(103c) Numerical Study on Clustering of Polydisperse Group-A Particles in Riser Flow

Particle clustering plays a crucial role in the performance of circulating fluidized bed (CFB) risers. Over the past few decades, numerous studies have been conducted to characterize clusters and their effect on mixing, with the majority of them focusing on monodisperse particles. In this work, numerical simulations of riser flow made up of a continuous distribution of Geldart Group-A particles are conducted. Particle diameters are sampled from a lognormal size distribution, informed from dynamic image analysis of zeolite particles, spanning 21 to 140 um with a mean of 58 um. Highly-resolved simulations of fully-developed particle-laden pipe flow are conducted using the open-source multiphase flow solver NGA2. The volume-filtered Eulerian-Lagrangian equations are coupled with an immersed boundary method to efficiently handle the gas-particle flow. Models are employed to capture subgrid-scale velocity fluctuations induced by particles (termed pseudo-turbulence). The flow begins perfectly mixed and relatively uniform. Strong momentum coupling between the phases results in the spontaneous generation of dense clusters that fall near the riser walls (annulus). It is hypothesized that smaller particles are more likely to rise fast with the high-speed flow in the relatively dilute core, while larger particles are more susceptible to gravity and remain in the slower moving dense annulus. Size segregation during the initial transient will be quantified. Two-phase flow statistics at steady state will be studied to reveal correlations between voidage and velocity with particle size. Finally, a conclusion is drawn on the effect of polydispersity on the size and the shape of the clusters.