Experimental and numerical studies of local current
mapping on a PEM fuel cell
J.J. Hwnag
a, W.R. Chang
b,*
, R.G. Peng
c, P.Y. Chen
d, A. Su
eaDepartment of Environment and Energy, National University of Tainan, Tainan 700, Taiwan bDepartment of Landscape Architecture, Chung-Hua University, Hsinchu 300, Taiwan c
Department of Mechanical Engineering, National Chiao Tong University, Hsinchu 300, Taiwan
dDepartment of Power Mechanical Engineering, National Tsing Hua University, Hsinchu 300, Taiwan eDepartment of Mechanical Engineering and Fuel Cell Center, Yuan Ze University, Taoyuan, Taiwan
a r t i c l e
i n f o
Article history: Received 16 April 2008 Received in revised form 6 July 2008
Accepted 8 July 2008
Available online 14 September 2008 Keywords:
PEM fuel cell MEA
Hall-effect sensor Current
a b s t r a c t
Local current distribution on a PEM fuel cell has been mapped experimentally by using a special-designed single cell fixture. It is composed of a composite cathodic flow-field plate, a membrane electrode assembly (MEA) and a stainless-steel anodic flow-field plate. An array of 16 individual conductive segments was distributed on the composite plate. A self-made MEA is in direct contact with the segmented current collectors. Regional-aver-aged current through each segment is determined by using the Hall-effect sensor. To ensure the data reliability, a comparison of polarization curves was made between the composite field plate and the conventional field plate. Then, the effects of flow-field patterns, dew points of the cathodic feedings and cathodic stoichiometrics on the local current distribution were examined. The transient variation of the local current distribution on the cathode under supersaturated conditions was further visualized to illustrate the flooding phenomena in different flow patterns. This technique developed by the present work has contributed to knowledge and understanding the local current distributions in a PEM fuel cell that is helpful in designing the fuel-cell components. ª2008 International Association for Hydrogen Energy. Published by Elsevier Ltd. All rights
reserved.
1.
Introduction
Owing to many advantages accompanied by proton
exchange membrane (PEM) fuel cells, such as non-pollution, high efficiency, low noise and quick startup, they have been regarded as the most potential power system for the ultimate eco-vehicles. In the past decade, almost all major car manufacturers in the world have devoted themselves to the development of PEM fuel cell vehicles. However, there still have been some challenges about PEM fuel-cell tech-nologies that need to be overcome, especially in prolonging the lifetime of the PEM fuel cell[1].
It is well known that the lifetime of a PEM fuel cell depends strongly on the local current distribution on the electrode. Uneven distribution of local current on the electrode means irregularity of the electrochemical reaction on the reacting site. It not only reduces the cell performance but also shortens its lifetime. The local current distribution on a PEM fuel cell will be affected by many factors, such as cell temperature, degree of humidification, partial pressure of the feeding gases, configuration of the flow channels, microstructures and characteristics of the electrodes[2–4], etc. Therefore, it is very important to understand the above effects on local current distribution for designing and operating a PEM fuel cell.
* Corresponding author. Tel.: þ886 3374281. E-mail address:ruby@chu.edu.tw(W.R. Chang).
j o u r n a l h o m e p a g e : w w w . e l s e v i e r . c o m / l o c a t e / h e
0360-3199/$ – see front matter ª 2008 International Association for Hydrogen Energy. Published by Elsevier Ltd. All rights reserved. doi:10.1016/j.ijhydene.2008.07.035
In the past, many experimental researches have published concerning the local current distribution in a PEM fuel cell [5–19]. Some works rely on the numerical simulation[20–28]. However, studies reported that both experimental and numerical results are rather sparse, which motivates the present work to map the local current on the PEM fuel cell numerically and experimentally. Experimental evidence not only confirms the numerically calculated results but also gives a better understanding of the nature of the electro-chemical reactions occurring inside the cell. Thus, the local current distribution by direct measurement can offer essen-tial references for designs of cells, stacks and even systems of the PEM fuel cell. In the present work, a special-designed fixture is used to in-situ map the local current distribution on the electrode of a PEM fuel cell. An array of 4 by 4 segmental current collectors is distributed on the cathodic flow-field plate. Each conductive segment is insulated mutually. Current through each current collector is measured individ-ually by using a Hall-effect sensor, which well represents the regionally averaged data on its counterpart. A self-made MEA is integrated into the above single-cell fixture for local
current measurement. Parametric studies include the
cathodic flow-field patterns (parallel, serpentine, interdigi-tated and biomimic flow fields), the cathodic-feeding stoi-chiometries (xc¼2.5–3.5) and the dew point of cathodic
feedings (Tdp,c¼40–70C). First, the experiments are
compared with the numerical simulation results to verify the reliability of the instruments. In addition, to ensure the data consensus, a comparison of polarization curves is made between the composite flow-field plate and the conventional flow-field plate. Moreover, the evolution of local current distribution on the cathode of the parallel and serpentine flow fields under supersaturated conditions are visualized and compared thereafter. It can give an evidence to explain the reason why the parallel flow field is easier flooded in the electrode than the serpentine flow field. Finally, the influ-ences of flow-field pattern and stoichiometry variation of
feeding gases are discussed in terms of the measured local current distribution in the segmented single cell.
2.
Experiment
Fig. 1 (a) shows a photo of the special-designed single-cell fixture for local current measurements, whileFig. 1(b) is the corresponding assembly diagram. The fixture consists of a cathodic flow-field plate, an anodic flow-field plate, an MEA and other assisting parts (such as fittings, screws, gaskets, etc.). The anodic flow-field plate is made of stainless steel, while the cathodic counterpart is a composite plate that is made of a polyoxymethylene (POM) frame infixed a number of stainless-steel current collectors. As shown inFig. 2, on the composite plate, the active surface is divided by 4 by 4 segments. Sixteen pieces of stainless-steel blocks of dimen-sion 9 mm 9 mm 12 mm are inserted into a pre-machined POM frame with a space of 2.5 mm in between. They are electrically insulated from each other. All stainless-steel blocks are coated with a thin layer of titanium on surface of to prevent the possible corrosion. In addition, to ensure leakage free of the composite plate, the epoxy resin is used to fill gaps between the POM frame and the stainless steel blocks. After the epoxy is dried, the flow-field channels are carved on the surface of the composite plate using the CNC machine. As shown in Fig. 3, four kinds of flow fields are tested in the present work, i.e., a parallel flow field, a serpentine flow field, an interdigitated flow field, and a biomimic flow field, respectively. The depth of flow channels is 1.0 mm and the reactive surface is 5.0 cm 5.0 cm. Note that the open-area ratios of the above four flow fields ranged from 53% to 58%. Under the same clamping pressure, the Ohmic resistance for current passing through interface between the gas diffusion layers and the flow-field plate can be regarded as the same [13]. As for the anodic plate, the surface is machined with a serpentine flow field that is the same as the cathodic Nomenclature
cO2 concentration of oxygen (mol m
3)
cH2O concentration of water vapor (mol m
3)
DH2O diffusivity of water vapor (m
2s1)
DH2O;eff effective diffusivity of water vapor (m
2s1)
DO2 diffusivity of oxygen (m
2
s1) DO2;eff effective diffusivity of oxygen (m
2
s1)
F faradays constant (96 487 C mol1)
i local current density (A m2)
io exchange current density (A m2)
ict local transfer current density
p pressure (Pa)
R universal gas constant (W mol1K1)
T temperature (K)
M molecular weight (kg mol1)
Sa surface area-to-volume ratio of the porous matrix
(m1)
U operational voltage of the single cell (V)
u velocity vector (m s1) x, y, z coordinates system (m) Greek symbols x reactant stoichiometry k permeability (m2) e cathode porosity r density (kg m3) h overpotential (V) s tortuosity a symmetric factor m dynamic viscosity (m s2) 4 potential (V) u mass fraction Suffix a anode c cathode cell cell dp dew point eff effective in inlet out outlet ref reference
counterpart. Its surface is also coated with a thin layer of titanium to prevent the possible erosion. Note that the anode keeps its flow pattern when the effect of the flow-field on the cathodic plate is examined.
The MEA used in the present work is prepared as the following processes[14]. First, the membranes Nafion112 are soaked in a mixture of H2O2/H2O at approximately 80C for
2 h, and then are washed with distilled water to remove the organic and mineral impurities. Subsequently, the trans-parent Nafionmembranes are soaked in a 1 M boiling H
2SO4
solution for 2 h. Finally, they are rinsed again to remove excess H2SO4, and thus the membranes, being ready to use,
are stored in distilled water. Subsequently, the commercially available electrocatalyst, 20% platinum on Vulcan XC-72
carbon (from E-Tek), is suspended in an aqueous Nafion
solution, which is then well mixed by using an ultrasonic bath.
The above mixture is then printed directly onto two sides of the membrane as active layers, rather than deposed onto gas diffusion layers (carbon papers). The amounts of Pt and Nafion on the electrode were about 0.2 and 0.6 mg/cm2,
respectively. As for the gas diffusion layers, composed 20% Pt/ C and 0.2 Pt mg/cm2, for both electrodes is added to the carbon papers. The five-layer MEA is assembled conventionally, using a hot pressing process conducted at 140C and 8.0 Mpa for
90 s. The active surface area of the MEA is 25 cm2.
As shown inFig. 4, the current on each segment is con-ducted to a multiplexer consisted of 16 Hall-effect sensors. The assembled single-cell fixture is then installed on a commercial test stand (APFCT FCED 200) that controls all operational conditions such as dew points of feeding gases,
cell temperature and backpressure. Table 1 lists the MEA
configurations and details of the operation conditions. Fig. 1 – Test fixture of local current measurement, (a) photo of the test fixture and (b) assembly diagram of the test fixture.
3.
Numerical model and scheme
To realize the electrochemical reactions within the flow field of the fuel cell, a numerical approach is conducted to compare with the experiment data. The details of the model have been discussed elsewhere and only some important features are described here.
The oxygen reduction reaction (ORR) takes place on the three-phase boundary (TPB) of the porous cathode. It can be represented by the following equations.
O2þ4Hþþ4e/2H2O (1)
Note that the breathing holes are not only the entrance of the fresh air but also the exit of the products (water vapor) of the electrochemical reaction.
Fig. 2 – Locations of the current collectors and their corresponding number on the composite flow-field plate.
The momentum and mass conservations for the gas flow mixture in the porous cathode are described as follows: 1 eV,ðruuÞ ¼ 3Vp þ V,ðmVuÞ þ 3mu k (2) VðruÞ ¼ictSaM 4F (3)
where r is the density of the gas mixture; u, the velocity vector; k, the permeability; 3, the porosity; m, the viscosity and p is the pressure. As for the mass conservation, the electro-chemical reaction occurring on the TPB acts as a sink of oxygen and, simultaneously, a source of water vapor. Sais the
surface area-to-volume ratio of the porous matrix; ict, the local
transfer current density; M, the mole masses of gas mixtures and F is the Faraday’s constant.
The species conservations for oxygen, water vapor and nitrogen in the multi-component system are expressed respectively as V,ruuO2 ¼ rDO2;effV2uO2þ ruO2 M DO2;effVM j Dj;effV Muj þSO2 (4) V,ruuH2O ¼ rDH2O;effV 2u H2Oþ ruH2O M DH2O;effVM X j Dj;effV Muj þSH2O (5) uN2¼1 uO2 uH2O (6)
In Eqs. (4) and (5), the first term on the right-hand side represents the Fickian diffusion, while the second and third terms are the modification to enforce the Stefan–Maxwell equations on the multi-component diffusive system. The source terms for the mass balances for oxygen and water vapor are given, respectively, by
SO2¼ ictSaMO2 4F (7) SH2O¼ ictSaMH2O 2F (8)
As for the effective diffusivities of the oxygen ðDO2;effÞand
water vapor ðDH2O;effÞin the porous cathode, they follow the
Bruggemann model, i.e., DO2;eff¼ 3 sD O2 (9) DH2O;eff¼ 3 sD H2O (10)
The local transfer current density ictinEqs. (3), (7) and (8)is
depicted by the Butler–Volmer correlation[12].
ict¼io c O2 cO2;ref exp 4aF RTðfl fsÞ cH2O cH2O;ref 2 exp 4ð1 aÞF RT ðfl fsÞ (11)
where a is the symmetric factor and T is the temperature. cO2and
cH2Oare the concentrations of oxygen and water vapor, respectively.
4l 4s is the potential difference between the electrolyte
phase and the catalyst phase in the porous electrode. It drives the transfer current density from the electrolyte phase to the catalyst phase that keeps the electrochemical reaction continuously.
The above governing equations are numerically solved by using a commercial code[29,30]. Computations are performed on 2252 unstructured meshes. Additional runs for the coarser meshes, 1832, and the finer meshes, 2682, are taken for a check of grid independence. The program gives results within 1% of each other on the finest meshes used. A typical simulation requires about 300 min of central processing unit time on a Pentium IV 2.8 GHz PC.
4.
Results and discussions
4.1. Validation of the measurement technique
Before the discussion of the experimental results, it is neces-sary to validate the present instruments for local current Fig. 4 – Configuration of the 16-channel multiplexer for
local current measurements.
Table 1 – .Parametric conditions of the experiment
Cell PEM Nafion112
Reaction area 25 cm2
Pt loading 0.2 mg/cm2for anode
and cathode GDL E-Tech for anode
and cathode Cell temperature, Tcell 40, 50, 60C Anode feeding H2 Pure (>99.99%)
Dew point, Tdp,a 40C
Stoichiometry (flow rate), xa 1.5 Cathode feeding O2 Pure
Dew point, Tcelldp,c 40, 50, 60, 70C
Stoichiometry (flow rate), xc
measurements. Fig. 5 compares the polarization curves between the composite single cell (dashed line) and the conventional single cell (square symbol). The cell operational temperature is fixed Tcell¼40C. The stoichiometries for the
anodic and cathodic feeding are fixed at xa¼1.5 and xc¼2.5,
respectively. It is seen that these two polarization curves almost fall into a single curve. The consensus of data between these two techniques means the present instruments are reliable. Fig. 6 further confirms the present technique by comparing the local current distributions between the numerical predictions and the experimental measurements at the cell voltage of U ¼ 0.5 V. Qualitatively, the local current distributions are largely similar between measured and pre-dicted results. Both results reveal a local maximum near the entrance of the flow-field plate, and a local minimum near the outlet of the flow-field plate. In quantity, the predicted and measured averaged current densities are 142 and 158 mA/cm2,
respectively, with an error about 10%. Generally, the agree-ment between the above results is satisfactory.
4.2. Evolution of current distribution
Fig. 7shows a comparison of transient developments of local current distributions on two different flow-field plates under supersaturated conditions, i.e., the parallel flow-field plate and the serpentine flow-field plate. It can provide a compara-tive assessment of the capability of water management between these two flow fields. The cell voltage is fixed at U ¼ 0.5 V, while the stoichiometries of anodic and cathodic feedings are xa¼1.5 and xc¼2.5, respectively. The cell
opera-tional temperature is fixed at Tcell¼40C. The anodic feeding
is saturated with a dew point of 40C, while the cathodic
feeding has a dew point of Tdp,c¼70C.
As shown inFig. 7(a), at the initial stage (t ¼ 0 min), the current distribution reveals significantly high values near the inlet of the parallel flow channels, which is resulted from the incoming fresh oxygen. With increasing time, the local maximum current drops gradually; thus, the current distri-bution is leveled off. The phenomena may be explained as follows. When the oxygen of dew point Tdp,c¼70C enters
a cold environment (i.e., Tcell¼40C), the water vapor
becomes supersaturated. Therefore, a large amount of water Fig. 5 – Comparison of the polarization curves between the
present composite cathodic flow-field plate and the conventional flow-field plate, Tcell[ 40 8C, xa[ 1.5 and xc[ 2.5.
Fig. 6 – Comparison of the numerical predictions and the experimental measurement, (a) flow direction, (b) numerical results and (c) experimental results, U [ 0.5 V, Tcell[ 40 8C, xa[ 1.5 and xc[ 2.5.
vapor is condensed to liquid form. In addition, the ORR generates some liquid water. All liquid water in the electrode should be removed to keep the electrochemical reaction. However, the pressure gradient for flow across the parallel channels is too small to drive out the liquid water. The liquid
water blocks the flow and thus retards the electrochemical reaction. In contrast, as shown in Fig. 7 (b), the current distribution on the serpentine flow-field plate is relatively uniform at the initial stage. Then it decreases slightly and soon reaches to a steady distribution. The slight decline in
Y-Axis 1 2 3 4 X-Axis 1 2 3 4 Current / A 0 0.2 0.4 0.6 Current / A0.7 0.6 0.5 0.4 0.3 0.2 0.1 0 Y-Axis 1 2 3 4 X-Axis 1 2 3 4 Current / A 0 0.2 0.4 0.6 Current / A0.7 0.6 0.5 0.4 0.3 0.2 0.1 0 Y-Axis 1 2 3 4 X-Axis 1 2 3 4 Current / A 0 0.2 0.4 0.6 Current / A0.7 0.6 0.5 0.4 0.3 0.2 0.1 0 Y-Axis 1 2 3 4 X-Axis 1 2 3 4 Current/ A 0 0.2 0.4 0.6 Current / A0.7 0.6 0.5 0.4 0.3 0.2 0.1 0 Y-Axis 1 2 3 4 X-Axis 1 2 3 4 Current / A 0 0.2 0.4 0.6 Current / A 0.7 0.6 0.5 0.4 0.3 0.2 0.1 0 Y-Axis 1 2 3 4 X-Axis 1 2 3 4 Current / A 0 0.2 0.4 0.6 Current /A0.7 0.6 0.5 0.4 0.3 0.2 0.1 0 Y-Axis 1 2 3 4 X-Axis 1 2 3 4 Current / A 0 0.2 0.4 0.6 Current / A 0.7 0.6 0.5 0.4 0.3 0.2 0.1 0 Y-Axis 1 2 3 4 X-Axis 1 2 3 4 Current / A 0 0.2 0.4 0.6 Current / A0.7 0.6 0.5 0.4 0.3 0.2 0.1 0 Y-axis 1 2 3 4 X-Axis 1 2 3 4 Current / A 0 0.2 0.4 0.6 Current / A0.7 0.6 0.5 0.4 0.3 0.2 0.1 0 Y-Axis 1 2 3 4 X-Axis 1 2 3 4 Current / A 0 0.2 0.4 0.6 Current / A0.7 0.6 0.5 0.4 0.3 0.2 0.1 0 0 min 1 min 15 min 30 min
Parallel flow field Serpentine flow field
45 min
Outlet inlet Outlet inlet
Outlet inlet Outlet inlet
Outlet inlet Outlet inlet
Outlet inlet Outlet inlet
Outlet inlet Outlet inlet
a
b
Fig. 7 – Transient developments of the local current distribution, (a) parallel flow field and (b) serpentine flow field, U [ 0.5 V, Tcell[ 40 8C, xa[ 1.5 and xc[ 2.5.
local current in the startup period (from 0 min to 1 min) is because of an increase in the oxygen usage. The quickly stabilized current distribution is due to the good water management in the serpentine flow field. In general, a signif-icant pressure gradient in the serpentine flow channels can keep the water in good movement. Thus, the performance remains stable without water accumulation in the electrode even after a long-term test. Note that the steady-state distri-butions of local current shown in Fig. 7(b) reveal a slight bumper near the middle of the reaction region where is the straight portions of the serpentine flow channels. That is, the local current around the 180oreturn regions is smaller than
that on the straight counterparts. This is because the liquid water is trapped easily as the flow turns around the 180o
returns[31], and thus reduces the electrochemical reaction. This is caused mainly by the longer residence times of the velocity distribution of the core gas flow. Particle image velocimetry (PIV) measurements[32]in the U-shaped channel identified the regions of recirculation and flow separation. 4.3. Parametric studies
Fig. 8further shows a comparison of the effect of cathodic-feeding dew point (Tdp,c¼40–70C) on the cell performance
between the parallel flow field and the serpentine flow field. Both the dew point of the anodic feedings and the cell
oper-ational temperature are fixed at 40C. The anodic and
cathodic flow rates are fixed at stoichiometry xa¼1.5 and
xc¼2.5, respectively. The operational voltage of the cell is
fixed at U ¼ 0.5 V. It is seen fromFig. 8that at the low dew point of the cathodic feedings, Tdp,c¼40C, the current
delivered by the single cell with a parallel flow field is almost the same as that with a serpentine flow-field. As the cathodic-feeding dew point increases, the current for the parallel flow field decreases drastically, while, in contrast, the current in the serpentine flow-field changes slightly. As mentioned above, when the saturated reactants with a high dew point (e.g., 70C) enters a cold flow field (T
cell¼40C), a large
amount of water vapor will be condensed. The excess of liquid water not only blocks the flow channels along with the gas diffusion layer that increases the concentration polarization, but also covers the reacting site in the catalyst layer. Both effects will reduce the electrochemical reaction rate in a PEM fuel cell. As shown inFig. 8, in the parallel flow channels, an increase in cathodic-feeding dew point from Tdp,c¼40C to
70C results in a decrease in current from 8.2 to 5.5 A, with
a reduction of the electrochemical performance about 30%. Using the serpentine flow channels instead of the parallel ones, the current reduction is only 5%. This is because the serpentine flow channels have high-pressure drops along the flow direction that can drive the liquid water downstream. Contrarily, the pressure drops across the parallel flow chan-nels cannot well remove the water from the porous cathode. Thus, it can be concluded that the serpentine flow channels perform better than the parallel flow channels due to their superior in water management.
Fig. 9shows the local current distributions on the biomimic flow-field plate under three different cell voltages, i.e., 0.3, 0.5 and 0.7 V. The cell operational temperature and the dew point for both feedings are fixed at 60C. The stoichiometries of
hydrogen and oxygen are maintained at xa¼1.5 and xc¼2.5,
respectively. It is seen from the figure that the current distri-bution is raised by the decrease in cell operational voltage. At the high cell voltage of 0.7 V, the local current distribution is rather uniform. However, at low operational voltage of 0.3 V, the distributions of the local current on the cathodic flow-field plate are relatively uneven.
Fig. 10 shows the effect of stoichiometry of the cathodic feedings on the local current distributions on the biomimic flow-field plate. The cell operational temperature and voltage are fixed at Tcell¼60C and U ¼ 0.6 V, respectively. The dew
point for both feedings is fixed at 60C while the hydrogen
stoichiometry is kept at xa¼1.5. It is seen from this figure that
the effect of oxygen flow rate on the local current distribution is negligible as the cathodic stoichiometry is varied from xc¼2.5 to xc¼3.5.
Fig. 8 – Effect of dew point of cathodic feedings on the current and power for parallel and serpentine flow fields, U [ 0.5 V, Tcell[ 40 8C, xa[ 1.5 and xc[ 2.5. Y-Axis 1 2 3 4 X-Axis 1 2 3 4 Current/A 0 0.2 0.4 0.6 0.3 V 0.5 V 0.7 V 0.8 Current/A 0.8 0.7 0.6 0.5 0.4 0.3 0.2 0.1 0 Outlet Inlet
Fig. 9 – Local current distributions under different operating voltages, biomimic flow-field plate, Tcell[ 40 8C, xa[ 1.5 and xc[ 2.5.
Fig. 11 shows a comparison of the polarization curves among the three different flow fields, i.e., interdigitated, serpentine and biomimic flow fields. Both the dew point of the cathodic feedings and the cell operational temperature are 60C. The anodic and cathodic feedings have flow rates of
xa¼1.5 and xc¼2.5, respectively. It is clearly seen that the
serpentine flow field performs best among the three flow fields as the cell operational voltage is less than 0.8 V. At high oper-ational voltage (U > 0.8 V), the performance of the biomimic flow field is slightly better than the serpentine flow field. Noteworthy from the above discussion that using the same MEA under the same operational conditions, switching the flow-field plate means altering the mass transfer mechanisms in the fuel cell. That is, the local current distribution is affected
by the flow-field patterns via altering the convection/diffusion mechanisms. The present results have demonstrated that the serpentine flow field is more favorable than other types of flow fields due to its stronger mass transfer ability.
5.
Conclusions
This paper has demonstrated a new technique for determi-nation of regionally averaged current distributions in a PEM fuel cell. The technique is mapping the current distribution on a special-designed composited plate with an array of segmented current collectors. A 16-channel multiplexer with an array of Hall-effect current sensors simultaneously detects transient distribution of the local currents on the composite plate. A self-made membrane electrode assembly is sand-wiched between the cathodic composite plate and an anodic stainless-steel plate. The effects of the cathodic flow-field patterns (parallel, serpentine, interdigitated and biomimic flow fields), the cathodic-feeding stoichiometries (xc¼2.5–3.5)
and the dew point of cathodic feedings (Tdp,c¼40–70C) on
the current distribution on a PEM fuel cell have been exam-ined. Results showed that the local current distributions obtained by the present composite plate are compared well with the numerical results. The higher the operational voltage is, the more uniform the local current distribution become. The evolution of local current distribution further showed that the serpentine flow field performs best among the flow fields investigated due to its superior in mass transfer capa-bility as well as in water management. Moreover, from the local current distribution in the serpentine flow field, it is found that the liquid water entraps easily in the corners of the 180oreturn that reduces the electrochemical reaction.
Acknowledgements
This work was partly sponsored by the National Science Council of Taiwan under contract No. NSC 95-2212-E-216-003.
a
ξ =2.5
b
ξ =3.0
c
ξ =3.5
Y-Axis 1 2 3 4 X-Axis 1 2 3 4 Current/A 0 0.2 0.4 0.6 0.8 Current/A Y-Axis 1 2 3 4 X-Axis 1 2 3 4 Current/A 0 0.2 0.4 0.6 0.8 Current/A Y-Axis 1 2 3 4 X-Axis 1 2 3 4 Current/A 0 0.2 0.4 0.6 0.8 Current/A0.8 0.7 0.6 0.5 0.4 0.3 0.2 0.1 0 0.8 0.7 0.6 0.5 0.4 0.3 0.2 0.1 0 0.8 0.7 0.6 0.5 0.4 0.3 0.2 0.1 0 Outlet Inlet Outlet Inlet Outlet InletFig. 10 – Effect of stoichiometry of cathodic feedings (oxygen flow rate) on the local current distribution on biomimic flow field, (a) x [ 2.5, (b) x [ 3.0 and (c) x [ 3.5, U [ 0.6 V, Tcell[ 60 8C.
Fig. 11 – Comparison of performance among three different flow fields, Tcell[ Tdp,c[ 60 8C, xa[ 1.5 and xc[ 2.5.
r e f e r e n c e s
[1] Hwang JJ, Wang DY, Shih NC. Development of a lightweight fuel cell vehicle. J Power Sources 2005;141:108–15.
[2] Noponen M, Mennola T, Mikkola M, Hottinen T, Lund P. Measurement of current distribution in a free-breathing PEMFC. J Power Sources 2002;106:304–12.
[3] Ghosh PC, Wu¨ster T, Dohle H, Kimiaie N, Mergel J, Stolten D. In situ approach for current distribution measurement in fuel cells. J Power Sources 2006;154:184–91.
[4] Dutta S, Shimpalee S, Van Zee JW. Numerical prediction of mass-exchange between cathode and anode channels in a PEM fuel cell. Int J Heat Mass Transfer 2001;44: 2029–42.
[5] Stumper J, Campbell SA, Wilkinson DP, Johnson MC, Davis M. In-situ methods for the determination of current
distributions in PEM fuel cells. Electrochim Acta 1998;43: 3773–83.
[6] Cleghorn SJC, Derouin CR, Wilson MS, Gottesfeld S. A printed circuit board approach to measuring current distribution in a fuel cell. J Appl Electrochem 1998;28:663.
[7] Yoon YG, Lee WY, Yang TH, Park GG, Kim CS. Current distribution in a single cell of PEMFC. J Power Sources 2003; 118:193–9.
[8] Sun H, Liu H, Guo LJ. PEM fuel cell performance and its two-phase mass transport. J Power Sources 2005;143:125. [9] Hakenjos A, Hebling C. Spatially resolved measurement of
PEM fuel cells. J Power Sources 2005;145:307–11.
[10] Ghosh PC, Wuster T, Dohle H, Kimiaie N, Mergel J, Stolten D. In situ approach for current distribution measurement in fuel cells. J Power Sources 2006;154:184–91.
[11] Eckl R, Grinzinger R, Lehnert W. Current distribution mapping in polymer electrolyte fuel cells – a finite element analysis of measurement uncertainty imposed by lateral currents. J Power Sources 2006;154:171–9.
[12] Strickland G, Litster S, Santiago JG. Current distribution in polymer electrolyte membrane fuel cell with active water management. J. Power Sources 2007;174:272–81.
[13] Wahdame, Candusso D, Francois X, Harel F, Pera M, Hissela D, et al. Comparison between two PEM fuel cell durability tests performed at constant current and under solicitations linked to transport mission profile. Int J Hydrogen Energy 2007;32:4523–36.
[14] Scott Weil K, Xia G, Yang ZG, Kim JY. Development of a niobium clad PEM fuel cell bipolar plate material. Int J Hydrogen Energy 2007;32:3724–33.
[15] Sun H, Zhang G, Guo L, Dehua S, Liu H. Effects of
humidification temperatures on local current characteristics in a PEM fuel cell. J Power Sources 2007;168:400–7.
[16] Sauer DU, Sanders T, Fricke B, Lofer TB, Wippermann K, Kulikovsky AA, et al. Measurement of the current
distribution in a direct methanol fuel cell – confirmation of
parallel galvanic and electrolytic operation within one cell. J Power Sources 2008;176:477–83.
[17] Wang L, Liu H. Separate measurement of current density under the channel and the shoulder in PEM fuel cells. J Power Sources 2008;180:365–72.
[18] Sun H, Zhang G, Guo LJ, Liu H. A novel technique for measuring current distributions in PEM fuel cells. J Power Sources 2006;158:326–32.
[19] Hartnig C, Manke I, Kardjilov N, Hilger A, Grunerbel M, Kaczerowski J, et al. Combined neutron radiography and locally resolved current density measurements of operating PEM fuel cells. J Power Sources 2008;176:452–9.
[20] Hwang JJ, Liu CC, Savinell RF, Wainright J. A three-dimensional numerical simulation of the transport phenomena in the cathodic side of a PEMFC. J Appl Electrochem 2004;34:217.
[21] Hwang JJ, Chen PY. Heat/mass transfer in porous electrodes of fuel cells. Int J Heat Mass Transfer 2006;49: 2315–27.
[22] Wang Y, Wang CY. A nonisothermal, two-phase model for polymer electrolyte fuel cells. J Electrochem Soc 2006;153: A1193–200.
[23] Pasaogullari U, Wang CY. Liquid water transport in gas diffusion layer of polymer electrolyte fuel cells. J Electrochem Soc 2004;151:A399–406.
[24] Jung HM, Lee WY, Park JS, Kim CS. Numerical analysis of a polymer electrolyte fuel cell. Int J Hydrogen Energy 2004;29: 945–54.
[25] Hwang JJ. A complete two-phase model of a porous cathode of a PEM fuel cell. J Power Sources 2007;164:174–81. [26] Hwang JJ, Chao CH, WU W. Thermal-fluid transports in
a five-layer membrane-electrode assembly of a PEM fuel cell. J Power Sources 2006;163:450–9.
[27] Eckl R, Grinzinger R, Lehnert W. Current distribution mapping in polymer electrolyte fuel cells – a finite element analysis of measurement uncertainty imposed by lateral currents. J Power Sources 2006;154:171–9.
[28] Shimpalee S, van Zee JW. Numerical studies on rib & channel dimension of flow-field on PEMFC performance. Int J Hydrogen Energy 2007;32:842–56.
[29] Hwang JJ. Thermal-electrochemical modeling of a PEM fuel cell. J Electrochem Soc 2006;153:A216–24.
[30] Hwang JJ, Chao CH. Modeling of two-phase temperatures in a two-layer porous cathode of polymer electrolyte fuel cells. Int J Hydrogen Energy 2007;32:405–14.
[31] Martin J, Oshkai P, Djilali N. Flow structures in a U-shaped fuel cell flow channel: quantitative visualization using particle image velocimetry. ASME J Fuel Cell Sci Technol 2005;2:70–80.
[32] Spernjak D, Prasad AK, Advani SG. Experimental investigation of liquid water formation and transport in a transparent single-serpentine PEM fuel cell. J Power Sources 2007;170:334–44.
