Abstract
A theoretical formulation of ionic fluxes set up in spinel crystals during dislocation climb is given, with the ultimate aim to achieve a realistic description of climbing dislocations (local core charge and local core composition) and climb rates (i.e. strain rates). Quantities such as the effective diffusion coefficient, the climb velocity, the electric charge per unit length of dislocation, and the spinel composition (n in (Al2O3) nMgO) at the core, are derived in terms of the acting stresses, the temperature, the composition n, and the element diffusivities. The most interesting effect is the possible composition deviation Δ n at the core of a climbing dislocation. Under some conditions detailed in the paper, it is suggested that deviations as large as 0.5-1 could be built up in particular for large n values (e.g. n = 3.5) and /or under the high internal stresses acting during usual climb annealings. Such deviations could explain in turn non-planar faulted ribbons (or ribbons lying apparently on almost any plane in the lattice), or could trigger the microprecipitation of a second phase onto dislocation cores