Sie sind auf Seite 1von 12

Composite Structures 79 (2007) 180191

www.elsevier.com/locate/compstruct

Analysis of RC beams strengthened with mechanically fastened


FRP (MF-FRP) strips
Lawrence C. Bank *, Dushyant Arora
Department of Civil and Environmental Engineering, University of Wisconsin-Madison, Room 2206, 1415 Engineering Drive,
Madison, WI 53706, USA
Available online 27 January 2006

Abstract
The analytical model for reinforced concrete beams strengthened using mechanically-fastened ber reinforced polymer (MF-FRP)
pultruded strips is discussed in this paper. In this method, FRP strips, reinforced with a combination of carbon and E-glass unidirectional bers and continuous strand mats, are fastened to the concrete with steel powder-actuated (PA) fasteners and expansion anchors
(EA). The model predicts the ultimate strength and failure modes of MF-FRP strengthened beams and was developed based on results of
tests on over 75 RC beams of many dierent sizes using the MF-FRP method, that have been conducted over the last ve years. These
tests have explored numerous dierent failure modes and factors aecting the behavior of MF-FRP strengthened beams. The analytical
model can be used to proportion the strengthening system for an RC beam so as to cause the beam to fail in a unique ductile failure
mode. This procedure was used to proportion strengthening systems for large-scale beams (7.3 m long by 51 cm 51 cm) that were used
to verify the analytical procedure. The strengthened RC beams were designed to fail in a ductile manner. In the test results presented in
this paper, the strengthened beams showed increases in yield and ultimate moments of up to 25% and 58%, respectively over an
unstrengthened beam. All strengthened beams failed, as intended, in a ductile manner with the ultimate failure mode due to concrete
compression failure at large deections with the FRP strip still rmly attached. Comparisons between the analytical predictions and
the experimental results show good agreement.
 2006 Elsevier Ltd. All rights reserved.
Keywords: Analytical modeling; Beams; Fiber-reinforced polymers (FRP); Fasteners; Strengthening; Testing; Verication

1. Introduction
Strengthening, retrotting and repairing of reinforced
concrete structures using ber reinforced polymer (FRP)
composite materials has become an accepted engineering
practice. In recent years, guidelines for the design and construction of FRP strengthening systems for concrete structures have been published throughout the world by many
organizations, including the ACI [1] and the b [2]. Many
more design guides and specications are currently in production worldwide. The current state-of-the art has been

Corresponding author. Tel.: +1 608 262 1604; fax: +1 608 262 5199.
E-mail address: bank@engr.wisc.edu (L.C. Bank).

0263-8223/$ - see front matter  2006 Elsevier Ltd. All rights reserved.
doi:10.1016/j.compstruct.2005.12.001

reviewed in monographs by Teng et al. [3] and Holloway


and Leeming [4]. A survey of research on FRP strengthening of concrete beams was recently complied by Bonnaci
and Maalej [5]. Most current work related to the strengthening, retrot and repair of concrete structures with FRP
materials utilizes adhesive bonding of the FRP to the concrete substrate. This method is known in many parts of the
world as the externally-bonded (EB) method of
strengthening.
The installation of an EB-FRP system often requires
time-consuming and specialized surface preparation of
the concrete to provide a smooth surface needed to develop
adequate bond strength between the FRP strip and the
concrete substrate. The concrete typically needs to be sandblasted, cleaned and ground smooth prior to the application of the strips [6,7]. The concrete surface must also be

L.C. Bank, D. Arora / Composite Structures 79 (2007) 180191

dry and moisture free. This is often dicult to achieve in


the eld due to ambient weather conditions. The EB-FRP
systems typically require time (up to seven days) to gain
their full strength, which can delay use of the strengthened
structure [7]. Bonded systems are susceptible to brittle failures of the strengthened member by strip detachment [8,9].
The study by Bonnaci and Maalej [5] showed as many as
69% of the FRP bonded beams surveyed failed by debonding of the FRP strips. These delamination failures in
bonded systems can limit the use of the EB-FRP method
and often require complex analysis [3]. Research suggests
that a high strain gradient is found in the adhesive layer
where strips are terminated without enough anchorage
or length in the shear span [10]. To remedy this problem,
a number of authors have recommended the use of steel
anchor bolts and cover plates or composite wraps at the
strip ends [11,12].
An alternative method for strengthening reinforced concrete members using mechanical fasteners has recently been
developed [1316]. Experimental laboratory studies have
been conducted on 35 small 1.2 m long rectangular beams,
25 moderate-length 3.7 m long rectangular beams, and 15
large 8.8 m long T-beams [1316]. In addition, a 1930s
era reinforced concrete at-slab bridge in Wisconsin with
a 7.3 m span was strengthened with the MF-FRP method
and tested to failure [17,18]. Recently, four additional
bridges were strengthened and load tested in Missouri
using the MF-FRP method [19]. With the MF-FRP
method, the strengthening is obtained by attaching FRP
strips to concrete members using closely spaced steel power
actuated (PA) fasteners and/or steel expansion anchors
(EA). The method is rapid, uses conventional typically
available hand-tools, lightweight materials and unskilled
labor. In addition, unlike the bonded method, the MFFRP method requires minimal surface preparation and
permits immediate use of the strengthened structure. Previous research on this method on a variety of beam sizes has
shown promising results in terms of installation eciency,
level of strengthening achieved, and preventing strip
delamination before concrete crushing [1316].
2. Research signicance and objectives
The mechanically-fastened FRP (MF-FRP) strengthening method for reinforced concrete members has advantages of rapid installation, minimal surface preparation,
and ductile behavior of the strengthened RC member.
Extensive laboratory testing and a number of eld applications have proved the eectiveness of the method. This
paper provides an analytical method for predicting the load
carrying capacity and failure modes of RC beam strengthened with the MF-FRP method. Results of a series of test
on 7.3 m long RC beams are also provided to demonstrate
that the MF-FRP method can be used to produce a ductile
progressive failure of a FRP strengthened beam. The objective of this research was to develop an analytical model for
predicting the ultimate capacity and failure modes of MF-

181

FRP strengthened concrete beams. The analytical method


was developed based on the results of prior developmental
tests [1316]. To demonstrate the application of the analytical method concrete beams strengthened with MF-FRP
strips were tested in the laboratory. In the laboratory tests,
the load transfer mechanisms between the fasteners, the
concrete, and the FRP strips were studied in order to demonstrate that a highly-desirable ductile failure mode could
be achieved in a consistent fashion. Observations from
these studies were used to develop detailing recommendations for the MF-FRP system.
3. Analytical model for MF-FRP strengthened beams
An analytical model was developed to predict the ultimate strength of reinforced concrete members strengthened
with the MF-FRP system. The model utilizes the strain
compatibility, equilibrium and constitutive relations of
the materials. The following assumptions, many of which
are identical to those made in routine RC analysis, were
made in developing the model for the mechanically fastened FRP strips:
(1) Plane sections remain plane and full composite
action exists at the ultimate condition. This implies that a
linear strain variation exists through the depth of the
cross-section up to the ultimate stage. This assumption is
supported by strain distributions seen during the laboratory testing and reported in detail [17,18]. Although the
FRP strips are attached at discrete locations, composite
action is developed throughout the length of the beam.
Furthermore, this assumption implies that there is no relative slip between the concrete and the FRP strip when the
ultimate capacity of the section is reached. This assumption
is made for convenience of calculations of ultimate capacity. It is recognized that the bearing failure of the FRP strip
will cause the strip to slip relative to the concrete, however,
this will not signicantly aect the ultimate exural
strength of the member since the progressive bearing failure (or slotting) of the strip will maintain a constant load
in the FRP strip.
(2) After the cracking moment the concrete carries no
tensile stress, and compressive failure of concrete occurs
at a compressive strain of 0.003.
(3) The steel reinforcing bars have elastic-perfectly plastic behavior and the contribution of the compression steel
in the section is ignored.
(4) A uniform stress distribution exists in the FRP strip
which provides only a tensile force component. This
assumption is valid considering the low moment of inertia
of the strip as compared to the concrete element. This
assumption also ignores shear lag eects implying no variation in stress across the width of the strip.
(5) The FRP strip has a pseudo-elastic-perfectly plastic
behavior when attached with discrete mechanical fasteners.
This assumption implies that the FRP strip behaves in a
linear elastic fashion up until the bearing failure of the
FRP strip and thereafter the strip carries a constant stress.

182

L.C. Bank, D. Arora / Composite Structures 79 (2007) 180191

Fig. 1. Typical sustained bearing failure of the FRP strip.

While the FRP strip by itself is a linear, elastic material, the


mechanically fastened FRP strip connection has an elastopseudo-plastic behavior. This is quite dierent from the
behavior of an externally-bonded FRP strip that has a linear elastic behavior up until strip rupture or delamination.
This ductile behavior is unique to the MF-FRP method
and gives the strengthened beam its ductility. The sustained
bearing stress, fbu, is the stress at which a progressive and
sustained shear-out slotting failure is achieved and is a
key design property of the FRP strip for the MF-FRP
method. Fig. 1 shows the typical sustained bearing failure
(or slotting) in the tested specimens. Fig. 2 shows the load
versus displacement plots for typical specimens tested in
pin bearing. The bearing test data show that the FRP strip
has a pseudo-plastic response once the sustained bearing
failure is initiated. The bearing stress, fb, is calculated as
fb

Pf
d f tfrp

where, Pf is the load at which the stress is calculated, df is


the fastener diameter and tfrp is the strip thickness. When
the bearing strength, fb,u, of the strip is reached and the
slotting failure beings to develop in the strip the constant
sustained force transferred, Pf,u, to the strip is given as

Fig. 2. Load versus displacement for bearing test coupons.

P f;u fb;u d f tfrp

(6) Preexisting strains in the concrete substrate are


ignored. This assumption implies that the element is not
subjected to any loads when the FRP strip is applied.
The eect of preexisting strains can be added as explained
in ACI 440.2R [1].
(7) The mechanically fastened FRP strip has no eect on
the shear strength of the section. This implies that there is
no benecial eect due to dowel action, nor any detrimental eect due to the embedment of the PA fasteners and
anchors.
(8) The tensile force in the FRP strip at any point along
the length of the strip is equal to the bearing force carried
by the PA fasteners, Ppaf, and the end anchors, Pea, in the
shear span (or shear transfer length) of the beam. If the
shear force in the shear span is constant (i.e., as occurs in
four-point bending) the force transferred by all equal diameter fasteners is equal since the force transferred by each
fastener is a function of the bearing stress and the bearing
area as shown in Eq. (2). This means that the force transferred to the FRP strip, Tfrp, is directly proportional to
the number of fasteners, N, and the number of end
anchors, M, available at a point along the length of the
strip and is given as:
XN
XM
T frp
P paf
P ea
3
1
1
This assumption is based on equilibrium and is only strictly
valid in the elastic range assuming that all fasteners are
snug-t from the onset of the loading (which is probably
not realistic, since the hole in the FRP strip is typically
slightly larger than the fastener and some slip must occur
for all fasteners to bear on the strip.) This assumption
can, however, be reasonably extended to inelastic situations if the load transfer behavior is elastic-perfectly plastic
due to sustained bearing failure of the FRP strip. Experimental tests on multibolted joints in pultruded FRP plates
with two and three bolts have shown that the load distribution between the bolts is uniform when slotting occurs
[20,21]. The maximum uniform tensile stress in the FRP
strip, ff, in the strip can be related to the tensile force in
the strip and the cross-sectional area of the strip, Astrip,
as follows:

L.C. Bank, D. Arora / Composite Structures 79 (2007) 180191

PN
ff

PM
P paf 1 P ea
Astrip

When sustained bearing force exists in all the PA fasteners,


Ppaf,u, and end anchors, Pea,u, at the ultimate condition
then stress in the strip, ff,b, is constant and is calculated as
PM
PN
1 P paf;u
1 P ea;u
ff;b
5
Astrip
(9) When concrete compression failure occurs at the
maximum load (ultimate capacity), after the steel has
yielded, the Whitney rectangular stress distribution can
be used to determine the compressive force resultant in
the concrete.
3.1. Flexural capacity of the MF-FRP strengthened
section

3.3. Mode I sustained bearing failure in the FRP


strip (ff,b 6 ff 6 f0)
If the stress in the FRP strip calculated from Eq. (12) is
larger than the stress that will cause sustained bearing failure ff,b calculated using Eq. (5) the solution is not valid
(since the strip cannot attain the stress calculated). If
ff = ff,b and less than the open-hole strength f0, then a balanced sustained bearing condition exists where simultaneous bearing failure of the FRP strip at the fasteners
and concrete compression occurs. This is a highly desirable
failure mode since it results in a ductile failure of the
strengthened RC beam as will be shown in the test results
that follow. The location of the neutral axis, c = cb, at
which this sustained bearing failure occurs can then be
directly calculated from
cb

Assuming that the internal tension steel yields and the


concrete fails in compression prior to the FRP strip failing
or detaching, the nominal moment capacity, Mn, of the
MF-FRP strengthened RC beam can be determined from


a
a
M n As fy d s 
Afrp ff d frp 
6
2
2
where, As is the area of the steel tensile reinforcement, fy, is
the yield stress of the steel, ds is the distance to the centroid
of the steel from the top of the beam, Afrp, is the total area
of the MF-FRP strips, ff, is the stress in the MF-FRP strip,
dfrp is the distance to the centroid of the MF-FRP strip
from the top of the beam, and a is the depth of the Whitney
stress block. The depth of the neutral axis, c = a/b1, is
determined from
Ac2 Bc C 0

where,
A 0:85fc0 b1 b
B As fy Afrp Efrp ecu
C Afrp Efrp ecu d frp

8
9
10

ecu is the compressive failure strain in the concrete, assumed to be 0.003, and b is the beam width. Once location
of the neutral axis c is determined, the strain, efrp, and the
corresponding stress in the FRP strip may be calculated
from


d frp
efrp 0:003
1
11
c
ff Efrp efrp

12

183

As fy Afrp ff;b
0:85fc0 b1 b

13

Eq. (13) can be written in terms of the sustained load per


fastener using Eq. (5) as
 PN

PM
As fy n
1 P paf;b
1 P ea;b
cb
14
0:85fc0 b1 b
where n = Afrp/Astrip and is equal to the number of equal
width strips used. Once the location of the neutral axis cb
is calculated, the nominal capacity can be calculated using
Eq. (6) with ff = ff,b.
3.4. Mode II Rupture of the FRP strip at the net-section
(ff P ff,b P f0)
In this case the stress predicted in the FRP strip is greater
than the stress than will cause sustained bearing failure in
the FRP strip, however, due to an excessive number of fasteners the open-hole strength of the FRP strip controls with
ff,b > f0 leading to a tension failure (rupture) at the net section at the hole before bearing failure is initiated. The location of the neutral axis, c = c0, at which the open-hole
tensile failure occurs can be directly calculated from
c0

As fy Afrp f0
0:85fc0 b1 b

15

Once location of the neutral axis c0 is calculated, the nominal capacity can be calculated using Eq. (6) with ff = f0.
The strain in the steel should be checked to ensure that it
is greater than the steel yield strain. In this mode it is possible that the FRP strip will rupture before the concrete
fails in compression, however, if the steel has yielded the
above calculated nominal moment should be reasonably
accurate as discussed by ACI 440.2R [1].

3.2. Prediction of FRP failure modes


The calculated stress in the FRP determines the condition of the FRP strengthening strip at failure and the failure mode of the strengthened beam. Four modes are
possible as detailed below.

3.5. Mode III no failure in the FRP strip (ff 6 ff,b and
ff,b < f0)
This case may occur if the stress in the FRP strip calculated from Eq. (12) is less than the stress than will cause

184

L.C. Bank, D. Arora / Composite Structures 79 (2007) 180191

sustained bearing failure ff,b of the FRP strip and less


than f0. The stress in the strip in this situation is still
in its linear range when concrete compression failure
occurs. The nominal strength is obtained using Eq. (6)
(assuming the steel has yielded). The location of the
neutral axis is calculated from Eqs. (7)(10). This condition occurs when too large an area of FRP is used (i.e.,
too many strips). This will typically result in an uneconomical design but may have certain constructability
benets.
3.6. Mode IV delamination of the FRP strip due to
fastener Pryout (ff 6 ff,b 6 f0)
This is a unique case where closely spaced fasteners can
produce high stresses in the concrete substrate leading to
failure of the substrate. This may occur even if the stress
in the strip is below the limit that causes bearing in the
FRP strip ff 6 ff,b. It may also occur prior to yielding in
the concrete. Previous research conducted on large Tbeams with a large compression zone and closely spaced
fasteners (and having no end anchors) have revealed this
failure mode [16]. A pull-out cone-shaped failure surface (envelope) develops in the concrete when this occurs.
These surfaces overlap if the spacing of the fasteners is
too close to one another, which leads to a delamination
failure of the FRP strip. This failure can initiate either
in the interior of the beam or near the support where it
depends on the end distance of the strip. This is a similar
failure mode to that seen in externally-bonded FRP strips
where concrete delamination failure occurs. The design of
mechanical anchors in concrete is discussed in ACI 318-02
Appendix D [22]. Further research is needed to apply
these design models to predict concrete pryout failure in
MF-FRP reinforced beams and the moment capacity at
the point of pryout failure. In the absence of an analytical
model, at this time, this failure mode can be avoided with
appropriate detailing. Based on prior testing a minimum
spacing of 76 mm (3 in.) between fasteners, and a
minimum edge distance to the edge of the concrete
of 76 mm (3 in.), is recommended to avoid this failure
mode. The use of multiple strips in layers is also not recommended at this time [17]. The use of multiple lines of
fasteners in a single 102 mm (4 in.) should also be
avoided.

4. Laboratory verication study


Four beams 7.32 m (24 ft) long with nominal cross-section measurements of 508 508 mm (20 20 in.) were
designed and tested at the University of Wisconsin-Madison in order verify the proposed analytical model. A control beam (C-l) and three strengthened beams (FRP1,
FRP2SS and FRP3) were tested. Beams FRP 1 and FRP
2SS were tested to determine the exural behavior of the
beams when strengthened with two FRP strips fastened
using galvanized steel fasteners and stainless steel fasteners,
respectively. FRP 3 was tested to determine the exural
behavior of the beam strengthened with 3 strips and to
assess the performance of the MF-FRP system under
repeated loads. Table 1 shows the details of the four beams.
A concrete mix with a nominal strength of 28 MPa
(4000 psi) was used for the beams. The day of test concrete strength was close to 42 MPa (6000 psi) and is shown
in Table 1. The beams were reinforced with 3 #7 Grade 60
main bars (qs = 0.5%) and #4 Grade 60 closed stirrups at
200 mm (8 in.) spacing along the full length of the beam.
2 #4 top bars were used to stabilize the cage. The shear
reinforcement was designed to prevent shear failure of
the beam. Fig. 3 shows the steel reinforcement details.
A Hilti DX A41 powder actuated fastening system was
used to insert the powder actuated fasteners. The DX
A41 is an indirect-action gunpowder charged system. Similar, powder actuated fastening (PAF) systems are manufactured by vendors such as Ramset, Powers and
Simpson. The system consists of fasteners (called pins)
and a tool which uses a 6.8 mm (0.27 in.) gunpowder
charge to shoot the fastener into the base material. A minimum of 30 mm (1.2 in.) embedment of the fastener is recommended to achieve the design shear and normal force
capacity of the embedded fastener [23]. In previous studies
the eects of fastener type, washer, diameter, length,
embedment depth and the use of pre-drilled holes were
investigated [1316]. Based on these studies hardened galvanized steel fasteners with neoprene backed steel washers,
and 12 mm (0.5 in.) pre-drilled holes are recommended for
use with the FRP strip. Use of shallow pre-drilled holes
was shown to signicantly reduce spalling during fastener
driving and to increase tensile and shear capacity of the
embedded fasteners, and is recommended by manufacturers of PAF systems.

Table 1
Properties of laboratory beams
Beam

Width, b
(mm (in.))

Height, H
(mm (in.)

Concrete
strength, fc0
(MPa (psi))

Dist. to FRP,
dfrp (mm (in.))

Area of FRP,
Afrp (mm2 (in.2))

Fastener and
anchor type

C-1
FRP 1
FRP 2 SS
FRP 3

508 (20)
508 (20)
508 (20)
508635 (2025) Varies

501.6
514.4
514.4
514.4

41.1
41.5
41.5
41.2

515.6 (20.3)
515.6 (20.3)
515.6 (20.3)

645.2 (1.0)
645.2 (1.0)
967.8 (1.5)

X-AL-H47 KBII CS
X-CR-S44 KBII SS
X-AL-H47 KBII CS

(19.8)
(20.25)
(20.3)
(20.3)

(5960)
(6015)
(6015)
(5970)

Note: For all beams As = 1161 mm2 (1.8 in.2); distance to steel ds = 470 mm (18.5 in.); fastener spacing s = 76.2 mm (3 in.) 80 PA fasteners and 4 end
anchors per FRP strip; d1 = 25.4 mm (1 in.); S1 = 50.8 mm (2 in.); and S2 = 76.2 mm (3 in.).

L.C. Bank, D. Arora / Composite Structures 79 (2007) 180191

185

# 4 Stirrups @203.2 mm

(8") O.C.

508 mm (20")
(2) # 4 GR 60

50.8 mm
(2")
38.1mm
(1.5")

470 mm
(18.5")

508 mm
(20")

C-1
(3) # 7 GR 60

152.4mm (6")

Fig. 3. Beam dimensions and reinforcement details.

Fig. 4. End geometry of strengthened beams.

Table 2
Fasteners used in laboratory beams
Type

Material

Hilti X-AL-H 47

Zinc plated
hardened steel
Stainless
steel HRC 52
Carbon steel
Stainless steel

Hilti X-CR-S 44
Hilti Kwik Bolt II
Hilti Kwik Bolt II 304SS

Shank diameter
mm (in.)

Length
mm (in.)

4.5 (0.177)

47 (2.875)

4.0 (0.158)

44 (1.75)

12.7 (0.5)

70 (2.75)

Of the four beams tested, two were strengthened using


zinc plated (galvanized) hardened steel X-AL-H 47 fasteners and one was strengthened using X-CR-S 44 stainless
steel fasteners. In addition, either carbon steel or stainless
steel end anchors (Hilti Kwik Bolt (KB) II Expansion
Anchors) were used at the strip ends. Table 2 gives the
types and dimensions of the fasteners used. The powder
charge and the fastener length were chosen such that
30 mm (1.2 in.) of the fastener was embedded in the
concrete.
The strengthening of the beams was done in-place from
underneath the beams while they were simply-supported on
0.9 m (3 ft) high concrete blocks. Prior to fastening the
FRP strips to the beams, 4.8 mm (0.188 in.) holes were
drilled in the strips at the desired fastener locations of
76 mm (3 in.) along the length of the strip. The FRP strips,
with the fastener holes, were then positioned and held in
place on the beam sot with carpenters clamps. The concrete, was then drilled to a depth of 12 mm (0.5 in.) using a
bit with a stop, at each desired fastener location. The FRP
strip with the pre-drilled holes served as a convenient template for drilling the hole in the concrete. The PA fasteners
were then driven through the hole in the FRP strip into the
concrete using the PAF tool. A total of 80 fasteners and 4
expansion anchors were used to fasten each FRP strip. The

expansion anchors were inserted into 12.7 mm (1/2 in.)


drilled holes and tightened in the usual manner. The termination lengths and the anchor locations are illustrated in
Fig. 4. The entire attachment process was completed by
one worker in approximately 90 min per beam.
The pultruded FRP strips used with the MF-FRP system were designed to have a high longitudinal tensile
strength and stiness and to provide a ductile bearing failure at the locations of the fasteners. The FRP strips were
produced by Strongwell (Chateld, MN) under the name
of SafStripTM. The FRP strips are 102 mm (4 in.) wide
and 3.2 mm thick (1/8 in.). The reinforcing bers in the
FRP strip are a combination of unidirectional carbon
and E-glass bers and E-glass continuous strand mats.
Details of the exact ber architectures have been published
elsewhere [1719]. The resin system is a vinylester system
formulated for the pultrusion process.
To characterize the FRP strips for the MF-FRP
method, longitudinal tension and longitudinal bearing tests
were performed at the University of Wisconsin. Longitudinal tension tests (ASTM D 3039 [24]) were conducted on
25 mm (1 in.) wide at-sided strips to obtain the longitudinal elastic modulus and longitudinal tensile strength.
Open-hole longitudinal tension (OHT) tests (ASTM D
5766 [25]) were conducted to obtain the reduced longitudinal strength in the presence of a 4.8 mm (0.188 in.) diameter hole (i.e., the strength reduction due to the hole),
which exists in the strip when it is attached with a fastener.
The open-hole tensile strength is considered to be the
eective tensile strength of the FRP strip when using
the MF-FRP method as the strip will always fail at the
net-section due to the presence of the hole. Longitudinal
bearing tests were conducted on FRP coupons to determine the bearing strength of the FRP strips. Bearing tests

Table 3
FRP strip tension test data
Test

ASTM D3039
ASTM D5766

Longitudinal strength MPa (ksi)

Longitudinal failure strain

Longitudinal modulus GPa (ksi)

Average

SD

COV

Average

SD

COV

Average

SD

COV

844 (122.4)
640 (92.8)

77 (11.2)
48 (7.0)

9.1
7.5

0.0138
0.0118

0.0007
0.0009

5.1
7.6

61 (8892)

6 (800)

9.8

186

L.C. Bank, D. Arora / Composite Structures 79 (2007) 180191

Table 4
FRP strip bearing test data
Maximum load N (lbs)

Ultimate bearing stress MPa


(ksi)

Sustained bearing load N (lbs)

Sustained bearing stress MPa


(ksi)

Average

SD

COV

Average

SD

COV

Average

SD

COV

Average

SD

COV

3990 (897)

418 (94)

10.5

264 (38)

28 (4)

10.5

3550 (798)

151 (34)

4.3

234 (34)

10 (1.5)

4.3

(ASTM D 5961 [26]) were conducted in double shear on an


FRP strip having a thickness of (0.125 in.) 3.2 mm and
with a hole diameter of 4.8 mm (0.188 in.) at the center.
ASTM D 5961 denes the ultimate bearing strength as
the maximum load divided by the bearing area. For the
purposes of this study the ultimate bearing strength and
a sustained bearing stress calculated over the displacement interval from 1.3 mm (0.05 in.) to 5 mm (0.20 in.)
for each specimen was calculated. Results of the tests are
shown in Tables 3 and 4.
5. Test results
The beams were tested at the Structures and Materials
Testing Laboratory at the University of Wisconsin-Madi-

Fig. 5. Locations of strain gages of FRP strips.

son in four-point bending. A 445 kN (100 kip) servo


hydraulic MTS actuator with an MTS TestStar IIs controller was used to load the beams. The actuator was attached
to a spreader beam, which distributed the load to the two
load points. All beams had a clear span of 6.5 m (256 in.)
with 254 mm (10 in.) wide support at each end, and a
moment span of 762 mm (30 in.). All the tests were run
in displacement control. Two linear string potentiometers
on either side of the beam, and the actuator stroke were
used to measure the deection of the beam. All deection
measurement were found to be very close in value and no
twisting of the beam occurred.
Strain gages were used to measure the strain in concrete
and the FRP strips at dierent locations. For the control
beam, two strain gages were used in the moment span to
measure compressive strain in the concrete. For the
strengthened beams, an additional 8 strain gages were used
on the FRP strips at various locations near the ends, and
the midspan of the beam. Locations of the strain gages
on the FRP strips under the beams are shown in Fig. 5.
All beams, except for FRP 3 were tested under monotonic
load at a rate of 2.54 mm/min (0.1 in./min). FRP 3 was
tested under repeated load cycles to assess the performance
of the fasteners and the FRP strip under repeated loadings
in the elastic and inelastic ranges. All beams were tested to
displacements well beyond the ultimate (maximum) load to
obtain their complete load-deformation history and to
observe the FRP strip failure modes.

Displacement (in.)
0

10

480
FRP 3

440
400

320

FRP 2SS

280
FRP 1

320

240

280

200
C- 1

240

160

200

120

160
120

80

80
40

40
0
0

40

80

120

160

200

240

Displacement (mm)
Fig. 6. Moment versus midspan deection for tested beams.

0
280

Moment (k-ft)

Moment (KN-m)

360

L.C. Bank, D. Arora / Composite Structures 79 (2007) 180191

187

Table 5
Summary of beam test results
Beam

Yield moment
kN m (k-ft)

%Inc.

Ultimate moment
kN m (k-ft)

%Inc.

Moment @L/64
kN m (k-ft)

%Inc.

Ultimate
failure mode

FRP strip
failure mode

C-1

194.7 (143.6)

283.0 (208.7)

240.3 (177.2)

FRP1

230.3 (169.8)

18.2

375.9 (277.2)

32.8

359.5 (265.1)

49.6

FRP2 SS

240.0 (177.0)

23.2

389.0 (286.9)

37.5

371.4 (273.9)

54.6

FRP3

244.1 (180.0)

25.2

447.3 (329.9)

58.1

410.5 (302.7)

70.8

Concrete crushing
after steel yielding
Concrete crushing
after steel yielding
Concrete crushing
after steel yielding
Concrete crushing
after steel yielding

Fastener pullout
FRP rupture
Anchorage failure

Unstrengthened
%Increase StrengthenedControl
 100.
Control Unstrengthened

Table 6
Beam displacement and ductility data
Beam

dy mm (in.)

du mm (in.)

df mm (in.)

/ult

%Dierence

/f

%Dierence

C-1
FRP 1
FRP 2 SS
FRP 3

35
34
33
34

224
120
123
135

247 (9.73)
182 (7.16)
237 (9.32)

6.4
3.5
3.8
4.0

45.3
40.6
37.5

6.4
7.2
5.6
7.0

(1.38)
(1.35)
(1.28)
(1.33)

(8.80)
(4.74)
(4.85)
(5.31)

12.5
12.5
9.4

Unstrengthened
%Difference StrengthenedControl
 100.
Control Unstrengthened

The displacement, d, and the load, P, were recorded and


the moment versus displacement of each beam was plotted
as shown in Fig. 6. The moment M, was calculated by multiplying half the total load by the shear span of 2870 mm
(113 in.). The yield moment My (and yield deection dy)
at which the loaddeection curve deviated from its initial
linear regime, the ultimate moment Mu (and ultimate
deection du) at which the maximum load was reached,
and the moment at a displacement of 102 mm (4 in.) corresponding to a deection ratio of L/64 (noted for comparison purposes with the control beam) are listed in Table 5.
The yield and ultimate deections as well as, df, the nal
deection (the point at which FRP strengthening system
became ineective) are listed in Table 6. Ductility ratios,
calculated to compare the ductilities of the beams, are also
shown in Table 6. The ultimate ductility ratio /ult and the
nal ductility ratio /f were dened as
/ult

du
dy

16

and,
/f

df
dy

17

6. Discussion of beam test results


The ultimate failure mode (i.e., the failure mode at maximum load) in all the tested beams was concrete crushing
after yielding of the tensile steel. The strengthened beams
exhibited this ultimate failure mode at a signicantly higher
load than the control beam. The detachment of the FRP
strengthening system in the strengthened beams occurred

long after the concrete compression failure at large


displacements.
Beam C-1. The control beam, C-1, was tested to obtain
the unstrengthened capacity of the beam as a basis for
comparison with the strengthened beams. The beam failed
due to concrete compression at a large displacement. The
large failure displacement and the 30% increase in the
post-yield strength of the control beam is attributed to
the low reinforcement ratio and the strain hardening of
the steel reinforcement.
Beam FRP 1. Beam FRP 1 was strengthened with 2 FRP
strips using X-AL-H fasteners and carbon steel KBII
anchors. During the initial loading of the strengthened
beam exural cracks developed in the moment span.
Beyond yield, the beam continued to gain strength until
concrete compression failure occurred near one of the load
points. At rst, local crushing was observed under one of
the load points. This then developed into concrete compression failure in the entire moment region. A load drop
was seen at this point. The loading was continued after
the concrete compression failure and a nonlinear load
increase was seen. The load stabilized for a period of time
until a second decreasing load range was seen. The load
nally dropped o when the FRP strengthening system
detached when a number of PA fasteners pulled out in
the interior near one of the load points. The strengthened
beam showed an increase of 18% in the yield and 33% in
the ultimate capacity over the control beam. In addition
to the improved yield and ultimate capacities, FRP 1
showed increased ductility. A signicant level of load was
sustained by the beam well after the concrete had crushed
(see Fig. 6). Fig. 7 shows the strain gage data. The strain
distribution clearly indicates that tensile load is carried

188

L.C. Bank, D. Arora / Composite Structures 79 (2007) 180191

Fig. 7. Strains in concrete and FRP strip in beam FRP1.

by the FRP strip throughout the load history both prior to


and after steel yielding. Strain gage C2 shows concrete
compression strain failure at 0.5% near the load point,
while the gage C1 in the middle shows concrete compression failure at 0.3%. The maximum strain seen in the
FRP strip was 0.8% (gage F3). At the ends of the beam
(gages F1, F2, and F7, F8) small strains were recorded as
expected. A small negative strain corresponding to compression in the FRP strip near the one support is seen. Similar strain distributions in the FRP strips were obtained for
all strengthened beams [17,18].
After the test, the beam was turned upside down to view
the failure modes of the FRP strips. The FRP strips
showed a sustained bearing failure at all the fastener locations and at the end anchors. Fig. 8 shows the bearing and
shear-out failure at one end of beam FRP 1. The slot
length was maximum at the ends of the strip and
decreased toward the center of the beam. A complete
shear-out failure in one of the FRP strips at one end of
the beam was observed. At the middle of the span, under
one of the load points, fastener failure due to pullout was
seen. It is important to note that these failures of the

MF-FRP strip occurred at a large displacement of the


beam; well after the concrete compression failure.
Beam FRP 2SS. Beam FRP 2SS was strengthened with 2
FRP strips using X-CR-S stainless steel fasteners and KBII
stainless steel anchors. The beam was tested to determine if
the stainless steel fasteners, which have a dierent geometry
and washer from the X-AL-H fasteners, could be used with
the MF-FRP method. No previous testing has been conducted using stainless steel fasteners whose corrosion resistance may be of interest in applications of the MF-FRP
method, and have recently been used to strengthen a bridge
structure [17,18]. The yield strength increase was 23% while
the ultimate strength increase was 38%. The loading was
continued after the concrete compression failure and the
load stabilized for a period of time until a secondary load
drop was seen.
After the test, the beam was turned over to view the failure of the FRP strip. Similar to FRP 1, slotting or bearing
of the strip near the fasteners was clearly visible increasing
towards the end of the strip near the supports. The FRP
strip failure was due to rupture of the FRP strip at the midspan (well after the ultimate load). It is believed that at the

Fig. 8. Close-up of sustained bearing failures at PA fasteners in the FRP strip after failure of beam FRP 1.

L.C. Bank, D. Arora / Composite Structures 79 (2007) 180191

time of the installation of the stainless steel fasteners, the


stainless steel washer damaged the FRP strip due to the
lack of a neoprene backing. Once bearing failure began
in the FRP strip, an embedded washer caused the strip to
split in a horizontal plane, which caused a through-thethickness delamination of the strip. Consequently, use of
the PA fasteners without neoprene backed washers is not
recommended for applications. FRP 1 and FRP 2SS failed
in concrete compression at almost identical displacements.
FRP 2SS showed a small increase in ultimate capacity over
FRP 1, but showed a slightly reduced ductility, which cannot be considered to be statistically signicant.
Beam FRP 3. Beam FRP 3 was strengthened with 3 FRP
strips using X-AL-H fasteners and KBII anchors. The
beam was tested to determine the eect of increased FRP

25

Total load (kips)

20
15
10
5
0
0

0.2

0.4

0.6

0.8

Displacement (in.)
Fig. 9a. Beam FRP 3 response during load cycles in the elastic range.

60

Total load (kips)

50
40

189

reinforcement and to obtain data on the eect of repeated


load cycles in the elastic and inelastic ranges on the fasteners and the FRP strip. The beam was loaded up to 89 kN
(20 kips) (below yield) and then for 10 cycles at the rate
of 25.4 mm/min (1 in./min) at a mean load of 49 kN (11
kips) and an amplitude of 40 kN (9 kips). The test was then
resumed at 89 kN (11 kips) under a displacement rate of
2.5 mm/min (0.1 in/min) until the yielding of the steel reinforcement. At 178 kN (40 kips) (beyond yield) 10 more
loading cycles were applied at the rate of 25.4 mm/min
(1 in./min) with a mean load of 93 kN (21 kips) and amplitude of (85 kN) 19 kips. Thereafter, the monotonic loading
was resumed until failure of the beam occurred. Figs. 9a
and 9b show the beam response for the elastic and inelastic
cycles, respectively. A very small displacement increment is
seen by the 10 cycle in both cases indicating that the
strengthened beam had reached an elastic state with a permanent set added to the overall deection. The behavior of
beam FRP 3 was similar to that of the other strengthened
beams. The ultimate failure mode was due to concrete compression failure in the moment span. After concrete compression, the beam continued to carry considerable load.
At a later stage a concrete cover failure near one of the supports was seen. Following the cover failure, an FRP strip
detached due to anchor pullout. It must be noted that this
failure was observed well after the concrete crushed. An
increased stiness in the elastic and inelastic regions of
the loading was seen for FRP 3 over FRP 1 and FRP
2SS. The yield strength increase was 25% while the ultimate
strength increase was 58%.
After the test, the beam was turned over to view the failure mode of the FRP strip. Sustained bearing failure or
slotting in three FRP strips was observed. Fig. 10 shows
the slotting in one of the FRP strips.
6.1. Comparison between analytical predictions and test
results

30
20
10
0
0.5

0.75

1.25

1.5

1.75

Displacement (in.)
Fig. 9b. Beam FRP 3 response during load cycles in the inelastic range.

The analytical model presented above was used to calculate the ultimate capacity of the laboratory beams. Results
of the calculations and comparisons with the test data are
shown in Table 7. All beams were predicted to fail due to
concrete compression with sustained bearing in the FRP
strips (i.e., Mode I). In all cases the sustained bearing stress

Fig. 10. Close-up of sustained bearing failure at end anchors in the FRP strip after failure of beam FRP 3.

6.9

in the FRP strip was less than the open-hole strength of


the strip. The analytical predictions compare well with
the test data. The negative dierence between the predicted results and the test results is conservative and demonstrates that the model can be used to predict the
ultimate capacity of MF-FRP strengthened beams.

421 (61)
9452 (2125)
2

3345 (752)

638 (92.5)

968 (1.5)

416 (307)

447 (330)

7. Conclusions

35

moment
 100.
%Difference from experimental CalculatedExperimental
Experimental moment
a
In shear span of each beam.

283 (209)
376 (277)
389 (287)
219 (162)
352 (259)
340 (251)

645 (1.0)
645 (1.0)

421 (61)
383 (56)

9452 (2125)
9452 (2125)

35
35

C-1
FRP 1
FRP
2 SS
FRP 3

2
2

3345 (752)
2989 (672)

638 (92.5)
638 (92.5)

Ultimate moment
(experimental),
M exp
(kN m (k-ft))
n
Ultimate moment
(calculated), M calc
n
(kN m (k-ft))
Area of FRP
strips, Afrp
(mm2 (in.2))
Open hole strength
of FRP strip, f0
(MPa (ksi))
Sustained
bearing stress
in each FRP
strip, ff,b
(MPa (ksi))
Anchor bearing
load, Pea,b
(N (lbs))
Fastener
bearing
load, Ppaf,b
(N (lbs))
Number of
anchorsa, M
Number of PA
fasteners a, N
Beam

Table 7
Comparison between model predictions and test results

22.6
6.5
12.7

L.C. Bank, D. Arora / Composite Structures 79 (2007) 180191


%Dierence

190

An analytical model has been developed to predict the


strength of MF-FRP strengthened beams. The model
accounts for the unique failure modes that occur in MFFRP strengthened beams. In particular, the model
accounts for the pseudo-plastic behavior of the FRP strip
when it fails in a progressive fashion in sustained bearing.
This sustained bearing failure of the mechanically fastened FRP strip is the desired design failure mode and
leads to a ductile response of the strengthened reinforced
concrete beam. Predictions of the analytical model compare well with the test results both in terms of strength
capacity and in terms of mode of failure observed.
The results of tests on concrete beams presented in this
paper and elsewhere demonstrate that powder-actuated
pins and expansion anchors can be used to attach an
FRP strip to an existing reinforced concrete beam for
the purpose of increasing the exural capacity of the
beam. In order for the method to be used, the FRP strip
must have a high longitudinal bearing strength and a high
open-hole tensile strength. Currently available commercially produced unidirectional carbon ber/epoxy pultruded strips cannot be used with the MF-FRP method,
as they do not have sucient bearing and open-hole
strength to enable forces to be transferred into the FRP
strip by bearing. An FRP strip for use with the MFFRP method, called SafStripTM, is commercially available.
Fasteners are available from many vendors, although,
tests to-date have only been conducted using fasteners
produced by Hilti.
As demonstrated by the test results, full-size concrete
beams strengthened with the MF-FRP method can reach
levels of strengthening comparable to those of externallybonded FRP systems (greater than 50% strengthening).
When appropriately designed, the MF-FRP method can
yield a ductile response of the strengthened beam with
concrete compression failure with the FRP strip remaining attached through very large displacements.
Acknowledgements
Support for this research was provided by the Wisconsin Department of Transportation through the Wisconsin
Highway Research Program (WHRP) and the US Army
Corps of Engineers. The contents of this paper reect
the views of the authors who are responsible for the facts
and the accuracy of the data presented herein. The contents do not necessarily reect the ocial views of the
Wisconsin Department of Transportation or the US

L.C. Bank, D. Arora / Composite Structures 79 (2007) 180191

Army Corps of Engineers. Mr. Peter Gulbrandsen is


thanked for conducting the bearing tests.
References
[1] ACI 440.2R. Guide for the design and construction of externally
bonded FRP systems for strengthening concrete structures. Farmington Hills, MI: American Concrete Institute; 2002.
[2] b (Federation Internationale du Beton). Externally bonded FRP
reinforcement for RC structures. Switzerland: International Federation for Structural Concrete (b); 2001.
[3] Teng JG, Chen JF, Smith ST, Lam L. FRP-strengthened RC
structures. New York: Wiley; 2001.
[4] Holloway LC, Leeming MB, editorsStrengthening of reinforced
concrete structures. UK: Woodhead Publishing Limited; 1999.
[5] Bonacci JF, Maalej M. Behavioral trends of RC beams strengthened
with externally bonded FRP. J Compos Construct 2001;5(2):10213.
[6] Stallings JM, Tedesco JW, El-Mihilmy M, McCauley M. Field
performance of FRP bridge repairs. J Bridge Eng 2000;5(2):10713.
[7] Hag-Elsa O, Alampalli S, Kunin J. Application of FRP laminates
for strengthening of a reinforced concrete T-beam bridge structure.
Compos Struct 2001;52:45366.
[8] Chajes MJ, Finch WW, Januszka TF, Thomson TA. Bond and force
transfer of composite material plates bonded to concrete. ACI Struct
J 1996;93(2):20817.
[9] Bizindavyi L, Neale KW. Transfer lengths and bond strengths for
composites bonded to concrete. J Compos Construct
1999;3(4):15360.
[10] Fanning PJ, Kelly O. Ultimate response of RC beams strengthened
with CFRP plates. J Compos Construct 2001;4(2):1227.
[11] Spadea G, Bencardino F, Swamy RN. Structural behavior of
composite RC beams with externally bonded CFRP. J Compos
Construct 1998;2(3):1327.
[12] Garden HN, Hollaway LC. An experimental study of the inuence of
plate end anchorage of carbon bre composite plates used to
strengthen
reinforced
concrete
beams.
Compos
Struct
1998;42:17588.
[13] Lamanna AJ, Bank LC, Scott DW. Flexural strengthening of
reinforced concrete beams using fasteners and ber-reinforced polymer strips. ACI Struct J 2001;98(3):36876.

191

[14] Lamanna AJ, Bank LC, Scott DW. Flexural strengthening of RC


beams by mechanically attaching FRP strips. J Compos Construct
2004;8(3):20310.
[15] Bank LC, Lamanna AJ, Ray JC, Velazquez GI. Rapid strengthening
of reinforced concrete beams with mechanically fastened, berreinforced polymeric composite materials. US Army Corps of
Engineers, Report Number ERDC/GSL TR-02-4, 2002.
[16] Bank LC, Borowicz DT, Lamanna AJ, Ray JC, Velazquez GI. Rapid
strengthening of full-size reinforced concrete beams with powderactuated fastening systems and ber-reinforced polymer (FRP)
composite materials. US Army Corps of Engineers, Report Number
ERDC/GSL TR-02-12, 2002.
[17] Arora D. Rapid Strengthening of reinforced concrete bridge with
mechanically fastened-ber reinforced polymer strips. M.S. thesis,
University of Wisconsin Madison, 2003.
[18] Bank LC, Oliva MG, Arora D, Borowicz DT. Rapid strengthening of
reinforced concrete bridges, Wisconsin Highway Research Program,
Report No. 03-06. Wisconsin Department of Transportation, Madison, WI, 2003.
[19] Bank LC, Nanni A, Rizzo A, Arora D, Borowicz DT. Pultruded FRP
strips, mechanically fastened, give concrete bridges strength, Composites Fabrication, September 2004, pp. 3239.
[20] Abd-El-Naby SFM, Holloway L. The experimental behavior of
bolted joints in pultruded glass/polyester material. Part 2: Two-bolt
joints. Composites 1993;24(7):53946.
[21] Rosner CN, Rizkalla SH. Bolted connections for ber-reinforced
composite structural members: experimental program. J Mater Civil
Eng 1995;7(4):22331.
[22] ACI 313-02. Building code requirements for structural concrete,
American Concrete Institute, Farmington Hills, MI, 2002.
[23] Hilti. North American Product Technical Guide. Tulsa, OK: Hilti
Inc..
[24] ASTM D 3039. Standard test method for tensile properties of
polymer matrix composite materials. West Conshohocken, PA:
American Society for Testing and Materials.
[25] ASTM D 5766. Standard test method for open hole tensile strength of
polymer matrix composite laminates. West Conshohocken, PA:
American Society for Testing and Materials.
[26] ASTM D 5961. Standard test method for bearing response of
polymer matrix composite laminates. American West Conshohocken,
PA: Society for Testing and Materials.

Das könnte Ihnen auch gefallen