Composite Structures 92 (2010) 1200–1211

Contents lists available at ScienceDirect

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

Seismic evaluation of FRP strengthened RC buildings subjected to near-fault

ground motions having fling step
A. Mortezaei a, H.R. Ronagh b,*, A. Kheyroddin c
a
School of Civil Engineering, The University of Queensland, Brisbane, Australia
b
School of Civil Engineering, The University of Queensland, Brisbane, QLD 4072, Australia
c
Faculty of Civil Engineering, Semnan University, Semnan, Iran

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

Article history: Recordings from recent earthquakes have provided evidence that ground motions in the near field of a
Available online 11 November 2009 rupturing fault differ from ordinary ground motions, as they can contain a large energy, or ‘‘directivity”
pulse. This pulse can cause considerable damage during an earthquake, especially to structures with nat-
Keywords: ural periods close to those of the pulse. Failures of modern engineered structures observed within the
Rehabilitation near-fault region in recent earthquakes have revealed the vulnerability of existing RC buildings against
Retrofit pulse-type ground motions. This may be due to the fact that these modern structures had been designed
Near-fault earthquake
primarily using the design spectra of available standards, which have been developed using stochastic
FRP
Dynamic analysis
processes with relatively long duration that characterizes more distant ground motions. Many recently
RC building designed and constructed buildings may therefore require strengthening in order to perform well when
subjected to near-fault ground motions. Fiber Reinforced Polymers are considered to be a viable alterna-
tive, due to their relatively easy and quick installation, low life cycle costs and zero maintenance require-
ments. This paper presents the results of a study of the response of typical existing RC buildings to
near-fault ground motions and the potential improvements achievable after FRP retrofitting of the build-
ings. Results demonstrate the successful implementation of FRP with an improvement in stiffness,
strength and lateral displacement capacity of the rehabilitated structure. It is demonstrated that
strengthening with FRP is very effective in reducing drift demands for structures for a wide range of nat-
ural periods.
Ó 2009 Elsevier Ltd. All rights reserved.

1. Introduction quake, engineers and seismologists realized the potential damage

Impulsive type motions can cause considerable damage during
an earthquake, especially to structures with natural periods close
to those of the pulse. Near-fault effects can be broken down into
three types depending on the pulses, i.e. whether they are of accel-
eration, velocity, or displacement type. The velocity pulse motion,
sometimes referred to as ‘‘fling,” represents the cumulative effect
of almost all of the seismic radiation from the fault [1]. From a seis-
mological perspective, the velocity pulse is more commonly found
in earthquake records than are acceleration or displacement
pulses. Although, from an engineer’s perspective, the velocity pulse
is a better indicator of damage than the acceleration pulse, the
damage potential is also dependent on the peak displacement dur-
ing the pulse [2].
The displacement pulse without the high velocity pulse does
not have a high damage potential because the structure has time
to react to the displacements. After the 1971 San Fernando earth-
* Corresponding author. Tel.: +61 7 33659117; fax: +61 7 33654599.
E-mail address: h.ronagh@uq.edu.au (H.R. Ronagh).
that may occur due to the effects of near-fault ground motions
on structures. The damage observed during the 1994 Northridge,
California, the 1995 Kobe, Japan, the 1999 Izmit, Turkey, and the
2003 Bam, Iran earthquakes proved the engineer’s hypothesis that
structures located within the near-fault area suffered more severe
damage than structures located outside of this zone. These earth-
quakes provided a wealth of new information about the behaviour
of engineered structures because the respective epicenters were in
urban settings. Based on the data collected, building designers
started studying the near-fault effects on buildings. Their research
and findings led to the implementation of design factors in the
1997 Uniform Building Code [3] that began to account for near-
fault motions. Additional design factors to more accurately model
near-fault effects were implemented in the 2000 International
Building Code [4].
Although design codes and provisions introduce site-source and
distance-dependent near-source factors, the effectiveness of con-
stant amplification factors in providing adequate ductility levels
to structures located in the proximity of fault zones is question-
able. This is because current design spectra were developed using

0263-8223/$ - see front matter Ó 2009 Elsevier Ltd. All rights reserved.
doi:10.1016/j.compstruct.2009.10.017
 A. Mortezaei et al. / Composite Structures 92 (2010) 1200–1211
stochastic processes having relatively long duration that character-
izes more distant ground motions. On other hand, failures of mod-
ern engineered structures observed within the near-fault region in
recent earthquakes revealed the vulnerability of existing RC build-
ings against pulse-type ground motions.
Extensive investigations into the seismic performance of rein-
forced concrete frame buildings have been performed [5–9]. These
studies showed at a local level that most of the damage is likely to
occur in the beam-column joint panel zone and that the formation
of soft-story mechanisms can greatly impair the global structural
performance of the systems. A comprehensive overview of tradi-
tional seismic rehabilitation techniques was presented by FEMA-
547 [10]. Conventional techniques which utilize braces, jacketing
or infills, as well as more recent approaches including base isola-
tion and supplemental damping devices, have been considered to
strengthen the building and as such improve its behaviour. In the
past decade, an increased interest in the use of advanced non-
metallic materials has been reported.
The objective of this research is to use the wealth of recent
ground motion data to improve the understanding of the response
of typical reinforced concrete buildings to pulse-type ground mo-
tions that result from fling-step effects, and also to discuss the fea-
sibility and efficacy of a retrofitting intervention using FRP
composite materials in order to upgrade a far-fault earthquake de-
signed RC building to a near-fault one.
2. Research significance
Reinforced concrete is the material of choice for the majority of
structures in many places around the world. Because the design,
construction and maintenance of a building have been governed
and controlled by building codes and standards, the seismic vul-
nerability of an existing structure will be, to a certain extent, re-
lated to the seismic requirements of the codes used in its design.
One of the most effective ways of minimizing potential earth-
quake-related losses is to conduct reliable assessments of the vul-
nerability of existing structures and to develop and implement
effective ways of rehabilitating structures identified as being haz-
ardous. External bonding of fiber reinforced polymer (FRP) com-
posites is now a well-established technique for the rehabilitation
of RC structures.
In the current study, the results of 168 nonlinear time-history
(NTH) analyses that were conducted on six RC buildings are pre-
sented. In the presentation, the common failure modes exhibited
by these RC buildings during far-fault earthquakes will be dis-
cussed. The buildings were then strengthened using FRP compos-
ites and the analyses were re-done. The results show the
potential strengthening benefits and the changes in the structural
characteristics of the buildings.
3. Seismic risk reduction and rehabilitation of existing
buildings
Past earthquakes have demonstrated the power of nature and
the catastrophic impact of such power upon urban populations.
Protecting existing buildings through retrofitting is a challenge
for the engineers involved in the mitigation of seismic hazards.
Seismic retrofitting of buildings is a relatively new occupation for
most structural engineers. The retrofitting of a building requires
appreciation of the technical, economic and social aspects of the is-
sue at hand. In general, the practice of seismic risk reduction is to
follow a three-step process of screening, evaluation and
rehabilitation.
Screening entails assessing buildings to ascertain their level of
seismic risk, by following a simplified procedure whose main
1201
objective is to determine if the building should or should not be
subjected to a more detailed investigation. In the evaluation pro-
cess, a detailed investigation is performed on buildings with med-
ium to high priority as a result of the screening exercise.
Rehabilitation, i.e. seismic retrofit, becomes necessary if it is shown
that, through a seismic performance evaluation, the building does
not meet minimum requirements of the current building code, and
that it may suffer severe damage or even collapse during a seismic
event.
Once the potential losses in an existing building have been
quantified through a detailed seismic evaluation, a decision must
be made whether to: (1) take no action, (2) do minimum repair
and/or modifications, (3) retrofit, or (4) demolish the building
and replace it with a new building.
The decision on the course of action is usually complex because
it involves many factors, which may include:
(A) Technical considerations: level of expected damage (if col-
lapse may occur, options (1) and (2) are eliminated);
required level of safety; availability of reliable rehabilitation
techniques; and availability of technical expertise and
equipment.
(B) Economic considerations.
(C) Social and political considerations.
The selection of adequate rehabilitation strategies depends on
the particular building in question. Some of the main factors affect-
ing this selection are: (1) fundamental period of vibration; (2) ac-
tual strength; (3) initial stiffness; (4) deformability; (5) type of
mechanism; (6) local site conditions; (7) level of safety required;
and (8) level of seismic hazard.
Traditionally, steel plates were used for strengthening RC mem-
bers. The technique of steel plate bonding has proved to be a useful
means in structural strengthening works [11]. However, the sus-
ceptibility of steel to corrosion, the difficulties of installing heavy
steel plates in congested and confined spaces, and the restrictions
imposed by simple lifting and handling devices make it desirable
to seek other viable alternatives. The application of fiber reinforced
polymer (FRP) materials has opened an exciting new path for
structural repair and strengthening. The prime material types that
are useful as reinforcing fibbers for strengthening applications are
glass, carbon and aramid. FRPs have very low self weight but high
strength-to-weight ratio and do not exhibit any corrosion prob-
lems. This results in low maintenance costs. Such versatility may
predict a rapid increase in their usage.
4. Characteristics of near-fault ground motions
The near-fault of an earthquake can be defined as the area in the
close vicinity of the fault rupture surface. Besides strong shaking,
the characteristics of near-fault ground motions are linked to the
fault geometry and the orientation of the traveling seismic waves
[12]. Vertical strike–slip faults can produce a directivity effect,
and dip–slip faults can produce directivity effects as well as hang-
ing wall effects. Hanging wall effects are felt on the hanging wall of
a fault (the earth above a vertically dipping fault), and are due to
the proximity of much of the fault to hanging wall sites. Directivity
effects can be classified as forward, reverse, or neutral. Fig. 1 por-
trays the three zones of directivity, with the star representing the
epicenter and the black line indicating the fault.
The rupture directivity pulse is oriented in the strike-normal
direction and the static ground displacement (fling-step) is oriented
parallel to the fault strike. Fling step is a result of the evolution of
residual ground displacement due to tectonic deformation associ-
ated with rupture mechanism. This phenomenon is characterized
1202 A. Mortezaei et al. / Composite Structures 92 (2010) 1200–1211
Fig. 1. Zones of directivity [13].
by a unidirectional large amplitude velocity pulse and a monotonic
step in the displacement time history [13]. It arises in strike–slip
faults in the strike parallel direction or in the strike-normal direc-
tion for dip–slips faults. In general, the permanent displacement
is considered as the pseudo static deformation and its frequency
is smaller than that of the forward directivity pulse.
The velocity and displacement time histories of typical near-
fault ground motions having fling step (Sakarya record of 1999
Kocaeli earthquake) are compared to those histories of ordinary
far-fault motion (Taft record of 1952 Kern County earthquake) in
Fig. 2. This figure shows that the fault parallel component of the
Sakarya record exhibits apparent tectonic deformation at the end
of the displacement time history that is a typical signature of the
fling step.
5. Description of buildings used for evaluation
To select an appropriate retrofitting method, an accurate evalu-
ation of both the seismic performance and the condition of an
existing structure is necessary. Based on this evaluation, engineers
can choose the most effective retrofit among the various interven-
tion techniques and optimize the improvement in seismic perfor-
mance for an existing structure. Seismic deficiencies should first
be identified through a seismic evaluation of the structure.
Six existing reinforced concrete special moment-resisting frame
buildings of 3, 6, 10, 14, 16 and 19 stories were selected as repre-
sentative case studies to evaluate their seismic demands when
subjected to near-fault ground motions with fling step, and to com-
pare their respective responses to typical far-fault ground motions.
Buildings with three, six and ten stories possess a moment-resist-
ing frame system and the buildings with fourteen, sixteen and
nineteen stories, a wall-frame system. The rectangular plan of all
buildings measures 30 m  25 m. The floor plans and elevation
views of the buildings with beam and column sections are shown
in Fig. 3. The columns are embedded into grade beams and an-
chored to the top of the pile cap, essentially restraining displace-
ments and rotations in all directions. The buildings are assumed
to be fixed at the base with a damping ratio of 5% in all modes,
and the floors as rigid diaphragms with infinite in-plane stiffness.
The sections of structural elements are square and rectangular
and their dimensions are changed at different stories. The slab
thickness is 100 mm. For the sake of clarity, the column and beam
dimensions and reinforcement of the 10-story building are pro-
vided in Tables 1 and 2. Story heights of buildings are assumed
to be constant with the exception of the ground story. The modulus
of elasticity (Young’s modulus) E = 30 kN/mm2, Poisson’s ratio
m = 0.20 and the mass density q = 24 kN/m3 are assumed in all
models. The uniaxial strength for nonlinear modeling of the con-
crete is considered to be 35 MPa. The rebar is modeled as steel with
yield strength of 400 MPa and an ultimate strength of 600 MPa.
Permanent and imposed loads are assumed to be: dead load of
story level, 5.5 kPa; dead load of roof, 6 kPa; dead load of parti-
tions, 1 kPa; dead load of external walls, 2.5 kPa; live load of story
levels, 2 kPa; and live load of roof, 1.5 kPa.
Strengths of materials were: compressive strength of cylindrical
concrete specimen, 25 MPa; yielding strength of bending reinforce-
ment, 400 MPa; and yielding strength of transverse reinforcement,
30 MPa.
6. Ground motion database
The ground motion database compiled for nonlinear time-his-
tory (NTH) analyses constitutes a representative number of far-
fault and near-fault ground motions from a variety of tectonic envi-
ronments. A total of 14 records were selected to cover a range of fre-
quency content, duration and amplitude. Near-fault records were
chosen so as to consider the presence of fling-step effects. Hence
the assembled database can be investigated in two sub-data sets.
The first set contains seven ordinary far-fault ground motions re-
corded within 90 km of the causative fault plane from earthquakes
in the magnitude (MW) range of 6.5–7.4. The second set includes se-
ven near-fault ground motions characterized with fling-step effect.
These records come from earthquakes having a magnitude (MW)
range of 6.5–7.4, and recorded at closest fault distance of 0.0 to
10 km. Information pertinent to the ground motion data sets,
including station, component of earthquake and peak ground accel-
eration (PGA), peak ground velocity (PGV), and peak ground dis-
placement (PGD), are presented in Tables 3 and 4 and their elastic
acceleration response spectra are shown in Figs. 4 and 5.
Utilized in this study is a data processing technique proposed in
Iwan et al. [14] and refined in Iwan and Chen [15] to recover the
long period components from near-fault accelerograms. This pro-
cess has been extensively elaborated in Boore [16] and Boore
et al. [17].
7. Nonlinear dynamic analysis program
Nonlinear dynamic analyses were performed using a program
previously developed by the authors [18]. This program can be
used to predict the nonlinear behaviour of any plain, reinforced
or prestressed concrete structure that is composed of thin plate
members with plane stress conditions. This includes beams, slabs,
shells, girders, shear walls, or any combination of these structural
elements. Time-dependent effects such as creep and shrinkage can
also be considered.
7.1. Concrete properties
The concrete behaves differently under different types and
combinations of stress conditions due to the progressive micro-
cracking at the interface between the mortar and the aggregates
(transition zone). The propagation of these cracks under the ap-
plied loads contributes to the nonlinear behaviour of the concrete.
As shown in Fig. 6a, the uniaxial stress–strain curve of concrete
adopted in this study comprises two parts. The ascending branch
up to the peak compressive strength is represented by the equation
proposed by Ashour and Morley [19]:
E0 e
r¼     2 ð1Þ
1þ E0
2 e þ e
Esc emax emax
where E0 is the initial modulus of elasticity of the concrete, Esc is the
secant modulus of the concrete at the peak stress, r is stress, e is
 A. Mortezaei et al. / Composite Structures 92 (2010) 1200–1211 1203

200
150
Velocity [cm/sec] 100
50
0
-50
-100
-150
-200
0 5 10 15 20 25 30 35 40 45 50
Time [sec]

200
Displacement [cm]

150
100
50
0
-50
-100
-150
-200
0 5 10 15 20 25 30 35 40 45 50
Time [sec]

(a) Taft record of 1952 Kern County earthquake

200
150
Velocity [cm/sec]

100
50
0
-50
-100
-150
-200
0 5 10 15 20 25 30 35 40 45 50 55
Time [sec]

200
Displacement [cm]

150
100
50
0
-50
-100
-150
-200
0 5 10 15 20 25 30 35 40 45 50
Time [sec]

(b) Sakarya record of 1999 Kocaeli earthquake

Fig. 2. Typical velocity and displacement time histories of: (a) far-fault, (b) near-fault (fling-step) ground motions.

Fig. 3. Structural configuration of: (a) 3-, 6- and 10-story; and (b) 14-, 16- and 19-story buildings (units: meter).
1204 A. Mortezaei et al. / Composite Structures 92 (2010) 1200–1211

Table 1
Dimensions and amount of reinforcement of columns and shear walls in the 10-story building.

Building Story Corner column Perimeter column Internal column Shear wall
2 2 2
Dim.(cm) Reo. (mm ) Dim. (cm) Reo. (mm ) Dim. (cm) Reo. (mm ) Thickness (cm)
10-story 1 60  60 2100 60  60 1880 60  60 1930 –
2 60  60 1200 60  60 1200 60  60 1200 –
3 60  60 1200 60  60 1200 60  60 1200 –
4 60  60 1200 60  60 1200 60  60 1200 –
5 50  50 910 50  50 1335 50  50 950 –
6 50  50 840 50  50 910 50  50 840 –
7 50  50 840 50  50 900 50  50 840 –
8 40  40 775 40  40 1070 40  40 840 –
9 40  40 600 40  40 880 40  40 600 –
10 40  40 600 40  40 880 40  40 600 –

Table 2
Dimensions and amount of reinforcement of beams in 10-story building.

Story Beams of external frames Beams of internal frames

Type 4 Type 3 Type 2 Type 1
2 2 2
Dim. (cm) h.w. Reo. (mm ) Dim. (cm) h.w. Reo. (mm ) Dim. (cm) h.w. Reo. (mm ) Dim. (cm) h.w. Reo. (mm2)
Bot. Top Bot. Top Bot. Top Bot. Top
1 60  50 1316 1664 60  50 1108 1583 60  50 1176 1670 60  50 1104 1593
2 60  50 1685 2071 60  50 1488 1972 60  50 1532 2093 60  50 1484 1981
3 60  50 1728 2141 60  50 1566 2054 60  50 1568 2176 60  50 1562 2059
4 60  50 1764 2181 60  50 1564 2043 60  50 1600 2223 60  50 1557 2055
5 50  45 1484 1931 50  45 1276 1823 50  45 1298 1965 50  45 1270 1835
6 50  45 1393 1860 50  45 1221 1770 50  45 1203 1903 50  45 1214 1778
7 50  45 1287 1737 50  45 1076 1607 50  45 1103 1779 50  45 1066 1625
8 40  40 947 1424 40  40 711 1334 40  40 738 1449 40  40 702 1352
9 40  40 665 1157 40  40 529 1100 40  40 570 1190 40  40 533 1110
10 40  40 495 760 40  40 473 727 40  40 502 773 40  40 492 757

Table 3
Far-fault ground motion database.

No. Earthquake Year Station Comp. MW Dis. (km) PGA (g) PGV (cm/s) PGD (cm)
1 Kern County 1952 Taft 111 7.4 81 0.17 17.47 8.83
2 Tabas 1978 Dayhook TR 7.4 107 0.4 26.17 9.1
3 Imperial Valley 1979 Calexico 225 6.5 90.6 0.27 21.23 8.98
4 Loma Prieta 1989 Presidio 000 6.9 83.1 0.099 12.91 4.32
5 Loma Prieta 1989 Cliff House 90 6.9 84.4 0.107 19.78 5.06
6 Manjil 1990 Abbar L 7.3 74 0.51 42.46 14.92
7 Kocaeli 1999 Ambarli 90 7.4 78.9 0.18 33.22 25.84

Table 4
Near-fault ground motion database.

No. Earthquake Year Station Comp. MW Dis. (km) PGA (g) PGV (cm/s) PGD (cm)
1 Kocaeli 1999 Sakarya 90 7.4 3.1 0.37 79.49 70.56
2 Chi-Chi 1999 TCU052 N 7.6 0.24 0.41 118.51 246.27
3 Chi-Chi 1999 TCU052 W 7.6 0.24 0.34 159.04 184.51
4 Chi-Chi 1999 TCU068 N 7.6 1.09 0.46 263.1 430.0
5 Chi-Chi 1999 TCU068 W 7.6 1.09 0.56 176.65 324.27
6 Chi-Chi 1999 TCU072 W 7.6 1.79 0.3 112.47 89.23
7 Chi-Chi 1999 TCU128 W 7.6 9.7 0.139 73.06 90.66

Data source: PEER (http://peer.berkeley.edu/smcat).

strain and emax is the strain at peak stress. The descending, or the the orthotropic constitutive relationship developed by Darwin
strain-softening, branch is idealized by the Bazant et al. model [20]: and Pecknold [21] is used for modeling the concrete using the
  smeared cracking idealization. The constitutive matrix, D, is given
e
r ¼ rc exp ð1  e=emax Þ ð2Þ by:
emax
2 pffiffiffiffiffiffiffiffiffiffi 3
where rc is compressive strength of the concrete. For uniaxially E1 m E1 E2 0
1 6 pffiffiffiffiffiffiffiffiffiffi 7
loaded concrete, rc is equal to fc0 . D¼ 4 m E E
1 2 E2 05 ð3Þ
ð1  m2 Þ p ffiffiffiffiffiffiffiffiffi
ffi
For analysis of most plane stress problems, concrete is assumed 0 0 14 ðE1 þ E2  2m E1 E2 Þ
to behave as a stress-induced orthotropic material. In this study,
 A. Mortezaei et al. / Composite Structures 92 (2010) 1200–1211 1205

2.5 normal to the cracks for the cracked concrete. The concept of the
Imperial Valley ‘‘equivalent uniaxial strain” developed by Darwin and Pecknold
Response Acceleration (g)

Kern County [21] is utilized to relate the increments of stress and strain in the
2 principal directions. Therefore, stress–strain curves similar to the
Kocaeli
Loma Cliff uniaxial stress–strain curves can be used to formulate the required
1.5 Loma Presidio stress–strain curves in each principal direction.
Manjil The strength of concrete, rc, and the values of E1, E2 and m are
TabasDayhook functions of the level of stress, and the stress combinations. The
1 concrete strength when subjected to biaxial stresses is determined
using the failure envelope developed by Kupfer et al. [22]. The val-
0.5
ues of E1 and E2 for a given stress ratio (a = r1/r2) are found as the
slopes of the r1  e1 and r2  e2 curves, respectively. For the
descending branches of both compression and tension stress–
0 strain curves, Ei is set equal to a very small number, 0.0001, to
0 1 2 3 4 avoid computational problems associated with a negative and zero
Period (sec) values for Ei. The concrete is considered crushed once the equiva-
lent compressive strain in the principal directions exceeds the ulti-
Fig. 4. Elastic acceleration response spectra of far-field ground motion recordings mate compressive strain of the concrete, ecu. For determination of
used in the evaluation of each building.
the concrete ultimate compressive strain, ecu, two models for
unconfined high and normal-strength concrete (Pastor [23]) and
confined concretes (Chung et al. [24]) are implemented into the
1.6 program.
Kocaeli In order to eliminate numerical problems after crushing (e > ecu)
1.4
Response Acceleration (g)

TCU52-N and cracking of the concrete (e > etu), a small value is assigned to
1.2 TCU52-W the compressive and tensile stresses as a fraction of concrete
TCU68-N strength, cc fc0 and ct ft0 , at a high level of stress (Fig. 6a), where
1
TCU68-W parameters cc and ct define the remaining compressive and tensile
strength factors, respectively.
0.8 TCU72-N
Regarding the stress–strain relationships for cyclic loading, it is
TCU128
0.6 important to distinguish between the unloading paths before and
after the compression strength (Fig. 6b) is exceeded. In the first
0.4 case, the unloading path is a straight line defined by the elastic
modulus E0 and tensile stresses are still possible. In the second
0.2 case, the unloading path does not reach the tensile region. After
exceeding the maximum tensile strength, cracks occur perpendic-
0
ular to the principal stress direction. Based on a smeared crack
0 1 2 3 4
model a smeared crack width is then calculated.
Period (sec)
7.2. Crack modeling techniques
Fig. 5. Elastic acceleration response spectra of near-fault ground motion recordings
used in the evaluation of each building.
Cracking of the concrete is one of the important aspects of non-
linear material behaviour of the concrete. Besides reducing the
in which E1 and E2 are the tangent moduli in the directions of the stiffness of the structure, cracks have resulted in redistribution of
material orthotropy, and m is the Poisson’s ratio. The orthotropic stresses into the reinforcing steel as well as increasing the bond
material directions coincide with the principal stress directions stress at the steel–concrete interface [25]. Cracking of the concrete
for the uncracked concrete and these directions are parallel and is idealized using the fixed smeared cracking model and is assumed

Fig. 6. The stress–strain relationship of concrete for: (a) monotonic loading; (b) cyclic loading.
1206 A. Mortezaei et al. / Composite Structures 92 (2010) 1200–1211

to occur when the principal tensile stress at a point (usually a

Gauss integration point) exceeds the tensile strength of the con-
crete. After cracking, the axes of orthotropy are aligned parallel
and orthogonal to the crack. The elastic modulus perpendicular
to the crack direction is reduced to a very small value, close to zero,
and the Poisson effect is ignored. The effect of the crack is smeared
within the element by modifying the [D] matrix. If r1 exceeds the
tensile strength of concrete, ft0 , the material stiffness matrix is de-
fined as (one crack is opened):
2 3
0 0 0
6 7
½D ¼ 4 0 E 2 0 5 where 0 < b  1:0 ð4Þ
0 0 bG
Once one crack is formed, the principal directions are not al-
lowed to rotate, and a second crack can form only when r2 > ft0 ,
in a direction perpendicular to the first crack. Then,
2 3
0 0 0
6 7
½D ¼ 4 0 0 0 5 where 0 < b  1:0 ð5Þ
0 0 bG
The shear retention factor, b, with a value of less than unity,
serves to eliminate the numerical difficulties that arise if the shear
modulus is reduced to zero, and more importantly, it accounts for
the fact that cracked concrete can still transfer shear forces
through aggregate interlock and dowel action. Due to the bond be-
tween the concrete and the steel reinforcement, a redistribution of
the tensile stress from the concrete to the reinforcement will occur.
In fact, the concrete is able to resist tension between the cracks in Fig. 8. Some typical finite elements in developed program.
the direction normal to the crack; this phenomenon is termed ten-
sion–stiffening.
uncracked, partially cracked, fully cracked, non-yielded, yielded
7.3. Reinforcing bar properties and crushed – depending on the stress or strain conditions.
Eight-node brick elements were used to model the concrete. These
The reinforcing bars are modeled as an elastic strain-hardening elements include a smeared crack analogy for cracking in tension
material as shown in Fig. 7. The reinforcing bars can be modeled zones and a plasticity algorithm to account for the possibility of
either as smeared layers or as individual bars. concrete crushing in compression regions.

7.4. Finite element formulation 7.5. Nonlinear solution

The element library includes plane membrane and plate bend-
ing as well as a facet shell element which is a combination of the
plane membrane and plate bending elements. Fig. 8 shows some
of these elements and the associated degrees of freedom. The pro-
gram employs a layered finite element approach. The structure is
idealized as an assemblage of thin constant thickness plate ele-
ments with each element subdivided into a number of imaginary
layers as shown in Fig. 8c. Each layer can assume any state – being
In nonlinear analysis, the total load applied to a finite element
model is divided into a series of load increments called load steps.
At the completion of each incremental solution, the stiffness ma-
trix of the model is adjusted to reflect nonlinear changes in struc-
tural stiffness before proceeding to the next load increment. The
Newton–Raphson equilibrium iterations for updating the model’s
stiffness were used in the nonlinear solutions. Prior to each solu-
tion, the Newton–Raphson approach assesses the out-of-balance
load vector, which is the difference between the restoring forces
and the applied loads. Subsequently, the program carries out a lin-
ear solution using the out-of-balance loads and then checks for
convergence. If convergence criteria are not satisfied, the out-of-
balance load vector is re-evaluated, the stiffness matrix is updated,
and a new solution is carried out. This iterative procedure contin-
ues until the results converge.

8. Seismic response evaluation of buildings

In total, 168 nonlinear time-history (NTH) analyses were con-

Fig. 7. The stress–strain relationship of steel.
ducted on the six buildings. Inter-story drift ratio (IDR), defined
as the relative displacement between two consecutive story levels,
displacements at different story levels, base shear force, base bend-
ing moment and shear forces at different story levels are used as
the primary measure of seismic demand. Additional demand mea-
sures, such as component and story ductility were also investi-
gated. In general, there is a reasonable correlation between the
 A. Mortezaei et al. / Composite Structures 92 (2010) 1200–1211 1207

inter-story drift demands and component/story-level ductility de-
mands; hence these results have not been included here. The peak
inter-story drift profiles obtained from NTH analyses of the build-
ings subjected to the two sets of ground motions (i.e. far-fault mo-
tions, near-fault motions with fling-step) are presented in Fig. 9.
For the 3-story building, far-fault motions produce nearly uni-
form inter-story drift demands for most records, with the excep-
tion of the Loma-Cliff record. In case of near-fault records, they
impose higher demands in comparison to far-fault records, though
the maximum drift is generally concentrated at the middle story
levels. The largest demand was caused by the TCU052-W record,
which produced 34.4 mm inter-story drift at the second story.
For the 6-story building, the maximum story demand for far-
fault records is observed to be either at the second or third story
levels and depends on the frequency content of the motion. For
near-fault records, the demands at the upper levels are much high-
er. Of the entire data set, the TCU068-W record generated the high-
est demand (31.8 mm inter-story drift) at the third story.
For the 10-story building, the TCU052-W record generated the
highest demand (35 mm inter-story drift) at the seventh story level.
Higher-mode effects are predominant in many of the near-fault re-
cords causing a shift in demands from the lower to upper stories.
For the 14-story building, the maximum story demand for far-
fault records is observed to be either at the 7th or 8th story levels

3-story building
4
4

Imperial Valley Kocaeli

3 3
Kern County TCU052-N

Kocaeli TCU052-W
Story

Story
2 Loma-Cliff 2 TCU068-N

Loma-Presidio TCU068-W

Manjil TCU072
1
1 TCU128
Tabas-Dayhook

0
0 0 10 20 30 40
0 10 20 30 Drift (mm)
Drift (mm)
(a) (b)

6-story building
7 7

6 6
Kocaeli
Imperial Valley
5 Kern County 5 TCU052-N
Kocaeli TCU052-W
4
Story

4
Story

Loma-Cliff TCU068-N
3 Loma-Presidio
3
TCU068-W

2 Manjil 2 TCU072
Tabas-Dayhook TCU128
1 1

0 0
0 5 10 15 20 0 10 20 30 40
Drift (mm) Drift (mm)
(a) (b)

10-story building
12
12

10 Imperial Valley
10 Kocaeli
Kern County TCU052-N
8 8
Kocaeli TCU052-W
Story
Story

Loma-Cliff 6 TCU068-N
6
Loma-Presidio
TCU068-W
4 Manjil 4
TCU072
Tabas-Dayhook
2 TCU128
2

0 0
0 5 10 15 20 25 0 10 20 30 40
Drift (mm) Drift (mm)
(a) (b)
Fig. 9. Maximum inter-story drift for each building subjected to: (a) far-fault earthquakes, (b) near-fault earthquakes with fling-step.
1208 A. Mortezaei et al. / Composite Structures 92 (2010) 1200–1211

14-story building
15 15
14 14
13 13
12 Imperial Valley 12 Kocaeli
11 Kern County 11 TCU052-N
10 10
Kocaeli
Story

9 9 TCU052-W

Story
8 Loma-Cliff 8 TCU068-N
7 7
Loma-Presidio 6 TCU068-W
6
5 Manjil 5 TCU072
4 Tabas-Dayhook
4
TCU128
3 3
2 2
1 1
0 0

0 5 10 15 0 5 10 15 20
Drift (mm) Drift (mm)
(a) (b)

16-story building
17 17
16 16
15 15
14 Imperial Valley 14 Kocaeli
13 13
12 Kern County 12 TCU052-N
11 11
Kocaeli Story 10 TCU052-W
Story

10
9 Loma-Cliff 9 TCU068-N
8 8
7 Loma-Presidio 7 TCU068-W
6 Manjil
6
5 TCU072
5
4 Tabas-Dayhook 4 TCU128
3 3
2 2
1 1
0 0
0 5 10 15 0 5 10 15 20 25
Drift (mm) Drift (mm)

(a) (b)

19-story building
20 20
19 19
18 18
17 17
16 16
15 Kocaeli
15
14 Imperial Valley 14 TCU052-N
13 Kern County 13
12 12 TCU052-W
Story

Kocaeli
11
Story

11 TCU068-N
10 Loma-Cliff 10
9 Loma-Presidio 9 TCU068-W
8 8
7 Manjil TCU072
7
6 Tabas-Dayhook 6 TCU128
5 5
4 4
3 3
2 2
1 1
0 0
0 5 10 15 20 0 10 20 30
Drift (mm) Drift (mm)

(a) (b)
Fig. 9 (continued)

and depends on the frequency content of the motion. Though sim-
ilar observations hold for near-fault records, the demands at the
upper levels are much higher. Of the entire data set, the TCU052-
N record generated the highest demand (18.84 mm inter-story
drift) at the roof story. Similar results were observed for 16- and
19-story buildings. In both buildings, the TCU052-N record gener-
ated the highest demand (21.1 mm and 24.5 mm inter-story drift
for 16- and 19-story buildings, respectively).
The variation in story demand for the far-fault records is less
significant. While higher-mode effects are expected to contribute
to the response of the 6- and 10-story buildings, the response of
the 3-story building demonstrates that even for low-rise buildings,
higher-mode effects could be significant.
9. Seismic retrofitting with fiber reinforced polymer
Fiber reinforced polymer (FRP) materials are composites con-
sisting of high strength fibers embedded in a polymeric resin
(Fig. 10). Fibers in an FRP composite are the load-carrying elements,
 A. Mortezaei et al. / Composite Structures 92 (2010) 1200–1211
Fig. 10. Representation of FRP material.
while the resin maintains the fibers’ alignment and protects them
against the environment and possible damage. Among commer-
cially available fibers, those made from carbon exhibit the highest
strength and stiffness when compared with steel. The type of fiber
is selected based on mechanical properties and durability require-
ments, while the type of resin depends upon environmental and
constructability needs.
Recent research and development efforts have led to many
applications of composite materials for strengthening existing
reinforced concrete structures. Ease of construction and broader
applications have made fiber composite sheets a more popular
choice than plates. While plates are appropriate for flat surfaces
and beams, sheets can be used on round (such as columns) and lar-
ger (such as walls) surfaces more efficiently and effectively. The
primary load-carrying element within a composite is the fiber.
Consequently, the fiber has a strong influence on the mechanical
characteristics of the composite such as strength and elastic mod-
ulus. The resin provides a mechanism for the transfer of load
among the fibers. It also protects the fibers from abrasion and other
environmental and chemical effects. The fibers can be oriented in a
single direction (unidirectional) or several directions, to optimize
the performance of the composite. Fiber strengthening technique
(by wrapping the element with fiber composite sheets) is a rela-
tively simple process. Fiber strengthening technology is among
the most efficient and effective new technologies for seismic retro-
fitting of RC structures. Its application is quite simple, very non-
intrusive to the building’s occupants, and not labor intensive –
making it one of the more desirable alternatives for the seismic ret-
rofit of existing buildings. The carbon fiber’s non-corrosive charac-
teristics and resistance to most chemicals give the carbon fiber
strengthening system a considerably longer life than the alterna-
tive systems.
Fig. 11. Formation of plastic hinge in upper levels of 10-story RC buildings.
1209
The above-mentioned structures were loaded by earthquake
loads, up to the development of a clear plastic collapse mechanism.
After that, these structures were retrofitted. The main aim of the
retrofitting system design was to change the failure mode from a
column-type to a beam-type collapse mechanism, forcing plastic
hinges to form in the beams. The FRP materials used for this reha-
bilitation had the Young modulus of 165 GPa.
The analytical results showed that typical RC buildings can be
subjected to large displacement demands at the arrival of the
velocity pulse, which require the structure to dissipate consider-
able input energy in a single or relatively few plastic cycles. This
demand will impact those structures with limited ductility capac-
ity. In contrast, far-fault motions build input energy more gradu-
ally, and the displacement demands are on average lower than
the demands in near-fault records.
Dynamic analysis demonstrated the migration of demands from
lower to upper stories. Therefore, according to the nonlinear finite
element analysis, a few plastic hinges had developed in the panel
zone and columns at higher levels.
For all of the near-fault earthquakes investigated in this study,
the severity of the demand is controlled by nonlinear finite ele-
ment analysis. Results show that near-fault motions characterized
by fling-step effects are potentially more damaging in the upper
one-third of buildings (Fig. 11). Therefore, in this paper, all panel
zones and top and bottom of columns were strengthened with
FRP sheets (Fig. 12).
10. Results of analyses
All buildings were analysed under the near-fault ground mo-
tions. The shear strength of a column tended to degrade faster than
its flexural strength with cycling of the lateral load. For beam
members, flexural cracks were observed. In some cases, flexural
cracks occurred on the bottom face of the beams. Shear cracking
was observed in shear walls and beam-column joints during the
near-fault earthquakes.
For the sake of clarity, results of the demands on the interior
frame of the 10-story building experiencing the largest demand
among each ground motion category are presented. For the 10-
story building, plastic hinges occurred at the upper levels of the
building (Fig. 11). Due to the significant contribution of joint fail-
ures to the collapse of buildings during earthquakes, it is necessary
to rehabilitate the joints. These locations were selected for rehabil-
itation with FRP.
On average, the rehabilitated buildings had a total shear force
capacity V, 1.5 times that of the original buildings as shown in
1210 A. Mortezaei et al. / Composite Structures 92 (2010) 1200–1211

200000

175000

Cumulative energy (kN-m)

150000

125000

100000

75000

50000

25000 Original
Rehabilitated
0
0 2 4 6 8 10
Drift ratio (%)

Fig. 14. Energy dissipation curves for original and rehabilitated buildings.

11. Buildings with shear wall

In earthquake resistant design, shear walls are common lateral

load resistance systems found in many reinforced concrete struc-
tures. Since the 1950s, shear walls have been accepted as effective
alternatives to moment resistant frames for providing the main
earthquake resistance mechanism in the seismic design of concrete
Fig. 12. Scheme of FRP strengthening.
structures. The results of nonlinear dynamic analyses show that,
for wall-frame system buildings, which are strengthened by shear
walls, the maximum drift demands are reduced. The analyses show
that some cracks have occurred in the walls at contra-flexure point
10 Story
900 of buildings.
Many different methods of seismic strengthening and repair of
800
shear wall structures have been developed and tested in the last
700 30 years. These techniques include the strengthening of existing
Total Shear (kN)

600 shear walls by the application of shotcrete or ferrocement, filling

500
400
300
200
Original
100
Rehabilitated
0 the lateral stiffness of damaged and undamaged shear walls, it is
0 50 100 150
Roof Displacement (mm)
Fig. 13. Relationship between total shear and roof displacement in 10-story
building.
Fig. 13. Application of FRP increased the stiffness and ductility of
the rehabilitated buildings. The rehabilitated buildings had an elas-
tic stiffness which was 1.4 times that of the original building.
Fig. 14 shows the energy dissipation curves for the original and
rehabilitated buildings. The cumulative energy dissipation for
rehabilitated specimens was 2.3 times that of the original building,
on average. This demonstrates that the energy dissipation capacity
of the frames increased when they were rehabilitated using FRP.
A comparison of the results of the 3-, 6- and 10-story analyses
shows that the effect of FRP rehabilitation in increasing the story
shear capacity is more significant for the high-rise building than
for the low-rise building.
in openings with reinforced concrete and so on [26]. Research
studies have shown that the concrete jacketing of a damaged shear
wall can restore the gravity load-carrying capacity of the wall, but
is not as effective when the objective is to restore the lateral stiff-
ness of the structure. For the purpose of strengthening, an alterna-
tive upgrade is the addition of steel bracings to the shear walls.
Although this method is effective in recovering and enhancing
200 architecturally unappealing since it changes the exterior and/or
interior layout of the structure, and results in a significant reduc-
tion of the building’s usable space. It may also add significant
weight to the structure and thus alter the magnitude and distribu-
tion of the seismic loads. Recently, applications of composite mate-
rials have been introduced as effective alternatives to conventional
techniques for the retrofit and repair of RC shear walls.
The repair by FRP sheets results in an increase of the load-car-
rying capacity compared to that of the original wall-frame build-
ing. The analytical results conclude that the application of
externally bonded carbon fiber sheets is an effective seismic
strengthening and repair procedure for reinforced concrete shear
walls. The carbon fiber repair system can be used to recover the
initial elastic stiffness and to increase the yield load and ultimate
flexural capacity of seismically damaged walls. In strengthening
applications, the carbon fiber sheets can be used to increase the
pre-cracked stiffness, the secant stiffness at yield and the cracking
load, the yield load and the ultimate flexural capacity of undam-
aged walls.
 A. Mortezaei et al. / Composite Structures 92 (2010) 1200–1211
12. Conclusion
The study summarized in this paper is intended to explore FRP
strengthening technique for RC buildings subjected to near-fault
earthquakes. Many reinforced concrete (RC) frame structures de-
signed according to the current codes are susceptible to abrupt
strength deterioration once the shear capacity of the columns is
reached. Fiber composites are used to increase the shear strength
of existing RC columns and beams by wrapping or partially wrap-
ping the members. Increasing the shear strength can alter the fail-
ure mode to a more ductile mode with higher energy dissipation
and inter-story drift ratio capacities.
The objective of this study was to analytically evaluate the ef-
fect of fiber reinforced polymer (FRP) rehabilitation on the seismic
performance of six existing RC buildings when subjected to near-
fault ground motion records. The study investigates the effect of
the FRP strengthening on the capacities of the structures with re-
spect to maximum inter-story drift ratio, maximum shear story
and maximum energy dissipation. It was found that the maximum
inter-story drift value decreases with the FRP strengthening, which
leads to a more ductile behaviour.
In general, columns are critical elements in any structural sys-
tem and their performance during a seismic event can dominate
the overall outcome of the structure. The rehabilitated buildings
possessed an elastic stiffness 1.4 times that of the original build-
ings. The energy dissipation of the rehabilitated buildings was
2.3 times that of the original buildings.
References
[1] Somerville P. The characteristics and quantification of near-fault ground
motion. In: Proceedings of the FHWA/NCEER workshop on the national
representation of seismic ground motion for new and existing highway
facilities. Burlingame, California; May 1997.
[2] Hall JF, Heaton TH, Halling MW, Wald DJ. Near-source ground motion and its
effects on flexible buildings. Earthquake Spectra 1995;11:569–605.
[3] International Conference of Building Officials. Uniform Building Code, Whitter,
California; 1997.
[4] International Conference of Building Officials. International Building Code,
Whitter, California; 2000.
[5] Hadi MNS. Retrofitting of shear failed reinforced concrete beams. Compos
Struct 2003;62(1):1–6.
[6] Pham H, Al-Mahaidi R. Assessment of available prediction models for the
strength of FRP retrofitted RC beams. Compos Struct 2004;66(1–4):601–10.
1211
[7] El-Dakhakhni WW, Hamid AA, Hakam ZHR. Hazard mitigation and
strengthening of unreinforced masonry walls using composites. Compos
Struct 2006;73(4):458–77.
[8] Colomb F, Tobbi H, Ferrier E, Hamelin P. Seismic retrofit of reinforced concrete
short columns by CFRP materials. Compos Struct 2008;82(4):475–87.
[9] Promis G, Ferrier E, Hamelin P. Effect of external FRP retrofitting on reinforced
concrete short columns for seismic strengthening. Compos Struct
2009;88(3):367–79.
[10] Techniques for the Seismic Rehabilitation of Existing Buildings, FEMA 547;
October 2006.
[11] Daudey X, Filiatrault A. Seismic evaluation and retrofit with steel jackets of
reinforced concrete bridge piers detailed with lap-splices. Can J Civil Eng
2000;27:1–16.
[12] Somerville P. Characterization of near field ground motions. US–Japan
workshop: effects of near-field earthquake shaking, San Francisco; March
2000.
[13] Orozco G, Ashford SA. Effects of large pulses on reinforced concrete bridge
columns, Pacific Earthquake Engineering Research Center. PEER Report 2002/
23, Berkeley: College of Engineering, University of California; 2002.
[14] Iwan WD, Moser MA, Peng CY. Some observations on strong-motion
earthquake measurements using a digital accelerograph. Bull Seismol Soc
Am 1985;75:1225–46.
[15] Iwan WD, Chen XD. Important near-field ground motion data from the Landers
earthquake. In: Proceedings of the 10th European conference on earthquake
engineering, Vienna, Austria; 1994.
[16] Boore D. Effect of baseline correction on displacements and response spectra
for several recordings of the 1999 Chi-Chi, Taiwan, earthquake. Bull Seismol
Soc Am 2001;91(5):1199–211.
[17] Boore D, Stephens CD, Joyner WB. Comments on baseline correction of digital
strong motion data: examples from the 1999 Hector Mine California
earthquake. Bull Seismol Soc Am 2002;92(4):1543–60.
[18] Mortezaei A. A program for three-dimensional nonlinear dynamic analysis of
reinforced concrete buildings, School of Civil Engineering. Brisbane
(Queensland): The University of Queensland; 2009.
[19] Ashour AF, Morley CT. Three-dimensional nonlinear finite element modelling
of reinforced concrete structures. Finite Elem Anal Des 1993;15(1):43–55.
[20] Bazant ZP, Belytschko T, Yul-Woong H, Ta-Peng C. Strain-softening materials
and finite-element solutions. Comput Struct 1986;23(2):163–80.
[21] Darwin D, Pecknold DA. Nonlinear biaxial stress–strain law for concrete. ASCE
J Eng Mech Div 1977;103(EM4):229–41.
[22] Kupfer HB, Gerstle KH, Rusch H. Behavior of concrete under biaxial stresses.
ACI Struct J 1969;66(8):656–66.
[23] Pastor JA. High-strength concrete beams, Ph.D. Thesis. Ithaca (New York):
Cornell University; 1986.
[24] Chung H, Yang K, Lee Y, Eun H. Stress–strain curve of laterally confined
concrete. Eng Struct 2002;24(9):1153–63.
[25] Sundara Raja Iyengar KT, Raviraj S, Jayaram TN. Analysis of crack propagation
in strain-softening beams. Eng Fract Mech 2002;69(6):761–78.
[26] American Society of Civil Engineers (ASCE). Prestandard and commentary for
the Seismic rehabilitation of buildings, prepared for the SAC joint venture,
published by the Federal Emergency Management Agency, FEMA-356,
Washington (DC); 2000.