Kinetic theory based solutions for particle clustering in turbulent flows

Abstract

Kinetic theory provides an elegant framework for studying dispersed particles in turbulent flows. Here the application of such probability density function (PDF)-based descriptions is considered in the context of particle clustering. The approach provides a continuum representation for the particle phase in which momentum conservation identifies two fundamental contributions to the particle mass flux. These take the form of an additional body force, which emerges from inhomogeneities in the sampling of turbulence by particles, and a component of the particle phase stress tensor associated with turbophoresis. Remarkably, these contributions are frequently overlooked in the specification of mean-field models for dispersed particle flows. To assess the relative importance of these mass flux contributions a kinematic simulation study has been performed, making use of a specially constructed inhomogeneous flow field designed to mimic the dynamics of particle pair behaviour. Whilst the turbophoretic contribution always acts to increase the clustering of particles, both the direction in which the additional body force acts and the relative importance of the two mass flux contributions are found to vary with particle inertia and turbulence intensity. In some regimes the body force is also the dominant mechanism responsible for particle clustering, demonstrating its importance within model formulations. It is further highlighted that the evolution of the radial distribution function which describes particle clustering is represented as a balance between convection and diffusion, and only inclusion of the identified mass flux contributions within this balance enables correct prediction of the particle pair concentration profile observed in simulations.