Title: Functionally graded porous electrode design of proton exchange membrane fuel cells via mechanistic and data-driven hybrid modelling approach
Time: 2023-10-25 19:00 - 20:00
Venue: Institute for Energy Research, 1517
Organizer: Institute for Energy Research
Abstract: Performance, durability, and cost are the trilemma for the commercialisation of proton exchange membrane (PEM) fuel cells. The use of Pt-based alloys as catalysts for the oxygen reduction reaction (ORR) and the nonuniform distribution of current density inside a membrane electrode assembly (MEA) result in high cost and low durability, which strongly hinders the wide adoption of PEMFCs in industrial and civil applications. For PEM fuel cells operated at various loads, the required activities and mass transport rates are different because the reactant and product are nonuniformly distributed inside the MEA. Thus, a rational design for an MEA, especially the porous electrode, with a spatial distribution of functional components is helpful for reducing the usage of precious components, improving cell performance, and achieving uniform distributions of current density and heat. To reduce the cost and improve the current homogeneity without sacrificing the cell performance, we studied the graded design of the functional components within the gas diffusion layer (GDL) and catalyst layer (CL), e.g., Pt loading, ionomer loading, electrode porosity, etc., along both the through-plane and in-plane directions and the spatial variation of operating parameters, i.e., temperature and pressure, as well as the interaction of multiple design variables. We developed the physics-based sophisticated model first, which is then used to generate the database for data-driven modelling after experimental validation. Therefore, physics-based simulation and machine learning (ML) based surrogate modelling are integrated to build a sophisticated hybrid model, in which multi-physics and multi-phase flow simulation, ML-based surrogate modelling, multi-variable and multi-objects optimisation are combined. The novel approach has proved to be effective and efficient for the functionally graded design of the porous electrode of PEM fuel cells.
