Pore network modelling has traditionally been used to study displacement processes in idealized porous media related to geological flows, with applications ranging from groundwater hydrology to enhanced oil recovery. Very recently, pore network modelling has been applied to model the gas diffusion layer (GDL) of a polymer electrolyte membrane (PEM) fuel cell. Discrete pore network models have the potential to elucidate transport phenomena in the GDL with high computational efficiency, in contrast to continuum or molecular dynamics modelling that require extensive computational resources. However, the challenge in studying the GDL with pore network modelling lies in defining the network parameters that accurately describe the porous media as well as the conditions of fluid invasion that represent realistic transport processes. In this work, we discuss the first stage of developing and validating a GDL-representative pore network model. We begin with a two-dimensional pore network model with a single mobile phase invading a hydrophobic media, whereby the slow capillary dominated flow process follows invasion percolation. Pore network geometries are designed, and transparent hydrophobic microfluidic networks are fabricated from silicon elastomer PDMS using a soft lithography technique. These microfluidic networks are designed to have channel size distributions and wettability properties of typical GDL materials. Comparisons between the numerical and experimental flow patterns show reasonable agreement. Furthermore, the fractal dimension and saturation are measured during invasion, revealing different operating regimes that can be applied to GDL operation. Future work for model development will also be discussed.
Fluid Meniscus Algorithms for Dynamic Pore-Network Modeling of Immiscible Two-Phase Flow in Porous Media
