Effectiveness of Lean/Sweep and Boundary Layer Suction Combination on Compressor Cascade Performance

The flow in the endwall regions of compressors is complex, with various vortex structures interacting and inducing corner separation phenomena. This interaction generates significant aerodynamic losses and adversely affects the operational efficiency of the compressor. In order to further enhance the performance of axial-flow compressors, this study employs computational fluid dynamics to investigate the combined effects of lean/sweep profilings and boundary layer suctions on the internal flow dynamics of a low-speed compressor cascade under different operating conditions. Using an RBF neural network as a surrogate model, with the total pressure loss in design and near stall conditions as the objective functions, optimization is performed through NSGA-II multi-objective genetic algorithm to obtain three optimal lean/sweep profilings. The lean/sweep profiling suppresses the development of flow separation by reducing the reverse pressure gradient at the mid-chord position on the suction surface. Spearman correlation analysis indicates that the positive vorticity in the cascade passage substantially increases the total pressure loss of the cascade. Moreover, under the near-stall condition, the correlation coefficients between secondary flow intensity and positive vorticity, as well as between secondary flow intensity and reverse flow volume, are higher than those under the design condition. This suggests that at high angles of attack, the secondary flow intensity is more conducive to the development of passage vortices and is also more prone to induce reverse flow phenomena. The optimized lean/sweep profilings are coupled with various endwall boundary layer suctions to further reduce the total pressure loss. Compared to the original cascade, the total pressure loss in both design and near-stall conditions is reduced by over 20% with the combined lean/sweep profiling and boundary layer suction scheme. The lean/sweep profiling significantly improves the flow field at midspan but slightly increases losses near the endwall. Conversely, the endwall suction slots reduce losses near the endwall by removing the boundary layer, albeit with minor effects on the flow field at midspan. The dissipation function has been utilized to quantitatively measure the loss intensity in four regions: the leading edge, trailing edge, blade surface, and passage. Under the design condition of the combined sweep/lean profiling and boundary layer suction scheme, the dissipation intensity at the leading edge and trailing edge is reduced by more than 25%, and the passage loss is reduced by 7.7%. However, under near-stall conditions, due to the influence of secondary flows, the dissipation loss on the blade surface increases by 8.8%, while the passage vortex intensity significantly decreases, resulting in a reduction of passage dissipation loss by 11.5%. Compared to a single flow control scheme, the combined scheme is more effective in reducing internal airflow blockage and enhancing the adaptability under multiple operating conditions.