An Identification Strategy for Lode Parameter-Dependent Constitutive Models Under Complex Stress States and Large Deformation

Accurate identification of constitutive model parameters is critical for reliable simulation of metal plastic forming processes involving complex stress states and large deformations. This study presents a parameter identification method tailored for such conditions, incorporating the Lode parameter to account for loading path dependence. Five large-deformation experiments under representative stress states are conducted, and data are corrected and optimized to enhance identification accuracy. A stepwise identification strategy is adopted, where the hardening parameters and Lode-dependence parameters are determined separately. The experimental data used in both stages capture the material's authentic large deformation behavior. By integrating this approach with an inverse optimization technique, the constitutive parameters of 20CrMnTiH steel are systematically identified. The method is validated under various stress paths, demonstrating its robustness and effectiveness. The identified model is further applied to a T-shaped cylindrical compression test, where numerical simulations show excellent agreement with experimental measurements, confirming the method's strong applicability and predictive capability in complex metal forming processes.