Interplay between magnetism and short-range order in bcc Fe-V alloys

The aim of this research is to investigate short-range chemical order (SRO) in body-centered cubic Fe-V alloys. To accomplish this, we will employ a machine learning approach based on a force field generated by a trained neural network (NN), using the DeePMD software package. We will compute the Warren-Cowley parameters for the first two coordination shells and demonstrate the presence of a significant repulsive force between vanadium atom pairs in close proximity. We will also explore the relationship between magnetic interactions of nearest neighbors and SRO. Results suggest a strong correlation between these two factors, which arises from increased frustration as vanadium concentration rises. This frustration arises from competing Fe-V and V-V interactions, with no tendency for vanadium atoms to form clusters, which was confirmed by the analysis of data using unsupervised learning techniques. These findings can be utilized to develop vanadium steels with improved properties.