Numerical studies on ejector in proton exchange membrane fuel cell system with anodic gas state parameters as design boundary

摘要

A comprehensive entrainment performance evaluation system of the ejector was built including four indexes. An ejector's Computational Fluid Dynamics (CFD) model was established, and the sensitivity analysis of the entrainment performance to four key geometry parameters of the ejector, namely, the nozzle diameter (D-n), the primary nozzle exit position (NXP), the mixing tube diameter (D-m), and the secondary flow inlet diameter (D-s) was performed. Based on the quantified boundary conditions obtained by the Simulink simulation, the ejector structure was optimized with a new method. It is found that the total recirculation ratio increases but the hydrogen recirculation ratio decreases with the increase of the relative humidity of the secondary flow. The hydrogen recirculation ratio shows a unidirectional increase tendency with the increase of D-s and the decrease of D-n and NXP. The hydrogen recirculation ratio increases firstly and then decreases with the increase of D-m. High hydrogen recirculation ratio with low primary hydrogen flow rate, corresponding to low current operation point of fuel cell system and low sensitiveness to the changing relative humidity are usually incompatible. The hydrogen recirculation ratio with low primary flow rate degrades significantly when D-n increases and D-m is larger than a certain value. The ejector with smaller D-s shows lower sensitiveness to the changes of relative humidity, while the hydrogen recirculation ratio with low primary hydrogen flow rate is not affected badly. When matching with a specific system, it is necessary to balance the ejection performance at low current density operating points and the sensitiveness to the changes of relative humidity in combination with the anodic gas state at each operating point, so as to find the optimal structural parameters. The optimization sequence of structural parameters should follow: D-n is selected firstly, then NXP and D-s are optimized, and finally D-m is chosen. (C) 2021 Hydrogen Energy Publications LLC. Published by Elsevier Ltd. All rights reserved.