Deep learning for full-feature X-ray microcomputed tomography segmentation of proton electron membrane fuel cells

This study demonstrates the benefit of convolutional neural networks to accurately classify the different materials of proton exchange membrane fuel cells using X-ray micro-computed tomography. Nineteen 2D orthoslices from a 3D tomography dataset were segmented with high quality and used to train a novel UResNet convolutional neural network (CNN) to segment the complete volume. The results were compared with a 3D manual segmentation performed under time constraints. The CNN segmented all phases with equal or greater accuracy in comparison to the manual segmentation. In particular, the CNN excelled in separating the carbon fibres and binder phase in the gas diffusion layer, which is usually completely avoided due to difficulty. Further, permeability calculations were performed on the binder void space for both segmentations, with the CNN displaying realistic results. Therefore, CNNs have been shown to be a viable and valuable method in segmenting such fuel cells with increased efficiency and accuracy. (c) 2022 Elsevier Ltd. All rights reserved.

Automated summary

This study demonstrates the benefits of using convolutional neural networks (CNNs) in accurately classifying different materials of proton exchange membrane fuel cells using X-ray micro-computed tomography. The study shows that a novel UResNet CNN can effectively segment the complete volume of the fuel cells with high accuracy. The CNN outperforms the manual segmentation, especially in separating carbon fibres and binder phase in the gas diffusion layer. Additionally, the CNN provides realistic permeability calculation results for the binder void space.