An Atomic Level Investigation of Grain-Size-Dependent Thermal Conductivity of Polycrystalline Argon by Molecular Dynamics

Abstract

The thermal conductivity of polycrystalline solid argon with different grain sizes is investigated by molecular dynamics simulation. The three-dimensional Voronoi tessellation method is employed to generate periodic polycrystalline structures consisting of randomly shaped grains with controlled size. The Green–Kubo method is then used to calculate the thermal conductivity according to the fluctuation–dissipation theorem. The effect of the random configurations of the grains with the same average grain size on the thermal conductivity is examined and is found to be less than 15 %. Configurations of polycrystalline argon with an average grain size from about 2 nm to 15 nm are then generated to obtain the relationship between the thermal conductivity and grain size. There exists a critical grain size at a temperature beyond which the thermal conductivity increases with average grain size, and below which the thermal conductivity decrease with the grain size. The thermal resistance of the grain boundary is found to be large when the grain size is comparable with the phonon mean free path (MFP), but smaller when the grain size is near the lattice constant of argon or larger than the phonon MFP. Both the thermal conductivity and the grain boundary resistance of polycrystal argon depend on temperature.