[SY-I1] Symposium I-1

[SY-I1] Estimation of Grain Boundary Anisotropy using Multi-phase-field Model based on the Ensemble Kalman Filter

^○Akinori Yamanaka, Yuri Maeda, Kengo Sasaki (Tokyo University of Agriculture and Technology, Japan)

Grain growth is one of the most fundamental microstructure evolutions in metallic materials. Because the size of crystal grains strongly affects mechanical properties of materials, it is essential to predict the grain growth behavior during polycrystalline grain growth and recrystallization. Recently, the multi-phase-field model has been widely used for simulating the grain growth behavior and predicting average grain size and grain size distribution. In order to simulate a realistic grain growth behavior by the multi-phase-field simulation, we need to use accurate data of grain boundary properties, for example, anisotropic grain boundary mobility and grain boundary energy. However, it is difficult to identify the anisotropic grain boundary properties from experimental results by trial-and-error. This study proposes a data assimilation methodology for estimating grain boundary properties by incorporating experimental results into the multi-phase-field simulation of polycrystalline grain growth. We construct the data assimilation methodology using the Ensemble Kalman filter. We demonstrate that the anisotropic grain boundary energy can be successfully estimated from three-dimensional distributions of polycrystalline microstructure by the data assimilation method proposed in this study.