Fast and Stable Quasi-Static Bending Simulations in LS-DYNA: Identification of Optimal Finite Element Model Parameters 2016-01-1392

The quality of material model input files for finite element analysis (FEA) is a fundamental factor governing the fidelity and accuracy of simulations at a sub-system or a vehicle level, dictating an investment of due diligence in developing and validating the material models. Several material models conventionally employed for FEA typically allow accounting for only uniaxial tensile behavior of the material; however, the models may be required to predict component-level response in a complex loading scenario. Therefore in developing LSDYNA material input files for such models, it becomes critical to validate their performance in alternative loading scenarios. For out-ofplane loading, typically a three or four-point bending load-case is used for validation. Simulating three point bending (TPB), particularly in the quasi-static regime, requires detailed representation of the moving pin impacting the specimen, and sliding of the specimen on the stationary pins. This in turn requires optimization of contact parameters, thereby introducing an additional factor that adds uncertainty to validation of the material-model in bending scenarios.

This paper identifies the key parameters that underlie the robustness of quasi-static bending simulation and identifies optimal settings for these parameters to ensure fast, accurate, and stable bending simulations, which can provide an unbiased evaluation of the performance of the material model in out-of-plane loading scenarios. Using the conventional TPB fixture mounted on a universal testing machine (UTM), bending tests were carried out on a thermoplastic material. In bending simulations the material is represented using the piecewise linear plasticity model. The optimal settings for numerical and contact parameters, identified using a detailed simulation design of experiments (DoE), are also validated for an alternate thermoplastic material.