The atom sampling method for mesoscale molecular dynamics and its application to graphene assemblies

Several coarse-grained (CG) molecular dynamics models have been developed to predict the mechanical behaviors of mesoscale graphene structures because of the enormous promise of graphene in various carbon-based nanostructures and the spatiotemporal limitations of experimental testing and atomistic modeling. Although the CG models reduced the number of degrees of freedom and speeded up simulations, the increased mass of the beads caused a significant mismatch in the system's total kinetic and potential energy. In this paper, we propose an atom sampling method for mesoscale molecular dynamics simulations, which could effectively mitigate the consequences arising from the increased mass of the beads in the CG models while still maintaining the computational efficiency of CG models. Then, the Tersoff sampling model of graphene using the atom sampling method is developed to coarsen graphene in planar directions. The mechanical responses of monolayer graphene and multilayer graphene (MLG) assemblies are simulated by the Tersoff sampling model, and simulation results obtained with the all-atom (AA) Tersoff model and CG models are compared. The results show that the Tersoff sampling model can reduce the misfit of kinetic and potential energy due to the fact that temperature changes more accurately reproduce the mechanical behaviors of graphene in the elastic and fracture zones compared to the CG models. At the same time, it further validates the atom sampling method for mesoscale simulations. The atom sampling method can also be easily applied to isotropic fluids and ordered crystal structures (including established coarse-grained models of these substances).