Surface microstructure predictions from atomistic rule set cellular automata

When a specific microstructure is required, if the preparation variables increase beyond a few, it is very difficult and expensive to determine the optimum conditions experimentally. Consequently there is considerable interest in predicting conditions via computer simulations. Since ultimately, microstructure depends on processes occurring at the atomistic level, to be fully transferable, it is desirable that such a model is atomistically-based. This should also allow us to include the role of all types of chemical and crystallographic defects explicitly. In this study, we begin by calculating the energetics associated with the way in which individual gas atoms interact with a specific metal surface. Both perfect and defective metal surfaces are considered. The energetics are translated into rule sets which form the basis of the cellular automata. The rule sets involve both temperature and gas atom flux as variables. The result is a model which can quickly, explicitly describe the evolution of 104 surface sites over 10^-6 seconds with very modest computing facilities. In the simulations, the formation and growth of domains which exhibit critical behavior are observed. That is, the rate of growth is not a well-behaved function of temperature or flux but exhibits a region in which the rate of growth suddenly falls to zero. Surface defects are also predicted and have a dramatic effect on growth rates