Key points are not available for this paper at this time.
Since the BGK model is based on the assumption of constant collision frequency, and since this model has been found inadequate in describing some experimental data, the numerical study of a variable collision frequency model proposed earlier by Cercignani and Loyalka and Ferziger is described. Specifically, the Kramer’s problem for this model is solved, and it is found that the ’’velocity defect’’ in the Knudsen layer is quite sensitive to the velocity dependence of the collision frequency. In fact, for the hard sphere collision frequency, the present results agree reasonably well with the recent experimental data of Reynolds, Smolderen, and Wendt.
Sudarshan K. Loyalka (Mon,) studied this question.