Computational study of thermal dependence of accommodation coefficients in a nano-channel and the prediction of velocity profiles

A molecular dynamics (MD) study of the flow of rarefied gas in a nano-channel with infinite parallel wall is reported. Nano-channels are considered to be more appropriate option in the cooling of microelectronic devices in the near future. The momentum and the energy transfer between the gas and the surface is a significant parameter which account for many important properties. The study aims at the calculation of energy and momentum accommodation coefficients for different gas-wall temperature combinations, and also its dependence on Knudsen number and gas-wall temperature difference. The different gas-wall temperature combinations are achieved by controlling the wall temperature as well as the gas temperature. The incident and reflected velocity distributions are obtained and the reflected distributions in different directions are predicted using Maxwell-type boundary conditions. It is shown that the partial energy accommodation coefficients are sufficient to predict the velocity distribution in tangential and normal directions. The statistical deviation of the predicted distribution from the distributions obtained from MD simulation is found to be very low. The accommodation coefficients are found to vary linearly with the natural logarithm of gas-wall temperature difference. The tangential components of momentum showed a strong correlation with the incident speed of the particles. (C) 2012 Elsevier Ltd. All rights reserved.