go to HTM21 Home Page

go to NIMS Superalloy home

DESIGN AND DEVELOPMENT OF NI-BASE SINGLE-CRYSTAL SUPERALLOYS ON ATOMIC BASIS

COMPARISON BETWEEN APFIM ANALYSIS AND NUMERICAL ESTIMATES

Ni-base single crystal superalloys, which have remarkable properties at elevated temperatures, have been designed for use as turbine blades in aeroengines. In order to develop higher efficiency engines, there are still considerable efforts being devoted to enhance the temperature capabilities of superalloys. Common superalloys consist of g and g' phases, which have the face-centred cubic and L12 structures, respectively. Since the nature and interactions between these two phases to a large extent affect the properties of Ni-base single crystal superalloys, understanding the atomic structure of the g and g' phases is of the utmost importance in the development of novel alloys with superior mechanical properties. Atom probe field ion microscope (APFIM) , which has spatial and chemical resolution at an atomic level, is an appropriate method for investigating the atomic structure of the two phases. Numerical methods have been proposed to predict the atomic structure and mechanical properties of Ni-base superalloys. Our research group have established an alloy design program (ADP), based on a semiemprical rationalisation of available data [6-8]. It has a demonstrated ability to predict the compositions of the g and g' phases, g / g' lattice misfits, creep rupture life etc. Using this method, we have developed several Ni-base superalloys with superior mechanical properties at elevated temperatures [1-5]. TMS-63 and TMS- 71, the compositions of which are listed in table 1, are tintroduced in this article.

As a more fundamental approach, we have also applied the cluster variation method (CVM) to Ni-base superalloys. This method employs Leonard-Jones pair potentials, and allows the equilibrium phase chemistries, volume fractions and lattice parameters to be estimated as a function of alloy composition, temperature and pressure. Some important factors affecting the mechanical properties of nickel-base superalloys can be estimated with the Two numerical methods described above. In FIG.1, the composition data of TMS-63 obtained from CVM and ADP calculations are summarized and compared against APFIM experiments. It is obvious that experiment and theory compare well for all the alloys examined.

It is also important to investigate the site occupancy of alloying elements in the g' phase since whether the alloying element prefers the Ni sites or the Al sites drastically alters the g' volume fraction, eg. alloying elements which are likely to substitute for the Al site increase the g' volume fraction whearas those likely to substitute for the Ni site decrease the volume fraction. The "layer-by-layer analysis" using APFIM makes it possible to experimentally determine the substitution site of solute atoms in the g' phase. FIG.2 shows the summarized site occupancy probabilites derived from APFIM analyses and CVM calculations in the case of TMS-63. The experimental results confirm that Ta and Mo prefer to substitute for the Al sites, while Cr prefers the Ni sites. These site occupation tendencies are generally good agreement with theoretical prediction derived from CVM.

VARIATIONS IN LOCAL CHEMISTRY AT g / g' INTERFACES

Composition changes at the g / g' interfaces can be determined by APFIM analysis. These interface analysis has been conducted for TMS-71 with 1300x4h~AC solution treated and 1100x1h~AC ageing heat treatments. FIG. 3 is the typical example of the composition change at the interface.


Fig. 3 Compositional changes at a g and g' interface , showing enrichment of Re in the g phase adjacent to the interface.

Re enrichment of the g phase adjacent to the interface was observed, which can be explained as in FIG.4.

Under thermodynamically equilibrium condition, the volume fraction of the g' phase is higher when temperature is low. It is thus expected that the g' phase tends to increase in size during cooling process. Since Re has strong preference to partition into g phase, Re atoms are depleted from the g' phase. Because of its low mobility, Re atoms are likely to enrich at the interface. This tendency of Re may play a role in retarding the coarsening of g' phase during cooling, which is advantageous for practical heat treatment of the alloys. FIG. 5 shows the temperature capability of alloys as a function of lattice misfit. It is suggested in this figure that the temperature capability can be improved by changing the lattice misfit towards negative direction. However, if the lattice misfit is too large negative value, the coherancy easily breaks down during heat treatment, or even during cooling process. In practice, the alloy TMS-64 is still very sensitive to heat treatment so as to maintain coherency although this alloy has demonstrated superior temperature capabilities under optimum heat treatment procedures. It is thus expected that TMS-71, which has slightly negative misfit at room temperature, can be higher temperature capabilities than any orher alloy existing and less sensitive to cooling process.

FIG.5 The predicted and experimentally observed temperature capabilities of some Ni-base single crystal superalloy as a function of lattice misfit. Re containing TMS-71 is a promising material with good mechanical properties which are less sensitive to its thermal histroy.

REFERENCES
  1. H.Harada and M.Yamazaki, Tetsu to Hagane, 65, (1979) 1059.
  2. H.Harada, K.Ohno, T.Yamagata, T.Yokokawa and M.Yamazaki, Proc. of the 6th International Symposium on superalloys, Seven Springs, (1988) 305.
  3. H.Harada, T.Yamagata, S.Nakazawa, K.Ohno and M.Yamazaki, Proc. of a Conference "High Temperature Materials for Power Engineering 1990", held in Liege, Belgium, (1990)1319.
  4. T.Yokokawa, K.Ohno, H.Harada, S.Nakazawa, T.Yamagata and M.Yamazaki, Proc. of the 5th inter. conf. on Creep and Fracture of Engineering Materials and Structures, Swansea, U.K,.(1993) 245.
  5. H.Harada, T.Yamagata, T.Yokokawa, K.Ohno and M.Yamazaki, Proc. of the 5th inter. conf. on Creep and Fracture of Engineering Materials and Structures, Swansea, U.K,.(1993) 255.
  6. M.Enomoto and H.Harada, Metall. Trans., 20A, (1989) 649.
  7. M.Enomoto, H.Harada and M.Yamazaki, CALPHAD, 15, (1991) 143.
  8. M.Enomoto, H.Harada and H.Murakami, Tetsu-to-Hagane, 80, (1994), 57.
PUBLICATIONS
  1. H.Harada, A.Ishida, H.Murakami, H.K.D.H.Bhadeshia and M.Yamazaki, Atom-probe microanalysis of a nickel-base single crystal superalloy, Appl. Surf. Sci., 67, (1993) 299.
  2. H.Murakami, H.Harada and H.K.D.H.Bhadeshia, The location of atoms in Re- and V- containing multicomponent nickel-base single- crystal superalloy, Appl. Surf. Sci., 76/77 (1994) 177.
  3. H.Murakami, P.J.Warren and H.Harada, Atom-probe microanalyses of some Ni-base single crystal superalloys, Proc. of the 3rd Inter. Charles Parsons Turbine Conference, Newcastle upon Tyne, (1995) 343.
Contact MURAKAMI.Hideyuki@nims.go.jp

Back to Parent Page