Threshold fine-tuning and 3D characterisation of porous media using X-ray nanotomography

A common challenge in the X-ray nanotomography of porous media, such as fuel cell gas diffusion layers (GDLs), is to binarize nanotomography greyscale images in order to differentiate between solids and voids for structural characterisation and numerical flow analysis. In the process threshold determination is critical. This paper presents a study on determination of and fine-tuning threshold value based on comparison of material porosity and average fibre diameter obtained from nanotomography images with porosity data from density experiments and average fibre diameter achieved from scanning electron microscopy images respectively. The more accurate 3D reconstructed model is then used to calculate pore size distribution and average pore size, while the gas permeability of the representative 3D binary images are calculated using a single phase Lattice Boltzmann (LB) model in the D3Q19 regime.