Unsteady 3D CFD Simulation of a PEM Fuel Cell: Efficient Parameterization and Simulation using Numerically Reduced Models

Abstract

The reliable and robust model parameterization of polymer electrolyte membrane fuel cells (PEMFCs) is challenging. It is difficult to find appropriate material data in the literature, and the fuel cell is scarcely accessible for measurements. Further, the dependence on many material parameters, e.g., temperature, water content, and pressure, makes the model setup even more challenging. The goal of the present study was to perform an unsteady 3D-CFD simulation of an automotive PEMFC to investigate membrane humidification and current homogeneity during transient operation. A bottom-up methodology was chosen to find suitable material parameters. First, a real-time capable fuel cell model in AVL Cruise was adapted to available experimental data using numerical optimization methods. Eight unknown material parameters characterizing the catalytic layer, membrane, and gas diffusion layers were calibrated. In the two following steps, the model parameters were implemented in an AVL Fire CFD model of a single channel and, finally, in the model of the entire fuel cell using the homogenized channel approach. During these two steps, the material parameters were fine-tuned to match the experimental data. With the obtained parameter values, a steep load step was simulated with the 3D-CFD model. The liquid water formation and self-humidification strategy of the PEMFC were investigated in detail and could be explained. With an excellently calibrated base model, further investigations concerning degradation will be possible. This will be the next step in follow-up research work.