Data assimilation for phase-field simulations of the formation of eutectic alloy microstructures

The phase-field (PF) method can effectively predict the formation of microstructures of eutectic alloys. However, numerous simulation parameters must be determined correctly for each alloy system to reproduce the experimentally observed microstructures. In this study, we present a data assimilation method based on an ensemble Kalman filter to determine PF simulation parameters for the directional solidification of eutectic alloy by optimizing the conditions for data assimilation. Numerical twin experiments revealed that eutectic microstructures can be reproduced, although four PF simulation parameters remained unknown. We also investigated appropriate experimental observation conditions for estimating the simulation parameters and found that the sufficient frequency of observations can be determined from the solid–liquid interfacial velocity. Our results provide guidance for data assimilation combined with the PF simulations of eutectic alloys. Moreover, our study provides a deeper understanding of the formation mechanisms of various types of eutectic microstructures.