Numerical study on springback prediction of aged steel based on quasi-static strain-hardening material model

Abstract

Material model of aged steel, the material’s stress and strain behavior in the region of yielding, plays a critical role in springback prediction of FEA simulation. In most of the FEA simulations, the material is modeled as a linear elastic and then after yielding, a linear strain-hardening model or a power-law strain-hardening model. In the prediction of springback by FEA simulation, there is always a problem that the prediction is far away to the true springback of the products or samples. This paper employs a quasi-static strain-hardening material model which gives a more accurate description of the material properties of aged steel for FEA simulation of the forming. A typical V-bending for aged steel is simulated with finite element software, ABAQUS, to predict the springback. A 2D model which consists of multiple rigid bodies and a deformable body is performed in this study adopting four-node linear plane-strain elements. Meanwhile, a 2D model that applies a power-law strain-hardening material model is implemented as a contrast. By comparing the results from the simulations based on different material models with those from experiments, a better correlation between the simulations using a quasi-static strain-hardening material model and the experiments is achieved. Besides, the bending process for aged steel is also analyzed utilizing quasi-static strain-hardening material model with variable ageing coefficients. It is drawn that the greater the ageing coefficient within certain range, the greater the springback simulation accuracy by comparison between simulation and experiment results. In addition, as the upper yield stress is not easy to be measured directly, a method for reverse calculation of upper yield stress through springback ratio and quasi-static flow stress is proposed.