Prediction of Low Speed Compressor Rotor Flowfields With Large Tip Clearances

Abstract

Rotor tip modeling fidelity, grid resolution, and near wall modeling have been examined to determine the requirements for an accurate prediction of the effects of large tip clearance in a low-speed axial compressor rotor. The effort, using a Reynolds-Averaged Navier-Stokes (RANS) solver, aimed to obtain the most accurate predictions from a three-dimensional, steady, single blade row simulation. A recently tested, modern low speed rotor, was used as the test geometry; the measured pressure rise characteristic as well as detailed data near stall was used to evaluate the ability of different modeling strategies to capture the correct flow structure. The leakage flow was quantified to show that a wide range of tip blockage could be obtained for different simulations of the same geometry and conditions. The results show that using a square tip and gridding to fully resolve the real tip gap was better able to capture the effects of loading on the leakage flow than either of the approximate models studied. Sufficient clustering near the casing to capture the shear layers was also found to be critical. While wall integration provided the best results in simultaneously improving the prediction of pressure rise characteristics and flow range, higher fidelity wall modeling and a casing y+ of approximately 3 were found to provide similar benefits.