Artificial neural network-based prediction of stacking fault energy in Fe-Cr-Mn-C-N steels

Abstract

This study develops an artificial neural network (ANN) model to systematically investigate the influence of alloying elements on the stacking fault energy (SFE.) in Fe-Cr-Mn-C-N steels. SFE is a key factor in determining these materials' mechanical properties and deformation mechanisms. The ANN model demonstrates excellent predictive accuracy, with an error of less than 4% and an R-2 value of 93%, significantly outperforming traditional empirical equations and thermodynamic models. Additionally, our analysis establishes a clear qualitative hierarchy among the alloying elements influencing SFE, with nitrogen exerting the strongest effect, followed by carbon, manganese, and chromium (N > C > Mn > Cr). These insights provide a deeper understanding of SFE in Fe-Cr-Mn-C-N steels and offer a strategic framework for optimizing alloy compositions, supporting the development of high-performance austenitic steels.