Coupled optimization of auxiliary channels and porosity gradient of GDL for PEMFC

In this study, we proposed a novel under-rib auxiliary channel and developed a systematic methodology combining computational fluid dynamics (CFD), an artificial neural network (ANN), and a particle swarm optimization (PSO) algorithm for the coupled optimization of the auxiliary-channel structure and gas diffusion layer (GDL) porosity gradient in a proton exchange membrane fuel cell (PEMFC). First, we developed a three-dimensional multiphase numerical model of a non-isothermal parallel flow-field PEMFC. ANN was then applied to obtain a mapping relationship between auxiliary channels structure, GDL porosity gradient and output performance of PEMFC with finite CFD data. Finally, PSO was employed to perform an optimization search to determine the optimal parameter pairing, resulting in a 16.7 % increase in maximum power density and a 20.6 % increase in the current operating range. In addition, the optimized PEMFC exhibited superior mass transfer and water removal capabilities, where the average oxygen concentration in the intermediate section of cathode GDL was increased by 33.3 % and the maximum liquid water saturation was reduced by 23.5 %. Moreover, the coupled optimization of the auxiliary channels and the GDL porosity gradient can increase power density by at least 8.7 % and current density by 9.1 % compared to single component optimization.