A widely applicable dislocation model of creep

Publisher Website
Google Scholar
Abstract
A physical model, based on the average behaviour of large numbers of dislocations, is proposed to explain a wide variety of creep curves. The model, which is consistent with others in the literature, predicts that the rate of change of mobile dislocation density obeys zero, first and second-order kinetics. Whereas other models consider the mean velocity of mobile dislocations to be independent of dislocation density, here the velocity is allowed to decrease with increasing density to allow for work-hardening. Arguments are developed to show the form of this dependence. The resultant creep equation is able to predict most of the shapes of creep curves observed in practice, including an incubation period. Comparison of the expression with experimental data for a variety of materials (both single and polycrystalline) and conditions shows very good quantitative agreement indicating the validity of the assumptions made. Further confirmation of the model is provided by the fact that the analysis predicts a linear relation between dislocation density and strain for small strains, an observation frequently made in practice.