Constitutive model of 25CrMo4 steel based on IPSO-SVR and its application in finite element simulation

Abstract

Optimizing the forming process of high-speed railway axles requires establishing an accurate model of the flow behavior of 25CrMo4 steel. In this regard, this study investigated the thermal deformation behavior of 25CrMo4 axle steel by conducting isothermal compression tests using a Gleeble-1500 thermal simulator within a certain temperature range (1323–1423 K) and strain rate range (0.01–1.0 s−1). In order to establish the constitutive relationship model of 25CrMo4 steel at high temperature, this study uses an intelligent algorithm (IPSO-SVR) combining improved particle swarm optimization and support vector regression. The study also analyzed the predictive performance of the model and examined the impact of key parameters of PSO and SVR on prediction accuracy. Subsequently, the study utilized the predicted data of the IPSO-SVR model as input to the finite element software to numerically simulate high-temperature deformation. The research results indicate that the IPSO-SVR constitutive model has high prediction accuracy and can effectively characterize the dynamic mechanical properties of 25CrMo4 steel. Therefore, this method provides a feasible and effective solution for finite element simulation.