Diffusion creep in ceramics

At sufficiently low stresses, where intragranular dislocation mobility is rather limited, plastic deformation can occur solely by the diffusion of vacancies either through the matrix (Nabarro-Herring) or along grain boundaries (Coble). The process of diffusion creep has been modelled theoretically for over 50 years now, although there still remain some doubts over the experimental validation of the models. It has been suggested that low intragranular dislocation mobility leads to more frequent observations of diffusion creep in ceramics compared to metals. In ceramics, the diffusion creep process is more complex than in metals due to the need to account for charge balance and the transport of two or more ionic species along two different paths (lattice or grain boundary). The report critically evaluates the experimental observations of diffusion creep in some oxide-based ceramic systems. In addition, the process of ambipolar diffusion in ceramics is examined with two different considerations: (a) the total flux to grain boundaries is in the appropriate stoichiometric ratio, so that the cations and anions may be transported along different paths, and (b) the flux along each transport path is in the appropriate stoichiometric ratio. It is demonstrated that the above two considerations lead to substantially different predictions on rate controlling processe