Elias Baltic

Water Content Measurement of Gas Diffusion Media and Membranes by Neutron Radiography

Neutron radiography is an excellent tool for measuring water in metallic housings; in stable conditions measurement sensitivity down to 4 μm water thickness in a 127 μm pixel has been demonstrated (1). Recent advances in detector technology have produced neutron imaging detectors with spatial resolution approaching or less than 10 μm, enabling direct measurement of the water content in gas diffusion media (GDM) and membranes with a through-plane thickness greater than 100 μm. This improvement in spatial resolution is demonstrated in Figure 1 for a proton exchange membrane fuel cell (PEMFC) operated nominal under the same conditions. This article will focus on the performance of high resolution neutron imaging detectors and image analysis methods for fuel cell research.

Improving material identification by combining x-ray and neutron tomography

X-rays and neutrons provide complementary non-destructive probes for the analysis of structure and chemical composition of materials. Contrast differences between the modes arise due to the differences in interaction with matter. Due to the high sensitivity to hydrogen, neutrons excel at separating liquid water or hydrogenous phases from the underlying structure while X-rays resolve the solid structure. Many samples of interest, such as fluid flow in porous materials or curing concrete, are stochastic or slowly changing with time which makes analysis of sequential imaging with X-rays and neutrons difficult as the sample may change between scans. To alleviate this issue, NIST has developed a system for simultaneous X-ray and neutron tomography by orienting a 90 keVpeak micro-focus X-ray tube orthogonally to a thermal neutron beam. This system allows for non-destructive, multimodal tomography of dynamic or stochastic samples while penetrating through sample environment equipment such as pressure and flow vessels. Current efforts are underway to develop methods for 2D histogram based segmentation of reconstructed volumes. By leveraging the contrast differences between X-rays and neutrons, greater histogram peak separation can occur in 2D vs 1D enabling improved material identification.