A multi-objective parametric study of the claw hydrogen pump for fuel cell vehicles using taguchi method and ANN

As hydrogen circulation pump is one of the key components for hydrogen circulation in Fuel Cell Vehicles (FCVs), the optimization of a hydrogen circulation pump is critical for the performance improvement of a FCV. This study focuses on six factors that have impacts on the performance of a claw-type hydrogen pump, including the rotating speed, pressure ratio, inlet pressure, the radial clearance between the rotor and the casing (RC1), the radial clearance between the rotors (RC2), and the axial clearance (AC). By using the Taguchi method, a series of CFD simulation cases with different levels of the six factors were carried out to determine the optimum conditions for the volumetric efficiency and the shaft power. With the analysis of variance (ANOVA) method, the quantitative contribution of these six factors to volumetric efficiency and shaft power were obtained separately. According to the results, pressure ratio (36.2%), AC (29.4%) and rotating speed (21.5%) had the greatest impact on volumetric efficiency, while shaft power was more sensitive to the pressure ratio (64.6%), rotating speed (23.0%) and inlet pressure (5.2%). Based on the analysis of the ANOVA method, the most significant factors on volumetric efficiency and shaft power were chosen to be input factors of the Neural Network and the data of simulation were adopted to train the Neural Network to predict the performance of a claw pump. The accuracy of the Neural Network was tested and validated. The results can be used as guidelines for the design and selection of claw pumps used for fuel cell systems.