Measurement of material deformation behavior under high-temperature conditions and parameter estimation of material models by Bayesian data assimilation

During plastic forming of automotive parts, a plastic work caused by the forming increases temperature of the material. The increase of temperature changes mechanical properties and makes it difficult to predict the deformation behavior of material. Therefore, it is crucial to identify material model parameters that depend on temperature. The parameters of thermo-elastoplastic material models have been identified based on material test results obtained under various temperature conditions. However, many high-temperature tests are required to identify accurate model parameters. In this study, we developed a Bayesian data assimilation method to estimate the material model parameters for high-temperature condition using uniaxial tensile test results. The full-field strain data obtained by the uniaxial tensile test and the digital image correlation method (DICM) are used as observation data for the data assimilation. It was demonstrated that the optimally estimated parameters obtained by the proposed data assimilation allow a more accurate reproduction of the strain distribution than the initially estimated parameters based on experimental data.