The document describes an experimental setup to measure water pressure-water saturation curves for gas diffusion layers (GDLs) using X-ray tomography. The experimental results show default pressure-saturation curves differ from measurements of specific GDLs, and adding a microporous layer decreases water saturation at a given pressure. Computational fluid dynamics simulations using a Leverett function with default parameters versus parameters fitted to the specific GDL measurements show significant differences in predicted water saturation distributions.

1. JHZSWJuni14
An Experimental setup has been developed
to measure water pressure–water saturation
curves based on X-ray tomography. With this
setup parameters for GDLs can be
calculated for an adapted Leverett-function.
J. Haußmann*1), C. Bergbreiter1), M. Ender1), H. Markötter2), T. Arlt2), I. Manke2), J. Scholta1)
1) Zentrum für Sonnenenergie- und Wasserstoff-Forschung (ZSW), Ulm, Germany
2) Helmholtz-Zentrum Berlin für Materialien und Energie (HZB), Berlin, Germany
*email: jan.haussmann@zsw-bw.de, Tel.: +49 731 9530-209
Experimental Method
Conclusions
Introduction
The authors like to thank H. Riesemeier (BAM Federal Institute for Material Research and Testing). Funding of the project GECKO
(grant number: 03SF0454D ) by the Federal Ministry of Education and Research (BMBF) is gratefully acknowledged.
1. Water pressure is applied to GDL and
MPL to simulate water production at
catalyst layer
2. Water pressure is stepwise increased
until saturation is stable
3. Tomography is conducted for each step
4. Water saturation is calculated with
respect to porosity of dry GDL / MPL:
Optimizing the liquid water transport in
porous media of a PEM fuel cell is of high
interest. Especially the control of the water
saturation within GDL and MPL is a big
issue. Thereby, the porous materials have to
accomplish two oppositional tasks:
1. High membrane hydration
 High proton conductivity
2. Low GDL water saturation
 High gas permeability
For optimization of the GDL water transport
CFD simulations are conducted using water
saturation analysis.
CFD simulation settings
For the analyzed GDL types SGL 25 BA and
25 BC a strong difference to the default
parameters is found. Besides, an additional
MPL causes a water pressure barrier. It
results in a lower water saturation of the
GDL, if the same pressure is applied.
The CFD simulations show a drastic change
in water saturation, if calculated with specific
GDL parameters. Especially the saturation
under the land increases almost five times.
Hence, it is suggested to measure Leverett-
parameters specifically for each kind of GDL.
Improvement of CFD simulations
with respect to a GDL specific Leverett-function
References
[1] M. C. Leverett, Capillary behavior in porous solids, A.I.M.E. Transactions 142 (1941) 152–168.
[2] K. S. Udell, Heat transfer in porous media considering phase change and capillarity - the heat pipe effect, International Journal of Heat and Mass Transfer 28 (2) (1985) 485–495.
ANSYS® FLUENT® Fuel Cell Module is used
to predict the water saturation of a SGL
25 BC GDL with MPL on the cathode side.
Simulation is set with the following
Operating conditions
- Temperature: 70 °𝐶
- Relative Humidity: 100 %
Material Parameters SGL 25 BC
- Compressed porosity: 68 %
- Contact angle: 92°
- Water permeability: 2.55 ∙ 10−13
𝑚2
- Surface tension: 0.0644 𝑁𝑚−2
Fig. 3: Predicted water saturation of GDL SGL 25BC with
default parameters and adaption according to experiment
Acknowledgements
Fig. 1: Experimental Setup for measuring the water
saturation of the GDL with synchtron X-ray tomography
Fig. 2: Water pressure – water saturation curves for
SGL 25 BA and 25 BC as well as with default parameters [2]
Experimental Setup
- =
Experimental Results
- Default pressure – saturation curve
differs strongly from measured values for
SGL 25 BA and 25 BC
- Applying an MPL on the GDL substrate
results in a strong decrease of water
saturation in matters of pressure
- New parameters for GDLs 25BA & 25BC:
Parameter Default 25 BA 25 BC
a 1.417 0.052 0.22
b 2.12 0.019 0.11
c 1.263 0.0022 0.017
CFD simulation results
Dry
GDL
GDL &
Water
Water
The two simulation cases show a strong
difference in water saturation:
- Default case shows always a too low
water saturation, especially under the
land
- Case with new 25 BC parameters reveals
a much higher difference in saturation
between channel and land area
Average water saturations are calculated:
Case Channel area Land area
Default 4.9 % 8.5 %
SGL 25 BC 6.3 % 41.6 %
An experimental setup is developed to
determine parameters by measuring water
pressure – water saturation curves (Fig. 1).
In ANSYS® FLUENT® CFD simulation software
the liquid water transport is described by a
function developed by Leverett [1]:
𝑝𝑐 = 𝜎 𝑐𝑜𝑠𝜃
𝐾
𝜀
𝐽 𝑠
𝑝: 𝑐𝑎𝑝𝑖𝑙𝑙𝑎𝑟𝑦 𝑝𝑟𝑒𝑠𝑠𝑢𝑟𝑒, 𝜎: 𝑠𝑢𝑟𝑓𝑎𝑐𝑒 𝑇𝑒𝑛𝑠𝑖𝑜𝑛
𝐾: 𝑤𝑎𝑡𝑒𝑟 𝑝𝑒𝑟𝑚𝑒𝑎𝑏𝑖𝑙𝑖𝑡𝑦, 𝜀: 𝑝𝑜𝑟𝑜𝑠𝑖𝑡𝑦
𝐽 𝑠 : 𝐿𝑒𝑣𝑒𝑟𝑒𝑡𝑡 − 𝑓𝑢𝑛𝑐𝑡𝑖𝑜𝑛, 𝑠: 𝑤𝑎𝑡𝑒𝑟 𝑠𝑎𝑡𝑢𝑟𝑎𝑡𝑖𝑜𝑛
If contact angle: 𝜃 < 90° (hydrophilic)
𝐽 𝑠 = 𝑎 1 − 𝑠 − 𝑏 1 − 𝑠 2
+ 𝑐 1 − 𝑠 3
𝑤𝑖𝑡ℎ 𝑎 = 1.417, 𝑏 = 2.12, 𝑐 = 1.263 [2]
Parameters are expected to vary depending
on the type of GDL and MPL (if applied).