Proceedingsofofthe
theASME
Seventh
International
ASMEConference
Conferenceon
onNanochannels,
Nanochannels,Microchannels
Microchannelsand
andMinichannels
Minichannels
Proceedings
2009
7th International
ICNMM2009
ICNMM2009
June22-24,
22-24,2009,
2009,Pohang,
Pohang,South
SouthKorea
Korea
June
ICNMM2009-82140
IN SITU CHARACTERIZATION OF TWO-PHASE FLOW IN CATHODE CHANNELS OF AN
OPERATING PEM FUEL CELL WITH VISUAL ACCESS
Jacqueline M. Sergi*, Zijie Lu, Satish G. Kandlikar
Mechanical Engineering Department
Kate Gleason College of Engineering
Rochester Institute of Technology, Rochester, NY USA
*[email protected]
ABSTRACT
Water management is a critical factor in the performance
and durability of a proton exchange membrane (PEM) fuel cell.
In situ experiments are needed to gain a better understanding of
water transport within the channels of the cell during operation.
In this work a 50 cm2 fuel cell with optical access is designed
and tested in an in situ experimental facility. Two-phase flow
in the cathode channels of the cell is observed, and flow
patterns are characterized. Three primary two-phase structures
are identified – slug flow, film flow, and mist flow – and a flow
pattern map is developed. A comparison between in situ and
ex-situ flow pattern maps shows that ex-situ experimentation
can be used to predict some in situ flow characteristics, but
cannot capture the effects of reaction kinetics or relative
humidity. The total pressure drop signature is seen to be a
useful parameter for predicting two-phase flow dynamics in the
gas channels. In addition, channel to channel flow variation
caused by the presence of liquid water in the cathode channels
is investigated using entrance region pressure drop
measurements.
NOMENCLATURE
PEMFC – Proton Exchange Membrane Fuel Cell
PEM – Proton Exchange Membrane
GDL – Gas Diffusion Layer
DOE – Department of Energy
CCM – Catalyst Coated Membrane
RH – Relative Humidity
1. INTRODUCTION
Water management within a proton exchange membrane
fuel cell (PEMFC) is regarded as a critical parameter affecting
fuel cell performance and longevity [1]. Without sufficient
hydration, proton conductivity cannot be maintained within the
proton exchange membrane and the ionomers in catalyst layers,
but an excess amount of water can lead to flooding of the cell
and result in reduced reactant flow and increased mass transport
losses. As a result of electrochemical reactions and humidified
inlet gases, two-phase flow is typically present in the gas
channels of the fuel cell. Ex-situ studies cannot capture the
inherent cell dynamics present in situ, and there exists a need to
establish a fundamental understanding of the two-phase flow
dynamics within the channels of an operating fuel cell.
Due to the complexity of a fuel cell, there are limited
experimental techniques available for characterizing this twophase flow.
Two practical techniques that have been
investigated are neutron radiography and optical imaging.
Neutron radiography allows for non-invasive visualization of
water transport within the cell, and has been used by several
groups to visualize and quantify liquid water presence in cell
channels, within the GDL, and underneath land area [2-4].
There are some technical challenges that limit the use of
neutron radiography, including its two-dimensional nature
which can make it difficult to distinguish cathode side water
from anode side water, current temporal resolution (up to 30
Hz) which is not capable of resolving water transport dynamics,
and high cost and limited number of facilities. Recent work
using high resolution neutron radiography [5, 6] shows some
promising advances in this field.
Optical visualization requires the development of a fuel
cell with transparent components to make viewing the internal
channels possible. This is a relatively low cost alternative to
neutron radiography, and makes it a practical option for
investigating in situ flow characteristics. In addition, optical
visualization offers a higher spatial and temporal resolution.
Tuber et al. [7] was the first group to utilize this method to
study the buildup of water in a cathode channel at low
temperatures. This method was later applied across a wide
variety of fuel cell conditions to investigate water transport
within the gas channels [8-15]. The simultaneous use of optical
visualization and neutron radiography has recently been
explored by Spernjak et al. [16]
Flooding can be a serious detriment to PEMFC
performance, and can occur across a wide range of current
densities. Typically the cathode side is studied due to the water
1
Copyright © 2009 by ASME
production at the cathode electrode, but the anode side is also
prone to flooding [3, 4]. Pressure drop measurement has
recently emerged as a diagnostic tool for fuel cell water
management, and has been proposed as a method to detect
flooding [17-20]. Liu et al. [13, 15] investigated PEMFC
performance and its relationship to pressure drop and water
transport dynamics within a cell. Although Pressure drop
measurements across a cell can be useful in assessing liquid
water presence, they cannot capture the highly localized twophase flow dynamics present in the gas channels of a PEM fuel
cell.
The instantaneous measurement of flow rates in
individual cathode channels can provide a more sensitive
insight to liquid water holdup within the channels and its
effects on flow maldistribution, which can directly affect fuel
cell performance. Recently, a method for measuring channel
flow rates using entrance region pressure drop was developed
by Kandlikar et al. [21] and applied to ex-situ experimental
work investigating two phase flow in fuel cell gas channels
[22].
This work focuses on experimental investigation of the two
phase flow characteristics within the cathode channels of an
operating 50 cm2 fuel cell. The pressure drop, instantaneous
channel flow rates and two-phase flow structures are studied in
combination with high speed optical visualization. Results
obtained from these in situ experiments provide further insight
into the relationships between pressure drop, flow structure,
and parallel channel flow maldistribution in PEM fuel cells.
2. EXPERIMENTAL DESIGN
2.1 Test Cell and In Situ Facility
2
A transparent 50 cm test cell was developed for in-situ
experiments. The test section is comprised of anode and
cathode flow fields which sandwich the gas diffusion layers
(GDLs) and catalyst coated membrane (CCM). The flow fields
are made of gold plated copper plates, and also function as the
current collectors. Copper was chosen for its high electrical
conductivity and receptiveness to gold plating, which was
applied to minimize corrosion. The anode and cathode gas
channels are machined through the copper plates which have a
thickness of 0.4 mm (channel depth). The rectangular cathode
channels are 183 mm long, 0.7 mm wide, with 0.5 mm spacing
between adjacent channels. Anode channel geometry is
identical except for a width of 1.5 mm between adjacent
channels. A 50 cm2 active area contains 22 such cathode
channels and 11 anode channels.
In order to avoid
misalignment effects and mechanical shearing of the GDL
caused by straight channels, a wavy channel pattern was
implemented, as seen in Figure 1. The channel geometries and
dimensions were designed by Owejan et al. [23] in order to
meet DOE targets for volumetric power density [24].
In order to measure the instantaneous flow distribution in
each parallel channel, the entrance region pressure drop method
developed by Kandlikar et al. [21] is used. This method
involves using the pressure drop measurement in the entrance
region of each individual channel (where single phase flow is
present) to determine the flow rates in each channel. A straight
runner header design was used, and pressure taps were placed
in the Lexan® support piece corresponding with each channel’s
straight entrance region. The holes were staggered in three
rows to allow clearance for attachment of pressure taps and
tubing to the pressure transducers. The entrance region is
before the introduction of product water (due to
electrochemical reaction) so that single-phase gas flow is
assured in this region. Details of the cathode header region are
shown in Figure 2. The total pressure drops across the anode
and cathode channels are also measured with pressure taps
placed at the respective inlet and outlet headers. Dowel pin
holes were included for proper alignment of each component
during assembly.
(a)
0.5 mm
0.7 mm
0.4 mm
(b)
Fig. 1: a) Section of wavy cathode flow field, channels
designed with 11° angular channel switchback every 5 cm
[23], b) cathode channel profile and dimensions. &ot to
scale.
The assembled test section comprises of a catalyst coated
membrane (CCM) sandwiched by gas diffusion layers, anode
and cathode flow fields and the Lexan® pieces, with appropriate
Teflon® gaskets. The CCM was fabricated by W.L. Gore, Inc.
on an 18 µm thick perfluorosulfonic acid (PFSA) membrane,
with a platinum loading of 0.2/0.3 mg/cm2 for anode and
cathode, respectively. The gas diffusion layer used in this work
was made of Mitsubishi MRC-10 carbon paper, and was wetproofed and coated with a thin micro-porous layer by General
Motors. This GDL is considered to be the Baseline sample for
in situ testing. The Baseline GDL is a bi-layer structure and is
PTFE treated, which makes it hydrophobic and aids in water
removal. The GDL has a thickness of approximately 230 µm.
An exploded view of the test section assembly can be seen in
Figure 3.
a
b
c
d
Fig. 2: Detailed view of cathode header region. a) header
cavity in Lexan® support piece, b) pressure tap bores in
straight channel section, c) 22 cathode channels formed by
gold plated copper current collectors, d) o-ring groove
2
Copyright © 2009 by ASME
Due to very small channel dimensions of the cathode
current collector and lack of structural support along the
channel length, the flow fields are carefully assembled to
ensure channel width uniformity. The test section is housed on
both sides by machined 6061 aluminum blocks which contain
the inlets and outlets for air and hydrogen and are also used for
cell compression. Each block possesses a series of five cavities
for viewing access that can be sealed with an additional Lexan®
window for temperature control. O-ring gaskets were used for
sealing between the aluminum blocks and test section. The cell
assembly was compressed to 200 psi, and pressure sensitive
film in place of the soft goods was used to verify even
compression distribution within the assembly prior to testing.
1
2
3
4
5
6
Fig. 3: Exploded view of test section assembly. 1) Lexan®
support piece, 2) o-ring groove, 3) current collector/flow
fields, 4) PTFE gasket, 5) GDL 6) CCM
The fuel cell is tested with a Hydrogenics G40 fuel cell test
station under current control mode. Air and hydrogen are
supplied to the cell through the test station from a Parker
Balston Zero Air Generator and H2-1200 Hydrogen generator.
A Siemens water de-ionization system provides water for the
test set-up. The temperature of the fuel cell is varied using
Watlow flexible silicone heaters on the surface of the aluminum
blocks housing the test section. The total pressure drop and
individual channel pressure drop in the entrance region are
measured using Honeywell FDW2AT differential pressure
transducers with an accuracy of ± 0.251%. Pressure data was
collected and recorded through a National Instruments NI-9205
DAQ with LabVIEW software.
For visualization of the fuel cell microchannels, a Photron
Fastcam 1024-PCI high speed camera was used, with a pixel
resolution of 1024x1024 for a frame rate range of 60-1,000 fps.
Primary visualization was performed using a Nikon 105 mm
AF Micro Nikko lens, which allows for simultaneous
observation of all 22 cathode channels. An Infinity K2/STM
long distance microscope lens was used for additional highresolution visualization. The camera was mounted to a Velmex
motorized 3-axis stage with positional repeatability of 5
microns. Fiber optic lighting was used to illuminate the cell
viewing windows. The cell assembly, pressure transducers, and
visualization equipment were mounted to a Newport ST-UT2
vibration isolation table.
A schematic of the in situ
experimental test set-up is shown in Figure 4.
Fig. 4: Schematic for in situ experimental set-up. &ot to
scale.
2.2 Experimental Procedure
Initial testing was performed with the cell mounted in a
vertical down orientation. Previous research has shown the
effect of gravity on the formation and detachment of water
droplets on the surface of the GDL to be insignificant [25],
however, gravity could be a contributing factor for some
mechanisms of water transport within the channels, especially
for a large amount of water. Different cell orientations are
being considered for future research. The cell temperature is
controllable in the range of 35 – 90 °C. This work focuses on
experiments performed at a cell temperature of 35 °C, and
effects of increased temperature are currently being
investigated. Stoichiometric (stoich) ratios (an:ca) of 1.5:2.5,
1.5:5, and 3:8 were tested at each temperature, with both dry
and 100% RH inlet gases. For each test, current density was
increased in increments of 0.05 A/cm2 from 0.05 A/cm2 to 0.4
A/cm2, and in increments of 0.1 A/cm2 for each additional test
point until cell failure. For each test point the total cathode
pressure drop and individual channel pressure drops were
recorded, with simultaneous visualization of two-phase flow
within the channels.
3. RESULTS AND DISCUSSION
3.1 Visualization of Two-Phase Flow
The two-phase flow structure (flow pattern) within the
cathode channels of an operating 50 cm2 fuel cell was studied
using high-speed imaging techniques. In this work only the
cathode channels are studied, and visualization of the anode
side is planned for future work. Two primary modes for liquid
water entering the gas channels were identified. The first mode
is the emergence of droplets through the gas diffusion layer
pores and into the channels. Product water is also rejected from
the catalyst layer in the form of water vapor, especially under
the dry inlet gas conditions, which is demonstrated by the
condensation in the gas channels at conditions with dry inlet
gases. When humidified gas streams are used, water
condensation also occurs from the humidified inlet gases.
By observing the water transport mechanisms within the
cathode channels, three key two-phase flow patterns – slug
flow, film flow, and mist flow – were identified. Similar flow
3
Copyright © 2009 by ASME
structures were reported by Lu et al. [22] in an ex-situ twophase flow study with identical cathode flow field geometry
and similar flow conditions.
Slugs
3.1.1 Droplet Formation
The formation of droplets on the surface of the gas
diffusion layer is a result of GDL hydrophobicity. By using a
long distance microscope lens it was possible to observe water
droplet emergence, growth and detachment on the GDL
surface. An image of water droplets on the GDL surface is
shown in Figure 5.
Fig. 6: Slug flow present in three cathode channels.
Baseline GDL, 1.5:2.5 (an:ca) stoich ratio, 0.1 A/cm2, 35°C
cell temp., 100% RH inlet gases. &ote the presence of
condensation on the channel surface due to humidified inlet
gases.
Fig. 5: Droplet formation on GDL surface observed with
long distance microscopic lens. Baseline GDL, 1.5:2 (an:ca)
stoich ratio, 25°C cell temp., 0.74V.
3.1.2 Slug Flow
Slug flow was characterized by a water column bridging
both sides of a gas channel. Although the slugs can appear to
be blocking the entire channel, flow in the channels may still be
present, indicating a semi-slug [26]. There are several factors
that can contribute to this. The difference in surface properties
between the GDL (which has a contact angle, θ, of
approximately 140°) and the gold-coated copper plate (θ ∼ 20°)
as well as Lexan® (θ ∼ 60°) may prevent the complete contact
of water with the entire GDL surface. In addition, gravity
could pull down large slugs before they fill the entire channel,
hindering the formation of fully developed slugs in this
experiment. Flow can also be present in channels containing
slugs due to gas flow across the GDL from adjacent channels.
Slug flow was observed to be the primary water transport
mechanism in channels at conditions of low air flow rate (lower
stoich ratio and current density). Figure 6 shows three slugs
that formed in the cathode gas channels.
Slug formation and removal is considered to be a
significant contributor to large fluctuations in the total pressure
drop signature and flow maldistribution among the parallel
channels due to larger quantities of liquid water and tendency
to restrict or block the flow of reactants in a channel.
Maldistribution caused by slug flow is discussed further in
section 3.4.
3.1.3 Film Flow
In film flow, liquid water was present on the channel walls,
but did not bridge the entire channel width. Film flow moved
through the channel in a variety of ways. It was typically
driven by air flow and maintained contact with one wall, which
lead to the development of a larger leading edge and thinner
trailing edge, as seen in Figure 7. Other mechanisms of
transport included the formation of slug flow from film flow,
which was caused by the film thickness growing to the size of
the channel width, and by means of intersection with another
slug/film from upstream flow. Since film flow does not block
the channels, the fluctuations in total pressure drop signature
associated with film flow were not as large as slug flow. Film
flow was usually present at intermediate air flow rates (stoich
ratios).
Trailing
Edge
Leading
Edge
Fig. 7: Film flow in multiple cathode channels. Baseline
GDL, 3:8 (an:ca) stoich ratio, 0.1 A/cm2, 35°C cell temp.,
100% RH inlet gases.
4
Copyright © 2009 by ASME
3.1.4 Mist Flow
Mist flow can be described as a water transport mechanism
in which very small liquid droplets are traveling in air at a high
velocity, rendering them difficult to be observed visually. The
water may also be completely evaporated. An image of mist
flow occurring in all 22 cathode channels is shown below in
Figure 8. Mist flow was most common at the highest stoich
ratio. It was also present at the intermediate stoich ratio at low
current densities, due to decreased water production. Mist flow
usually did not cause significant pressure drop fluctuations.
experiments, liquid water is injected into the test section at
desired flow rates (corresponding to the appropriate water
production rates by assuming that all water is removed through
the cathode channels). Because the in situ experiments were
conducted with an operational fuel cell, there are several
parameters affecting the flow pattern map that cannot be
captured with ex-situ simulations. This includes the ratio of
reactants supplied to the cell, the resulting product water and
heat generated by chemical reaction, and the effects of an actual
load being applied to the cell. Even with the vast differences in
factors affecting in situ and ex situ maps, the transition trends
still proved to be very similar. In both cases slug flow pattern is
the primary water transport mechanism at lower air flow rates,
followed by film flow pattern at intermediate and mist flow
pattern at the higher air flow rates.
Fig. 8: Mist flow. Baseline GDL at 3:8 (an:ca) stoich ratio,
0.4 A/cm2, 35°C cell temp., 100% RH inlet gases.
The primary two-phase flow structures that were identified
through in situ visualization using optical imaging techniques
were also identified in ex-situ experiments [22]. Figures and
two phase flow characteristcs for slug, film, and mist flow
presented in that work closely resemble in situ findings.
3.2 Flow Pattern Maps
The in situ behavior of channel air flow rate coupled with
optical visualization was used to identify two-phase flow
structures within the cathode channels of an operating fuel cell.
Flow pattern maps are a unique representation of the various
water transport methods present at certain test conditions. The
current density and air flow rates (logarithmic scale)
corresponding to the various stoich ratios tested were plotted
along with the flow characteristics present at each in situ test
condition. The resulting flow pattern map is shown in Figure 9.
This plot is useful in determining the conditions at which
different flow patterns are likely to emerge. As can be seen in
Figure 9, slug flow tends to be the primary water transport
mechanism at lower air flow rates. As the air flow rate
increases, the flow regime transitions to film at intermediate air
flow rates and ultimately mist flow at higher air flow rates. The
term “dominant” in the flow pattern map indicates that
additional flow regimes may have been present at these
conditions, but one was observed to be the most commonly
occurring transport mechanism. Artificial lines were added to
the flow pattern maps to aid in seeing the transition between
each two-phase flow structure.
For comparison, the ex-situ flow pattern map, which is
obtained with identical channel design and at similar flow
conditions, is also included as shown in Figure 10. In ex-situ
Fig. 9: In situ map of two-phase flow patterns as a function
of air flow rate and current density, 35°C cell temp., dry
gas.
Fig. 10: Ex-situ map of two-phase flow patterns as a
function of air flow rate and current density. &ote that
current density was not actually applied in ex-situ tests, but
calculated from amount of water artificially introduced into
the cell.
5
Copyright © 2009 by ASME
One example of a strong correlation between the two maps is
seen at a current density of 0.20 A/cm2. At this condition, each
in situ flow pattern identified at its respective air flow rate
correlates to the same flow pattern in the ex-situ map. This
points to the validation of using ex-situ experiments to predict
two-phase flow structure occurring at in-situ conditions.
3.3 Pressure Drop Measurements
The in situ experimental set-up presents the unique
capability of simultaneously measuring the total cathode
pressure drop and the individual channel pressure drops during
cell operation, which can then be used to calculate individual
channel flow rates. Previous experiments from Liu et al. [13]
and Ma et al. [18] have demonstrated that the total pressure
drop across the gas channels is function of the amount of liquid
water present in the channels, and increases with greater
amounts of liquid water. This increase in water can lead to
channel flooding and ultimately decrease fuel cell performance.
Analyzing the total pressure drop signature can thus be a useful
diagnostic tool for assessing the amount of liquid water present
in the channels, and ultimately the flow regime(s) present each
test condition.
The pressure drop across all cathode gas channels is
proving to be an effective means to monitor the two phase flow
during cell operation. The presence of liquid water in the
channels is a contributor to fluctuations in the overall pressure
drop. Figure 11 shows a comparison between total pressure
drop measurements taken at three different stoich ratios with a
cell temperature of 35 °C and inlet gas temperature of 45 °C. In
the first plot (a), dry air was supplied to the cathode side at flow
rates calculated to simulate stoich ratios of 1.5:2.5, 1.5:5 and
3:8, and a current density of 0.15 A/cm2. In these tests an
actual current density was not applied, and no hydrogen was
supplied to the anode, thus no water production was present in
the cathode gas channels. The pressure drop for these
conditions remained constant with no fluctuations. The second
plot (b) shows dry air tests run at the same stoic ratios with an
actual current density of 0.15 A/cm2 applied to the cell, and
hydrogen supplied to the anode. There is a notable change in
the both the magnitude and behavior of the pressure drop
signature at each stoich ratio, which can be attributed to the
presence of water in the cathode channels. The final plot (c)
shows results from the same stoich ratios and current density,
but with 100% RH inlet gases supplied to the cell. Again the
magnitude of the total cathode pressure drop increases for each
stoich ratio, and more severe fluctuations become present due
to larger amounts of water. This is an effect of both water
production due to cell reactions, and the increased water
present in the humidified gas streams.
(a)
(b)
3.4 Flow Maldistribution and Parallel Channel Interaction
The effects of flow maldistribution among fuel cell
channels, which is predominantly caused by slug flow
(although GDL intrusion and inlet manifold design are inherent
contributors), can be a major factor in reduced cell
performance. Flow maldistribution contributes to non-uniform
current density, localized hot spots, and material degradation.
Individual channel flow rates can provide insight into the effect
of liquid water in the channels on flow maldistribution and
(c)
Fig. 11: Total pressure drop signatures for cathode side of
fuel cell operating at 35°C cell temp., 45°C inlet gas temp,
stoich ratios (an:ca) of 1.5:2.5, 1.5:5, and 3:8 with (a) dry
air only, no H2, no current, (b) dry gases, H2 supplied,
current density of 0.15 A/cm2, (c) 100% RH inlet gases H2
supplied, current density of 0.15 A/cm2.
6
Copyright © 2009 by ASME
interaction between parallel channels. A novel method of
calculating individual channel flow rates using the entrance
region pressure drop method was developed by Kandlikar et al.
[21]. By using the pressure drop measurements obtained for
each channel during these experiments, and modifying the
entrance region pressure drop method for use with the in situ
set-up, instantaneous flow rates were calculated for each
channel.
Maldistribution and parallel channel interactions attributed
to the presence of liquid water were observed in these in situ
experiments. Figure 12 shows the fluctuations present in the
total pressure drop at a stoich ratio of 1.5:2.5 and a current
density of 0.15 A/cm2 for a period of 200 seconds. In Figure
13, the corresponding flow rates for 10 out of the 22 individual
cathode channels (for ease of viewing) are shown during this
same time period.
The channel flow rate plot in Figure 13 indicates the
compensation effect seen in parallel channels – with the flow
rate in some channels increasing at the same instant of
decreasing channel flow rate in others. The decreasing flow rate
is likely due to the presence of liquid water which causes
constricted reactant flow. As seen in the figure, from 100 to 125
seconds the flow rates in channels 10, 11, 12, 13, and 14
decreased, and simultaneously the flow rates in channels 17,
19, 20, 21, and 22 increased. This corresponds to an increase in
the cathode total pressure drop during the same time period. It
is important to note that that the magnitude of total pressure
drop fluctuations and channel flow variations are not a 1:1
relationship. Given the nature of the fluctuations in the total
pressure drop and the decreased flow rate in certain channels, it
is reasonable to assume that there is a significant presence of
liquid water in the cathode channels at this operating condition,
most likely in the form of slug flow in some of the channels.
This was confirmed by visualization, and Figure 14 shows a
representation of water in the channels at the same operating
conditions shown in Figures 12 and 13.
Slug
Flow
Fig. 12: Subplot of Fig. 11 plot (b), 1.5:2.5 stoich ratio.
Highlights fluctuations in total pressure drop due to
presence of liquid water in the channels.
Fig. 14: Presence of slug flow in cathode channels
corresponding to pressure drop fluctuations and channel
flow rates (Figures 12 and 13). Baseline GDL, 1.5:2.5
(an:ca) stoich ratio, 0.1 A/cm2, 35°C cell temp.
In situ experimentation was used to establish a relationship
between two-phase flow structure, individual channel flow rate
variations, and total pressure drop. The results correlate well
with similar investigations performed in an ex-situ facility [21,
22], and validate the use of pressure drop as a promising
diagnostic tool for assessing flow structure within fuel cell
channels.
3.4 Effects of Relative Humidity
Fig. 13: 10 channel flow rates corresponding to Fig. 12, with
current density of 0.15 A/cm2 and 1.5:2.5 (an: ca) stoich
ratio. Demonstrates maldistribution and interaction
between parallel channels due to presence of liquid water.
Other factors that affect in situ flow pattern maps are the
relative humidity of the inlet gases and the cell temperature.
The map from Figure 9 was formed from dry gas tests, and thus
the RH is not influencing the cathode channel water content.
The effect of RH on the flow pattern map is shown in Figure
15, where fully humidified gases are input at both the anode
and cathode. As can be seen, there is a major shift in the
transition lines due to the increased presence of liquid water at
each condition. This may be due to the strong condensation
from the gas stream. The effects of inlet gas RH and cell
7
Copyright © 2009 by ASME
temperature on in situ flow pattern maps are currently being
investigated.
Motors Fuel Cell Research Laboratory in Honeoye Falls, NY, is
gratefully acknowledged.
REFERENCES
[1] Kandlikar, S.G., and Lu, Z., 2009, “Fundamental
Research Needs in Combined Water and Thermal
Management Within a Proton Exchange Membrane Fuel
Cell Stack Under Normal and Coldstart Conditions,” J.
Fuel Cell Sci. Technology, in press.
[2] Trabold, T.A., Owejan, J. P., Jacobson, D.L., Arif, M.,
and Huffman, P. R., 2006, “In Situ Investigation of Water
Transport in an Operating PEM Fuel Cell Using Neutron
Radiography: Part I – Experimental Method and
Serpentine Flow Field Results,” Int. J. Heat and Mass
Transfer, 49, 4712-4720
Fig. 15: In situ map of two-phase flow patterns as a function
of air flow rate and current density, 35°C cell temp., 100%
RH inlet gases.
4. CONCLUSION
A 50 cm2 in situ fuel cell with optical access was
developed, and the two-phase flow characteristics of the
cathode channels were studied. The instantaneous flow rate in
each channel was measured, and simultaneous visualization of
two-phase flow structures present within the channels was
performed. Three primary flow patterns – slug, film, and mist
flow – were identified, and the in situ flow pattern map was
formed. Slug flow was found to be dominant at lower air flow
rates, while film and mist flow were commonly present at
intermediate and high air flow rates, respectively.
The pressure drop feature for each flow pattern was
described. Large fluctuations in total pressure drop were the
result of slug flow within the channels. Film flow caused slight
fluctuations in pressure drop. Almost no fluctuation in total
pressure drop was present during mist flow. It was established
that the total pressure drop signature can be used as a diagnostic
tool for two-phase flow dynamics in the gas channels.
Flow maldistribution caused by liquid water presence
(attributed mostly to slug flow) in the cathode channels was
demonstrated. Interaction among parallel channels was
observed, with an increased flow rate in some channels due to
compensation for a decreased flow rate in channels where the
presence of liquid water restricted reactant flow.
The effect of inlet gas humidification was also studied, and
the presence of humidified inlet gases showed significant
impact on the in situ flow pattern map. This effect is currently
being investigated along with cell temperature effects.
ACKNOWLEDGMENTS
This work was supported by the US Department of Energy
under contract No. DE-FG36-07G017018. Technical support
from Dr. Thomas Trabold and Mr. Jon Owejan of General
[3] Owejan, J. P., Trabold, T. A., Jacobson, D. L., Baker, D.
R., Hussey, D. S., and Arif, M., 2006, “In Situ
Investigation of Water Transport in an Operating PEM
Fuel Cell Using Neutron Radiography: Part 2 – Transient
Water Accumulation in an Interdigitated Cathode Flow
Field,” Int. J. Heat and Mass Transfer, 49, 4721-4731
[4] Turhan, A., Heller, K., Brenizer, J.S., and Mench, M.M.,
2006, “Quantification of Liquid Water Accumulation and
Distribution in a Polymer Electrolyte Fuel Cell Using
Neutron Imaging,” J. Power Sources, 160, 1195-1203
[5] Hussey, D.S., Jacobson, D.L., Arif, M., Owejan, J.P.,
Gagliardo, J.J., and Trabold, T.A., 2007, “Neutron Images
of the Through-Plane Water Distribution of an Operating
PEM Fuel Cell,” J. Power Sources 172, 225–228
[6] Hickner, M. A., Siegel, N. P., Chen, K. S., Hussey, D. S.,
Jacobson, D. L., and Arifc, M., 2008, “In Situ HighResolution Neutron Radiography of Cross-Sectional
Liquid Water Profiles in Proton Exchange Membrane
Fuel Cells,” J. Electrochem. Soc., 155, B427-B434.
[7] Tuber, K.., Pocza, D., and Hebling, C., 2003,
“Visualization of Water Buildup in the Cathode of a
Transparent PEM Fuel Cell,” J. Power Sources, 124, 403414.
[8] Yang, X.G., Zhang, F.Y., Lubawy, A.L., and Wang, C.Y.,
2004, “Visualization of Liquid Water Transport in a
PEFC,” Electrochemical and Solid-State Letters, 7, A408A411.
[9] Zhang, F.Y., Yang, X.G., and Wang, C.Y., 2006, “Liquid
Water Removal from a Polymer Electrolyte Fuel Cell,” J.
Electrochem. Soc., 153, A225-A232.
[10] Weng, F.B., Su, A., Hsu, C.Y., and Lee, C.Y., 2006,
“Study of Water-Flooding Behavior in Cathode Channel
of a Transparent Proton-Exchange Membrane Fuel Cell,”
J. Power Sources, 157, 674-680.
8
Copyright © 2009 by ASME
[11] Liu, X., Guo, H., and Ma, C., 2006, “Water Flooding and
Two-Phase Flow in Cathode Channels of Proton
Exchange Membrane Fuel Cells,” J. Power Sources, 156,
267–280.
[20] Barbir F., Gorgun, H., and Wang, X., 2005, “Relationship
Between Pressure Drop and Cell Resistance as a
Diagnostic Tool for PEM Fuel Cells,” J. Power Sources,
141, 96–101.
[12] Spernjak, D., Prasad, A.K., and Advani, S.G., 2007,
“Experimental Investigation of Liquid Water Formation
and Transport in a Transparent Single-Serpentine PEM
Fuel Cell,” J. Power Sources, 170, 334–344.
[21] Kandlikar, S.G., Lu, Z., Domigan, W.E., White, A.D., and
Benedict, M.W., 2008, “Measurement of Flow
Maldistribution in Parallel Channels and its Application
to Ex-Situ and In-Situ Experiments in PEMFC Water
Management Studies,” Int. J. Heat and Mass Transfer, 52,
1741–1752.
[13] Liu, X., Guo, H., Ye, F., and Ma, C.F., 2007, “Water
Flooding and Pressure Drop Characteristics in Flow
Channels of Proton Exchange Membrane Fuel Cells,”
Electrochim. Acta, 52, 3607–3614.
[14] Ge, S., and Wang, C.Y., 2007, “Liquid Water Formation
and Transport in the PEFC Anode,” J. Electrochem. Soc.,
154, B998–B1005.
[15] Liu, X., Guo, H., Ye, F., and Ma, C.F., 2008, “Flow
Dynamic Characteristics in Flow Field of Proton
Exchange Membrane Fuel Cells,” Int. J. Hydrogen
Energy, 33, 1040–1051.
[16] Spernjak, D., G. Advani, S.G., and Prasad, A.K, 2009,
“Simultaneous Neutron and Optical Imaging in PEM Fuel
Cells,” J. Electrochem. Soc., 156, B109-B117.
[17] He, W., Lin, G., and Nguyen, T.V., 2003, “Diagnostic
Tool to Detect Electrode Flooding in Proton-ExchangeMembrane Fuel Cells,” AIChE J., 49, 3221–3228.
[18] Ma, H.P., Zhang, H.M., Hu, J., Cai, Y.H., and Yi, B.L.,
2006, “Diagnostic Tool to Detect Liquid Water Removal
in The Cathode Channels of Proton Exchange Membrane
Fuel Cells,” J. Power Sources, 162, 469–473.
[19] Bosco, A.D., and Fronk, M.H., Fuel Cell Flooding
Detection and Correction. U.S. Patent 6103409; 2000.
[22] Lu, Z., Kandlikar, S.G., Rath, C., Grimm, M., Domigan,
W., White, A.D., Hardbarger, M., Owejan, J.P., and
Trabold, T.A., 2009, “Water Management Studies in PEM
Fuel Cells, Part II: Ex-Situ Investigation of Flow
Maldistribution, Pressure Drop and Two-Phase Flow
Pattern in Gas Channels,” Int. J. Hydrogen Energy,
doi:10.1016/j.ijhydene.2008.12.025.
[23] Owejan, J.P., Gagliardo, J.J., Sergi, J.M., Kandlikar, S.G.,
and Trabold, T.A., 2009, “Water Management Studies in
PEM Fuel Cells: Part I – Fuel Cell Design and In-Situ
Water Distributions,” Int. J. Hydrogen Energy,
doi:10.1016/j.ijhydene.2008.12.100.
[24] U.S. Department of Energy, “Hydrogen, Fuel Cells, and
Infrastructure
Technology
Programs:
Multi-Year
Research, Development and Demonstration Plan,”
Section
3.4
–
Fuel
Cells,
October
2007.
[http://www1.eere.energy.gov/hydrogenandfuelcells/myp
p/]
[25] Zhang, F.Y., Yang, X.G., and Wang, C.Y., 2006, “Liquid
Water Removal from a Polymer Electrolyte Fuel Cell,” J.
Electrochem. Soc., 153, A225-A232.
[26] Grolman, E., and Fortuin, J.M.H., 1997, “Liquid HoldUp, Pressure Gradient, and Flow Patterns in Inclined GasLiquid Pipe Flow,” Exp. Therm. Fluid Sci.,” 15, 174–82.
9
Copyright © 2009 by ASME