Multi-objective optimization and experimental validation of a mixed-flow impeller in compressed air energy storage system

In compressed air energy storage (CAES) systems, employing high-flow-coefficient mixed-flow compressors as a replacement for multi-stage axial compressors plays a significant role in improving system efficiency, enhancing off-design performance, and reducing system complexity and cost. However, as the flow coefficient and the isentropic pressure-rise coefficient increase, the difficulty of impeller optimization design escalates. To address this, a surrogate-based optimization framework is adopted in this study to perform multi-objective optimization of the full-range performance of the impeller. Under simultaneous constraints on the choking margin and the design-point pressure ratio, a 1.20% improvement in peak efficiency is achieved. Subsequently, both machining and experimental testing of the optimized impeller are completed. The experimental results show good agreement with the computational fluid dynamics (CFD) predictions, with the measured peak efficiency exceeding 87%, thereby validating the effectiveness of the proposed optimization method and the precision of the CFD simulations. Finally, flow field analysis indicates that fully free-form blades can significantly improve the incidence angle matching at the leading edge and reduce irreversible losses in that region. Moreover, the geometrically complex blade surface better conforms to the flow field, mitigating downstream secondary flow losses and the mixing loss between tip leakage flow and the mainstream. The numerical and experimental results presented herein provide valuable references for the design and optimization of high-flow-coefficient mixed-flow impellers in CAES systems.