The absence of standardized hardware hinders consistent membrane electrode assembly characterization in proton exchange membrane fuel cells. To address this, the European Joint Research Centre established two reference configurations: the Single-Serpentine Testing Hardware (SSHW) and the Zero ∇ Cell (ZGC). While experimental studies demonstrated the superior performance of the ZGC, the underlying physical mechanisms have not yet been fully clarified. In this work, three-dimensional computational fluid dynamics simulations were performed to reproduce and interpret the operation of both cells, elucidating operational aspects inaccessible for direct measurements. The models were initially validated against the available experimental data, and then used to provide spatially resolved insights into the interplay between electrochemical kinetics, mass transport, and water management. A novel flux-decomposition sub-model is implemented to decouple electro-osmotic drag from back-diffusion, enabling spatially resolved quantification of hydration dynamics inaccessible to direct measurement. Overpotentials are dissected to indicate the dominant loss type, with simulation results confirming that the ZGC configuration exhibits lower ohmic and mass transport losses than the SSHW, thanks to its more uniform hydration and reactant distribution combined with a larger area for current conduction, whereas the SSHW is subject to a relevant oxygen depletion and to intense anode dehydration, leading to increased ohmic and mass transport overpotentials. Beyond the comparison of the two hardware designs, this study focuses on the capability of the developed modeling framework to extend the understanding of cell’s hard-to-quantify quantities (e.g., spatial distributions, membrane water flux decomposition, overpotentials breakdown), offering a robust complementary methodology for the design, optimization, and standardization of testing protocols. • Spatially-resolved Electro-Osmotic Drag and Back-diffusion balance model. • 3D-CFD analysis explains performance gaps in JRC reference hardware. • Enhanced overpotential breakdown through simulation.
Marra et al. (Fri,) studied this question.