A universal two-stage framework for gas diffusion layer reconstruction in proton exchange membrane fuel cells: Toward high-fidelity through-plane porosity distribution

Traditional stochastic reconstruction method of paper-type gas diffusion layer (GDL) in proton exchange membrane fuel cell (PEMFC) faces the limitation of generation accuracy. Firstly, according to the derivation process of porosity, the coupling effect of carbon fiber porosity and structural element radius on the post-additive-addition structure porosity is proposed, along with an improved inverse derivation method for porosity distribution of GDL. According to the generation process, combined with the principle of morphological processing technology, the influencing factors and mechanism of the porosity of the generated structure are analyzed, and the fitting formula between the porosity and the influencing factors is proposed, with the R2 value of the fitting result larger than 0.98. Based on the above research, an improved two-stage reconstruction framework of GDL is proposed, which achieves high through-plane porosity distribution fidelity. The method shows good applicability for reconstruction across diverse carbon paper substrates with variations in thickness, porosity distribution, and PTFE content, providing an enhanced tool for precise control and optimal design of the structural and component distributions within GDLs.