Direct Simulation Monte Carlo Calculation: Strategies for Using Complex Initial Conditions

Modeling of phenomena is increasingly being used to obtain an understanding of important physical events as well as to predict properties that can be directly tied to experimental data. For systems with relatively low densities of particles, the Direct Simulation Monte Carlo (DSMC) method is well suited for modeling gases with non-equilibrium distributions, coupled gasdynamic and reaction effects, emission and absorption of radiation. On the other hand, if the density of particles is large such as in dense gases or condensed matter, the DSMC method is not appropriate and techniques such as molecular dynamics (MD) simulations are employed. There are phenomena such as laser ablation, however, in which the system evolves from a condensed state appropriate to be studied with MD to an expanding rarified gas appropriate to be studied with DSMC. The work presented here discusses the means of transferring information from a MD simulation of laser ablation to a DSMC simulation of the plume expansion. The presence of clusters in the MD output poses the main computational challenge. When the laser fluence is above the ablation threshold, the cluster size distribution is very broad (up to 10,'000's of particles per cluster) but there are relatively few of each cluster size. We have developed a method for statistical processing of the MD results and have represented the cluster size as a random variable. Various aspects of the coupling between the MD and DSMC models are discussed and several examples are presented.