(Invited) A Systematic Approach to Develop and Validate Models for the Design of Membrane Electrode Assemblies

Abstract

The modeling and simulation landscape for fuel-cells has increased dramatically over last years. A number of commercial and OpenSource (e.g. OpenFCST, FAST-FC, DuMux) software packages are available to the aspiring researcher or engineer in academia and industry. The complexity ranges from purely analytic models with O(1) number of parameters to fully numerically models with O(100) input parameters. However, despite the apparent abundance on model and simulation tools, their practical use for the design of next generation MEA components like the catalyst layers or porous transport layers has been very limited and progress has been mainly driven by large, slow and expensive experimental studies. This stands in stark contrast to combustion engine design where virtual prototypes are built, tested and optimized through numerical simulations before a physical prototype is built. In the fuel-cell community this lack of application tends to be explained by the large number of equations and the corresponding computational resources needed to tackle the problem. The fact that combustion engineering does not suffer this problem, despite an equally high level of complexity, seems to indicate that numerical complexity is not the root cause. Browsing the numerical codes and the modeling literature shows little consensus about the correct description of the individual physical effects. A few of the topics of disagreement are: Gas transport: Fickian diffusion, Maxwelll-Stefan diffusion, Dusty Gas models, ...Membrane water transport: concentration driven, liquid pressure driven, ...Reaction at the catalyst surface: macro-homogeneous models, ionomer filled agglomerate models, water filled agglomerate models,thin film models, ...Electrochemical reaction mechanisms: Tafel, Butler-Volmer, Double-Trap, Dual-Path, ...... In this tutorial, a consistent validation strategy with a focus on gas mass transport is outlined and demonstrated. We will take AFCCs baseline MEA model platform OpenFCST through a validation with ex-situ experiments: Permeability and Diffusivityvalidation through in-situ experiments: Limiting current methodsPolarization curves under different operating conditionsThickness variations validation through imaging techniques: Diffusion simulations on reconstructed catalyst layers The impact of different model configurations will be studied and we will investigate if a validation strategy like this can pinpoint the weaknesses or strengths of the model configuration chosen. References OpenFCST (www.openfcst.org)M. Bhaiya, A. Putz and M. Secanell, "Analysis of non-isothermal effects on polymer electrolyte fuel cell electrode assemblies", Electrochimica Acta, 147C:294-309, 2014. DOI: 10.1016/j.electacta.2014.09.051 Figure 1