MICROSCOPIC SIMULATION OF DILUTE GASES WITH ADJUSTABLE TRANSPORT-COEFFICIENTS

The Bird algorithm is a computationally efficient method for simulatin g dilute gas flows. However, due to the relatively large transport coe fficients at low densities, high Rayleigh or Reynolds numbers are diff icult to achieve by this technique. We present a modified version of t he Bird algorithm in which the relaxation processes are enhanced and t he transport coefficients reduced, while preserving the correct equili brium and nonequilibrium fluid properties. The present algorithm is fo und to be two to three orders of magnitude faster than molecular dynam ics for simulating complex hydrodynamical flows.