V. K. Sikka and M. L. Santella

, I PROCESSING AND OPERATING EXPERIENCE OF Ni,ALBASED INTERMETALLIC ALLOY IC221M V. K. Sikka and M. L. Santella Metals and Ceramics Oak Ridge National Laboratory P.O. Box 2008 Oak Ridge, Tennessee 37831 and J. E. Orth United Defense LP 2101 West loth Street Anniston, Alabama 36201 ABSTRACT The cast Ni,Albased intermetallic alloy IC221M is the most advanced in its commercial applications. This paper presents progress made for this alloy in the areas of: (1) composition optimization; (2) melting process development; (3) casting process; (4) mechanical properties; (5) welding process, weld repairs, and thermal aging response; and (6) applications. This paper also reviews the operating experience with several of the components. The projection for future growth in the applications of nickel aluminide is also discussed. INTRODUCTION Development of intermetallics for structural applications have been tallred about in the scientific literature [l141 for nearly two decades. From the thousands of possible intermetallic compounds [15] that exist based on elements in the periodic table, the aluminides of nickel, iron, titanium, and molydisilicide have shown the most potential for commercial applications. Among the aluminides, a cast Ni,Albased alloy, developed at the Oak Ridge National Laboratory ( O W ) and known as IC221M, has reached the most significant stage of commercialization. The purpose of this paper is to provide the processing details, properties, and operating experience for this alloy. *Research sponsored by the U.S. Department of Energy, Assistant Secretary for Energy Efficiency and Renewable Energy, Office of Industrial Technologies, Advanced Industrial Materials Program, under contract DEACO596OR22464 with Lockheed Martin Energy Research Corp. SP

DISCLAIMER This report was prepared as an account of work sponsored by an agency of the United States Government. Neither the United States Government nor any agency thereof, nor any of their employees, make any warranty, express or implied, or assumes any legal liability or responsibiuty for the accuracy, completeness, or usefulness of any information, apparatus, product, or proces disclosed, or represents that its use wouid not infringe privately owned rights. Reference herein to any specific conunem'al product, process, or service by trade name, trademark, manufacturer, or otherwise does not necessan'ly constitute or imply its endorsement, recommendation, or favoring by the United States Government or any agency thereof. The views and opinions of authors expressed herein do not necessarily state or reflect those of the United States Government or any agency thereof.

DISCLAIM= Portions of this document may be illegible in electronic image product& h a g s are produced from the best available original dor?umf!lll! '

COMPOSITION The IC221M is a cast alloy with composition [15161 based on controlled additions of boron, chromium, zirconium, and molybdenum to the Ni,Al base. The nominal composition of IC221M, given in Table 1, can be produced in the laboratory environment with good control of each element. However, the alloy production under commercial environment resulted in the following described issues which have been dealt with for commercial applications: 1. Melting procedures not known for such high alumiprwn content alloy. This issue stems from two aspects of commercial melting practices. First, only small aluminum additions are typically used in commercial alloys. Such aluminum additions are made to the melt near the end of melting for the purposes of deoxidation. Large additions of a l d u m such as in IC221M were feared to chill the melt and mostly oxidiz~and go to slag. The oxidation process was expected to cause p r chemistry control The second concern for melting was based on the assumption that the only way to melt an alloy with such high aluminum content is to load the entire aluminum content at the beginning of the melt. In this scenario, the large melting point difference between duminum and nickel will cause overheating of the aluminum prior to the start of the nickel melting. The overheated aluminum of very high fluidity will seep through any cracks in the furnace crucible and attack the induction coil. Such an attack of the induction coil will bring molten aluminum in contact with water and result in an explosion. Both of these concerns have been eliminated with the development [17191 of a melting process known as ExoMeltm. The key features of this melt practice are described in some detail in a later section. 2. Recovery of various elements. The oxidation of aluminum, zirconium, and boron during air melting were considered of concern for their recovery. However, our work with melting of many experimental heats have demonstrated that the use of the ExoMelt process results in essentially 100% recovery of all alloying elements except zirconium, which requires a loading of 120% in the melt stock. 3. Pickup of elements other than nominal from the commercial melt stock. The commercial grade melt stock tends to introduce impurities such as carbon and sulfur. A systematic study of a large number of heats to determine the effect of impurity elements on the weldability has allowed us to set acceptable limits of carbon, sulfur, silicon, and boron. The pickup of silicon and iron during the foundry practices is described in issue 2

No. 4 below. Since limited opportunity occurs to remove these impurities during airinduction melting, carbon and sulfur contents can be kept to a minimum by proper selection of the melt stock. The contents of carbon and sulfur may also be controlled by starting with a new furnace lining or using pure nickel wash heat prior to melting the nickelaluminide alloy. 4. Pickup of certain elementsfrom the foundry environment. The elements silicon and iron are commonly picked up when melting nickel aluminide in an ironbased alloy foundry. The silicon is picked up from the fact that nickel aluminide tends to m c t with the sand (SiO,), especially when extra metal is poured into sand pig molds. If the pig is used in melting revert heats without removing the sand, the aluminum and zirconium contents of nickel aluminide reduces SiO, to silicon, which is picked up by the alloy. The silicon pickup from this source can be reduced by two approaches: (1) to pour the extra metal in Zro, washcoated pig molds (such a wash eliminates the reaction of molten IC221M with sand); and (2) if sand is stuck to the pigs, it must be removed by grit blasting prior to melting the revert. Another source of silicon pickup is by the attack of molten nickel aluminide on SiO, in zirconia... crucibles. Silicon pickup from this source can be lllllllfillzed by two steps: (1) use A&03as a furnace crucible or lining rather than zirconia, and (2) minimize the time that molten IC221M stays in contact with the crucible. The contact time can be reduced by proper scheduling of melting and casting of heats so that the molten metal does not set in the furnace for long periods waiting for it to be poured. The iron pickup can occur in at least two different ways. First, being in an iron foundry, there is always a chance of a small piece of iron or steel from other heats getting into the nickelaluminide melt stock. Pickup &om this source can be minimized by careful controls in the foundry practice. The second source for the pickup of iron is the steel liner that is typically used for setting a new furnace lining. The best method to reduce such a pickup is to run a wash heat of nickel after setting the new lining and prior to melting IC221M. The pickup of iron over 1 wt % has the tendency to precipitate the beta phase, which lowers the hightemperature strength of IC221M. 5. Melting of foundry revert stock. The foundry revert stock consists of several items: (1) pigs cast from metal left over after casting all planned molds, (2) runners and risers removed from the castings, and (3) any I defective castings that are beyond weld repairing. Any of these three revert stocks have a potential of adding 3

impurities to the alloy. The potential of silicon from the pigs has already been described above. The runners and risers can also add silicon if all of the sand stuck to the castings is not completely removed. The remelting of the castings can also add iron impurities if some of the steel beads become trapped in the defective area. Proper removal of any sand from the runners and risers and trapped steel beads can minimize the pickup of silicon or iron fkom the foundry revert stock. 6. Oxidation of molten metal. The zirconium in the IC221M is the first element to oxidize followed by... aluminum,if molten metal is exposed to air for long periods of time. Such oxidation can be rmrlliz11zed using an argon cover during &induction melting. Similarly, the oxidation of molten metal flowing through the sand molds can be minimized by flushing the molds with argon prior to casting. MELTING PROCESS The most significant progress towards commercialization of IC221M has occurred through the development of the ExoMeltm process for melting of IC221M using the currently available melt technology at most foundries. The process is based on the fact that the formation of aluminides from their alloying elements is exothermic [17191 and uses a furnaceloading sequence to effectively use the heat of formation in the melting process. The most significant benefits of the ExoMeltm process for melting IC221M include: 1. It permits the commercial melting of IC221M and uses the currently available foundry equipment. 2. It produces reproducible alloy chemistries. 3. For virgin heats, it save 50% energy and 50% time as opposed to the conventional process, which industry was not willing to try for reasons listed above. 4. The energy and time savings and extended crucible life result in 50% lower cost than the conventional process. 5. One ORNL licensee has melted over 22,700 kg of the IC221M alloy in 1996, and two licensees have melted over 22,700 kg so far in 1997. 6. The use of the ExoMeltm process has helped accelerate the commercialization process. 4

CASTING PROCESSES The IC221M produced by the ExoMeltm process can be cast into complex shapes by the static sand mold casting process. It can be cast into tubes and pipe by the centrifugal process. The components of nickel aluminide can be produced by welding the centrifugal and static castings. The alloy can also be investment cast into ceramic molds. Examples of sand casting of IC22 1M are shown in Fig. 1. These castings vary sigruficantly in section thickness from 6.35 mm for the heattreating fmture to 203 mm for the forging die. The casting complexity also increases from a simple shape such as pallet tip to very complex heattreating fixtures. Not only can the complex shapes be cast, but they can also be produced with good dimensional control as demonstrated in Table 2. PROPERTIES The physical and mechanical properties and corrosion data on nickel aluminide alloys were compared with commercial alloys in a recent paper [20]. The characteristic physical and mechanical properties [21J of the IC221M alloy are listed for a range of temperahms in Table 3. WELDING The continuing effort at ORNL has demonstrated that the cast IC221M can be successfully welded for either the repair of castings or fabrication of components. The gas tungsten arc is the best welding process, and IC221LA and IC221W are the two recommended filler wires. The composition for IC221LA in weight percent is 4.5 Al16.0 Cr1.2 Mo1.5 Zr0.003B76.8 Ni, and the composition for IC221W in weight percent is 8.0 Al 7.7 Cr1.4 Mo3.0 Zr0.003 B79.9Ni. The welding process development has progressed to a point of being able to successfully weld largediameter centrifugalca&pipe to sandcast trunnions. The welding success of such large components has resulted from optimized weld design, a tight control of base metal chemistry, and the use of stainless steel backing ring to minimize root pass cracking. The weldment properties reported elsewhere [22] show that tensile properties are similar to base metal with somewhat lower ductility and creep rupture strengths, 70% of the base metal. 5

NICKELALUMINIDEAPPLICATIONS The applications of nickel aluminide continue to grow. They range from heattreating fixtures for carburizing and oxidizing environments, rolls for steelhardening furnaces, dies for hotforging applications, tube hangers for hightemperature processing furnaces, and components for calcination furnaces. A new application is developing in the area of radiant burner tubes for furnace heating. FUTUREPROJECTIONS The calendar year 1996 has been an excellent year for commercialization of nickel aluminides. Over 22,700 kg of the IC22 1M alloy were melted and cast into a variety of components. The performance of 1996 has extended into 1997. This is evidenced by the extension of a single foundry licensed in 1996 to three licensees at the time of this writing. The current application development activity suggests that the production of IC221M is expected to double in 1997 and has the potential to grow to five times during 1998. CONCLUSIONS The cast nickelduminide alloy IC221M is well on its way to commercialization. This paper has presented the compositions, melting, casting processes, properties, welding, applications, and future projections. ACKNOWLEDGMENTS The authors wish to thank C. Randy Howell for coordinating the testing and data analyses, Robert W. Swindeman and Craig A. Blue for paper review, and Millie L. Atchley for preparing the manuscript.

REFERENCES 1. C. C. Koch, C. T. Liu, and N. S. Stoloff, eds., High Temperature Ordered Intemtallic Alloys, 39, Materials Research Society, Pittsburgh, 1985. 2. N. S. Stoloff, C. C. Koch, C. T. Liu, and 0. Izumi, eds., High Temperature Ordered Intermetallic Alloys II,81, Materials Research Society, Pittsburgh, 1987. 3. C. T. Liu, A. I. Taub, N. S. Stoloff, and C. C. Koch,eds., High Temperature Ordered Intermetallic Alloys 111, 133, Materials Research Society, Pitsburgh, 1989. 4. L. A. Johnson, D. P. Pope, and J. 0. Stiegler, eds., High Temperature Ordered Intermetallic Alloys IV,213, Materials Research Society, Pittsburgh, 1991. 5. I Baker, R. Darolia,J. D. Whittenberger, and M. H. Yoo, eds., High Temperature Ordered Intermetallic Alloys V, 288, Materials Research Society, Pittsburgh, 1993. L 6. J. A. Horton, I. Baker, S. Hanada, R. D. Noebe, and D. S. Schwartz, eds., High Temperature Ordered Intermetallic Alloys VI,Materials Research Society, Pittsburgh, 1995. 7. S. H. Whang, C. T. Liu, D. P. Pope, and J. 0. Stiegler, eds., High Temperature Aluminides und Intermetallics,The Mineral, Metals and Materials Society, Warrendale, 1990. 8. S. H. Whang, D.P. Pope, and C. T. Liu, eds., HighTemperature Aluminides and Intermetallics, Proceedings of the Second TMS/ASM Symposium, Mater. Sci. Eng., A15WA153 (1992) 746. 9. Izumi, O., ed., Intermetallic Compounds Structure and Mechanical Properties, Japan Institute of Metals, Tokyo, Japan, 1991. 10. C. T. Liu, R. W. Cahn, and G. Sauthoff, eds., Ordered Intennekrllics Physical Metallurgy and Mechanical Behavior, Kluwer Academic Publishers, Dordrecht, The Netherlands, 1992. 11. Y. W. Kim and R. R. Boyer, eds., MicrostructurdPropertiesRelationships in Titanium Aluminides and Alloys, The Mineral, Metals and Materials Society, Warrendale, 1991. 12. R. Darolia, J. J. Lewandowski, C. T. Liu, P. L. Martin, D. B. Miracle, and M. V. Nathal, eds., Structural Intermetallics,The Mineral, Metals and Materials Society, Warrendale, 1993. 13. N. S. Stoloff and V. K. Sikka, eds., Physical Metallurgy and Processing of Intermetallic Compounds, Chapman and Hall,New York, 1995.

14. G. Welsch and P. D. Desai, eds., Oxidation and Corrosion of Intemtallic Alloys, Purdue University, West Lafayette, 1996. 15. K. Aoki and 0. IzUmi, Nippon Kinzoku Gakkaishi, 43 (1979) 1190. 16. R. L. Fleischer, Intermetallic Compounds Structure and Mechanical Properties, Proceedings of JIMIS6, ed. 0.Izumi (Tokyo: Japan Institute of Metals, 1991), 15763. 17. C. T. Liu, C. L. White, and J. A. Horton, Acta Metall., 33 (1985) 213. 18. V. K. Sikka, S. C. Deevi, and J. D. Vought, Adv. Mater. Process., 147 (1995) 29. 19. S. C. Deevi and V. K. Sikka, Intermetallics, 96 (1997) 17. 20. S. C. Deevi and V. K. Sikka,Intermetallics, 4 (1996) 357. 2 1. V. K. Sikka, Commercialization of Nickel and Iron Aluminides, Nickel and Iron Aluminides: Processing, Properties, and Applications, eds., S. C. Deevi, P. J. Maziasz, V. K. S u a, and R. W. Cahn (Materials Park ASM International, 1997), 36176. 22. M. L. Santella, An Overview of the Welding of Ni3AI and Fe3AI Alloys, Nickel and Iron Aluminides: Processing, Properties, andapplications, eds., S. C. Deevi, P. J. Maziasz, V. K. Sikka, and R. W. Cahn (Materials Park: ASM International, 1997), 32128.

1. Examples of sand casting of IC221M alloy. FIGURE CAPTION

f I

Table 1. Comparison of nominal chemical analysis of IC22 1M with the range observed for heats made using virgin and revert stock in a pilot commercial melt run of 94 heats carried out at Alloy Engineering & Casting Company Element Virgin heats (wt %) Nominal (wt %) Range Revert heats (wt %.) Average Range Average Al 8.0 7.5 8.2 7.86 7.3 8.3 7.74 Cr 7.7 7.63 8.11 7.8 1 7.56 8.5 7.88 Mo 1.43 1.38 1.50 1.45 1.34 1.56 1.43 zr 1.70 1.73 2.02 1.93 1.62 2.05 1.86 B 0.0080 0.004 0.0054 0.003 0.008 0.0054 0.008 C 0.012 0.032 0.022 0.01 0.05 0.024 Si 0.021 0.055 0.036 0.026 0.155 0.06 1 Fe 0.03 0.15 0.077 0.03 0.91 0.194 Ni 81.1 50%virgin and 50% revert. %dance. b 80.8 1 b 80.8 1

... Table 2. Variability of casting dimensions for a pilot commercial run of 63 sets of IC221M castings Average k standard deviation Oength) Average k standard deviation (width) Tray 672.7 k 0.7 672.5 2 1.2 Lower fixture 657.7 k 1.8 Component Upper fixture 632.8 f 2.2 657.4 f 1.9 632.3 f 2.0

.,I Table 3. Physical and mechanical properties of Ni,Albased cast alloy IC221M Temperature ("C) Room 200 400 600 800 900 lo00 1100 7.86 30 260 270 280 290 280 230 120 200 190 174 160 148 139 126 114 12.77" 13.08 13.72b 14.33 15.17 15.78 16.57 Thermal 11.9 13.9 16.7 20.3 25.2 27.5 30.2 0.2% Tensile yield strength 555 570 590 610 680 600 400 200 ultimate tensile strength Wa) Total tensile elongation (%) le h Rupture strength (MPa) Id h Rupture strength W a ) lo4h Rupture strength (MPa) CharPY impact toughness (J) Fatigue 106 cycle life (MPa) Fatigue lo7cycle life (MPa) 770 800 850 850 820 675 500 200 14 14 17 18 5 5 7 lo 252 124 55 28 172 83 36 18 124 55 24 11 15 10 Density (g/cm3) Hardness (R,) Microhardness @Ph) ModuIus (GPA) Mean Coeff. of the& expansion (lod/'c) conductivity (w/m k) (Mpa) _ 40 40 40 35 63@ "Roomtemperature to 100 C. Qmm temperature to specified temperature.?>ah at 650 Cfor investmentcast test bars. 550