Abstract-Weanalyze the growth of cavities along grain interfaces by the combined processes of grain boundary diffusion and plastic dislocation creep in the adjoining grains. It is shown that the coupling between the processes can be expressed in terms of a parameter L, which has the dimensions of length and which is a function of material properties, temperature and applied stress; L decreases with increasing temperature and stress and has, e.g., values in the range of 0.25 to 25 pm for various pure metals when stressed to 10m3 x shear modulus at 0.5 T,. The contribution of dislocation creep to the cavity growth rate is shown to be negligible when L is comparable to or larger than the cavity spacing, but significant interactions occur. leading to growth rates very much in excess of those predicted on the basis of boundary diffusion alone, when L is comparable to or smaller than the cavity size. The coupling occurs because extensive dislocation creep allows local accommodation of matter diffused into the grain boundary from the cavity walls. The cavity growth rate is analyzed by formulating a new variational principle that governs combined processes of grain boundary diffusion and non-linear viscous creep, and by implementing this principle through the finite-element method to obtain numerical solutions. Results for the cavity growth rate are presented for a wide range of ratios of L to cavity spacing. and of cavity radius to spacing. Also, results are presented for the total growth time of cavities from an initial size to final coalescence. [6,7].However, the works mentioned neglect the influen~ of plastic creep flow on the diffusive cavitation process. Our aim here is to model the combined effects of diffusion and creep flow on cavity growth. As will be seen, at stress levels of the order 10m3 p and higher (p = shear modulus) at 0.5 T,,, (T,,, = melting temperature) interactions between diffusive transport and creep flow are typically very important, although not at much iower stresses, of the order 10e4 p at this temperature. These combined effects lead to rates of cavity enlargement which can be appreciably greater than would be the case for either mechanism acting in isolation. Indeed, the possible importance of plastic creep flow to rupture at elevated temperature is suggested indirectly by the well-known MonkmanGrant [S] correlation, in which the product i,,t, (i, = steady state creep strain rate, t, = rupture time) is sometimes found to vary only slightly over ranges of stress and temperature which correspond to large changes in both factors.A previous attempt to model the combined effects of creep and diffusion on cavity growth was made by Beere and Speight [9]. Their model constitutes only a very rough approximation, and is based on the concept of each cavity being surrounded by a spherical she11 of effectively non-peeping material, within which a Hull-Rimmer diffusive cavitation process takes place, with these shells being embedded in a matrix of uniformly creeping material. The model has been modifie...