Abstract: To achieve high- efficient and precision simulation of proton exchange membrane (PEM) fuel cells and stacks, this work develops a three-dimensional coupled one-dimensional hybrid dimension modeling strategy based on the FLUENT solver. An in-depth analysis of the impact of the FLUENT built-in Global Summations macro (PRF_GRSUM) on the parallel solving efficiency of fuel cell simulation was conducted. Based on the data coupling characteristics of the hybrid dimension model, two new parallel strategies, including the data compression synchronization method and the data filter marker method, are proposed. The principles and processes of these three data synchronization methods are described in detail, and the impact of these methods on the solving efficiency of parallel computation of PEM fuel cell models is investigated. The results show that the data filter marker method can speed up the simulation to 112.5 times compared to using the PRF_GRSUM macro when simulating a PEM fuel cell stack with a hundred cells.

Key words: proton exchange membrane fuel cell, parallel computation, hybrid dimension model, data filter marker method