A SENSITIVE TEST FOR ACCURACY IN EVALUATION OF MOLECULAR COLLISION NUMBER IN THE DIRECT-SIMULATION MONTE-CARLO METHOD

A sensitive test is proposed for the accuracy in the evaluation of the molecular collision number in the direct-simulation Monte Carlo method: It is shown that the translational relaxation in a spatially homogeneous box, which is divided into a number of cells where the simulation of molecular collisions is carried out, is a simple but sensitive test problem for the evaluation of the lower limit of mean molecular number per cell, 〈ΔN〉 min, and the computation time, as well as for the verification of the accuracy in the evaluation of the collision number. It is confirmed from the test that 〈ΔN〉min is about 1 for the null-collision (NC) technique but about 30 for the time-counter (TC) technique, and the computation time for the NC technique is about 10%-40% larger than that for the TC technique in the simulation with the same molecular number. © 1990 American Institute of Physics.