Gas-surface interactions using accommodation coefficients for a dilute and a dense gas in a micro/nano-channel : heat flux predictions using combined molecular dynamics and Monte Carlo techniques

The influence of gas-surface interactions of a dilute gas confined between two parallel walls on the heat flux predictions is investigated using a combined Monte Carlo (MC) and molecular dynamics (MD) approach. The accommodation coefficients are computed from the temperature of incident and reflected molecules in molecular dynamics and used as effective coefficients in Maxwell-like boundary conditions in Monte Carlo simulations. Hydrophobic and hydrophilic wall interactions are studied, and the effect of the gas-surface interaction potential on the heat flux and other characteristic parameters like density and temperature is shown. The heat flux dependence on the accommodation coefficient is shown for different fluid-wall mass ratios. We find that the accommodation coefficient is increasing considerably when the mass ratio is decreased. An effective map of the heat flux depending on the accommodation coefficient is given and we show that MC heat flux predictions using Maxwell boundary conditions based on the accommodation coefficient give good results when compared to pure molecular dynamics heat predictions. The accommodation coefficients computed for a dilute gas for different gas-wall interaction parameters and mass ratios are transferred to compute the heat flux predictions for a dense gas. Comparison of the heat fluxes derived using explicit MD, MC with Maxwell-like boundary conditions based on the accommodation coefficients, and pure Maxwell boundary conditions are discussed. A map of the heat flux dependence on the accommodation coefficients for a dense gas, and the effective accommodation coefficients for different gas-wall interactions are given. In the end, this approach is applied to study the gas-surface interactions of argon and xenon molecules on a platinum surface. The derived accommodation coefficients are compared with values of experimental results