Neural network model of creep strength of austenitic stainless steels

The creep rupture life and rupture strength of austenitic stainless steels have been expressed as functions of chemical composition, test conditions, stabilisation ratio, and solution treatment temperature. The method involved a neural network analysis of a vast and general database assembled from published data. The outputs of the model have been assessed against known metallurgical trends and other empirical modelling approaches. The models created are shown to capture important trends and to extrapolate better than conventional techniques.