Prediction of Residual Stress Induced by Laser Shock Processing Based on Artificial Neural Networks for FGH4095 Superalloy

Abstract FGH4095 superalloy samples were single point treated by laser shock processing (LSP) with laser energy of 5–7 J, and shocked times of 1& 3, and laser profile of Gaussian distribution & Flat top distribution. The residual stress of treated samples were determined by PROTO LXRD stress device. Artificial neural network (ANN) was used to predict the residual stress induced by LSP. The experimental residual stresses with laser energy of 5 J and 7 J were used as training datasets, and that with laser energy of 6 J were reserved as test datasets to validate the trained network. And this work showed a good fitness of experimental results and predicted results, which can provide theoretical reference for predicting the mechanical properties and servery performance of materials by LSP.