Cluster models for random particle aggregates—Morphological statistics and collision distance

Abstract

In this paper, new models for particle–cluster and cluster–cluster-like aggregates are investigated through detailed Monte Carlo simulation in order to model aggregation of micro-sized particles in turbulent flows.These structures are built up by mechanisms which occur, for instance, in homogeneous isotropic turbulence flows. In contrast to other agglomeration models they are not based on Brownian motion and can be studied without a predefined motion field. The models are compared among each other in two and three dimensions with respect to several morphological properties, such as Feret diameter, radius of gyration, convexity, aspect ratio, fractal dimension, pair correlation and coordination number averaged over hundreds of thousands of aggregates of each size.It is shown that real particle aggregates generated in a turbulent regime can be modeled well using one of these approaches.An important part is devoted to the question how these aggregates can be approximated by spherical particles. For this, the collision distance to other particles is used, measured and approximated for both aggregate models. It can be useful for calculations related to agglomeration kernels in the framework of Smoluchowski’s coagulation equation.