A modified direct simulation Monte Carlo method for low-speed microflows

Numerical simulations on stationary and moving gasses (helium, argon and xe non) are carried out using direct simulation Monte Carlo (DSMC) methods. It is found that the statistical error of the results obtained by the DSMC me thod is quite large, and there exists a limit in reducing the statistical e rror through increasing the sample size. To reduce the statistical error, a modified DSMC method is presented for simulating the microflows with low v elocity in microelectromechanical systems. In the modified DSMC method, the random velocity of gas molecules as well as the stress and temperature of gas flow is split into two isolated parts. One part is determined through a macroscopic physical model and the other is obtained by the DSMC algorithm . Micro Couette flows for several types of monatomic gasses are simulated u sing both the original and present modified DSMC methods. The numerical res ults show that the results obtained by the present method have much smaller statistical error and higher accuracy than those by the original DSMC meth od. Moreover, the present method reduces significantly the computational ef fort.