Current manufacturing techniques used in the production of yttria-stabilized zirconia/porcelain dental prostheses lead to high rates of brittle failure from stress concentrations at porcelain defects. A high temperature dwell been proposed as a way to induce creep stress relaxation in the porcelain and reduce this failure mode. However, creep cavitation in porcelain is currently poorly understood. Therefore we propose performing SANS to identify the size, shape and orientation of creep induced defects as a function of temperature. Ex-situ creep specimens will be developed and the large dynamic Q range of Sans2d will be exploited to gain insight into the growth and nucleation of pores in the range 0.25-300nm. These results will be integrated into finite element models of the interface to predict the optimum processing parameters required to produce safer dental prostheses.