Experimental investigations on the preferential concentration and clustering phenomena of wall-bounded gas-solid flows

Abstract

The transport of solid particles by fluid flows are ubiquitous in many environmental, biomedical and industrial processes, yet a full understanding of the dynamics of these processes remain elusive. In this thesis, we present experimental investigations of two different phenomena that arises in particle-laden flows: the preferential concentration of particles in turbulent flows, and the clustering of particles in dense gas-solid flows. In the former case, we investigate inertial particles dispersed in a turbulent downward flow through a vertical channel. The working fluid is air laden with size-selected glass microspheres, having Stokes numbers St = O(10) and O(100) when based on the Kolmogorov and viscous time scales, respectively. Cases at friction Reynolds numbers 235 and 335 and solid volume fractions 3E-6 and 5E-5 are considered. Between the more dilute and denser cases, substantial differences are observed in all measured statistics e.g. the particle concentration profile, mean velocity profile and the velocity fluctuation levels; consistent with a scenario in which the increase in volume fraction from O(1E-6) to O(1E-5) triggers two-way and local four-way coupling effects. An analysis of the spatial distributions of particle positions and velocities in the higher volume fraction cases also reveals different behavior in the core and near-wall regions. In the channel core, dense clusters form that travel faster than the less concentrated particles; whereas in the near-wall region the particles arrange in highly elongated streaks that are associated with negative streamwise velocity fluctuations. In the denser regime, we present experimental observations on the velocities and spatial distribution of particles in a three-dimensional, gas-solid riser with particle volume fractions approaching 1%. The setup consists of a vertical square channel in which air flows upwards against falling 212 um glass spheres. We use a backlighting technique and high-speed imaging to quantify the spatial and temporally resolved particle concentration and velocity fields. By controlling the particle feed rate and the flow rate of the fluidizing air, volume fractions and bulk flow Reynolds number are adjusted independently. Results show that, in the present range of parameters, clustering of particles appear beyond a critical volume fraction regardless of fluidization velocities, influencing the mean and r.m.s. statistics and strongly modulating the two-point correlation statistics. Space-time autocorrelation analysis reveals the convection of structures in the velocity and concentration fields, and the fluctuations of velocities and concentrations are well-described by the classic gradient diffusion hypothesis. Particle-resolved measurements reveal that particles in the riser have a sub-Poissonian spatial distribution, and their streamwise velocity fluctuations are correlated in the streamwise direction. This indicates significant hydrodynamic interaction between the particles, especially in the direction of gravity.