Image-based and Multi-scale Oxygen Transport Modeling of a Polymer Electrolyte Membrane Fuel Cell

<p dir="ltr">Polymer electrolyte membrane fuel cells (PEMFC) are emerging as a promising solution for heavy-duty vehicles, particularly due to their scalability and efficiency over long distances. However, achieving widespread adoption requires significant improvements to PEMFC performance. </p><p dir="ltr">This thesis aims to address recent design limitations by focusing on oxygen transport through advanced electrochemical modeling. Specifically, it introduces novel modeling approaches to analyze both idealized and non-idealized geometries across multiple length scales, providing deeper insights into transport phenomena within the fuel cell structure. In one case, a machine-learning approach is applied to multi-scale modeling to capture the full range of length-scales in a computationally efficient framework. In the second, an image-based approach is applied utilizing X-ray computed tomography imaging to incorporate the true three-dimensional geometries of mass-manufactured fuel cells. This thesis investigates the formation of electrode structures by developing an idealized representative geometry for the carbon-supported catalyst particles. Based on this model, the properties of ionomers, ionomer surface coverage, and ionomer platinum separation are analyzed to enhance oxygen transport. Literature-reported structures of low surface area and high surface area carbon inform the modeling approach, ensuring both physical relevance and scalability. A hybrid framework that combines customizable solvers with tailored code enables the detailed simulation of electrochemical reactions and species transport. To address the high computational cost associated with the extensive design space and operating conditions of PEMFCs, machine learning models are employed to efficiently link the idealized geometry of the catalyst to the full cell scale. This approach not only accelerates simulation workflows but also simplifies the modeling process, improving stability while maintaining high fidelity in capturing the underlying transport-reaction physics. </p><p dir="ltr">In addition to idealized models, this thesis investigates non-idealized geometries through two complementary approaches. First, it introduces simulated geometric defects into idealized structures to study their impact on oxygen transport and overall cell performance. Second, it incorporates realistic flow field geometries derived from radiographs obtained via X-ray microcomputed tomography (micro-CT) of an automotive fuel cell stack. These 3D reconstructions capture structural features, e.g., curvature and thickness variation, not typically represented in idealized models. Once processed, the real-world geometry is integrated into a transport-reaction framework, enabling further electrochemical analysis. </p><p dir="ltr">By conducting simulations for different carbon-support and ionomer binders for typical fuel cell operating conditions, this work reveals key relationships between electrode structures and fuel cell performance. Notably, the idealized porous electrode model suggests that ionomer-related resistance contributes minimally to mass activity losses. In contrast, simulations using an idealized ionomer indicate that high oxygen permeable ionomer (HOPI) significantly enhances mass activity in non-porous electrodes. For a non-idealized electrode model with structural defects, the results show that agglomerate size contributes less to oxygen than uniform ionomer dispersion. Additionally, non-idealized channel geometry demonstrates that convective oxygen flow toward the electrode increases local oxygen concentration, thereby enhancing overall cell performance. These results provide valuable insights into the underlying electrochemical mechanisms and oxygen transport processes within electrodes and flow fields, enabling the evaluation of performance under various PEMFC operating conditions. The findings from this work are expected to support the design of high-performance electrodes and contribute to broader applications of PEMFC technology.</p>