Answers are grounded in this paper and cite their sources.
Vaned diffusers are widely employed in centrifugal compressors to improve stage aerodynamic efficiency. However, their presence leads to a reduction in the stable operating range due to the increased tendency for flow separation within the diffuser passages. Consequently, identifying an optimal diffuser vane geometry is critical to achieve a compromise between aerodynamic performance and operational stability. In the present study, a three-dimensional compressible Reynolds-averaged Navier-Stokes (RANS) framework is adopted to investigate the effects of vane surface geometry modifications on the aerodynamic performance and stall margin of a centrifugal compressor. The baseline compressor is a transonic compressor with a relative inlet Mach number of 0.85 and a Reynolds number of approximately 1.5 × 10 6 at the design point. A novel modal shape parametrization technique was introduced to systematically modify the diffuser vane surface, providing a compact physics-informed representation of complex geometric variations. Fourteen design configurations were generated using an optimal space-filling (OSF) approach, based on two governing geometric parameters: mode number and mode amplitude. A surrogate model was constructed using a genetic aggregation algorithm (GA) technique trained with the numerical dataset, and a multi-objective genetic algorithm (MOGA) was then employed to identify optimal vane configurations that simultaneously enhance efficiency and extend stall margin. The results indicated that most of the modified designs enhanced both efficiency and operating range compared to the baseline case. These improvements are primarily attributed to the alteration of flow separation within the diffuser passage, a reduction in the incidence angle, and improved flow uniformity at the diffuser inlet. The optimal vane geometry exhibited a consistent aerodynamic performance improvement over the entire operating range, achieving an efficiency increase of approximately 4.1% at the design point, along with a 2.722% improvement in stall margin.
No takes yet. Share an insight, caveat, or question.
Zamiri et al. (2026) studied this question.
Synapse has enriched 10 related papers. Consider them for comparative context: