Optimization design of aviation centrifugal pump based on synthetical loss model

Structural design optimization for the centrifugal pumps is a promising way for achieving configurations with high mechanical efficiency and stability, but the accurate synthetical loss assessment of the centrifugal pump has been a great challenge during this process. The aim of this work is to develop models and methods for hydraulic optimization of the centrifugal pump. First, an optimization platform is established based on a Synthetical Loss Assessment (SLA) model, which consists of parametric modelling, computation of synthetical efficiency, validation of CFD numerical simulation, as well as intelligent optimization module. With the SLA model, the functional relationship of the loss with design parameters has been established. With the synthetical efficiency as target, an evolutionary optimization algorithm is performed together with the SLA model for searching the optimal structural configuration. Using CFD simulation, the pump performance is examined and compared for the structural configurations before and after optimization. By experimental examination, it is shown that the prediction accuracy of the proposed SLA model is 1.5% higher than that of CFD computation. It is also concluded from the CFD simulation results that, the optimized pump show more uniform static pressure distribution, weaker axial dissymmetry and smaller area of low pressure in the entrance; and volutes with either rectangular cross-section or trapezoidal cross-section result in significant reduction of backflow at the volute tongue compared to the prototype pump. It is ultimately concluded that, the configuration of optimization design shows higher working and suction performances than the prototype one.