Closed Loop Identification in Diffusion-limited Aggregation Processes
Jack C. H. Chung
Maria Litt
Gary G. Leininger
Harvey Scher
The Standard Oil Company, Research and Development Center,
Information Sciences Laboratory, 4440 Warrensville Center Road,
Cleveland, OH 44128, USA
Abstract
Diffusion-limited aggregation (DLA) is a useful model for studying such common physical phenomena as dust clustering, unstable fluid flow, chemical species precipitation, and crystal growth. Simulating the DLA processes using electrostatic analogy is very computation-intensive. Elimination of closed loops in DLA images can significantly reduce the dimensionality of the problem and minimize the computational time required. In this paper, we describe both non-recursive and recursive techniques for the identification of closed loops in DLA processes. The recursive algorithm developed can identify closed loops without rescanning the complete image at each stage of the aggregation process. Computer simulations indicate that the recursive algorithm is several orders more efficient than the non-recursive one.
