Stress Triaxiality in Damage Models

Abstract

High strain rate of loading combined with multiaxial effects are important factors, which we need to consider in understanding the response of structures subjected to hard-body impact, blast, and detonation. Better design and load mitigation schemes against these factors can be developed based on the simulations that should be more reliable in predictive with such an approach. An important step toward this is a systematic validation of the simulations and various factors explained above and incorporation of these factors into the modeling schemes. Here, we consider the problem of modeling the impact dynamic response of metallic components undergoing hard-body contact and nonlinear elasto-plastic deformation. We consider the influence of triaxiality in the damage models with an example of the Johnson–Cook failure model, which is extensively used in industry for simulating impact, blast response, and crash. The material parameters involved in both the constitutive model and the failure model are estimated from one-dimensional tensile tests with coupons. Two problems arise, one is the material behavior characterized by the yield and plastic flow/hardening and the other is the two- and three-dimensional equivalent from the one-dimensional tests and parametric representation such that an accurate response can be simulated predictably and not merely by adjusting the material parameters in every simulation by knowing an actual test result in advance. The Johnson–Cook failure model has a five-parameter function to describe the plastic strain at failure. It includes the damage state in terms of the stress triaxiality, strain rate, and temperature effects. In the present work, the material constants and damage parameters are experimentally determined using uniaxial tensile tests. Finite-element model of ASTM standard tensile test coupon is modeled using Johnson–Cook material in LS-DYNA and are validated with experimental results for three different strain rates. ASTM standard flexural model is further simulated and validated with experimental results. Varying aspect ratios in flexural models are simulated and are experimentally compared. We observe that material parameters extracted under uniaxial tensile test conditions are not sufficient to model complex phenomenon such as blast, impact, etc., as these models cannot accurately predict damages and failure occurring in mixed mode loading conditions.