Вы находитесь на странице: 1из 7

ARTICLE IN PRESS

Ocean Engineering 35 (2008) 287293


www.elsevier.com/locate/oceaneng

Testing and evaluation of GFRP composite deck panels


P. Alagusundaramoorthy, R. Veera Sudarsana Reddy
Composites Technology Centre, IIT Madras, Chennai 600 036, India
Received 1 June 2005; accepted 22 November 2007
Available online 21 December 2007

Abstract
Fiber reinforced polymer composite deck panels are effectively used in the construction of offshore structures such as pontoons,
oating docks, oil drilling platforms, ocean thermal energy conversion (OTEC) systems and harbor structures due to their excellent
corrosion and fatigue resistance, high strength to weight ratio and stiffness to weight ratio and less maintenance cost. The main objective
of this investigation is to study the loaddeection behavior of glass ber reinforced polymer (GFRP) composite deck panels under static
loading. Three prototype GFRP composite deck panels each with a size of 3000 mm  1000 mm  300 mm were fabricated using hand
lay-up process and tested under a factored load of AASHTO HS20/IRC Class A wheeled vehicle. The deck panels were analyzed using
the standard FE software, ANSYS. Maximum deection and strain at factored load, and exural and shear rigidities were calculated in
the FE analysis and compared with the experimental data, and also with the specications given by the Ohio Department of
Transportation (ODOT), USA. From this study, it is concluded that the fabricated GFRP deck panels satised the performance criteria
specied by ODOT and can be used in berthing structures, bridges in coastal regions, offshore oil platforms, OTEC systems and also in
seismic prone areas.
r 2007 Elsevier Ltd. All rights reserved.
Keywords: Static behavior; GFRP composite deck panels; Factored load; Finite element analysis; Maximum deection and strain; Flexural and shear
rigidities

1. Introduction
The corrosion of steel is a signicant problem in offshore
structures like pontoons, oating docks, oil platforms and
ocean thermal energy conversion (OTEC) systems and in
berthing structures leading to concrete/steel deterioration
and loss of serviceability. The resistance to corrosion in
aggressive environments, good fatigue strength, high
stiffness to weight ratio, high strength to weight ratio and
capability to mold into any shape of ber reinforced
polymer (FRP) composites make them as an alternative to
conventional construction materials and structural steel in
the construction of marine structures such as offshore and
berthing structures. FRP composites will not chip like
concrete or rust like steel and the maintenance associated
with structures made of FRP composites is completely
eliminated. Since FRP deck panels are signicantly light in
weight, higher live load levels through replacement of
Corresponding author. Tel.: +91 44 2257 4276; fax: +91 44 2257 0545.

E-mail address: aspara0@iitm.ac.in (P. Alagusundaramoorthy).


0029-8018/$ - see front matter r 2007 Elsevier Ltd. All rights reserved.
doi:10.1016/j.oceaneng.2007.11.007

heavier reinforced concrete (RC)/steel decks are possible in


berthing structures and offshore oil platforms. In addition,
they have strong potential for use in earthquake prone
zones and in places where longer unsupported spans are
necessary. Despite the overall benets of using FRP deck
panels, they are not fully commercialized. The material
properties and behavior of FRP composites are not as
completely understood like conventional materials and no
standardized code is available for application. The use of
FRP composite systems in offshore and onshore structures
requires a thorough knowledge about the structural
behavior of these systems under the particular loading
and environmental conditions.

2. Review of literature
Alagusundaramoorthy et al. (2003) analyzed single and
double span pultruded FRP deck panels using ANSYS and
compared the analytical results with the experimental data
(Harik et al., 1999a ,b ,c) and the performance criteria

ARTICLE IN PRESS
P. Alagusundaramoorthy, R. Veera Sudarsana Reddy / Ocean Engineering 35 (2008) 287293

specied by Ohio Department of Transportation (ODOT).


Aref and Parsons (1999) presented a simplied procedure
for an optimum design of ber reinforced composite bridge
decks. Aref and Sreenivas (2001) conducted eld tests and
studied the dynamic response of rst FRP composite
bridge built in USA. The authors developed a nite
element model using MSC-PATRAN and analyzed using
ABAQUS. Davalos et al. (2001) presented a combined
analytical and experimental characterization of FRP
honeycomb deck panels. The authors concluded that the
equivalent orthotropic properties developed in their study
could be used for the analysis and design of the FRP
sandwich panels. Ehlen (1999) examined the life cycle cost
effectiveness of three FRP deck panels and compared with
the conventional concrete materials. Gan et al. (1999)
evaluated various cross sectional proles for pultruded
deck panels with reduced local stress and improved
buckling strength. Harik et al. (1999ac, 2000) conducted
(i) static tests on three concrete deck panels made of
concrete reinforced with glass ber reinforced polymer
(GFRP) rebars and externally bonded with GFRP
rectangular tubular sections, (ii) three ber glass composite
bridge deck panels fabricated using the cell core technology
in conjunction with SCRIMP, (iii) three pultruded FRP
deck panels and (iv) three single span contact molding
hand lay-up ber glass composite deck panels and
concluded that all FRP deck panels performed satisfactorily for AASHTO standard HS25 truck wheel load and the
factor of safety of all deck panels was found to be greater
than 5. Hayes et al. (2000) conducted static and fatigue
tests on composite deck systems that were assembled from
glass/polyester pultruded components. The authors found
that the deck system met the necessary strength performance criteria and observed that strength and stiffness of
deck did not change even after 3 million cycles of fatigue
load. Veera Sudarsana Reddy and Alagusundaramoorthy
(2004) carried out the characterization of different resin
systems, reinforcements and FRP composites, and suggested guidelines for the selection of resin and reinforcement for making FRP deck panels. In literature only few
researchers made an attempt to study the static behavior of
pultruded multicellular FRP composite deck panels. The
cost of pultruded deck systems is approximately ve times
the cost of hand lay-up deck systems. Since the cost of
multicellular hand lay-up FRP composite deck systems is
much cheaper than the deck systems manufactured by
other processes, and also not much information is available
on the static behavior of prototype hand lay-up FRP
composite decks, it is necessary to fabricate and test
multicellular FRP composite deck systems under static and
fatigue loading.
The main objective is to study the static behavior of
hand lay-up multi-cellular prototype GFRP composite
deck panels under static loading both experimentally and
analytically. Three prototype GFRP composite deck panels
were fabricated using hand lay-up process at Composites
Technology Centre, IIT Madras, India and tested under a

factored wheel load of standard AASHTO HS20/IRC class


A wheeled vehicle. The static load/deection behavior of
the decks was analyzed using nite element method (FEM)
and the results were compared with the experimental data
and also with the performance criteria specied by ODOT,
USA. A set of conclusions was drawn based on the
experimental and nite element analysis (FEA).
3. Experimental study
3.1. Fabrication of GFRP deck panels
Three prototype GFRP deck panels are fabricated using
hand lay-up process that is the simplest method used for
the fabrication of small to large size FRP composite
systems without any infrastructure and sophisticated
machinery. The cross section of a GFRP deck panel is a
multicellular rectangular section with additional stiffeners
connecting the web to the top ange (Fig. 1). The length of
deck was kept constant as 3000 mm. The width and depth
of the deck were 1000 and 300 mm, respectively. The
thickness of top ange, bottom ange and the web was
30 mm and the thickness of additional stiffener was
20.5 mm. ECR glass ber woven roving mat (WRM) of
610 gsm with vinyl ester resin was used for the fabrication
of composite deck panels. The characterization of liquid
resin, cast resin, glass ber reinforcements and GFRP
composites was carried out as per American Society for
Testing and Materials (ASTM) standards D695M-91,
D2343-67 and D2583-95, British Standards (BS) 27821970 Parts 1 and 3, and International Organization for
Standardization (ISO) standards 62-1980(E), 178-1975(E),
179-1982(E) and 180-1982 (E). The properties of GFRP
composites and resin obtained from the characterization of
materials are shown in Table 1.
Wooden molds of size equal to the inner dimensions of
cell1, cell2, cell3 and truncated triangular cells (Fig. 1) were
made using plywood. The surfaces of the wooden molds
were polished using rough and smooth les and with sand
grit paper sheets. Duco putty was applied on the polished
wooden surfaces for leveling. The surfaces of the molds
were polished again with smooth emery paper to get a
smooth nished surface. Releasing agents such as wax and
polyvinyl alcohol (PVA) were applied on the nished
surfaces of the molds and allowed to dry for fteen
Top skin

30 mm

Additional Stiffener

20

.5

30 mm

mm

30 mm

web
cell 1

cell 2

cell 3

30 mm
1000 mm

Bottom skin

Fig. 1. Cross section of deck panel.

300 mm

288

ARTICLE IN PRESS
P. Alagusundaramoorthy, R. Veera Sudarsana Reddy / Ocean Engineering 35 (2008) 287293

289

Table 1
Properties of GFRP composites and vinyl ester resin
Sl. no.

Material properties

Value

1
2
3
4
5
6
7
8
9
10
11

Longitudinal modulus of elasticity (E1)


Transverse modulus of elasticity (E2)
Shear modulus (G12)
Ultimate tensile strength
Shear strength
Major poissons ratio (n12)
Ultimate strain (eult)
Gel time of resin
Heat distortion temperature (HDT)
Peak exotherm of resin
Volume fraction of bers (vf)

27,619 MPa
22,959 MPa
2631 MPa
368 MPa
50 MPa
0.24
0.012
19 min
130 1C
118 1C
0.38

minutes. Accelerator, promoter and catalyst like cobalt


naphthanate, dimethyl amine and methyl ethyl ketone
peroxide (MEKP), respectively, were mixed with vinyl ester
resin and a layer of catalyzed resin was applied on the
surfaces of releasing agent with resin application rollers. A
layer of WRM 610 gsm was laid on the resin and on this
WRM; a layer of catalyzed resin was applied. The
entrapped air in the lay-up was removed using paddle
rollers (laminating rollers). The procedure of laying the
reinforcement, application of catalyzed resin and paddle
rollers were repeated till reaching the half of the actual
required skin thickness of each cell. After curing for 24 h,
cell1, cell2, cell3 and truncated triangular cells were
released from the respective molds. The cross section of
the deck panel was arrived by bonding cell1, cell2, cell3 and
truncated triangular cells using vinyl ester resin and the
assembly was clamped for 24 h. The thickness of additional
stiffeners was obtained while bonding and no additional
lay-up was required. The total thickness of top and bottom
skins, and end webs of the deck panel was obtained by
doing complete wrapping on the assembled cross section
of the deck panel. Fiber to resin ratio of 1: 0.85 was
maintained during the fabrication of the decks. The
fabricated GFRP bridge deck panels (Fig. 2) were
designated as FRPBD1, FRPBD2 and FRPBD3.
3.2. Testing of GFRP deck panels
GFRP deck panels are tested in the test set-up as shown
in shown in Fig. 3. The span (3000 mm) of the GFRP deck
panel was kept perpendicular to the trafc direction and
was simply supported on shorter dimensions (1000 mm).
The dead load (DL) on the deck due to future surface
wearing course was calculated as (1.65  103 MPa) 4950 N
and the same was simulated by keeping slotted weights on
the deck panel (Fig. 3). A patch load was applied (Fig. 4)
using a hydraulic jack of 1000 kN capacity. A steel plate of
size of 500 mm  250 mm was kept below the ram of the
jack and the load from the ram was distributed on the deck
through the steel plate. This load distribution represented
the wheel load on the deck. A neoprene rubber-bearing pad
with the same dimensions of the steel plate was placed

Fig. 2. Fabricated GFRP deck panels.

between the GFRP deck panel and the steel plate in order
to minimize the abrasion between the steel plate and the
deck panel while loading. The out-of-plane deections at
the top and bottom skins were measured using dial gauges
with an accuracy of 0.01 mm. Electrical resistance strain
gauges were xed on both top and bottom skins of the deck
panel at midspan, and quarter spans from either side of the
supports. Strain gauges were also xed at critical locations
at the cross section of the deck panel. The position of the
dial gauges is shown in Fig. 5. The strain gauges were
connected to a data logger and the data were recorded and
stored in a computer at each incremental load. The deck
panels were initially loaded gradually up to 10 kN and then
the load was released. This operation was repeated twice in
order to ensure the loading edges and the supports
remained in proper contact with the specimen. AASHTO
HS20/IRC Class A wheel load of 57 kN was considered as
the live load. The GFRP deck panels were gradually loaded
from zero to a factored load of 83 kN (wheel load of
57 kN+30% of impact factor+DL of future wearing
surface) and then unloaded to zero. The measured values of
maximum deection and strain during testing are given in
Tables 2 and 3.
4. Finite element analysis of deck panels
GFRP deck panels were analyzed using ANSYS, the
standard FE software. SHELL93 elements with orthotropic material properties were selected for modeling. The
element is having the capability of plasticity, creep,
swelling, stress stiffening, large deection and large strain.
The deck panel was simply supported on shorter spans and
a rectangular patch load that represents AASHTO HS20/
IRC Class A wheeled vehicle was applied on an area of
500 mm  250 mm at the center. The discretization of the
deck panel is shown in Fig. 6. The maximum deections
and strains at DL and at factored load obtained using
from FEA are shown in Tables 2 and 3, respectively.

ARTICLE IN PRESS
290

P. Alagusundaramoorthy, R. Veera Sudarsana Reddy / Ocean Engineering 35 (2008) 287293

Jack

Slotted
weights

GFRP Deck
Panel

Support

Dial Gauge

Fig. 3. Test set-up.

W = 1000 mm

Patch load

250 mm
Line support

Line support

500 mm

300 mm

Patch load

L = 2700 mm

150 mm

150 mm

Fig. 4. Patch load on GFRP deck panel.

L = 2700 mm

D1
( D20)

D3
( D12)

D2
( D11)

250 mm

W = 1000 mm

D4
( D13)

L /4

L/ 4

D6
( D16)
500 mm

D9
( D19)

D22
( D15)

D8
( D18)

D5
( D14)

D7
( D17)

150 mm

D10
( D21)

W/4

L /4

L /4

W/4

150 mm

3000 mm

Fig. 5. Position of dial gauges at the top and bottom surfaces of deck panel, Note: Dial gauges within the parentheses were xed at the bottom of deck.

ARTICLE IN PRESS
P. Alagusundaramoorthy, R. Veera Sudarsana Reddy / Ocean Engineering 35 (2008) 287293

291

Table 2
Maximum deection in GFRP deck panels under factored load
Sl. no.

1
2
3

Maximum deection at a factored load of 83 kNa

Specimen

FRPBD1
FRPBD2
FRPBD3
a

Experimental (de) (mm)

FEA (da) (mm)

ODOT (dODOT) (mm)

da/de

da/dODOT

2.404
2.233
2.141

2.193
2.193
2.193

3.375
3.375
3.375

0.912
0.982
1.024

0.650
0.650
0.650

AASHTO HS20/IRC Class A wheel load+30% (impact factor)+dead load including future wearing surface.

Table 3
Maximum strain in GFRP deck panels at dead load and factored load
Sl. no.

1
2
3

Specimen

FRPBD1
FRPBD2
FRPBD2

Maximum strain at dead load

Maximum strain at factored load

FEA

ODOT

Experimental (ee)

FEM (ea)

ODOT (eODOT)

ea/ee

ea/eODOT

0.000012
0.000012
0.000012

0.001200
0.001200
0.001200

0.000301
0.000294
0.000266

0.000224
0.000224
0.000224

0.002400
0.002400
0.002400

0.744
0.762
0.842

0.093
0.093
0.093

Fig. 6. Discretization of GFRP deck panel.

Fig. 7. Deection contour of GFRP deck panel.

The deection contour of specimen FRPBD1 subjected to


a factored load of 83 kN is shown in Fig. 7.

panel. The maximum allowable shear for a factored load of


1.3 [1.67(LL+IM)+DL] should be less than 45% of the
ultimate shear load capacity for FRP deck panels.

5. Performance criteria
6. Results and discussions
ODOT, USA specied the deection, exure and shear
criteria for FRP deck panels. The deection of the deck is
limited to span/800. The maximum allowable strain is
limited to 20% of ultimate strain under factored load of
live load (LL)+dynamic allowance factor (IM)+DL. The
maximum allowable DL strain is limited to 10% of
ultimate strain. This includes the weight of future surface
wearing course. The maximum factored load of 1.3
[1.67(LL+IM)+DL] should be less than 50% of ultimate
load capacity for FRP deck panels. Shear capacity should
be equal to or greater than that of a RC conventional deck

Load/deection curves at midspan and quarterspan are


drawn up to the factored load from the experimental data
and FEA (Fig. 8). The pattern of the curves indicate that
the analytical study under-estimates the deection upto
factored load at quarter span and at midspan. The load/
strain curves at midspan and at quarter span are drawn and
shown in Fig. 9. The maximum deections and strains at
factored load obtained from FEA are compared with the
experimental data and also with ODOT specications
(Tables 2 and 3). The maximum deections obtained from

ARTICLE IN PRESS
P. Alagusundaramoorthy, R. Veera Sudarsana Reddy / Ocean Engineering 35 (2008) 287293

292

FEA at factored load for the deck panels FRPBD1 and


FRPBD2 are 29% lower than the experimental value
except FRPBD3. The maximum deection is within the
allowable limits specied by ODOT. The maximum DL
strain of the deck panels satised the limits by ODOT. The
maximum strain obtained from FEA under factored load
for the deck panels was 1626% lower than the corresponding experimental data. The exural (EI) and shear
rigidities (GAw) of the deck panels are calculated from the
experimental data and FEA by solving
d1=2 PL3 =48 EI PL=4 GAw ,

d1=4 11PL3 =768 EI PL=8 GAw ,

(2)

where d1/2 is the deection at a distance of L/2 from the


support, d1/4 is the deection at a distance of L/4 from the
support, P is the patch load distributed over a region of
500 mm  250 mm at the center (Fig. 4) and L is the
effective span. The calculated values of exural rigidity (EI)
and shear rigidities (GAw) are shown in Table 4. The EI of
deck panels calculated using FEA is 1127% higher and
GAw is 1427% lower than the corresponding values
obtained from experimental data.

(1)
7. Summary and conclusions

90
80
70

Load (kN)

60

Deflection at mid span


FRPBD1
FRPBD2
FRPBD3
FEM

50
40

Deflection at quarterspan
30

FRPBD1
FRPBD2
FRPBD3
FEM

20
10
0
0

0.5

1.5

2.5

Deflection (mm)

Three prototype multicellular GFRP composite deck


panels with a size of 3000 mm length, 1000 mm width, and
300 mm depth were fabricated using hand lay-up process.
A rectangular patch load that represents the AASHTO
HS20/IRC Class A wheeled vehicle was applied at the
center of each deck panel and its static behavior up to
factored load was studied. The static load/deection
behavior of the decks was analyzed using FEM and the
results were compared with the experimental data and also
with the performance criteria specied by ODOT, USA.
Flexural (EI) and GAw of GFRP composite deck panels
were calculated from the experimental data and FEA. The
following major conclusions are drawn based on FEA and
the tests conducted:

Fig. 8. Load/deection curves up to factored load.

90
80
70
Load (kN)

60
Strain at mid span

50

FRPBD1
FRPBD2
FRPBD3
FEM

40
30

Strain at quarterspan
FRPBD1
FRPBD2
FRPBD3
FEM

20
10
0
0

50

100

150

200

250

300

Microstrain

Fig. 9. Load/strain curves up to factored load.

350

(i) The maximum strain at factored load obtained from


FEA and experimental data are within 20% of the
ultimate strain specied by ODOT, USA.
(ii) The maximum deection under factored load obtained from both FEA and experimental data are
within the value of span/800 as specied in the
deection criteria by ODOT, USA.
(iii) The maximum deection in FRPBD1 and FRPBD2
obtained from FEA under the factored load is 9% and
2% lower than the experimental data, and 2% higher
than the experimental data in FRPBD3.
(iv) The maximum strain obtained from FEA under
factored load is 1626% lower than the experimental
data.
(v) The exural rigidity calculated using FEA is 1127%
higher than the values obtained from the experimental
data.

Table 4
Flexural and shear rigidities of GFRP deck panels
Sl. no.

1
2
3

Specimen

FRPBD1
FRPBD1
FRPBD3

Flexural rigidity (kN mm2)

Shear rigidity (kN mm2)

Experimental
(EIe) (  1010)

FEA (EIa)
(  1010)

Experimental
(GAwe)

FEA (GAwa)

2.14147
2.19676
2.45918

2.73302
2.73302
2.73302

68,770
81,948
74,010

59,119
59,119
59,119

EIa/EIe

GAwa/GAwe

1.276
1.244
1.111

0.860
0.721
0.799

ARTICLE IN PRESS
P. Alagusundaramoorthy, R. Veera Sudarsana Reddy / Ocean Engineering 35 (2008) 287293

(vi) The shear rigidity calculated using FEA is 1427%


lower than the values obtained from the experimental
data.
(vii) The fabricated GFRP deck panels using hand lay-up
process can be used effectively in offshore structures
such as pontoons, oating docks, oil drilling platforms and OTEC systems.

References
Alagusundaramoorthy, P., Veera Sudarsana Reddy, R., Lakshman
Kumar, K.A.B., 2003. Fiber reinforced polymer composite bridge
deck panels, In: Proceedings of the International Conference on
Recent Trends in Concrete Technology and Structures, Coimbatore,
India, pp. 693698.
Aref, A.J., Parsons, I.D., 1999. Design optimization procedures for ber
reinforced plastic bridges. Journal of Engineering Mechanics 125 (9),
10401047.
Aref, A.J., Sreenivas, Alampalli, 2001. Vibration characteristics of a ber
reinforced polymer bridge superstructure. Composite Structures 52,
467474.
Davalos, J.F., Pizhong Qiao, X., Frank, Xu, Justin, Robinson, Karl, E.
Barth, 2001. Modeling and characterization of ber reinforced plastic
honeycomb sandwich panels for highway bridge applications. Composite Structures 52, 441452.
Ehlen, M.A., 1999. Life-cycle costs of ber-reinforced-polymer bridge
decks. Journal of Materials in Civil Engineering 11 (3), 224230.

293

Gan, L.H., Lin, Ye., Yu-Wing, Mai., 1999. Design and evaluation of
various section proles for pultruded deck panels. Composite
Structures 47, 719725.
Harik, I., Alagusundaramoorthy, P., Siddiqui, R., Anido, R.L.,
Morton, S., Dutta, P., Shahrooz, B., 1999a. Testing of concrete/FRP
composite deck systems. In: Lawrence C.B. (Ed.), Proceedings of the
Fifth ASCE Materials Engineering Congress, Materials and Construction, Cincinnati, OH, USA, pp. 351358.
Harik, I., Alagusundaramoorthy, P., Siddiqui, R., Anido, R.L.,
Morton, S., Dutta, P., Shahrooz, B., 1999b. Static testing on FRP
bridge deck panels. In: Leslie Jay Cohen et al., (Eds.), Proceedings of
the 44th International Sample Symposium and Exhibition, Long
Beach, CA, USA, pp. 16431654.
Harik, I., Alagusundaramoorthy, P., Siddiqui, R., Anido, R.L.,
Morton, S., Dutta, P., Shahrooz, B., 1999c, Testing of berglass
composite bridge deck panels. In: Chen, C.S., et al., (Eds.),
Proceedings of the Thirteenth International Conference on Boundary
Element Technology, Las Vegas, NV, USA, pp. 663672.
Harik, I., Alagusundaramoorthy, P., Siddiqui, R., Anido, R.L.,
Morton, S., Dutta, P., Shahrooz, B., 2000. Testing of contact molding
hand lay-up FRP deck panels. In: Markus A. E. (Ed.), Proceedings of
the 21st International Conference of SAMPLE Europe, Paris, France,
pp. 425434.
Hayes, M.D., Don, Ohanehi, John, J.L., Thomas, E.C., Dan, W., 2000.
Performance of tube and plate berglass composite bridge deck.
Journal of Composites for Construction 4 (2), 4855.
Veera Sudarsana Reddy, R., Alagusundaramoorthy, P., 2004. Characterization of FRP composite materials, In: Proceedings of the National
Conference on Advances in Materials and Their Processing (AMTP),
Bagalkot, India, pp. 8791.

Вам также может понравиться