Multiobjective Optimization of Low-Specific-Speed Multistage Pumps by Using Matrix Analysis and CFD Method

Abstract

The implementation of energy-saving and emission-reduction techniques has become a worldwide consensus. Thus, special attention should be provided to the field of pump optimization. With the objective of focusing on multiobjective optimization problems in low-specific-speed pumps, 10 parameters were carefully selected in this study for an $L_{27}$(3¹⁰) orthogonal experiment. The parameters include the outlet width of the impeller blade, blade number, and inlet setting angle of the guide vane. The numerical calculation appropriate for forecasting the performance of multistage pumps, such as the head, efficiency, and shaft power, was analyzed. Results were obtained after calculating the two-stage flow field of the pump through computational fluid dynamics (CFD) methods. A matrix method was proposed to optimize the results of the orthographic experiment. The optimal plan was selected according to the weight of each factor. Calculated results indicate that the inlet setting angle of the guide vane influences efficiency significantly and that the outlet angle of blades has an effect on the head and shaft power. A prototype was produced with the optimal plan for testing. The efficiency rating of the prototype reached 58.61%; maximum shaft power was within the design requirements, which verifies that the proposed method is feasible for pump optimization.