(189w) Study of Different Approaches for Modeling Cyclones Using Cfd

Due to their relative robustness and low cost, cyclones are widely used in industries as gas-solid and liquid-solid separators. Their performance is usually measured by two parameters: pressure drop and collection efficiency. While this equipment is simple in concept, studies have shown that the confined vorticial flow inherent to cyclones is in fact quite complex. Because of this, computational fluid dynamics (CFD) has been employed to model flows and solve them numerically, and has shown relative success in dealing with phenomena such as vortex breakdown, reversal flow, and high turbulence intensity.

Cyclone simulations are usually expensive in terms of computer processing, mainly due to the fact that an anisotropic turbulence model must be used. Several studies have addressed that issue, and Reynolds Stress Model (RSM) has been considered by many researchers as being the most suitable model. In order to reduce computational times while still acquiring reliable results, other aspects of the simulation must be studied, such as the modeling approach to the particulate phase, which is the subject of this work.

While the Eulerian approach is generally applied for the gas, different schemes can be considered for the particulate phase. In this work, two different methods are explored: the classic Eulerian-Eulerian (E-E), where a mean diameter is considered to calculate the entirety of the flow field of the particulate phase; and Eulerian-(Eulerian)n (E-En), where particles with different diameters are included in the same simulation, each as a distinct phase. The fractional collection efficiency curve is obtained for both methods. An analysis is made regarding accuracy and computational costs.

The simulations make use of ANSYS CFX, a commercial software package, running in parallel mode in a high performance computing cluster. The fractional and overall collection efficiencies comparison with the experimental data showed that E-En achieved more accurate results.

KEYWORDS: CFD, cyclone, gas-solid