Evaluations and modifications on Reynolds stress model in cyclone simulations

The prediction performance of the Reynolds stress model (RSM) on the flow field in a cyclone has been validated by using particle imaging velocimetry (PIV) experimental results. The validations mainly focus on two features of the flow information with averaged and fluctuating flow fields, which can mostly reflect the turbulent flow properties in the cyclone. The comparisons between predictions and measurements show that the RSM has good performance on the averaged flow field prediction, especially on the prediction of the mean tangential velocity. However, for the fluctuating flow field, the prediction performance of RSM is rather poor. The predicted root mean square (RMS) velocities are greatly underestimated. The standard model coefficients used in RSM have to be modified to enhance its accuracy when used to predict the fluctuating structure of the strong swirling flow in the cyclone. The recommended model coefficients give much better predictions on the fluctuating flow field than the standard coefficients, without decreasing the prediction accuracy on the mean flow field.