Data for: Accurate interatomic force field for molecular dynamics simulation by hybridizing classical and machine learning potentials

Published: 28 August 2018| Version 1 | DOI: 10.17632/ycd9nstg6p.1
Hongtao WANG, Peng Wang, Yecheng Shao, Wei Yang


The source code and examples. The usage is provided in the Readme.txt file. We copy it below: Usage: The multisclae pair (pair_multiscale.h, pair_multiscale.cpp) has been implemented in LAMMPS. Compile: cp pair_multiscale.h pair_multiscale.cpp ~/lammps-31Mar17/src make ubuntu Usage: Example: pair_style multiscale pair_coeff * * Al.eam.alloy Al.agni Al Al Al 10 100 5 6 1 1 2 3 4 5 6 7 8 9 10 10 coefficients: 1: filename and path for EAM-FS potential 2: filename and path for AGNI potential 3-5: element name 6: Nevery, the CS parameter is computed every Nevery timesteps 7: Nfreq, the CS parameters are averaged every Nfreq timesteps based on the calculated CS parameter in the preceding portion of the simulation every Nevery timesteps. 8: Threshold value, the atoms in simulation are divided into two sub-regions based on the comparison between averaged CS parameters and the threshold value option-coeff: 9: The thickness of the transition region 10: Whether to dump the atomic sub-regions to atomic types. For more questions, please feel free to contect



Machine Learning, Molecular Dynamics, Multiscale Analysis