MD data for the study of molecular dynamics inferred transfer learning models for finite‐strain hyperelasticity of monoclinic crystals: Sobolev training and validations against physical constraints

Published: 27 April 2022| Version 1 | DOI: 10.17632/tgyg9xvck5.1
Puhan Zhao


The set contains data from 15 iaothermal molecular dynamics simulations of uniaxial compression/tension, biaxial compression, and pure shear. This data set has been used as training data for a machine learning model. Details can be found in the literature: 10.1002/nme.6992. Note: the potential energy in the dataset is the total potential energy of the system. Number of molecules for each system can be found in the literature.


Steps to reproduce

The MD system contains beta-HMX crystal in the P21/n space group setting. Details of the simulations can be found in the literature: 10.1002/nme.6992.


University of Missouri Columbia, Columbia University


Molecular Dynamics Study, Data-Driven Learning