Composite Structures 122 (2015) 456–467

Contents lists available at ScienceDirect

Composite Structures
journal homepage: www.elsevier.com/locate/compstruct

Retrofitting of interior RC beam–column joints using CFRP strengthened

SHCC: Cast-in-place solution
Esmaeel Esmaeeli a,⇑, Joaquim A.O. Barros a, Jose Sena-Cruz a, Luca Fasan b, Fabio Raimondo Li Prizzi b,
José Melo c, Humberto Varum d
a
ISISE, Dept. Civil Eng., University of Minho, Guimarães, Portugal
b
Dept. Civil Eng., University of Padova, Padova, Italy
c
Dept. Civil Eng., University of Aveiro, Portugal
d
Dept. Civil Eng., University of Porto, Portugal

a r t i c l e i n f o a b s t r a c t

Article history: The effectiveness of a repair strategy, for damaged RC beam–column joints, that combines strain harden-
Available online 12 December 2014 ing cementitious composite (SHCC) and laminates of carbon fiber reinforced polymers (CFRP laminates) is
assessed in the present work. According to this technique, the existing concrete cover in the joint zone of
Keywords: the frame is replaced by a self-compacting SHCC. This thin layer of SHCC is reinforced with CFRP
Strain hardening cementitious composite laminates that are bonded into the saw cut grooves. Two full-scale severely damaged interior RC
(SHCC) beam–column joints were retrofitted using two different configurations of this technique: (i) applying
Near surface mounted carbon fiber
the strengthening system to only the front and rear faces of the specimen; (ii) jacketing all sides of the
reinforced polymer (NSM-CFRP)
Interior reinforced concrete (RC)
elements of the specimen with the strengthening system. The effectiveness of these retrofitting configu-
beam–column joint rations is assessed and compared by evaluating experimentally the hysteretic response, the dissipated
Cyclic behavior energy, the degradation of secant stiffness, the displacement ductility and the failure modes of each
Retrofitting repaired specimen, and also using the values of these indicators obtained in the virgin state of the spec-
imens. This comparison revealed that the adopted retrofitting strategies can restore and even enhance
the performance of this type of structural elements, mainly when the solution based on four-sided
jacketing is used.
Ó 2014 Elsevier Ltd. All rights reserved.

1. Introduction the beam–column joints. Continuous damage due to aging effects,

even in those structures designed based on seismic oriented codes,
Seismic deficiencies of RC structures designed based on pre- also makes them vulnerable against earthquakes.
seismic provisions, such as pre-1970th buildings, is figured out in Several strategies for seismic retrofitting of these group of
both experimental tests [1,2] and also post-earthquake observa- structures are available, such as steel jacketing, cast-in-place con-
tions (e.g. Turkey 1999 and Italy 2009). These vulnerabilities are crete/RC jacketing [3], shotcrete jacketing [4], epoxy injection
mostly due to the lack of seismic design and detailing of these repair [5], application of Fiber Reinforced Polymers (FRPs) [6–9].
structures. Among the structural components of a framed-struc- Li et al. [10] showed that the interior RC beam–column joints
ture, beam–column joints play the most significant role in the lat- can be strengthened by using a ferrocement jacket as the replace-
eral stability, since a brittle failure at the joint region may result in ment of the exiting concrete cover at critical regions of the framed
a progressive collapse of a building. Therefore, both energy dissipa- elements along with embedding inclined bars in the joint region.
tion and ductility capacities of these structures, when a seismic High performance fiber reinforced composites (HPFRC) were used
event occurs, highly depend on the stability and deformation of by Shannag and Alhassan [11] for the strengthening of 1/3 scale
interior beam–column joints containing vulnerable detailing
against seismic actions. A 25 mm thick jacket of HPFRC covering
⇑ Corresponding author at: ISISE, Dep. Civil Eng., School Eng., University of Minho, critical regions of column-joint and extending up to a 100 mm on
Campus de Azurém, 4800-058 Guimarães, Portugal. Tel.: +351 917 40 90 75. the beams of the specimens were the adopted strengthening con-
E-mail addresses: [email protected] (E. Esmaeeli), barros@civil.
figuration. The results of this experimental program have revealed
uminho.pt (J.A.O. Barros), [email protected] (J. Sena-Cruz), morefasan@gmail.
com (L. Fasan), [email protected] (F.R. Li Prizzi), [email protected] (J. Melo), that HPFRC jackets can significantly improve the seismic response
[email protected] (H. Varum). of deficiently detailed interior beam–column joints. However, this

http://dx.doi.org/10.1016/j.compstruct.2014.12.012
0263-8223/Ó 2014 Elsevier Ltd. All rights reserved.
 E. Esmaeeli et al. / Composite Structures 122 (2015) 456–467 457

jacketing solution has increased the dimensions of the cross sec- The idea of reinforcing SHCC layer with bonded CFRP laminates
tions of the elements in 25% to 47%, which can be a real obstacle into the saw cut grooves benefits the progressive increase in tensile
on the use of this technique in certain applications. The experimen- strength of the SHCC, at least up to the rupture strain of CFRP lam-
tal program performed by Tsonos [4] was focused on the strength- inate, generally 1.5% to 1.6%. This provides strain compatibility
ening of 1/2 scale exterior beam–column joints by adding a new between these two composites while ductility of SHCC in combina-
steel cage reinforcement that was covered either with shotcrete tion of high tensile strength of CFRP laminates may produce a
or cast-in-place cement based materials. Based on the results of strengthening scheme with a high toughness. Also bonding CFRP
this experimental program, Tsonos [4] stated that both cast-in- laminates as the supplementary tensile reinforcement to the
place and shotcrete solutions provided a significant improvement exposed face of the SHCC layer, in the hardened state, minimizes
in the seismic response of this type of structures. A superior perfor- the obstacles during placing the fresh SHCC. A better bond quality
mance of the cast-in-place solution in respect to the shotcrete control between SHCC and CFRP laminates can be expected too.
technique, mainly in terms of energy dissipation capacity, was Finally, the mechanism of formation of numerous diffused micro
observed and attributed to a better covering of the added steel cracks and opening and closing of those during reversal cyclic loads
bar cage that was assured by the former technique. Considering results in a high capacity of energy dissipation which is the most
an increase of 140 mm in each side of the column’s cross section desired character for seismic load resisting elements.
of the 1/2 scale specimens, inacceptable interference can result
in terms of architectural and functional requisites. Wang and Hsu 2. Experimental program
[12] a satisfactory performance of the strengthening effectiveness
of RC jacketing of columns of beam–column assemblies with shear The experimental program is composed of retrofitting two
deficiency in the joint region. For this case, the adopted thickness severely damaged full-scale interior beam–column joints. The ret-
of the RC jacket resulted in an increase up to 67% in the dimensions rofitting methodology was based on replacing concrete cover with
of the column section. SHCC in the joint region and along the critical lengths of beams and
Strain hardening cementitious composite (SHCC) is a class of columns. To enhance the tensile strength of this ductile layer, CFRP
Fiber Reinforced Concretes (FRCs), with the character of developing laminates were bonded into grooves cut on hardened SHCC along
a continuous increase of post-cracking tensile capacity up to the both the longitudinal and the transverse directions. Two different
stress localization at one of the multiple formed cracks for a rela- depths, 10 and 20 mm, for the grooves were adopted in order to
tively high tensile strain. The formation of multiple diffused hair- allow the arrangement of the CFRP laminates in two different ori-
line cracks through all the loaded length of the specimen during entations. An X-shaped configuration of these CFRP laminates,
the hardening stage assures levels of ductility not possible to attain bonded in two different levels, was used for the shear strengthen-
in conventional FRCs. By testing in bending masonry elements ing of the joint region.
strengthened with a thin layer of SHCC applied to their tensile face, The difference between the adopted retrofitting schemes for the
Esmaeeli et al. [13] demonstrated that higher load carrying capac- tested specimens was the number of faces of the framed elements
ity and ductility is achievable when compared to flexural strength- that was retrofitted. While in one specimen only the front and rear
ening methodologies based on the use of thicker layers of faces of beams, columns and joint were retrofitted, for the other
self-compacting steel FRC. Recently Esmaeeli et al. [14] developed specimen all the external faces of the mentioned elements were
a thin prefabricated hybrid composite plate (HCP), composed of jacketed.
SHCC and CFRP laminate, with a high durability potential. By After retrofitting, these specimens were subjected to the same
performing some experimental tests, they demonstrated the high test setup and loading pattern that were used to characterize their
efficacy of the HCP for the repair and the strengthening of the dif- lateral load–displacement response in their virgin state. To evalu-
ferent types of the RC elements including retrofitting of damaged ate the efficacy of the adopted retrofitting strategies the results
beam–column joints. obtained from these experiments were then compared to the cor-
In this paper an experimental program for the assessment of the responding results of their virgin states.
effectiveness of a retrofitting technique for damage interior RC
beam–column joints by casting-in-place SHCC and further 2.1. Damaged specimens
reinforcing that with CFRP laminates is described, and the main
obtained results are presented and discussed. Two severely damaged interior RC beam–column joints, desig-
Both the fine graded matrix and the high content of fly ash in nated as JPA3 and JPB, were selected from a group of specimens
the skeleton of SHCC can promote the formation of a relatively high that were tested in their virgin state in the ambit of an experimen-
bond quality at the interface between SHCC and existing concrete. tal research program of a PhD thesis [20]. Both specimens were
A high strain capacity (strain at maximum tensile strength of com- identical in terms of the lengths and cross section geometry of
posite) and a tight crack width, often and in average smaller than their framed elements. The only difference between these speci-
100 lm up to the ultimate tensile capacity, is known as a durable mens was the number of longitudinal rebars: more 4 longitudinal
composite cover and expected sufficient confining pressures at rel- rebars were used in the column of JPB. The lengths of the beams
atively high strains. In fact, the results of previous studies showed and columns of these specimens were taken as the mid-span and
that for the concrete patched with SHCC, a single crack with a large the mid-storey of a common RC building built before 1970th,
width formed in substrate transforms to a multiple diffused fine respectively. The mid-length of the elements were used to facili-
crack in the patch layer which are typically impermeable and tate the simulation of the boundary conditions in the experimental
assure the durability of the repaired substrate [15–17]. When SHCC test since moment inflection point of a RC frame under a lateral
is used to repair RC elements with progressive corrosion of their loading is expected to occur in the these zones.
steel rebars, the risk of splitting and spalling of this ductile retrofit- According to the configuration of the most buildings con-
ting cover due to the expanded volume of the rusted bars is mini- structed before 1970th, plain steel bars were used as the longitudi-
mized [18,19]. Moreover, the tensile strain ductility of SHCC results nal and transverse reinforcement for both beams and columns.
in a high potential of stress redistribution at the bearing zones, There were no transverse reinforcement in the joint region and
therefore avoids premature failure at this region when anchors 90° bended end was used for the stirrups and hoops of the beams
used to enhance shear stress transference between the retrofitted and columns, respectively. More details about the configuration of
layers. the selected specimens for the retrofit, JPA3 and JPB, are shown in
458 E. Esmaeeli et al. / Composite Structures 122 (2015) 456–467

Fig. 1. Adopting a shorter length for the inferior column was due to tensile strength of the steel longitudinal reinforcement, respec-
the limitation imposed by the test setup which is discussed further tively, with an elasticity modulus of 198 GPa.
in this section. A lateral reversal displacement history was imposed to the top
The average compressive strength, measured in 150 mm cubes of the superior column at the presence of a constant axial load of
of concrete, was equal to 23.8 MPa with an estimated characteristic 450 kN. This axial force introduces a gravity load corresponding
compressive strength of 19.8 MPa, corresponding to the C16/C20 to an axial compressive stress of 21.3% of the average concrete
concrete strength class according to the classifications of the compressive strength. The lateral load was constituted of a series
EC2-1992-1-1 [21]. By performing tensile tests, average values of of displacement-controlled cycles, in push (positive displacement)
590 and 640 MPa were determined for the yield and the ultimate and pull (negative displacement) direction, with an incremental

2150 300 2150

Superior column
4Ø12(JPA3)
8Ø12(JPB)

Column's lateral face Column's lateral face

1600

B B
Beam's top face

Left Beam 2Ø12 A 2Ø12 Ø8@200

400

4Ø12 A 4Ø12 Right Beam

- The reinforcement showed in Beam's bottom face
1000

dashed line only exist for the case

of JPB. Ø8@250
Inferior column
- All units are in mm.

Top face Front face Front face

2Ø12 300 300
300

Lateral face
Lateral face

Lateral face

Ø8 Ø8
Front face

Ø8
Rear face

300
300

20

20
4Ø12
400

20

8Ø12 4Ø12
Bottom face Rear face Rear face
SECTION A-A SECTION B-B (JPB) SECTION B-B (JPA3)

Fig. 1. Details of adopted configurations for the interior RC beam–column connections.

Cycle Numbers (mm) Max drift (%)

120 N 4.0
(+) 1 ±1 0.033
100
Fc dc 3.3
2 ±2 0.067
80 2.7 3 ±4 0.133
60 2.0
4 to 6 ±6 0.20
7 to 9 ±10 0.33
Displacem ent, d c (m m )

40 1.3 10 to 12 ±15 0.50

20 0.7 13 to 15 ±20 0.67
Drift (% )

16 to 18 ±25 0.83
0 0.0
19 to 21 ±30 1.00
-20 -0.7 22 to 24 ±40 1.33
-40 -1.3 25 to 27 ±50 1.67
28 to 30 ±60 2.00
-60 -2.0
31 to 33 ±70 2.33
-80 -2.7 34 to 36 ±80 2.67
-100 -3.3 37 to 39 ±90 3.00
40 to 42 ±100 3.33
-120 -4.0
43 to 45 ±110 3.67
Step 46 to 48 ±120 4.00
p
Fig. 2. Loading history adopted for the lateral displacement cycles (dc : peak displacement for the corresponding cycle or set of cycles).
 E. Esmaeeli et al. / Composite Structures 122 (2015) 456–467 459

magnitude up to 4% interstory drift. After three cycles of loading and test results of the virgin specimens can be found elsewhere
that introduced a drift level of 0.13%, each level of displacement [20].
was repeated three times, as it is shown in Fig. 2. The specimens
were tested in a horizontal position according to the test setup 2.2. Retrofitting strategy
illustrated in Fig. 3. As it is shown in this figure, the shorter length
of the inferior column of the specimen is connected to a steel ele- According to the adopted retrofitting strategy, the concrete
ment with equivalent stiffness, to accommodate the load cells and cover at critical regions of the damaged beam–column joints is
pin connection at the bottom of this column. replaced with a thin layer of a casted-in-place SHCC. Afterward,
The maximum load carrying capacity of 43.2 and 39.5 kN was this layer of the SHCC was reinforced with CFRP laminates bonded
registered for JPA3 and JPB, respectively, at the drift levels of to the saw cut grooves on that according to the NSM technique.
2.7% and 2.3%, correspondingly. Chemical anchors were used to improve inter-laminar shear stress
As shown in Fig. 4, in both specimens the extents of the dam- transference between the SHCC and the concrete substrate. The
ages included concrete crushing and spalling at the intersections rheology of the SHCC material, used in this study, is tailored to pro-
of the beams and the columns, and sever sliding of longitudinal duce a highly fluid and self-compacting fresh state behavior so that
reinforcement due to significant bond deterioration. Flexural this composite can easily flow and fill narrow spaces between the
cracks on the right beam of both specimens were localized at the formworks and the existing concrete (gaps of less than 25 mm).
beam–joint interface. The main crack at the right beam of JPA3 To the retrofitted JPA3 and JPB specimens, the nomination of
and JPB specimens is at a position of 120 and 170 mm far from the JPA3-R and JPB-R was attributed, respectively. As mentioned
beam–joint interface, respectively. There were minor flexural before, the adopted retrofitting schemes for the specimens differed
cracks at the column-joint interfaces of JPB. Specimen JPA3 also according to the number of faces of their elements which was ret-
has experienced severe damages concentrated in the joint region, rofitted. While in JPA3-R only the front and rear faces of beams,
where two wide diagonal cracks have formed and concrete cover columns and joint were retrofitted, in JPB-R all the external faces
has spalled. Additional information about experimental program of the mentioned elements were jacketed.
The retrofitting process was applied with the specimens posi-
tioned horizontally and in two steps: (i) before and (ii) after turn-
ing the specimens. Following the details of each step of the
N retrofitting strategy are described.
Hydraulic
Actuators 2.2.1. Concrete cover removal and replacement
Details of the retrofitting schemes are presented in Fig. 5. The
retrofitting length for both beams and columns was taken as twice
Fc dc of the section depth of the corresponding element. Therefore, using
(-) (+) a jackhammer concrete cover was removed in the joint region and
Superior

also in all lateral faces of the beams and columns of both speci-
column
1500 mm

mens for a length of 800 and 600 mm, respectively. The concrete
cover was initially removed up to a depth to expose the longitudi-
nal reinforcements. Afterward, in an effort of increasing the inter-
Left beam Right beam
face area between casted-in-place materials and existing steel bars,
the removal of the concrete cover continued up to attain approxi-
1500 mm

Inferior
column

mately half of the diameter of the longitudinal bars. To seal the

Connector existing cracks, boreholes were drilled through the cracked sec-
piece tions. After cleaning the holes using compressed air, small diame-
ter pipes were placed inside them, then the exposed crack
development at the concrete substrate was sealed and epoxy resin
Load cells SikaDur-52 was injected. After turning the specimen, the injection
was repeated to assure that the cracked section was sealed as
much as possible.
Wooden formworks with interior varnished faces were installed
to cast the cement based materials. The lateral faces of columns
2000 mm 2000 mm and the top and bottom faces of the beams of JPA3 were casted
using a mortar that was then cured for 7 days (see Fig. 1 for the
Fig. 3. Test setup adopted for the horizontally placed specimens [20].
nomination of the faces of the elements of the beam–column

Fig. 4. The extent of damages before retrofitting (a) JPA3 and (b) JPB.
460 E. Esmaeeli et al. / Composite Structures 122 (2015) 456–467

JPA3-R JPB-R
Superior Column
Superior Column

CFRP
CFRP@100mm
5

200

200
CFRP

600
300

300
CFRP @100mm CFRP
CFRP
Lateral Face
Anchors(Ø10) A A´ CFRP@100mm
112 247 327 347 303 67

5
50

50
Right Beam
Left Beam

CFRP

CFRP
40

400
300

300
50

50
40 68 282 317 332 278 67

100
A A´ CFRP@100mm
CFRP
Anchors(Ø10)
300

300
Note:

600
B B´
Since anchors of the
125

125
column are not placed in
50

50
symmetric positions, all of
them are shown
800 150 150 800

det. A Anchor (Ø10) det. A´

Anchor (Ø10)
det. B
Anchor (Ø10) det. B´
55 45 105 105 45 55
106 45 50

Anchor (Ø10)
145

75
145
410
400

310

145 145
50 45 106

310
160
75
40 110 110 40
65 90 90 65
155 155 300 310 45 75 70 75 45
310
310

SECTION A-A SECTION B-B SECTION A´-A´ SECTION B´-B´

CFRP Laminate CFRP Laminate

CFRP Laminate Epoxy Resin (longitudinal) Epoxy Resin (longitudinal)
Epoxy Resin (longitudinal) SHCC CFRP Laminate
40 SHCC (transverse)
CFRP Laminate
Mortar SHCC 40 (transverse)
40
40

Epoxy Resin Epoxy Resin

40

40

Mortar SHCC

CFRP Laminate CFRP Laminate

Beam Section Column Section CFRP Laminate Column Section
(transverse) (transverse)
Beam Section
40 (longitudinal) 40
CFRP Laminate CFRP Laminate
SHCC (longitudinal) SHCC (longitudinal) Epoxy Resin

Detail A Detail A´ Detail B Detail B´

Fig. 5. Details of the schemes used for the retrofitting of the damaged specimens (dimensions in mm).

joints). After this period of curing, the top edges of the hardened thickness required to accommodate two layers of CFRP laminates
mortar were roughened and fresh SHCC was placed. in the SHCC layer, a 5 mm higher finishing level for the SHCC
For the case of JPB, a continuous placing of SHCC starting from was adopted, as measured from the level of the existing concrete
lateral faces of the columns and the top and bottom faces of the cover at the extremities of the retrofitted regions.
beams, and then moving to the front face of the specimen was The self-compacting character of the SHCC and its high fluidity
followed. eliminated the need to any external vibration. Only the exterior
Considering the variation in the thickness of the existing con- face of the fresh SHCC was levelled by using a thin long metal
crete cover, between 16 and 20 mm, and a minimum of 20 mm bar, with a rectangular cross section, for the finishing purpose.
 E. Esmaeeli et al. / Composite Structures 122 (2015) 456–467 461

Column Column
(a) (b)
Fig. 6. View of the retrofitted specimens (a) JPA3-R and (b) JPB-R.

It should be noted that before casting the cement based materi- which is a fast curing injectable bonding agent. An embedded
als, the concrete substrate was saturated with water in order to length of 145 mm was assured for the anchors, measured from
assure a better interface bond and a lower risk of developing the finished surface of SHCC. A torque of 30 N  m was applied to
shrinkage cracks. fasten the nuts and partially confine the concrete substrate. Fig. 6
One day after casting the SHCC the formworks were removed. A shows a view of the specimens after have been repaired.
wet curing procedure was followed for at least 7 days as it was
reported the most appropriate curing for SHCC [22]. After at least 2.3. Material properties of retrofitting system
17 days of casting the SHCC, grooves were executed on the SHCC
according to the configurations showed in Fig. 5. These grooves The self-compacting SHCC was composed of a cementitious
had 5 mm of width, and 10 or 20 mm of depth, depending on the mortar reinforced with 2% of volume short discrete PVA fibers.
level adopted for the installation of the CFRP laminates. Before The PVA fiber used in this study had a length of 8 mm and was pro-
inserting the CFRP laminates, the grooves were cleaned by using duced by Kuraray Company with the designation of RECs 15  8.
compressed air, and then filled with epoxy resin S&P 220 as the The average tensile stress at crack initiation and the average tensile
bonding agent. Afterward, CFRP laminates that were previously strength of the SHCC was 2.43 and 3.35 MPa, respectively, with a
cleaned with acetone, were placed inside the grooves. minimum tensile strain capacity of 1.3%. More details on mixture
After turning the specimens the same retrofitting process was ingredients, mixing process and test setup of the SHCC can be
applied to the rear face, namely: removal of the concrete cover, found in [10,13,22]. From uniaxial tensile tests carried out accord-
sealing of the cracks, roughening the top edges of newly casted ing to the recommendations of ISO 527-2:1996 [24] on seven days
materials, placing the fresh SHCC, curing of SHCC, cutting the cured of six dumbbell-shaped S&P 220 epoxy resin, an average ten-
grooves and inserting CFRP laminates. sile strength of 18 MPa and average modulus of elasticity of
For the case of JPB the grooves were also cut on the SHCC casted 6.8 GPa were obtained. Tensile properties of CFRP laminate
on the lateral faces of columns and the top and bottom faces of the (S&P laminate CFK 150/2000) with a cross section of 1.4  10 mm2
beams, and pair of CFRP laminates was bonded into these grooves were characterized following the procedure proposed in ISO
according to Fig. 5. Therefore, for the case of JPA3-R, the longitudi- 527-5:2009 [25]. From the tests executed in six coupons, average
nal reinforcement comprised pairs of continuous laminates on values of 2689 MPa, 1.6% and 165 GPa were obtained for the tensile
each of the front and rear faces of the beams and columns strength, strain at CFRP rupture and modulus of elasticity,
(see Fig. 5), while JPB had a similar CFRP strengthening but also respectively.
with extra pairs of CFRP laminates bonded to the each of the lateral
faces of its columns, and the top and bottom faces of its beams. 2.4. Test setup and loading pattern
CFRP laminates bonded to the lateral faces of the beams and col-
umns were continued beyond the interface of these elements with The test setup, lateral load history and gravity load used for
the joint region, where the occurrence of the largest bending testing virgin specimens were adopted for testing the retrofitted
moments is expected (moment critical sections). For this purpose, specimens.
an inclined drilling was used to execute the holes. After placing the
CFRP laminates, the epoxy resin was injected. The bond length of 3. Results and discussion
100 mm was adopted for these CFRP laminates after moment crit-
ical section (anchorage length), since a minimum of 90 mm is char- 3.1. Hysteresis response
acterized as the required bonding length to fully mobilize potential
tensile strength of this type of CFRP laminates [23]. Fig. 7 shows the hysteretic responses of both virgin and retrofit-
The adopted spacing for transverse CFRP laminates in both ted specimens in terms of lateral load versus lateral displacement
JPA3-R and JPB-R was 100 mm (Fig. 5). In an attempt to increase (and drift) registered at the top of the superior column. Both retro-
the shear resistance of the joint region, a pair of CFRP laminates fitting techniques resulted in stable loops with smooth decay of
with an X shape configuration was applied on each front and rear load carrying capacity in the post-peak stage of the structural
face of the joint region of both specimens. response. The values registered for the maximum lateral load (FP)
and its corresponding drifts (dP) for specimens in the retrofitted
2.2.2. Installing chemical anchors and virgin states are listed in Table 1. The increase level in terms
Chemical anchors were installed before and after turning the of lateral peak load after retrofitting is also indicated in this table.
specimens, when the SHCC was cured at least 20 days. These According to the obtained results, the retrofitting technique
10 mm diameter anchors (HIT-V-8.8 M10X190) were mounted adopted for JPA3-R recovered up to 93% of the maximum lateral
inside the holes perforated on the beams, columns and on the joint load carried out by this specimen in its virgin state, calculated as
region, at the positions represented in Fig. 5. Before mounting the the average load in the positive and negative directions. Applying
anchors, the holes were partially filled with Hilti Hit-HY 200-A, the retrofitting technique to all lateral faces of the framed
462 E. Esmaeeli et al. / Composite Structures 122 (2015) 456–467

Drift (%) Drift (%)

-4 -3 -2 -1 0 1 2 3 4 -4 -3 -2 -1 0 1 2 3 4

60 JPA3-R 60 60 JPB-R 60
JPA3 JPB
40 40 40 40

Lateral load (kN)

Lateral load (kN)

20 20 20 20

0 0 0 0

-20 -20 -20 -20

-40 -40 -40 -40

-60 -60 -60 -60

-120 -90 -60 -30 0 30 60 90 120 -120 -90 -60 -30 0 30 60 90 120
Lateral displacement (mm) Lateral displacement (mm)

Fig. 7. Hysteretic responses of the specimens in the strengthened and virgin states.

Table 1
Maximum lateral load capacity and the corresponding drifts of the specimens in the repaired and virgin states.

Specimen Negative direction Positive direction Negative direction Positive direction


F
p (kN) dp (%) Fþ
p (kN)
þ
dp (%) Increase in peak load

JPA3-R 38.0 1.65 +40.9 +1.65 9.3% 5.5%

JPA3 41.9 2.31 +43.3 +2.60
JPB-R 52.7 1.62 +57.14 +2.33 +48.9% +44.5%
JPB 35.4 1.99 +39.55 +2.24

elements, as was done in JPB-R, resulted in a significant increase in 3.2. Damage evolution and failure modes
terms of lateral load carrying capacity. This increase was +48.9%
and +44.5% for negative and positive directions, respectively, when Fig. 8 shows the pattern of the developed micro cracks, and
compared to the corresponding values recorded in the virgin state major damages registered at the end of the test on the front faces
of this specimen (JPB). of both specimens. The surface of the SHCC was painted with a
For both strengthening techniques the average value of the drift transparent concrete varnish before testing the specimens. At the
corresponding to the maximum lateral load, in negative and posi- end of the test this surface was sprayed by a penetrating liquid
tive direction, has decreased. This can be attributed to a lower to reveal micro cracks difficult to detect at necked eyes. The sche-
shear deformation at the joint region due to the contribution of matic representation of these damages is showed at the left side of
the strengthening scheme in confining the concrete of the joint the corresponding photo for the purpose of better assessment of
core, and also in increasing the shear stiffness of the joint panel, the developed damage. The damage evolution is described in the
up to the peak load. following paragraphs.
For each specimen, the residual lateral load carrying capacity at JPA3-R: The first series of cracks has initiated at the cycles cor-
4% drift (F4%) was compared to the registered peak of the lateral responding to 0.33% of drift. These cracks were formed at the top
 
load (Fp) according to a ¼ 1  ðF 4% =F p Þ %. The degradation of the face of the left and the right beams at a distance of 100 mm from
peak load (a) was calculated for both virgin (av) and retrofitted the lateral faces of the column. At cycles corresponding to 0.5% of
specimens (aR). The amount of this degradation was 21.85% and drift, cracks at the bottom faces of both left and right beams, sym-
25.6% for JPA3-R and JPB-R, respectively, which are the average val- metric to the cracks on top face, were observed. Some relative slid-
ues for negative and positive loading directions. While at the same ing between retrofitting layer and concrete substrate was observed
drift level, JPA3-R had almost the same peak load degradation as when cycles of 0.83% drift were reached.
JPA3 (22.35%), corresponding value for JPB-R was much higher The first series of the inclined cracks at the junction of the
than JPB (3.7%). Larger degradation in the peak load of JPB-R, as beams and columns was observed in all four corners at the cycles
compared to JPB, is associated to different damage evolution and corresponding to 1% of drift. Further increase in the lateral displace-
failure modes of these specimens. In fact, in comparison with ment at the top of the superior column resulted in the progress of
JPB, JPB-R attained higher level of lateral load; therefore, higher these cracks into the interface of the epoxy adhesive/SHCC of the
shear stresses were applied to its joint region at the ultimate state. bonded X shape CFRP system at the joint region. Thus, for any larger
This resulted in an eventual damage concentration in the joint displacement demand, damages were localized at the joint region
region. In the other hand, the lateral load carrying capacity of JPB in the form of progressive separation between the epoxy adhesive
was limited by the premature flexural capacity of the beams, and the SHCC. Finally, at drift cycles of 1.67%, due to the load rever-
which was caused by the sliding of their longitudinal rebars that sal effects, the debonding was almost progressed along the entire
is expected to have smoother load degradation. length of the elements of the X shape CFRP configuration. As a
 E. Esmaeeli et al. / Composite Structures 122 (2015) 456–467 463

(a) JPA3-R

(b) JPB-R
Fig. 8. Damage propagation and concentration at the failure of (a) JPA3-R and (b) JPB-R.

consequence of this debonding, a total loss in contribution of these MR  ML

Vc ¼ ð1Þ
inclined CFRP laminates as a part of shear resisting mechanism of Lc
the joint region was occurred. Thus, shear failure of the joint region
where Vc is the shear force in the column, MR and ML are the values
was the governing failure mode of JPA3-R.
of the internal bending moment developed at the beam–column
JPB-R: The onset of the first series of cracks was at the set of
interfaces of the right and the left beam, respectively. The sign of
cycles corresponding to 0.5% of drift. These cracks were formed
the bending moment is assumed positive when the bottom face of
at the top and bottom faces of the left and right beams in a dis-
the beam is in tension and negative when this face is in compres-
tance of approximately 90 mm far from lateral faces of the col-
sion. In Eq. (1), LC is the total length of the column between its lat-
umn. The inclined cracks at the junction of the beams and
eral supports (1.5 + 1.5 m). According to Eq. (1), any reduction in the
columns were initially formed at cycles corresponding to a drift
flexural capacity of the left or right beams may result in the loss of
level of 0.83%. Similar to the case of JPA3-R, these set of cracks
lateral capacity of the beam–column assembly, unless this reduc-
resulted in a progressive debonding along the interface of epoxy
tion could be compensated through the moment redistribution to
adhesive/SHCC of the X-shaped CFRP system at the joint region.
other parts of the structure.
At drift cycles of 1.67% this debonding was already progressed
The maximum moments of each of the left and right beams ver-
along the entire length of the inclined CFRP laminates. At the
sus the drift demands were calculated, at a distance 50 mm far from
same cycles, the longitudinal steel bars at the top face of the
beam–column interfaces, by considering the force values registered
right beam started to have significant sliding, so that the con-
in the load cells and equilibrium conditions, and the obtained
crete cover perpendicular to the bended end of these bars was
results are illustrated in Fig. 9. Note that in this figure, for the con-
cracked. As it will be discussed in the next section, sliding of
venience of understanding, the multiplied value of ML by 1 is pre-
these rebars resulted in degradation of flexural capacity of the
sented. Thus, the beams’ bending moments corresponding to the
beams when the top face of them was in tension. The non-sym-
negative and the positive loading directions are presented in the
metrical response of JPB-R, in negative and positive loading
first and the third quadrants of Cartesian system, respectively.
directions, can be caused by this phenomenon. At the next sets
According to Fig. 9a, the maximum bending moments developed
of the cycles, corresponding to 2% of drift, the already cracked
in the left (ML) and the right (MR) beams of JPA3-R, during the neg-
concrete cover over the bended region of these bars was spalled
ative displacement, were +65.94 and 39.6 kN  m both at a drift
off. Afterward, any further increase in drift demand just followed
level of 1.64%. During the positive displacement, the left and the
by widening of the existing X-shaped cracks at the joint region.
right beams reached their maximum bending moment, 43.04
Therefore, the shear failure at the panel of the joint resulted in
and +71.17 kN  m, at drift levels of +1.65% and +2.65%, respectively.
degradation of lateral load resistance of JPB-R in both negative
As depicted in Fig. 9b, the values of maximum bending
and positive loading directions.
moments for JPB-R in the left and the right beams, during the neg-
ative displacement were +108.81 kN  m at a drift level of 2.62%
3.3. Flexural strength of beams
and 57.16 kN  m at 1.62% of drift, respectively. The developed
maximum bending moment for the positive displacement, in the
Eq. (1) represents the static equilibrium between the maximum
left and the right beams were 55.64 and +107.46 kN  m at drift
developed moments at the left and the right beams with respect to
levels of +1.66% and +2.33%, respectively. A sudden reduction
the lateral force at the top of the column.
464 E. Esmaeeli et al. / Composite Structures 122 (2015) 456–467

80 120

60
80
ML Pull Cylces MR ML MR
Flexural Strength (kN.m) 40 Pull Cylces

Flexural Strength (kN.m)

40
20
-(ML) (MR) -(ML) (MR)
0 0

-20
-40
ML Push Cylces MR ML MR
-40 Push Cylces
-80
-60

-120
-80
-4 -3 -2 -1 0 1 2 3 4 -4 -3 -2 -1 0 1 2 3 4
Drift (%) Drift (%)
(a) (b)
Fig. 9. Development of the resisting bending moment at the interfaces of the beams with columns (a) JPA3-R and (b) JPB-R.

Table 2
Maximum bending moments developed in the beams of the repaired and the virgin specimens.

Specimen Negative direction Positive direction Negative direction Positive direction

Absolute values Variation

Mþ
L (kN  m)
M
R (kN  m) M
L (kN  m) Mþ
R (kN  m) Mþ
L
M
R M
L Mþ
R

JPA3-R +65.94 39.6 43.04 +71.17 13.07% 1.39% +8.03% 10.98%

(1.64)* (1.64) (+1.65) (+2.65)
JPA3 +75.85 40.16 39.84 +79.95
(2.32) (2.32) (+2.59) (+2.59)
JPB-R +108.81 57.16 55.64 +107.46 56.16% 79.35% 62.22% 41.81%
(2.62) (1.62) (+1.66) (+2.33)
JPB +69.68 31.87 34.30 +75.78
(4.0) (1.99) (+2.58) (+2.44)
*
Values in parentheses indicate the corresponding drift in percentage at maximum bending moment.

observed in bending moment capacity of the right beam during capacity, as well. Therefore, for SHCC material with the potential
negative loading at drift cycle of 1.67% (Fig. 9b) was caused by a of formation multiple diffused micro cracks, a high level of energy
significant sliding of longitudinal bars at the top face of the right dissipation under cyclic loadings is expected. The amount of dissi-
beam, as discussed in the previous section. pated energy per cycle, Ei, can be calculated from the enclosed area
The registered maximum bending moments for these in each loading cycle, as presented by the hysteresis response of
specimens during both the positive and the negative loading lateral load versus lateral displacement. Summation of the dissi-
displacements are also indicated in Table 2. In this table pated energy with respect to the increment in lateral drift results
Mþ  þ 
L ; M L ; M R and M R indicate the positive and negative bending in cumulative dissipated energy up to each given level of interstory
moments in the left or right beams. Corresponding values for their drift. The evolution of the dissipated energy for retrofitted and cor-
virgin state and the percentage of the increase in their flexural responding specimen in virgin state is presented in Fig. 10. During
capacities achieved after the retrofitting are also reported in this all loading steps, both retrofitting solutions have provided a cumu-
table. According to this data, after retrofitting, in average and for lative dissipated energy higher than the one registered in their cor-
the positive bending moments, up to 88% of flexural capacity of responding virgin state. In this respect, the retrofitting solution
the beams of JPA3 was recovered. For the negative bending applied in JPB specimen was more effective. In fact, at 4% of drift
moments, the flexural capacities of the beams in virgin state were the cumulative dissipated energy of JPA3-R was 44.4 kN  m, which
fairly restored. The retrofitting system adopted for JPB, however, was only 5% larger than the corresponding value in JPA3, while the
provided a much larger increase in resisting bending moments of JPB-R reached 53.4 kN  m indicating an increase of 95% comparing
the beams. Based on this retrofitting technique an average increase to value calculated for JPB.
of 49% and 71% for the positive and negative moments were
obtained, respectively. 3.5. Secant stiffness
It should be noted that the values registered for flexural resis-
tance of both retrofitted specimens do not necessarily represent Degradation in the stiffness of a beam–column joint can pro-
the flexural capacity of the beams, since the degradation in gressively occurs when it is subjected to reversal cyclic loading.
beam–column joint shear capacity was the prevailing failure To assess the stiffness degradation, the secant stiffness, Ks, is esti-
modes of both specimens. mated during the drift evolution, and its relationship is repre-
sented in Fig. 11, for both the specimen in the retrofitted and
3.4. Dissipated energy virgin states. The secant stiffness is taken as the slope of the
straight line which connects the peak loads at the positive and
Energy dissipation capacity of a RC element is the consequence the negative displacements of the load versus displacement
of inelastic deformation and damage propagation. Opening and envelop at the first cycle of each level of imposed drift. According
closing of cracks contribute significantly to the energy dissipation to this figure, the retrofitting technique adopted for JPA3-R has just
 E. Esmaeeli et al. / Composite Structures 122 (2015) 456–467 465

Drift (%) Drift (%)

1 2 3 4 1 2 3 4
JPA3-R JPAB-R
50 50 JPB
JPA3

Dissipated energy (kN.m )

Dissipated energy (kN.m)
40 40

30 30

20 20

10 10

0 0
0 10 20 30 40 50 0 10 20 30 40 50
Cycle number Cycle number
(a) (b)
Fig. 10. Evolution of the dissipated energy during the cyclic loading (a) JPA3-R and JPA3, and (b) JPB-R and JPB.

6000
JPA3-R JPB-R
5000
JPA3 JPB
Secant stiffness (kN/m) 5000
Secant stiffness (kN/m)

4000
4000

3000
3000

2000
2000

1000 1000

0 0
0 1 2 3 4 0 1 2 3 4
Drift (%) Drift (%)
(a) (b)
Fig. 11. Secant stiffness evolution in (a) JPA3-R and JPA3, and (b) JPB-R and JPB.

restored 82% of the initial secant stiffness of this specimen in its

virgin state, while the technique applied on the JPB-R has almost
restored the initial secant stiffness registered in JPB (its virgin
state). This can be explained by a less effective bond between the
casted mortar and the old concrete of JPA3-R.
Considering the degradation of the secant stiffness at the end of
each sets of loading cycles, JPA3-R had greater secant stiffness than
JPA3 between loading cycles corresponds to 0.13% and 1.67%. After
1.67% the secant stiffness of the retrofitted and virgin state was
fairly similar. For the case of JPB-R, after 0.13% of drift, the adopted
retrofitting scheme resulted in a slower degradation in secant stiff-
ness than its virgin state.

3.6. Displacement ductility

Ductility is the potential of a lateral load resisting system to

undergo large inelastic deformation during its post-peak regime Fig. 12. Schematic representation of the definition of the equivalent bilinear curve
for the evaluation of the displacement ductility index.
with only slight reduction in its ultimate lateral load carrying
capacity. The ductility is generally quantified as a normalized dis-
placement or a rotation index depending if the ductility is aimed to a load level in the post-peak response of the specimen that is a
be assessed in terms of global or local behavior, respectively. For fraction of the peak load (Fp). According to the available literature,
the case of the present study, the displacement ductility index this ratio can be taken between 10% and 20% [26–28]. The yield
(lD) is calculated as the ratio of the ultimate lateral displacement displacement can be obtained from a bi-linear curve assuming
(du) and the displacement at the yield point (dy), Fig. 12. The ulti- equivalent elastic–perfectly plastic response. To estimate this bi-
mate point can be defined as the displacement corresponding to linear curve, two conditions should be fulfilled: (i) the area under
466 E. Esmaeeli et al. / Composite Structures 122 (2015) 456–467

60 60
JPA3-R JPB-R
JPA3 JPB
40 40

20 20
Force (kN)

Force (kN)
0 0

-20 -20

-40 -40

-60 -60
-5 -4 -3 -2 -1 0 1 2 3 4 5 -5 -4 -3 -2 -1 0 1 2 3 4 5
Drift (%) Drift (%)
(a) (b)

Fig. 13. Envelope of the load versus drift for both the repaired and virgin specimens along with the equivalent elastic–perfectly plastic curves (a) JPA3-R and JPA3, and (b) JPB-
R and JPB.

Table 3
Data for the evaluation of displacement ductility factor.
 
Specimen Negative direction Positive direction lD lRD lVD
lVD
  þ þ
dy (mm) du (mm) dy (mm) du (mm)

JPA3-R 22.5 (0.75)* 79.2 (2.64) +16.5 (+0.55) +102.9 (+3.43) 4.88 +56%
JPA3 34.5 (1.15) 110.5 (3.68) +34.5 (+1.15) +105.2 (+3.51) 3.13
JPB-R 25.5 (0.85) 93.3 (3.11) +31.5 (+1.05) +97.2 (+3.24) 3.37 +12%
JPB 34.5 (1.15) 87.6 (2.92) +31.5 (+1.05) +108.9 (+3.63) 3.00
*
Values in parentheses indicate the corresponding drift in percentage at maximum bending moment.

this curve should be equal to the one of the envelope of load versus 4. Conclusions
lateral displacement, and (ii) the deviation between these two
curves, measured based on the absolute sum of the areas enclosed The effectiveness of retrofitting methodologies by jacketing the
between these curves, should be the minimum (see Fig. 12). The critical regions of two full-scale severely damaged reinforced con-
displacement ductility index is then calculated as the ratio crete (RC) interior beam–column joints was experimentally inves-
between the ultimate and the yield displacements. In this context tigated. Cast-in-place strain hardening cement composites (SHCC)
it was assumed for the ultimate displacement the one correspond- reinforced with carbon fiber reinforced polymer (CFRP) laminates
ing to 10% loss of the peak load (0.9Fp). The envelope of the load according to the near surface mounted (NSM) technique forms
versus drift, and also the equivalent elastic–perfectly plastic curves the main concept of the adopted retrofitting strategy. Two varia-
estimated for both the retrofitted and virgin specimens are pre- tions of this retrofitting technique were applied, where the main
sented in Fig. 13. Table 3 also indicates the yield and the ultimate difference is restricted to the number of the retrofitted sides of
displacements obtained for the calculation of the displacement the sections of the elements (2 sides in the JPA3-R and 4 sides in
ductility index for the positive and negative loading, where JPB-R specimens). Chemical anchors were used to improve shear
lVD and lRD are the ductility for the specimen in the virgin and ret- stress transference between retrofitting scheme and the existing
rofitted state, respectively. The reported ductility index is calcu- concrete substrate.
lated as the average ductility using the corresponding values of The developed SHCC was able to easily flow and fill the rela-
displacement ductility in both positive and negative displace- tively small gaps between formworks and the substrate without
ments. It is verified that, for both retrofitted specimens the average the need of any vibration, which is an important requisite for a cast
of the yield displacements, in negative and positive directions, has in place retrofitting intervention. Based on the results obtained
decreased when compared to the average value registered for their from experimental tests where cyclic lateral loading under a con-
corresponding specimens in the virgin state. The reduction of the stant column axial force was applied, the following conclusions
yield displacement is a consequence of lower stiffness degradation can be pointed out:
assured by the retrofitting system, mainly during the cycles up to
1.15% of drift. According to the results included in Table 3, and 1. Two-sided retrofitting system applied to the severely damaged
comparing to the displacement ductility registered in the speci- JPA3 specimen was capable of restoring the lateral load carrying
mens in its virgin state, the retrofitting strategy has assured an capacity and energy dissipation performance, and increase the
increase of 56% and 12% in displacement ductility of JPA3-R and ductility registered in the virgin state of this specimen. The ini-
JPB-R, respectively. The higher increase in displacement ductility tial secant stiffness of this specimen in its virgin state was, how-
of JPA3-R can be attributed to the larger sliding between the ever, not totally recovered (82%).
retrofitting scheme and the concrete substrate, and also due to 2. The four-sided retrofitting system applied in the severely dam-
the existence of larger damages before retrofitting of this aged JPB specimen assured a significant increase in terms of lat-
specimen. eral load capacity and energy dissipation when the
 E. Esmaeeli et al. / Composite Structures 122 (2015) 456–467 467

corresponding values registered in this specimen in its virgin [6] Esmaeeli E, Danesh F. Shear strengthening of 3D corner beam–column
connection using bidirectional GFRP layers. In: Motavalli M, editor.
state are considered for comparison purposes. A higher increase
Proceeding of fourth international conference on FRP composites in civil
of the flexural resistance for the beams was also obtained due to engineering (CICE2008). Zurich, Switzerland: International Institute for FRP in
the presence of CFRP laminates in the top and bottom faces Construction (IIFC); 2008.
of the beams, which has contributed to decrease the sliding of [7] Danesh F, Esmaeeli E, Alam MF. Shear strengthening of 3D RC beam–column
connection using GFRP: FEM study. Asian J Appl Sci 2008:1.
the flexural steel reinforcement of these beams. This technique [8] Ha G-J, Cho C-G, Kang H-W, Feo L. Seismic improvement of RC beam–column
has also decreased the rate of the stiffness degradation during joints using hexagonal CFRP bars combined with CFRP sheets. Compos Struct
the cyclic loading process, and assured a higher increase of 2013:95.
[9] Lee WT, Chiou YJ, Shih MH. Reinforced concrete beam–column joint
ductility than the two-sided retrofitting configuration. In com- strengthened with carbon fiber reinforced polymer. Compos Struct 2010:92.
parison with the substantial enhancement attained for these [10] Esmaeeli E, Barros JAO, Baghi H. Hybrid composite plates (HCP) for shear
mentioned seismic characters, the increase in displacement strengthening of RC beams. In Minho Uo, editor. 11th international
symposium on fiber reinforced polymers for reinforced concrete structures
ductility was only 12%. (FRPRCS11), Guimaraes; 2013. p. 11–4.
3. Although the governing failure mode for both specimens was [11] Shannag MJ, Alhassan MA. Seismic upgrade of interior beam–column
joint shear capacity deterioration, no brittle response was subassemblages with high-performance fiber-reinforced concrete jackets.
ACI Struct J 2005;102:8.
observed. [12] Wang Y-C, Hsu K. Shear strength of RC jacketed interior beam–column joints
4. Considering that the progress of the inclined cracks in the joint without horizontal shear reinforcement. ACI Struct J 2009;106:11.
region resulted in debonding failure between the adhesive of [13] Esmaeeli E, Manning E, Barros JAO. Strain hardening fibre reinforced cement
composites for the flexural strengthening of masonry elements of ancient
the X-shaped CFRP laminates and the SHCC, effectiveness of this
structures. Constr Build Mater 2013;38:1010–21.
configuration of CFRP laminates in the joint region is under ques- [14] Esmaeeli E, Barros JAO, Baghi H, Sena-Cruz J. Development of hybrid composite
tion. Therefore, bonding a horizontal or vertical arrangement of plate (HCP) for the repair and strengthening of RC elements. In: 3rd
transverse CFRP laminates at this region is recommended. international RILEM conference on strain hardening cementitious
composites (SHCC3-Delft). Dordrecht, Netherlands: RILEM Publications SARL;
5. A high capacity of stress redistribution in SHCC resulted in mul- November 3–5, 2014.
tiple crack formation around anchored regions, but no bearing [15] Lim YM, Li VC. Durable repair of aged infrastructures using trapping
failure was observed. mechanism of engineered cementitious composites. Cem Concr Compos
1997:19.
6. The final geometry of the retrofitted specimens was almost not [16] Sahmaran M, Li M, Li VC. Transport properties of engineered cementitious
affected by the proposed retrofitting interventions, but the seis- composites under chloride exposure. ACI Mater J 2007:104.
mic performance of these specimens was significantly improved. [17] Rokugo K, Kunieda M, Lim S. Patching repair with ECC on cracked concrete
surface. In Proc of ConMat05, Vancouver, Canada [CD-ROM]; 2005.
[18] Kobayashi K, Iizuka T, Kurachi H, Rokugo K. Corrosion protection performance
of high performance fiber reinforced cement composites as a repair material.
Acknowledgments Cem Concr Compos 2010:32.
[19] Sahmaran M, Li VC, Andrade C. Corrosion resistance performance of steel-
reinforced engineered cementitious composite beams. ACI Mater J 2008:105.
The study presented in this paper is a part of the research pro- [20] Fernandes CAL. Cyclic behaviour of RC elements with plain reinforcing bars
ject titled ‘‘PrePam–Pre-fabricated thin panels by using advanced [PhD]. Aveiro, Portugal: University of Aveiro; 2012.
[21] EN1992-1-1. Eurocode 2: Design of Concrete Structures. Part 1–1: General
materials for structural rehabilitation’’ with reference number of
Rules and Rules for Buildings. Brussels, Belgium: European Committee for
PTDC/ECM/114511/2009. The first author acknowledges the grant Standardization; 2004.
SFRH/BD/65663/2009 provided by FCT. The authors also thank [22] Esmaeeli E, Barros J, Mastali M. Effects of curing conditions on crack bridging
the collaboration of the following material suppliers: SIKA, Dow response of PVA reinforced cementitious matrix. In 8th RILEM international
symposium on fibre reinforced concrete: challenges and opportunities
Chemical Co., ENDESA Compostilla Power Station, S&P, Hilti, and (BEFIB2012). Guimaraes, Portugal; 2012.
also Artecanter for transportation of the specimens. [23] Costa IG, Barros JAO. Evaluation of the influence of adhesive properties and
geometry of carbon fiber laminates using pull-out tests. Department of Civil
Engineering, University of Minho; 2012. p. 16.
References [24] International Organization for Standardization – ISO 527–2:1996. Plastics –
Determination of tensile properties – Part 2: Test conditions for moulding and
[1] Hakuto S, Park R, Tanaka H. Seismic load tests on interior and exterior beam– extrusion plastics. London South Bank University.
column joints with substandard reinforcing details. ACI Struct J 2000;97:15. [25] Verderame GM, Fabbrocino G, Manfredi G. Seismic response of R.C. columns
[2] Pampanin S, Calvi GM, Moratti M. Seismic behaviour of R.C. beam–column with smooth reinforcement. Part II: Cyclic tests. Eng Struct 2008:30.
joints designed for gravity loads. In: 12th European conference on earthquake [26] Paulay T. Equilibrium criteria for reinforced concrete beam–column joints. ACI
engineering. London: Elsevier Science Ltd; 2002. Struct J 1989;86:8.
[3] Tsonos AG. Effectiveness of CFRP-jackets and RC-jackets in post-earthquake [27] EN 1998–3. Eurocode 8: Design of structures for earthquake resistance – Part
and pre-earthquake retrofitting of beam–column subassemblages. Eng Struct 3: Assessment and retrofitting of buildings; 2005.
2008:30. [28] Fardis M, Biskinis D. Deformation capacity of R.C. members, as controlled by
[4] Tsonos A-DG. Performance enhancement of R/C building columns and beam– flexure or shear. In: Performance based engineering for earthquake resistant
column joints through shotcrete jacketing. Eng Struct 2010:32. reinforced concrete structures: a volume honoring Shunsuke Otani. University
[5] Karayannis CG, Sirkelis GM. Strengthening and rehabilitation of RC beam– of Tokyo; 2003.
column joints using carbon-FRP jacketing and epoxy resin injection.
Earthquake Eng Struct Dyn 2008:37.