Based on the cluster multiple labeling technique,a novel cluster detection algorithm is presented as an analysis subroutine in two-and three-dimensional molecular dynamic simulations of ejecta that take place as a planar shock wave encounters a free metal surface.The algorithm is described,tested,and used to detect cluster distribution of ejecta from copper and aluminum under a shock loading.The information obtained about the size,distribution,evolution of the cluster is helpful in the understanding of ejection.

A scalable parallel algorithm especially for large-scale three dimensional simulations with seriously non-uniform molecular distributions is presented. In particular, based on cell-block data structures, this algorithm uses space filling curve to convert three-dimensional domain decomposition for load distribution across processors into one-dimensional load balancing problems for which run-time measurement-based multilevel averaging weights method can be applied successfully. Parallel numerical experiments for simulations of three-dimensional metal-eject physical models with 210 million particles using 500 processors of one massively parallel processors have shown that this algorithm has achieved speedup about 420.