X-ray tomography dataset of gas diffusion layers with different coating percentages along with their corresponding segmented images

• https://doi.org/10.1021/acsaem.2c00023
• Download

Abstract — We inject water at a low flow rate through gas diffusion layers containing different percentages of polytetrafluoroethylene (PTFE) coating: 5, 20, 40, and 60%. We use high-resolution three-dimensional X-ray imaging to identify the arrangement of fibers, water, and air in the pore space. We also quantify the contact angle and meniscus curvature once the water has spanned the layer, flow has ceased, and water has reached a position of equilibrium. The average contact angle and water pressure at breakthrough increase with the amount of coating, although we see a wide range of contact angles with values both above and below 90°, indicating a mixed-wet state. We identify that the menisci form minimal surfaces (interfaces of zero curvature) consistent with pinned gas-water-solid contacts. Scanning electron microscopy images of the fibers show that the coated fibers have a rough surface. Between 93 and 100% of the contacts identified were found on the rough, hydrophobic, coated fibers or at the boundary between uncoated (hydrophilic) and coated (hydrophobic) regions; we hypothesize that these contacts are pinned. The one exception is the 60% PTFE layer, which shows distinctly hydrophobic properties and a negative capillary pressure (the water pressure is higher than that of air). The presence of minimal surfaces suggests that the water and gas pressures are equal, allowing water to flow readily without pressure build-up. From topological principles, the negative Gaussian curvature of the menisci implies that the fluid phases are well connected. The implication of these results is explored for the design of porous materials where the simultaneous flow of two phases occurs over a wide saturation range.

Abstract — Wettability design is of crucial importance for the optimization of multiphase flow behaviour in gas diffusion layers (GDLs) in fuel cells. The accumulation of electrochemically-generated water in the GDL will impact fuel cell performance. Hence, it is necessary to understand multiphase displacement to design optimal pore structures and wettability to allow the rapid flow of gases and water in GDLs over a wide saturation range. This work uses high-resolution three-dimensional X-ray imaging combined with a pore network model to investigate the breakthrough capillary pressure and water saturation in gas diffusion layers manufactured with different degrees of polytetrafluoroethylene coating: 5, 20, 40, and 60%, making them more hydrophobic. We first demonstrate that the pore network extraction method provides representative networks for the fibrous porous media examined. Then, using a pore-network flow model we simulate water flooding in initially gas-filled fibrous media, and analyse the effect of wettability on breakthrough capillary pressure and water saturation. With an appropriate pore-scale characterization of wettability, a pore network model can match experimental results and predict displacement behaviour.

• https://doi.org/10.1016/j.fuel.2023.128180
• Download

Abstract — High-resolution X-ray computed tomography (micro-CT) has been widely used to characterise fluid flow in porous media for different applications, including in gas diffusion layers (GDLs) in fuel cells. In this study, we examine the performance of 2D and 3D U-Net deep learning models for multiphase segmentation of unfiltered X-ray tomograms of GDLs with different percentages of hydrophobic polytetrafluoroethylene (PTFE). The data is obtained by micro-CT imaging of GDLs after brine injection. We train deep learning models on base-case data prepared by the 3D Weka segmentation method and test them on the unfiltered unseen datasets. Our assessments highlight the effectiveness of the 2D and 3D U-Net models with test IoU values of 0.901 and 0.916 and f1-scores of 0.947 and 0.954, respectively. Most importantly, the U-Net models outperform conventional 3D trainable Weka and watershed segmentation based on various visual examinations. Lastly, flow simulation studies reveal segmentation errors associated with trainable Weka and watershed segmentation lead to significant errors in the calculated porous media properties, such as absolute permeability. Our findings show 43, 14, 14, and 3.9% deviations in computed permeabilities for GDLs coated by 5, 20, 40, and 60 w% of PTFE, respectively, compared to images segmented by the 3D Weka segmentation method.