Prediction on Turbomachinery Flows Using Advanced Turbulence Models

Gas turbine engine is a very crucial technology for civil and military applications. Accurate prediction is very important for the numerical design process of a new turbomachine or for improving the performance of an existing design. It can largely lower the cost of expensive rig tests. In a numerical design process for turbomachine, 3D CFD is the last step prior to rig tests. If it is not accurate enough, lots of money and resources will be wasted in expensive rig tests. Therefore, the predictive accuracy of 3D CFD analysis is of great importance. In this study, a 3D CFD procedure is used to compute complicated multi-stage rotor/stator flows in turbomachines with an advanced eddy-viscosity turbulence model V₂f. Linear EVM and k-ω SST models are also used for comparison. V₂f model has been proven to return much more accurate results in many simple and yet important flows than other models, such as k-ε eddy-viscosity models, that are widely used in industry today. V₂f model can predict stress anisotropy, transition, and effects associated with streamline curvature, hence, it is expected to perform equally well in complicated turbomachinery flows. In addition, the computational cost of V₂f model is not much higher than those simple linear EVMs, making it an ideal model for turbomachinery flow simulation. The results prove that the predictive accuracy of V₂f model is much better than linear EVMs in those cases investigated here.