Multi-objective optimization of centrifugal pump impeller for conceptual design

Two multi-objective optimization methods were applied to design the centrifugal oxidizer pump impeller for liquid propellant rocket engine. Major geometric parameters of pump impeller were optimized for the targeted delivery head, efficiency and mass of pump. The Pareto Front of multi-objective optimization was gained by Neighborhood Cultivation Genetic Algorithm (NCGA); the super-transitive approximation method was introduced to acquire optimal weights of individual objectives. Then, Evaluation Function method was used to obtain optimal design of multi-objective optimization. The effects of major structural parameters on pump performance parameters were also analyzed.