Volume 01 - Properties and Selection Irons, Steels, and High-Performance Alloys Part 16 potx

Duhl, The Development of Single-Crystal Superalloy Turbine Blades, in Advanced Temperature Alloys: Processing and Properties, American Society for Metals, 1986, p 41-49 High-5.. Yaker,

At 815 °C (1500 °F) At 870 °C (1600 °F) At 980 °C (1800 °F) At 1095 °C (2000 °F)

100 h MPa (ksi)

1000 h MPa (ksi)

100 h MPa (ksi)

1000 h MPa (ksi)

100 h MPa (ksi)

1000 h MPa (ksi)

100 h MPa (ksi)

Fig 1 Progress in the high-temperature capabilities of superalloys since the 1940s Source: Ref 2

The development of new polycrystalline alloys continued through the 1970s, however, at a more moderate rate Attention was concentrated instead on process development, with specific interest directed toward grain orientation and directional-solidification (DS) turbine blade and vane casting technology (Fig 2)

Fig 2 Advances in turbine blade materials and processes since 1960 Source: Ref 4

Applied to turbine blades and vanes, the DS casting process results in the alignment of all component grain boundaries such that they are parallel to the blade/vane stacking fault axis, essentially eliminating transverse grain boundaries (Fig 3) Because turbine blades/vanes encounter major operating stress in the direction which is near normal to the stacking fault axis, transverse grain boundaries provide relatively easy fracture paths The elimination of these paths provides increased strain elasticity by virtue of the lower <001> elastic modulus, thereby creating opportunities for further exploitation of the nickel-base alloy potential

Fig 3 The evolution of the processing of nickel-base superalloy turbine blades (a) From left, equiaxed,

directionally solidified, and single-crystal blades (b) An exposed view of the internal cooling passages of an aircraft turbine blade Source: Ref 5

The logical progression to grain-boundary reduction is the total elimination thereof Thus, single-crystal turbine blade/vane casting technology soon developed, providing further opportunity for nickel-base alloy design innovation

The late 1970s and the 1980s have, therefore, been a productive development period for nickel-base alloys designed specifically for directionally solidified columnar-grain and single-crystal cast components These new process technologies, which are more fully discussed in the article "Directionally Solidified and Single-Crystal Superalloys" in this Volume, have contributed to dramatic improvements in gas turbine engine operating efficiency

References

1 R.W Fawley, Superalloy Progress, in The Superalloys, C.T Sims and W.C Hagel, Ed., John Wiley & Sons,

1972, p 12

2 R.F Decker, Superalloys Does Life Renew at 50?, in Proceedings of the Fourth International Symposium

on Superalloys, American Society for Metals, 1980, p 2

3 Appendix B: Superalloy Data, in Superalloys II, C.T Sims, N.S Stoloff, and W.C Hagel, Ed., John Wiley &

Sons, 1987, p 575-597

4 M Gell and D.N Duhl, The Development of Single-Crystal Superalloy Turbine Blades, in Advanced Temperature Alloys: Processing and Properties, American Society for Metals, 1986, p 41-49

High-5 L.E Dardi, R.P Dalal, and C Yaker, Metallurgical Advancements in Investment Casting Technology, in

Advanced High-Temperature Alloys: Processing and Properties, American Society for Metals, 1986, p 25-39

Polycrystalline Cast Superalloys

Gary L Erickson, Cannon-Muskegon Corporation

Superalloy Design

Nickel-base superalloys have microstructures consisting of an austenitic face-centered cubic (fcc) matrix (γ) dispersed intermetallic fcc γ' Ni3(Al,Ti) precipitates coherent with the matrix (0 to 0.5% lattice mismatch), and carbides, borides, and other phases distributed throughout the matrix and along the grain boundaries These complex alloys generally contain more than ten different alloying constituents Various combinations of carbon, boron, zirconium, hafnium, cobalt, chromium, aluminum, titanium, vanadium, molybdenum, tungsten, niobium, tantalum, and rhenium result in the commercial alloys used in today's gas turbine engines

Some alloying elements have single-function importance, whereas others provide multiple functions For example, chromium is primarily added to nickel-base alloys for sulfidation resistance (Cr2O3 protective-scale formation), whereas aluminum not only is a strong γ' former but also helps provide oxidation resistance when present in sufficient quantity by forming a protective Al2O3 scale

Many of the other alloying elements also have multiple roles Titanium, while primarily partitioning to the γ', also participates in the formation of primary (MC) carbides, the hexagonal close-packed (hcp) eta (η) phase, and undesirable nitride and carbosulfide formation Molybdenum, tungsten, tantalum, rhenium, cobalt, and chromium additions promote solid-solution strengthening, but it is known that tantalum, tungsten, and rhenium may also partition to the γ' to varying degrees and that tantalum and rhenium may also be beneficial to environmental resistance properties

Vanadium is a γ' partitioner, but it also promotes the formation of M3B2-type borides Niobium forms the intermetallic phases delta (δ) (orthorhombic Ni3Nb) and γ'' (body-centered tetragonal Ni3Nb), but it is also involved in the formation of Laves (Fe,Ni2Nb) phase, carbides, borides, and/or nitrides Hafnium is a strong carbide former that is added to polycrystalline alloys to improve grain-boundary ductility However, at the same time, it increases the volume fraction of γ/γ' eutectic and increases oxidation resistance Carbon, boron, and zirconium are used at varying levels for grain-boundary strengthening

All of these constituents interact in various ways to provide high tensile, creep, and fatigue strengths, plus oxidation and sulfidation resistance Proper control of the cast microstructure and subsequent solutioning and aging treatments generally result in satisfactory component performance

Under the extreme temperature/stress conditions in which superalloy components operate, however, microstructural features change, often with attendant property changes The microstructural instabilities that may occur include:

Although it occurs during both solidification and heat treatment, carbide precipitation is generally promoted during component heat treatment to effect an optimum grain-boundary carbide morphology and population Discrete, blocky

M23C6 particles distributed in a discontinuous fashion are preferred High-temperature, stressed exposure tends to cause carbide degeneration, often resulting in grain-boundary overload and compromised rupture strength

MC-type carbides generally occur during alloy solidification They are titanium-rich (MC-1) or tantalum-rich (MC-2) and may partially degenerate with high-temperature exposure to form hafnium-rich (MC-3) carbides and/or M23C6, M7C3, and

M6C carbides (secondary carbides); the specific type depends upon alloy chemistry and exposure temperature The chromium-rich M23C6 generally forms at the grain boundaries in polycrystalline materials; when present as discrete, discontinuous particles, it provides the grain-boundary strength and resistance to fracture needed to prolong service life

On the other hand, carbide degeneration also releases titanium and tantalum to the solid-solution matrix, resulting in further matrix saturation Oversaturation can result in the formation of undesirable secondary phases such as (tungsten- and/or molybdenum-rich), α-W, α-Cr, and or M6C carbides, making chemistry balancing and controlled thermal treatment necessary for ultimate success

Superalloys are, indeed, complex However, careful alloy design and processing will provide the desired results Simply stated, superalloy property attainment is principally a function of the amount and morphology of the γ', grain size and shape, and carbide distribution Early superalloys contained less than 25 vol% γ' However, commercial vacuum induction refining and casting provided the opportunity for greater γ' volume fraction, to the extent that today's commercial superalloys generally contain approximately 60 vol% γ'

This increased level of γ' results in greater alloy creep strength (Fig 4), but it can be fully exploited only in single-crystal components, where full γ' solutioning is generally possible For polycrystalline superalloy components, high-temperature strength is affected by the condition of the grain boundaries and, in particular, the grain-boundary carbide morphology and distribution Optimized properties can be achieved if solutioning and aging treatments are developed to attain discrete, globular carbide formation along the grain boundaries in conjunction with the optimized γ' volume fraction/morphology and component grain structure Representative stress-rupture curves for selected nickel-base superalloys are shown in Fig 5 Table 5 also provides stress-rupture data

Fig 4 The relationship between γ' volume percent and stress-rupture strength for nickel-base superalloys Source: Ref 6

Fig 5 Stress-rupture curves for selected superalloys (a) and (b) Nickel-base superalloys 1000 h (c)

Cobalt-base superalloys 1000 h Source: Ref 3 (d) Larson-Miller stress-rupture curves for selected nickel-Cobalt-base superalloys Source: Ref 7 (e) Larson-Miller stress-rupture curves for selected cobalt-base superalloys Source: Ref 8

Cobalt-base alloys (see Table 2) are designed around a cobalt-chromium matrix with chromium contents ranging from

18 to 35 wt% The high chromium content contributes to oxidation and sulfidation resistance, but also participates in carbide formation (Cr7C3 and M23C6) and solid-solution strengthening Carbon content generally ranges from 0.25 to 1.0%, with nitrogen occasionally substituting for carbon

Cobalt-base alloys are often designed with significant levels of both nickel and tungsten The addition of nickel helps to stabilize the desired fcc matrix, while tungsten provides solid-solution strengthening and promotes carbide formation Molybdenum also contributes to solid-solution strengthening but is less effective and potentially more deleterious than a tungsten addition Other alloying elements contributing to the solid solution and/or carbide formation are tantalum, niobium, zirconium, vanadium, and titanium

These additions provide strength by means of solid-solution and second-phase strengthening No intermetallic precipitated phase has been discovered to equal the benefit imparted by γ' nickel-base superalloys Solid-solution strengthening results principally from the chromium, tantalum, niobium, and tungsten additions, while second-phase strengthening is obtained primarily from the carbides and carbonitrides formed with chromium The multiple-composition complex carbides may be present as MC, MC, MC , M C, and MC

As with nickel-base superalloys, carbides must be precipitated at grain boundaries to control gross grain-boundary sliding and migration Optimum mechanical properties are obtained through the careful balancing of the carbides at grain boundaries and within the matrix When carbides are present in sufficient quantity, the skeletal carbide network that results can contribute to component strength much like the strengthening that is achieved in a composite (Ref 9)

Most cobalt-base alloy aerospace castings are not heat treated apart from a relatively low-temperature stress-relief anneal Carbide distribution is, therefore, determined during solidification, thereby high-lighting the need for stringent control of the alloy pouring temperature and cooling rate during and after solidification Exceptions to this may be found in medical applications, where cobalt-base alloy castings for orthopedic implants are sometimes solution treated Representative stress-rupture curves for selected cobalt-base alloys are shown in Fig 5(c) and 5(e) Table 6 also provides stress-rupture data

References cited in this section

3 Appendix B: Superalloy Data, in Superalloys II, C.T Sims, N.S Stoloff, and W.C Hagel, Ed., John Wiley &

Sons, 1987, p 575-597

6 R.F Decker, "Strengthening Mechanisms in Nickel-Base Superalloys," Paper presented at the Steel Strengthening Mechanisms Symposium, Zurich, 1969

7 W Betteridge, Nickel and Its Alloys, Ellis Horwood, 1984

8 W Betteridge, Cobalt and Its Alloys, Ellis Horwood, 1982

9 M.J Donachie, Introduction to Superalloys, in Superalloys Source Book, American Society for Metals,

1984, p 9

Polycrystalline Cast Superalloys

Gary L Erickson, Cannon-Muskegon Corporation

Vacuum Induction Melting of Superalloys

Commercial vacuum induction melting was developed in the early 1950s, having been stimulated by the need to produce superalloys containing reactive elements within an evacuated atmosphere The process is relatively flexible, featuring the independent control of time, temperature, pressure, and mass transport through melt stirring As such, VIM offers more control over alloy composition and homogeneity than all other vacuum melting processes

The primary purification reaction occurring in the process is the removal of melt-contained oxygen by means of a

reaction with carbon to form carbon monoxide (CO) The reaction occurs most readily at or near the melt surface with the reaction kinetics being affected by crucible geometry and melt stirring The removal of oxygen from the melt as CO is favored by decreased melt chamber pressure, elevated bath temperature, and increased carbon activity (Ref 10)

The melting crucible material is not inert and is actually another source of oxygen and other impurities, depending on refractory type and condition Therefore, both melt refining temperature and refining duration are carefully scrutinized Proper melt stirring is integral to the deoxidation process and must be optimized through proper furnace power frequency and application procedure to prevent refractory lining erosion, a potential problem particularly during the controlled but more vigorous CO boiling portion of the process

Vacuum induction melting deoxidation, that is, the generation of CO gas, proceeds as CO bubbles are nucleated heterogeneously along the walls and, sometimes, bottom of the melt/lining-refractory interface This occurs preferentially

at small crevices existing in the lining, with the bubbles growing during movement toward the molten metal/vacuum interface (Ref 11, 12, 13) Actual bubble formation is dependent on the number of gas molecules present; the pressure in the liquid at the level of the bubble; the temperature of the gas; and, for very small bubbles, the interfacial tension between the gas and the liquid metal Figure 6 shows bubble formation during the VIM process

Fig 6 Vacuum induction refining process

Following formation, bubble growth and mass transport within the liquid toward the liquid/vacuum interface is dependent on:

melt stirring

The relatively vigorous, but controlled, portion of the boiling process results in the greatest CO removal Concurrently, a slight nitrogen loss is realized because of scavenging associated with the CO bubbles, and a slight sulfur reduction may occur during the CO supersaturation stage via sulfur dioxide (SO2) evolution Minor tramp elements such as lead, silver, bismuth, selenium, and tellurium, which are deleterious to alloy elevated-temperature rupture strength and ductility (Ref

14, 15), are partially evaporated during this period as well as throughout the entire refining process (Ref 16, 17) Some undesirable elements, however, such as arsenic and tin, must be controlled through raw material selection because they are not removed by vacuum refining Figure 7 shows the effects of VIM time and temperature on tramp element concentration Once the boiling subsides, surface desorption of additional CO occurs, and it is during this nonboiling period that nitrogen removal (desorption) is most effective (Ref 18)

Fig 7 Evaporation of elements from an 80Ni-20Cr alloy during VIM

Refractory Materials. Superalloy melting is generally undertaken in a relatively unreactive, bond strength, purity MgO-Al2O3 spinel refractory lining Refractories may be monolithic or brick and mortar, with the former providing the greater potential for alloy quality By minimizing extremes in thermal cycling; optimizing alloy sequencing; and refining temperature, time, and pressure, alloys practically void of any lining-related nonmetallics can be produced

high-The types of raw materials and the melt procedure vary depending on the quality of alloy being produced Alloys destined for critical application, the components of which may be difficult to cast, are produced using the highest-quality raw materials commercially available, in conjunction with sophisticated melt processing Lower-quality raw materials, such as GMR-235 or IN-713 C, are used for commercial application, such as nonrotating, noncritical, or nonaerospace use, for example, turbocharger wheels, because a higher-quality alloy product is not necessary The more sophisticated alloy systems are generally produced to premium-quality or integral wheel quality levels, with special attention given not only

to cleanliness but also to specific microstructural characteristics and/or chemistry requirements to assist castability, weldability, and component mechanical properties

The Melt Process. Base charge materials are layered in the relatively warm furnace, in a manner which recognizes and accommodates the elemental melting point of the material and bridging tendency Only those materials with oxides that are relatively easily reduced for the encountered melt conditions are placed in the initial furnace charge along with a small, controlled carbon addition Also, those elements that have a particularly strong affinity for nitrogen may be withheld from the base charge because they lower the activity of the dissolved nitrogen

Following furnace evacuation and particular heat-up cycles that ensure proper closure of any refractory lining cracks prior

to metal liquation, optimum temperature and vacuum pressure, consistent with promoting a somewhat vigorous CO boil,

is attained Bath refining is undertaken at a temperature and duration long enough to reach the so-called system equilibrium conditions, the assurance of which is provided by the attainment of consistent furnace leak-up rates At this point, those elements that were held from the base charge because of their relative reactivity toward oxygen, for example, aluminum, titanium, zirconium, and hafnium, are added with an associated solutioning and homogenization procedure

Dip sample alloy chemistry is checked in a relatively short time period with any necessary chemistry adjustment undertaken A similar analytical check is undertaken prior to pouring

Pouring proceeds once the correct chemistry is ensured, the bath is properly solutioned/homogenized, and the proper pour temperature is attained It is generally undertaken under high vacuum condition and proceeds from the furnace crucible into a relatively sophisticated multicompartment tundish, thereby ensuring that extensive time for flotation is achieved

and also that laminar flow conditions prevail in the final separation and pour compartments The flow-rate controlling, high-alumina tundish system results in a relatively slow pouring rate, effectively maximizing alloy cleanliness

Filters. One of the most critical stages with respect to cleanliness is the pouring of the melt Ceramic foam filters are used in some master metal operations to remove relatively large melt inclusions by means of entrapment Foam filters are most effective where extremely high pour rate conditions and gross cleanliness problems prevail Filter performance often varies because of the occasional use of filters with poor mechanical strength and/or thermal shock resistance Foam cell particle breakage often results from handling during shipment or tundish installation and, if undetected, results in filter particulate in the alloy bar stock and subsequently cast components Optimized VIM technology and practice without filters provide a clean alloy, without the inherent risks associated with filter use when applied to master metal production

Alternative Melt Techniques. Electron beam, cold hearth refining systems have been shown to be effective for nonmetallic-inclusion removal from superalloys Initial problems occurring during the process that are associated with analytical control, for example, chromium evaporation, have almost been resolved However, the process is not currently cost effective for virgin cast superalloy production, although use may be found for powder metallurgy (P/M) superalloy production and/or revert scrap reclamation in the future Vacuum arc skull melting and casting technology may also be used for superalloy castings

A plasma melting/refining facility for the production of superalloy powder metal is currently being commissioned Plasma melt investment casting trials have been undertaken, but no data are currently available Skull casting development would also help pace the interest in this secondary refining technology Each of the above-mentioned

techniques is described in detail in Casting, Volume 15 of ASM Handbook, formerly 9th Edition Metals Handbook

References cited in this section

10 D.R Gaskell, Introduction to Metallurgical Thermodynamics, Scripta Publishing, 1973, p 268-273

11 D Winkler, Thermodynamics and Kinetics in Vacuum Metallurgy, in Vacuum Metallurgy, O Winkler and

R Bakish, Ed., Elsevier, 1971, p 42-54

12 J.S Foster, "Liquid Metal-Gas Systems and Kinetics of Metal Degassing," Metallurgical Kinetics, Michigan Technological University course material, 1974

13 J.F Elliot, "Metal Refractory Reactions in Vacuum Processing of Steel and Superalloys," Paper presented at the AIME Electric Furnace Conference, American Institute of Mining, Metallurgical, and Petroleum Engineers, 1971

14 D.R Wood and R.M Cook, Effects of Trace Elements on the Creep Rupture Properties of Nickel Base

Alloys, Metallurgia, Vol 7, 1963, p 109

15 W.B Kent, Trace Element Effects in Vacuum Melted Alloys, J Vac Sci Technol., Vol 11 (No 6),

Nov/Dec 1974, p 1038-1046

16 Evaporation of Elements From 80/20 Nickel-Chromium During Vacuum Induction Melting, in

Transactions of the Vacuum Metallurgy Conference, American Vacuum Society, 1963

17 R.E Schwer, M.J Gray, and S.F Morykwas, "Trace Element Refining of Ni-Base Superalloys by Vacuum Induction Melting," Paper presented at the Vacuum Metallurgy Conference, American Vacuum Society, Battelle Memorial Institute, Columbus, OH, June 1975

18 V.M Antipov, Refining of High-Temperature Nickel Alloy in Vacuum Induction Furnaces, Stal', Feb 1968,

p 117-120

Polycrystalline Cast Superalloys

Gary L Erickson, Cannon-Muskegon Corporation

Quality Considerations

Cast cobalt-base superalloys have large application in the aerospace, medical, and chemical industries Nickel-base superalloys are used in a wide range of applications also, although most are related to the aerospace industry For aerospace applications, the quality level required for static components differs tremendously from that required for critical, highly stressed rotating components Thus, it is useful to classify superalloy products according to quality level

Superalloy quality can be categorized as commercial grade, aerospace quality, or premium grade Commercial-grade alloys may be produced with select materials, foundry revert, and/or lower-quality elemental raw materials Aerospace quality materials may also be produced with select, revert, and/or virgin materials However, the resulting quality, as measured by cleanliness and gas and tramp element content, is greater Premium grades are produced for blade/vane airfoil, wheel, and structural-part applications using top-quality raw materials and sophisticated VIM procedures Typical gas and tramp element level data for the three grades are presented in Table 7

Table 7 Quality level (allowable tramp element concentrations) for selected nickel-base superalloys

IN-718/MAR-M 247, typical alloy concentration

Commercial grade

Element

Tooling applications Other

Aerospace quality Premium quality

(a) (a) + = or higher

Stringent control of alloy gas and tramp element content is paramount to achieving satisfactory foundry performance and component mechanical property response Low gas content is desired because oxygen in the alloy is present as nonmetallic stable oxide inclusions, which affect weldability and mechanical properties Nitrogen also is controlled because it induces microporosity in castings and, when sufficiently high, causes agglomerated microporosity Nitrogen promotes alloy/crucible wetting during precision investment casting operations and may also be involved in deleterious titanium carbonitride particle formation

Superalloy silicon content is controlled to low levels because it adversely affects mechanical properties and may cause hot-short cracking during welding operations In the case of IN-718, silicon is avoided because it also impedes the rate of Laves phase transformation Low sulfur content is also desired because it tends to migrate to grain boundaries, thereby decreasing hot ductility and promoting cracking In addition, it can be deleterious to fatigue strength and contribute to increased alloy/crucible wetting during precision casting

Zirconium is added to superalloys to control grain-boundary strength However, it is minimized in alloys that undergo extensive welding because it increases the tendency for cracking in the weld and heat-affected zone of the base material

Polycrystalline Cast Superalloys

Gary L Erickson, Cannon-Muskegon Corporation

Investment Casting of Superalloys (Ref 19)

A number of casting processes can provide near-net shape superalloy cast parts, but essentially all components are produced by investment casting (Fig 8) The characteristic physical and mechanical properties and complex, hollow shape-making capabilities of investment casting have made it ideal for amplifying the unusual high-temperature properties of superalloys

Fig 8 Basic steps in the investment casting process See Fig 9(a) for a close-up of an automated slurry

coating process Source: Ref 4

Cast superalloys are made in a wider range of compositions than are wrought alloys Creep and rupture properties of a given superalloy composition are maximized by the casting and heat-treatment processes Ductility and fatigue properties

of polycrystalline materials are generally lower in castings than in their wrought counterparts of similar composition The gap, however, is being reduced by new technological developments to eliminate casting defects and refine grain size

Investment Casting Process

Patterns, Cores, and Molds. The first step in the investment casting process is to produce an exact replica or pattern

of the part in wax, plastic, or a combination thereof Pattern dimensions must compensate for wax, mold, and metal shrinkage during processing If the product contains internal passages, a preformed ceramic core is inserted in the die cavity, around which the pattern material is injected Except for large or complex castings, a number of patterns may be assembled in a cluster and held in position in order to channel the molten metal into the various mold cavities Design and positioning of the runners and gating is critical to achieving sound, metallurgically acceptable castings Today the molds are produced by first immersing the pattern assembly in an aqueous ceramic slurry (Fig 9) A dry, granular ceramic stucco is applied immediately after dipping to strengthen the shell These steps are repeated several times to develop a rigid shell After slow, thorough drying, the wax is melted out of the shell, and the mold is fired to increase substantially its strength for handling and storage An insulating blanket is tailored to the mold configuration to minimize heat loss during the casting operation and to control solidification More information on the production of patterns, cores, and

shells for investment casting is available in the article "Investment Casting" in Casting, Volume 15 of ASM Handbook, formerly 9th Edition Metals Handbook

Fig 9 (a) Automated slurry coating of an investment casting mold (b) Cutaway view of a shell mold for an

air-cooled turbine blade casting Source: Ref 5

To make equiaxed-grain castings, the mold is preheated to enhance mold filling, control solidification, and develop the proper microstructure For vacuum casting, the alloy charge is melted in an isolated chamber before the preheated mold is inserted, and the pressure is maintained at about 1 μm for pouring After casting, exothermic material is applied as a hot top for feeding purposes, and the mold is allowed to cool A different procedure is followed in the production of directionally solidified (DS) and single-crystal (SC) superalloy castings; this is described in the article "Directionally Solidified and Single-Crystal Superalloys" in this Volume and in the article "Directional and Monocrystal Solidification"

in Casting, Volume 15 of ASM Handbook, formerly 9th Edition Metals Handbook

The Casting Process. Most superalloys are cast in vacuum to avoid the oxidation of reactive elements in their compositions Some cobalt-base superalloys are cast in air using induction or indirect arc rollover furnaces The vacuum casting of equiaxed-grain products is usually done in a furnace divided into two major chambers, each held under vacuum and separated by a large door or valve The upper chamber contains an induction-heated reusable ceramic crucible in which the alloy is melted Zirconia crucibles are commonly employed; single-use silica liners may be specified when alloy cleanliness is especially critical

The preweighed charge is introduced through a lock device and is melted rapidly to a predetermined temperature, usually

85 to 165 °C (150 to 300 °F) above the liquidus temperature Precise optical measurement of this temperature is crucial Metal temperature during casting is much more critical than mold temperature in controlling grain size and orientation; it also strongly affects the presence and location of microshrinkage When the superheat condition is satisfied, the preheated mold is rapidly transferred from the preheat furnace to the lower chamber, which is then evacuated The mold is raised to the casting position, and the molten superalloy is quickly poured into the cavity; speed and reproducibility are essential in order to achieve good fill without cold shuts and other related imperfections Precise mold positioning and pour rates also are imperative For maximum consistency, melting and casting are automated with programmed closed-loop furnace control The filled mold is lowered and removed from the furnace

Shrinkage during solidification is minimized in part by maintaining a head of molten metal to feed the casting; this is achieved by adding an exothermic material immediately after mold removal from the furnace

Because of thermal expansion differences, the shell mold usually fractures upon cooling, facilitating its removal by mechanical or hydraulic means Before grit- and sand-blasting operations, the individual castings are separated from the cluster by abrasive cutoff After shell removal, the cluster is checked by one of several commercially available emission

or x-ray fluorescence instruments to verify the alloy identification

A major portion of the casting cost is in the finishing operations, which remain labor intensive Superficial surface defects are blended out abrasively within specified limits, and the castings may require mechanical straightening operations before and after heat treatment to satisfy dimensional requirements

References cited in this section

4 M Gell and D.N Duhl, The Development of Single-Crystal Superalloy Turbine Blades, in Advanced Temperature Alloys: Processing and Properties, American Society for Metals, 1986, p 41-49

High-5 L.E Dardi, R.P Dalal, and C Yaker, Metallurgical Advancements in Investment Casting Technology, in

Advanced High-Temperature Alloys: Processing and Properties, American Society for Metals, 1986, p

25-39

19 W.R Freeman, Jr., Chapter 15 in Superalloys II, C.T Sims, N.S Stoloff, and W.C Hagel, Ed., John Wiley

& Sons, 1987, p 411-439

Polycrystalline Cast Superalloys

Gary L Erickson, Cannon-Muskegon Corporation

Control of Casting Microstructure (Ref 20)

The solidification of investment cast superalloy components is precisely controlled so that the microstructure, which ultimately determines mechanical properties, remains consistent For example, once the process for a particular component has been defined, the production of these components does not deviate from the agreed-upon steps for the entire production run, which may last many years without proper approval If steps are changed, it must be shown that the new steps do not cause a degradation in the properties of the component

To control the solidification of equiaxed-grain castings, the investment caster has several tools at his disposal: facecoats that encourage grain nucleation, pour temperature of the metal, preheat temperature of the shell, shell thickness, part orientation, part spacing, gating locations, insulation to wrap the shells, pouring speed, and shell agitation However, the investment caster must first fill the shell cavity, prevent hot tears or other cracks, and minimize porosity If the first two objectives can be met, the investment caster has some freedom to produce the desired structure If the desired structure still cannot be made, other more complex techniques may be employed, including changing the thermal conductivity of the shell

Dendrites are probably the most visible microstructural feature in superalloy castings Primary and secondary dendrite arm spacings are controlled by the cooling rate As the dendrite arm spacing is reduced, segregation in the dendrite core and interdendritic regions is also reduced, thereby benefitting mechanical properties

Carbides. Conventional equiaxed-grain nickel-base superalloys typically have 0.05 to 0.20 wt% C, while cobalt-base alloys contain up to about 1.0% C Both alloy systems may use carbon to increase grain-boundary strength Cobalt-base alloys require more because internal carbides are one of the primary strengthening mechanisms

Carbide morphology is controlled by solidification or composition For example, by increasing the cooling rate, more discrete, blocky-type MC carbides are formed in IN-713 C, and this often results in an improvement of at least two times

to low-cycle fatigue (LCF) properties (Ref 21, 22) If it is not possible to influence the cooling rate of a casting significantly, adding small amounts of magnesium, calcium, cerium, or other rare earth metals acting as nucleating agents will assist in carbide shape control

Eutectic Segregation. By the very nature of solidification, segregation is introduced into the component Important segregants of interest in cast superalloys are eutectics, which often are found in interdendritic or intergranular regions In nickel-base alloys, eutectic pools are the last constituents to solidify and have a cellular appearance The composition of the eutectic pools varies, but they typically contain excess γ', carbides, borides, and low melting point phases Control of the eutectic pool is done primarily through composition However, it has been shown that while the volume fraction of

eutectic remained constant near 0.10 vol% in IN-713, the size of the eutectic pool increased from 11 to 19 μm as the cooling rate decreased from 0.56 to 0.036 °C/s (1 to 0.065 °F/s) (Ref 20)

In cobalt-base alloys, the eutectics typically form lamellar γ and M23C6 pools or colonies Heat treatment between 1150 and 1230 °C (2100 and 2250 °F) for 4 h resolutions these eutectic colonies, redistributing much of the carbon

Porosity. It is important to minimize the porosity in castings because the pores serve as initiation sites for fracture, especially fatigue cracks There are three primary sources of porosity in superalloy investment castings: undissolved gas, microshrinkage caused by poor feeding between dendrites, and macroshrinkage caused by inadequate gating Undissolved gas is gas that has come out of solution but, with today's vacuum technology, it is seldom experienced Usually made up

of oxygen, nitrogen, or hydrogen, this gas can form spherical voids up to two or more times the diameter of the dendrite arm spacing Gas porosity can be essentially eliminated by maintaining a vacuum during remelting and casting

Microshrinkage (microporosity) is inherent to castings that experience dendritic solidification The pores are spherical, but they typically have a diameter less than the dendrite spacing Microshrinkage forms just ahead of the advancing solidus interface because liquid metal feeding is impeded by the tortuous path through and around the secondary dendrite arms (a fluid flow problem)

The 2 to 6% shrinkage experienced upon solidification by superalloy castings makes macroshrinkage (solidification shrinkage) a problem This type of porosity tends to be confined within the thickest section of the casting, where the last solidification takes place

The investment caster can control solidification shrinkage to a great extent by gating or feeding those areas that are the last to solidify With complex geometries and the desire to produce net-shape castings, the size and placement of gating is based on experience, which necessitates experimentation before the process can be defined Modeling with computers, however, is changing this practice Recently, it has become possible to model the solidification process of simple castings

by taking into account the thermal properties of the metal and the shell Thus, the areas of solidification shrinkage can be predicted Once the shrinkage areas are located, various gating schemes can be evaluated until the shrink within the part is pulled into the gate At this point, the model is verified by an experiment, significantly reducing the overall time it takes

to design gating configurations More information on the use of modeling to predict solidification shrinkage and other

casting variables is available in the Section "Computer Applications in Metal Casting" in Casting, Volume 15 of ASM

Handbook, formerly 9th Edition Metals Handbook

Grain Size. The control of grain size is an important means for developing and maintaining both physical and mechanical properties Generally, a number of randomly oriented equiaxed grains in a given cross section is preferred to provide consistent properties, but often this is difficult to achieve in thin sections To meet this objective, mold facecoat nucleants, mold and metal temperature, and other parameters are chosen to accelerate grain nucleation and solidification

Finer grain size generally improves tensile, fatigue, and creep properties at low-to-intermediate temperatures (Fig 10, 11, 12) The finer grain size produced by relatively rapid solidification is accompanied by a finer distribution of γ' particles and a tendency to form blocky carbide particles The latter morphology is preferred to the script-type carbides produced

by slow solidification rates, particularly for a fatigue-sensitive environment Under these conditions, the carbide particles

do not contribute to superalloy properties As the service temperature increases, they impart important grain-boundary strengthening, provided that continuous films or necklaces are avoided

Fig 10 Influence of casting variables on intermediate-temperature (760 °C, or 1400 °F, at 690 MPa, or 100

ksi) stress-rupture properties of a cast nickel-base superalloy Pour temperature and mold temperature affect solidification and thus grain size of the component Source: Ref 19

Fig 11 Effect of grain size control and HIP processing on the strain-controlled (axial) low-cycle fatigue of CM

247 LC

Fig 12 Influence of grain refinement on fatigue crack growth rate Source: Ref 19

For high-temperature rupture performance, slower solidification and cooling rates are preferred to coarsen both the grain size during solidification and the γ' precipitated during cooling While this benefits high-temperature strength through a reduction in grain-boundary content, more property scatter can be expected due to (random) crystallographic orientation effects For turbine blades, the desired microstructure is difficult to achieve because the thin airfoils operating at the highest temperatures should have coarse grains, and the heavier-section root attachment area, being less rupture dependent, should have a fine-grain microstructure Where conventional practice fails, a gate, or gutter, along the airfoil edges may be employed, through which metal is caused to flow, thereby creating deliberate hot spots to retard the local solidification rate

A significant foundry advancement has been the development of processes to produce fine-grain superalloy castings In the late 1960s, experiments were conducted on a grain refinement technique for integral turbine wheels The technique used the mechanical motion of a mold to shear dendrites from the solidifying metal These dendrites then acted as nucleation sites for additional grains However, the process was not commercially introduced because it produced castings with unacceptable levels of porosity

In the mid-1970s, developmental work on this process resumed when it was realized that hot isostatic pressing could be used to eliminate residual casting porosity The Howmet Corporation process that developed from this work, known as Grainex, results in ASTM grain sizes as fine as No 2 A further Howmet Corporation development, the Microcast-X process, has led to a greater refinement in grain size (ASTM No 3 to 5) Figure 13 compares the microstructures of grain-

refined rotors with those of a conventionally cast part References 21, 22, 23, 24, 25 provide additional information on grain size control/property relationships and on fine-grain casting process development work performed by others in the precision investment casting industry

Fig 13 Structure of conventionally cast turbine wheel (a) compared to wheels cast using the Grainex (b) and

Microcast-X (c) processes Courtesy of Howmet Turbine Corporation

References cited in this section

19 W.R Freeman, Jr., Chapter 15 in Superalloys II, C.T Sims, N.S Stoloff, and W.C Hagel, Ed., John Wiley

22 M Lamberigts, S Ballarati, and J.M Drapier, Optimization of the High Temperature, Low Cycle Fatigue

Strength of Precision Cast Turbine Wheels, in Proceedings of the Fifth International Symposium on Superalloys, American Institute of Mining, Metallurgical, and Petroleum Engineers, 1984, p 13-22

23 B.A Ewing and K.A Green, Polycrystalline Grain Controlled Castings for Rotating Compressor and

Turbine Components, in Proceedings of the Fifth International Symposium on Superalloys, American

Institute of Minning, Metallurgical, and Petroleum Engineers, 1984, p 33-42

24 M.J Woulds and H Benson, Development of a Conventional Fine Grain Casting Process, in Proceedings of the Fifth International Symposium on Superalloys, American Institute of Mining, Metallurgical, and

Petroleum Engineers, 1984, p 3-12

25 S.J Veeck, L.E Dardi, and J.A Butzer, "High Fatigue Strength, Investment Cast Integral Rotors for Gas Turbine Applications," Paper presented at the TMS-AIME Annual Meeting, The Metallurgical Society, Dallas, Feb 1982

Polycrystalline Cast Superalloys

Gary L Erickson, Cannon-Muskegon Corporation

Postcasting Processing

Heat Treatment. The heat treatment of cast superalloys in the traditional sense was not employed until the mid-1960s

Before the use of shell molds, the heavy-walled investment mold dictated a slow cooling rate with its associated aging effect on the casting As faster cooling rates with shell molds developed, the aging response varied with section size and the many possible casting variables These factors, coupled with significant γ' alloying additions, provided the opportunity

to minimize property scatter by heat treatment The combination of hot isostatic pressing plus heat treatment has also greatly enhanced properties

Generally, heat treating cast superalloys involves homogenization and solution heat treatments or aging heat treatments A stress-relief heat treatment may also be performed in order to reduce residual casting, welding, or machining stresses Detailed information can be found in Ref 20

Cast cobalt-base superalloys are not usually solutioned However, they may be given stress relief and/or aging treatments When required, aging is generally done at 760 °C (1400 °F) to promote the formation of discrete Cr23C6 particles Higher-temperature aging can result in acicular and/or lamellar precipitate formation

Cobalt-base alloy heat treatment may be done in an air atmosphere unless unusually high-temperature treatments are required, in which case vacuum or inert gas environments are used Conversely, nickel-base alloys are always heat treated

in a vacuum or in an inert gas medium

Polycrystalline cast nickel-base superalloys may or may not be given solution treatment Because alloys respond differently to γ' solutioning, some are given only aging treatment For those that do respond to partial solutioning, the treatment is performed at a temperature safely below the alloy's incipient melting point, for times ranging from 2 to 6 h at temperature Rapid cooling from high temperature is necessary to ensure that the γ' precipitate is fine sized, thereby maximizing strength potential

Solution heat-treating procedures must be optimized to stabilize the carbide morphology High-temperature exposure may cause extensive carbide degeneration, resulting in grain-boundary carbide overload and compromised mechanical properties Many polycrystalline materials are used in the as-cast plus aged condition

A typical aging cycle involves heating to 980 °C (1800 °F), holding for 5 h, and air cooling, followed by heating to 870

°C (1600 °F), holding for 20 h, and air cooling An alternative is heating to 1080 °C (1975 °F), holding for 4 h, and air cooling, followed by heating to 870 °C (1600 °F), holding for 20 h, and air cooling The 980 °C (1800 °F) and 1080 °C (1975 °F) exposures are carried out in conjunction with protective coating diffusion treatments It is important that the cooling rate from the 980 °C (1800 °F) and 1080 °C (1975 °F) treatments be rapid to maintain the optimum γ' size However, a furnace cool from 870 °C (1600 °F) treatment is acceptable

Hot Isostatic Pressing. Hot isostatic pressing (HIP) subjects a cast component to both elevated-temperature and isostatic gas pressure in an autoclave The most widely used pressurizing gas is argon For the processing of castings, argon is applied at pressures between 103 and 206 MPa (15 and 30 ksi), with 103 MPa (15 ksi) being the most common Process temperatures of 1200 to 1220 °C (2200 to 2225 °F) are common for polycrystalline superalloy castings

When castings are hot isostatically pressed, the simultaneous application of heat and pressure virtually eliminates internal voids and microporosity through a combination of plastic deformation, creep, and diffusion The elimination of internal defects leads to improved nondestructive testing ratings, increased mechanical properties, and reduced data scatter (Fig

14 and 15)

Fig 14 Effect of hot isostatic pressing on stress-rupture properties of cast IN-738 Test material was hot

isostatically pressed at 1205 °C (2200 °F) and 103 MPa (15 ksi) for 4 h (a) Test conditions: 760 °C (1400 °F) and 586 MPa (85 ksi) (b) Test conditions: 980 °C (1800 °F) and 152 MPa (22 ksi) Source: Howmet Corporation

Fig 15 Improvement of fatigue properties by the elimination of microporosity through HIP processing Source:

Ref 19

In the past 15 years, hot isostatic pressing has become an integral part of the manufacturing process for high-integrity aerospace castings The growth of hot isostatic pressing has paralleled the introduction of advanced nickel-based superalloys and increasing complex casting designs, both of which tend to increase levels of microporosity In addition, to optimize mechanical properties, turbine engine manufacturers have become more stringent in allowances for microporosity The requirement for reduced porosity levels and increased mechanical properties has been achieved in many cases through the use of hot isostatic pressing

In selecting HIP process parameters for a particular alloy, the primary objective is to use a combination of time, temperature, and pressure that is sufficient to achieve closure of internal voids and microporosity in the casting There are

also material considerations for avoiding such deleterious effects as incipient melting, grain growth, and the degradation

of constituent phases such as carbides

If encountered, incipient melting can be avoided by pre-HIP homogenization heat treatments or by lowering HIP temperatures If the temperature is lowered, an increase in processing pressure may be required to obtain closure in certain alloys For example, hafnium-bearing nickel-base superalloys such as C 101 and MAR-M 247, when cast with heavy sections (for example, integral wheels), have been found to undergo incipient melting when hot isostatically pressed at 1205 °C (2200 °F) and 103 MPa (15 ksi) for 4 h To prevent incipient melting and still obtain closure, the HIP parameters were changed to 1185 °C (2165 °F) and 172 MPa (25 ksi) for 4 h This trade-off between temperature and pressure can also be used to prevent grain growth and to prevent or limit carbide degradation while obtaining closure of microporosity

Time at temperature and pressure will obviously affect proccessing cost For most alloys, 2 to 4 h is sufficient Exceptions are massive section sizes, which require additional thermal soaking time

References cited in this section

19 W.R Freeman, Jr., Chapter 15 in Superalloys II, C.T Sims, N.S Stoloff, and W.C Hagel, Ed., John Wiley

& Sons, 1987, p 411-439

20 G.K Bouse and J.R Mihalisin, Metallurgy of Investment Cast Superalloy Components, in Superalloys, Supercomposites and Superceramics, Academic Press, 1989, p 99-148

Polycrystalline Cast Superalloys

Gary L Erickson, Cannon-Muskegon Corporation

References

1 R.W Fawley, Superalloy Progress, in The Superalloys, C.T Sims and W.C Hagel, Ed., John Wiley &

Sons, 1972, p 12

2 R.F Decker, Superalloys Does Life Renew at 50?, in Proceedings of the Fourth International Symposium

on Superalloys, American Society for Metals, 1980, p 2

3 Appendix B: Superalloy Data, in Superalloys II, C.T Sims, N.S Stoloff, and W.C Hagel, Ed., John Wiley

& Sons, 1987, p 575-597

4 M Gell and D.N Duhl, The Development of Single-Crystal Superalloy Turbine Blades, in Advanced Temperature Alloys: Processing and Properties, American Society for Metals, 1986, p 41-49

High-5 L.E Dardi, R.P Dalal, and C Yaker, Metallurgical Advancements in Investment Casting Technology, in

Advanced High-Temperature Alloys: Processing and Properties, American Society for Metals, 1986, p

25-39

6 R.F Decker, "Strengthening Mechanisms in Nickel-Base Superalloys," Paper presented at the Steel Strengthening Mechanisms Symposium, Zurich, 1969

7 W Betteridge, Nickel and Its Alloys, Ellis Horwood, 1984

8 W Betteridge, Cobalt and Its Alloys, Ellis Horwood, 1982

9 M.J Donachie, Introduction to Superalloys, in Superalloys Source Book, American Society for Metals,

1984, p 9

10 D.R Gaskell, Introduction to Metallurgical Thermodynamics, Scripta Publishing, 1973, p 268-273

11 D Winkler, Thermodynamics and Kinetics in Vacuum Metallurgy, in Vacuum Metallurgy, O Winkler and

R Bakish, Ed., Elsevier, 1971, p 42-54

12 J.S Foster, "Liquid Metal-Gas Systems and Kinetics of Metal Degassing," Metallurgical Kinetics,

Michigan Technological University course material, 1974

13 J.F Elliot, "Metal Refractory Reactions in Vacuum Processing of Steel and Superalloys," Paper presented

at the AIME Electric Furnace Conference, American Institute of Mining, Metallurgical, and Petroleum Engineers, 1971

14 D.R Wood and R.M Cook, Effects of Trace Elements on the Creep Rupture Properties of Nickel Base

Alloys, Metallurgia, Vol 7, 1963, p 109

15 W.B Kent, Trace Element Effects in Vacuum Melted Alloys, J Vac Sci Technol., Vol 11 (No 6),

Nov/Dec 1974, p 1038-1046

16 Evaporation of Elements From 80/20 Nickel-Chromium During Vacuum Induction Melting, in

Transactions of the Vacuum Metallurgy Conference, American Vacuum Society, 1963

17 R.E Schwer, M.J Gray, and S.F Morykwas, "Trace Element Refining of Ni-Base Superalloys by Vacuum Induction Melting," Paper presented at the Vacuum Metallurgy Conference, American Vacuum Society, Battelle Memorial Institute, Columbus, OH, June 1975

18 V.M Antipov, Refining of High-Temperature Nickel Alloy in Vacuum Induction Furnaces, Stal', Feb 1968,

22 M Lamberigts, S Ballarati, and J.M Drapier, Optimization of the High Temperature, Low Cycle Fatigue

Strength of Precision Cast Turbine Wheels, in Proceedings of the Fifth International Symposium on Superalloys, American Institute of Mining, Metallurgical, and Petroleum Engineers, 1984, p 13-22

23 B.A Ewing and K.A Green, Polycrystalline Grain Controlled Castings for Rotating Compressor and

Turbine Components, in Proceedings of the Fifth International Symposium on Superalloys, American

Institute of Minning, Metallurgical, and Petroleum Engineers, 1984, p 33-42

24 M.J Woulds and H Benson, Development of a Conventional Fine Grain Casting Process, in Proceedings

of the Fifth International Symposium on Superalloys, American Institute of Mining, Metallurgical, and

Petroleum Engineers, 1984, p 3-12

25 S.J Veeck, L.E Dardi, and J.A Butzer, "High Fatigue Strength, Investment Cast Integral Rotors for Gas Turbine Applications," Paper presented at the TMS-AIME Annual Meeting, The Metallurgical Society, Dallas, Feb 1982

Directionally Solidified and Single-Crystal Superalloys

K Harris, G.L Erickson, and R.E Schwer, Cannon-Muskegon Corporation

Introduction

THE PRIMARY GOALS in the continuing development of the aircraft gas turbine are increased operating temperatures and improved efficiencies A more efficient turbine is required to achieve lower fuel consumption Higher turbine inlet temperature and increased stage loading result in fewer parts, shorter engine lengths, and reduced weight Engine operating costs can be reduced if higher temperatures are possible without increasing part life-cycle costs

Critical turbine components include high-pressure turbine blades, vanes, and disks During the last 15 years, turbine inlet temperatures have increased by 278 °C (500 °F) About half of this increase is due to a more efficient design for the air cooling of turbine blades and vanes, while the other half is due to improved superalloys and casting processes (Ref 1) The cooling that is now possible with serpentine cores and multiple shaped-hole film cooling (Fig 1) enables high-pressure turbine blades and vanes to operate with turbine inlet temperatures of typically 1343 °C (2450 °F), which is above the melting point of the superalloy materials Turbine inlet temperatures as high as 1571 °C (2860 °F) are current parameters for several advanced fighter engines (Ref 2) It is forecast that by the mid to late 1990s, fabricated single-crystal airfoils with ultra-efficient transpiration cooling schemes will be capable of operating in gas temperatures greater than 1649 °C (3000 °F) with sufficient durability and reliability for man-rated flight turbine engines These advanced superalloy single-crystal diffusion-bonded airfoils will present a major challenge to the emerging ceramic composite component technology

Fig 1 Shaped holes, turbulators, pin fins, and other techniques used in turbine rotor blade cooling

For the past 28 years, high-pressure turbine blades and vanes have been made from cast nickel-base superalloys The higher-strength alloys are hardened by a combination of approximately 60 vol% γ' [Ni3(Al,Ti)] precipitated in a γ matrix, with solid-solution strengthening provided by the powerful strengtheners tantalum, tungsten, and molybdenum The γ phase, which has an ordered face-centered cubic structure, is coherent with the γ matrix, their lattice parameters being almost identical (<1% mismatch) This allows homogeneous nucleation of the precipitate with low surface energy and

long-time stability at temperature, ensuring the potential usefulness of the alloys at elevated temperatures up to 0.85 Tm

(melting point) for extended periods of time Tantalum, tungsten, and hafnium substitute for some of the aluminum and titanium in the γ', thus stiffening this phase because of their relatively large atomic size Initially, the blades were made as

isotropic polycrystal or equiaxed castings Under aerospace turbine engine operating conditions, failure of these grain components usually occurred at the grain boundaries from a combination of creep, thermal fatigue, and oxidation

equiaxed-Development of the directional solidification (DS) casting process to produce blades and vanes with low-modulus oriented columnar grains aligned parallel to the longitudinal, or principal-stress, axis (Fig 2) resulted in significant improvements in creep strength and ductility as well as in thermal fatigue resistance (5× improvement) Pratt and Whitney Aircraft (PWA) pioneered this process (Ref 3, 4), as well as its turbine engine application, and has accumulated 18 years

(100)-of production experience with over 25 million flight hours with DS blades and vanes (Ref 5)

There has been recent interest in DS blades not only for small- to medium-size airfoils for industrial turbines that burn natural gas but also for large base-load electricity-generating machines Improved fuel efficiency requirements, along with the desire for high-temperature exhaust gases from the gas turbine (to produce steam suitable for co-generation electricity production), have resulted in the development and application engineering of DS blades with component lengths in the range of 305 to 635 mm (12 to 25 in.)

Single-crystal (SX) casting technology was pioneered in the mid-1960s by PWA (Ref 5, 6) However, there was limited interest in the development of single-crystal blades because the conventional heat treatments being applied to MAR-M 200-type single-crystal components did not produce improvements in creep strength, thermal fatigue strength, and oxidation resistance that significantly exceeded the results achieved with the DS columnar-grain MAR-M 200

Hf Only ductility and transverse creep resistance were improved Around 1975, the beneficial role of γ' solutioning heat treatment applied to DS MAR-M 200 Hf was shown by PWA (Ref 7) It was found that creep strength was a direct function of the volume fraction of solutioned and reprecipitated fine γ' (Fig 3) Experimental work by PWA showed that the elimination of grain-boundary strengthening elements (boron, hafnium, zirconiun, and carbon) resulted in a substantial increase in the incipient melting temperature of the alloy (Ref 6) Consequently, the complete solutioning of the γ' phase, with appreciable solutioning of the γ-γ' eutectic phase, became possible without provoking incipient melting of the alloy

Fig 2 Directionally solidified turbine blade CM 247 LC

Fig 3 Rupture life versus volume fraction of fine γ' at a fixed total amount of fine and coarse γ' for DS MAR-M 200 Hf

alloy

Single-crystal alloy PWA 1480 (Table 1) offered a 25 to 50 °C (45 to 90 °F) temperature capability improvement in terms

of time-to-1% creep, compared to the extensively used DS MAR-M 200 Hf alloy (Ref 5) The creep property improvement, which increases with temperature, depended on optimized single-crystal microstructures with full solutioning of the as-cast coarse ' The PWA 1480 alloy was developed to utilize the relatively low thermal gradient, single-crystal casting facilities already available as DS production units, without the freckling problems of alloy 444 (single-crystal MAR-M 200 with no carbon, boron, hafnium, zirconium, or cobalt) (Ref 5) Alloy PWA 1480, with its high tantalum (12%) and low tungsten (4%) contents, proved to be unique with this castability feature Multistep homogenization/solutioning treatments with tight temperature control were developed to completely solution the ' PWA

1480 without inducing incipient melting Since 1982, PWA has had more than 5 million flight hours of successful experience using turbine blade and vane parts of single-crystal alloy PWA 1480 in commercial and military engines (Ref 8)

Table 1 First-generation single-crystal superalloys

Nominal composition, wt%

Alloy

Cr Co Mo W Ta V Nb Al Ti Hf Ni

Density, g/cm 3

References

1 F.E Pickering, Advances in Turbomachinery, Cliff Garrett Award Lecture, Aerosp Eng., Jan 1986, p 30-35

2 Snecma Advances M88 Demonstrator, Flight Int., 22 March 1986, p 26

3 B.J Piearcey and F.L VerSnyder, J Aircr., Vol 3 (No 5), 1966, p 390

4 F.L VerSnyder and M.E Shank, Malter Sci Eng., Vol 6 (No 4), 1970, p 321

5 M Gell, The Science & Technology of Single Crystal Superalloys, in Proceedings of Japan-U.S Seminar on Superalloys, International Iron and Steel Institute, 1984

6 M Gell, D.N Duhl, and A.F Giamei, The Development of Single Crystal Superalloy Turbine Blades, in

Proceedings of the Fourth International Symposium on Superalloys (Seven Springs, PA), American Society

for Metals, 1980, p 205-214

7 J.J Jackson, M.J Donachie, R.J Henricks, and M Gell, The Effects of Volume % of Fine ' on Creep in DS

MAR M 200 Hf, Metall Trans A, Vol 8A (No 10), 1977, p 1615

8 A.D Cetel and D.N Duhl, Second-Generation Nickel-Base Single Crystal Superalloy, in Sixth International Symposium on Superalloys (Seven Springs, PA), The Metallurgical Society, 1988, p 235-244

Directionally Solidified and Single-Crystal Superalloys

K Harris, G.L Erickson, and R.E Schwer, Cannon-Muskegon Corporation

Directionally Solidified Superalloys

Chemistry and DS Castability. Early work with directionally solidified columnar-grain turbine blades in the 1960s involved the superalloys used for conventionally cast, equiaxed blades containing approximately 60 vol% γ', such as IN

100 and MAR-M 200 The problems encountered ranged from little longitudinal stress-rupture improvement with IN 100

to the lack of transverse ductility and DS grain-boundary cracking with MAR-M 200

Pioneering work by Martin Metals resulted in the addition of hafnium to conventionally cast equiaxed superalloys to improve 760 °C (1400 °F) stress-rupture ductility and castability For directional solidification, PWA added hafnium to MAR-M 200, which reduced DS grain-boundary cracking and increased transverse ductility Although hafnium levels of

up to 2% and greater in MAR-M 200 Hf combatted DS grain-boundary cracking, increasing levels of hafnium also increased the DS airfoil component rejection rate and the number of quality assurance problems This was due to the occurrence of HfO inclusions that usually resulted from hafnium-ceramic reactions (core, shell-mold) Other first-generation DS alloys that were successfully and extensively adopted by turbine engine companies included René 80H (René 80 + Hf) by GE, MAR-M 002 by Rolls-Royce, and MAR-M 247 by Garrett Both MAR-M 002 and MAR-M 247 were originally developed by Martin Metals to contain hafnium for optimized equiaxed turbine blade mechanical properties and castability The nominal compositions of these first-generation DS superalloys are listed in Table 2 Directionally solidified superalloy turbine blades employed in large commercial turbofan engines for long-distance flights have been used for up to 15,000 h with high reliability

Table 2 First-generation DS superalloys with extensive turbine engine airfoil applications

MAR-M 247 0.15 8 10 0.6 10 3.0 5.5 1.0 0.015 0.03 1.5 bal

Continuing improvements in airfoil cooling techniques have usually led to significant gains in gas turbine operating efficiencies However, these cooling techniques often result in very complex cored, thinwall (0.5 1 mm, or 0.02 to 0.04 in.) airfoil designs, which can be susceptible to grain-boundary cracking during the DS casting of high-creep-strength alloys, particularly with modern high-thermal-gradient casting processes Thus, the need for improved DS castability resulted in the development by the Cannon-Muskegon Corporation of CM 247 LC, a second-generation alloy from the MAR-M 247 composition (Ref 9) The nominal composition of this superalloy, which is also known as René 108, is given in Table 3 The CM 247 LC alloy has particularly excellent resistance to DS grain-boundary cracking and is capable of essentially 100% γ' solutioning to maximize creep strength without incipient melting or deleterious M6C platelet formation upon subsequent high-temperature stress exposure, but with adequate transverse ductility retention

Table 3 Second-generation DS and SX superalloys

Nominal composition, wt%

Alloy

C Cr Co Mo W Ta Re Al Ti B Zr Hf Ni

Density, g/cm3

DS alloy

CM 247 LC 0.07 8 9 0.5 10 3.2 5.6 0.7 0.015 0.010 1.4 bal 8.54

SX alloys

PWA 1484 (Ref 8) 5 10 2 6 9 3 5.6 0.1 bal 8.95

With respect to DS grain-boundary cracking, zirconium and silicon are generally known to be bad actors Small amounts

of a brittle, hafnium-rich eutectic phase containing high concentrations of zirconium and silicon have been found in DS crack-prone tests (Ref 9) It was observed that very small reductions in zirconium and titanium contents, combined with a very tight control of silicon and sulfur, dramatically reduced the DS grain-boundary cracking tendency of a high-creep-strength superalloy such as MAR-M 247 (Ref 11) The major microstructural effect of the lower titanium content in CM

247 LC, compared to MAR-M 247, is to significantly reduce the size of the γ/γ' eutectic nodules as well as to lower the volume fraction of the eutectic from approximately 4 vol% in MAR-M 247 to 3 vol% in CM 247 LC DS components This factor is also believed to be significant in reducing the DS grain-boundary cracking tendency of CM 247 LC

Heat Treatment and Mechanical Properties. Multistep solutioning techniques based on a slow temperature increase between steps and temperatures up to 1254 °C (2290 °F) are used to supersolution heat treat CM 247 LC DS airfoil components to attain microstructures such as those shown in Fig 4 Resultant stress-rupture property improvements are illustrated in Fig 5

Fig 4 CM 247 LC directionally solidified turbine blade, as-cast, and supersolutioned microstructures, heat

V6692 Micrographs taken from airfoil, transverse orientation (a) As-cast 90× (b) As-cast 905× (c) Supersolutioned 90× (d) Supersolutioned 905×

Fig 5 Larson-Miller stress-rupture strength of DS CM 247 LC versus DS and equiaxed MAR-M 247 MFB,

machined from blade; GFQ, gas fan quenched; AC, air cooled

The advent of single-crystal technology is not likely to preempt the need for DS airfoil components in the intermediate term Directionally solidified airfoils will continue to be used for vane segments and low-pressure blades in advanced turbine engines because of the producibility of the components, which makes them cost effective

Several third-generation DS superalloys containing rhenium have been developed that have stress-rupture strength values close to those of the first-generation single-crystal alloys (Ref 12) These new alloys are particularly useful for DS vanes where load-bearing capability, such as to support a bearing, is an important design consideration

References cited in this section

8 A.D Cetel and D.N Duhl, Second-Generation Nickel-Base Single Crystal Superalloy, in Sixth International Symposium on Superalloys (Seven Springs, PA), The Metallurgical Society, 1988, p 235-244

9 J.J Burke, H.L Wheaton, and J.R Feller, Paper presented at the Annual TMS-AIME Meeting, Denver, CO, The Metallurgical Society, 1978

10 K Harris, G.L Erickson, and R.E Schwer, Development of CMSX-4 for Small Gas Turbines, Paper presented at the TMS-AIME Fall Meeting, Philadelphia, The Metallurgical Society, Oct 1983

11 K Harris, G.L Erickson, and R.E Schwer, MAR M 247 Derivations: CM 247 LC DS Alloy CMSX Single

Crystal Alloys, Properties and Performance, in Proceedings of the Fifth International Symposium on Superalloys (Seven Springs, PA), The Metallurgical Society, 1984, 221-230

12 K Harris, G.L Erickson, and R.E Schwer, CMSX Single Crystal, CM DS and Integral Wheel Alloys:

Properties and Performances, in Cost 50/501 Conference on High Temperature Alloys for Gas Turbines and Other Applications (Liege), Reidel, 1986

Directionally Solidified and Single-Crystal Superalloys

K Harris, G.L Erickson, and R.E Schwer, Cannon-Muskegon Corporation

Fig 7 Progress in turbine airfoil metal temperature capability Source: Ref 13

Fig 8 f-100 paired single-crystal vane cast in PWA 1480

Other pioneering single-crystal alloy development work resulted in the derivation of several single-crystal compositions from MAR-M 247 during the Garrett/NASA Materials for Advanced Technology Engines (MATE) program, which began in 1977 (Ref 14, 15) The two alloys studied extensively were NASAIR 100 and NASAIR Alloy 3; the latter contained a minor hafnium addition

The compositions of the first-generation single-crystal superalloys, many of which are being used in turbine engine applications, are shown in Table 1: René N-4 developed by General Electric (Ref 16); SRR 99 and RR 2000 by Rolls-Royce plc (Ref 17); AM1 by the Office National d'Etudes et de Recherches Aerospatiales (ONERA) and Snecma (Ref 18); and CMSX-2, CMSX-3 (Ref 19), and CMSX-6 (Ref 20) by Cannon-Muskegon Corporation These alloys are characterized by approximately the same creep-rupture strength (density corrected) but have differing SX castabilities, grain qualities, solution heat treatment windows, propensities for recrystallization upon solution treatment, environmental oxidation and hot corrosion properties, and densities Typical stress-rupture properties are shown in Fig 9 and 10

Fig 9 Larson-Miller stress-rupture strength of CMSX-2/CMSX-3 versus DS MAR-M 247, using 1.8 mm (0.070

in.) specimens machined from blades

Fig 10 Larson-Miller specific stress-rupture strength of CMSX-6 versus CMSX-2/3 MFB, machined from blade;

GFQ, gas fan quenched; AC, air cooled

References cited in this section

13 M Gell, D.N Duhl, D.K Gupta, and K.D Sheffler, Advanced Superalloy Airfoils, J Met., July 1987, p

11-15

14 T.E Strangman, G.S Hoppin III et al., Development of Exothermically Cast Single Crystal Mar M 247 and Derivative Alloys, in Proceedings of the Fourth International Symposium on Superalloys (Seven Springs,

PA), American Society for Metals, 1980, p 215-224

15 G.S Hoppin III and W.P Danesi, Manufacturing Processes for Long Life Gas Turbines, J Met., July 1986

16 C.S Wukusick, Final Report, Contract N62269-78-C-0315, Naval Air Systems Command, 25 Aug 1980

17 D.A Ford and R.P Arthey, Development of Single Crystal Alloys for Specific Engine Applications, in

Proceedings of the Fifth International Symposium on Superalloys (Seven Springs, PA), The Metallurgical

Society, 1984, 115-124

18 E Bachelet and G Lamanthe, AM1I, High Temperature Superalloy for Turbine Blades, Paper presented at the National Symposium on Single Crystal Superalloys, Villard-de-Lans, France, Feb 1986

19 K Harris, G.L Erickson and R.E Schwer, "Development of the Single Crystal Alloys 2 &

CMSX-3 for Advanced Technology Turbines,&#CMSX-34; Paper 8CMSX-3-GT-244, American Society of Mechanical Engineers

20 J Wortmann, R Wege, K Harris, G.L Erickson, and R.E Schwer, Low Density, Single Crystal Superalloy

CMSX-6, in Proceedings of the Seventh World Conference on Investment Casting (Munich, West

Germany), European Investment Casters' Federation, 1988

Directionally Solidified and Single-Crystal Superalloys

K Harris, G.L Erickson, and R.E Schwer, Cannon-Muskegon Corporation

Chemistry and SX Castability

Alloy CMSX-2 was developed in 1979 from the MAR-M 247 composition using some of the experience of the Garrett/NASA MATE program (Ref 14) A multidimensional development approach was used to achieve a high level of balanced properties (Fig 11) The chemistry modifications applied to MAR-M 247 to develop CMSX-2 (Table 1) are summarized below with respect to function and objectives:

achieve a very high incipient melting temperature (1335 °C, or 2435 °F)

• Partial substitution of tantalum for tungsten (CMSX-2 has 6% Ta) for good single-crystal castability, high γ' volume fraction (68%), improved γ' precipitate strength, microstructural stability (freedom from α-tungsten and tungsten, molybdenum-rich μ phases), good oxidation resistance, and coating stability

window (difference between the γ' solvus and the incipient melting temperature), of at least 22 °C (40

°F)

close-packed phases

Fig 11 CMSX-2 alloy development goal

Figure 12 shows the relative potency of tantalum, tungsten, and molybdenum as solid-solution strengtheners in binary nickel alloys, where tantalum is the most powerful strengthener on an atomic percent basis An increase in the lattice parameter of the γphase due to alloy additions increases the solid-solution strengthening Tantalum also partitions strongly to the γ' phase, increasing the volume fraction and stiffening the γ' due to its relatively large atomic size The strength of the γ' phase is important in superalloys with a high volume fraction of γ' (>50%) because γ' shearing is the primary strengthening mechanism With the mean free edge-to-edge distance in the matrix between the precipitates

being smaller than the average precipitate size itself, dislocation shearing of the γ' particle is favored over Orowan dislocation looping around the γ' particles

Fig 12 Influence of alloying elements on the lattice parameter of binary nickel alloys Source: Ref 21

Detailed transmission electron microscopy studies of dislocation movement in cast high-strength superalloys, such as MAR-M 002 (Table 2) and its single-crystal derivative SRR 99 (Table 1), have shown the importance of ensuring that the antiphase boundary (APB) energy is high, so that the stacking fault mode of creep deformation occurs at temperatures up

to 850 °C (1562 °F), thus ensuring high creep strength (Ref 17) Tantalum additions raise the APB energy relative to the stacking fault energy (Ref 17), leading to the increased tendency for stacking faults to be formed at lower temperatures The CMSX-2 alloy is designed to provide good SX foundry performance because castability is a crucial alloy performance criterion for any complex, thin-wall turbine blade or vane component, a characteristic sometimes given

limited attention in alloy design It affects not only the yield and cost of components but also the defect level and therefore component performance Single-crystal casting defects of concern are:

• Freckling: A spiral of equiaxed grains caused by elemental segregation in the liquid state

• Slivers: Moderate-angle grain defects

• Microporosity: A uniform distribution of interdendritic micropores

• Spurious grains: High-angle grain boundaries

• Stable oxide inclusions: Al2 O3

• Carbides: TiC

The partial substitution of tantalum for tungsten in the CMSX-2 alloy, compared to the MAR-M 247 chemistry, helps overcome the freckling problems inherent in the low-tantalum, high-tungsten single-crystal alloys The strong γ'-forming elements, aluminum and titanium, which are also low density, tend to segregate to the last liquid to solidify in the interdendritic spaces during the SX solidification process This can create density changes and consequential flow in the liquid metal close to the solidification front, which can nucleate freckle trails of equiaxed grains This can occur particularly under conditions of low or changing thermal gradients Tantalum, which is a strong γ'-forming element of high density, also tends to segregate to the last liquid to solidify in the interdendritic spaces and thus evens out these density changes in the liquid, or mushy, zone and reduces freckling tendencies

Several studies undertaken in the United States, Europe, and Japan confirm that high [N] and [O] levels in single-crystal superalloy ingot adversely affect SX casting grain yield, supporting the importance for low [N] and [O] levels in the master alloy Carbon, sulfur, and [O] master alloy impurities are shown to transfer nonmetallic inclusions, such as Al2O3, (Ti,Ta) C/N, and (Ti,Ta)x S, to SX parts (Ref 22) Grain defects can nucleate on these inclusions

Several second-generation, rhenium-containing, single-crystal superalloys have been developed for turbine engine applications Two typical compositions are given in Table 3 Rhenium partitions mainly to the γ matrix; this retards coarsening of the γ'-strengthening phase and increases γ/γ' misfit (Ref 23) Atom-probe microanalysis of rhenium-containing modifications of the PWA 1480 and CMSX-2 alloys reveals the occurrence of short-range order in the matrix with small rhenium clusters (~1.0 nm, or 10 Ao , in size) detected in the γ in the alloys (Ref 24) The rhenium clusters can act as more efficient obstacles against dislocation movement compared to isolated solute atoms in the γ solid solution; therefore, they play a significant role in improving the creep strength The Larson-Miller stress-rupture comparison of CMSX-4 and CMSX-2/3 is shown in Fig 13 The stress-rupture temperature capability advantage of CMSX-4 over CMSX-2/3 is 27 °C (48 °F) (density corrected) in the 248 MPa/982 °C (36 ksi/1800 °F) region In the 103 MPa/1121 °C (15 ksi/2050 °F) region, the stress-rupture temperature capability advantage is 30 °C (54 °F) (density corrected) The data also indicate that CMSX-4 has a potential peak-use temperature under stress of at least 1149 °C (2100 °F)

Fig 13 Larson-Miller stress-rupture strength of CMSX-4 versus CMSX-2/3

Single-Crystal Casting Techniques. A variety of single-crystal airfoil component-casting techniques have been

developed to production status around the world in the last 10 years Most involve a withdrawal-type vacuum induction casting furnace with mold susceptor heating Cooling plate sizes range in diameter from 140 to 610 mm (51

2 to 24 in.) Some of the developed SX casting techniques are presented in Ref 12, 13, 25, and 26

The modern helicopter engine turbine vane shown in Fig 14 represents a difficult cored configuration The large shrouds and core make this vane susceptible to shrinkage, grain nucleation, and recrystallization during solution heat treatment Single-crystal casting processes developed by the Allison Gas Turbine Division of General Motors Corporation result in high yields for this vane in CMSX-3 Similar yields have been demonstrated with CMSX-4 using the same Allison casting process

Fig 14 SX turbine vane cast in CMSX-4

Single-Crystal Heat Treatment and Microstructures. With regard to solutioning, the latest multistep ramped cycles developed for single-crystal components are designed to completely solution the γ' and most of the γ/γ' eutectic without incipient melting An additional benefit of the high-temperature cycles is the element homogenization effect, as shown in Fig 15 Alloy CMSX-4, which is solutioned at a maximum temperature of 1321 °C (2410 °F) in commercial vacuum heat treatment furnaces, readily attains the 99%+ (<1% remnant γ/γ' eutectic) solutioned microstructure, as illustrated in Fig 16