An examination of a substructure-invariant model for the creep of metal matrix composites

The creep behavior of metal matrix composites may be analyzed using two different approaches: (i) by assuming flow is controlled by creep of the matrix material so that the true stress exponent, n, is ∼3 or ∼5, or (ii) by invoking a substructure-invariant model requiring that n=8. A detailed analysis of published creep data for aluminum-based composites suggests that the substructure-invariant model is untenable because it leads consistently to activation energies for creep which are higher than the value anticipated for lattice self-diffusion in aluminum.