Short-range order and its impacts on the BCC MoNbTaW multi-principal element alloy by the machine-learning potential

We utilize a machine-learning force field, trained by a neural network (NN) with bispectrum coefficients as descriptors, to investigate the chemical short-range order (SRO) influences on the BCC MoNbTaW alloy strengthening mechanism. The NN interatomic potential offers a transferable force field that exhibits accuracy comparable to density functional theory. This innovative NN potential is employed to examine the SRO effects on various aspects such as elasticity, vibrational modes, plasticity, and strength in the MoNbTaW multi-principal element alloy (MPEA). The findings reveal a significant attraction between Mo-Ta pairs, resulting in the formation of locally ordered B2 clusters. These clusters can be adjustedvia temperature and enhanced by Nb content. The presence of SRO leads to an increase in high-frequency phonon modes and introduces additional lattice friction to dislocation motion. This approach facilitates efficient compositional screening, paving the way for computational-guided materials design of novel MPEAs with enhanced performance. Furthermore, it opens up avenues for tuning the mechanical properties through optimization of the processing parameters.

Automated summary

We use a machine-learning force field and neural network to investigate the influence of chemical short-range order on the strengthening mechanism of the MoNbTaW alloy. The findings reveal the formation of locally ordered B2 clusters due to a significant attraction between Mo-Ta pairs. The presence of short-range order increases high-frequency phonon modes and introduces additional lattice friction to dislocation motion.