The role of intergranular precipitates in controlling creep cavitation
A satisfactory model for cavitational failure in creep must account for the fact that fracture can occur under a very low stress, for example, only 0.7 MPa for a solid solution magnesium alloy. A mechanism for growth based on the transfer of vacancies from high angle grain boundaries to intergranular cavities satisfies this low stress requirement for it converts a relatively high fraction of the work done by the applied load into surface energy of fracture. However, for such growth to proceed, cavity nuclei of radius greater than a critical value, r c , must exist on those grain boundaries which are approximately normal to the applied tensile stress axis. It can be shown quite simply that r c = 2y/o, where y is the surface energy per unit area and o the applied tensile stress. A typical value for r c is 1 pm which is far too large to occur spontaneously by chance accumulation of vacancies. It is in fact generally agreed that cohesion is lost owing to the concentration of stress at some small obstacle in a sliding grain boundary. These cavities are nucleated along the boundary under applied stresses which are lower than those needed to cause triple point cracking where the whole of the length of the boundary is available to concentrate stress. This was a puzzle until Smith & Barnby (1967) demonstrated that the stress concentrated at a small obstacle in a sliding boundary was far higher than that concentrated at a very large obstacle as incorporated in, for example, the Stroh derivation.