Applying a new constitutive model to analyse the springback behaviour of titanium in bending and roll forming

Abstract This paper is an extension to previous work that dealt with the development of a strain path dependent constitutive model to describe the inelastic behaviour of Ti-6Al-4V at room temperature based on the homogenous yield function combined with the anisotropic hardening characteristics, the so-called HAH model. The present work is to apply and verify the accuracy of the proposed model in the finite element (FE) analysis of springback in bending dominated forming processes such as the V-die bending and the roll forming process. In addition, the model is applied to develop a greater insight into the nature of springback in the roll forming process where springback is generally lower compared to that found for simple bending. For this the hardening characteristics of Ti-6Al-4V were identified applying an inverse analysis approach in Abaqus Standard and the model used to describe the evolution of the anisotropic yield surface during non-proportional strain path deformation; this included the cyclic pure bending and cyclic tension–compression tests to generate experimental target curves. The constitutive model parameters were optimised to capture the cyclic behaviour of Ti-6Al-4V in both the pure bending and tension—compression and incorporated into the numerical models of a V-die bending test and a roll forming procedure to analyse springback. The model achieved good agreement with experimental results and reproduced the lower springback observed in roll forming compared to simple bending. In contrast to this a conventional isotropic hardening model significantly overestimated springback for the roll forming process. This suggests that the lower level of springback observed in roll forming compared to V-die bending may be due to kinematic hardening effects. In addition, a lower level of accumulated effective stress and the presence of redundant shear strain was numerically observed in the bending regions of the roll formed section which also may have contributed to the reduction of springback. The results of this study suggest that for the accurate numerical prediction of springback in roll forming kinematic hardening effects need to be accounted for. For this the study presents an effective numerical approach and prove its applicability by numerical roll forming and V-die bending studies performed on high strength Ti-6Al-4V at room temperature and verified with experimental results.