Stochastic Microstructural Modeling of PEFC Gas Diffusion Media

Abstract

A novel microstructural modeling approach was developed to stochastically generate a realistic 3D structure of carbon fiber paper based gas diffusion media used in polymer electrolyte fuel cells (PEFCs). The in-house developed algorithm features a stochastic structure generator that allows creating full-length non-overlapping fibers, carbonized binder, and PTFE as distinct phases according to the carbon paper manufacturing process. The resulting 3D structure consists of a high-aspect ratio paper sheet which is large enough to contain an accurate representation of the variability expected from the manufacturing process, as confirmed by thorough validation with experimental data. Three characteristic subdomains with low, average, and high porosity were selected from the sheet to analyze the distribution of transport properties and determine correlations to structural features such as porosity, fiber alignment, and orientation. The range of simulated transport properties were found to be in good agreement with experimental data and provides a practical data set including bulk average and statistical variation for each property. The stochastic modeling framework reported herein is based entirely on in-house developed algorithms that can be applied as a design tool for development and analysis of new gas diffusion materials for fuel cells.