Evaluation of various turbulence models in predicting screw compressor flow processes by CFD

Computational Fluid Dynamics (CFD) plays an important role in modelling and design of positive displacement screw machines. It is used mainly to identify and quantify effects of internal flows which cannot be modelled by use of other methods. Although significant improvements have been made to achieve accurate prediction of the overall performance, there are still aspects which have not been fully evaluated, such as the influence of turbulence modelling on local velocities. The usual practice is to use k-ε model but it sometime fails to predict local velocity values in the compression chamber and ports measured by Laser Doppler Velocimetry (LDV). This paper evaluates and compares various readily available turbulence models in screw compressor performance predictions. Modelling of the screw compressor with three different turbulence models on three different mesh sizes is performed by use of commercial CFD software ANSYS CFX v11. These predictions are then compared with the measurements obtained by LDV. It is shown in the paper that the flexibility of the developed procedures together with appropriate turbulence modelling creates further opportunities for a broader use of CFD for analysis of twin screw machines in a range of new applications.