Mechanical behavior of multilayer graphene reinforced epoxy nano-composites via a hierarchical multi-scale technique

In this paper, a multi-scale technique is presented to study the mechanical behavior of nano-composites by linking the atomistic information from lower scale to the continuum model in upper scale. Molecular dynamics simulations are employed to compute the material properties of graphene/epoxy nanocomposites using the COMPASS interatomic force field. In order to obtain the atomistic stress surfaces used to evaluate the mechanical properties of material in upper scale, the biaxial loading is applied to different representative volume elements. On the continuum level, the hyperelastic strain energy functions are utilized to calculate the material parameters using the hyperelastic functions from atomistic data. The stress and elasticity tensors are obtained by computing the first and second order derivatives of hyperelastic functions with reference to the components of the right Cauchy-Green deformation tensor. The stress-strain surfaces of hyperelastic functions in lower scale are used to calculate the properties of nano-composite material in upper scale. The efficiency and applicability of the proposed technique is presented through various numerical examples. It is shown that the proposed multi-scale technique is able to solve large problems within acceptable computational time, which is not possible using conventional molecular dynamics approaches. Crown Copyright (c) 2021 Published by Elsevier Ltd. This is an open access article under the CC BY-NC-ND license (http://creativecommons.org/licenses/by-nc-nd/4.0/)

Automated summary

This paper presents a multi-scale technique to study the mechanical behavior of nano-composites by linking atomistic information from lower scale to continuum model in upper scale. Molecular dynamics simulations are used to compute material properties, while hyperelastic strain energy functions are utilized on the continuum level to calculate material parameters and stress and elasticity tensors. The proposed technique is shown to efficiently solve large problems within acceptable computational time, demonstrating its superiority over conventional molecular dynamics approaches.