Modeling The Strain-Rate Dependent Yielding And Failure Behavior Of High Strength Steel Sheets

Sheet metals are one of the most impoartant structural materials in automobiles. Generally, the characterization work for such materials involves three important aspects, i.e. plastic anisotropy, facture criterion and strain rate effect. These three aspects are covered in turn in the present study on a DP780 steel sheet.Firstly, the plastic anisotropy of the material was investigated under quasi-static condition. Test results of uniaxial tension tests in seven different directions showed an obvious anisotropy in plasticity. Moreover, it turned out that the traditional Hill48 model which is based on associated flow rule was insufficient to describe the anisotropic properties of the flow stress and the Lankford r-value simultaneously. Consequently, the non-associated Hill48 model was adopted, in wich the yield function and plastic potential function are determined by flow stress and Lankford r-value respectively. The model was implemented in Abaqus/explicit by user subroutine VUMAT, and was then applied to simulaiting the calibration tests and the verification tests(shear and punch). Simulation result could exactly regenerate the experimental data, in both force response and local strain response. Meanwhile, the model was compared with three associated ones, and several suggestions were made about the selection and calibration of plasticity models.Subsequently, the fracuture behavior of the DP780 steel sheet was studied. A hybrid experimental/numerical inverse method was exploited on the shear test to determine the post-necking hardening curve firstly. Accordingly, the punch test was simulated, and the the accuracy of the hybrid method was validated. Subsequently, fracture tests in five different stress states were simulated and satisfactory simulation results were obtained. So based on this work, the evolution curves of the stress state and the equivalent plastic strain at the fracture-onset position were extracted from the simulaton of different tests. Using all the data, the MMC(Modified Mohr-Coulomb) frature criterion was calibrated with three different strategies. Characterization results by these three strategies were relatively close to each other, and of satisfactory accuracy.Lastly, the strain-rate effect was characterized. Uniaxial tension tests of the studied material were carried out under seven different strain-rates, and an obvious positive strain-rate dependence was observed. The Johnson-Cook model and the Khan-HuangLiang model were employed to characterize the hardening curves of the senven strain-rates, but it was found that the accuracy of the characteriaton was not acceptable. Therefore, the two models were modified, and their performance was greatly improved. Furthermore, the present study adopted a new data processing method, which useed the necking onset as a characteristic point to normalize the hardening curve. And it turned out that the normalized hardening curve of different strain-rates were almost identical to each other. Accordingly, the 2D-scaling model was proposed, the accuracy of which was later verified by a rate-jump test. Several other materials were investigated and it was found that the 2D-scaling model had a very good applicability for the DP steels while the accuracy of its modeling on the other materials were limited.