Pultruded FRP Plank as Formwork and Reinforcement

for Concrete Members

by
Lawrence C. Bank, Michael G. Oliva, Han-Ug Bae, Jeffrey W. Barker
and Seung-Woon Yoo
Reprinted from
Advances in Structural Engineering
Volume 10 No. 5 2007
MULTI-SCIENCE PUBLISHING CO. LTD.
5 Wates Way, Brentwood, Essex CM15 9TB, United Kingdom
ASE 10-5_Bank 15/10/07 11:46 am Page 1
1. INTRODUCTION
In the past decade there have been numerous applications
of FRP composite materials as internal reinforcements
(typically in the form of FRP rebars) for concrete
structures (ACI 2006). FRP reinforcing materials are
generally used to prevent corrosion in reinforced concrete
structures that plague conventional steel reinforced
concrete structures. In addition there have been
widespread applications of bonded FRP materials for
repairing or strengthening concrete structures. The use of
hybrid FRP/concrete members with a dual purpose of
both formwork and reinforcement, has been considered
in some studies and has been applied in a small number
of bridge decks (Deskovic et al. 1995; Hall and Mottram
1998; Hullat et al. 2003; Dieter et al. 2004; Matta et al.
2006; Cheng et al. 2005; Berg et al. 2006). A combined
formwork and reinforcement system can facilitate rapid
construction of concrete members since no conventional
formwork is needed, which requires time consuming
assembly and dismantling. In order for a smooth
pultruded FRP plank to act compositely with the
concrete, the surface of the FRP plank needs to be treated
to increase its bond properties.
This particular study was primarily motivated by the
desire to use a commercially produced pultruded FRP
Advances in Structural Engineering Vol. 10 No. 5 2007 525
Pultruded FRP Plank as Formwork and Reinforcement
for Concrete Members
Lawrence C. Bank
1,*
, Michael G. Oliva
1
, Han-Ug Bae
1
, Jeffrey W. Barker
1
and Seung-Woon Yoo
2
1
Department of Civil and Environmental Engineering,
University of Wisconsin, Madison, WI 53706, U.S.A.
2
Department of Civil Engineering,
Kwandong University, Kangwon 215-802, Korea
(Received: 21 November 2006; Received revised form: 29 June 2007; Accepted: 18 July 2007)
Abstract: A feasibility study in which the use of a commercially produced pultruded
fiber reinforced polymer (FRP) plank for both permanent formwork and secondary or
primary tensile reinforcement of a concrete structural member is described in this
paper. To achieve satisfactory bond at the interface between the smooth surface of the
FRP plank and the concrete, two kinds of aggregate, gravel and sand, were epoxy
bonded to the planks. Concrete beams using the aggregate-coated FRP planks were
fabricated and tested. Satisfactory bond between the FRP plank and the concrete was
developed which was evidenced by numerous well-distributed flexural cracks, and
ultimate capacities of the aggregate coated FRP plank specimens greater than the steel
rebar reinforced control specimen. ACI 440 equations were found to provide good
predictions of the flexural strengths but poor predictions of the shear strengths of the
FRP plank reinforced beams. ACI 318 equations, however, provided good shear
strength predictions.
Key words: pultrusion, fiber reinforced polymer, formwork, reinforcement, concrete structures, aggregate coating,
bridge decks.
* Corresponding author. Email address: bank@engr.wisc.edu; Fax: +1-608-262-5199.
05_06-518-Bank 15/10/07 10:26 am Page 525
Pultruded FRP Plank as Formwork and Reinforcement for Concrete Members
526 Advances in Structural Engineering Vol. 10 No. 5 2007
plank as a stay-in-place form and for secondary
reinforcement for flexural crack control in a new bridge
deck that was constructed without any reinforcing bars
in the concrete deck (also known as reinforcement-
free or steel-free) in Black River Falls, Wisconsin,
USA in June and July 2007. The deck was supported on
wide-flange bulb T precast concrete girders at 2.1 m
spacing that had a 914 mm gap between edges of the
flanges. A number of reinforcement free decks have
been constructed in Canada in recent years based on the
philosophy of arching action in the decks (Newhook
and Mufti 1996; Bakht and Lam 2000). It has been
shown that the ultimate failure mode of these decks is
punching shear and not flexure. In these designs no
negative moment reinforcing is provided in the slab over
the girders, however, fiber reinforced concrete is
typically used to control shrinkage cracking in the
deck. Nevertheless, it has been found that unsightly
longitudinal cracks have developed on the undersides of
the slabs between the girders in these decks under
service loads (Bakht and Lam 2002). This has led to the
use of secondary transverse reinforcement (typically
using FRP bars) to distribute the flexural cracks in the
decks. An objective of this study was to determine if the
FRP plank was able to serve as the stay-in-place form as
well as control the flexural cracking in a reinforcement
free deck, thus relieving the requirement for the
secondary crack control reinforcement. A complementary
motivation for the study was to determine if the FRP
plank could serve as the primary tensile reinforcement
of the deck if this was necessary for a deck design that
required non-metallic reinforcing bars for transverse
flexural strength to resist positive bending moment
between the girders. In this case it is expected that
conventional steel or FRP reinforcing bars would be
used to carry negative moment in the deck.
In this feasibility study, an inverted portion of a
commercially available FRP plank having integral
T-shaped ribs, shown in Figure 1, was used as the
formwork and the reinforcement for the concrete. The
FRP plank is produced in 305 and 610 mm widths and
is used in walkway or flooring applications with the flat
side facing up and has a smooth inner surface. (Silica
sand grit is often applied to the flat surface of the plank
to create a non-slip walking surface). Two types of stone
aggregate, sand and gravel, were bonded to the inner flat
surface of planks to improve the bond between the
concrete and the FRP plank. The aggregate was bonded
with epoxy to the FRP plank, which was cured, prior to
the placement of the concrete. No other flexural or shear
reinforcement was used in the beams. Two control
specimens were also tested. One control had no
aggregate bonded to the FRP plank and the other had
internal steel main reinforcing bars instead of the FRP
plank. Depending on the length of the beam, shear or
flexural failures occurred.
Two other options to achieve the composite action
between the FRP plank and the concrete were considered.
One was to apply epoxy to the surface of the FRP plank
and to place the concrete on the FRP planks before
epoxy cured. This option was felt to be unrealistic for
field applications for which the FRP plank is intended,
as there is insufficient time to coat the formwork with
epoxy during the deck casting and finishing operations.
In addition, the use of a liquid epoxy system would not
allow the workers to stand on the forms during the deck
pour which would make the casting operation almost
impossible. This was confirmed during the pouring of
the bridge deck in Black River Falls shown in Figure 2.
The need for the workers to stand on the forms between
the deck during the casting is evident. The other method
consider to develop bond was to use FRP cross rods
inserted through the webs of the FRP plank (similar to
FRP grating construction). This option was attempted in
Figure 1. 12 inch wide pultruded FRP plank
Figure 2. Casting of concrete bridge deck using pultruded FRP
planks as formwork and secondary reinforcing
05_06-518-Bank 15/10/07 10:26 am Page 526
Lawrence C. Bank, Michael G. Oliva, Hang-Ug Bae, Jeffrey W. Barker and Seung-Woon Yoo
Advances in Structural Engineering Vol. 10 No. 5 2007 527
a previous study (Ringlestetter et al. 2006) but was
found to be difficult and costly to implement in the
shop. Therefore, bonding the aggregate to the surface of
FRP plank prior to the placement of concrete was
selected as the preferable option for this study.
The feasibility of using this system was investigated
experimentally using beams of different lengths that
were tested in three point bending. The test results
from the FRP plank specimens were compared to code
predictions and to the results obtained from the control
specimens.
2. FABRICATION OF SPECIMENS
Five 230 mm wide by 178 mm deep beams (specimens
1-5) with different lengths (three beams with a 1.09 m span
and two beams with a 1.83 m span) were fabricated using
the aggregate coated FRP plank as the bottom formwork
for the concrete (see Table 1 for details). No other
tensile or shear reinforcement was used in the beams.
Longitudinal tension tests were conducted on the FRP
plank material yielding a longitudinal tensile strength,

L
=481.3 MPa and a longitudinal modulus of elasticity,
E
L
= 26,890 MPa (Ringelstetter et al. 2006). Figure 3
shows the approximate cross sectional dimensions of a
two-legged portion of the FRP plank that was used in the
tests. The 203 mm wide portion, shown in Figure 3, was
cut from the center portion of the 305 mm wide FRP
plank, shown in Figure 1, to produce a representative unit
width element of the FRP plank reinforced beam. Figure 4
shows the dimensions of the specimens.
The first step in the fabrication was to cut the FRP
planks to the appropriate dimensions. Thereafter, a
construction-grade epoxy (Concressive 1090) was
spread on the top (inner) surface of the inverted plank.
Epoxy was only placed between the ribs of the plank on
the horizontal surface. No epoxy or aggregate was
bonded to the protruding vertical ribs on either the
flanges or the webs of the ribs. The epoxy was placed
directly on the as-received pultruded material which
was not roughened or sanded. The gravel or sand was
scattered on the wet epoxy using a perforated bucket to
cover approximately 30% of the total area. To ensure
that the top surface of the aggregate was not covered by
epoxy a thin layer of epoxy was used. Sizes of the
aggregates were 3.18 to 6.35 mm for the gravel and 1.59
to 2.54 mm for the sand. Figure 5 shows the aggregate
coated FRP planks. Plywood forms were then
constructed around the FRP planks to form the sides of
the beam specimens which were then cast using
concrete from a local ready-mix commercial concrete
vendor. A Wisconsin Department of Transportation
Grade D, Size 1 (19 mm max. aggregate size) concrete
design mix, having a 28-day target compressive strength
of 27.6 MPa, was specified for all test specimens.
Two 203 mm wide by 178 mm deep beams
(specimen C1 and C2) were also fabricated as control
specimens for comparison. One of the control
specimens, C1, had a length of 1.09 m was fabricated
25
3
3
102 102
5
4
25
5
1
Figure 3. Approximate cross-sectional dimensions of potion of
FRP plank used in test beams (dimensions in mm)
Span 203
1
7
8
FRP plank
Figure 4. Dimension of fabricated FRP plank reinforced beam
specimens (dimensions in mm)
Figure 5. Aggregate coated FRP planks
05_06-518-Bank 15/10/07 10:26 am Page 527
Pultruded FRP Plank as Formwork and Reinforcement for Concrete Members
528 Advances in Structural Engineering Vol. 10 No. 5 2007
using the FRP plank as formwork, but no aggregate was
bonded to the top of the FRP plank. This control
specimen served as the control for FRP plank specimens
1-3 and was intended to show the difference between the
specimens with and without the aggregate when tested
over short spans that would be shear dominated. The
other control specimen, C2, was a conventional steel
reinforced concrete beam with a 1.83 m span, with no
shear reinforcement and no FRP plank. The beam was
reinforced with three 10 mm diameter steel rebars (414
MPa nominal yield strength) with 38 mm of clear
bottom cover. This reinforcement quantity was the
amount required for the bottom flexural transverse
reinforcement in the Black River Falls bridge deck when
a conventional flexural design approach according to
AASHTO is used (AASHTO 2007). The clear span
between girders for the bridge was 914 mm. This
control specimen was used to compare the performance
of the FRP plank beam to that of a conventional steel
reinforced beam of a longer span whose failure was
expected to be flexurally dominated. The purpose of this
investigation was to see if the aggregate coated FRP
plank could serve as the positive moment reinforcement
for the bridge deck, both from a strength (i.e., capacity)
and a serviceability (i.e., crack control) perspective.
Details of all the fabricated beams including control
specimens are summarized in the Table 1.
3. EXPERIMENTAL SETUP AND
INSTRUMENTATION
The beam test specimens were placed on two steel
bearing plates 51 mm wide in a simply supported
configuration. The center-to-center spacing of the
bearing plates was either 1.09 m (specimens 1-3 and C1)
or 1.83 m (specimens 4-5 and C2). A steel roller was
placed under the center of each bearing plate to allow
rotation at the supports. Deflections were recorded with
linear variable differential transformers (LVDTs) at the
center of the specimen. A load cell on the head of the
hydraulic actuator measured loads during the
experiments. Longitudinal strains of the FRP plank and
the concrete top surface were measured using electrical
resistance strain gauges. The strain gauges were offset
89 mm in the longitudinal direction from the center of
the beam to avoid interfering with the load head. Figure 6
shows typical experimental instrumentation used for the
testing.
The load was applied at the center of the specimen in
a three point bend test configuration by a hydraulic
actuator under manual displacement control. A 76 mm
wide steel plate was placed at the bottom of the loading
head. Gypsum cement was cast between the loaded top
surface of the specimen and the steel plate to transfer
load uniformly along the center line of the specimen.
Figures 7 and 8 show the test setup for the specimens 1
and 4, respectively.
4. EXPERIMENTAL RESULTS
4.1. Comparison Criteria
Three criteria were selected to evaluate the composite
action between the FRP plank and the concrete during
the experiments. The first criterion was the initial
cracking load of the specimen because cracking is an
important service load design criterion for bridge
decks. This was obtained from observations of the
changes in the strain gauge readings and the deflection
readings during the test. The second criterion was the
distribution of flexural cracks. In bridge decks it is
preferable to have numerous small cracks that wont
lead to deterioration, rather than a few large cracks.
Flexural (and shear) cracks were marked and counted
during the experiments. It was expected that the bond
between the FRP plank and the concrete provided by
the aggregate coating would result in closely spaced
flexural cracks having small widths would occur as
opposed to a few wide and sparsely-spaced cracks.
Table 1. Details of the fabricated beams
Compressive
Specimen Span strength, f
c
I.D. (m) Tensile Reinforcement Type (MPa)
1 1.09 Gravel coated FRP plank 25.8
2 1.09 Sand coated FRP plank 25.8
3 1.09 Sand coated FRP plank 24.5
C1* 1.09 FRP plank 25.8
4 1.83 Sand coated FRP plank 32.1
5 1.83 Sand coated FRP plank 33.5
C2* 1.83 Steel reinforcement (3-D10) 32.9
*Control specimen
05_06-518-Bank 15/10/07 10:26 am Page 528
Lawrence C. Bank, Michael G. Oliva, Hang-Ug Bae, Jeffrey W. Barker and Seung-Woon Yoo
Advances in Structural Engineering Vol. 10 No. 5 2007 529
The third criterion was the capacity of the specimen and
its failure mode, which was anticipated to be either a
flexural mode for the long beams or a shear mode for
the short beams.
4.2. Initial Cracking Capacity
The initial flexural cracking load was identified by
observing the change in slope of the load-strain curves.
The load at first cracking obtained from the strain gage
data was typically less than the cracking load obtained
by visual observation. The initial cracking loads and the
strains at the initial cracking loads are shown in Table 2.
All the initial cracks were observed near the center of
the specimens. The loads obtained from each test were
converted to give the cracking moment and normalized
with respect to a 27.6 MPa compressive strength
concrete to account for the differences in the strengths
of the concrete. The normalization was performed by
multiplying the calculated moment by .
The beams (specimens 1-5) with the aggregate coated
FRP planks showed higher initial cracking capacities
compared to the control specimen reinforced by the FRP
plank with the smooth surface (specimen C1) and
compared to the specimen with the steel reinforcing bars
(specimen C2). This indicates that a coated FRP plank
can produce an increase in the initial cracking moment
capacity of a concrete beam, which may provide a
serviceability benefit for a reinforced concrete section,
particularly a bridge deck. The beams with the sand
coated FRP plank (specimen 2 and 3) showed higher
initial cracking capacity than the specimen with the
gravel coated FRP plank (specimen 1), indicating that
sand coating may provide a better interfacial bond
mechanism than the gravel. However, as only one gravel
coated beam was tested no statistically conclusive
difference between the sand and gravel coated specimens
can be deduced from these results.
4.3. Distribution of Flexural cracks
Current design codes allow a concrete structure to crack
under service loads, however, the width of the cracks is
typically limited to a prescribed value or otherwise
controlled by reinforcing spacing requirements. Since a
larger number of cracks will lead to narrower crack
widths for each crack, a comparison of the number of
flexural cracks in the specimens was an important
evaluation criterion in this study
The number of the flexural cracks was counted after
the failure of each specimen (Table 2). In the short
beams, 9-11 small flexural cracks occurred in the
aggregate coated case (specimens 1-3) and only 3 large
cracks occurred in the case without the aggregate coating
(specimen C1). This clearly shows that aggregate
coating provides a mechanism to transfer bond stresses
at the interface between the FRP plank and the concrete
to distribute the cracks. The gravel coated beam
(specimen 1) showed slightly better performance than
the sand coated beams (specimen 2-3). Maximum crack
width versus load for the short beams are shown in
Figure 9. In the long beams (specimens 4 and 5), 22-29
flexural cracks were measured. This was similar to that
of the steel control beam (specimen C2) as shown in
Table 2. This demonstrates that the aggregate coated
FRP plank can serve as an effective tensile reinforcement
and can distribute flexural cracks in a similar manner to
internal steel reinforcements.
4.4. Capacities and Failure Modes
The specimens with the aggregate coated FRP planks
failed in a number of different modes. The failure modes
f
c
27 6 . MPa
Figure 6. Typical instrumentation for tested specimens
Figure 7. Experiment setup of specimen 1 prior to loading
Figure 8. Experiment setup of specimen 4 prior to loading
05_06-518-Bank 15/10/07 10:26 am Page 529
Pultruded FRP Plank as Formwork and Reinforcement for Concrete Members
530 Advances in Structural Engineering Vol. 10 No. 5 2007
Figure 9. Crack widths versus load for short beams
Table 2. Test results
# of
Initial Deflection flexural
cracking Normalized initial at failure cracks
Specimen Span load cracking Failure load load at both Failure
I.D. (m) (kN) moment
a
(kN-mm) (kN) (mm) side mode
1 1.09 15.6 4,395 63.2 3.56 11 Shear
2 1.09 20.9 5,909 66.7 4.22 9 Shear
3 1.09 22.2 6,451 57.4 4.22 9 Shear
C1 1.09 15.1 4,271 160.9 0.43 3 Flexural
(40.9
b
) (15.3
b
)
4 1.83 12.5 5,276 59.6 16.99 22 Hybrid
5 1.83 12.0 4,983 65.8 17.91 29 Hybrid
C2 1.83 8.0 3,356 38.3 54.48 22 Flexural
(27.8
c
) (8.56
c
)
Measured strain at initial Measured strain at
cracking load failure load
() ()
Specimen Concrete FRP Concrete FRP
I.D. (Compression) (Tension) (Compression) (Tension)
1 74 37 740 3,000
2 164 179 812 4,203
3 180 176 802 3,621
C1 100 139 187 619
(499
b
) (2835
b
)
4 197 151 2,091
d
6,610
5 188 243 3,047 6,496
C2 112 - 3,804
(1,038
c
) -
a
Normalized with respect to concrete with 27.6 MPa compressive strength
b
Value given for secondary peak
c
Value given for yield load
d
Concrete crushing did not occur in the gauge location and hence the gage showed a lower value than at the failure location.
05_06-518-Bank 15/10/07 10:26 am Page 530
Lawrence C. Bank, Michael G. Oliva, Hang-Ug Bae, Jeffrey W. Barker and Seung-Woon Yoo
Advances in Structural Engineering Vol. 10 No. 5 2007 531
were shear failure for specimens 1-3, flexural failure with
a significant slip at the interface of the FRP plank and the
concrete for specimen C1, flexural/shear failure for
specimens 4-5 and flexural failure with yielding of the
steel rebars followed by concrete crushing for specimen
C2. Capacities of the tested specimens are listed in Table 2.
Figures 10 and 11 show the specimens after the tests,
Figures 12 and 13 show the load vs. defection curves for
the tested specimens and Figures 14 and 15 show the load
vs. strain curves for the tested specimens. As can be seen
from Table 2, the control specimens showed less capacity
than the aggregate coated FRP plank specimens. The
moment capacities of the short beams at cracking (4,395,
5,909, 6,451, and 4,271 kN-mm)) can be compared with
the design service load moment for the Black River Falls
bridge which (4,316 kN-mm) which was calculated using
a AASHTO LRFD (2007) truck loading with a unit strip
width for a clear span of 914 mm.
(a) Shear failure of specimen 1
(b) Shear failure of specimen 2
(c) Shear failure of specimen 3
(d) Flexural failure of specimen C1
Figure 10. Specimens 1-3 and C1 after test
05_06-518-Bank 15/10/07 10:26 am Page 531
Pultruded FRP Plank as Formwork and Reinforcement for Concrete Members
532 Advances in Structural Engineering Vol. 10 No. 5 2007
Distributed flexural cracks occurred during the initial
loading stages of specimens 1-3. As the load approached
the maximum value, a number of the flexural cracks
developed into diagonal shear cracks. Finally a critical
diagonal shear crack formed that lead to the beams failing
in shear, as seen in Figures 10 (a)-(c). Partial debonding
between the FRP plank and the concrete was observed at
the midspan during the tests for all the beams. However,
there was no evidence of any slip of the FRP plank from
the concrete at the end of the beam following the shear
failure of the specimens. The compressive and tensile
strain levels recorded by the strain gauges (less than 0.001
for concrete in compression and 0.0045 for FRP in
tension) are evidence that there was no concrete crushing
or FRP tensile failure at the ultimate load.
The control specimen for specimens 1-3 (specimen
C1) failed in a flexural mode with slip between the FRP
plank and the concrete. A single wide flexural crack that
propagated almost to the top of the beam occurred at
midspan can be seen in Figure 10 (d). When specimen C1
was loaded after reaching its first peak, it continued to
resist additional load and a secondary peak was seen. This
secondary load carrying regime seemed to be due to the
flexural resistance of the FRP plank itself with some force
transfer to the concrete due to friction and the shape of the
plank. No shear cracks were observed in the specimen.
The specimen was felt to have lost its fully composite
action at the first load peak due to the previously noted
slip. The first peak load was, therefore, selected as an
ultimate capacity of this specimen. The load deflection
curves of the 1.09 m beams are shown in Figure 12.
Specimens 4-5 failed in a hybrid mode of shear and
flexure at the failure load, after the occurrence of the
distributed flexural cracking. This failure mode was felt
to have occurred since a significant diagonal crack
developed simultaneously with visible concrete crushing
near the loaded area, as seen in Figures 11 (a) and (b).
This can also be classified as a shear-compression
failure. At failure the compressive strain level at the top
surface of the concrete was close to or greater than 0.003
which is generally accepted to be near the failure strain
of concrete in compression. The control specimen
(specimen C2) for specimens 4-5 failed in a flexural
mode with yielding of steel reinforcement followed by
crushing of concrete which is the typical failure mode of
an under-reinforced steel reinforced beam, as seen in
Figure 11 (c). The ultimate capacity for the specimen C2
was 58~64% of specimens 4 and 5 and demonstrates that
the aggregate coated FRP plank can be used as tensile
reinforcement instead of the steel reinforcement for a
(a) Hybrid failure mode of specimen 4
(b) Hybrid failure mode of specimen 5
(c) Flexural failure of specimen C2
Figure 11. Specimens 4-5 and C2 after test
05_06-518-Bank 15/10/07 10:26 am Page 532
Lawrence C. Bank, Michael G. Oliva, Hang-Ug Bae, Jeffrey W. Barker and Seung-Woon Yoo
Advances in Structural Engineering Vol. 10 No. 5 2007 533
beam designed to achieve the capacity of the steel
reinforced beam. It can be seen from the load-deflection
plot in Figure 13 that even though the load at failure in
the FRP plank beam at failure was larger than that of the
steel reinforced beam at failure, the deflection at failure
in the steel reinforced beam was much larger than that of
the FRP plank reinforced beam. This phenomenon of a
higher load carrying capacity and a lower deformability
in FRP reinforced beams is a result of the linear elastic
nature of FRP reinforcing materials and is accounted for
in FRP design codes (ACI 2006).
5. COMPARISON OF EXPERIMENT TO
CODE PREDICTION
Test results were compared to current ACI 318 and ACI
440 code predictions and listed in Table 3. Comparison
with the AASHTO bridge design code was not possible
since the current AASHTO code does not have
provisions for FRP reinforcements. The prediction of
the capacities from the code equations was performed
based on assumptions that FRP plank did not cause an
increase in the cracking load of the beams and that full
composite action existed between the FRP plank and
Figure 12. Load-deflection plots for specimens 1-3 and C1
Figure 13. Load-deflection plots for specimens 4-5 and C2
05_06-518-Bank 15/10/07 10:26 am Page 533
Pultruded FRP Plank as Formwork and Reinforcement for Concrete Members
534 Advances in Structural Engineering Vol. 10 No. 5 2007
the concrete. ACI 318 predictions for the cracking
capacity were conservative except for the steel reinforced
specimen C2. Compared to the experimental results,
ACI 318 equations for shear capacity of the beams
predicted better values than those of ACI 440. The ACI
440 equation for the shear capacity is a function of the
distance from extreme compression fiber to the neutral
axis in the cracked elastic section while ACI 318
equation for the shear capacity is a function of the
distance from extreme compression fiber to the centroid
of the tensile reinforcement. When using ACI 318 or
ACI 440 for the calculation of the shear and flexural
capacities of the FRP plank reinforced beams the
resultant tensile force in the FRP was assumed to be
applied at the centroid of the FRP section which was
located at a height of 12.78 mm from the bottom of the
Figure 14. Load-strain plots for specimens 1-3 and C1
Figure 15. Load-strain plots for specimens 4-5 and C2
Table 3. Comparison of the test results to code predictions
Initial cracking
load Failure load
(kN) (kN)
ACI 318 ACI 440 ACI 440
Specimen (Shear (Shear (Flexural
I.D. Test ACI 318 Test failure) failure) failure)
1 15.6 12.5 63.2 56.5 36.5 109.4
(Shear failure)
2 20.9 12.5 66.7 56.5 36.5 109.4
(Shear failure)
3 22.2 12.0 57.4 55.2 36.0 105.4
(Shear failure)
C1 15.1 12.5 16.9 NA NA 12.5
a
(Flexural failure)
4 12.5 8.5 59.6 63.2 38.7 75.6
(Hybrid failure)
5 12.0 8.5 65.8 64.5 39.1 77.4
(Hybrid failure)
C2 8.0 8.5 38.3 52.0 - 24.5
b
(Flexural failure)
a
For C1 the flexural failure load is taken as the cracking load per ACI 318 or 440.
b
Calculated with the assumption of full composite action between the FRP plank and concrete. For C2 the flexural capacity was calculated using ACI 318.
05_06-518-Bank 15/10/07 10:26 am Page 534
Lawrence C. Bank, Michael G. Oliva, Hang-Ug Bae, Jeffrey W. Barker and Seung-Woon Yoo
Advances in Structural Engineering Vol. 10 No. 5 2007 535
plank. The total area of the FRP plank in the 203 mm
wide portion was 1,339 mm
2
. The design strength and
the design modulus were taken as, f
fu
= 481.3 MPa
and E
f
=26,890 MPa respectively (i.e., no environmental
or resistance factors were used in the calculations
and the average strength was assumed to be the
guaranteed strength obtained from testing of the
FRP). All the FRP plank beams were over-reinforced
beams in flexure.
6. CONCLUSIONS
A feasibility study was performed to investigate if an
aggregate coated FRP plank could serve as both a stay-
in-place form and either secondary or primary tensile
reinforcement for a concrete beam. The aggregate coated
FRP plank reinforced beams performed as well or better
than a steel reinforced control beam in terms of initial
cracking moment capacity, the ability to distribute
flexural cracks, and the ultimate load carrying capacity.
The epoxy bonded aggregate coating at the surface of the
FRP plank was essential for developing composite
action. The use of the FRP plank without the surface
treatment as a tensile reinforcement showed significant
slip between concrete and the FRP plank, no distributed
cracking and considerably less capacity. Sand and gravel
aggregate were used in this study and both were found to
be acceptable. Comparisons with code predictions
indicated that the ACI 440 design guide provides an
acceptable method to predict the flexural capacity of
aggregate coated FRP plank reinforced beams, while
ACI 318 provides an acceptable method to predict their
shear capacity. Since it has been demonstrated that the
aggregate coated FRP plank can be used as both a
secondary and primary tensile reinforcement, its
constructability advantages as a stay-in-place form can
also be realized. This will lead to construction efficiency
and may also produce a more durable bridge deck.
While this study only included a small number of
specimens and no definitive statistical conclusions can
be reached from the results they were sufficiently
convincing to enable the use of a FRP plank stay-in-
place form and reinforcement system in a pilot
reinforcement free bridge deck that was constructed in
Wisconsin in 2007.
ACKNOWLEDGEMENTS
Funding for this project was provided by the Federal
Highway Administration (FHWA) through the
Innovative Bridge Research and Construction Program
(IBRC) and by the Wisconsin Department of
Transportation (WisDOT). Special thanks are due to
Finn Hubbard and Scot Becker of WisDOT and to
Strongwell for supplying the FRP materials for the study.
REFERENCES
AASHTO (2007). AASHTO LRFD Bridge Design Specifications, 4
th
Ed., American Association of State Highway and Transportation
Officials (AASHTO), Washington D.C.
ACI (2005). Building Code Requirements for Structural Concrete,
ACI 318-05, American Concrete Institute, Farmington Hills, MI,
U.S.A.
ACI (2006). Guide for the Design and Construction of Structural
Concrete Reinforced with FRP Bars, ACI 440.1R-06, American
Concrete Institute, Farmington Hills, MI, U.S.A.
Bakht, B. and Lam, C. (2000). Behavior of transverse confining
system for steel-free deck slabs, Journal of Bridge Engineering,
ASCE, Vol. 5, No. 2, pp. 139147.
Bakht, B. and Lam, C. (2002). Closure to behavior of transverse
confining systems for steel-free deck slabs, Journal of Bridge
Engineering, ASCE, Vol. 7, No. 1, pp. 6870.
Berg, A.C., Bank, L.C., Oliva, M.G. and Russell, J.S. (2006).
Construction and cost analysis of an FRP reinforced concrete
bridge deck, Construction and Building Materials, Vol. 20, pp.
515526.
Cheng, L., Zhao, L., Karbhari, V.M., Hegemier, G.A. and Seible, F.
(2005). Assessment of a steel-free fiber reinforced polymer-
composite modular bridge system, Journal of Structural
Engineering, ASCE, Vol. 131, No. 3, pp. 498506.
Deskovic, N., Triantafillou, T.C. and Meier, U. (1995). Innovative
design of FRP combined with concrete: short-term behavior,
Journal of Structural Engineering, ASCE, Vol. 121, No. 7,
pp. 10691078.
Dieter, D.A., Dietsche, J.S., Bank, L.C., Oliva, M.G. and Russell,
J.S. (2002). Concrete bridge decks constructed with FRP stay-
in-place forms and FRP grid reinforcing, Transportation
Research Record No.1814, Journal of the Transportation
Research Board, pp. 219226.
Hall, J.E. and Mottram, J.T. (1998). Combined FRP reinforcement
and permanent formwork for concrete members, Journal of
Composites for Constructions, ASCE, Vol. 2, No. 2, pp. 7886.
Hulatt, J., Holloway, L. and Thorne, A. (2003). The use of advanced
polymer composites to form an economic structural unit,
Construction and Building Materials, Vol. 17, No. 1, pp. 5568.
Matta, F., Nanni, A., Ringelstetter, T.E. and Bank, L.C. (2006).
Rapid construction of concrete bridge deck using prefabricated
FRP reinforcement, CICE 2006, Third International Conference
on FRP Composites in Civil Engineering, December 1315,
Miami, FL, USA, pp. 151154.
Newhook, J.P. and Mufti, A.A. (1996). A Reinforcing steel-free
concrete deck slab for the Salmon River Bridge, ACI Concrete
International, Vol. 18, No. 6, pp. 3034.
Ringelstetter, T.E., Bank, L.C., Oliva, M.G., Russell, J.S., Matta, F.
and Nanni, A. (2006). Cost-effective, structural stay-in-place
formwork system of fiber-reinforced polymer for accelerated and
durable bridge deck construction, Transportation Research
Record No.1976, Journal of the Transportation Research Board,
pp. 183189.
05_06-518-Bank 15/10/07 10:26 am Page 535