Pores have been observed to move bodily within solid materials. Such motion may occur in a more-or-less random manner when no driving forces are imposed on the specimen or it may be biased due to imposed gradients. The possible mechanisms for movement are surface diffusion, volume diffusion through the matrix and volume diffusion through the vapor phase (vapor transport). Important driving forces include: thermal gradient, strain energy gradient, bowed dislocation, curved grain boundary and electric potential gradient (for ionic crystals). A unified treatment for the calculation of driving forces and diffusion-controlled mobilities for the various mechanisms is presented, which points out certain errors and unnecessary complexities introduced by previous authors. Especially highlighted are the complementary approaches using atomic or pore mobilities. Comparisons are then made between the theoretically-derived expressions and available experimental observations, both with and without imposed driving forces. The overall impression garnered from such comparisons (some of which involve rather indirect experimental observations) is one of substantial agreement between theory and experiment. Possible explanations are offered for the apparent discrepancies between theory and experiment, which exist primarily in the cases of no imposed driving forces; in particular, the effects of faceting, dislocations and grain boundaries in impeding pore motion are discussed. © 1969.
