ABSTRACT
A cellular automaton consists of a discrete lattice of sites. Each site carries a discrete value chosen from a small set of possibilities. Cellular automata are thus, by construction, almost ideal for simulation on digital electronic computers. Detailed studies have demonstrated that many of the phenomena seen in actual fluid experiments can accurately be reproduced by this simple cellular automaton model. At a theoretical level, cellular automaton fluid models can be analyzed by much the same methods of statistical mechanics as have been used in trying to derive the Navier-Stokes equations for physical fluids from the microscopic dynamics of real molecules. The cellular automaton evolution thus acts like a pseudorandom number generator: even though a simple seed is given, the algorithm yields sequences whose simple origins cannot be discerned. Statistical descriptions of cellular automaton fluid models are close in form to explicit finite difference approximations to partial differential equations.
