A novel numerical simulation method to verify turbulence models for predicting flow patterns in control valves

Abstract

Control valves have been widely used in thermal and nuclear power plants to regulate the flow entering the turbine. The safety and stable operation of the turbine is threatened by the control valves' vibration and noise, which are directly related to the flow pattern. Many turbulence models have been used to show the mechanism in previous research, but few studies compared their abilities. In this study, four turbulence models, realizable k-ε, SST, DES based on SST and DES based on realizable k-ε, are selected to predict the flow pattern. A novel simulating method with a continuously changing pressure ratio is employed in the numerical simulation. Using a numerical method, the turbulence models are tested by two challenges in aerodynamic flow:(I)growth and separation of the boundary layer and (II)boundary layer reattachment. Compared to the traditional method with a constant pressure ratio, the new method offers more clear and credible results. The results are successfully verified by the sound mutation in the experiment. It is found that the DES based on the realizable k-ε model is the most powerful turbulence model for the two processes of boundary layer separation and reattachment.