Self-Diffusion during Plastic Deformation

Attention is directed to the uncertain state-of-affairs surrounding the influence of plastic deformation on concommitant tracer-diffusivity in metals. We now offer some clarification of this elusive solid-state transport problem.It is concluded that strain-enhanced diffusion is a real phenomenon, and in fact, it can be of substantial magnitude if the strain-rate is high enough and/or the temperature low enough. However, the dynamic-diffusivity measurements can be obscured by spurious mass movements in the diffusion zone arising from inadvertent specimen roughening or nonuniform bulk flow. The diffusivity enhancement per se occurs during compressive or tensile plastic deformation in poly- or mono-crystals, and in face-centered or body-centered cubic metals.Arguments are presented against the vacancy model as an explanation of the observed deformation/diffusion effects; instead, the dynamic diffusivity is accounted for by the presence of moving dislocation which provide high-diffusivity paths for the tracer atoms. The concentration profile remains Gaussian, indicating that the overall process is essentially one of random-walk under the prevailing conditions. Then, the temperature-dependence of the strain-enhanced diffusivity simply becomes the activation enthalpy for normal lattice diffusion minus that for dislocation-pipe diffusion. This interpretation can be checked by independent diffusion measurements under static conditions.The “window” on the experimental determinations of strain-enhanced diffusivities is limited by (a) the requirement for a steady-state structure during the deformation/diffusion run, in order to achieve a constant dynamic diffusivity, (b) the selection of strain-rate/temperature combinations which will permit the dynamic diffusivity to be significantly larger than the static diffusivity, and (c) the avoidance of recrystallization which serves to eliminate the operative dislocation substructure.