Genetic Algorithm Optimization of Langevin Thermostat and Thermal Properties of Graphene-Aluminum Nanocomposites: a Molecular Dynamics

Abstract

The thermal properties of a laminated structure of graphene-coated aluminum composite nanomaterial were investigated through non-equilibrium molecular dynamics (NEMD) simulations to address the problem of temperature deviation in the thermostat volume applied. This paper presents a new insight into the best values of timestep and Langevin thermostat damping parameters for each atom in the nanomaterial with different size configurations using the genetic algorithm (GA) method by considering the timestep and thermostat damping parameters for each atom type, as well as the thickness of the nanomaterial, the thermostat, buffer, and heat flow lengths. The initial population results indicate that the thermostat temperature deviation increases with higher thermostat damping coefficients and timestep. However, the deviation decreases significantly with increased heat flow and thermostat lengths. Variations in buffer length and aluminum thickness do not have a significant effect on temperature. The application of a GA for optimization leads to a decrease in thermostat temperature deviation. The optimized parameters resulted in better thermostat temperature deviations when analyzing the temperature, aluminum thickness, and both buffer and thermostat lengths. Additionally, the thermal conductivity of aluminum-graphene nanomaterial decreases with increasing temperature, buffer length, and aluminum thickness, but increases by up to 9.85% with increasing thermostat length.