Modifications of RHT material model for improved numerical simulation of dynamic response of concrete

Sophisticated numerical models are increasingly used to analyze complex physical processes such as concrete structures subjected to high-impulsive loads. Among other influencing factors for a realistic and reliable analysis, it is essential that the material models are capable of describing the material behaviour at the pertinent scale level in a realistic manner. One of the widely used concrete material models in impact and penetration analysis, the RHT model, covers essentially all macro features of concrete-like materials under high strain rate loading. However, the model was found to exhibit undesirable performance under certain loading conditions and some of the modeling issues have been discussed within a recent review paper by the authors. The present paper provides a more in-depth evaluation of the RHT model and proposes modifications to the model formulation to enhance the performance of the model as implemented in the hydrocode AUTODYN. The modifications include Lode-angle dependency of the residual strength surface, tensile softening law and the dynamic tensile strength function. The improvement of the performance of the modified RHT model is demonstrated using numerical sample tests, and further verified via simulations of two series of physical experiments of concrete penetration/perforation by steel projectiles. The results demonstrate an overall improvement of the simulation with the modified RHT model. In particular, the depth of penetration, projectile exit velocity and the crater size are predicted more favourably as compared to the test data. It is also shown that the modeling of the concrete tensile behaviour can affect sensibly the predicted perforation response (e.g., the projectile exit velocity), as is generally expected when the impact velocity exceeds the ballistic limit.