Study of Particle Dispersion Models in a Cyclone Separator with Hybrid Turbulence Models

Hydrocyclone is a widely used industrial equipment for separation and classification of particles. It has the advantage of centrifugally separating particles without any moving parts. k-ω SST IDDES hybrid models have been used to predict hydrodynamics and flow quantities. Multiphase flow in the cyclone separator is studied using different particle submodels. Hydrodynamics serves as an input to these particle sub-models, including the dispersion model. The present work focuses on the applicability of different dispersion models in combination with the k-ω SST IDDES hybrid turbulence model. Numerical simulations are carried out with particles of different sizes. From the results, we find that the MOB dispersion model is the best among the different stochastic dispersion models considered. The other models under discussion are CRW and DRW. For sub-micron parcels, the isotropic DRW shows better prediction than CRW. For intermediate ranges of particle sizes, the order of better prediction is MOB>CRW>DRW. With MOB model predictions, the effect of Saffman lift force on efficiency prediction is studied, and it is reported that there is no conceivable change observed because of the inclusion of Saffman lift force in the given particle range. These results will be important in predicting the efficiency of the device.