Superalloy Development for Aircraft Gas Turbines

Transcription

1 $1.50 PER COPY 75C TO ASME MEMBERS The Society shall not be responsible for statements or opinions advanced in papers or in discussion at meetings of the Society or of its 69-GT-7 Divisions or Sections, or printed in its publications. Discussion is printed only if the paper is published in an ASME journal or Proceedings. Released for general publication upon presentation Copyright 1969 by ASME Superalloy Development for Aircraft Gas Turbines R. J. QUIGG Manager, R & D, TRW Metals Division, Minerva, Ohio. H. E. COLLINS Principal Engineer, TRW Equipment Laboratories Minerva, Ohio. The creep resistance of nickel-base superalloys has progressed consistently over the past two decades. These improvements have been accomplished through the use of three principal strengthening mechanisms, solid solution strengthening, intermetallic (gamma prime) strengthening, and carbide strengthening. Most recently statistical regression analysis has been employed to predict the influence of alloying ingredients in nickel, and an alloy (TRW-NASA VI A) has been designed through this procedure. Superalloys, along with appropriate cooling mechanisms, appear to be entrenched as the materials of construction for the hotter portions of aircraft gas turbines for at least the next ten years. Contributed by the Gas Turbine Division for presentation at the Gas Turbine Conference & Products Show, Cleveland, Ohio, March 9-13, 1969, of The American Society of Mechanical Engineers. Manuscript received at ASME Headquarters December 16, Copies will be available until January 1, THE AMERICAN SOCIETY OF MECHANICAL ENGINEERS, UNITED ENGINEERING CENTER, 345 EAST 47th STREET, NEW YORK, N.Y

2 Superalloy Development for Aircraft Gas Turbines R. J. QUIGG H. E. COLLINS INTRODUCTION The improvement in gas turbine performance over the past 20 years has been accomplished to a great extent by the materials developments in a class of materials called superalloys. These materials may be either nickel, cobalt, or iron base but in recent years emphasis has been on the nickel-base alloys. The origin of these materials is principally from three sources: (a) an outgrowth of stainless steels, (b) continuing modification of a basic 80 Ni - 20 Cr (nimonic 75) alloy, and (c) an outgrowth of dental casting materials such as Vitallium. The last item is particularly interesting as described by Thielemann (1). 1 In the first turbojet engines, standard high temperature steels such as the 17W alloy were found to be inadequate in their high temperature creep properties, causing the initiation of a rather sizeable investigation into cast alloys. Since the investment casting technology was most advanced for the dental industry, the aid and efforts of this industry were 1 Numbers in parentheses designate References at the end of the paper. solicited. From the resultant investigation, it was determined that a cobalt-base dental alloy, Vitallium, by far outperformed the competitive alloys. From this type material, over 40 million turbo supercharger blades were investment cast by the end of World War II (1). As the size of engines was increased and, hence, the size of the turbine blades and vanes increased, the mode of failure shifted from creep to fatigue. This was attributed to the coarse grain size of the investment cast part. As a result, emphasis was placed on the wrought materials, with the resultant development and utilization of wrought cobalt-base superalloys such as s816. These alloys which were air melted, generally contained tungsten for solid solution strengthening, chromium for corrosion resistance, and a relatively high carbon level to produce a high population of carbides for high temperature strengthening. The development of vacuum melting technology in the 1950's opened the door for the development of the advanced precipitation hardened nickel-base superalloys. Because of the high reactivity of Table 1 Alloy Cr Mo W Vitallium o S o Nimonic 80A 19.5 M Waspaloy Rene' Udimet AF Udimet Inca 713C IN Mar M B Mar M246 9.o TRW-NASA VI A 6, Nominal Compositions of Superalloys TaAl TiCb B Zr o o o o o CCo N1 Others.40 Bal Bal Fe Bal Bal Bal Bal Bal Bal 9.5 Fe Sal Bal Bal 1.0 V Bel Bal Bal Bel 0.4 Re 0.43 Hf 1

3 3 Fig.l Turbine blade in the macroetched condition showing the equiaxed grain structure titanium and aluminum, these elements could not be added in any quantity in air. With the advent of vacuum melting, quantities up to 7 wt percent (Ti + Al) could be introduced with the retention of adequate ductility enabling the nickel-base alloys to take full advantage of the Ni3(Al, Ti) gamma prime precipitation. What resulted was a series of alloys including M252, Waspaloy, Rene 41, Udimet 500, AF 1753, and Udimet 700 in the United States and Nimonic Series of alloys ranging from Nimonic 80 to Nimonic 115 in the United Kingdom. All of these alloys were classified as wrought materials. (In Russia, parallel nickelbase alloy development seemed to emphasize solid strengthening rather than emphasizing the gammaprime precipitation as was the case in the Free World.) Investment casting returned in the early s when it was determined that: (a) the most advanced superalloy composition could only be forged with great difficulty, (b) intricate cooling passages could best be cast, and (c) the cast structure inherently gave superior high temperature creep properties. Grain control had advanced to a sufficient degree in investment castings that producers would guarantee most any grain size ranging from salt and pepper to an equiaxed structure, Fig.l. This grain control capability has recently been expanded to include columnar and single crystal structures, Figs.2 and 3 (2,3). A whole new series of high strength-high temperature cast nickel-base alloys have evolved including Inc() 713C, IN-100, Mar-M200, B-1900, Mar-M246, and TRW-NASA VI A. Fig.2 Turbine vane in the macroetched condition showing the columnar grain structure mma, Fig.3 Single crystal turbine blade in the macroetched condition (3) Oxidation resistance over this sequence of The progression of superalloy improvement alloys has been slightly degraded because of the is best illustrated by the graph of Fig.4 in which reduced amount of chromium present (see the compothe stress rupture properties at a single parameter sition listed in Table 1); however, the oxidation are compared for the superalloys mentioned in the resistance of all these alloys remains quite good preceding discussion. As is evident, the high tem- up to 2000 F. Of greater importance than oxidaperature capability of these materials has been tion is the "hot corrosion" problem where the alincreased by several hundred degrees from these loys are attacked by a combination of sulfur and alloy developments. sea salt. Both oxidation and "hot corrosion" have

4 , n Fig.4 Comparison of the stress rupture properties Fig.6 Gamma prime formations in an advanced cast of commercial superalloys nickel-base superalloy. X10,000 t7 e TO 60 ', o 30 LA Solid Solution Strengthening A number of elements have a wide range of solid solubility in nickel. Cobalt is completely soluble and chromium, molybdenum, tungsten, vanadium, tantalum, and columbium exhibit wide solid solubility in nickel. The refractory metal solid solution hardness significantly improve the stress rupture properties, especially at 1800 F and above. Chromium and cobalt, while not particularly beneficial to high temperature strength, nevertheless, do have important effects: chromium on oxidation resistance and on gamma prime and carbide formation, and cobalt upon gamma prime stability, workability, and ductility. Fe Cb Ni BASE METAL Fig.5 Comparison of the percent of the melting Intermetallic Strengthening point where usable strength is retained for various The gamma prime intermetallic (Ni3A1) is by base metals been alleviated somewhat through the use of aluminide coatings in the hottest parts of the gas turbine. The remarkable level of achievement obtained in nickel base superalloys can be illustrated by comparing the percent of the melting point in degrees Rankine to which nickel-base alloys can be used with other alloy base systems, as shown in Fig.5. It is apparent that nickel-base alloys far exceed the other base metals shown including titanium, iron, and columbium (the latter base material being largely oxidation limited). STRENGTHENING MECHANISMS The excellent high temperature properties obtained with nickel-base superalloys are achieved through three principal strengthening mechanisms, namely solid solution, intermetallic, and carbide strengthening. These three mechanisms will be discussed separately. far the most important single factor in the achieving of high temperature strength in nickel-base superalloys. This intermetallic has only a slight mismatch with the gamma matrix and, hence, induces little actual hardening but rather the chief beneficial effect is in acting like an "inert particle," effective over a wide range of temperatures. Titanium can also be used to replace aluminum in this compound. Thus, a measure of the degree of intermetallic strengthening in a given superalloy can be ascertained from the (Ti + Al) level. In the most advanced cast superalloys, substantial quantities of gamma prime are found, as is evident in Fig.6. This structure is formed directly on cooling from the melt and no heat treatment other than perhaps a 1600 F age (to produce fine second generation gamma prime and discrete partial carbide structures for optimum intermedi-, ate temperature properties) is conducted. Commonly a segregated gamma prime formation, called "primary gamma prime" is found in cast superalloys. This type structure, shown in Fig.7, is generally considered undesirable and must be avoided or at least minimized. 3

5 " Fig.7 Eutectic structure of gamma prime and gamma Fig.9 MC carbide formations in a cast nickel-base phases (also called "primary gamma prime") in a superalloy. X10,000 cast nickel-base superalloy. X3400 Fig.8 Heat treated structure of a wrought nickelbase superalloy showing intragranular gamma prime and discrete particles of M23C6 carbides in grain boundary. X15,000 In wrought superalloys a solution heat treatment is required to produce optimum high temperature properties. The gamma prime phase is then re-precipitated during some intermediate temperature age (depending on the size and morphology desired). The resultant structure in a wrought alloy is shown in Fig.8. Carbide Strengthening While the level of carbon present in the typical nickel-base superalloy is not high (generally it is around 0.10 C), carbide formations, nevertheless, play an important role in the high temperature strengthening of these alloys. The carbide formations are made with the alloying elements and, therefore, vary considerably from alloy to alloy. In general, there are three main types of carbides present in nickel base superalloys. These are MC M6C, and M23C6. Fig.10 Platelet M23C6 carbide rormation in a heat treated wrought nickel-base superalloy. X15,000 MC carbides are generally massive relatively inert carbides which are soluble only at very high temperatures. A typical MC carbide is shown in Fig.9. In the compound, the M is generally titanium, tantalum, or columbium. M6C carbides are present in alloys which contain significant quantities of molybdenum or tungsten. This carbide generally forms in the 1700 F F range and may form on cooling through this range or upon aging in this range. M6C carbide formation may be desirable or undesirable depending on the morphology and mode of formation. In general, plate-like formations are not desirable, while massive M6C formations are desirable. M23C6 carbides are largely grain boundary formations where the M is principally chromium. M23C6 carbides generally form in the 1400 F F range and may be of the discrete particle type, Fig.8, or of the less desirable platelet type, Fig.10. 4

6 AN EXAMPLE OF ALLOY OPTIMIZATION found plus some M6C and small amounts of M23C6 (7). TRW-NASA VI A is a recently developed nickelbase superalloy. This alloy was developed utilizing a combined metallurgical and statistical approach. The statistical techniques employed are described in a separate publication (4), thus, the following discussion will be centered in the metallurgical effect of each element present in the alloy and the reason for its inclusion. The composition of TRW-NASA VI A is as follows: Cr Co Mo W Ta Re Hf Cb Al Ti B Zr C Ni Bal Thus, a total of 13 alloying elements are present. Chromium is added largely for corrosion resistance and is found both in solid solution and combined in the M23 :',6 carbides. Cobalt is completely in solid solution. Molybdenum, tungsten, and tantalum which are all mainly in solid solution have a considerable strengthening effect with tantalum and tungsten having the greatest effects. Rhenium and hafnium are not common alloy constituents in nickel base superalloys. This can be principally attributed to their expense. These elements are conceivably in solid solution; their solid solubility in nickel is not wide but probably adequate to cover the amounts present. The theory behind adding a large number of elements for solid solution strengthening rather than more of a single element is that the first portion of each element creates a greater strengthening effect than subsequent quantities (5). This reasoning also applies to the columbium addition. Aluminum and titanium are added for the gamma prime precipitate. The amounts of these elements are as high as possible considering the large quantities of refractory elements present. In most castings, however, a limited amount of "primary gamma prime" is still formed. Boron and zirconium are added principally for their grain boundary strengthening effects. It has been hypothesized that boron and zirconium both segregate to the grain boundary area thereby minimizing high temperature creep (6). Carbon is added as is conventional for nickel-base superalloys to provide for carbide strengthening. In this alloy considerable MC is FUTURE TRENDS IN SUPERALLOY DEVELOPMENT While there is certainly additional room for optimization of nickel base superalloys along the lines of the TRW-NASA VI A alloy, these alloys are, however, melting point limited. Most nickel base superalloys cannot be used above 2300 F without encountering incipient melting. Thus, to achieve 2500 F or 2600 F inlet temperatures in a gas turbine, complex cooling schemes will be required. These cooling schemes will be dependent upon the continuous flow of air (or other media) throughout the part. Clogged passageways cannot be tolerated; thus, it will appear likely that future superalloy development will emphasize corrosion resistance hopefully without penalty in creep resistance. To achieve this, future superalloys will be complex like VI A, with further optimized combinations of creep and corrosion resistance. REFERENCES 1 Thielemann, R. H., "History and Future of Gas Turbine Alloys," World Investment Casting Conference, Piearcey, B. J. and VerSnyder, F. L., "A New Development in Gas Turbine Materials, Properties and Characteristics of PWA-664," Pratt and Whitney Aircraft, April 21, Piearcey, B. J. and VerSnyder, F. L., "Monocrystaloys - A New Concept in Gas Turbine Materials, Properties and Characteristics of PWA 1409," Prat and Whitney Aircraft, Feb. 2, Collins, H. E., Dreshfield, R. L., and Quigg, R. J., "Development of Nickel-Base Superalloy Using Statistically Designed Experiments," accepted for publication in Transactions ASM. 5 Guard, R. W., "Alloying for Creep Resistance," Mechanical Behavior of Materials at Elevated Temperatures, John Dorn, ed., Interscience, New York, 1961, p Decker, R. F. and Freeman, J. W., "The Mechanism of Beneficial Effects of Boron and Zirconium on Creep Properties of Complex Heat Resistent Alloys," Transactions AIME, Vol. 218, 1960, p Collins, H. E., "Relative Stability of Carbide and Intermetallic Phases in Nickel-Base Superalloys," International Symposium on Structural Stability in Superalloys, Sept. 5,