Proton exchange membrane fuel cells, consuming hydrogen and oxygen to generate clean electricity and water, suffer acute liquid water challenges. Accurate liquid water modelling is inherently challenging due to the multi-phase, multi-component, reactive dynamics within multi-scale, multi-layered porous media. In addition, currently inadequate imaging and modelling capabilities are limiting simulations to small areas (<1 mm2) or simplified architectures. Herein, an advancement in water modelling is achieved using X-ray micro-computed tomography, deep learned super-resolution, multi-label segmentation, and direct multi-phase simulation. The resulting image is the most resolved domain (16 mm2with 700 nm voxel resolution) and the largest direct multi-phase flow simulation of a fuel cell. This generalisable approach unveils multi-scale water clustering and transport mechanisms over large dry and flooded areas in the gas diffusion layer and flow fields, paving the way for next generation proton exchange membrane fuel cells with optimised structures and wettabilities.
Pore‐scale digital images are usually obtained from microcomputed tomography data that has been segmented into void and grain space. Image segmentation is a crucial step in the process of digital rock analysis that can influence pore‐scale characterization studies and/or the numerical simulation of petrophysical properties. This is concerning since all segmentation methods have user‐selected parameters that result in biases. Convolutional neural networks (CNNs) provide a way forward since once trained, CNN can provide consistent and reliable image segmentation with no user‐defined inputs. In this paper, a CNN is used to segment digital sandstone data, and various ground truth data sets are tested. The ground truth images are created based on high‐resolution microcomputed tomography data and corresponding scanning electron microscope data. The results are evaluated in terms of porosity, permeability, and pore size distribution computed from the segmented data. We find that watershed‐based segmentation provides a wide range of possible petrophysical values depending on user‐selected thresholds, whereas CNN provides a smaller variance when trained on scanning electron microscope data. It can be concluded that CNN offers a reliable and consistent way to segment digital sandstone data for petrophysical analyses.