Evaluation of Surge Prediction Capabilities of Body-Force Model on a High-Speed Multi-Stage Axial-Radial Compressor

Abstract

Multi-stage axial-radial compressors are typical configurations of compression systems used in turboshaft engines. The aerodynamic instabilities encountered in this type of compressor are commonly identified as surge, which seriously threaten the operability and reliability of the compressor itself and even the entire engine. Therefore, correct prediction of surge characteristics and the aerodynamic loading are crucial during the design process. However, due to the complexity of compressor surge, high-fidelity numerical methods, like unsteady Reynolds-averaged Navier-Stokes (URANS) simulation, require enormous computational resources and time costs, which can barely be used in design iterations. Therefore, finding a more efficient way for surge prediction is essential. This paper describes a general method of surge prediction based on an in-house code of body-force model. A high-speed multi-stage axial-radial compressor is used to evaluate the capabilities of this method to predict surge characteristics against URANS. The run-time is reduced by approximately 2 orders of magnitude. Key features of surge (i.e., flow reversal, flow resumption, and repressurization) and the aerodynamic loading during surge are compared. Overall, the results from the two method show a close matching. Additional analyses are also made on the fidelity limitations of this method in the prediction of finer surge features, and the corresponding modifications are proposed.