The Microstructure of Superalloys

Madeleine Durand-Charre

• Editore: CRC Press
• Anno: 2017
• ISBN:
  9781351409810
• Edizione: 1
• Argomento: Technologie et ingénierie
• Categoria: Science des matériaux

Presents all the main aspects of the microstructure of nickel-base superalloys, and includes micrographs chosen from among a large range of commercial and academic alloys, from the as-cast product to in-situ components, worn from in-service use. Including more than 100 illustrations, the text explains all the transformation mechanisms involved in the origination (creation) of microstructures during solidification or heat treatments (crystallization paths, segregation, crystal orientation, precipitation, TCP, coarsening and rafting, etc.). It includes up-to-date information and data such as phase diagrams, crystallographic structures, and relationships with functional properties. Nearly 300 references provide a key to further investigation.

1 General Aspects of Superalloys

1.1 Historical Development Of Superalloys

The term “superalloys” is used somewhat loosely to describe a wide range of high performance materials which usually combine high strength and corrosion resistance at temperatures above the limits accessible to martensitic steels, i.e. 650°C and above. It is sometimes extended to include chemically or structurally related materials designed to fulfil more specific functions. A common feature of all these materials is the face-centered cubic “austenitic” crystal structure of the matrix phase. Strenghthening is obtained chiefly by solid solution hardening and carbides in the cobalt-based grades¹ and nickel-based sheet alloys, whereas other wrought and cast nickel-and iron-based alloys are also strengthened by precipitation of an intermetallic phase. The present work will be devoted essentially to this latter category, which is most characteristic of a “typical” superalloy.

The development of creep- and corrosion-resistant alloys dates back to immediatly before the first world war, driven essentially by the requirement of exhaust valves and superchargers for aircraft engines. However, in France, an Fe-60%Ni-ll%Cr alloy hardened by tungsten and carbon was patented by the company now called Imphy S.A.² and was designated ATG (alliage pour turbines à gaz), reflecting contemporary preoccupations with the as yet unsuccessful development of land-based gas turbines. This alloy was subsequently used for ammonia synthesis vessels, under the designation BTG. Many later valve and supercharger alloys were strengthened in the same way by tungsten additions. Another patent³ filed in 1917 by Chevenard at Imphy, corresponded to the alloy ATV (alliage pour turbines à vapeur), in response to intergranular corrosion problems encountered in steam turbine blading steels in impure steam. This stable austenite iron-base alloy contained 34% Ni and 11% Cr, to enable the addition of 0.3% C, necessary to acheive the required yield strength, while preventing the formation of iron carbides believed to be responsible for the corrosion in less stable Fe-Ni steels. However, it was soon realized that chromium carbide precipitation could also lead to corrosion problems. In order to enable the use of lower carbon levels, Chevenard sought another stengthening mechanism and imagined the possibility of using precipitation hardening phenomena similar to those already known in light alloys. This led to the discovery of the hardening effect of aluminium and titanium additions, and a patent was filed in 1929^4-6, covering austenitic Fe-Ni-Cr alloys with a wide range of nickel contents. A few months earlier, in the same year, Pelling, Merica and Merica⁷ had filed a similar patent in the USA, to be published only six years later, particularly claiming the effect of titanium, but also mentioning aluminium, in a nickel-chromium matrix of the type used for electrical heating alloys. Chevenard’s work, which led to the family of ATVS alloys, together with the superior creep strength of nickel-chromium matrices, formed the basis for the develpment of the Nimonic alloys by Pfeil et al.,⁸ in Great Britain, immediatly before and during the second world war. The nature of the hardening phase Ni₃(Al,Ti), known as gamma prime (γ’) phase, was only determined much later, in 1957, by Betteridge and Franklin⁹, using X-ray diffraction techniques. These materials maintain their strengths at temperatures above 650°C, where those of the best martensitic steels fall rapidly. This is particularly true in alloys containing little or no iron.

The guiding line in subsequent superalloy development was essentially a gradual increase in the amounts of aluminium and titanium, i.e. in the volume fraction of the γ’ hardening phase, together with the addition of solid solution strengthening elements. This evolution was accompanied by intensive constitutional research in order to carefully balance alloy chemistries and thus avoid the formation of undesirable embrittling phases^11,12. Since aluminium and titanium have high chemical reactivities, it became necessary to adopt vacuum melting techniques, followed by cold-crucible remelting to control solidification and reduce segregations. Furthermore, because of the high mechanical strength of these materials, they are difficult to hot work in the presence of γ’. For volume fractions greater than about 40-45%, the interval between the solvus and incipient melting temperatures, within which an alloy can be worked in the single phase field, becomes too narrow for all practical purposes. Components with volume fractions higher than this must be produced by casting^13,14, although the limit can be raised to a certain extent by the use of powder metallurgy processing, which refines the scale of segregation.

An important application of cast superalloys is for the production of turbine blades. It was found that the grain boundaries represented weak points and that performance could be significantly improved by directional solidification to produce columnar structures with the grains aligned parallel to the blade axis. Further progress was obtained when this technique was modified to eliminate grain boundaries entirely, leading to blades each consisting of a single crystal. Moreover, the absence of grain boundaries made it possible to remove boundary strengthening elements such as carbon, boron and zirconium, increasing the solidus temperature and providing scope for further alloy development¹⁵. This recent family of “single crystal superalloys” will be discussed in detail in the present monograph, and where appropriate will be used as a reference for the behavior of more conventional grades.

Over the last 15 years or so, the maximum metal temperature in turbine blades in service has been increased by nearly 300°C, half of this gain being due to blade design and half to progress in alloy development and processing, involving optimization of both composition and microstructure^16-18. The cost of a modern single crystal turbine blade is several hundred times that of an equivalent weight of microalloyed steel, reflecting not only the noble or rare elements of which it is composed, but more particularly the high degree of technological sophistication attained.

1.2 Strengthening Mechanisms In Superalloys

Nickel-base superalloys have a matrix, γ, with a face-centered cubic structure, containing a dispersion of ordered intermetallic precipitate particles of the type Ni₃Al (γ’). In order to increase the strength of the initial simple Ni-Cr-Al-Ti alloys, the approach has been to strengthen the matrix by adding other elements that can be taken into substitutional solid solution, and to increase the volume fraction of γ’¹⁹. The grain boundaries are reinforced by carbide precipitation and by the use of minor additions of boron and zirconium to increase boundary cohesion. The latter strengthening method obviously does not apply in the case of single crystal materials.

Because of its electronic structure, the f.c.c. nickel lattice has a large solubility for many other elements²⁰. Solid solution strengthening is caused partly by lattice distortion, and therefore increases with atomic size difference, up to a maximum of about 10%. High melting point elements provide strong lattice cohesion and reduce diffusion, particularly at high temperatures. Molybdenum and tungsten are thus particularly effective for both these reasons. Atomic clustering or short range order can also strengthen the matrix. This is an electronic effect and is observed with molybdenum and tungsten, together with chromium and aluminium. These elements thus produce greater hardening than iron, titanium, cobalt or vanadium. However, strengthening due to short range order generally diminishes rapidly above about 0.6 Tm, due to increased diffusion.

The origin of precipitation hardening is complex. The size and spacing of the particles and therefore their volume fraction are important factors²¹. Pure γ’ is quite unusual in that its intrinsic strength is low at low temperatures and increases with temperature up to a maximum at about 700-750°C, probably due to the formation and interaction of complex faults. At low temperatures, pure γ’¹ can be sheared relatively easily by super-dislocation-stacking fault systems. However, in two-phase materials, for matrix dislocations to penetrate a γ’ particle, it is first of all necessary to create the corresponding stacking faults or antiphase boundaries, and this requires considerable expenditure of energy. Matrix/particle interfaces therefore play an important role. For low volume fractions and fine particles, the leverage available can be sufficient to produce shearing. When the inter-particle spacing is large and the precipitates are coarse (low volume fractions) matrix dislocations can bend round the particles by the Orowan mechanism. However, even in single crystal superalloys, where volume fractions exceed 60%, dislocations are often seen to bow into the narrow matrix corridors, pile-ups of several dislocations then being necessary to attain the required stress²², which nevertheless remains insufficient to cause shearing of the γ’. The ability of γ’ to resist shearing depends on its composition, and this is one of the important factors in alloy optimization (cf. § 7.2.5)

1.3 Manufacturing Processes

As already mentioned, the progress in superalloy performance would not have been possible without the parallel development of processing technologies. Thus, the use of vacuum induction melting (VIM) and vacuum arc remelting (VAR) enabled the control of undesirable residual elements and non-metallic inclusions, together with the refinement of ingo...