Microstructure Reconstruction and Characterization of the Porous GDLs for PEMFC Based on Fibers Orientation Distribution

Abstract The 3D microstructure of the gas diffusion layers (GDLs) is generated, using a stochastic reconstruction approach. The method uses basic input parameters and fibers orientation distribution and is capable to model carbon fiber and binder phases of all types of carbon fiber GDLs with different structural parameters. Morphological operators of image processing are used to add the binder to the fibrous skeleton in three dimensions. Binder impregnation, pore size distribution and tortuosity factor of the reconstructed GDL are evaluated and compared with literature. Twenty and forty percentages of binder volume fraction are used to reconstruct the microstructure. To further investigate the reliability of our approach, mass transport properties of the reconstructed microstructure, namely absolute permeability, effective diffusivity of the pore space and effective thermal conductivity, are investigated through a finite volume equation solver in AVIZO. Only 4% of discrepancy between the absolute permeability calculation and the value reported by the manufacturer is observed in the through‐plane direction. Finally, in order to show the flexibility of the methodology the microstructure of a commercial Toray GDL is also reconstructed and characterized. The proposed method is proved to be a high‐speed and versatile tool for research and development in the GDL material design.