Adding ecology to particle capture models: numerical simulations of capture on a moving cylinder in crossflow.

Abstract

The particle capture efficiency, η, of systems that remove suspended particles from ambient flow (e.g. suspension feeding, abiotic pollination) has been studied using static collectors in steady flows. Particle deposition on collectors moving due to fluid flow remains largely unknown, despite its ecological relevance. We used numerical modeling to simulate particle deposition on a 2D circular cylinder subject to flow-induced oscillation in a cross flow. Using parameter values relevant to wind pollination and other natural biological systems, we examined the influence of the direction, amplitude and frequency of the oscillation, the Stokes number (Stk=0.01-5, characterizing particle behavior), as well as the Reynolds number (Re=662 and 3309, characterizing flow regime) in steady and unsteady flow, on η. The numerical model was validated with empirical results for parts of the parameter space. Particle capture occurred via "inertial impaction", "direct interception" and "leeward deposition", as well as via a new mechanism, "collector chasing" for moving collectors. The η of an oscillating cylinder varied significantly relative to a static cylinder, depending on the parameters used, and on the magnitude of a numerical error that caused loss of particles. This variance of η was due to a change in relative momentum between the particle and the moving collector, which depends on Re, Stk and the oscillation parameters. Collector oscillation transverse to oncoming flow direction strongly increased η, whereas collector motion parallel to flow had little effect on capture efficiency. The oscillation also changed leeward capture significantly in some cases. For most conditions, however, leeward deposition was small. Results suggest that collector motion could have significant influence on the particle capture efficiency of natural systems, which indicates the need to incorporate these ecologically more relevant findings into current models. Empirical studies, however, are still necessary to validate these results and provide reliable data.