Bio-inspired trailing edge optimization for a double-suction pump impeller: a hybrid NSGA-II and CFD approach

Energy consumption reduction in pump systems has attracted increasing attention. Improving the efficiency of double-suction pumps can contribute to reducing energy consumption due to their extensive applications. This paper aims to optimize the double-suction pump by introducing the bionic structure inspired by the pectoral fin structure of humpback whales. In particular, this paper targeted the design improvement of the trailing edge of the double-suction pump impeller based on the tubercle structure on the pectoral fin. A combined surrogate model with a genetic algorithm was employed to predict and optimize the external characteristics of double-suction pumps with different wavelength and amplitude. Subsequently, the obtained optimal model performance was validated with numerical simulation. The results show that the pump's head was increased by 4.3%, and the efficiency increased by 1.3% under design conditions. The numerical model was validated against experimental measurements, with maximum deviations of 3.57% for head and 2% for efficiency. Furthermore, the Kriging surrogate model achieved R2 values of 0.93 and 0.92 for head and efficiency prediction, respectively, demonstrating the reliability of the proposed optimization framework. Also, a comparison was made between the pressure distribution, turbulent kinetic energy distribution, and vortex core structure distribution of the blade before and after optimization. The optimized structure reduced areas with high turbulent kinetic energy and vortex core structures and high-velocity vortices at the trailing edge, thereby improving the head and efficiency of the double-suction pump. It was concluded that the implementation of a bionic sine-curve structure at the blade trailing edge, optimized via a genetic algorithm, provides a new approach for improving the external characteristics of the double-suction centrifugal pump.