Integrated Phase Field and Machine Learning Study of Microstructure Evolution during Interface-Controlled Spinodal Decomposition

Abstract

This study leverages artificial intelligence (AI) to advance materials science, focusing on microstructural evolution in binary alloys during spinodal decomposition. Following the formulation of Zhu et al., we explore the microstructure evolution during interface-controlled spinodal decomposition. A comprehensive dataset captures the dynamic microstructural changes, highlighting the model's efficiency in analyzing complex data. The innovative use of an Autoencoder- ConvLSTM model enables precise, low-error microstructural transformation predictions, demonstrating AI’s potential in materials science research. This work provides a deeper understanding of material behaviors and offers new research directions.