Neural network potential for Zr–Rh system by machine learning

Abstract

Zr–Rh metallic glass has enabled its many applications in vehicle parts, sports equipment and so on due to its outstanding performance in mechanical property, but the knowledge of the microstructure determining the superb mechanical property remains yet insufficient. Here, we develop a deep neural network potential of Zr–Rh system by using machine learning, which breaks the dilemma between the accuracy and efficiency in molecular dynamics simulations, and greatly improves the simulation scale in both space and time. The results show that the structural features obtained from the neural network method are in good agreement with the cases in ab initio molecular dynamics simulations. Furthermore, we build a large model of 5400 atoms to explore the influences of simulated size and cooling rate on the melt-quenching process of Zr77Rh23. Our study lays a foundation for exploring the complex structures in amorphous Zr77Rh23, which is of great significance for the design and practical application.

First-principles based deep neural network force field for molecular dynamics simulation of N–Ga–Al semiconductors

“…In addition, the RESMs of the NNPs in the literature are also summarized, which are in the ranges from 1.6 to 17 meV per atom for energy and from 0.01 to 0.25 eV Å −1 for force. 25,42–45 The RESMs in our work have relatively lower values compared with those in the literature, and indicate that our NNP is adequate for the MD simulations of the N–Ga–Al semiconductor systems.…”

Atomic structure of intermetallic compound Nb5Si3 by new cluster transformation analysis method

“…On the other hand, the atomic structure is an indispensable tool for the atomic simulations of physical properties which have no reliable measuring technique, such as solidliquid interfacial energy and diffusion coefficient [10][11][12]. With the help of novel simulation techniques, including classical molecular dynamics (MD), first principal calculations, and deep-learning MD, it is now possible to obtain the numerous coordination data of atoms with high accuracy and efficiency [13][14][15][16]. Subsequently, extracting and characterization of the local structure information at the atomic level from the numerous data have come to the forefront.…”