An Analytical Study of An FRP Deck On A Truss Bridge
An Analytical Study of An FRP Deck On A Truss Bridge
An Analytical Study of An FRP Deck On A Truss Bridge
)
FEM
Load Test
FIGURE 2. Comparison of Analytical and Experimental Results
Failure Mode Predictions
In order to predict the failure modes of the bridge, overload static, and buckling analyses were
conducted independently. The results showed that by having two trucks on the bridge side-by-side
would maximize the longitudinal strains. As the result, the same truck configurations were used in
the overload static and buckling analyses.
FRP Deck Failure
In the overload static analysis, all truck axle loads were applied on the FRP deck with an incre-
ment of one wheel load index. At each increment, Tsai-Hill failure criterion (Jones 1999) was applied
to the composite deck at different ply levels. It was assumed that initial failure of a layer indicates
laminate failure. This definition leads to a conservative assessment of capacity. The results showed
that as the applied axle loads increased, Tsai-Hill failure index approached unity, where the first-ply
failure was expected to occur.
In Figure 3, Tsai-Hill failure index for FRP deck core increased as the applied load increased.
This indicates that the FRP deck would reach its failure in the core material (transverse direction)
at the wheel load index of 7. The results also show that there would be no ply failure taken place in
both faceskins at this load level. This type of failure is entirely based on the FRP deck only.
Longitudinal and Transverse Panel Joints
Joint failure was another mechanism that must be considered for this type of deck system.
5
Theoretically, the joints were designed to transfer forces across from one panel to the others. The
results show that the combined stresses induced in the composite splice-plate at the deck failure load,
i.e. wheel load index = 7, would not result in the joint failure. The maximum interlaminar shear
stresses along the adhesive material at the wheel load index of 7.0 was approximately 13 MPa (see
Figure 4).
Wheel Load Index
1 2 3 4 5 6 7
1.2
1.0
0.8
0.6
0.4
0.2
0.0
0
FRP Core (Tran.)
FRP Core (Long.)
FRP Top Skin
FRP Bottom Skin
Tsai-Hill Failure
Buckling Failure T
s
a
i
-
H
i
l
l
F
a
i
l
u
r
e
I
n
d
e
x
(Wheel Loading / Initial Wheel Loading)
FIGURE 3. Tsai-Hill Failure Index of FRP Deck in Static Overload Analysis
0
5
10
15
20
25
0 2 4 6 8 10 12 14 16 18 20
Interlaminar Shear Stress,
13
(MPa)
D
e
p
t
h
(
m
m
)
13
(over epoxy)
Epoxy
Splice-Glue Interface
Glue-Epoxy Interface
Splice-Plate
Glue
FIGURE 4. Interlaminar Shear Stress on the Joint at Wheel Load Index = 7
Steel Floor Beams
There are two different types of failures which are possible in the floor beams. They are local
buckling failure, and member failure due to flexure and shear stresses. The latter type of failure will
be addressed first as the result of the applied static load. Considering the state of stresses in the iso-
tropic material, Von Mises failure criterion was employed. At the wheel load index of 7.0, Von Mises
6
stresses in the floor beam lies in the elastic range. Therefore, the floor beams were not expected to
reach the failure.
Finally, local buckling failure can be predicted by performing a linear buckling analysis (ABA-
QUS 2000). The same truck axle configurations were imposed on the FRP deck. Five buckling
modes of the bridge are depicted in Table 2. The first buckling mode would occur when the bridge
is subjected to the wheel load index of 1.46, which is much smaller than the FRP deck failure load.
Local buckling failure was expected to occur in the top flange of the second floor beam near the south
abutment. In addition, the interlaminar shear stress at the wheel load index of 1.46 (buckling failure)
would only reach the maximum value of 2.5 MPa (Aref & Chiewanichakorn 2001).
TABLE 2. Buckling Analysis Results
Mode 1 2 3 4 5
Eigenvalue 1.46 1.54 2.27 2.46 2.53
0.0
0.2
0.4
0.6
0.8
1.0
1.2
0 100 150 200 250 300 350 400 450
Burning Truck Transient Temperature Gradient (
o
C)
T
s
a
i
-
H
i
l
l
F
a
i
l
u
r
e
I
n
d
e
x
FRP Core (Tran.)
FRP Core (Long.)
FRP Top Skins
Tsai-Hill Failure
50
FRP Bottom Skins
FIGURE 5. Tsai-Hill Failure Index of FRP Deck at Different Temperature Gradient
Thermal-Stress analysis
At a high ambient temperature gradient (60
o
C), Tsai-Hill failure indices for the top and bottom
skinfaces were relative small. However, the maximum Tsai-Hill failure index occurred in the FRP
core (transverse direction) was approximately 0.113, which implied a large reserved capacity. From
engineering point of view, the bridge will hardly fail under the ambient thermal-induced stress condi-
tion.
As the temperature gradient increased at a certain location in the event of a fire, Tsai-Hill failure
index for the top faceskin rises more dramatically than the FRP cores and bottom faceskin. At the
stage near the glass transition temperature of vinylester resins, i.e. 504
o
C (939
o
F), Tsai-Hill failure
index for the FRP top skinface approached 0.619. The rapid rise in the failure index was resulted
by an extremely high transient temperature gradient, which was constrained to a confined area, i.e.
a burning truck. The results of the different applied transient temperature gradients can be shown
7
in Figure 5. The Tsai-Hill failure index of 0.619 combined with a truck at high temperature may lead
to a failure of the FRP deck in that specific location.
CONCLUSION
Failure mechanism in the FRP deck was verified by the concept of First-ply failure, which is
defined as a point where Tsai-Hill failure index of any single ply layer reaches unity. The analysis
showed that the initiation of failure would occur in the FRP web-core in the transverse direction. A
wheel load index associated with this failure mode was approximately 7.
According to a linear buckling analysis, the local buckling at the top flange of the second floor
beam was found to govern the overall failure mechanism of the bridge superstructure. Local buckling
failure would be triggered at the wheel load index of 1.46, while the FRP deck would be able to resist
the live-load up to the wheel load index of 7. However, this prediction was based on the assumption
that the actual shear strength of the adhesive at the field-splice was adequate.
Thermal-stress analysis indicated that the FRP deck would have a sufficient reserve capacity to
carry the live-load during a high ambient temperature at 60
o
C. In an event of fire, the top skinface
would be subjected to the higher thermal-induced stresses, which resulted in the Tsai-Hill failure
index to increase rapidly compared to any other FRP deck components. The combination of the truck
live-load with the most severe thermal effect could lead the FRP deck to reach its failure.
ACKNOWLEDGEMENTS
The work in this paper was conducted in collaboration with New York State Department of Trans-
portation (NYSDOT). The authors would like to thanks Dr.Sreenivas Alampalli for his guidance and
supporting throughout the project.
REFERENCES
Alampalli, S. and J.Kunin (2001), Load Testing of an FRP Bridge Deck on a Truss Bridge, Special
Report 137, Transportation Research and Development Bureau, NYSDOT, Albany, New York.
Aref, A.J. and Chiewanichakorn, M. (2001), The Analytical Study of Fiber Reinforced Polymer
Deck on an Old Truss Bridge, Report submitted to New York State Department of Transporta-
tion, Transportation Research an Development Bureau and Transportation Infrastructure Re-
search Consortium.
Hibbitt, Karlsson and Sorensen (2000), ABAQUS/Standard Users Manual (Version 6.1), Hab-
bitt, Karlsson & Sorensen, Inc.
Hyer, M.W. (1997), Stress Analysis of FiberReinforced Composite Materials, McGraw-Hill,
627p.
Jones, R.M. (1999), Mechanics of Composite Materials, 2nd Edition, Taylor & Francis, 519p.
MSC. Software Corp. (1999), MSC. Patran Version 9 Users Guide.