Analysis of Turbulence Models Performance for the Predictions of Flow Yield, Efficiency, and Pressure Drop of a Gas-solid Cyclone Separator

This paper presents the results obtained from the application of computational fluid dynamics (CFD) to modelling the flow field of a Lapple cyclone and to optimizing the cyclone based upon its geometrical parameters. A pre-processor software GAMBIT was employed to set up the configuration, discretisation, and boundary conditions of the cyclone. The characteristics of the cyclone being studied was 0.2 m in diameter, receiving a gas flow rate of 0.1 m3/s with a particle mass loading of 0.01 kg/m3. A commercial CFD code FLUENT 6.2.16 was employed to simulate the flow field and particle dynamics in the cyclone. The objective of this research was to investigate the performance of a number of turbulence models on the prediction of the flow field, collection efficiency and pressure drop in the Lapple cyclone. A number of five turbulence models under Reynolds Averaged Navier Stokes (RANS) category, including Spallart-Allmaras, standard k-ε model, RNG k-ε model, standard k-ω model, and Reynolds Stress Model (RSM) were examined in the simulation of the flow field and particle dynamics inside the cyclone. A validation of all calculation was performed by comparing the predicted results in terms of axial and tangential velocities, efficiency and pressure drop against experimental data of a Lapple cyclone taken from literature. The results of the investigation show that out of five turbulence models being tested, the RSM presented the best predicted results. The predictions of axial and tangential velocities as well as cyclone efficiency by this model are in excellent agreement with the experimental data. Although the pressure drop in the cyclone is under-predicted, the RSM predictions are far better than those of other model. Other turbulence models are over-predicted and under-predicted the axial and tangential velocity, respectively. With respect to efficiency and pressure drop of the cyclone, other models are capable of following the trend of the experimental data but they failed to agree with the experimental values. These results suggest that the RSM is the most suitable turbulence model to represent the flow field and particle dynamics inside a cyclone gas-solid separator