See discussions, stats, and author profiles for this publication at: https://www.researchgate.

net/publication/257930218

Seismic Retrofit of Beam-to-Column Joints with High-Strength Fiber Jackets

Article · January 2002

CITATIONS READS

5 2,830

2 authors:

Alexander- Dimitrios GEORGE Tsonos Kosmas C. Stylianidis

Aristotle University of Thessaloniki Aristotle University of Thessaloniki
58 PUBLICATIONS   753 CITATIONS    32 PUBLICATIONS   491 CITATIONS   

SEE PROFILE SEE PROFILE

Some of the authors of this publication are also working on these related projects:

Stability of Structures View project

All content following this page was uploaded by Alexander- Dimitrios GEORGE Tsonos on 20 October 2015.

The user has requested enhancement of the downloaded file.

 Engineering Structures 30 (2008) 777–793
www.elsevier.com/locate/engstruct

Effectiveness of CFRP-jackets and RC-jackets in post-earthquake and

pre-earthquake retrofitting of beam–column subassemblages
Alexander G. Tsonos
Aristotle University of Thessaloniki, School of Engineering, 541 24 Thessaloniki, Greece

Received 28 February 2006; received in revised form 14 March 2007; accepted 7 May 2007
Available online 2 July 2007

Abstract

This paper presents the findings of an experimental study to evaluate retrofit methods which address particular weaknesses that are often found
in reinforced concrete structures, especially older structures, namely the lack of sufficient flexural and shear reinforcement within the columns
and the lack of adequate shear reinforcement within the joints. Thus, the use of a reinforced concrete jacket and a high-strength fibre jacket for
cases of post-earthquake and pre-earthquake retrofitting of columns and beam–column joints was investigated experimentally and analytically. In
this paper, the effectiveness of the two jacket styles was also compared.
c 2007 Elsevier Ltd. All rights reserved.

Keywords: Evaluation and retrofit; Buildings; Structural response concrete; Composite materials; Cement grout

1. Introduction four identical reinforced concrete exterior beam–column-

Damage caused by earthquakes over the years, has
indicated that some reinforced concrete buildings designed
and constructed in the 1960s and 1970s were found to have
serious structural deficiencies. These deficiencies are primarily
a consequence of a lack of capacity design approach and/or
poor detailing of reinforcement. As a result, lateral strength
and ductility of these structures were minimal [10,11,14].
The wrapping of reinforced concrete members with fiber-
reinforced polymer (FRP) sheets including carbon (C), glass
(G), or aramid (A) fibres, bonded together in a matrix made
of epoxy, vinylester or polyester, has been used extensively
throughout the world in numerous retrofit applications in
reinforced concrete buildings. These are recognized as alternate
strengthening systems to conventional methods, such as steel
plate bonding and shotcreting [2,3,8,15,18,21].
The feasibility and technical effectiveness of the high-
strength fibre jacket system and the reinforced concrete
jacket system both in the cases of post-earthquake and
pre-earthquake retrofitting of columns and beam–column
joints is investigated and presented in this paper. Thus,
E-mail address: tsonosa@civil.auth.gr.
slab-transverse beam subassemblages (O2 , P2 , F1 and S1 )
were constructed with non-optimal design parameters such as
flexural strength ratio and joint shear stress, with less column
transverse reinforcement than that required by the modern
Codes [6,7,9] and without joint transverse reinforcement,
representing the common construction practice for column and
beam–column joints in older structures built in the 1960s and
1970s.
The subassemblages F1 and O2 were subjected to cyclic
lateral load histories so as to provide the equivalent of
severe earthquake damage. The damaged specimens were
then strengthened by high-strength fibre jacketing or by four-
sided reinforced concrete jacketing. These jackets were applied
in the columns and in b/c joint regions of the damaged
subassemblages. The subassemblages S1 and P2 represent
parts of an old frame structure which has been upgraded to
resist strong future earthquakes. These two subassemblages
were tested only after strengthening by high-strength fibre-
jacketing or by four-sided reinforced concrete jacketing. These
jackets were also applied in the columns and b/c joint regions
of the subassemblages S1 and P2 . The four repaired and
strengthened subassemblages were subjected to cyclic lateral
load history so as to provide the equivalent of severe earthquake
damage.

c 2007 Elsevier Ltd. All rights reserved.

0141-0296/$ - see front matter

doi:10.1016/j.engstruct.2007.05.008
778 A.G. Tsonos / Engineering Structures 30 (2008) 777–793
Fig. 1. Dimensions and cross-sectional details of original specimens O2 , P2 , F1 and S1 (dimensions in m).
A direct comparison of the load deflection envelopes of
the original and retrofitted subassemblages was provided in
the paper. The effectiveness of the two jacket styles was also
compared. It is worth noting that studies referring to this
significant comparison are scarce in the relevant literature.
2. Description of the specimens
2.1. Original test specimens O2 , P2 , F1 and S1
Four identical test specimens (O2 , P2 , F1 and S1 ) were
constructed using normal weight concrete and deformed
reinforcement. Deformed bars are high-bond bars with ribs.
All the specimens were typical of existing older structures
built in the 1960s and 1970s. In “Recommendations for Design
of Beam–Column Joints in Monolithic Reinforced Concrete
Structures (ACI 352R-02)”, The ACI-ASCE Committee
specifies the
p maximum allowable joint shear stresses in the
form of γ f c0 MPa, where joint shear stress factor γ is a
function of the joint type (i.e. interior, exterior and so on)
and of the severity of the loading, and f c0 is the concrete’s
compressive strength. The flexural strength ratio M R is the sum
of the flexural capacity of columns to that of beam(s). The lower
limits of the flexural strength ratio M R and joint transverse
reinforcement are also specified by the Committee. Thus, for
the beam–column connections examined in this investigation,
the lower limits of M R and γ are 1.40 and 1.00 respectively.
In Fig. 1 the dimensions and cross-sectional details of
specimens O2 , P2 , F1 and S1 are shown. All these specimens
had less column transverse reinforcement than that required by
the new Greek Code for the Design of Reinforced Concrete
Structures (C.D.C.S.-2000) [9] or by Eurocode 2-2003 [6]
and Eurocode 8-2004 [7]. In addition, these specimens did
not have any joint transverse reinforcement (often ties in
the joint region were simply omitted in the construction
process in the past because of the extreme difficulty they
created in the placing of reinforcement), whereas the values
of flexural strength ratio were less than 1.40, p and those
of the joint shear stress were greater than 1.0 f c0 MPa
 A.G. Tsonos / Engineering Structures 30 (2008) 777–793
Table 1
Flexural strength ratio M R and the joint shear stresses factor γ of
subassemblages O2 , P2 , F1 , S1 , SO2 , SP2 , FRPF1 and FRPS1
Specimen MR a γa
O2 0.90 (1.40) 2.00 (1.00)
P2 0.90 (1.40) 2.00 (1.00)
F1 0.95 (1.40) 1.70 (1.00)
S1 0.95 (1.40) 1.70 (1.00)
SO2 2.40 (1.40) 0.54 (1.00)
SP2 2.40 (1.40) 0.65 (1.00)
FRPF1 1.95 (1.40) 1.70 (1.00)
FRPS1 1.95 (1.40) 1.70 (1.00)
a Numbers outside the parentheses are the provided values, numbers inside
the parentheses are the values required by the ACI–ASCE Committee 352-
02 [1].
for all the specimens O2 , P2 , F1 and S1 (see Table 1).
Thus, the beam–column connections of the original specimens
could be expected to fail in shear. The dimensions of the
test specimens were primarily dictated by the availability of
formwork and laboratory testing capacities, resulting in a
beam-to-column subassemblage model of approximately 1:2
scale. Low concrete compressive strength (measured on 150 ×
300 mm cylinders) was intentionally selected. At the time
of testing, the concrete compressive strengths of specimens
O2 , P2 , F1 and S1 were 16.20 MPa, 16.00 MPa, 22.00 MPa
and 21.80 MPa, respectively. Approximately 10 electrical-
resistance strain gauges were bonded in the reinforcing bars of
each specimen, O2 , P2 , F1 and S1 .
2.2. Strengthening technique: Specimens SO2 , SP2 , FRPF1
and FRPS1
Both original specimens F1 and O2 had experienced brittle
shear failure in the joint region. The strengthening of specimen
SO2 involved encasing the original beam–column joint and
the columns of O2 with a four-sided cement grout jacket
reinforced with additional collar stirrups in the joint region and
additional ties in the columns. To support the transverse steel,
additional longitudinal reinforcement was placed in each corner
of the jacket, which was then welded to the existing column
reinforcement. The welds were designed to carry 100% of the
additional bars strength. To improve the bond between the old
and new concrete, and to weld the new reinforcement to the
existing reinforcing bars, the concrete cover of the original
specimen O2 was chipped away and their surface roughened
by light sandblasting. A premixed, nonshrink, rheoplastic,
flowable and nonsegregating mortar of high strength with 0.95
cm maximum size of aggregate was used for the construction
of the cement grout jacket. Using wooden formwork, the
specimen was jacketed by an experienced contractor. The forms
used were rigid, sufficiently tight-fitting, and sealed to prevent
leakage. As shown in Fig. 2a, specimen SO2 had a four-sided
cement grout jacket, plus 14 longitudinal bars at each corner
of the column connected by 8 supplementary ties at 7 cm. All
longitudinal bars in the jacket extended through the joint region
of the subassemblage.
779
The beam-to-column joint is undoubtedly the most difficult
structural member to strengthen because of the great number of
elements assembled in this region. For the sake of simplicity in
construction, collar stirrups were used – instead of hoops – in
the buildings repaired in Greece [4,5]. For this reason, collar
stirrups – instead of hoops – were also used in the joint of the
strengthened specimen SO2 to increase its shear strength. These
collar stirrups were inclined bars 14 bent diagonally across
the joint core of SO2 , as shown in Fig. 2a. To support the collar
stirrups, 14 bar segments were used, which were welded
along the longitudinal bars of the jacket, as shown in detail
(1) of Fig. 2a. Two welded segments were used for each collar
stirrup. Further information regarding the collar stirrups used is
shown in detail (2) of Fig. 2a. The collar stirrups used in the
joint of specimen SO2 had the same amount of reinforcement
as that of the hoops required by the ACI-ASCE Committee 352-
02.
The columns of the strengthened specimen SO2 satisfied
all the requirements of the Greek Code for the Design
of Reinforced Concrete Structures (C.D.C.S.-2000) [9] and
those of Eurocode 2-2003 [6] and Eurocode 8-2004 [7].
Similarly, the b/c joint region of this specimen satisfied all
the requirements of the ACI-ASCE Committee 352-02 (see
Table 1). The subassemblage SO2 could, therefore, be expected
to develop flexural hinges in the beams without severe damage
concentrations in the joint regions.
The repair measures implemented in specimen F1 consisted
of: (1) the removal and replacement of all loose concrete in the
joint region by a premixed, nonshrink, rheoplastic, flowable and
non-segregating mortar of high strength, and (2) a high-strength
fibre jacketing in the joint region and on the columns (see
Fig. 2b). The repaired and subsequently strengthened specimen
was named FRPF1 . The design for the retrofit process with
carbon fibre-reinforced polymer sheets (CFRPs) was based on
E f = 230 GPa, t f = 0.165 mm (t f = layer thickness) and
ε f u = 1.5% (ε f u = ultimate FRP strain). The bonding of the
composite materials took place at a concrete age of about 40
days. Before the use of the CFRP sheets, careful preparation of
the specimen was made by grinding several areas to achieve a
fully smooth surface and a rounding of the corners (at a radius
of approximately 30 mm). CFRP sheets were then applied to
subassemblage FRPF1 in the following sequence:
Application of ten layers of CFRP sheets of 250 mm in
width to increase the shear strength of the joint, with the
fibres orientated in parallel to the longitudinal axis of the
beam (detail ¬ in Fig. 2b). As pointed out in the study by
Tsonos and Stylianidis [21], the anchorage of FRP sheets used
for the shear strengthening of joints is crucial. Debonding of
these sheets in this study took place after a few cycles of
loading, despite their sufficient anchorage length, which was
followed by rapid joint shear failure. Thus, in order to prevent
premature debonding of these sheets in specimen FRPF1 , the
existing slab was drilled close to the joint region, and the
beam of FRPF1 was fully wrapped with four layers of CFRP
sheets going through the slab opening. These sheets were
100 mm in width, with the fibers orientated perpendicular to the
longitudinal axis of the beam (see details ­ and ® in Fig. 2b).
780 A.G. Tsonos / Engineering Structures 30 (2008) 777–793

Fig. 2a. Jacketing of column and beam–column connection of subassemblages SO2 , SP2 (dimensions in m).

Next, nine layers of CFRP sheets of 200 mm in width were flexural strengths and, consequently, the flexural strength ratio
applied on the two opposite faces of the columns meeting at of specimen FRPF1 . The fibres of these CFRP sheets were
the joint, corresponding to tension and compression behaviour orientated in a vertical direction, parallel to the columns’
(see detail ° in Fig. 2b), in order to increase the columns’ longitudinal axis. The columns were also fully wrapped with
 A.G. Tsonos / Engineering Structures 30 (2008) 777–793 781

Fig. 2b. Jacketing of column and beam–column connection of subassemblages FRPF1 and FRPS1 (dimensions in m).

seven layers of CFRP sheets 450 mm in width to increase their
shear strength, with the fibres orientated perpendicular to the
longitudinal columns axis (see detail ¯ in Fig. 2b).
The bonding of the sheets took place in several main
stages, which included: (a) the application of a two-part epoxy
adhesive on the concrete; (b) the bonding of the first CFRP
layer; (c) the application of epoxy and the impregnation of the
sheet using a plastic roller; (d) the application of the next layer
of sheet; and so on up to the application of the last layer.
The subassemblages S1 and P2 represent parts of an old
frame structure which was upgraded to resist strong future
earthquakes. Therefore, the specimen P2 was tested after its
782 A.G. Tsonos / Engineering Structures 30 (2008) 777–793

Table 2
Original and strengthened specimens’ steel yield stress

Bar diameter Steel yield stress (MPa)

6 560
8 605
14 540

strengthening by reinforced concrete jacketing, as specimen

SP2 . On the other hand, the specimen S1 was tested after
its strengthening by high-strength fiber jacketing, as specimen
FRPS1 . The strengthening schemes of specimens SP2 and
FRPS1 were the same as those of specimens SO2 (Fig. 2a) and
FRPF1 (Fig. 2b) respectively. However, it is obvious that the
strengthening schemes of specimens SP2 and FRPS1 do not
include the removal and replacement of the loose concrete in
the joint regions with a premixed, high-strength mortar, as were
included in the strengthening schemes for specimens SO2 and
FRPF1 . Fig. 3a. Test setup (dimensions in m).
The original specimens F1 , S1 , O2 and P2 and the
strengthened specimens SO2 and SP2 were constructed
using deformed reinforcement (Note: 6, 8, 14 = bar
with diameter 6 mm, 8 mm, 14 mm respectively). The
subassemblages’ steel yield stresses are shown in Table 2. The
concrete compressive strength of the jackets of SO2 and SP2
were 40.70 MPa and 41.00 MPa respectively. Approximately
10 electrical-resistance strain gauges were bonded in the
reinforcing bars of each strengthened subassemblage, SO2 ,
SP2 , FRPF1 and FRPS1 .
In order to compare the effectiveness of the two jacket
styles, the corresponding strengthened subassemblages should
be of the same strength. Thus, each structural member (column,
joint) of specimens FRPF1 and FRPS1 were of almost the same
flexural and/or shear strength as that of specimens SO2 and SP2
respectively. Exactly for this reason, both specimens FRPF1 and Fig. 3b. Lateral displacement history.
FRPS1 had such a large number of layers of CFRPs (Fig. 2).
For the sake of brevity, all the computations relating to the imposed by a hydraulic jack, (see Fig. 3a). All the specimens
strengthening of specimens SO2 , SP2 , FRPF1 and FRPS1 are in O2 , SO2 , SP2 , F1 , FRPF1 and FRPS1 , were loaded transversely
Reference [23] and are therefore not included here. The basic according to the load history shown in Fig. 3b.
equations for computing the strengthening parameters, can be
found in [8], and in [14]. 4. Test results

3. Test setup-loading sequence The connections of both original subassemblages F1 and O2

The tests on specimens O2 , SO2 , SP2 , F1 , and FRPF1
were conducted in the testing frame shown in Fig. 3a. The
column inflection points are simulated with the aid of specific
arrangements that are connected to the testing frame and with
the free ends of the specimen columns by hinges. Thus, the
horizontal and vertical displacement of the columns’ ends
are restrained while able to rotate freely. Due to the manner
in which the specimens were attached to the test frame, no
magnification of the loads due to P–∆ effects was introduced.
Lateral loading was applied to the beam’s end by two one-
way actuators. The load applied by the actuators was measured
with two load cells attached to the specimens. The load point
displacement was measured by a potentiometer. A constant
axial load (150 kN) in the column of all the specimens was
exhibited, as expected, premature shear failure during the early
stages of seismic loading. Damage occurred both in the joint
area and in the critical regions of the columns. The beams in
both specimens F1 and O2 remained intact at the conclusion
of the tests (Fig. 4a). The failure mode of specimens SO2 ,
SP2 , FRPF1 and FRPS1 involved, as expected, the formation
of a plastic hinge in the beam near the column juncture. One
difference between the failure modes of specimens SO2 , SP2 ,
FRPF1 and FRPS1 was that more damage was concentrated in
the joint region of specimens FRPF1 and FRPS1 , as opposed
to that of specimens SO2 and SP2 . Views of the collapsed
subassemblages O2 , F1 , SO2 , SP2 , FRPF1 and FRPS1 are
shown in Fig. 4a. In order to detect the failure modes of
subassemblages FRPF1 and FRPS1 , the strengthening layers
of FRPs in both beams and beam–column joints were cut and
 A.G. Tsonos / Engineering Structures 30 (2008) 777–793 783

Fig. 4a. Views of the collapsed subassemblages O2 , F1 , SO2 , SP2 , FRPF1 and FRPS1 .

Fig. 4b. Post-damage views of the collapsed subassemblages FRPF1 and FRPS1 following removal of the reinforcing sheets.

subsequently removed. Thus, Fig. 4b reveals the damage pattern
that developed in subassemblages FRPF1 and FRPS1 .
Plots of applied shear-versus-drift angle for all the
specimens O2 , F1 , SO2 , SP2 , FRPF1 and FRPS1 are shown in
Fig. 5. Strengthened specimens SO2 and SP2 exhibited stable
hysteresis up to the 8th cycle of drift angle R of 5.0% and up
to the 4th cycle of drift angle R of 3.0% respectively. Beyond
that, a considerable loss of strength, stiffness and unstable
degrading hysteresis can be observed, especially in specimen
SP2 (Fig. 5b).
Subassemblages FRPF1 and FRPS1 , strengthened with
CFRP layers exhibited stable hysteresis up to the 5th cycle of
drift angle R of 3.5% and up to the 6th cycle of drift angle R
of 4.0%, respectively. Specimen FRPS1 showed a considerable
784 A.G. Tsonos / Engineering Structures 30 (2008) 777–793
Fig. 5a. Plots of applied shear-versus-drift angle for specimens O2 , F1 .
loss of strength, stiffness and unstable degrading hysteresis
beyond drift angle R ratios of 4%, while specimen FRPF1 did
not show any unstable degrading hysteresis (Fig. 5c).
In order to study the effectiveness of reinforced concrete
jacketing and fiber carbon/epoxy jacketing in improving the
earthquake resistance of columns and beam–column joints in
the cases of post-earthquake and pre-earthquake strengthening,
the seismic behavior of the strengthened specimens SO2 , SP2 ,
FRPF1 and FRPS1 was compared to that of the original ones
O2 and F1 . The original specimens O2 and F1 are, for all
practical purposes, the counterparts of the original specimens
P2 and S1 respectively (Table 1). Since subassemblages P2
and S1 had not been subjected to any lateral load before
strengthening, and in order to study the effectiveness of the
jackets in the case of pre-earthquake strengthening, it was
decided to compare the seismic behaviour of the strengthened
subassemblages SP2 , and FRPS1 with that of the original ones
O2 and F1 , respectively. Thus, O2 was used as the “control”
specimen for both strengthened specimens SO2 and SP2 . In
addition, subassemblage F1 was used as the “control” specimen
for both strengthened specimens FRPF1 and FRPS1 . Figs. 6 and
7 summarize the comparisons of the seismic behaviour of the
strengthened specimens SO2 and SP2 with that of the original
one O2 , while Figs. 8 and 9 summarize the comparisons of the
seismic behaviour of the strengthened specimens FRPF1 and
FRPS1 with that of the original one F1 (specimen S1 is similar
to specimen F1 , as shown in Fig. 1). As comparison parameters,
Fig. 5b. Plots of applied shear-versus-drift angle for specimens SO2 , SP2 .
those most critical in terms of the seismic behaviour of a R/C
substructure, such as stiffness, energy dissipation capacity and
strength, were chosen. Figs. 6–9 show comparisons of the peak-
to-peak stiffness (Figs. 6a–9a), energy dissipation capacity
(Figs. 6b–9b), and peak strength (Figs. 6c–9c) observed for
each load cycle of the relevant specimens.
The comparison of the performance of specimens SO2
(strengthened by reinforced concrete jackets in a post-
earthquake case) and SP2 (strengthened by reinforced concrete
jackets in a pre-earthquake case) with that of the original
specimen O2 indicated that the strengthened specimens
achieved significantly increased strength, stiffness and energy
dissipating capacities as compared to the original specimens,
even in large displacement amplitude cycles. Specimen SO2
shows up to 100% higher stiffness (Fig. 6a), up to 350% more
energy dissipated (Fig. 6b) and up to 200% higher strength
(Fig. 6c) than O2 , while specimen SP2 shows up to 80%
higher stiffness, 170% more energy dissipated and 120% higher
strength than O2 (Fig. 7). Specimen FRPF1 showed up to
50% higher stiffness, up to 135% higher energy dissipation
capacity and up to 170% higher strength than specimen F1
(Fig. 8). Subassemblage FRPS1 demonstrated up to 70% higher
stiffness, up to 200% higher energy dissipation capacity and up
to 190% higher strength than subassemblage F1 (Fig. 9).
In order to compare the effectiveness between the two
jacket styles it is worth studying: a. the comparison between
the strength, stiffness and energy dissipation capacity of the
specimen SO2 strengthened by reinforced concrete jacket
 A.G. Tsonos / Engineering Structures 30 (2008) 777–793
Fig. 5c. Plots of applied shear-versus-drift angle for specimens FRPF1 and
FRPS1 .
Fig. 6a. Stiffness comparison of the strengthened specimen SO2 to the original
one O2 .
(in a post-earthquake case) with those of the specimen
FRPF1 strengthened (also in a post-earthquake case) by high-
strength fibre jacket (Fig. 10), and b. the comparison of the
strength, stiffness and energy dissipation capacity between the
corresponding specimens (strengthened in a pre-earthquake
case) SP2 and FRPS1 (Fig. 11).
Moreover, in order to compare the effectiveness between the
pre-earthquake and post-earthquake type of strengthening, it
is interesting to compare a. the strength, stiffness and energy
dissipation capacity between specimens SO2 and SP2 (Fig. 12),
785
Fig. 6b. Energy dissipation comparison of the strengthened specimen SO2 to
the original one O2 .
Fig. 6c. Strength comparison of the strengthened specimen SO2 to the original
one O2 .
Fig. 7a. Stiffness comparison of the strengthened specimen SP2 to the original
one O2 .
and b. the above-mentioned parameters between specimens
FRPF1 and FRPS1 (Fig. 13).
From the diagrams in Fig. 10, it is evident that specimen
SO2 achieved a significant superiority in strength, stiffness
and energy dissipating capacities as compared with FRPF1 .
However, the seismic performance of specimen FRPS1 –
strengthened in a pre-earthquake case by a high-strength
fibre jacket – was almost the same as that of specimen SP2
strengthened in a pre-earthquake case by a reinforced concrete
jacket (Fig. 11). The reinforced concrete jackets and the CFRP
jackets act as external layers increasing the strength of the
786 A.G. Tsonos / Engineering Structures 30 (2008) 777–793
Fig. 7b. Energy dissipation comparison of the strengthened specimen SP2 to
the original one O2 .
Fig. 7c. Strength comparison of the strengthened specimen SP2 to the original
one O2 .
Fig. 8a. Stiffness comparison of the strengthened specimen FRPF1 to the
original one F1 .
upgraded structural members. As mentioned previously, each
structural member (column, joint) of specimens SO2 and SP2
was designed to have almost the same flexural and/or shear
strengths as those of specimens FRPF1 and FRPS1 respectively.
Thus, it was expected, theoretically speaking, that each pair of
subassemblages (SO2 , FRPF1 ) and (SP2 , FRPS1 ) would exhibit
almost the same seismic performance. This was confirmed by
a comparison between the seismic performance of SP2 and
FRPS1 (see Fig. 11). However, the removal and replacement
of the loose concrete in the joint region of subassemblage
SO2 by high strength mortar and the subsequent construction
of the reinforced concrete jacket for SO2 using the same
Fig. 8b. Energy dissipation comparison of the strengthened specimen FRPF1
to the original one F1 .
Fig. 8c. Strength comparison of the strengthened specimen FRPF1 to the
original one F1 .
Fig. 9a. Stiffness comparison of the strengthened specimen FRPS1 to the
original one F1 .
mortar, resulted in a beam–column joint in SO2 with new
dimensions of 340 mm × 340 mm × 300 mm of high-strength
(40.70 MPa). This was the reason why subassemblage SO2
exhibited superior seismic behaviour as compared to that of
subassemblage FRPS1 . If the damage in the joint region in O2
was not as severe (e.g. if the joint region demonstrated only
hairline cracks after the first seismic loading), then only epoxy
injections would be implemented for the repair of damage in
the joint. Thus, removal and replacement of damaged material
would not be carried out. In that case, the seismic behaviour
 A.G. Tsonos / Engineering Structures 30 (2008) 777–793 787

Fig. 9b. Energy dissipation comparison of the strengthened specimen FRPS1 Fig. 10b. Energy dissipation comparison between the strengthened specimens
to the original one F1 . SO2 and FRPF1 .

Fig. 9c. Strength comparison of the strengthened specimen FRPS1 to the

original one F1 . Fig. 10c. Strength comparison between the strengthened specimens SO2 and
FRPF1 .

Fig. 10a. Stiffness comparison between the strengthened specimens SO2 and
FRPF1 .
Fig. 11a. Stiffness comparison between the strengthened specimens SP2 and
FRPS1 .
of both SO2 and FRPF1 would have been almost the same.
Due to the removal of all loose concrete in the joint region of
subassemblage SO2 and its replacement by a premixed mortar
where γ (SO2 ) = 0.54 is lower than γ (SP2 ) = 0.65). This was
of high strength (40.70 MPa), and because of the particular
the reason why the seismic performance of SO2 (strengthened
construction of the reinforced concrete jacket, the new joint
region of SO2 (340 mm×340 mm×300 mm) was considerably in a post-earthquake case) was better than that of specimen SP2 ,
strengthened. This resulted in a significant decrease in the value strengthened in a pre-earthquake case (Fig. 12). The seismic
of the joint shear stress factor γ for specimen SO2 as compared behaviour of specimens FRPF1 and FRPS1 was almost the same
with the value of this factor for specimen SP2 (see Table 1, (Fig. 13).
788 A.G. Tsonos / Engineering Structures 30 (2008) 777–793

Fig. 12b. Energy dissipation comparison between the strengthened specimens

Fig. 11b. Energy dissipation comparison between the strengthened specimens
SO2 and SP2 .
SP2 and FRPS1 .

Fig. 12c. Strength comparison between the strengthened specimens SO2 and
Fig. 11c. Strength comparison between the strengthened specimens SP2 and
SP2 .
FRPS1 .

Fig. 13a. Stiffness comparison between the strengthened specimens FRPF1

Fig. 12a. Stiffness comparison between the strengthened specimens SO2 and
and FRPS1 .
SP2 .

where Vcd is the shear capacity of the concrete compression

5. Theoretical considerations zone according to [6,7], and Vwd is the shear carried by the web
reinforcement through the truss mechanism according to [7].
5.1. Specimens strengthened by reinforced concrete jackets The shear capacities of the strengthened beam–column joints
(SO2 and SP2 ) can be calculated as follows:
V Rd = Vcd + Vinclined (2)
The shear capacities of the strengthened columns can be
calculated as follows: where Vcd is the shear capacity of the concrete compression
zone according to [6,7], and Vinclined is the shear carried by the
V Rd = Vcd + Vwd (1) truss mechanism utilizing inclined bars (Fig. 2a). After the first
 A.G. Tsonos / Engineering Structures 30 (2008) 777–793
Fig. 13b. Energy dissipation comparison between the strengthened specimens
FRPF1 and FRPS1 .
Fig. 13c. Strength comparison between the strengthened specimens FRPF1 and
FRPS1 .
yielding, the shear carried by the inclined bars is
VS X = 2As f y sin θ
where As is the area of the inclined bars, f y is the yield stress
of these bars and θ is the inclination of these reinforcing bars to
the column axis. Strain gages data for the inclined bars of both
specimens SO2 and SP2 indicated that these bars yielded during
the first cycle of loading but remained active during the whole
test.
5.2. Specimens strengthened by high-strength fibre jackets
(FRPF1 and FRPS1 )
The shear capacities of the strengthened columns and
beam–column joints can be calculated as follows:
V Rd = Vcd + Vwd + VFRP
where Vcd is the shear capacity of the concrete compression
zone according to [6,7], Vwd is the shear carried by the web
reinforcement through the truss mechanism according to [7],
and VFRP is the FRPs contribution to shear capacity, which can
be written in the following form:
VFRP = 0.9ε f,e E f ρ f bw d
where d is the effective depth of the cross-section, bw is the
minimum width of the cross section over the effective depth,
789
ρ f is the FRPs reinforcement ratio equal to (2t f /bw ) for
continuously bonded shear reinforcement of thickness t f , E f
is the elastic modulus of FRPs in the principal fibre orientation,
and ε f,e is the design value of effective FRP strain, which is
given by the following expression for fully wrapped or properly
anchored FRPs [8]
 !0.3 
2/3
f cm
ε f,e = min 0.17ε f u , 0.006 (6)
Ef · ρf
where f cm is the mean value of the concrete compressive
strength.
5.3. Proposed shear strength formulation
A new formulation published in recent studies [20–22],
predicts the beam–column joint ultimate shear strength, and
was used in the present study to predict the actual values of
connection shear stress of the subassemblages O2 , F1 , SO2 ,
SP2 , FRPF1 and FRPS1 . A summary of this formulation is
presented in the following. The validity of the formulation
was checked using test data for more than 120 exterior and
interior beam–column subassemblages that were tested in the
Structural Engineering Laboratory at the Aristotle University
of Thessaloniki, as well as using data from similar experiments
carried out in the United States, Japan and New Zealand.
Fig. 14a shows a reinforced concrete exterior beam–column
joint for a moment resisting frame (As1 and As2 are the top
and bottom longitudinal beam reinforcement respectively, Mc1
and Mc2 are the moments acting on the top and bottom column
respectively framing at the joint, Mb is the moment acting on
(3)
the beam, Pu and Pu0 are the axial loads acting on the top and
bottom column respectively, Vcol and Vcol 0 are the shear forces
acting on the top and bottom column respectively, and Vb is
the beam shear force). Fig. 14a also shows the two mechanisms
of shear transfer and detail of the truss mechanism in the joint
core.
The shear forces acting in the joint core are resisted: (i)
partly by a diagonal compression strut and (ii) partly by a truss
mechanism formed by horizontal and vertical reinforcement
and concrete compression struts [12]. Both mechanisms depend
on the core concrete strength. Thus, the ultimate concrete
strength of the joint core under compression/tension controls
the ultimate strength of the connection. After failure of the
(4) concrete, strength in the joint is limited by gradual crushing
along the cross-diagonal cracks and especially along the
potential failure planes (Fig. 14a).
For instance, consider the section I-I in the middle of the
joint height (Fig. 14a). In this section, the flexural moment is
almost zero. The forces acting in the concrete are shown in
Fig. 14b. Ti are the forces acting in the longitudinal column
bars between the corner bars in the side faces of the column.
(5)
These bars compress the joint core through equal and opposing
directional forces. Each force acting in the joint core is analysed
into two components along the X and Y axes (Fig. 14b). Thus,
790 A.G. Tsonos / Engineering Structures 30 (2008) 777–793

Fig. 14a. External beam–column connection, the two mechanisms of shear transfer (diagonal concrete strut and truss mechanism) and detail of truss mechanism in
the joint core.

the vertically acting forces are: where h 0c and bc0 are the length and the width of the joint core
respectively (Fig. 14a).
The relationship between the average normal compressive
(7) stress σ and the average shear stress τ are shown in
Eq. (11):
where V jv is the vertical joint shear force [7].
The sum of the horizontally-acting forces also gives the V jv
horizontal joint shear force as σ = ·τ (11)
Vjh
Dcx + (D1x + · · · + Dvx ) = V j h . (8)
where
The normal vertical compressive stress σ and the shear stress
τ uniformly distributed over the whole section are given by the V jv hb
= =α [7] (12)
Eqs. (9) and (10): Vjh hc
Dcy + Dsy V jv where h b is the total depth of the beam and h c is the total depth
σ = 0 0
= 0 (9)
h c × bc h c × bc0 or width of square column (Fig. 14a). The principle stresses
Vjh (σ I = maximum, σ II = minimum) are given by Mohr’s circle
τ= (10) (Fig. 14c)
h 0c × bc0
 A.G. Tsonos / Engineering Structures 30 (2008) 777–793 791

Fig. 14b. Forces acting in the joint core concrete through section I-I from the two mechanisms.

hoop fibres are similar to conventional hoop reinforcing steel.

Confinement results in an increase in the apparent strength of
the concrete.
For a square or rectangular section wrapped with an
FRP-jacket and with corners rounded with a radius R,
the following equation gives the increased joint concrete
compressive strength due to confining [16,19]:
 0.7
tf
fc = f c0 + 6 2a f f d,c (16)
D
where t f is the jacket thickness, f f d,c = 0.95 f f k (where
f f k is the characteristic value of the FRP tensile strength),
a = 0.4+1.2( D R
) (where R/D is the ratio of the radius R to the
equivalent diameter D). The values of a should be reduced to
(2/3a) when the confining FRP-layers are greater than 5.0. The
b2 h2
equivalent diameter D is given by the expression D = 2h + 2b ,
where h, b are the section dimensions of the column or the
beam–column joint [4,19].
Fig. 14c. Stress state of element of the studied region and representation of √Substituting Eqs. (11)–(13) into Eq. (14) and using τ =
concrete biaxial strength curve by a parabola of 5th degree. γ f c gives the following expression:
" !#5 !
αγ
s r r
σ σ 4τ 2 4 5αγ 4
σI,II = ± 1+ . (13) √ 1+ 1+ 4 +√ 1+ 2 −1 =1
2 2 σ2 2 fc α fc α
(17)
Eq. (14) was suggested for representing the concrete biaxial
strength curve by a parabola of fifth degree [20, Fig. 14c]: Assuming here that:
σI
 5
σII α
−10 + =1 (14) x= √ (18)
fc fc 2 fc
where f c is the increased joint concrete compressive strength and
due to confining from the steel hoops, which is given by the
α
r
4
model of Scott et al. [17], according to the equation ψ= √ 1+ (19)
2 fc α2
f c = K · f c0 . (15a) then expression (17) can be transformed into
Also, f c0
is the concrete compressive strength and K is a
parameter of the model of Scott et al. [17] expressed as (x + ψ)5 + 10ψ − 10x = 1. (20)

ρs · f yh The solution of the system of equations (18)–(20) gives the

K =1+
f c0
where ρs is the volume ratio of transverse reinforcement, and
f yh its yield strength.
Confining a concrete member with an FRP-jacket is
accomplished by orientating the fibres transverse to the
longitudinal axis of the member. In this orientation, the
(15b) beam–column joint ultimate strength.
6. Comparison of predictions and experimental results
In the original subassemblages O2 and F1 , both the columns
and the beam–column joints are poorly detailed. Both these
structural elements have been identified as critical structural
792 A.G. Tsonos / Engineering Structures 30 (2008) 777–793

Table 3
Experimental and predicted values of the strength of subassemblages O2 , F1 , SO2 , SP2 , FRPF1 and FRPS1
h τ
Specimen Joint aspect ratio α = hb γcal γexp γult Predicted shear strength τpred a Observed shear strength τexp b µ = τpred
exp
c
√ √
O2 1.50 2.00 0.81 0.80 0.80 f c 0.81 f c 0.99
√ √
F1 1.50 1.70 0.87 0.93 0.93 f c 0.87 f c 1.06
√ √
SO2 0.88 0.50 0.56 0.98 0.54 f c 0.56 f c 0.96
√ √
SP2 0.88 0.65 0.60 0.82 0.65 f c 0.60 f c 1.08
√ √
FRPF1 1.50 0.94 0.85 1.67 0.94 f c 0.85 f c 1.10
√ √
FRPS1 1.50 0.95 0.90 1.67 0.95 f c 0.90 f c 1.05
a For γ √ √
cal ≥ γult , γpred = γult and τpred = γult f c MPa. For γcal √
< γult , γpred = γcal and τpred = γcal f c MPa. An overstrength factor a0 = 1.25 for the beam
steel is included in the computations of joint shear stress τcal = γcal f c MPa.
bτ √
exp = γexp f c MPa.

elements, which appear to fail prematurely, thus performing as
“weak links” in RC frames. In the retrofitted subassemblages
SO2 , SP2 , FRPF1 and FRPS1 both the columns and the
beam–column joints were strengthened and their strengthening
schemes were designed according to modern codes. Thus, both
these structural members do not perform as “weak links” in the
RC frames.
Consequently, the question arises as to how a model
which gives the ultimate strength of a reinforced concrete
beam–column joint and which predicts the actual value of the
joint shear stress can also be used for the prediction of the actual
value of the column shear stress and, more generally, for the
prediction of the actual values of shear forces and moments
developed in the beam–column subassemblages of the present
study during the tests. The answer can be found in Paulay and
Priestley [13], who clearly demonstrated that the shear forces
acting in the beam–column joints are significantly higher than
those acting in their adjacent columns. Thus, the joints fail
earlier than the columns during a strong earthquake motion.
Consequently, a model predicting the actual value of the
joint shear stress could also predict the shear stress of the
adjacent columns of a subassemblage and could also predict
the actual values of shear forces and moments resisted by the
subassemblages of the present study during the tests.
The comparison between experimental and predicted results
by the preceding methodology for all the specimens in the
present study is shown in Table 3. A particularly close
correlation can be observed.
It is worth mentioning here that the prediction of the
actual values of connection shear stress during an earthquake
also involves the prediction of the actual values of the
subassemblages’ M R ratio with the same degree of accuracy.
7. Conclusions
Based on the results described in this paper, the following
conclusions can be drawn.
1. Original specimens O2 and F1 representing an existing
beam–column subassemblage designed in accordance with
older codes, performed poorly under reversed cyclic lateral
deformations. The connections of these subassemblages
exhibited premature shear failure during the early stages of
seismic loading, and damage to both subassemblages was
concentrated in the joint region.
2. The retest of the failed beam–column subassemblages,
repaired and strengthened with reinforced concrete jacketing
or with fibre carbon/epoxy jacketing, showed that both of the
employed repair and strengthening techniques were effective
in transforming the brittle joint shear failure mode of original
specimens (O2 and F1 ) into a more ductile failure mode
with the development of flexural hinges into the beams.
Damage of the strengthened specimens SO2 and FRPF1 was
concentrated in both the beam’s critical region and in the
joint area.
3. The effectiveness of the reinforced concrete jacket system
and the high-strength fibre jacket system was demonstrated
in the cases of both post-earthquake and pre-earthquake
retrofitting of reinforced concrete columns and beam column
joints.
4. The seismic performance of specimen SP2 , strengthened in
the case of pre-earthquake by a reinforced concrete jacket,
was inferior to the performance of SO2 , strengthened by the
same upgrading scheme in a post-earthquake case.
5. It seems that the reinforced concrete jacket is more effective
in a post-earthquake retrofitting of columns and b/c joints
than the high-strength fibre jacket, while, in the case of pre-
earthquake strengthening, both a reinforced concrete jacket
and a high-strength fibre jacket seem to be equally effective.
6. A new formulation which predicts the beam–column joint’s
ultimate shear strength was used to predict the actual values
of the connection shear stress of all the subassemblages
investigated in the present study. In all cases, the observed
capacity was predicted to within approximately 10% of
that computed using the joint shear strength formulation
(Table 3).
Acknowledgements
The experimental part of this research investigation
was sponsored by the Earthquake Planning and Protection
Organization (E.P.P.O.). The author gratefully acknowledges
the support provided by the sponsor.
References
[1] ACI-ASCE Committee 352-2002. Recommendations for design of
beam–column joints in monolithic reinforced concrete structures (ACI
352R-02). American Concrete Institute. 37 pages.
 A.G. Tsonos / Engineering Structures 30 (2008) 777–793
[2] ACI Committee 440. State of the art report on fiber reinforced plastic
reinforcement for concrete structures (ACI 440R-96). Detroit: American
Concrete Institute; 1996. 68 pages.
[3] Antonopoulos C, Triantafillou T. Experimental investigation of FRP-
strengthened RC beam–column joints. Journal of Composites for
Construction 2003;7(1):39–49.
[4] Dritsos S. Repair and strengthening of reinforced concrete structures.
Patras. 2001. 309 pages [in Greek].
[5] Dritsos S. Seismic retrofit of buildings a greek perspective. Bulletin of
New Zealand Society for Earthquake Engineering 2005;38(3):165–81.
[6] Eurocode No 2. Design of concrete structures—Part 1-1: General rules
and rules of buildings, prEN 1992-1-1: 2003E. Commission of the
European Communities; 2003.
[7] Eurocode No 8. Design of structures for earthquake resistance, part I:
General rules, seismic actions and rules for buildings, EN 1998-1:2004E.
Commission of European Communities; 2004.
[8] FIB (CEB-FIP). Externally bonded FRP reinforcement for RC structures.
Technical report. Bulletin 14. 2001. 131 pages.
[9] Greek code for the design of reinforced concrete structures, (C.D.C.S.-
2000). Ministry of environment, land planning and public works, General
Secretariat of Public Works. Athens. 2000. 497 pages.
[10] Hakuto S, Park R, Tanaka H. Seismic load tests on interior and exterior
beam–column joints with substandard reinforcing details. ACI Structural
Journal 2000;97(1):11–25.
[11] Karayannis C, Chalioris C, Sideris K. Effectiveness of RC beam–column
connection repair using epoxy resin injections. Journal of Earthquake
Engineering 1998;2(2):217–40.
[12] Park R, Paulay T. Reinforced concrete structures. New York: John Wiley
Publications; 1975. 769 pages.
[13] Paulay T, Priestley MJN. Seismic design of reinforced concrete and
masonry buildings. John Wiley & Sons; 1992. 767 pages.
[14] Penelis GG, Kappos JA. Earthquake-resistant concrete structures. E & FN
793
SPON, An Imprint of Chapman & Hall; 1997. 572 pages.
[15] Priestley MJN, Seible F, Calvi GM. Seismic design and retrofit of bridges.
New York: A Wiley-Interscience Publication, John Wiley & Sons Inc.;
1996. 687 pages.
[16] Samaan M, Mirmiran A, Shahawy M. Model of concrete confined by
fiber composites. ASCE Journal of Structural Engineering 1998;124(9):
1025–31.
[17] Scott BD, Park R, Priestley MJN. Stress–strain behavior of concrete
confined by overlapping hoops at low and high strain rates. ACI Journal,
Proceedings 1982;79(1):13–27.
[18] Thermou GE, Elnashai AS. Seismic retrofit schemes for RC structures
and local–global consequences. Progress in Structural Engineering and
Materials 2006;8(1):1–15.
[19] Triantafillou T. Design of reinforced concrete and masonry structures
strengthened with FRP. In: Strengthening of reinforced concrete structures
with FRP. Meeting organized by the Technical Chamber of Greece.
Athens. 2000 [in Greek].
[20] Tsonos AG. Lateral load response of strengthened reinforced concrete
beam-to-column joints. ACI Structural Journal Proceedings 1999;96(1):
46–56.
[21] Tsonos AG, Stylianides K. Seismic retrofit of beam-to-column joints
with high-strength fiber-jackets. Journal of European Association for
Earthquake Engineering 2002;(2):56–72.
[22] Tsonos AG. Seismic repair of reinforced concrete beam–column
subassemblages of modern structures by epoxy injection technique.
Structural Engineering and Mechanics, An International Journal 2002;
14(5):543–63.
[23] Tsonos AG. Effectiveness of CFRP-jackets and RC-jackets in post-
earthquake and pre-earthquake retrofitting of beam–column subassem-
blages. Final report on research conducted under grant no. 100/11-
10-2000 from the Earthquake Planning and Protection Organization
(E.P.P.O.). 2003. 167 pages [in Greek].