On the stress dependence and activation area for creep

Abstract

The formalism of reaction rate theory with respect to stress dependence and activation area for activated dislocation motion is reviewed. Some errors of previous treatments are corrected. In particular the forward and backward activation areas are clearly distinguished and shown to be in general quite different. Also, the prediction of a hyperbolic sinelaw stress dependence for all mechanisms is shown to be incorrect and the correct expression is derived. It is further shown that any and all mechanisms yield stress-dependencies which vary from linear at low stresses, to approximate “power-law” behavior at intermediate stresses, to approximately exponential at high stresses. Thus the often-seen practice of equating such changes in stress-dependence, a priori, with changes in rate-controlling mechanism is unwarranted. Specific models for diffusion-controlled creep are examined with regard to stress-dependence and activation areas, several errors concerning the predicted stress-dependence for the jogged-screw model are corrected and this model is shown to be incompatible with experimental results. The rate-controlling dislocation mechanism(s) for diffusion-controlled creep remains obscure.