Optimization of Component Material Properties and Flow Channel Geometric Parameters in a PEM Fuel Cell Cathode

PEM fuel cells are generally composed of components which have spatially uniform material properties. In other words, the material properties in one location in a fuel cell component are the same as those in a different location in the same component. Such components are cost-effective and easy to manufacture and model. However, recent advances in additive manufacturing at the micro- and nano-scale level have enabled the creation of parts and components whose material properties are spatially non-uniform. Components with spatially non-uniform material properties could potentially improve the performance of a PEM fuel cell.

To further illustrate the potential benefit of fuel cell components with spatially non-uniform material properties, consider the cathode side of a fuel cell with a serpentine flow channel. At the inlet, the oxygen concentration is highest, while at the outlet, the oxygen is depleted due to electrochemical reactions. However, the material properties of the typical gas diffusion layer are the same near the inlet as they are near the outlet. Due to the non-uniformity of the reactant concentration profile, it stands to reason that the performance of the fuel cell could be impacted by a gas diffusion layer which has a spatially variable effective diffusivity.

In this work, we use STAR-CCM+ to simulate heat transfer/production, flow, species transport, electrodynamics, and electrochemical reactions in the cathode side of a fuel cell with a serpentine flow channel. We use Optimate+, a STAR-CCM+ embedded optimization package, which uses a proprietary hybrid-adaptive algorithm named SHERPA, to maximize the total current generated in the cathode side of the fuel cell by changing the porosity profiles in the gas diffusion layer and catalyst layer, the platinum loading profile in the catalyst layer and the geometric parameters of the flow channel design. We will present results from our simulations and optimization study in addition to discussing the limitations of our physical model and future topics for research.