Machine learning–assisted ECNLE theory for predicting dynamical properties of FeCoCrMoCBTm metallic glass

In this work, we present an integrated modeling framework that combines machine learning (ML) and the Elastically Collective Nonlinear Langevin Equation (ECNLE) theory to investigate the glass transition dynamics of metallic glass systems. We focus on the Fe42Co6Cr15Mo14CxB21-xTm2 alloy family. The glass transition temperatures Tg of this alloy family are predicted using machine learning models trained on experimental datasets. These ML-predicted-Tg values are then used as input to the ECNLE theory to compute the temperature dependence of structural relaxation times, dynamic fragility, and diffusion coefficients. Our predictions of Tg and dynamic fragility quantitatively agree with the experimental data. This shows that the combined ML-ECNLE approach provides a practical and scalable tool for characterizing the relaxation behavior and diffusion dynamics metallic glasses when experimental data remain limited.