Skip to main content

Lund University Publications

LUND UNIVERSITY LIBRARIES

Numerical Reconstruction of Proton Exchange Membrane Fuel Cell Gas Diffusion Layers

Yang, Danan LU ; Garg, Himani LU orcid ; Beale, S. B. and Andersson, Martin LU (2023) ECS Meeting p.49-61
Abstract
Stochastic reconstruction is widely employed for effective and flexible imitation of Gas Diffusion Layers (GDLs), e.g., to facilitate the study of their properties. However, the reconstruction often overlooks crucial factors such as fiber curvature, fiber stack arrangement, and fiber anisotropy. Consequently, the impact of these structural characteristics remains poorly understood. In this study, an in-house reconstruction procedure is developed based on the periodic surface model. This procedure enables the generation of GDLs with either straight or curved fibers, layer-by-layer or random arrangement, and different probabilities of through-plane fiber orientation angles. The porosity, domain size, and fiber diameter are extracted from an... (More)
Stochastic reconstruction is widely employed for effective and flexible imitation of Gas Diffusion Layers (GDLs), e.g., to facilitate the study of their properties. However, the reconstruction often overlooks crucial factors such as fiber curvature, fiber stack arrangement, and fiber anisotropy. Consequently, the impact of these structural characteristics remains poorly understood. In this study, an in-house reconstruction procedure is developed based on the periodic surface model. This procedure enables the generation of GDLs with either straight or curved fibers, layer-by-layer or random arrangement, and different probabilities of through-plane fiber orientation angles. The porosity, domain size, and fiber diameter are extracted from an experimental image-based GDL and utilized as input data for the reconstruction. Furthermore, the different GDLs are compared in terms of pore size distribution and through-plane porosity distribution. It is concluded that introducing proper selections of these fiber features gives the reconstruction more realistic properties. (Less)
Please use this url to cite or link to this publication:
author
; ; and
organization
publishing date
type
Contribution to conference
publication status
published
subject
pages
49 - 61
conference name
ECS Meeting
conference location
Gothenburg, Sweden
conference dates
2023-10-08 - 2023-10-14
DOI
10.1149/11204.0049ecst
language
English
LU publication?
yes
id
294027ab-7fc1-4701-801c-3151ccd0a307
date added to LUP
2024-02-12 10:26:57
date last changed
2024-02-16 13:50:41
@misc{294027ab-7fc1-4701-801c-3151ccd0a307,
  abstract     = {{Stochastic reconstruction is widely employed for effective and flexible imitation of Gas Diffusion Layers (GDLs), e.g., to facilitate the study of their properties. However, the reconstruction often overlooks crucial factors such as fiber curvature, fiber stack arrangement, and fiber anisotropy. Consequently, the impact of these structural characteristics remains poorly understood. In this study, an in-house reconstruction procedure is developed based on the periodic surface model. This procedure enables the generation of GDLs with either straight or curved fibers, layer-by-layer or random arrangement, and different probabilities of through-plane fiber orientation angles. The porosity, domain size, and fiber diameter are extracted from an experimental image-based GDL and utilized as input data for the reconstruction. Furthermore, the different GDLs are compared in terms of pore size distribution and through-plane porosity distribution. It is concluded that introducing proper selections of these fiber features gives the reconstruction more realistic properties.}},
  author       = {{Yang, Danan and Garg, Himani and Beale, S. B. and Andersson, Martin}},
  language     = {{eng}},
  pages        = {{49--61}},
  title        = {{Numerical Reconstruction of Proton Exchange Membrane Fuel Cell Gas Diffusion Layers}},
  url          = {{http://dx.doi.org/10.1149/11204.0049ecst}},
  doi          = {{10.1149/11204.0049ecst}},
  year         = {{2023}},
}