Monte Carlo Simulations On Micro- And Nanoscale Gas Flow And Heat Transfer

The investigation of micro- and nanoscale gas flow and heat transfer is the key for the further developments of MEMS and nano science and technology. The Knudsen number becomes large enough due to the decreased characteristic length. The classical simulation methods are no longer suitable, and the flow must be described from a molecular point of view. The direct simulation Monte Carlo method (DSMC) is the most suitable and reliable numerical method for high-Knudsen- number flows. Based on theoretical analyses, the present dissertation studied the micro- and nanoscale gas flows and heat transfer by using the DSMC method, and then discussed their mechanism.In order to make the DSMC method applicable for the simulations of the micro and nanoscale gas flow and heat transfer, a new method was developed to deal with the pressure boundary implements. The new method avoided the computational divergence in the previous methods, and a higher convergence speed was obtained. The inlet and outlet boundary implements for a non-straight channel were detailedly presented. With these new methods, gas flows in a micro channel with a 90 degree corner were simulated, the results agreed with the existed experimental data qualitatively and no vortex flow appeared at the corner.The similarity was theoretically analyzed between microscale gas flows and normal-scale rarefied gas flows. The similarity qualifications were summarized as: similar geometries and boundary conditions, equal homonymous dimensionless criterions, and thermal perfect gas. This similarity rule was also numerically verified using the DSMC simulations for different scale flows.The disadvantages of the existing methods for dense gas flow simulation, in which the thermal perfect gas assumption breaks down, were analyzed and the mechanisms were presented. To correctly simulate the van der Waals gas flow, an Enskog-equation-based new algorithm, generalized Enskog Monte Carlo method (GEMC), was proposed. In GEMC, a generalized collision model was introduced to embody the effect of intermolecular attractive potentials on the collision cross section. The intermolecular collision rate was enhanced by considering the dense gas effect. A new internal energy transition model affected by the intermolecular attractive potential was also introduced into the new algorithm. By comparing the transport coefficients with the experimental data, the new algorithm was perfectly verified. The dense gas flows in micro and nanoscale channels were studied using the GEMC method, and the results showed that the van der Waals effect decreased the skin friction coefficient.To implement the particle-only coupling simulations of multi-flow-regime flows, the present paper developed a new method, relax time Monte Carlo method (RTMC), based on the BGK equation. The RTMC method is more suitable for near continuum flows than DSMC.Gas flows and heat transfer in micro and nanoscale gas slider bearings and micronozzles were simulated and analyzed using DSMC. The optimizations were suggested based on the numerical results and the design requirements.