Parameter Optimization of Centrifugal Pump Splitter Blades with Artificial Fish Swarm Algorithm

Abstract

Low specific speed centrifugal pumps typically suffer from low efficiency and severe backflow; adding optimally structured splitter blades can play a role. In this paper, the distribution of pressure and velocity in the flow channel is analyzed using CFD simulation for a low specific speed centrifugal pump. The geometric parameters of the splitter blade are optimized using an orthogonal test and an artificial fish swarm algorithm; then the optimal splitter blade structure is obtained. Results showed that the splitter blade not only effectively solves the backflow of the flow channel and compresses the range of the trailing vortex, but it also alleviates the cavitation at the inlet of the main blade. When considering the best head, the order of influence of each factor is: Splitter blade thickness > Splitter blade inlet diameter > Splitter blade inlet width. At this time, the thickness of the splitter blade is 4.5 mm, splitter blade inlet diameter is 155 mm (0.775) and Splitter blade inlet width is 23 mm. Through the closed pump experimental system, it is confirmed that hydraulic performance has been improved.