Abstract
Previous numerical calculations for a model of nickel alloys, employing the coherent potential approximation (CPA) and the random phase approximation (RPA), are extended to iron alloys. It is shown that often the effective scattering potentials are sufficiently weak for a rigid band calculation of the spin wave stiffness constant D to be a good guide to the more elaborate CPA calculations. The case of NiFe alloys is re-examined in the light of a recent calculation by Riedinger and Nauciel-Bloch (ibid., vol.5, p.732 (1975)). It is shown that D is very sensitive to the assumed exchange splitting in pure nickel and that within the RPA reasonable agreement with experiment is obtained only by assuming an unusually large value for this parameter.

This publication has 33 references indexed in Scilit: