Numerical study of the effects of particle size and polydispersity on the agglomerate dispersion in a cyclonic flow

This paper presents an investigation of powder dispersion based on a combined computational fluid dynamics (CFD) and discrete element method (DEM) approach. Agglomerates of different particle sizes and polydispersities are dispersed in a cyclonic flow at different flow velocities. The analysis of flow field and agglomerate properties indicates that the dispersion is governed by two competitive interactions, i.e. particle–particle cohesion and particle–wall impact, with the latter related to the air flow or the particle–fluid interaction. The internal shearing induced by air flow, however, only plays a minor role in powder dispersion for the system considered. The dispersion performance is described by fragment size distribution and fine particle fraction (FPF) which is the weight percentage of fragments with size less than 4.5 μm. Agglomerates of finer particles are more difficult to disperse at low flow rates due to strong particle–particle cohesion, but their dispersion becomes more efficient at high flow rates. Agglomerates with narrower particle size distributions tend to have better dispersion. An index based on the ratio of particle–wall impact energy and agglomerate cohesion energy is proposed to provide a quantitative characterization of dispersion performance.