Validation of pore network simulations of ex-situ water distributions in a gas diffusion layer of proton exchange membrane fuel cells with X-ray tomographic images

Abstract

OATAO is an open access repository that collects the work of Toulouse researchers and makes it freely available over the web where possible. This is an author-deposited version published in : http://oatao.univ-toulouse.fr/ Eprints ID : 17329 Pore network extraction parameters were benchmarked using full morphology. Keywords:Polymer electrolyte membrane fuel cell Gas diffusion layer X-ray tomographic microscopy Pore network Full morphology Liquid water a b s t r a c t Understanding and modeling two-phase flows in the gas diffusion layer (GDL) of proton exchange membrane fuel cells are important in order to improve fuel cells performance. They are scientifically challenging because of the peculiarities of GDLs microstructures. In the present work, simulations on a pore network model are compared to X-ray tomographic images of water distributions during an ex-situ water invasion experiment. A method based on watershed segmentation was developed to extract a pore network from the 3D segmented image of the dry GDL. Pore network modeling and a full morphology model were then used to perform two-phase simulations and compared to the experimental data. The results show good agreement between experimental and simulated microscopic water distributions. Pore network extraction parameters were also benchmarked using the experimental data and results from full morphology simulations.

Virtual Liquid Water Intrusion in Fuel Cell Gas Diffusion Media

“…Although the CFM predicted capillary pressure-saturation curve for the sample used in this study is quite different from the μ-CT and literature curves, a FM model has been shown to be able to predict the capillary pressuresaturation curve for SGL 24BA. 62 Therefore, analysis of different samples might be needed to improve the limitations of the CFM model. Schulz et al 53 proposed an improvement to the FM model by creating fictitious spheres with diameters increased by the cosine of the contact angle to account for the wettability of the walls.…”

“…Vogel et al 60 and Kim et al 61 however did not study fibrous media, such as gas diffusion electrodes. Agaesse et al 62 used μ-CT reconstructions of partially saturated GDLs at different capillary pressures to validate numerically simulated liquid water distributions using a pore network model and a full morphology approach. Their work demonstrated the capability of the numerical models to predict capillary pressure-saturation curves and saturation profiles similar to those obtained from the μ-CT reconstructions.…”

Versatile and efficient pore network extraction method using marker-based watershed segmentation

“…The median filter is actually recommended when preserving edges is desired, which is not the case here. Agaesse et al [35] applied a gray-scale reconstruction, sometimes referred to as an H-minima filter, that converts peaks to plateaus to remove small local peaks [44]. As discussed below, however, plateaus are also problematic and should be avoided.…”

“…Rounding out the methods is the watershed segmentation of the pore space, that was originally investigated by Thompson et al. [32] and Sheppard et al [33] over a decade ago, but has received surprisingly little attention until recently [34,35]. The common thread running through the existing network extraction algorithms is that they are essentially optimized for rock and soil, due to the importance of these media in oil recovery and aquifer management.…”

“…The algorithm begins by finding all the peaks that would lead to an oversegmented watershed, then progressively reducing the number of local maxima by ruling out several classes of peaks, producing a subnetwork of the oversegmented watershed (SNOW). Agaesse et al [35] have presented a rough outline of a similar extraction algorithm applied to fibrous media, but they only explored two alternative methods for removing spurious peaks and the results were not scrutinized. The SNOW algorithm, meanwhile, uses different methods, and combines several of them to arrive at a more complete solution that is validated against several well-studied materials.…”