We investigated a multi-objective optimization technique that can control a trade-off relationship between the efficiency and the stall margin of two-dimensional impellers using only a calculation result at the design flow rate. The suction flow coefficient at the design flow rate of the impeller we designed was 0.04. We used a multi-objective genetic algorithm and three-dimensional steady Reynolds Averaged Navier-Stokes (RANS) simulation with a blade-to-blade passage in the optimization. To determine a flow field index that can evaluate the stall margin of an impeller, the inner flows of two impellers were investigated. Results showed that the throat deceleration ratio correlates with the inception flow rate of the flow separation that occurs on the blade suction surface in the shroud side near the throat surface. Here, the throat deceleration ratio means a ratio of the relative velocity at the inlet to that at the throat surface in the shroud side of the impeller. We selected the adiabatic efficiency and the throat deceleration ratio at the design flow rate as the objective functions and performed an optimization. A 1.8% improvement in the efficiency and a 5% improvement in the stall margin were demonstrated in the optimized impeller from the detailed calculations. Moreover, it was found that an additional objective function was required to control the flow separation caused by the excessive increase of the blade loading in the rear part of the impeller.
展开▼
