Microscale gas flows are the subject of increasingly active research due to the rapid advances in micro-electro-mechanical systems (MEMS). The rarefied phenomena in MEMS gas flows make molecular-based simulations desirable. However, it is necessary to reduce the statistical scatter in particle methods for microscale gas flows. In this paper, the development of an information preservation (IP) method is described. The IP method reduces the statistical scatter by preserving macroscopic information of the flow in the particles and the computational cells simulated in the direct simulation Monte Carlo (DSMC) method. The preserved macroscopic information of particles is updated during collisions and is then modified to include the pressure field effects excluded in the collisions. An additional energy transfer model is proposed to describe the energy flux across an interface, and a collision model is used to redistribute the information after both particle-particle and particle-surface collisions. To validate the IP method, four different flows are simulated and the solutions are compared against DSMC results. The results from the IP method generally agree very well with the DSMC results for steady flows and low frequency unsteady flows ranging from the near-continuum regime to the free-molecular regime. © 2002 Elsevier Science (USA).
Mendeley saves you time finding and organizing research
Choose a citation style from the tabs below