Abstract
There are two methods for obtaining information on the transfer coefficients of gases and liquids, namely experimental measurement and molecular modeling. The kinetic theory has explicit formulas for transport coefficients only for a dilute gas in the simplest cases. The method of molecular dynamics has to be used to calculate transport coefficients in dense gases and fluid. This method, however, requires enormous computing resources. Creating an alternative less costly method of molecular modeling is very important. In this paper we propose a stochastic method of molecular modeling of transport coefficients in dense gases. The method is based on the idea of splitting the movement of gas molecules in the process of free shift and impact. Free shift of molecules is carried out depending on their speeds at a given time, and collisions run randomly. To do this it is first necessary to determine whether the molecule can collide or not. If a collision is realized, a random molecule with which it collides is selected from the remaining ones. As a result of simulation data on the coordinates and impulses of all the molecules of the system at successive times is obtained. Then, using methods of non-equilibrium statistical mechanics, transport coefficients are calculated based on this data. They are obtained by averaging the dynamical variables ensemble of independent phase trajectories. The efficiency of the algorithm is demonstrated by modeling of the argon viscosity coefficient in a wide range of pressures (densities). It is shown that the accuracy of the method increases when the number of molecules and the number f phase trajectories used for averaging in the simulated system increases.
Keywords: transport processes, dense gas, stochastic simulation, molecular modeling, viscosity, statistical simulation.