Simulation of cyclic behavior of mild steel with empirical model parameter equations

Abstract

Mild steel is a structural material widely used in seismic force-resisting systems due to its superior ductility. When conducting numerical analyses of structures made of mild steel for seismic loads, it is essential to use a material model that can simulate the cyclic behavior of mild steel, including its peculiar characteristics such as cyclic hardening and the Bauschinger effect. In this study, a combined hardening model containing one isotropic and three kinematic hardening models was constructed to accurately simulate the cyclic behavior of mild steel. In addition, empirical equations were proposed to conveniently calculate constituent modeling parameters of the combined hardening model using mechanical properties obtained from monotonic tensile tests. For this purpose, the life cycle fatigue (LCF) test results of 92 mild steel coupons were collected in this study. To determine the coefficients of the proposed equations, three regression algorithms were considered. The particle swarm optimization (PSO) produced the smallest error; thus, the PSO was used to determine the coefficients. To validate the accuracy of the proposed material model, the measured and simulated cyclic curves of two steel welded unreinforced flange-welded web connection specimens were compared. The constructed combined hardening model and proposed empirical equations reliably predict the cyclic behavior of mild steel; thus, numerical analyses (finite element analyses) with the constructed material model can be used to simulate the cyclic behavior of mild steel members such as welded unreinforced flange-welded web (WUF-W) connections.