTq - Kj T — T — — — — T —^—^—^—— 6 15 5 HY80, % Q1(N), % QT35, % Typical ladle analysis for plate Maximum specified ladle analys less than 44mm (1. 75in. ) thick is for 44mm (1. 75in. ) thick plate 0.71 0. 95 0.71 0.64 0. 95 0. 75 Calculated minimum preheat temperatures to avoid HAZ cold cracking in plates having typical analyses, assuming a restrained weld of thermal severity 18 and poor fit -up. Carbon equivalent, % NCRE Welding Institute approach , approach , oc °C UK MOD (Navy) recommended pre- heating range, °C QT35 0. 64 HY80 0. 71 Ql (N) 0.71 136 - 5 133 - 10 144 - 5 156 - 10 144 - 5 156 - 10 120-150 120-150 120-150 Table 3 . Typical Chemical Compost t ions of X-70 Pipe Steel Type Gage C Mn SI Ho N1 Cu Cr M> V T1 N B Nb-V 18 0.09 1.50 0.15 . 0.16 0.10 0.12 0.04 0.08 0.016 0.005 Mn-Mo-Nb 22 0.05 1.70 0.20 0.26 - . - 0.06 . 0.018 0.001 . Mn-Nb-V 22 0.08 1.90 0.27 0.24 0.20 0.24 . 0.05 0.06 . 0.007 . Nb-N1 (14) 18 0.07 1.60 0.20 . 0.55 _ . 0.11 . 0.019 0.004 . Nb-V-Cr 15 0.09 1.52 0.22 , - - - 0.35 0.04 0.08 . 0.008 . Nb-Mo-Cr 15 0.08 1.50 0.20 0.18 - . 0.30 0.05 . . 0.010 . Mo-Nb-Cu 15 0.06 1.45 0.20 0.25 - 0.30 . 0.025 . . 0.009 . Ti-Nt 15 0.08 1.58 0.30 - 0.26 . . . . 0.070 0.004 . Ti-Cu-Ni (10) 20 0.09 1.61 0.30 - 0.12 0.21 . . . 0.063 . . Low C-Nb 20 0.02 1.78 0.13 - - - . 0.10 . 0.015 0.004 . Low C-Nb-B (15) 20 0.02 1.89 0.13 - - - - 0.05 - 0.016 0.002 0.001 Low C-Nb-Mo-a (15) 22 0.02 2.01 0.16 0.30 - . . 0.05 - 0.018 0.002 0.001 C-Mn-Nb (Q i T) 20 0.09 1.37 0.27 0.16 1.00 - . 0.02 - 0.020 0.005 . C-Mn-Nb 20 0.08 1.55 0.60 - . . . 0.06 . . 0.010 . (Controlled Cooled) *0.?5 for gages above 18-20 urn. All Alumi nun leve Is .02 - .05 All Sul fur level s .002 - .007 P36-8 Table 4" Typical Chemical Compositions of ULCB Steel for Various Grades U) Steel Grade C Si Mn P S , Ni Mo Nb Ti B dV -£' UB-1 X-65 0.02 0.14 1 .59 0.018 0.003 - - 0.04 0.017 0.001 0.29 0.11 UB-2 X-65 0.03 0. 16 0. 14 0.15 1 .61 1 .91 0.016 0.018 0.003 0.003 0.17 - 0.05 0.05 0.016 0.018 0.001 0.001 0.31 0.34 0.12 0.14 UB-3 X-70 0.03 UB-4 X-70 0.01 1.87 0.022 *3) 0.007 - - 0.04 0.020 0.001 0.32 0.11 UB-5 X-80 0.02 0.26 1.95 0.022 0.003 0.38 0.31 0.04 0.019 0.001 0.43 0.16 M) Ceq = C ♦ -^ 6 O + Mo + V *2) P CM = C ♦ Mn+Cu+O 20 Si 30 Ni+Cu 15 V Mo 10 + 15 Ni 60 + 5B '3) High-sulfur content Table 5 Mechanical Properties of Pipe Body Steel Grade Pipe size (mm) Direct. of test Tensile properties Charpy V-notch Impact properties B0WTT Y.S. (MPa) T.S. (MPa) El. in 50.8mm (X) Yield ratio (%) Energy at -20°C (J) 501 Shear FATT (°C) Shear fracture area at -20°C (%) 80* Shear FATT PC) UB-1 X-65 1420 0D x 16 WT Trans- verse 500 583 42 85 369 < -80 100 -50 UB-2 X-65 1220 00 x 25 WT Trans- verse 493 602 45 82 324 <-80 95 -30 UB-3 X-70 1420 00 x 20 WT Trans- verse 542 641 42 85 206 <-80 100 -40 UB-4 X-70 762 00 x 22 WT Trans- verse 551 622 41 89 159 < -80 98 -35 UB-5 X-80 1420 00 x 20 WT Trans- verse 653 732 33 89 178 <-80 100 -50 Mot c : 7oWs, S 4-73 ^Pcv • SoVc*; 3 55o tofc ; OoW^i = C\<\ Kr* ; looks 6M H Table G Mechanical Properties of Submerged-Arc Weld Joint Steel Grade Pipe size Transverse weld- tensile properties Cha impac rpy V-notch t properties T.S. (MPa) Location of fracture Test position Energy at -20°C (J) 50% Shear FATT (°C) UB-1 X-65 1420 0D x 16 WT 595 Base metal Weld metal 164 -40 HAZ 248 -50 UB-2 X-65 1220 0D x 25 WT 617 Base metal Weld metal 191 <-60 | HAZ 272 1-40 UB-3 UB-4 UB-5 X-70 X-70 X-80 1420 0D x 20 WT 762 0D x 22 WT 1420 OD x 20 WT 665 633 727 Base metal Weld metal 148 -40 HAZ 182 -35 Base metal Weld metal 183 -45 HAZ 62*) -20 Base met. a 1 Weld metal 189 -60 -20 HAZ 130 *) Low energy due to high-sulfur content P36-9 Table 7. Chemical Composition and Mechanical Properties of 0.5% Ni, 0.1% Nb Steel C Si Mn P S Al Ni Nb 0.077 0.36 1.56 0.017 0.005 0.02 0.47 0.13 Y.S. UTS C y (+20°C) 50% CVN-FATT 85% BDWTT-FATT 70.1 ksi 84.5 ksi 78J -85°C -55°C 56 inch diameter SAW pipe, 17.5 mm wall thickness P36-10 F&.i. X70 HSLA LlN£P\PE STEELS lot •°Z Bam fc~TT P36-11 9o 1 to "5 t 40 C 5 Bqse w«M wel4 mtfel T«sr- Tenp«roI > ur« - C 2.0 Ftfr.X ~JouaU*esS proper Wps of Nu- Nb Sutler qe«A - ^c u>*U«A oipe (n«S w*. W.fr./ »ji Opportunities for Surface Modification Technology in Conservation of Chromium by Peter G. Moore* Although there are many other uses for chromium, three principal uses are to improve the corrosion-, oxidation- or wear-resistance of structures and materials. Because these properties are surface related, coatings can be used in many cases to reduce the amount of chromium used and to allow the surface properties of structures (as opposed to materials) to be tailored to surface requirements without sacrificing bulk properties. A number of directed energy tsjam (DEB) processing techniques have been developed in recent years which provide new opportunities for the conservation of chromium through re- duced usage and through substitution for chromium by other elements. Today, I will just describe in general terms two classes of processing techniques (ion beam processing and laser beam processing) which produce novel corrosion- and wear-resistant surface layers and then describe several cases in a bit more detail. With ion beam processing, the surface of the structure is bombarded by ions with energies in the range of 10 to 200 keV. These ions are stopped within 1 pm of the surface of the solid material as a result of a series of collisions. These collisions alter the microstructure of the surface layer and the implanted ions alter the composition of the surface layer. This ion implantation requires the production of an ion beam of the desired species. Ion mixing and ion plating are variations where a thin layer of the desired material is first coated onto the structure and then an argon or xenon ion beam is used to intermix the coating and the substrate atoms. * Peter G. Moore is with the Naval Research Laboratory P37-1 With laser processing, the surface of the structure is rapidly heated by a high power laser beam until the surface melts. After the beam is turned off, the liquid rapidly resolidifies and is quenched as the heat is conducted away to the bulk of the structure. Variations on this laser-surface melting process include shock and transformation hardening (where no melting takes place) , sur- face alloying and cladding (where alloying elements are added) and particle in- ject- Cohere wear resistant particles are incorporated into the melt) . There are also many analogous processes using electron beams. Directed energy beam (DEB) processed coatings in general have several advan- tages over traditional coatings: (1) a wide range of controlled coating thick- nesses is possible, (2) coatings are laetallurgically bonded to the substract, (3) there are few thermodynamic constraints for laser processing and fewer still for ion beam processing, and (4) structures and properties can be graded with depth from the surface in order to optimize performance. There are also several characteristics of DEB processing which restrict its use. Almost all DEB processing is either done in vacuum or in a helium which add to the cost of processing. Because all of the techniques are line-of -sight processing, structures which can be processed by such techniques are restricted to geometries which are relatively open and free of recesses. In addition, pro- cessing is limited to relatively small areas and structures, because of the vacuum requirement for ion beam processing and because of warpage due to residual stresses for laser beam processing. Ion Implantation As an example of the potential for solving materials problems through the use of ion implantation, consider the study of the effect of implanting various species on the resistance of M50 bearing steel to the pitting corrosion which results in the salt water contaminated aircraft lubricants. P37-2 Test specimens of M50 steel were implanted with 150 keV chromium ions to obtain a peak chromium concentration of 24% at a depth of about 400 A below the surface. Electrochemical polarization measurements and simulated field service tests indicated an improved resistance to pitting corrosion and as a result, an improved field performance. Laser Processing Three examples illustrate the potential of laser processing for solving materials problems in the areas of corrosion, oxidation and wear. These are the surface melting of 304 stainless steel to improve resistance to localized corrosion, the surface alloying of chromium into steel substrate to produce a stainless steel surface layer, and the production of wear resistant layers by the particle injection method. Recent work has attempted to improve the resistance of 304 stainless steel to pitting corrosion by laser surface melting. The resolidified material exhibited a very fine dendritic subgrain structure which indicated cooling rates in excess of 10 K/sec. These materials were electrochemically characterized by potentiodynamic polarization experiments in 0.1 M NaCl which indicated an improved resistance to pitting compared to wrought 304 and comparable to that of 316 stainless steel. A number of possible reasons exist for this improved behavior as a result of the redistribution of both major and minor alloying elements through homogonization during melting and segregation during and after solidification. At this time, it is felt that the improvement is due to the eliminiation of large precipitates which cause the local passive film to be less protective. Chromium steel surface alloys have been produced on various steel sub- strates using a variety of processing techniques. The electrochemical behavior P37-3 of these alloys has verified that the surface layers passivate in the manner of stainless steels. The broader significance of this and the previous re- sults is that a stainless steel surface layer, suitable to the chemical en- vironment can be rapidly and efficiently produced on the surface of a steel structure . A third type of laser processing, particle injection, is performed in the fashion of surface melting except that hard, wear-resistant particles are injected into the melt pool. This results in a metal -matrix composite surface layer which is metallurgically bonded to the substrate. Abrasive wear tests of layers produced by the injection of TiC and WC into aluminum-, titanium-, nickel- and iron-base alloys result in performance during abrasive wear tests which are comparable to that of specialty wear coating materials. An especially good example of the potential for this technique is the produc- tion of a TiC-aluminum matrix composite on an aluminum substrate; a wear re- sistant surface combined with a high thermal conductivity high specific- strength base metal which cannot normally be considered for applications where wear is a problem. Conclusion Ion implantation and laser processing are both attractive techniques which, when applied to appropriately, can be used to solve wear, corrosion and oxida- tion problems. The examples presented here have illustrated situations where chromium can be conserved by using these coating techniques, and performance comparable to or better than that of bulk alloys using chromium is obtained. In these applications, chromium was used in a more or less traditional manner and the role of the chromium in enhancing the performance has been the same as its role in conventional alloys. By the nature of these techniques, there are also many possibilities for enhancing performance of structures by pro- P37-4 ducing novel materials. It is this two-pronge potential of DEB processing technique which will make them invaluable in the event of critical shortages of key materials. P37-5 SALT BATH TREATING AS AN ALTERNATIVE FOR CHROMIUM PLATING Nitrogen/Oxygen Synergism The potential for replacing chromium plating by an oxidized nitrided surface (Fig. 1) evolved from a completely unrelated event. Actually the concept of following a liquid nitriding treatment by quenching in an oxidizing fused salt was originally devised to destroy any slight traces of cyanide developed during nitriding and the substantial amount of cyanate present in the salt as the active agent (Fig. 2). The corrosion resistance of the resultant combination diffusion and conversion coating is truly a synergistic effect. The Aerated Liquid Nitriding Process A more thorough understanding of the surface produced can be achieved by a brief consideration of the original nitriding process. The aerated liquid nitriding process was introduced into the United States in the late 1950' s and early 1960's. The treatment requires immersion in a cyanide/cyanate fused salt at 1060°F (570°C) for a 60 to 180 minute period. The salt is aerated by a sparging ring immersed in the bottom of the fused mixture. Make-up salt which consists of sodium and potassium cyanide is added and reacts with the bubbled air to produce sodium and potassium cyanate according to the following reactions: KCN + 1/2 2 = KCNO (1) NaCN + 1/2 2 = NaCNO (2) The formation of the cyanate compounds in situ in conjunction with a 45 to 50% cyanide content and a 45 to 50% cyanate content is the basis for U.S. Patent No. 3,022,204. During the nitriding action on carbon steels a 0.0004 in (.01 mm) to 0.0006 in (.015 mm) compound zone is formed in con- junction with a 0.018 in (0.45 mm) diffusion zone, (Fig. 3). The compound zone is composed of single phase epsilon iron nitride (Fe^N) (Fig. 4) responsible for its yery unique wear resistance and low coefficient of friction properties. The nitrogen dissolved in the diffusion zone (Fig. 5) is responsible for substantial increases in endurance values in both notched and unnotched ferrous structural parts. In addition, the single phase epsilon iron nitride surface exhibits excellent corrosion resistant properties and in many cases the nitriding treatment is performed with this factor as a primary consideration. P38-1 -2- Sub-Critical Heat Treatment Aerated liquid nitriding has the unique characteristic of being a sub- critical single heat treatment capable of increasing corrosion resistance, endurance values and wear resistance. The development of the epsilon iron nitride surface and the nitrogen diffusion at temperatures substantially below the critical temperatures of structural steels permits the elimination of excess material and the subsequent required finishing operations to com- pensate for growth and distortion during processing. The process is thus ideally suited for treating cylindrical shapes such as cylinder liners and rocker arm shafts, where final finishing costs can be excessive. This capability of retaining dimensional stability is also of vital importance where corrosion protection is dependent upon thin films developed by heat treatment thereby precluding final machining. Process Changes Required Because of Environment For many years this specialized heat treatment was documented, tested and specified for crankshafts, gears, rocker arm shafts, connecting rods, valves, valve guides and numerous other engine and chassis parts in all types of industrial applications. Environmental pressures eventually restricted the use of cyanide compounds for metal processing in certain areas and it became necessary to develop a relatively cyanide free liquid nitriding process. At the same time the stated goal was to be able to produce a single phase epsilor iron nitride compound zone (Fig. 6) with a nitrogen diffusion zone that could produce corrosion resistance, endurance values and wear resistance equivalent to those obtained in the existing process. The records accumulated during years of testing of the original process made interchangeability a desired asset. New Liquid Nitriding Process The aerated liquid nitriding process introduced in the 1970's met these requirements in all respects. Unfortunately, under certain conditions the fused salt also had the capability of generating small amounts of cyanide during processing, according to: KCNO + 2FeO = KCN + Fe^ (3) and therefore could not be specified as a complete cyanide free salt. Overcoming this objection involved post treatment in an oxidizing media which began as a standard fused sodium and potassium nitrate/nitrite mixture and then finally evolved into a sodium and potassium hydroxide, carbonate and nitrate composition capable of total destruction of all traces of cyanide and cyanate, (Fig. 7). The hardness patterns (Fig. 8) and (Fig. 9), wear characteristics (Fig. 10) of the quenched nitrided parts exhibited minimal change but the corrosion properties were substantially increased (Fig. 11). Laboratory analyses of the quenched parts revealed that the unexpected corrosion resistant results obtained were not a function of the jet black surface (Fig. 12) but a more complex combination of nitrogen, oxygen and iron. P38-2 Auger profiles substantiated the fact that the diffusion of oxygen into the epsilon iron nitrided compound zone was responsible for the corrosion resistance. Studies eventually revealed that optimum dwell times in the fused quenching salt would result in maximum corrosion resistance, although this work is still in progress. The complete mechanism involved in con- verting a liquid nitrided surface into an effective corrosion barrier in carbon and alloy steels is not completely understood. Test results how- ever continue to indicate that the majority of situations favor the substitution of this treatment for chromium plating and this trend should continue if the projected costs and availability of chromium metal eventually become a reality. Surface Compatibility with Bearings and Seals Aerated liquid nitriding in conjunction with oxidizing salt quenching may be a viable substitute for chromium plating but the process as performed lacks compatibility in contact with many non-metallic seals and bushings. With this in mind a third and fourth step was added to the treatment when the finished product is to be subjected to the above conditions. Basically the objection to a standard nitrided and quenched part is the surface residue or roughened finish resulting from nitriding, which can be a lap in contact with softer nonmetallic materials. Fortunately this difficulty can be easily overcome by polishing. Method of polishing is completely optional providing surface removal is uniform and restricted to 0.00005 in (0.0013 mm) per surface. Automated fixtured polishing with 320 grit paper has been used on bearing and contact surfaces of rotating or sliding shafts. Vibratory finishing is completely acceptable where configuration prohibits fixturing. Following the polishing operation the parts are immersed once again in the quench bath for an oxygen diffusion period of twenty minutes. The effect of this sequence of operations is shown in the bar graph of (Fig. 13). Starting from the outside left bar, the finished carbon steel has yery low corrosion resistance and a standard commercial finish. After aerated liquid nitriding, the corrosion resistance has increased dramatically, but directly above, the surface roughness has also increased substantially. The third set of bars illustrates how fine polishing reduces the corrosion resistance slightly but effectively produces a surface finish suitable for bearing contact. Finally, requenching in the oxidizing salt bath develops the ultimate combination of surface finish and corrosion resistance. The final process with oxygen diffused nitrided surfaces operating against elastomeric and nonferrous seals or bearings is as follows: 1. Aerated liquid nitriding in a cyanide free fused salt -- 60 to 120 minutes at 1070°F (577°C). 2. Quench in controlled rate oxidizing salt at 750°F (400°C). 3. Water rinse. P38-3 -4- 4 ' KMM t n To e m3 r ^ e f1n1Sh w1th max1mum stock 5. Reversion in controlled oxidation bath, dwell time 20 minutes. 6. Water rinse and oil. Corrosion and Wear Evaluating Any system that is developed as a <:.ihcti+,,+« * must be subjected to extensive laboratory It : I? r " Pr ° Ven existin 9 Product H~^ carbon steel are subjected treatments (Fig. 14). The ox d zed c ^ *! SuUable Protective test duration reflected less than a S - Urface after the no ™al 88 hour less than the 70% figure show for both Z™™ \ rea . Which ** significantly nickel plating. Continuation of fhS 2« h I° n "" m P atin 9 and electroless produce failure) deve oped a 336 Int rl ° n - the nitrided shafts ™^ (to significantly lesl than the result? h r^S" area ° f 5% to 10% wh ^ is Plated shafts during a 7 hour L ^ fro T the chro ™ and nickel were performed in .Jco'^^S^ B ft .Til^S 8 ^ *f ^•M^ 1?*«** of a supplier surface treatments, chromium p atina and n ; i 6 J: Tw ° rods with different subjected to an accelerated life left Sf 300 nm* U l Uld ^J^ns. were applied, followed by 96 hours of ASTM i I ° CyCle f Wlth a s1de load imparts a high pressure contact \onI h!l a ll S F/ ay test ' The fading rod in a limited area The floure ?i w?! th « °- r1n S se *l and the piston has numerous corrosion sites on the lurfarf^-^'L^ 6 ? hrom1um P lated ™d shows little effect f nm\ the surface while the oxidized nitrided rod Plated 1? J thelssemb y' ™ *1 lOTurf&fifff" S?"* r r ° dS > "™ are shown in (Fig. 17). q nitrided followed by quenching, foreground, applied through a rotating SrtldTcvTlnSiTSfi (Flg> 8) ' The test load was valve stem, oriented at riaM *l„L/l tl m ^ ng contact with a stationary on both the intake valves 9 MSM54 and fl'Vl""^ a \ 1s ' Tests were P erf °™« minute test period usinq prescribed i«i L X 5 auSt valve 21 " 2N - After a one loss of weight in the Zn ^componenjs (Fig th ?9) e9ree ° f Wear was measured « P38-4 -5- Summary In summary, the oxygen diffused nitriding process has the capability of producing a surface comparable or superior to chromium plating with respect to wear resistance and corrosion resistance in numerous industrial appli- cations. Unlike chromium plating, there is no requirement for the utilization of strategic materials. Also, unlike electroplating processes, there is no water or air pollution problems involved when the oxidized liquid nitriding process is practiced. The materials required to produce both of the fused salts involved in the process are readily available common chemicals and none of the compounds present in the effluent rinse waters are presently categorized on a restricted discharge basis. Unit costs of any process are usually suspect after a very short period of time because of our rapidly changing economic conditions but on a comparative basis the total costs involved in chromium plating and the oxygen diffused nitriding process are quite similar and are likely to be that way even with future changes in the Cost of Living Index. P38-5 Fig. 1. Ball studs, aerated liquid nitrided and salt bath quenched. Decomposition of Cyanide and Cyanate in Quench Salt CN" in mg/kg CNO" in mg/kg Cyanide Analyzed after 3 minutes 5 minutes 10 minutes - 30 minutes ,L XU_ 200 250 300 350 200 250 300 350 Temperature of Quench Salt in °C Fig. 2. Chemical destruction of cyanide and cyanate residues, after salt bath quenching, related to time and temperature. P38-6 jgg*^^***^?^'^* 1 COMPOUND I ZONE IRON NITRIDE NEEDLES CYANIDE-FREE LOW TEMP. LIQUID NITRIDED 90 MINUTES AT 1060F (570C) DIFFUSION ZONE Fig. 3. Compound zone and diffusion zone developed by low temperature liquid nitriding. Note the needle like structure of Fe.N developed in the diffusion zone by aging. 1600 1400 1200 1000 °F 800 < \ IRON -NITROGEN |\ EQUILIBRIUM DIAGRAM V 1 ^ i i i *l ■/■ !*N \ '1 : \ J —i *~«. V \ e oC-< h ? i &\ L ■ i h WT. % N 10 Fig. 4. The epsilon iron nitride phase in aerated liquid nitriding is developed, as shown, slightly to the right of the intersection of the 8% nitrogen line with the 1060°F (570°C) temperature line. The stoichiometric ratio of nitrogen to iron is very close to one to three. P38-7 ■ 2, 1.5 .15 1- -V IISTRISUTfON sac ws 179 MINUTES 1075F/IMC WATER QUENCH .05- m% Nitrogen .01 HON-WHiUTIRG LIQUID NITRIOING .02" DISTANCE FROM SURFACE .04" .06" Fig. 5. Nitrogen distribution in the diffusion zone of aerated liquid nitrided carbon steel. It is this nitrogen that is responsible for increases in endurance values from 20% to 100%. X-RAY DIFFRACTION SCAN o low Ttmpti rtun ■ Liquid NHridad a Standard Chwnfctiy 3 2 1 1 «F. - B B • 1 fl 1 It B | B -* II .,., B B II ■ ■ | 111 ff - ■ B B 111 ill ■ B n " u m Wj 11 ■ B ■ B fl B B . 74 72 70 M W 64 62 60 58 565! zv Mngu Mf D+0I6M Fig. 6. X-ray diffraction pattern of aerated liquid nitrided carbon steel. The three typical peaks of epsilon iron nitride are present while the absence of other compound peaks indicate that the zone is single phase. P38-8 - Cyanide Analysed after; • 3 minutes x5 « — o» a30 300 350 200 250 Temperature of AK-Bath in X Decomposition of cyanide and cyanate in AK-Bath Fig. 7. Chemical destruction of cyanide and cyanate in an oxidizing media. Fig. 8. Hardness pattern resulting from aerated liquid nitriding of plain carbon steel. The sharp drop below the surface measures the depth of the compound zone. P38-9 Fig. 9. Hardness pattern developed by aerated liquid nitriding of alloy steel. The higher surface hardness and the measurable retention through the diffusion zone results from alloy influence, RESULTS OF WEAR TEST 2 3 RUNNING TIME IN HOURS Fig. 10. Wear test comparison of untreated, carburized and aerated liquid nitrided specimens. P38-10 f ( Fig. 11. Salt water corrosion tests on various surface treatments. The SBN-SBQ notation refers to salt bath nitriding followed by salt bath quenching. The other treatments are standard. Fig. 12. The top gas spring has an aerated liquid nitrided shaft compared to the lower spring with a chromium plated shaft. P38-11 Fig. 13. The effect of sequential surface treatments on surface finish and corrosion resistance. The acronyms for aerated salt bath nitriding, SBN, and oxidizing salt bath quenching, SBQ, are used. Fig. 14. Automotive spool shafts subjected to salt spray testing according to ASTM B-117. The nitrided/ quenched surface is indicated by QPQ. P38-12 Fig. 15. Resistance of various surface treated spool shafts exposed to corrosion testing according to ASTM B117. QPQ refers to the quenching and polishing of aerated liquid nitrided specimens, Corrosion Resistance Piston Rods Test Conditions: 300,000 cycles endurance tests Hard Chroma Plato .Corrosion Sites Failed; more than ten corrosion sites per sq. In. Passed; no evidence of corrosion Fig. 16. The aerated liquid nitrided piston rod after salt bath quenching, polishing, and requenching exhibits no evidence of corrosion after endurance testing followed by 96 hours of salt spray. P38-13 Fig. 17. Shock absorber assembly and rod. Combinations of endurance properties and corrosion resistance are required. Plat Fig. 18. Surface treated engine valves, P38-14 Wear Resistance Evaluation Surface Treatments Test Conditions: Engine valve stems, cross cylinder wear test, 1 minute duration Log Weight Loss (grams) .2000- S8N-SBQ Chrome Plate Zero Load 25 ib. Load Intake Valve Stem Zero Load 25 lb. Load Exhaust Valve Stem Fig. 19. Comparison of the wear properties of hard chromium plated engine valve stems with stems treated by aerated liquid nitriding followed by salt quenching. P38-15 CLAD METALS - MATERIAL CONSERVATION THROUGH DESIGN FOR CORROSION CONTROL AND HIGH PERFORMANCE James T. Skelly Metallurgical Materials Division Texas Instruments Incorporated Attleboro, MA 02703 Abstract Growing concern over the continued availability of critical materials at a reasonable cost has spurred the development of new materials, product redesign, and improved conservation and recycling methods. The pursuit of new materials that reduce or even eliminate reliance upon critical materials can offer tremendous payoffs. Numerous research facilities are concentrating on new alloy development, surface coatings, surface modification, ceramics, and composites, all of which hold promise in reducing dependence upon unstable foreign sources for critical materials. Clad metals provide another approach. By precisely engineering a clad metal system, unique functional benefits can be achieved that are otherwise unavailable or available only with heavy reliance on critical materials. This paper will focus on the design, development, application, and performance of precision clad metals. Specific unique properties and corrosion control benefits will be discussed as will feasible alloy combinations. Cladding Technology The modern process for production of clad metal strip is focused upon continuous cold roll bonding of two or more strips. Prior to bonding, the individual strips of metal are rolled to precise gauges and extensively cleaned to provide contaminant- P39-1 free surfaces. When passed through the specially-designed high pressure rolling mill, a composite material is formed, deriving its bond integrity from the merging of the atomic lattices of the two metals into a common structure. Subsequent thermal treatment is performed to promote diffusion, improve bond strength, and provide stress relief for further cold work operations. Finish- ing operations, such as rolling to intermediate and final gauge, annealing, buffing or polishing, and slitting ensure matching clad metal properties to specific and exact product requirements. These precision clad metals are manufactured by a highly refined process which is based upon the technology of solid state welding. No intermediate brazing alloys or adhesives are used to achieve a permanent bond. It is important to distinguish clad metals from plated metals. Most importantly, the layers in the clad metal system are wrought metal as opposed to thin and porous plating layers. Also, the thickness of the individual clad metal layers can be precisely controlled over the entire composite metal thickness as opposed to the well-known practical limitations to plating thickness. Figure 1 compares the qualities of cladding versus plating. Generally, most metals and alloys can be clad. Figure 2 illustrates that many ductile metals and alloys available in wrought form are in production today. Numerous other combinations are technically feasible but require development before commercialization. The Clad Metal Approach At first glance, direct substitution of a clad metal system P39-2 to conserve critical materials can be envisioned as a surface layer of an alloy containing critical materials clad to a widely available, low cost base metal. In fact, origins of modern clad metal manufacture addressed conservation of precious metals such as gold or silver. Direct substitution of clad metals for critical materials, however, can only offer a partial solution. The usage of the critical material is only reduced and not eliminated in the end product. The process to clad the critical material will offset much of the apparent economic advantage of the critical material content reduction. We should not be looking for substitutes but should be looking for alternate approaches. Optimal application of clad metals for conservation of critical materials is derived from obtaining a longer useful life via corrosion control or improved performance of the end product without the heavy reliance upon critical materials. Metals Design for Corrosion Control Controlling the effects of corrosion with the use of clad metals necessitates utilization of the galvanic series, which is based upon the electrochemical potential of a metal and a reference cathode. Figure 3 illustrates a typical galvanic series in seawater where a less noble metal would corrode at accelerated rates when coupled electrically to a more noble metal. The application of the galvanic series in design of clad metal systems can result in enhanced corrosion control performance. The corrosion control mechanisms and example applications will be discussed below. 1 . By cladding a thin layer of highly corrosion-resistant P39-3 metal over a less corrosion-resistant and lower cost metal, it is possible to achieve enhanced and reliable corrosion resistance. The less noble core layer provides the mechani- cal strength at a low cost. The system differs from plating in the thickness and nonporosity of the noble layer. An example of this approach, called noble metal cladding, is stainless steel clad aluminum for truck bumpers. The highly corrosion-resistant type 301 stainless steel is 10 times the thickness of chromium plating over a steel or aluminum substrate. As shown in Figure 4, laboratory and field testing has confirmed that defects in the chromium plating layer or damage during installation or service which results in a small anode to cathode ratio, can lead to shorter life for a conventional plated steel or aluminum bumper. The stainless steel clad aluminum bumper has demonstrated beyond a reasonable doubt that it can meet the million-mile-life demands of the heavy duty truck bumper. 2. The combination of a more noble metal clad to a less noble metal, where the purpose of the less noble metal is to protect the more noble metal by corroding preferentially, is an example of a sacrificial corrosion system. This principle has been widely applied in the automotive industry where stainless steel clad aluminum trim has replaced solid stainless steel trim. As illustrated in Figure 5, the attachment of stainless steel directly to the automotive body will promote accelerated corrosion of the body steel. Conversely, as shown in Figure 6 the aluminum P39-4 cladding protects the automotive body from corrosion, by corroding preferentially to both the steel and the stainless steel, thereby extending the life of the vehicle. Solid aluminum trim could provide similar corrosion protection of the body steel, but aluminum itself is far less corrosion resistant than stainless steel, and would rapidly oxidize, detracting from the vehicle's appearance. 3. By cladding a less noble layer over a more noble layer, a corrosion barrier is formed where the more noble inner layer begins to corrode only after the outer layer has been extensively corroded. By designing sufficient mechanical strength for the application into the inner layer, a material system has been created that greatly exceeds the corrosion resistance of the individual components. A clad system of carbon steel clad to both sides of a core of stainless steel for automotive brake tubing demonstrates this corrosion barrier mechanism. Figure 7 compares the corrosion behavior of carbon steel tubing and carbon steel clad stainless steel tubing. While the conven- tional terne coated carbon steel tubing deteriorates rapidly due to localized corrosion, the corrosion in the clad tubing is arrested at the stainless steel layer due to the galvanic protection provided by the carbon steel. 4. As an insert between dissimilar metals, a clad tran- sition material eliminates the opportunity for destructive corrosion at the joint. It also simplifies attachment by welding. The joining of aluminum to steel, difficult by welding and doomed to crevice corrosion failure, can be P39-5 accomplished by placing an insert of aluminum clad steel. The aluminum component is welded to the aluminum layer and the steel is welded to the steel layer. The presence of the metallurgical bond eliminates the opportunity for crevice corrosion. 5. For a material system to withstand differing corrosive environments on each side of a strip requires a monometal or alloy to resist the corrosive effects of both environments. An alternate approach is to develop a clad metal system with each layer designed to resist the corrosive effects of the individual environment. An example of a complex multilayer clad system has been developed for anode caps for battery button cells. The inner copper layer is compatible with the internal chemistry of the battery cell while the outer nickel layer resists the corrosive effects of the external environment. The middle stainless steel layer provides the formability and mechanical strength. As illustrated in Figure 8, another example is a three-layer system of titanium clad copper clad nickel for a bipolar electrode in fuel cells. The nickel is required for its resistance to hydrogren permeability on the cathode side. The highly corrosive environment necessitates titanium on the anode side. The copper core is required for its thermal and electrical conductivity to carry electrical current and dissipate heat. Through appropriate design, the five clad metal approaches to corrosion control described above can greatly extend the useful P39-6 life of end products. This extended life itself can have dramatic reduction in the demands for critical materials. Clad Metal Design for High Performance In a similar fashion to corrosion control with clad metals, the combination of materials into a clad metal system can often result in a unique combination of physical properties or properties that might otherwise be achieved: only through the use of expensive or critical materials. Properties, such as strength, ductility, electrical and thermal conductivity, thermal expansion, magnetism, and surface appearance can be tailored to a specific requirements in a clad metal system with exacting precision. A key tool in the development of clad metal sytems is the rule of mixtures, which is a series of calculations used to determine the new composite's properties, based upon the known properties of the individual metals. The rule of mixtures equation can provide a good approximation of the composite's properties in order to narrow the field of potential clad metal systems during the design phase, without actually fabricating the composite . The general equation which can be applied to calculate properties such as density, lateral thermal conductivity, electrical conductivity, and thermal expansion through the thick- ness of the composite is shown below: D rAi x D -J + A2 x D2 + .... An x Dn 100 Where A-|, A2» etc., are the thicknesses, in percent, of the P39-7 components of the clad metal. D-|, D2» etc., are properties of the same components. Properties such as thermal conductivity normal to the surface, thermal expansion along the length, and modulus of elasticity can be calculated with different equations. For fatigue and mechanical strength properties, testing of actual sample specimens or prototypes is recommended. 1. An example of functional benefits offered by a precision clad metal is high thermal and electrical conductivity copper clad to both sides of a low thermal expansion rate 64% iron-36% nickel alloy known as Invar. As illustrated in Figure 9> it is being successfully applied as a printed wiring board core material for direct surface mounting of ceramic chip carriers. In order to achieve more dense electronic packaging, ceramic chip carriers were developed to replace the existing dual inline packages (DIP), thereby reducing component area by factors of 2-5:1. Because the solder joint is both the mechanical and electrical attachment medium for these components, an exact match is required in thermal coefficient of expansion (TCE) between the ceramic chip carrier and the printed wiring board. This is especially critical when wide operating temperatures are encountered such as in military, automotive, and telecommunication applications. Also, the greater packaging density achievable with chip carriers can generate more heat per unit area, which needs to be dissipated to prevent high temperature failure of the integrated circuit. Additionally, strength, P39-8 stiffness, and flatness are key criteria in selection of a printed wiring board material system. Unlike other material systems, copper clad Invar offers a combination of properties that allow the electronics packaging engineer to achieve all the required performance levels. Copper clad Invar is now being designed into appli- cations which had previously utilized cobalt-containing Kovar due to its low thermal expansion rate. With copper clad Invar, it is possible to achieve a thermal expansion rate comparable to Kovar as well as superior thermal and electrical conductivity properties. Figure 10 compares the thermal properties for copper clad Invar to other printed wiring board materials. 2. Another example of the synergistic attributes of clad metals is the application of stainless steel clad aluminum for aircraft firewalls. Titanium had been considered due to its excellent strength-to-weight characteristics and its high temperature performance. However, titanium's cost and availability make it an undesirable choice. Stainless steel clad aluminum provides strength-to- weight properties similar to that of titanium and has passed the stringent FAA flame test (15 minutes at 2000°f). Best of all, this material is readily available at a fraction of the cost of titanium. 3. Another high performance clad metal system is copper clad to both sides of a ferritic stainless steel core for integrated circuit lead frames. Copper clad stainless steel lead frame material has the ductility and strength P39-9 equivalent to or better than the traditional lead frame material, Alloy 42 (nickel-iron alloy), yet provides six times the thermal conductivity of Alloy 42. With integrated circuit devices becoming increasingly complex, they need improved thermal dissipation to avoid high temperature failures. The copper layer provides the excellent lateral thermal conductivity while the stainless steel layer provides the strength and ductility. Copper alloys, which are also being used in high power IC devices, have marginal mechanical properties for many lead frame applications. 4. Heat exchangers for industrial and transportation appli- cations have traditionally been made by inserting copper shims between the structural layers and furnace brazing the entire assembly. The approach can be highly labor intensive in compactly designed units and can also lead to defective braze joints due to improper location of the copper shims. As illustrated in Figure 11, the clad metal approach utilizes a composite of copper clad to both sides of a core of carbon steel or stainless steel. The copper acts as the brazing alloy and thus eliminates the cost of manufacturing and handling the copper shims. The clad metal system also ensures a sufficient presence of copper for requisite braze joint strength. Application of Clad Metals for Critical Material Conservation New alloy development, which holds promise in reducing dependence upon materials such as chromium, cobalt, manganese, and titanium, requires up to ten years and often increases P39-10 material cost or reduces performance. Innovative design and application of clad metal systems can sharply reduce the demand for critical material without sacrificing performance and cost. Additionally, clad metal development can take as little as one year. In most cases, readily available materials can be combined into clad metal systems that yield unique functional properties and corrosion control capabilties. When the U. S. Mint was searching for alternate material systems to silver bearing coinage, cupronickel clad copper clad cupronickel was developed in approximately one year. Furthermore, the clad coins met strict Mint requirements for corrosion resistance, formability, appearance, abrasion resistance, Rho density, reclaimable scrap, and availabilty. Innovative design and production of clad metals have resulted in widespread use in numerous industrial and consumer applications. The capability to engineer a clad metal to operate in hostile, corrosive environments and meet exacting performance requirements can be readily applied to the conservation of critical materials. Our engineering community, designing tomorrow's products, must include potential clad metal systems early in the design and evaluation phase of these new products. When designed in from the beginning, the full payback in terms of material conservation, optimal performance, and inherent corrosion control can be achieved through clad metal technology. Many engineers, unaware of the large volume availability and advanced technology of clad metals, do not consider clad metal systems but specify alloys that consume large amounts of critical P39-11 materials. Often design engineers make trade-offs among engineering properties by accepting some properties that are marginal to retain other more critical properties, or by overde- signing the entire system often using critical materials. This overuse or misuse of materials can led to marginal performance of the end product, shortened life in the operating environment as well as increased cost. An example of this is the use of Kovar (17% Co/29% Ni/ balance Fe) in numerous electronic components utilizing glass-to- metal seals due to its low TCE and glass-matching capability up to 450°C. Kovar, however, is extremely costly, consumes precious cobalt, and is a poor thermal conductor. Copper clad Invar offers a similarly low TCE through 225°C, improved thermal conductivity, a solderable surface, and eliminates the consumption of cobalt. Process modifications to allow copper clad Invar usage might necessitate a shift to a solder seal approach to accommodate the transition of Invar to a higher expansion rate at 225°C. However, it is these kinds of process changes that can reduce demand for critical materials with the added benefit of reducing cost. Summary Material conservation begins with the product design and evaluation stage. The design engineer must comprehend not only the initial product design but also its life cycle-performance requirements. The design engineer must seek innovative designs and processes that reduce dependence upon critical materials. The competitive marketplace demands such innovation; the motivation is lower cost as well as improved performance. P39-12 COMPARATIVE QUALITIES OF CLAD VERSUS ELECTROPLATED COATINGS Clad Laminate Electroplate Advantages Clad layers have low porosity Composition is precisely controlled. Alloys and active metals (Al, Ti, etc.) can be clad. Thickness of cladding is continuously variable and precisely controlled. Properties do not vary with cladding thickness. Complex, multilayer systems can be produced. Limitations Fabricated parts cannot be clad. Cut edges may require special treatment. Brittle metals cannot be clad. Width of stock is limited. Advantages Fabricated parts can be plated. Brittle materials can be plated. Finish brightness can be controlled. Selective areas can be plated. Limitations Deposited metal may be porous. Codeposition (with S, H2, etc- ^ is common. Active metals (Al, Ti, etc.) cannot be deposited Substrates with stable oxides cannot be plated. Thickness is limited by plating time. Properties can vary with plating thickness. Thickness and uniformity can vary. Figure 1 P39-13 GALVANIC SERIES IN SEAWATER NOBLE END Platinum Gold Silver Titanium 316 Stainless Steel (passive) 304 Stainless Steel (passive) 410 Stainless Steel (passive) Nickel (passive) Cupronickel Bronzes Nickel (active) Tin Lead 316 Stainless Steel (active) 304 Stainless Steel (active) Lead-Tin Solder 410 Stainless Steel (active) Cast Iron Steel Cadmium Aluminum Zinc Magnesium ACTIVE END Figure 3 P39-15 P39-16 J *• ^s w 2 'm. C o III o © u *© 15 w m in tr -r-f fa P39-17 ..... . . 1 f* ^m§m. P39-18 CYCLIC IMMERSION (DIP AND DRY) CORROSION TEST OF BRAZED AND TERNECOATED BRAKE LINE TUBING IN SIMULATED ROAD SALT ENVIRONMENT LCS/SS/LCS LCS INITIAL 20,000 CYCLES 40,000 CYCLES 120,000 CYCLES Figure 7 P39-19 -,- WmM Wzmm 1 m ., I i "tr IHS pilli Si ^Hj^j P39-21 ,■■. / ■--■■■.[■-■! ■ '■'■.;-;. y y- ■■':■. ■■■■'■'-.•■■■:>■■-;-■ t-: : .. ..(■■■' ^ y'-yyyyyy "Ji" m ^- to « o « CO 10 OV CD in \6 m zz'ki •*■» "**& : «"MB ■ ' ' / k-:' : -y*:- V"-;- m IBi — X H q ^•■^-^ ■:;:■ fflgit* P39-23 References 1) Hodesblatt, S. , Alternatives to Strategic Materials, Design Engineering, January, 1982, pp 44-49 2) Dance, F.J., Clad Metal Circuit Board Substrates for Direct Mounting of Ceramic Chip Carriers, Electronic Packaging and Production, January, 1982 3) Dance, F.J., Mounting Leadless Ceramic Chip Carriers Directly to Printed Wiring Boards - A Technology Review and Update, Nepcon West, February 24, 1982 4) Bowers, E. W., the U.S. Engaged in Another Kind of War - Strategic Minerals, Iron Age, April 14, 1982, pp 38-41 5) Hurlich, A., Strategic Materials - Technology Trends, Mechanical Engineering, July, 1982, pp 44-53 6) Baboian, R., Designing Clad Metals for Corrosion Control, SAE Trans., 8_[, pp 1763, 1973 7) Baboian, R., Clad Metals in Automotive Trim Applications, Paper 710276 SAE Automotive Engineering Congress, January, 1971 8) Baboian, R., Corrosion Resistant, High Strength Clad Metal System for Hydraulic Brake Line Tubing, SAE Trans., 8j_, pg 1117, 1973 9) Baboian, R., and Haynes, G., Joining Dissimilar Metals with Transition Materials, paper 760714, SAE Automotive Engineer- ing Congress, January, 1976 10) Baboian, R. and Haynes, G., Corrosion Barrier Materials for Communication Industry, Materials Performance, J_6, pg 30, 1977 11) Baboian, R., Controlling Galvanic Corrosion, Machine Design, October 11, 1979 12) Delagi, R., Designing with Clad Metals, Machine Design, November 20, 1980 13) Regan, R., Metals Industries Waiting for Reagan's Minerals Policy, Iron Age, December 16, 1981, pp 31-33 14) Cassidy, R., Strategics: The New Gold?, Sky, August, 1981, pp 10-16 15) Baboian, R., Clad Metals Respond to a Changing Automotive Environment, Body Engineering, pp 69-74, Spring, 1981 16) Verink, E.D., Our Next National Crisis: Materials, Mechanical Engineering, pg 42, August, 1980 P39-24 17) Baboian, R., Conservation of Critical Materials with Clad Metal Systems, Electrochemical Society Transactions 18) Gomez, P. J., A Systems Approach to the Economic Use of Natural Resources, Wire Association Convention, October, 1966 19) Quimby, W. G. , Production Methods, Properties and Applica- tions for Clad Metals, American Society for Metals Meeting, March 21, 1979 P39-25 Autobiography James T. Skelly is a graduate of the University of Rhode Island where he received his B.S. in Industrial Engineering. He also has an MBA from Northeastern University. He has been employed by Texas Instruments in finance, production control in the Control Products Division, and automotive program manager for the Metallurgical Materials Division. Jim is currently product manager in the Metallurgical Materials Division. P39-26 ABSTRACT It has long been known that carbon steel exposed to geothermal brine is aggressively attacked and large corrosion allowances must be made in the design of vessels and piping used in such environments. The economics of geothermal power presents a real need for the use of functional low-cost materials as liners for exposed carbon steel surfaces. Polymer concrete (PC) has been identified as a promising liner material. PC is defined as a concrete in which the aggregate is bound in a dense matrix with a polymer binder. Several high- temperature PC systems have been formulated and tested in the laboratory in brine, flashing brine, and steam at temperatures up to 260°C. Results are also available on field exposures lasting up to 36 months from at least one of six geothermal test sites. Good durability is indicated in all tests. A study has indicated that the use of PC liners as a replacement for the corrosion allowance in carbon steel components for a 50-MWe geothermal plant will reduce the cost of electrical generated power by ~6.2 mills per kilowatt hour. P40-1 INTRODUCTION The demand for the development of geothermal energy increased as the transitory era of cheap imported fuel came to an end. The constant increase in global energy consumption has made even greater demands for the production of a cheaper energy source. One such source being investigated is the extraction of heat energy from the earth, i.e., geothermal energy. Worldwide commercial utilization of geothermal energy was reported to be over 7000 megawatts in 1979. A 1980 survey performed by EPRI indicated that current geothermal electricity generation in North America was 663 megawatts and projected a growth to about 10,000 megawatts by the year 2000. In addition, potential direct utilization of geothermal energy by a wide range of industries was also indicated. * The development of geothermal energy has encountered some serious technical problems in the handling, of hot brine and steam. Hot brine and other aerated geothermal fluids are highly corrosive and they chemically attack most conventional construction materials. Corrosion and scale encrustations have been encountered in all geothermal plants, and to various degrees they adversely affect plant lifetimes and power output. To date, carbon steel has been the primary material of construction but the use of expensive materials such as chrome-moly steel and titanium base alloys may be required for long term operation. Corrosion studies of several different steels exposed to high-temperature geothermal brines were made by the Lawrence Livermore Laboratory at the Salton Sea Geothermal Field. Six-month exposure studies indicated corrosion rates of 1.65 mm/year for mild steel (AISI-1009) while chrome-moly steel (ASTM-A387-9) had a corrosion rate of 0.13 mm/year. Chrome-moly steels, however, cost about four to six times as much as the carbon steels, thereby substantially increasing plant construction costs. P40-2 The availability of low-cost materials with suitable properties will enhance the development of a stable geothermal industry. Brookhaven National Laboratory (BNL) under contract tc the Department of Energy (DOE) has been developing non metallic construction materials suitable for use with geothermal fluids. Calculations indicate that if successfully implemented, the use of materials such as polymer concrete (PC) in geother- mal electrical generating processes could result in a 10 to 20% cost reduc- tion. On the basis of projections of USA direct utilization applications by the year 2000, energy cost savings equivalent to an annual saving of $50xl0 6 to $300xl0 6 have also been estimated. Development of non metallic construction materials for use in geother- mal systems was initiated in the early 70s. Since that time, several high-temperature PC systems have been formulated. Laboratory and field tests have been performed in brine, flashing brine, and steam at tempera- tures up to 250°C. Laboratory data for exposure times of up to 2 years are available. Field test data from six geothermal sites with exposures of up to 3 years are also available. Durability of the PC has been good in almost all cases. The results to date have indicated the potential for the successful use of a specially formulated PC as a lining material for process piping and vessels in geothermal power systems. PC glass filament wound pipe, with low thermal conductivity, can be made for process heat and district heating applications.^ Economic studies performed concurrently with the research program have identified several cost-effective uses for PC in geothermal processes. One study has indicated that the use of PC liners as a replacement for the corrosion allowance in carbon steel components for a 50-MWe geothermal plant will reduce the cost of power by ~6.2 mills per kilowatt hour. Greater savings are indicated if PC is substituted for stainless steel, titanium, or Hasteloy in acid-handling systems, condensate-piping systems, reinjection lines, and steel separators. Uses in cooling towers, district heating systems, and the protection of concrete surfaces also appear to be cost effective. P40-3 PRODUCTION METHODS Polymer concrete is defined as a concrete in which the aggregate is bound in a dense matrix with a polymer binder. The techniques used for mixing and placement are similar to those used for portland cement con- crete, and after curing a high-strength durable material is produced. The most important process variables are monomer and aggregate composition and the aggregate particle-size distribution. Specimens can be produced with compressive strengths up to 207 MPa. Full strength is attained immediately after the polymerization reaction is completed. Polymerization can be accomplished with initiators and promoters at ambient temperatures or with initiators in conjunction with heat. Combinations of both methods are also used to ensure complete polymerization of the monomers used. A. Monomer Formulations Several monomer systems that can be used in high-temperature PC formu- lations have been developed. Some of these will be briefly reviewed. Long-term test data are available for the following systems: 1. 60 wt% styrene, 40 wt% trimethylolpropane-trimethacrylate (TMPTMA) 2. 50 wt % styrene, 33 wt% acrylonitrile, 17 wt % TMPTMA 3. 55 wt % styrene, 36 wt % acrylonitrile, 9 wt % TMPTMA These systems can be polymerized using chemical initiators (benzoyl peroxide, di-tert-butyl peroxide) and heat or by chemical initiators (benzoyl peroxide) and promoters (dimethyl aniline). The styrene-TMPTMA mixture is suitable for the manufacture of PC which is stable to 150°C, while the last two systems can be used at temperatures up to 250°C. B. Aggregate Selection The durability of PC to geothermal fluids is highly dependent upon the composition of the aggregate. Materials such as quartz, silica, flyash, and portland cement have been investigated. All of the aggregates have P40-4 been used successfully In materials at temperatures <210°C. Above this temperature, only PC materials containing a mixture of silica sand and Portland cement have been durable when subjected to brine and steam. Experimental work at BNL indicated that composites formed with a mono- mer system containing styrene(S), acrylonitrile (ACN), and TMPTMA, in con- junction with an aggregate system containing silica sand and portland cement (10 to 40% by weight of aggregate mixture), were stable in 25% brine solutions and steam at temperatures up to 250°C« Bonding of the monomers to the silica sand was enhanced by the use of a silane coupling agent. It was also observed that Type III portland cement gave better results than Type I or IV. This led to the belief that a bond was formed with the vinyl monomers and the tricalcium silicate (highest in Type III cement) which resulted in increased thermal and chemical stability. »-* Later experi- mental work at BNL verified that a bond does exist between vinyl-type monomers and the cement phase, thus cement-containing PC would be expected to have higher thermal stability and long-term durability. The compressive strengths of PC specimens tested at the Geysers, California (Figure 1), and duplicated in laboratory autoclave tests show that they decay for the first 30 to 60 days and then remain constant. Test data have been collected for periods up to 2 years, and some of the data are given in Table 1. Additional data collected from specimens exposed to 25% brine solutions at 177°C in laboratory autoclaves are given in Table 2. The data indicate that the strengths of the samples are essentially constant after the first 60 days of exposure, with samples containing acrylonitrile having higher compressive strengths. C . Manufacturing PC-lined pipes were manufactured at BNL and installed at the U.S. Bureau of Mines Corrosion Facility at Niland, California, in the Salton Sea geothermal field. This is considered to be the most aggressive corrosive fluid known to date. It contains 280,000 ppra dissolved solids in the brine at wellhead temperatures of 240° to 260°C The lined pipe was installed in P40-5 line of the corrosion facility and was in continuous service for ths (Figure 2). The pipe was 1.2 m long and had an inside diameter of 7 . 6 cm . The liner was made of PC containing 55 wt% S, 36 wt% ACN, 9 wt% The aggregate system was 70 wt% graded silica sand, 30 wt% Type Portland cement. The inside of the steel pipe was sandblasted to ove any scale and/ or rust. The inner core was made of rubber (7.6 cm m) and installed in the center of the steel pipe to leave an annulus of Benzoyl peroxide 98 (1%) and di-tert-butyl peroxide (0.5%) irs were dissolved in the monomer mixture. The aggregate system was loaded in a Day rotating blade mixer and the monomer mixture idded. When the composite was thoroughly mixed the PC mortar was in the annulus while the pipe was mounted on a table vibrator, or larger-diameter pipes would have additional vibrators on the pipe When the annulus was filled, the PC was allowed to cure in situ . he initial cure, the inner rubber hose was removed and the lined pe placed in an oven at 150°to 190°C for a final cure. Numerous steel pipes have been lined with PC and tested at various •thermal sites. Pipe diameters have ranged from 7.6 to 30.5 cm and from 1.5 to 2.5 m. At East Mesa, California, PC-lined pipes ire 3) were in service for up to 3 years with no apparent decay. Polymer Concrete Research, Inc. under contract to BNL has developed a fugal casting process that can be used for the commercial manufacture ined steel pipe.' A steel shell is placed in the centrifugal :ing machine and using a lance the inside of the shell is cleaned by asting (Figure 4). Preweighed PC batches are mixed and placed the pipe. End caps are attached and then the pipe is spun, first at pm to distribute the material and then at 1700 rpm for 2 minutes to the PC (Figure 5). One end cap is removed and any excess slurry is A prelilrainary cure at ambient temperature is accomplished with md caps in place. The pipe is then placed in a hot air circulating gure 6) at 190°C and a final cure is attained in ~8 hours. The PC er can be seen in Figure 7 . P40-6 D. Materials Properties Testing Since the steel shell of the pipe can he expected to carry all or most of the mechanical stresses, high physical strength in the PC are not of primary concern. There are several properties that PC must have to provide protection for the carbon steel shell. These include low permeability, good bond strength to the steel inner surface, good corrosion and acid resistance, and low erosion rates. 1. Permeability . Since the purpose of the liner is to provide a barrier between the geothermal fluid carried in the pipe and the mild steel shell, the permeability coefficient of the lining material is important. Studies were made on cylindrical samples 10 mm thick in the apparatus shown in Figure 8. An average permeability coefficient was calculated, using Darcy's Law, to be 22xl0~-'-2 cm/sec. This is about 23 times less perme- able than portand cement concrete. Sections of 15.2-cm-i.d. pipe with a nominal 2.54-cm wall thickness have been cast in PC with formulations previously mentioned. Specimens have maintained fluid pressures as high as 2.75 MPa before bursting or showing any signs of leakage." Attempts at measuring the permeability coefficient of centrifugally cast PC pipes have been unsuccessful because of leakage at the gaskets between the pipe ends and the fixture (Figure 9). 2. Bond Strengths . Bond strength between PC liner and the steel pipe has also been tested. Bond strengths are measured on sections of PC-lined steel pipe in a fixture as shown in Figure 10. Laboratory-cast sections have shown bond strength to average ~5.5 MPa. Sections of a PC-lined pipe that had been exposed to the 280,000-ppm flowing brine at 235° to 250°C for 90 days at the Bureau of Mines Materials Testing Facility at Niland, California, were tested at 25°, 100° , and 200°C. The bond strengths were 10.3, 10. A, and 8.9 MPa respectively. A section of the pipe after the bond test is shown in Figure 11. The ability of the PC liner to withstand the high force required to disbond it from the P40-7 steel and the absence of salts or corrosion products on the steel are indi- cations of the high strength and impermeability of the liner. 3. Corrosion and Acid Resistance . Testing of PC samples in the laboratory and in the field in various geotherraal environments has shown the material able to withstand the most adverse conditions. Loss of strength is generally noted for 30 to 60 days, then the materials stabilize and exposures of up to 3 years in several geothermal environments indicate good durability. The feasibility of using PC materials in high temperature - low pH environments has also been investigated. PC samples have been exposed to a pH 1 HC1 solution at 90°C for more than 440 days. No evidence of deterior- ation of the PC as determined by weight loss or volume change has been detected. Additional tests have been conducted in an environment of pH 1 HC1 at 200°C. PC samples manufactured with several monomer systems with a silica sand - portland cement aggregate system have been exposed for 180 days with no apparent deterioration (Figure 12). 4. Erosion Resistance . No definitive studies have been conducted to determine the erosion capabilities of PC. However, several PC-lined steel pipes have been exposed to flowing brine for periods of up to three years at the East Mesa geothermal site near El Centro, California. The brine flow rate was 226-302 lit/rain at a temperature of 160°C No deterioration, erosion, or scale accumulation was detected upon completion of the test. E. Economic Assessment The Bechtel Corporation has performed conceptual designs for two 50- MWe power plants operating on geotherraal brines. One plant utilized a moderate-temperature low-salinity brine as produced near Heber, California, in the Imperial Valley, and the other utilized a high- temperature, high- salinity brine as produced in Niland, California, near the Salton Sea. In both plants extensive use was made of carbon steel for vessels and P40-8 piping, with Type 316 L stainless steel or titanium used only when carbon steel was considered to be unsuitable. Burns and Roe Industrial Services Corporation (BRISC) reviewed the two designs with the view of substituting, where possible, PC materials for the metals used by Bechtel.-*-^ Each plant was considered separately and estimates of cost savings which could be made in the capital cost of the two plants were calculated. The capital cost estimates covered the battery limits plants and the brine supply and reinjection costs as delineated in the report for the Heber plant. No estimate of the piping was included for the Niland plant because of the limited amount of detail provided in the original report. In the case of the Heber plant, it was shown that a savings of ~$880,000 can be achieved, based on the equipment and lines within the battery limits, and ~$2,800,000 for brine supply and reinjection lines outside the battery limits (Table 3). In the Niland case a savings of ~$550,000 for equipment and lines within the battery limits was shown. These savings in capital costs have a direct impact on the cost of electric power and could result in a cost reduction of 2.72 mills per kilowatt hour for the Heber plant. A similar savings is expected for the Niland plant. It was also estimated that the use of PC-lined vessels and piping would allow for an on-stream time of ~74% versus 70% using unlined carbon steel components. This higher on-stream availability could result in additional savings of ~3.46 mills per kilowatt hour. Thus a net savings of ~6.2 mills per kilowatt hour could be realized by using PC material^ wherever possible. CONCLUSIONS 1. The durability of PC in geothermal fluid environments has been demonstrated by extensive long-terra tests conducted at BNL and at six geothermal sites. P40-9 2. PC may be used in the design of geotherraal or saline water facili- ties to counter corrosion at temperatures up to 250°C. 3. The cost of PC liners in most cases is less than other alternative methods of providing corrosion protection. A. PC, made with properly selected monomers and aggregates, can protect carbon steel surfaces at temperatures above the expected geothermal operating range and preserve structural integrity without the use of corro- sion allowances. 5. The need for exotic materials such as titanium, molybdenum, or austenitic stainless steels (Type 316 L) is reduced or eliminated by using PC-lined carbon steel. 6. Material costs for vessels and piping within the Heber plant battery limits results in a 44% savings through the use of PC linings. 7. Brine supply and reinjection piping material costs can be reduced by 40% using PC linings. 8. The use of PC materials at the Heber Plant can result in a reduc- tion in the cost of electric power production by 2.72 mills per kilowatt hour. 9. A 4% increase in on-stream availability for the Heber plant, due to the use of PC liners, will result in a net cost reduction of ~3.46 mills per kilowatt hour. 10. The reduction in capital cost and anticipated improvements in on- stream availability through the use of PC in the Heber plant will result in an overall saving of ~6.2 mills per kilowatt hour. P40-10 References 1. Anderson, David N. , and Lund, John W. , Editors, Direct Utilization of Geothermal Energy: A Technical Handbook , Geotherraal Resources Council Special Report No. 7, ISSN 0149-8991, ISBN "0-934412-07-03 , 1979. 2. McCright , R.D. , Frey, W. F. , and Tardiff, G. E., Localized Corro- sion of Steels in Geotherraal Steam/ Brine Mixtures, Geothermal: Energy for the Eighties, Transactions, Volume 4, Geotherraal Resources Council, ISSN 10193-5933, ASBN 0-934412-54-5, Davis, California, 1990. 3. Schroeder, J.E., Design and Fabrication of Polymer Concrete Pipe for Testing in Geothermal Energy Processes , Final Report, July 1981. 4. Kukacka, L. E., Fontana, J., Auskern, A., Concrete Polymer Materials for Geothermal Applications, Report No. 6, BNL 20571, July 1975. 5. Kukacka, L. E., et al., Alternate Materials of Construction for Geothermal Applications, Progress Report No. 14, BNL 50751, September 1977. 6. Sugama, T. and Kukacka, L.E., Cement and Concrete Research, Vol. 9, pp. 69-76, 1979. 7. Kaeding, A., Design and Fabrication of Polymer Concrete-Lined Pipe for Testing in Geothermal Energy Processes, Final Report, December 1981. 8. Kukacka, L.E., et al., Concrete-Polymer Materials for Geothermal Applications, Progress Report No. 9, BNL 21665, June 1976. 9. Conceptual Design of Commercial 50 MWe (net) Geothermal Power Plants at Heber and Niland, California. October 1976, Energy Research and Development Administration, Division of Geothermal Energy, SAN-1124-1. 10. Economic Assessment of Polymer Concrete Usage in Geothermal Power Plants, November 1977, Prepared by Burns and Roe Industrial Services Corporation, BNL 50777. P40-11 o o w — 1 5> CO CO D> 3 C~) »«4 r~ -o z: II r~ r- — i i-h 31 II CO o -o 3> > -< > c-> II o 33 CO CO 33 m > T3 H -< z z m 33 r~ m a o *— i o i 1— 1 2 z 2 i— ' 2 m ►"■ « o CO m H H z z X 33 o X. CO -< *— 1 2 H r* r m fr-9 o o m 33 o r- ■ — 1 z -o 2 -< H 33 O "0 > O z m i— i T3 o z > 2 ►— < m Z m ►- * a m 33 • H— * > i o z H 33 > 33 H cr> •—i r cr> 2 > 33 m z m H a cn X > > o H o m m 33 2 -< m CO r- z -< > H CO H • H m m 3; o 33 O O o CO 2 CO "0 1 o r CO >— < m z a 7<: i—< o z TJ CD CjO > O en m * z H H &•* cn cn ic U3 cn o o o o o o o iji ui in ui ui ui ui U1 O Ul O O U"l CD CO 1 CO 1 CO 1 CO 1 CO 1 CO 1 CO v^M Osl V-M V-M Osl V>J 1 cn V^J cn U>J V^J cn Osl > 3s» 3> 3» D> 3> o O C~> O O 3> z: i z: 1 z: i z: i z: i z: I o z: U3 I— • ^vl LO uo o i 3 —\ 3 -H — 1 13 ^>J -o 3 -a 3 3 tj —H -o — i -o -a — I — I 3 -H s: -H — I 3 3: 3> 31 3> 3 31 3> "D o VjsJ VjsJ V-M Vjsl Usj ^j «^j O O -Cr O I— ' un U1 X o O CO > c -< 33 co m 2 m v. X. 3! m O *-* z CD o X 2 H m 33 TJ m CO 33 -< O CO m H z m H 2 >» C - ) o CO 2 H T3 33 33 3 m m -o z CO > o CO H 1— 1 X < v. m HO o r~ •< 2 m 33 o o z o 33 m H m co •v m o > •- H 2 m H Z x co m m CD X m ■< -o o CO CO m m 33 a co > O f .-. o -n s: • »— i > z o t3 33 -< CO H m > N3 OO O O > r m P40-12 oo CO 3> "T3 -o H -o CD O O 33 m CD m o 33 z: z: z 1—4 m o o CD 3 CD • • H m > I z H K5 »— • co CO m v. 2 M m O co V. * s: c s: H H 33 1— ' H &-H ^ m CD ** a CO CO 3 CO H H > 3 •—4 ■< •< H r~ 33 33 a M m m r-o hH O z z CO > > m m o 3 o CO l i • X > z Osl -Cr *wM a CM O 00 i s: « 3 H H 3 o fc-fl fr* |- > — 1 O s: O 13 z H 33 -a CD a^ < -H • 1- ~g -H O 3> -< z • "0 ►— 1 m H 33 i—4 1—4 1—4 r- 1—4 m -o i o 33 i— • H *vl r- > s: z H a fr* o — 1 m 3 3 -o m — 1 z 3 H 3> ID 4? M H en cn oo 4=- en o ai o nj w oj uj w ho w cn j: O O sj h cn ui ui Ui ^ j: N O W N3 N3 oo Ui X "0 o CO c 73 m 3 m > •< CO -a o o o 3 TJ 73 m CO CO < m CO H 73 m z CD *■ -o o is -o z: > o N3 CD O 3 ■v 73 m CO CO m CO H 33 m z CD o Tl CD H m 73 m x -o o CO c 73 m -o m 33 o m 33 •-^ z m > O C~> > CO r* m rs3 P40-13 CD 73 CO 73 GO 73 CD c > *-* 1— 1 H Z z Z CO o m m H— 1 a — 1 7X3 CO 3 m o m c: M H ¥—4 -o H CO > -H Z -a CO > — r~ O c_ r* H o H m -< H H > o m > r H ** 70 •< r CD O z cc • z m • CO CO r- • r* • r— »— * • z m co r™ m o DO > Z < rn T| cz H m m c_ r- 1—4 H CO co z a m > T3 m c H rn o CO 3 73 CO z H X m -< co CO O z H o m 73 -H H |— o TO TO CO > •— t H C (A z 3 > CO > :* i—i r~ CO m 70 Z o o o CO H CO U1 I— ' l—' CD 1— ' oo 1— • v. «. v. o CD en 1— ' V-M i— ' CO 4S o o 73 i»h CD i—i Z > VJ- CD ■"^J CD I—" v Si s. Si NO rvo 1— ' o K5 ho OJ CO J=- -«=■ jr o Si Si s. Si VJSI CD oo o »— ■ jr CD oo U1 O CD o v> Si v. CD CD CD 4=- CD OO Si U1 CO V-M ^4 N3 CD Si o Si CD O Si 00 4^ CD CD 4=: OO o o z o TO m H m "a o i- -< 3 m 73 ** •t* ■W- un XT v-M s s. v 4=- Us) CD ^4 OO CD U1 1— ' U1 ^4 4=- *^J Si Si v Si un N3 *— ' o H- « t— • oo CO CD Ul *vj oo <*♦ ■fe* ■to* W Usl ro N> \i «• ». ^4 00 4S 04 VM un OO -vj OO -^i o •^4 v Si s. s. ^ ^4 CD O CD NJ ^>1 U1 CD V41 Osl N3 ■to* ■f» I—' I—* Si Si t-> O h-> 1— 1 o I—" OO h-" CD Si Si Si V/J >>l v-n CD CD Vjs) 1— ' V>J oo 1 4=" un !—• 1— • Sj v. w N3 KJ1 CD V-M ■^-4 CD U1 NO NO <»* <*♦ OO 4=r 4S- oo CD *»4 t— ' CD 4=- Si Si Sj CD UsJ **4 "^J 45- NO 4=r CD Ul CO > < 1—1 z CD CO H X -o o m 73 o o z o 73 m H m > r* m V-M co > < 4=- I *vj OO • • Wl 4? OO CD 4=- CD V>J CD • CD 4S- CD V>J 4=- CD 4=" 4=" oo CD OO CD v» CD &^ 73 co m > a < c *- o Z H o *-* CO o P40-14 List of Figures Figure 1 - PC samples tested at the Geysers, Calif, in flowing dry steam at ~235°C Figure 2 - PC lined pipe installed at the Bureau of Mines Corrosion Facility in Niland, Calif. Flowing geotherraal brine >280,000 ppm dissolved solids at ~240°C. Figure 3 - PC lined pipe installed at East Mesa Geothermal Site near El Centro, Calif, flowing brine ~160°C. Figure 4 - Sandblasting inner surface of steel pipe prior to application of PC liner. Figure 5 - Centrifugal casting of PC liner in steel shell. Figure 6 - Curing of PC liner in hot air circulating oven, Figure 7 - Centrifugal cast PC liner. Figure 8 - Permeability fixture for flat plates. Figure 9 - Permeability fixture for PC pipe. Figure 10 - Fixture used to measure bond strength of PC liner to steel sheel. Figure 11 - PC liner removed from steel pipe after testing at Bureau of Mines Corrosion Facility. Figure 12 - PC samples exposed to pH 1 HC1 at 200°C. P40-15 & c M P40-16 CM o M P40-17 >) »i —i a— i*i CO w Pi :=> o M P40-18 o I* -7 ;*'•/',- *•'«•?; :: -; _ P40-19 m w o M P40-20 W 5 o M P40-21 M P4 P40-22 00 o M Ix. P40-23' Pressure Gau ge fr co Q- 8 -J CO o a. o. ID CO 1 — - r CTS E3 // W ' J r / /'zzrz o u u o a. t- to o 0> r— IS) (o ro x: or- *4— +-> C »r- C o c o a> 00 •—• •— » I n n n i< .n L»_ Q_ T3 -J a: »— w erf ^> o H p4 ro P40-25 O M P40-26 «S M P40-27 ELECTROLESS NICKEL AS A SUBSTITUTE FOR CHROMIUM PLATING IN INDUSTRIAL APPLICATIONS Ronald N. Duncan ELNIC, Inc. Nashville, Tennessee In the last decade American industry has become almost totally dependent upon foreign supplies for strategic metals and minerals. While presently we import only 42 percent of the petroleum we use, more than 80 percent of the chromium, cobalt, manganese, niobium, tantalum ^ and platinum required for industry is supplied by foreign producers. Of these strategic metals, chromium is probably the most critical. Chromium is indispensable for the manufacture of stainless and alloy steels and is essential to the metal finishing industry. The United States presently _ consumes 25 percent of the world's production of chromium and has no domestic reserves. Worse yet, 98 percent of the world's chromium supply is located in southern Africa, Because of this area's political instability, a stoppage or reduction of its production is very possibled ,2) . In the National Materials Advisory Board's study on "Contingency Plans for Chromium Utilization" (3) , electro- less nickel was cited as the only true substitute for chromium. All other suggestions were techniques to merely reduce, not eliminate, chromium usage. Electro- less nickel was also described in this report as a probable improvement over chromium in many applications This paper describes this coating and its properties and discusses how it can be used as a substitute for industrial chromium coatings and for other more expense or less readily available materials. AN AMORPHOUS MATERIAL Electroless nickel is an alloy of nickel and phosphorus. Those coatings used for functional (rather than decorative) applications typically contain 10 to 11 percent phosphorus dissolved in nickel and less than 0.05 percent other impurities. Unlike chromium or other electrolytic coatings, electroless nickel is completely amorphous. P41-1 It has no crystal structure and contains no segregation or separate phases. Its lack of long range order has been confirmed by electron diffraction studies at magnifications of 150,000(4). An example of this coating is the 75 pm (3 mil) thick deposit shown in Figure 1. It is the lack of structure of this alloy which produces its unusual properties and makes it well suited for protection against corrosion, erosion and wear. Some of the properties of electroless nickel are compared to those of commercial hard chromium coatings in Table 1. Because of their purity and homogeneity, the internal stress of high phosphorus electroless nickel coatings is very low on most substrates. As shown by Figure 2(5), on steel the coating is typically compressively stressed at 4 MPa (0.5 ksi) . This helps to ensure the deposits integrity and the coatings performance. The internal stress of commercial hard chromium coatings is always very highly tensile, and usually exceeds 200 MPa (30 ksi). This, combined with chromium's brittleness, causes the coatings to be cracked and often produces fissures through its thickness to the substrate. An example of the cracks in hard chromium deposits is shown in Figure 3. A similar problem occurs with electroless nickel coatings containing less than 10 percent phosphorus. As illustrated by Figure 2, the internal stress level of low phosphorus deposits is also high. This also causes these deposits to be cracked and porous. The discontinuities present in both chromium and low phosphorus, electroless nickel coatings reduce their strength, ductility, and wear and corrosion resistance. UNIFORMITY PROVIDES MANY BENEFITS Unlike electroplated coatings, electroless nickel is applied without an electric current. Instead, the coating is produced by autocatalytic chemical reduction. Electroless nickel is plated onto a substrate by reducing nickel ions to metallic nickel with sodium hypophosphite. This chemical process avoids the non-uniformity associ- ated with most other metallic coatings. The thickness of electroplated coatings like chromium will vary significantly depending upon the part's configuration and proximity to the anodes. The coating tends to buildup on corners, edges and the like, and to be reduced on internal surfaces. Not only do these variations effect P41-2 the ultimate performance of the coating, but they can also cause additional finishing or machining to be needed after plating. With electroless nickel, the coating thickness is the same on any area of the part that is exposed to fresh plating solution. Grooves, slots, blind holes, and even the inside of tubing will have the same amount of coating as the outside surface of the part. The benefits of electroless nickel's uniformity are illus- trated by its substitution for chromium on many of the cylinders and rolls used in the printing and textile industries. This change has reduced chromium usage, and has also significantly reduced the finishing costs of these components. Previously the cylinders had to be ground, plated to a thickness of about 250 ym (12 mils) , and ground a second time, before they could be balanced and installed. With electroless nickel they are now ground only once, balanced, plated to the desired diameter -- usually with 25 to 37 ym (1 to 1% mil) -- and installed. Not only has this reduced the cost of plating by 40 percent, but 55 percent of the grinding time is also saved, freeing the machines for the production of new parts and increasing productivity (6). The thickness of electroless nickel coatings can be controlled to suit the application. Coatings as thin as 2% ym (0.1 mil) are commonly applied to electronic compo- nents, while those as thick as 75 ym (3 mils) are typical for chemical or petroleum equipment. Coatings thicker than 250 ym (10 mils) are also easily applied, but because of cost are normally used only for salvage or repair of worn or mis-machined parts. GLASS-LIKE PROPERTIES The mechanical and physical properties of electroless nickel deposits resemble those of other glasses. They have high strength, limited ductility and relatively low conductivity. The ultimate tensile strength of high phosphorus deposits exceeds 700 MPa (100 ksi) . This is equal to many alloy steels and 5 to 10 times higher than chromium. This allows the coating to withstand consider- able abuse without damage. Accordingly, it is commonly used to protect compressor blades, turbines, valves, pumps, extruders, blowers, and the like, and has often replaced stainless steel and exotic alloys. The ductility of electroless nickel is only about 1 to 1% percent. While this is less than that of most alloys, it is adequate for most coating applications and much superior to chromium. Thin films of the deposit can be bent P41-3 completely upon themselves without fracture, and the coating has been used successfully for springs and bellows. Electroless nickel, however, should not be applied to articles which subsequently will be bent or drawn. Severe deformation will crack the deposit, reducing corrosion and abrasion resistance. The electrical and thermal conductivity of electroless nickel is low. Its conductivity can be increased by heat treatment, but is still much less than that of conventional conductors like copper or silver. Because of the relatively thin layers used, however, for most applications its resis- tance is not significant. Electroless nickel coatings are being successfully used for such applications as exchanger tubing and electrical switches and contacts. Electroless nickel can be easily soldered, braised, and bonded and is often used to ease joining of non-metals and aluminum and stainless steel. This, combined with its uniformity and corrosion resistance, have made electroless nickel an ideal coating for electronic components. Accord- ingly, it is being increasingly used to reduce or eliminate gold and precious metal requirements in the electronic industry (7 ) . PROVIDES EXCELLENT RESISTANCE TO WEAR One of the most important properties of electroless nickel for many industrial applications is its hardness and wear resistance. As deposited, high phosphorus coatings have a microhardness of 480 to 500 VHN100. approximately equal to 48 HRC. This is similar to many hardened steels. Heat treatments, similar to age hardening procedures for aluminum alloys, can produce significant increases in coating hardness. As shown by Figure 4, hardness values as high as 1100 VHN100 (approximately 69 HRC) can be produced. This is equal to the hardness of commercial hard chromium and comparable to that of some hard facing alloys and ceramics . Hardening of electroless nickel is due primarily to the formation of nickel phosphide particles within the alloy. At temperatures above 260°C (500°F) coherent and then distinct particles of M3P begin to form, and at tempera- tures above 320°C (600°F) the glass begins to crystallize. This causes its hardness and wear resistance to increase rapidly. Maximum hardening is obtained through treatments at 400°C (750°F). P41-4 Both heat treated and non-heat treated electroless nickel coatings are commonly used to minimize the effects of erosion, abrasion and wear. Laboratory tests have shown fully hardened coatings to have wear resistance equal to hard chromium. This is illustrated by Table 2, which shows the results of Taber Abraser Wear tests of electro- less nickel coatings, and compares them to electroplated nickel and chromium(o). The excellent resistance of electroless nickel often allows it to replace high alloy materials, hard chromium, and even hard facings. An example of the ability of electroless nickel to make common materials behave like superior ones is the poly- ethylene pelletizer bowl shown in Figure 5. This bowl, with a 125 ym (5 mil) thick high phosphorus coating, was in service for two months with no measurable loss or attack. Previous uncoated aluminum bowls failed in less than three weeks after 40 percent of their weight was eroded away. Other typical wear applications include feed screws and extruders , computer drive mechanisms, textile and fiber equipment, hydraulic cylinders, molds and dies, and packaging equipment. Another property related to wear is lubricity. Under sliding or abrasive conditions, low friction surfaces minimize heat buildup and experience less scoring and galling than higher friction surfaces. Because of the phosphorus they contain, electroless nickel deposits have a low coefficient of friction, typically 0.13 (lubricated) to 0.4 (unlubricated) . This is approxi- mately 20 percent lower than chromium, one-half of that of steel, and significantly better than aluminum or stainless steel. One disadvantage of electroless nickel is that, unlike chromium, it is not hydrophilic. Electroless nickel lacks the crack pattern present in chromium which can hold an oil or water film. Thus for applications like ink rollers, which depend upon surface wetting, electro- less nickel may not always be suitable. ADHESION STRENGTH IS OUTSTANDING The performance of a coating is often dependent upon its adhesion to its substrate. Without adequate adhesion even the most resistant coating can become dislodged or broken, allowing attack of the underlying metal. The adhesion of electroless nickel coatings to steel, aluminum, copper and their alloys normally exceeds the sheer strength of the substrate. P41-5 The high bond strength of these coatings is due to the ability of the plating solution to completely remove microscopic contaminants from the substrate prior to the deposition of the first nickel-phosphorus layer. This allows the coating to develop both mechanical and metallic bonds with the substrate metal. The .adhesion of electroless nickel coatings to steel and aluminum is typically 300 to 400 MPa (40 to 60 ksi) . This is similar to and often superior to that of chromium coatings. PROVIDES SUPERIOR CORROSION RESISTANCE One of the most important differences between electroless nickel and chromium coatings is their corrosion resis- tance. Both are barrier coatings. Both protect the underlying metal by sealing it off from the environment rather than by sacrificial action. Thus, to be completely protective the coatings must be defect free. Because of the cracks present through even thick deposits, hard chromium coatings offer only limited protection against corrosion. These cracks offer pathways through the coating for a corrosive environment to reach and attack the substrate. Accordingly, chromium coatings will oftentimes develop a network of rust spots or rust lines across its surface. In addition, unlike electroless nickel, chromium coatings are subject to interface or underdeposit attack. With chromium, corrosion not only will extend into the substrate at the bottom of cracks or pores, but will also travel out from these areas along the coating-to-substrate interface, loosening and lifting the coating(9) . Because of its homogeneity and freedom from defects, electroless nickel coatings provide a true barrier to corrosion. They do not offer any pathways to the substrate, The metallic bonds these coatings form with their substrate also prevent underdeposit attack. Even if a pore were to be produced by improper processing or mechanical damage, in most environments it rapidly fills with corrosion product, stifling further attack. Corrosion will not spread out from the defect and thus it is contained. High phosphorus electroless nickel coatings are almost totally resistant to alkalis, like caustics and potash; to salt solutions and brines, like sea water or those present in food or chemical environments; to acid gas environments, like those in the petroleum industry; and to all types of organic media and solvents. The coating also has good resistance to ammonia solutions; to organic P41-6 acids, like lactic or acetic; and to reducing acids, like hydrochloric or sulfuric. It is only signifi- cantly attacked by strongly oxidizing solutions like concentrated sulfuric or nitric acid(lO) . In most environments, the corrosion resistance of hard chromium is much less than that of electroless nickel. Chromium is rapidly attacked by reducing environments and is subject to pitting and localized attack in halogens, especially oxidizing halogens like ferric chloride. This is illustrated by Table 3, which compares the corrosion of electroplated and cast chromium to that of electroless nickel in different environments (9 ,10 ,11) . The petroleum and chemical process industries are the largest users of electroless nickel for corrosion protection. Because of its superior resistance to attack in these environments, the coating ensures easy and reliable operation and extends the equipment life. Accordingly, it is often used in place of more critical or expensive materials, especially stainless steel. For instance, oil field valves coated with 75 ym (3 mils) of electroless nickel cost approximately one-third that of equivalent stainless steel valves and in most environ- ments provide equal protection(12) . Many petroleum components now coated with electroless nickel were originally plated with hard chromium. The poor performance of chromium in oil field environments, however, lead to a reduction in its use and to the increased specification of electroless nickel(13). Specific applications in these industries include ball, gate, plug and check valves, blow out preventers, chokes, heat exchange equipment, pumps, compressors, tubing, vessels, packers, and all types of down hole equipment (12) . CONCLUSION Electroless nickel has many unique properties which make it a superior engineering material. The coating offers high strength, excellent abrasion and wear resistance, lubricity, solderability , and superior corrosion resis- tance, together with ease of application and uniform thickness. Accordingly, electroless nickel has proved to be useful in improving reliability, in reducing cost, and in avoiding the use of critical or strategic materials. RND:SS October 4, 1982 P41-7 REFERENCES 1. Gray, A.G., Metal Progress , Vol. 117, No. 3, p. 33 (1980). 2. Swinburn, J., Materials Performance , Vol. 21, No. 1, p. 54 (1982). 3. National Materials Advisory Board, "Contingency Plans for Chromium Utilization", National Academy of Sciences, Washington, D.C., 1978. 4. Weil, R. , Stevens Institute of Technology, private communication, January 24, 1980. 5. Parker, K. and Shah, H. , Plating , Vol. 58, No. 3, p. 230 (1971). 6. Paliotta, J.V. , "Functional and Economic Impact of Electroless Nickel on the Printing Press Industry", Electroless Nickel Conference II, Cincinnati, March, 1981. 7. Bandrand, D.W. , Plating and Surface Finishing , Vol. 68, No. 12, p. 57 (1981). 8. Industrial Nickel Plating and Coating, International Nickel Company, New York, 1976. 9. Uhlig, H.H. , Editor, Corros ion Handbook , John Wiley and Sons, New York, 1948, p. 825-828. 10. Duncan, R.N. , "Corrosion Control With Electroless Nickel Coatings", AES Electroless Plating Symposium, American Electroplating Society, St. Louis, March, 1982. 11. LaQue , F.L., and Copson, H.R. , Corrosion Resistance of Metals and Alloys , Reinhold Publishing, New York, 1963, p. 448-449. 12. Duncan, R.N., "Performance of Electroless Nickel Coatings in Oil Field Environments", Corrosion/82 Conference, National Association of Corrosion Engineers, Houston, March, 1982. 13. King, J. A. , and Badelek, P.S.C., Oil and Gas Journal , Vol. 80, No. 28, p. 115 (1982). P41-8 TAPLE 1 COMPARISON OF ELECTROLESS NICKEL AND COMMERCIAL HARD CHROMIUM COATINGS PROPERTY ELECTROLESS NICKEL COMMERCIAL HARD CHROME MATERIAL STRUCTURE INTERNAL STRESS ON STEEL, MPa DENSITY, g/cm3 MELTING POINT, °C ELECTRICAL RESISTIVITY, wft -cm THERMAL CONDUCTIVITY , W/cm-°K MAGNETIC COERCITY TENSILE STRENGTH, MPa DUCTILITY, I ELONGATION MODULAS OF ELASTICITY, GPa COEFFICIENT OF THERMAL EXPANSION, ym/m/°C ADHESION STRENGTH, MPa HARDNESS, VHNiqo COEFFICIENT OF FRICTION VS STEEL (LUBRICATED) TABER WEAR RESISTANCE, . mg/1000 cycles CORROSION RESISTANCE Alloy of 10 to 11 percent dissolved in nickel. Amorphous; no phase structure, lamination or segregation. <7 7.75 890 90 0.08 Non-magnetic >700 1 to 1% 200 12 300-400 480 to 500, as deposited; heat treatable to 1100 0.13 15 to 20, as deposited; 2 to 9 after heat treatment Excellent resistance to attack by all but the most severely oxidizing environments. Chromium plus trace amounts of oxides and hydrogen. Crystalline; fine grained with numerous cracks. 200-300 6.90-7.18 1610 14-66 0.67 Non -magnetic <200 <<0.1 100-200 6 Good 800 to 1000 0.16 2 to 3 Poor due to cracks; resists oxidizing environments; attacked by halogens and reducing solutions . RND:ss October 4, 1982 P41-9 TABLE 2 COMPARISON OF THE TABER ABRASER RESISTANCE OF DIFFERENT ENGINEERING COATINGS. COATING HEAT TREATMENT TWI , mg/ 1000 CYCLES (1) Watts Nickel None 25 Electroless Nickel None 17 Electroless Nickel 300°C/1 hr 10 Electroless Nickel 500°C/1 hr 6 Electroless Nickel 650°C/1 hr 4 Hard Chromium None 2 (1) Taber Wear Index, CS-10 abraser wheels, 1000 gram load, determined as average weight loss per 1000 cycles for total test of 6000 cycles. RND:ss October 4, 1982 P41-10 TABLE 5 COMPARISON OF THE CORROSION BEHAVIOR OF CHROMIUM AND ELECTROLESS NICKEL IN DIFFERENT ENVIRONMENTS ENVIRONMENT TEMPERATURE 10% Acetic acid 10% Citric acid Cone. Hydrochloric acid 10% Hydrofluoric acid 10% Lactic acid 10% Malic acid 10% Nitric acid Cone. Nitric acid 10% Phosphoric acid 10% Sulfuric acid Cone. Sulfuric acid 10% Sodium hydroxide 10% Ammonium chloride 10% Cupric chloride 10% Cupric nitrate 10% Ferric chloride 10% Sodium chloride CORROSION RATE, ym/y CHROMIUM 12°C(D 16* ■C<2) nil 660 nil -- -- 100,000 25,000 -- nil -- 51 -- nil -- nil -- 25 5 280 -- 760 -- nil -- nil -- 380 -- 51 -- nil -- nil _ _ ELECTROLESS NICKEL 20°C(3) 25 19 46 30 19 17 44 >25,000 16 12 25 nil nil 25 12 780 0.5 (1) Commercial hard chromium deposit; corrosion rates less than 25 ym/y reported as nil. (2) Cast chromium metal. (3) Electroless nickel containing 10%% phosphorus and less than 0.05% other elements. RND-.ss October 4, 1982 P41-11 FIGURE 1 Typical high phosphorus electroless nickel deposit on a mild steel substrate. The coating thickness is 75 urn (3 mils) . Lighter layer at top is a copper overplate applied for edge resolution. 400X magnification. Nital etchant. FIGURE 2 Effect of phosphorus content on the internal stress of electroless nickel deposits on steel. (1 ksi = 6.89 MPa) o o o m cc t- v> -J < z cc 15 10 H -10 -15 Tensile ^^^^^^ Compressive 5 6 7 8 9 10 11 12 13 PHOSPHORUS CONTENT, PERCENT P41-12 FIGURE 3 Typical cracks in a commercial hard chromium deposit. 1000X magnification, Not etched. FIGURE 4 70- Effect of different 68- one hour heat 66- treatments on the hardness of electro- 64- less nickel containing 10%% phosphorus. « 62- o > 1 60- I 58 - 56- 54- 52- 50- 48-1^ 1 1 i i ■ i i i • 100 200300400500600700800 800100011001200 Temperature D e gree « F P41-13 FIGURE 5 Plated aluminum pelletizer bowl, after two months of exposure to high velocity, polyethylene pellets and Mississippi River water. Bowl is coated with 125 ym of high phosphorus electroless nickel. Previously, uncoated bowls failed in 3 weeks with weight losses of 40 percent. P41-14 DEVELOPMENT OF DUCTILE POLYCRYSTALLINE Ni 3 Al FOR HIGH-TEMPERATURE APPLICATIONS* C. T. Liu and C. C. Koch Metals and Ceramics Division Oak Ridge National Laboratory Oak Ridge, Tennessee 37830 ABSTRACT The nickel aluminide, Ni3Al, is strong and stable at elevated temperatures; however, low ductility and brittle fracture restrict its use for structural applications. Microalloying has been employed to overcome the embrittlement associated with grain-boundary separation. Among the dopants added, boron has been found to be most effective in improving fabricability and ductility of polycrystalline Ni3Al. Ten- sile elongations of >50% have been achieved by control of boron concentration and thermomechanical treatment. The B-doped Ni3Al shows excellent strength and oxidation resistance at high temperatures. Unlike conventional alloys, the yield strength of the aluminide increases rather than decreases with increasing test temperature. The potential for developing the aluminide as a substitute for Cr-containing heat-resistant alloys is assessed. INTRODUCTION Austenitic stainless steels and superalloys are the common heat- resistant materials for structural use at elevated temperatures. The alloys typically contain substantial quantities (15~30 wt %) of chromium for oxidation and corrosion resistance. Because of the strategic nature of the chromium supply, there has been an increasing interest in deve- lopment of aluminides as a substitute for Cr containing alloys. JL Research sponsored by the Exploratory Studies Program, Oak Ridge National Laboratory and Division of Materials Sciences, U. S. Department of Energy under contract W-7405-eng-26 with the Union Carbide Corporation. P42-1 Nickel and iron aluminides are in a class of materials usually referred to as intermetallic compounds. The aluminides are typically strong, stable, and resistant to oxidation and corrosion 1 at elevated temperatures. Unlike conventional alloys, the static strength of many aluminides shows an increase rather than a decrease with increasing test temperature. >° A major problem with using aluminides as structural materials is their reported brittle fracture and low ductility, par- ticularly at lower temperatures. '^ Because of the poor fabricability and low fracture toughness, the aluminides have not been considered as serious candidates for structural engineering applications. The nickel aluminide based on Ni3Al (Y") is an important strengthening constituent of commercial nickel-base superalloys. It has been known for years that single crystals of Ni3Al are quite ductile, but its polycrystalline forms are extremely brittle. > The brittleness of such poly-crystals is associated with weak grain boun- daries that cause brittle intergranular fracture without much plastic deformation within the grains. During the past 20 years, efforts have been spent on improving the ductility of aluminides, » 10 but significant progress has been achieved only recently through microalloying processes. Microalloying involves adding a small amount of dopants (usually <1%) for controlling chemistry and cohesion of the grain boun- daries. In this study, a microalloying approach was employed to alle- viate the grain boundary embrittlement problem of Ni3Al. Emphasis will be placed on the effect of boron additions on ductilization of Ni3Al alloys. EXPERIMENTAL PROCEDURES The Ni 3 Al alloys containing 26 and 24 at . % Al (13.9 and 12.7 wt % Al) were doped with a small amount of Ce (0.1 wt %) , Mn (2.1 wt %) , and B (0 to 0.1 wt %) . The alloys were prepared by arc melting and drop casting, using pure metal elements and Ni— 4 wt % B and Ni— 4 wt % Ce master alloys. There was no significant change in weight during the alloy preparation. The alloy ingots (2.5 x 5.3 x 0.6 cm) were first P42-2 homogenized at 1000-1200°C, and then fabricated to sheets by either hot rolling at 1200°C or cold rolling at room temperature with intermediate annealing at 1000°C. The mLcrostructure of Ni 3 Al alloys annealed for various times at 1000°C was examined metallographically . The specimens polished by stan- dard techniques were etched in a solution of 20 parts of H2O, 20 parts of HN0 3 y 10 parts of HF, 20 parts of H3P0i+, 10 parts of acetic acid, and 10 parts of HC1. The crystal structures of Ni 3 Al alloys were determined by x-ray diffraction using Cu Y^ radiation. Sheet specimens with a gage section of 12.7 x 3.2 x 0.7 mm were used for tensile and creep tests. Tensile tests were conducted on an Instron testing machine at a crosshead speed of 150 mm/s. To perform the tensile tests at elevated temperature, a water-cooled quartz-tube vacuum system was attached to the Instron machine, and specimens were heated inductively inside a tantalum susceptor. Creep tests were done in vacuum under dead-load arrangement. During the tests, the tem- perature was monitored by a Pt/Pt— 10% Rh thermocouple centrally located on the specimen. Fracture surfaces of selected tensile specimens were examined by a JSM-U3 scanning electron microscope (SEM) operated at 25 kV. RESULTS Undoped M3AI alloys containing 26 and 24 at. % Al cracked badly during hot and cold fabrication. Figure 1(a) shows part of the ingot hot rolled at 1200°C. The ingot was almost pulverized due to extensive cracking along the grain boundaries, indicating the grain-boundary brittleness in Ni 3 Al. The alloys doped separately with 0.1 wt % Ce, 0.1% B, and 2.1% Mn all cracked during fabrication except the B-doped Ni-24 at. % Al alloy (12.7 wt . % Al), designated as IC-6. It was possible to fabricate the IC-6 ingot into sheet by repeated cold rolling at room temperature and heat treatment at 1000° C. The amount of cold work was initially -15% reduction in thickness, and was gradually increased to 50% between each intermediate annealing. Figure 1(b) shows the 0.76-mm-thick sheet of IC-6 fabricated by cold rolling. P42-3 (a) CENTIMETER 1 lllll lH Il (b) Fig. 1. Comparison of fabricability of Ni 3 Al alloys containing 24 at. % Al; (a) A part of Ni 3 Al ingot hot rolled at 1200°C, showing extensive grain-boundary crack- ing; (b) A sheet of Ni 3 Al doped with 1000 ppm B, fabri- cated by cold rolling at room temperature. P42-4 To study the effect of B additions on fabricability , a series of alloys was prepared based on Ni-24 at. % Al, in which various levels of B were added. Table 1 summarizes the results. The alloys containing 100 ppm B or less cannot be fabricated by cold rolling. IC-18 con- taining 250 ppm B was fabricated into sheets, but it exhibited numerous shallow surface cracks. The alloys containing 400 ppm and above were fabricated into good quality sheets. Table 1. Effect of boron additions on fabricability a of Ni-24 at. % (Ni-12.7 wt %) Al ah B Addition ,,.,,. Alloy - . Fabricability IC-2 ' Cracked badly IC-20 50 Cracked badly IC-19 100 Cracked IC-18 250 Sheet fabricated shallow surfac IC-21 400 Sheet fabricated IC-15 500 Sheet fabricated IC-28 700 Sheet fabricated IC-6 1000 Sheet fabricated a Ingots were fabricated by repeated cold rolling (15 ~ 50% reduction in thickness) and softening heat treat- ment at 1000°C. The crystal structure in undoped and B-doped Ni3Al alloys was determined by x-ray diffraction. The superlattice lines, which were clearly visible, characterize the Ll2~type ordered crystal structure. Thus, the small amount of B does not affect the long-range ordered crystal structure in Ni3Al. The microstructure of B-doped Ni3Al alloys (24 at. % Al) was examined as a function of aging time at 1000°C. The alloys show a "regular" equiaxed grain structure with few annealing twins and second-phase particles (Fig. 2). Grain growth occurs at P42-5 (a) , -^ "^ a VMS' / O s a O -^ U 3 • «w >4k (b) ^ \ \ \ / / ~\ — S<: ** if 67^*7 Fig. 2. Microstructures of Ni 3 Al doped with 1000 ppm B. (a) 10 mln anneal at 1000°C; (b) 16 d anneal at 1000°C. P42-6 1000°C; however, there is no apparent difference in grain size for specimens containing 500 and 1000 ppm B after the same annealing treatment . Tensile properties of B-doped Ni3Al alloys were characterized as functions of B concentration, heat treatment, and test temperatures. Table 2 shows the room-temperature properties of the alloys doped with different levels of B. For the specimens annealed 30 min at 1000°C, the yield strength showed a linear increase with the B concentration [Fig. 3(a)] but the tensile strength was essentially insensitive to the B concentration. All the B-doped alloys exhibited transgranular ductile fracture [like Fig. 4(a)] with the ductility exceeding 40%. The tensile elongation increased with the B content [Fig. 3(b)] and reached 53.8% for the IC-6 specimen containing 1000 ppm B. After being annealed for 16 d at 1000°C, the IC-6 specimen showed a higher yield strength but distinctly lower ductility. In accompanying the decrease in ductility, the fracture mode changed from the transgranular to a mixed one [Fig. 4(b)]. On the other hand, the IC-15 specimen containing 500 ppm B displayed a lower yield strength with ductility remaining unchanged after the long term anneal. Figure 5 is a plot of yield strength as a function of test tem- perature for the B-doped Ni3Al (IC-15) and commercial fabricable alloys such as Hastelloy-X and type 316 stainless steel. Unlike the conven- tional alloys, the strength of IC-15 increases with increasing tem- perature and reaches a maximum around 600°C. As a result of this increase, the aluminide is much stronger than the commercial alloys at elevated temperatures. For example, the aluminide displays a yield strength of 550 MPa (80,000 psi) and a ductility of 45% at 600°C. In comparison, the yield strengths of Hastelloy-X and type 316 stainless steel are 210 and 120 MPa (31,000 and 17,000 psi), respectively. Thus, the aluminide has the superior high-temperature strength. Table 3 shows the results of preliminary measurements of creep pro- perties of the B-doped alloys tested at 760°C (1400°F) and 138 MPa (20,000 psi). The minimum creep rate of the aluminides is insensitive to the B concentration between 250 and 1000 ppm. IC-6 specimens exhi- bited a sharp drop in creep rate with an increase in grain size from P42-7 0) 4-t CO a) to 4->-*-. co B o o r-l m r-l .-I CO M-l H O CO O 00 H CI o rH W .£ 4J 60 C! ^ CU «H J-i CO 4J ^i cn v,/ cu CO rH cu •H 2 CO c CU H 4J C -H ai co ^ J*! 4J ^ W CO *T3 Pk -h s cu C o •H 4J CO 4-1 B c a cu a o ^ c o u « J-i >-, cu ■h B <3 !z 00 CO m 0> vD O m oo CO m CO m o 00 to o o o K "^ /-s CO CO 00 CM On i— 1 /-N • • • • • • 00 on 00 o 00 r-l ON • 00 00 CN 00 00 «tf on r-l r-l r-l 1-1 r-l r-l r-» s_/ V— ' S_/ s— ' v_^ \~s v—^ CO <• r». CO CO 00 • • • • • • • sr r^ «sf v£> CO r^ ON O ON r-l ON m o o o m o o o 00 i-4 r-l CM m l-H 00 CM \0 u C_> c_> U O m 1-4 vO u u P42-8 ORNL-DWG 82-1328 (a) (b) o Q. I I- o UJ or l- V) Q _J UJ > 35 < o z o -J UJ LU .J CO z UJ 400 5 300 200 100 60 40 20 Ni 3 AI (24at%AI) DOPED WITH B ADDITIONS. 1 200 400 600 800 1000 BORN CONCENTRATION (PPM) Fig. 3. Yield strength and tensile elongation of B-doped M3AI alloys (24 at. % Al) as a function of boron content. P42-9 10 (a) (b) Fig. 4. SEM fractographs of B-doped Ni3Al (1000 ppm B) tested in tension at room temperature. (a) Specimen annealed for 3 d and 15 h at 1000°C. (b) Specimen annealed for 16 d at 1000°C. P42-10 11 YIELD STRENGTH (MPa) 700 600 500 400 300 200 100 - 316 STAINLESS STEEL 200 400 600 800 TEMPERATURE (°C) 1000 100 90 80 70 60 YIELD STRENGTH 50 (ksi) 40 30 20 10 Fig. 5. Yield strength as a function of test temperature for Ni3Al (24 at. % Al) doped with 500 ppm B and commercial fabricable alloys Hastelloy X and type 316 stainless steel. P42-11 12 <=>- C co x: CD ^ U ui U I 00 -* CO 00 o o o • • • • • • i H I-l 3 On 1-1 •* CM CO 3 v ~' «tf CO m oo 2 o •d 2 r-t •H <0 iH X & CO JJ CO 5 T3 oo CO •H 4J rfi Z CO ^"\ CD o ■OU /^ CD « « « CN CD -H N /»-s r-\ ^•v i-l &.X CO ■H O CO S m u"» in v-/ o a CN CN 1-1 •o >^ 3. ^-> ^-/ N-* CD 1 o o C V CO WHO •H CD CD CD u ■H O CO C a C CO (4-10) « fc •H •H •H O 0*JO CO CM CO Cti ^ O ft Pn P* U CD 32 •H CO 4-> T3 Ph • p c g e O CD <0 o o • a oo •H O CJ OH CO ■U O o U CD i-l CO O O a- cd U <-^ i-l O 4-1 73 u B O O O O O an s c a ITl o O O 4J i-l CD <0 cd a CN m O O <0 CD CO o ^> i— i i-l ■U U CO y\ c C CO O CD Pn o B *C3 I-l o CJ <4-l C O O -H O « O vo CO •H >H J-l i-l iH r-l r-t com o CD CO CO a. vo CD CD CD m r» 4J c a o a i CO S S C_> >s CO CO 4J CO CO >. >» •X CD .O .O co 4J iH •H C •X3 *T3 • > •~\ •"N •H CD CD CD 00 m ^-v ^s CO O O H i 1—1 i-i vO vO 4J M 3 3 .O >> ( l | 1 CO *o *o CO O o o C_J CJ ►» O O H H M H (-4 M \£> o t-i P iH "w / ^^ V-* V-* i-l iH Ph P-. <3 CO i-H e i-c i— 1 <3 iH iH ■H CD <3 10 c a a a < Pseudoplastic Bingham Plastic St. Venant Body (Slope = -1) 10 Shear Rate 100 APPARENT VISCOSITY AS A FUNCTION OF SHEAR RATE FOR VISCOUS FLUIDS Figure 4. P43-22 EFFECT OF VISCOSITY ON THE MOLD FILLING CHARACTERISTICS Figure 5, P43-23 10 5 O Q_ V) O o to > 103 1 2l I L 57.0 v/o v/o 32.0 v/o -L_U L 10 100 1000 Shear Rate, Sec"^ 6a. Viscosity of Polyethylene at Different Levels of Silicon Nitride Loading 10 5 % 10* o Q_ w O o V) > 10 3 10 2 - /\ - v , No Modifying Agents >s "^^^^^^57.0 v/o _ % X ^With Modifying Agents X "^•^^^ 60.0 v/o — ^s» ^.With Modifying Agents - "^^^^ 57.0 v/o I I III I i i i I i iii! i 10 100 Shear Rate, Sec~1 1000 6b. Effects of Modifying Agents on Silicon Nitride Loading and Viscosity Figure 6. Viscosity Versus Shear Rate for Various Mixes P43-24 Q- Q_ o Figure 7a, P43-25 cc 111 o o LU _l O N LU o ~3 z z Figure 7b. P43-26 Plunger Fwd Mold Closed 13 sec. 22 sec. $ E z \ \ Sealing Point D F \V ' t Residual T \ \^ -X' B Mold Pressure s A Time >■ Figure 8 P43-27 SEM PICTURE OF INJECTION MOLDED AND SINTERED SILICON NITRIDE ON THE FRACTURE SURFACE Figure 9 P43-28 DEVELOPING AN INFORMATION STOCKPILE TO AID IN SUBSTITUTION PREPAREDNESS Robert T. Nash Vanderbilt University Introduction During the workshop the attendees had the opportunity to provide their written views on the acquisition, organization and dissemination of information describing alternative technologies which could be used if shortages of critical materials occurred. Thoughtful comments were obtained which in combination provide a coherent picture of the issue of stockpiling information. The following five elements of the information stockpile issue were identified in the responses from the workshop participants which are: Types of Information Explaining Alternative Technologies ° Uses and Sources of Information Explaining Alternative Technologies Channels for Disseminating Information ° Influential Organizations Qualification of a Material Each of these five elements will be discussed separately. Types of Information Explaining Alternative Technologies There was a consensus that the following types of information should be included in a substitution preparedness stockpile. Case Histories by Application Substitution Plans by Companies Structure - Property Relations Substitution Tables Invariably, changes in processes are associated with any change in the use of materials. Consequently, an alternative technology consists of both a change P44-1 in material and the associated processes which are employed in the production and subsequent fabrication of the material into a finished part. In addition to technical and economic information the participants felt that administrative information would be necessary. In particular, government agencies should keep industrial organizations informed about plans which would be employed during an emergency if a shortage of materials occurred. It was recognized that much useful information is proprietary and hence not likely to appear in any publication. Further, information concerning processes is usually not amenable to publication in any but the most rudimentary form. Hence, the workshop participants thought that it would be useful to develop lists of people that are expert on alternative technologies who could be contacted during an emergency. Uses and Sources of Information Explaining Technological Alternatives An important distinction exists between changes in materials for which qualification is required and those for which it is not. At one extreme a change can be made in a tool steel through a manufacturer's decision. At the other, materials used in nuclear power systems and aircraft can only be changed after lengthy qualification procedures. Only those technological alternatives which have undergone at least limited application in an industrial organization would be used by industry during an emergency. Therefore, a consensus supported the position that only a technology which has actually been employed in serial production, or at a minimum in prototype testing, should be included in the information stockpile. Any changes in the use of critical materials during an emergency must be simple, or they will not be implemented due to lack of time. P44-2 Technological alternatives which are used by industry often originate in government laboratories and universities. However, this work is only technological information in process until applied to the manufacture of actual materials and finished parts by industry. Laboratory results, no matter how promising, would be of little or no value in an emergency. Channels for Disseminating Information Existing channels of information transfer should be employed to the greatest extent possible in disseminating information on substitution preparedness. The regular sources include: Journals ° Conference Proceedings Handbooks ° Government Publications It was also the view of the workshop participants that these regular sources of information should be supplemented. The principal means would be to publish abstracts of papers already appearing in the open literature. These abstracts could be made available through existing electronic information retrieval systems as soon as they are prepared. In addition, hard copy summaries of the abstracts could be published annually. When a suitable body of information had been accumulated on a particular topic, publication of a monograph or handbook would become appropriate. Influential Organizations A number of public and quasi public organizations were mentioned which would play a role in the organization of a materials substitution stockpile. These include: NMAB Department of Commerce ASTM Department of Interior MPC (Bureau of Mines) AISI Department of Defense P44-3 To guide the development of an information stockpile, an advisory group could be organized which would include representatives from industry and each of the influential organizations listed above. This group could provide advice on the most effective means of organizing a stockpile of information pertinent to substitution preparedness. Qualification of a Material A crucial element in determining what information should be included in an information stockpile is the role of qualification procedures. These procedures help determine the length of time required to make any type of change in the use of a material. Therefore, the role of qualification in substitution preparedness should be evaluated carefully. The workshop participants felt that an alternative technology should only be included in the information stockpile after qualification of the material. Findings Concerning an Information Stockpile The following findings were obtained during the workshop concerning the organization of an information stockpile. 1) Since qualification procedures greatly influence the period of time required to change the use of a material, only that information concerning materials which have been qualified should be included in the stockpile where qualification is necessary. 2) Only that information concerning technologies which have been carried through prototype testing should be included in the stockpile where qualification is not necessary. 3) Existing channels for disseminating primary information should be employed to the greatest extent possible. The existing channels should P44-4 be augmented with Substitution Preparedness Abstracts which could be used to locate primary information. 4) Those organizations which can influence the adoption of new type' ">f materials should be instrumental in the development of the information stockpile. 5) Much information is never published. Hence, the names of people should be made available who would provide advice on alternative materials and processes for use during an emergency. 6) Information should only be included in the stockpile if the substitute technology can be implemented within a reasonable period of time. If too much time is required, the substitute technology will not be used in an emergency. P44-5 NBS-114A (rev. 2-ec) U.S. DEPT. OF COMM. BIBLIOGRAPHIC DATA SHEET (See instructions) ,. PUBLICATION OR REPORT NO. NBSIR 83-2679-2 2. Performing Organ. Report NoJ 3. Publication Date July 1983 4. TITLE AND SUBTITLE Technical Aspects of Critical Materials Use by the Steel Industry Volume II B: Proceedings of a Public Workshop; "Trends in Critical Materials Requirements for Steels of the Future; Conservation and Substitution Technology for Chromium". 5. AUTHOR(S) Edited by R. Mehrabian 6. PERFORMING ORGANIZATION (If joint or other than NBS. see instructions) NATIONAL BUREAU OF STANDARDS DEPARTMENT OF COMMERCE WASHINGTON, D.C. 20234 7. Contract/Grant No. 8. Type of Report & Period Covered 9. SPONSORING ORGANIZATION NAME AND COMPLETE ADDRESS (Street. City. State, ZIP) U.S. Department of Commerce, National Bureau of Standards U.S. Department of the Interior, Bureau of Mines U.S. Department of Defense, Army Research Office 10. SUPPLEMENTARY NOTES J Document describes a computer program; SF-185, FlPS Software Summary, is attached. 11. ABSTRACT (A 200-word or less (actual summary of most significant information. If document includes a significant bibliography or literature survey, mention it here) This volume presents papers given at a public workshop sponsored jointly by the National Bureau of Standards, Bureau of Mines, Army Research Office, and Vanderbilt University. The workship "Critical Materials Requirements for Steels of the Future; Conservation and Substitution for Chromium," was held at Vanderbilt University, October 4-7, 1982; it featured 50 presentations by technical authorities from steel producing and using industries. Volume I of this publication draws extensively on these proceedings and reviews technical opportunities for research in process improvement and in alternative materiald development that would reduce U.S. dependency in critical materials. The advanced technologies reviewed in Volume I in addition to their implications for critical materials conservation represent trends leading to better quality, lower cost steel products, and therefore they may contribute positively to the industry's more immediate concern for improved markets. 12. KEY WORDS (Six to twelve entries; alphabetical order; capitalize only proper names; and separate key words by semicolons) Advance coatings; alternative materials; chromium; processing; steel; strategic materials 13. AVAILABILITY | | Unlimited [XI Fo r Official Distribution. Do Not Release to NTIS ~ 1 Order From Superintendent of Documents, U.S. Government Printing Office, Washington, D.C. 20402. 3| Order From National Technical Information Service (NTIS), Springfield, VA. 22161 14. NO. OF PRINTED PAGES 415 15. Price USCOMM-DC 8043-P80 PENN STATE UNIVERSITY LIBRARIES A0D0071flfll5M c |