Abstract:
This thesis studies the high-strength steel Docol 600DP and the ultra-high-strength steel Docol 1200M, focusing on developing and validating constitutive and failure models using a combination of experiments and finite element (FE) simulations. The constitutive model uses an eight-parameter high exponent yield surface with a mixed isotropic-kinematic hardening model to capture the anisotropic behavior and non-linear strain paths observed in the materials. For failure prediction, three phenomena are considered: ductile fracture, shear fracture, and instability with localized necking. Ductile and shear fractures are described by the Cockroft-Latham and Bressan-Williams models, respectively, which are chosen for their ease of calibration using simple mechanical tests. The instability phenomenon is modeled directly by the constitutive laws and the FE models themselves, and therefore does not require additional experimental calibration. A significant portion of the work involves detailing the various mechanical tests performed to calibrate and verify the models, including tensile, shear, plane strain, and Nakajima tests. The FE simulations, performed using LS-DYNA with meshes created in TrueGrid, are shown to agree well with experimental results, particularly after modifying the hardening curve and introducing a shear correction factor to account for effects from the element formulation. The thesis concludes that using phenomenological failure models with a well-calibrated FE model is a promising approach to predicting failure, especially for complex loading paths, and that the predictions from the models generally show good agreement with experimental data. The work highlights the growing importance of simulation-based design for industries like automotive, enabling the creation of lighter and safer products.
