A method of distinguishing between diffusion creep and Harper-Dorn creep at low stress levels

Abstract

When polycrystalline pure metals and solid solution alloys are tested under creep conditions at elevated temperatures, the steady-state creep rate, {dot {epsilon}}, varies with the applied stress, {sigma}, raised to a power, n, which is typically in the range of {approximately}3--6 at high stresses but with a transition to Newtonian viscous flow with n = 1 at very loss stresses. Two distinct creep mechanisms have been proposed to account for creep at low stresses when n = 1: (1) diffusion creep where a permanent strain arises from vacancy flow either through the lattice (Nabarro-Herring diffusion creep) or along the grain boundaries (Coble diffusion creep) and (2) Harper-Dorn creep where strain is achieved through an intragranular dislocation process. There is considerable current debate concerning the relative significance of these two deformation processes. The present paper presents a solution to this apparent dichotomy. It is demonstrated in the following section that the processes of Harper-Dorn creep and diffusion creep will lead to significant differences in the appearance of marker lines which impinge on the grain boundaries at the specimen surface, thereby giving the potential for a measurable variation in the magnitudes of the estimated strains arising from grain boundary sliding during Harper-Dornmore » and diffusion creep. Furthermore, as demonstrated in the subsequent section, these anticipated differences are fully supported by measurements, already available in the published literature, taken under conditions of Harper-Dorn creep and diffusion creep, respectively. It is therefore concluded that both Harper-Dorn creep and diffusion creep are firmly established as potential deformation mechanisms in high temperature creep, and both processes should be taken into consideration in any detailed analysis of creep data at low stress levels.« less