COMPARISON OF COMPRESSION MOLDING AND SELECTIVE LASER SINTERING PROCESSES IN MAKING COMPOSITE BIPOLAR PLATES FOR PEM FUEL CELLS

psithurismaccountantUrban and Civil

Nov 29, 2013 (3 years and 8 months ago)

139 views


1


Proceedings of the

7
th

Annual

ISC

Graduate
Research Symposium

ISC
-
G
RS

20
1
3

April

24
, 201
3
,
Rolla
,
Missouri

COMPARISON OF COMPRE
SSION MOLDING AND SE
LECTIVE LASER SINTER
ING
PROCESSES IN MAKING
COMPOSITE BIPOLAR PL
ATES FOR
PEM
FUEL CELLS



Ehsan Taghipour

Department of
Mechanical and Aerospace Engineering

Missouri University of Science and Technology
,
Rolla, MO 65
409


ABSTRACT

Bipolar plates are key components of Proton Exchange
Membrane (PEM) fuel cells.

To

attain

electrical conductivity
and adequate

mechanical strength
,
graphite
-
polymer composite
plates were

manufactured using Compression Molding (CM),
whic
h is suitable for mass production, and Selective Laser
Sintering (SLS), which is suitable for making prototypes. In
this paper, the electrical conductivity and flexural strength of
the bipolar plates fabricated using the CM process versus
constitutive mate
rials were experimentally studied. The
properties of bipolar plates fabricated using three
series of CM
process

were compared with those of plates f
abricated using the
SLS process
,

and the SEM images were used to illustrate the
microstructures of selective

fabricated specimens. The results
showed that SLS was able to fabricate bipolar plates with very
good mechanical and electrical properties compared to the
CM
process
. Moreover,
mixing

solid polymer resin rather than
liquid epoxy resin with solid constitut
ive materials

and using
graphite with longer aspect ratio
enhanced the mechanical
strength significantly
, and
improved the electrical conductivity
of the composite plate
, respectively
.

1.

INTRODUCTION

Proton exchange membrane
(PEM) fuel cells

are attractive
o
ptions for alternative energy sources to fossil fuels

due to
promising features
that allow them to generate

electrical energy

without undergoing combustion
,

such as

their

high power
density, relatively low operating temperature,
and ability to
convert fuel

to water as the only byproduct [1,2,3].
In a fuel
cell stack, bipolar plates are key components as they

distribute
the fuel and oxidant within the cell
,
separate the individual cells
in the stack
,
carry current away from the cell
, and
withstand the
clamp
ing forces of the stack asse
mbly [3,4,5].

The b
ipolar plate is the heaviest component of the PEM
fuel cell
,

account
ing

for about 80% of the total weight of
a

fuel
cell
stack. It is also a very costly component. A
ccording to a
cost analysis

study
, about 45% of
a

fuel cell stack
’s

cost is
incurred by the bipolar plate [1].

Therefore, quality
improvement and weight and cost reductions are critical issues
in

t
he development of PEM fuel cell bipolar plates
.

The
p
rocessing methods and materials used
to

manufacture

bipolar
plates determine the final properties of the products, so they are
of great importance for any manufacturer or researcher

of the
bipolar plates
.

Graphite
-
polymer

composites are
promising

materials
for
making PEM fuel cell bipolar plate
s
because

they have the
advantages of
good

electrical,
mechanical,
and thermal
properties
;

they are also light weight and have

high corrosion
resistance

in the highly
-
acidi
c PEM fuel cell environment

[6,7].
C
onductive materials
such as

flaky natural graph
ite,
expanded
natural graphite,
synthetic graphite, and carbon black
were

used
by

many
researchers

to impart good electrical conductivity to
composi
te bipolar plates [3,5,6,8
-
14].
Moreover, carbon fiber
was

employed to improve the mechanical

integrity of b
ipolar
plates
.
The Department of Energy

(DOE)
proposed a technical
target
for

bipolar plates for the year 2010 [1
5
], in which the
main requirements
were to achieve

electrical conductivity
greater than 100 S/cm and flexural strength greater than 25
MPa.

Com
pression molding (CM) is
a

common method for
making polymer composite bipolar plates [16]
and

is very
suitable for mass production.

The effect
s

of
conductive and
reinforcing

particles on bipolar plate properties
including

electrical conductivity and flexural strength h
ave been studied
experimentally by various researchers [3,5,6,8
-
13]. Selective
Laser Sintering (SLS), an additive manufacturing

technique,
has been studied
in making

graphite composite bipolar plates
for PEM
fuel cell
s

[
14,17
-
19
].

In
the
SLS process,
a

mixture
of
graphite material and binder in a powder bed
is scanned by
laser
,

and the molten binder bonds graphite particles together to
form 3D parts layer by layer.

The SLS process is capable of

building

compl
ex flow fields for bipolar plates
and consuming
less time and
financial resources

when

using for making
prototypes
.

The goal of this study
was

to compare the
CM

process,
which

is
a conventional manufacturing method
suitable for
mass production, and
the
SLS

process, which is a modern
manufacturing method

(Additive Manufacturing)
suitable for
making prototypes
.
In
this paper,
three
series of the CM process

and the
SLS
process
were compared in terms of making
carbon
-
polymer composite bipolar plates. Hence,
t
he

electrical
conductivity and mechanical strength
results
obtained using the
CM process

were

compared with the results
previously

obtained
using

the SLS process

[14].
In this way, t
he effects of
constitutive materials
,

including natural graphite (NG),
expanded natural graphite (EG),
synthetic graphite (SG),
carbon black (CB), and carbon fiber (CF)
,

on the electrical
conductivity and mechanical strength of composite bipolar
plates for PEM fuel cel
ls
were

experimentally studied.


2


2.

MATERIALS AND PROCES
SES

The materials and compositions used for making composite
bipolar plates using the CM process are chosen to match those
used previously in the SLS process [14] in order to allow more
meaningful comparisons between the two processes
.


2.1
.

Conductive and re
inforcing constitutive materials

In the first series of the CM experiments and in the SLS
process, n
atural graphite (3610), synthetic graphite (4437),
carbon black
(5303) and carbon fiber (AGM99),
obtained from
Asbury Graphite Mills Inc. (New Jersey, USA)
, were
employed.

Then, one more CM process was performed using
e
xpanded natural graphite (3807).
Natural graphite is the
primary electrical conducting material, but it has poor
wettability with the liquid resin
.

A
sbury 3807 is manufactured
by expanding ex
pandable graphite, compressing the resultant
worms, and then milling this material.

Carbon black,
manufactured by the combustion or thermal decomposition of
hydrocarbon fuel under reducing conditions, has a very small
particle size and thus a ver
y large sp
ecific surface area.
Carbon
fiber is used to enhance the mechanical strength of composite
bipolar plates.

Properties of these materials, as provided by the
manufacturer, are given in Table 1.



Table
1

Properties of the
constitutive materials

Properties

NG(3610)

EG(3807)

SG(4437)

CB(5303)

CF(AGM99)

Particle size
(
𝜇𝑚
)

75
-
150

50

10
-
45

0.03


7 (dia.) x
150 (length)

Density (g/cc)

2.26

2.26

2.26

1.8

1.75

Surface Area
(
𝑚

/
𝑔
)

1.27

23.0

11.46

254

1.87

Bulk
conductivity
(
𝑆
/
𝑐𝑚
)

20.00

28.57

17.24

3
.
23

8.33


2.2
. Polymer

resin

The thermosetting epoxy resin is inexpensive and has good
mechanical strength, hardness, and thermal and chemical
resistance.
E
poxy EPON
TM

Resin 828, an undiluted, clear,
difunctional bisphenol
-
A/epichlorohydrin derived liquid epoxy
resin, was used in the first series of the CM experiments with
the addition of EPIKURE
TM

3230 curing agent. Then, two more
series of
CM experiment were conducted

using solid epoxy
EPON
TM

Resin 2002 with curing agent P
-
101 to make the
bipolar plate specimens. EPON
TM

Resin 2002 is a solid
bisphenol
-
A/epichlorohydrin epoxy resin.
The density of
EPONTM 828 is 1.16 g/c
c
, and the density of EPON
TM

2002 is
1.19 g/c
c
, acc
ording to the manufacturer, i.e. MOMENTIVE
Inc.


2.2. Manufacturing process

A 250 kN hydraulic press with hot platens bought from
MTI Corp. was used to compress the composite powder in a
circular mold. A temperature control unit on the apparatus
digitally

controlled the

temperature of the hot platens.
Two
cylindrical steel molds (of 3
-
inch diameter) and cylindrical
punches were used to create the

samples.
With a band saw, the
fabricated composite plates were cut into several specimens
with dimensions accor
ding to appropriate ASTM standards
.

In the first CM process, the solution mixing was used to
mix the graphite and carbon particles with the liquid epoxy
resin and its hardener.

Because a 35 vol.% binder was used in
all of the SLS experiments, a 35 vol.% re
sin was used in all of
the CM experiments. The epoxy resin and hardener were added
to an acetone solvent in a beaker and stirred by a mechanical
stirrer for several minutes. Next, the reinforcing constituents,
i.e.,
NG
,
SG
,
CF
, and
CB
, were added to the so
lution according
to the specific composition data. The mixture was stirred for
over one hour until the acetone evaporated. Then, the beaker
was kept under a fume hood for several hours to evaporate th
e
remaining acetone completely, and next

was maintained
inside
an oven at 100
o
C for half an hour to make sure that the
composite powder was completely dry and
ready to be
compression molded.
Compression molding was performed
under pressure of 54 MPa and temperature of 200
o
C for
3
0
minutes

of molding time
.


EPON
TM

2002 solid epoxy resin was used during the
second and third CM processes. The powder epoxy resin,
EPIKURE
TM

P
-
101 curing agent, and the powder material
constituents including the above
-
mentioned graphite and
carbon materials were added together acco
rding to the
composition prescription and then jar milled for over 12 hours
to create a homogenous mixture. The mixed powder was then
compression molded under pressure of 54 MPa and temperature
of 200
o
C for 10 minutes. In the third series of the CM
exper
iments,
EG

was used instead of
NG

and added to the solid
mixture.

The four
-
probe technique was employed for measuring
electrical conductivity according to the ASTM C611 standard.
Six specimens were selected for data measurement, and the
average values
we
re

reported for the conductivity. The three
-
point bending test was performed to measure flexural strength.
In accordance with ASTM D790
-
10, four tests were conducted
,
and the average values were reported.
The microstructures of
the samples were obtained us
ing the Hitachi S
-
4700 FE
-
SEM.

To distinguish each of the series of the CM experiments,
CM1 refers to the first experimentation, CM2 denotes the
second experimentation, and CM3 refers to the latest
experiment.

3.

RESULTS AND DISCUSSI
ON

The
effects

of the cons
titutive materials on the electrical
conductivity and flexural strength of the composite bipolar
plates are
evaluated as follows.


3.1.

Effect of synthetic graphite

The microstructures of parts
can be seen

in Fig.
1
, in which

the smaller
particles are SG. Also
, in Fig. 1 flaky shape of NG
can be seen.


3


(a)
(b)

Fig. 1

SEM images showing the microstructures of (a): a brown part
made using the SLS process

and

(b
)

a specimen made using CM2
process with 15 vol.% SG, 50 vol.% NG
and 35 vol.%
binder.


Electrical conductivity and flexural strength variations with
different volume fractions of SG are depicted in Figs.
2
(a) and
2
(b), respectively. Electrical conductivity is significantly
influenced by the size and shape of particles. In our
experiments, as

the volume fraction of graphite is high, the
intra
-
particle conductivity of large NG particles overcomes the
effect of direct contact between the small SG particles. Further,
as for the particle shape, flaky NG particles form surface
-
to
-
surface contacts,
but spherical SG particles have generally
point
-
to
-
point contacts, greatly lowering the electrical
conductivity [21]. Hence, the current conduction among NG
particles is reduced with the introduction of SG particles when
the two types of particles are mixe
d together, and as the volume
fraction of SG increases, the conductivity of the bipolar plate
decreases for the CM and SLS processes.

Increasing the SG content increases the flexural strength of
the samples made using the CM processes but decreases that o
f
the samples made using the SLS process. This is because in the
CM processes, the stress distribution around small spherical SG
particles is more uniform and the strain is smaller, so the
mechanical strength improves [7]. Moreover, the interface
adhesion
between the graphite particles and polymer resin is
increased by adding the interface area between them, or by
adding the specific surface area of the graphite particles [21].
Since SG has smaller size and bigger surface area than NG, it
can form more inte
rface adhesion with the epoxy resin. In CM3
experimentation, due to less stress concentration around SG and
that EG surface area is much larger than SG surface area,
adding SG to the composition slightly increases the flexural
strength eventually. On the o
ther hand, in the SLS method,
adding smaller SG particles fills up the big pores among larger
flaky NG particles, which reduces the porosity, so less resin can
infiltrate, leading to lower mechanical strength.

The experimental results indicate that the ele
ctrical
conductivity and flexural strength obtained when using CM1
process are much lower than the SLS results and are below the
DOE target values.

The maximum electrical conductivity value
with CM1 and CM2 processes is approximately 75 S/cm, while
the max
imum value wit
h SLS is approximately 380 S/cm.

(a)


(b)

Fig.
2

Variations in (a) electrical conductivity, and (b) flexural strength
versus changes in the volume fraction of SG for the CM and SLS
processes.


Moreover, the maximum flexural strength value
with CM1
process is approximately 9 MPa, while the maximum value
with SLS is approximately 37 MPa. T
he low electrical
conductivities

in the
composite
parts

made using CM

have been
ob
served by previous researchers [10,11,32].


The measured electrical conduc
tivity and flexural strength
in the bipolar plates made using the CM and SLS processes can
be attributed to the materials and preparation methods
employed in these processes. In CM1 and CM2 processes, the
epoxy resin covers the surface of the graphite part
icles during
mixing and thus hinders the electrical connection between the
graphite particles. In the SLS process, the phenolic binder in
the form of particles is mixed with the graphite particles, and
the polymer resin is infiltrated into the brown part,
imparting
mechanical strength to the composite plate. The phenolic binder
does not remain inside the part after sintering; instead, it is
converted to carbon ashes during sintering. The carbon ashes
fill up the pores and help to increase the electrical con
ductivity.
On the other hand, in the CM processes, the mixture does not
contain sufficient epoxy resin to impart mechanical strength to
the fabricated parts. Moreover, in CM1 process, there is local
mixing of the liquid epoxy resin with the composite plate

particles (rather than with the overall powder constituents
uniformly), so agglomerates form inside the mixture after the
acetone solvent is incorporated and the mixture is dried because
of the accumulation of epoxy resin in some regions. As a result,

4


the

electrical conductivity and flexural strength decrease as the
agglomerates introduce inhomogeneity inside the structure.

It was desirable to improve the electrical conductivity of
the composite plates made using CM1 and CM2 processes;
therefore, in CM3 p
rocess, NG was replaced with EG, which
has longer aspect ratio. EG is able to form more and better
conductive networks among the graphite particles and to lessen
the effect of the current concentration and type of insulating
polymer. CM3 results in Fig.
2

indicate that the maximum
electrical conductivity value improves and increases to
approximately 131 S/cm. This value is 75% higher than the
maximum value with CM1 and CM2 processes in which NG
was used, and it satisfies the DOE target. Fig.
2
(a) also
indic
ates that the conductivity of composite specimens made
using CM3 process become larger than the conductivity of the
specimens made using SLS when SG concentration reaches 15
vol.% and beyond. Reducing the polymer concentration and/or
changing the polymer t
ype to a less resistive one can help to
improve the electrical conductivity further.

Replacing the liquid
epoxy resin with the solid epoxy resin in CM2 and CM3
experimentations enables the composite plates to acquire
significantly higher flexural strengths
. Fig.
2
(b) indicates that
the maximum flexural strength value with CM2, which is
almost
31

MPa, is increased by approximately
240
%, and the
maximum flexural strength value with CM3, which is almost
26 MPa, is improved by approximately 190% compared to the

maximum flexural strength value with CM1 process. Using
solid epoxy resin caused the composite to have a more uniform
structure, because all the constitutive materials to be mixed
were solid, and no agglomerate was formed inside the mixture.


3.2.

Effect of ca
rbon fiber

Carbon fiber
was

used
to increase mechanical strength of
the bipolar plates. Fig.
3

shows the microstructures of a sample
made using CM2 process and a brown part made using the SLS
process, in which long, thin carbon fibers are mixed with
natura
l and expanded graphite particles.

(a)
(b)


Fig.
3

SEM images showing the microstructures of (a): a brown part
made using the SLS process
and
(b
)

a specimen made using CM2
process with 15 vol.% SG, 50 vol.% NG
and 35 vol.% binder


Variations in flexural strength versus changes in the CF
volume fractions are shown in Fig.
4
.

All CM and SLS
processes show the same trend in flexural strength.

In other
words, i
ncreasing the CF ratio increases the flexural strength of
the composite sampl
es, as expected.


Fig.
4

Variations in flexural strength versus changes in the
volume fraction of CF for the CM and SLS processes.


In Fig.
4
,
the

mechanical properties of CM1 fabricated
parts using NG, EG, CF, and epoxy resin materials are lower
than those of the SLS fabricated parts and the DOE targets.
This is due to the lack of homogeneity in the mixture of
materials and the formation of graphit
e
-
carbon
-
polymer
agglomerates inside the plate structure during preparation for
the CM process, as explained before. Fig.
4

shows that the
flexural strength improves significantly and reaches the DOE
target by employing the solid epoxy resin instead of the

liquid
epoxy resin because of mo
re uniformity in the composite.

3.3.

Effect of carbon black

Fig.
5
(a) shows the microstructure of a brown part made
using SLS and parts made using CM processes, all with 16
vol.% CB, 49 vol.% NG or EG, and 35 vol.% binder. The
s
urface of NG particles is covered by the nano
-
size CB
particles in the SLS sample shown in Fig.
5
(b). However, CB
particles are present mostly between NG and EG particles in
the CM
2

samples shown in Figs.
5
(
c
) and
5
(
d
) due to mixing
and applying high
compaction pressure.


(a)


(b)



(c) (d)

Fig.
5

SEM images showing the microstructures of (a): a brown part
made using the SLS process
an
d
(b): a specimen made using
CM1
2

process CM with 16 vol.% CB, 49 vol.% NG and 35 vol.%
binder.


5



Fig.
6

shows variat
ions in electrical conductivity
with
different volume fractions of CB. In the CM processes, adding
CB to NG and EG decreases the electrical

conductivity because
CB has less intrinsic electrical conductivity than NG and EG.
However, when the CB volume fraction continues to increase,
the electrical conductivity increases because nano
-
size CB
particles fill the voids between graphite particles.

In the SLS
process, as the CB content increases, electric conductivity
decreases. This is because CB has lower intrinsic conductivity
than NG.
Also, d
uring the ball
-
milling process, CB particles
with large surface areas tend to agglomerate and cover the
w
hole surface of NG particles, as shown in Fig.
5
(b), which
impedes contact between NG particles. This is different from
the CM processes, in which CB particles fill the small voids
between NG and EG particles rather than covering their
surface; thus, elect
rical conduction can pass through these small
CB particles in CM made bipolar plates. In CM1 and CM2
experimentations, the conductivity values are lower than SLS
values; however, by using EG in CM3 process, the composite
specimen shows higher conductivity
than SLS specimen after
the introduction of 5 vol.% of CB.



Fig.
6

Variations in electrical conductivity

versus changes in the
volume fraction of CB for the CM and SLS processes.


4
.

CONCLUSIONS


The

C
ompression
M
olding (CM) and
S
elective

Laser Sintering
(SLS) process

were compared in terms of their fabrication of
composite bipolar plates for PEM fuel cells. The results showed
that the trends of electrical conductivity variations versus
changes in the volume fractions of SG and CF were the

same
for the CM and SLS processes; however, increasing the volume
fraction of nano
-
size CB increased electrical conductivity in the
CM process but decreased electrical conductivity in the SLS
process.
In the CM process, t
he flexural strength improved
sign
ificantly by adding SG and CF. For SLS
-
fabricated parts,
only CF could improve the mechanical strength, while SG
reduced the flexural strength of the composite plates. On the
basis of the obtained results, the SLS process is capable of
making highly electr
ically conductive and mechanically strong
bipolar plates compared to the CM process

when

material types
and concentrations
explained in this paper were
employed.

Hence, it
can be used

for small production of PEM fuel cells.
Moreover, to
enhance the mechan
ical strength and
improve
electrical conductivity of the composite
bipolar
plate
s

made
using the CM process
, changing material types and
concentrations

are needed. In this study, the liquid epoxy resin
in the CM

process was replaced with solid epoxy resin
to
impart higher mechanical strength because of larger uniformity
in the composite plate
, and NG was changed to EG

to make
more conductive networks
among the graphite particles

and
increase the electrical conductivity.

5
.

REFERENCES

[1]

A. Hermann, T. Chaudhur
i, P. Spagnol, Bipolar plates
for PEM fuel cells: A review. Int. J. Hydrogen Energy,
2005, 30: 1297
-
1302.

[2]

R. O’Hayre, S.W. Cha, W. Colella, F.B. Prinz, Fuel cell
fundamentals, 2nd ed., John Wiley & Sons, New York,
2009.

[3]

B.K. Kakati, D. Deka
, Differences in physico
-
mechanical behaviors of resol(e) and novolak type
phenolic resin based composite bipolar plate for proton
exchange membrane (PEM) fuel cell, Electrochimica
Acta, 2007, 52: 7330
-
7336

[4]

J.S. Cooper. Design analysis of PEMFC bipolar pla
tes
considering stack manufacturing and environment
impact, J. Power Sources, 2004, 129: 152
-
169.

[5]

L. Du, S. C. Jana, Highly conductive epoxy/graphite
composites for bipolar plates in proton exchange
membrane fuel cells. J. Power Sources, 2007, 172:
734
-
741
.

[6]

R.B. Mathur, S.R. Dhakate, D.K. Gupta, T.L. Dhami,
R.K. Aggarwal, Effect of different carbon fillers on the
properties of graphite composite bipolar plate, Int. J.
Hydrogen Energy, 2008, 203: 184
-
192.

[7]

R. Antunes, M.C.L. de Oliveira, G. Ett, V. Ett, Carbo
n
materials in composite bipolar plate for polymer
electrolyte membrane fuel cells: A review of the main
challenges to improve electrical performance, J. Power
Sources, 2011, 196: 2945
-
2961.

[8]

J.H. Lee, Y.K. Jang, C.E. Hong, N.H. Kim, P. Li, H.K.
Lee, Effect

of carbon fillers on properties of polymer
composite bipolar plates of fuel cells, J. power
Sources, 2009, 193: 523
-
529.

[9]

R. Dweiri, J. Sahari, Electrical properties of carbon
-
based polypropylene composites for bipolar plates in
polymer electrolyte membran
e fuel cell (PEMFC), J.
power Sources, 2007, 171: 424
-
432.

[10]


B. Cunningham, D.G. Baird, The development of
economical bipolar plates for fuel cells, J. Materials
Chemistry, 2006, 16: 4385
-
4388.

[11]


I.U. Hwang, H.N. Yu, S.S. Kim, D.G. Lee, J.D. Suh,
S.H. Lee, B
.K. Ahn, S.H. Kim, T.W. Lim, Bipolar plate
made of carbon fiber epoxy composite for polymer
electrolyte membrane fuel cells, J. Power Sources,
2008, 184:90
-
94


6


[12]


I. Krupa, I. Chodak, Physical properties of
thermoplastic/graphite composites, European Polymer
J., 2001, 37: 2159
-
2168.

[13]


H. Wolf, M. Willert
-
Porada. Electrically conductive
LCP
-
carbon composite with low carbon content for
bipolar plate application in polymer electrolyte
membrane fuel cell. Journal of Power Sources,
2006(153): 41
-
46.

[14]


N. Guo, M.C. Le
u, Effect of different graphite
materials on the electrical conductivity and flextural
strength of bipolar plates fabricated using selective
laser sintering, Int. J. Hydrogen Energy, 2012, 37:
3558
-
3566.

[15]

http://www1.eere.energy.gov/hydrogenandfuelcells/my
p
p/pdfs/fuel_cells.pdf, 2007.

[16]


E. Middelman, W. Kout, B. Vogelaar, J. Lenssen, E.
de Waal, Bipolar plates for PEM fuel cells, J. Power
Sources, 2003, 118:44
-
46.

[17]


S. Chen, D.L. Bourell, K.L.Wood, Fabrication of PEM
fuel cell bipolar plates by indirect SLS, P
roceedings of
the International SFF Symposium, 2004: 244
-
256.

[18]


S. Chen, J. Murphy, J. Herlehy, D.L. Bourell, K.L.
Wood, Development of SLS fuel cell current
collectors, Rapid Prototyping J., 2006, 12(5): 275
-
282.

[19]


S. Chen, D.L. Bourell, K.L.Wood,
Improvement of
electrical conductivity of SLS PEM fuel cell bipolar
plates. Proceedings of the International SFF
Symposium, 2005:458
-
467.

[20]


S. Bhlapibul, K. Pruksathorn, Preparation of graphite
composite bipolar plate for PEMFC, Korean J. Chem.
Eng. 2008, 2
5(5):1226
-
1231.

[21]

S.I. Heo, J.C. Yun, K.S. OH, K.S. Han, Influence of
particle size and shape on electrical and mechanical
properties of graphite reinforced conductive polymer
composites for the bipolar plate of PEM fuel cells,
Adv. Composite Mater., 2006; 1
5(1):115
-
126.