论文和技术资料

Capturing microstructure details is essential to investigate transport of species and reactions taking place at the micro-scale level in a cathode catalyst layer of a PEM fuel cell. To this end, a three-dimensional geometry of the microstructure is generally created/reconstructed using dedicated software and then imported into COMSOL Multiphysics® software. Due to the complexity of the geometry, the size of the file to be imported can be quite large, and errors can arise during the importing operation. In addition, it usually requires time consuming extra steps to make the microstructure ready to be used in the model under development. By contrast, this work presents a workflow to seamlessly import ready to use microstructures into COMSOL Multiphysics® software, which also allows the user to perform simulations at the voxel level. To this end, a bespoke Python code has been created to read the microstructure from a stack of two-dimensional images and output a mesh file with a mphtxt extension. Each element in the mesh file corresponds to a voxel in the original three-dimensional microstructure. The Python code is optimised to minimise the computation time for large microstructures. The mesh file is then easily imported in COMSOL Multiphysics® software thanks to the import mesh capability. To test the proposed workflow, a simulation of oxygen transport in the void phase of the catalyst layer was simulated. The Transport of Diluted Species interface available in COMSOL Multiphysics® software was used to model the diffusion of oxygen through the porous structure of the representative elementary volume (REV) of the catalyst layer. The diffusion coefficient and the oxygen concentrations on two opposite faces of the REV were provided as input in the interface, together with symmetry boundary conditions on the remaining faces of the REV. The model outputs the oxygen concentration field in the REV, which is easily visualised using a 3D Plot Group in COMSOL Multiphysics® software. The results agreed with the expected oxygen distribution into the void phase, laying a solid foundation for implementing additional physics to capture the complex electrochemistry taking place in the cathode catalyst layer of a PEM fuel cell. Future work will therefore be focused on implementing the solid phase (ionomer, carbon, and catalyst) in the cathode catalyst layer model using the proposed workflow. Finally, the developed approach made it possible to import the microstructure for the void phase of a REV of cathode catalyst layer in two simple steps, involving running the developed Python code and the mesh import command in COMSOL Multiphysics® software.