Single inertial particle statistics in turbulent flows from Lagrangian velocity models

Abstract

Statistical descriptions of inertial particle motion are complicated by the random and multiscale structure of the turbulent velocity field. In this paper, we develop a stochastic model for inertial particles on the basis of ``filtered'' fluid tracer particles. Introducing an effective Stokes number allows us to capture crucial features of inertial particle motion such as the sharp decrease of acceleration variance for increasing Stokes numbers (acceleration trajectories are color-coded by maximum tracer acceleration). The proposed modeling approach could yield further insights on preferential concentration and a better experimental characterization of particle-laden flows.