Prediction of residual stress fields after shot-peening of TRIP780 steel with second-order and artificial neural network models based on multi-impact finite element simulations

Abstract

Shot-peening is a mechanical surface treatment widely employed to enhance the fatigue life of metallic components by generating compressive residual stress fields below the surface. These fields are mainly impacted by the selection of the process parameters. The aim of this work is to propose a hybrid approach to conduct two predictive models: second-order model and feed-forward artificial neural network model. For this purpose, a 3D multiple-impact finite element model coupled to a central composite design of experiments was employed. A parametric analysis was also conducted to investigate the effect of the shot diameter, the shot velocity, the coverage, and the impact angle on the induced residual stress profile within a TRIP780 steel. It was found that both models predict with good agreement, the residual stress profile as a function of the process parameters and can be used in shot-peening optimization due to their responsiveness.