Application of machine learning in the study of atomic rearrangement events in amorphous alloys
Abstract
<p indent="0mm">Amorphous alloys, as a novel class of metastable metallic materials, are characterized by high strength and excellent wear resistance. However, their brittleness at room temperature limits their practical applications. Deformation in amorphous alloys typically occurs through shear bands, accompanied by extensive atomic rearrangements. Overcoming the brittleness of amorphous alloys at room temperature hinges on the accurate identification of these atomic rearrangement events, which remains a significant challenge. In recent years, machine learning has shown great promise in this area. This review highlights the progress made in applying machine learning algorithms, such as artificial neural network (ANN), support vector machines (SVM), gradient boosting decision trees (GBDT), and extreme gradient boosting (XGBoost), to the study of atomic rearrangement events in amorphous alloys. It explores their role in uncovering the underlying micro-mechanisms and anticipates future developments in the application of machine learning to amorphous alloy deformation studies. By integrating machine learning with traditional materials science approaches, new insights are offered for the optimized design and performance prediction of amorphous alloys.</p>
