A neural network model to predict low cycle fatigue life of nitrogen-alloyed 316L stainless steel

Low cycle fatigue properties of nitrogen-alloyed 316L stainless steel (SS) has been studied at various temperatures between room temperature and 873 K. Four heats of the alloy with nitrogen contents of 0.042, 0.103, 0.131 and 0.151 wt% were tested in fully reversed loading conditions at a constant strain amplitude of ±0.5% and strain rate of 2 × 10−3 s−1. Using the data, an artificial neural network model was developed to predict fatigue life of nitrogen-alloyed 316L SS. The neural network model could predict fatigue life within a factor of 2.0 of the experimental values over the whole range of test temperature and nitrogen content. The model was expanded to develop a unified model to predict fatigue life of 316 SS grade of stainless steel with and without nitrogen, and with normal and low carbon contents. The fatigue test parameters covered a wide range of temperatures, strain ranges and strain rates. The unified model was found to predict fatigue life at any test condition within a factor of 2.