微尺度气体流动

MICRO-SCALE GAS FLOWS

摘 要 了解微尺度气体流动特点是微机电系统设计和优化的基础.有关的研究可以上溯到20世纪初Knudsen的平面槽道流动质量流量的测量和Millikan的小球阻力系数的测量,实验结果揭示了稀薄气体效应即尺度效应对气体运动的重要影响.由于流动特征长度很小,微尺度气流经常处于滑流区甚至过渡领域,流动的相似参数为 Knudsen数和 Mach数.因此可以考虑利用相似准则,通过增大几何尺寸、减小压力的途径,解决微机电系统实验观测遇到的困难.为解决直接模拟MonteCarlo方法分析微机电系统中低速稀薄气流遇到的统计涨落困难,我们提出了信息保存法（IP）,该方法能够有效克服统计散布,并己成功用于多种微尺度气流.

Understanding the features of micro-scale gas flows is essential to designing and optimizing micro-electro-mechanical systems (MEMS). Research on micro-scale gas flows originated from the famous experiments carried out by Knudsen and Millikan, respectively, i.e. the measurements of mass flux through a plane channel and drag coefficient of a small sphere, at the beginning of the 20th century. The experimental results revealed the significant effects of scale influence (rarefaction) on the gas flow characteristics. In the slip and transition regimes, the similar parameters of gas flows are the Knudsen number and Mach number, and therefore, there is a possibility to utilize the principle of similarity, i.e. increasing the geometric scale while decreasing the pressure, to solve the difficulty associated with the measurement in MEMS experiments. An information preservation (IP) technique was proposed by us to address the serious statistical scatter encountered in Monte Carlo direct simulation of low-speed gas flows. The IP technique may reduce the statistical scatter efficiently, and has been successfully applied to many micro-scale gas flows.