An intelligent CFD-based optimization system for fluid machinery: Automotive electronic pump case application

Abstract

Improving the efficiency of fluid machinery is an eternal topic, and the development of computational fluid dynamics (CFD) technology provides an opportunity to achieve optimal design in limited time. A multi-objective design process based on CFD and an intelligent optimization method is proposed in this study to improve the energy transfer eciency, using the application of an automotive electronic pump as an example. Firstly, the three-dimensional CFD analysis of the prototype is carried out to understand the flow loss mechanism inside the pump and establish the numerical prediction model of pump performance. Secondly, an automatic optimization platform including fluid domain modeling, meshing, solving, post-processing, and design of experiment (DOE) is built based on three-dimensional parametric design method. Then, orthogonal experimental design and the multi-island genetic algorithm (MIGA) are utilized to drive the platform for improving the eciency of the pump at three operating flowrates. Finally, the optimal impeller geometries are obtained within the limited 375 h and manufactured into a prototype for verification test. The results show that the highest eciency of the pump increased by 4.2%, which verify the eectiveness of the proposed method. Overall, the flow field has been improved significantly after optimization, which is the fundamental reason for performance improvement.