Two-phase flow and oxygen transport in the perforated gas diffusion layer of proton exchange membrane fuel cell

Abstract

Liquid water transport in perforated gas diffusion layers (GDLs) is numerically investigated using a threedimensional (3D) two-phase volume of fluid (VOF) model and a stochastic reconstruction model of GDL microstructures. Different perforation depths and diameters are investigated, in comparison with the GDL without perforation. It is found that perforation can considerably reduce the liquid water level inside a GDL. The perforation diameter (D = 100 lm) and the depth (H = 100 lm) show pronounced effect. In addition, two different perforation locations, i.e. the GDL center and the liquid water breakthrough point, are investigated. Results show that the latter perforation location works more efficiently. Moreover, the perforation perimeter wettability is studied, and it is found that a hydrophilic region around the perforation further reduces the water saturation. Finally, the oxygen transport in the partially-saturated GDL is studied using an oxygen diffusion model. Results indicate that perforation reduces the oxygen diffusion resistance in GDLs and improves the oxygen concentration at the GDL bottom up to 101% (D = 100 lm and H = 100 lm).

Water transport in the gas diffusion layer of proton exchange membrane fuel cell under vibration conditions

“…The domain dimension remains consistent with the previous study . As such, an 800 × 800 μm cross‐section area with a 192‐μm thick GDL has been reconstructed.…”

“…Here, the average superficial velocity is about 0.022 m s −1 . In the pressure range of 1000 to 6000 Pa, the average velocity varied from 0 to around 0.014 m s −1 , which correspond to the water production rate of fuel cell at a current of 0 to 7000 A cm −2 (normal PEMFC operates up to about 2 A cm −2 ). For the predicted water velocity larger than normal production rate, three important points were discussed to justify the selected range …”

“…In terms of the water transport, the reconstructed geometries act as the flow paths. The GDL is assumed to be formed by a stack of straight fiber layers . As have been mentioned, the stochastic method is adopted to generate the carbon fibers, and it has been introduced and validated in our previous studies .…”

“…25 As have been mentioned, the stochastic method is adopted to generate the carbon fibers, and it has been introduced and validated in our previous studies. 4,6,25 The working principle is using the known properties of a porous media (eg, fibrous diameter in GDL), combining the expected target statistical parameter (porosity, layer thickness), based on the stochastic method to generate a porous structure. Commonly, the carbon fiber is considered to be a long straight cylinder and randomly distributed in a plane.…”

“…The fiber diameter is chosen to be 8 μm, which is consistent with previous numerical and experimental studies. 25,26 The porosity of the present GDL follows the Toray carbon paper TGP-H-060, which changes in each layer and has been tested in the X-ray experiments. 27 The domain dimension remains consistent with the previous study.…”