Download as pdf or txt
Download as pdf or txt
You are on page 1of 42

2nd Reading

January 25, 2013 9:21 WSPC/S1793-4311/238-JET 1250026

Journal of Earthquake and Tsunami


Vol. 6, No. 4 (2012) 1250026 (42 pages)
c World Scientific Publishing Company
DOI: 10.1142/S1793431112500261

BEHAVIOR OF RC BUILDING COLUMNS UNDER


CYCLIC LOADING: EXPERIMENTAL STUDY

HUGO RODRIGUES
Department of Civil Engineering, University of Aveiro
Aveiro, 3810-193, Portugal
J. Earthquake and Tsunami Downloaded from www.worldscientific.com

ANTONIO ARÊDE
Department of Civil Engineering, Faculty of Engineering
University of Portol, Porto, 4200-465, Portugal
by WSPC on 01/24/13. For personal use only.

HUMBERTO VARUM and ANIBAL COSTA


Department of Civil Engineering, University of Aveiro
Aveiro, 3810-193, Portugal

Received 17 June 2010


Accepted 8 October 2010
Published 25 January 2013

An experimental campaign is underway in order to understand the behavior of rect-


angular columns under uniaxial and biaxial flexure. The present paper addresses the
initial part of that campaign, by reporting on two columns tested under uniaxial bend-
ing in one horizontal direction and then submitted to another uniaxial test in the other
direction. The test result analysis focuses on the contributions of flexure, shear and slip
components to the global deformation. The experimental results were compared with a
simplified method for column response. The Park and Ang damage index is calculated
and compared with the physical damage observed.

Keywords: RC columns; cyclic behavior; experimental study; flexural and shear defor-
mation; bar slippage; damage evolution.

1. Background and Objectives


The behavior of axially loaded reinforced concrete (RC) members combined with
uniaxial or biaxial bending is recognized as an important research topic. In fact,
the earthquake response of typical building structures is highly dependent on the
capacity of each structural member, but columns in particular play an essential
role in the final building capacity to withstand vertical loads in order to ensure life
safety compliance criteria.
Most of the present seismic design codes allow using non-linear analysis tools
for the safety assessment of existing reinforced concrete structures or for the design
of new ones. In order to perform non-linear analysis, it is essential to quantify

1250026-1
2nd Reading
January 25, 2013 9:21 WSPC/S1793-4311/238-JET 1250026

H. Rodrigues et al.

adequately some structural properties of RC elements, namely the initial stiffness,


the maximum strength, the local and global ductility capacities and, last but not
the least, their hysteretic behavior [Taucer et al., 1991].
Earthquake design of structures according to the capacity design philosophy
normally leads to dominant weak beam–strong column mechanisms, but it is also
worth reminding that several existing buildings, designed prior to 1980’s code stan-
dards (most of them already comprising seismic resistant provisions, as in the case of
Portuguese codes enforced after 1983), exhibit the opposite behavior where plastic
hinges develop in columns rather than in the beams, thus corresponding to strong
beam–weak column systems.
Indeed, recent earthquakes, have shown that some of the collapses or severe
J. Earthquake and Tsunami Downloaded from www.worldscientific.com

damage to buildings were due to columns’ inadequate response or failure, due


to insufficient shear strength of the columns, inadequate concrete core confine-
ment, or insufficient capacity for the bidirectional bending demands [Saatcioglu
and Ozcebe, 1989; Ozcebe et al., 2004; Rossetto et al., 2009; Eberhard et al. 2010].
by WSPC on 01/24/13. For personal use only.

The biaxial behavior of RC columns is therefore essential to be considered in the


seismic response analysis of buildings, even when composed by regular structural
systems.
Recent seismic codes and guidelines, e.g. EC8 [CEN, 2003], consider the three-
dimensional non-linear dynamic analysis as the most accurate one. However, many
difficulties are still recognized for this analysis approach and, therefore, research
must be pursued addressing the biaxial behavior of RC building columns, bearing
in mind that: (i) only sophisticated non-linear behavior models can provide accurate
representation of the real structural cyclic response of a given building and, (ii) these
non-linear hysteretic models have to be calibrated and validated with experimental
results obtained from bidirectional tests on RC columns [Qiu et al., 2002].
In this framework, the importance of experimental studies on cyclic behavior of
full-scale RC specimens, for both global structures and single members, has been
emphasized in studies by several other authors [Abrams, 1987a; Arêde, 1997; Lu
et al., 1999; Varum, 2003]. In fact, numerous experimental studies on the cyclic
uniaxial flexural response of RC columns, combined with constant axial load, have
been carried out in the last years, covering several distinct conditions concerning
the cross-section geometry and dimensions, the longitudinal reinforcement contents
and detailing, the confinement level, as well as the testing setup and load history
type [Mander et al., 1983; Atalay and Penzien, 1975; Abrams, 1987a; Low and
Moehle, 1987; Li et al., 1988; Ang et al., 1989; Bousias et al., 1992; Rodrigues,
2006; Delgado et al., 2009]. However, still the available experimental studies on
columns subjected to bidirectional loading are quite limited, for example [Priestley
et al., 1981; Park et al., 1982; Li et al., 1988; Saatcioglu and Ozcebe, 1989; Qiu
et al., 2002; Dhakal et al., 2006; Tirasit and Kawashima, 2007; Li et al., 2008].
In this context, the present study, refers to the first tests of a testing campaign
under development, aiming at contributing for reducing this gap. Therefore, test
results are herein presented and discussed in detail concerning the biaxial response

1250026-2
2nd Reading
January 25, 2013 9:21 WSPC/S1793-4311/238-JET 1250026

Behavior of RC Building Columns Under Cyclic Loading: Experimental Study

of two rectangular RC columns, first subjected to a uniaxial (1D) cyclic loading


in one direction and subsequently to another 1D cyclic loading in the orthogonal
direction.

2. Experimental Program and Testing Setup


2.1. Framework and general issues
The current experimental work is part of a large testing campaign undertaken by
the Laboratory of Earthquake and Structural Engineering (LESE), of the Faculty of
Engineering of Porto University (FEUP), for the study of RC columns (of buildings
and bridges) under horizontal cyclic loadings [Delgado et al., 2005, 2006]. The main
J. Earthquake and Tsunami Downloaded from www.worldscientific.com

objective of this experimental work is to study the cyclic behavior of RC building


columns, under uniaxial and biaxial horizontal cyclic loadings. The first biaxial
test series, reported in this paper, deals with the response of the columns in one
by WSPC on 01/24/13. For personal use only.

direction, after having been previously damaged in the other principal direction. In
the subsequent experimental studies, other biaxial load paths are to be considered
in accordance with load paths adopted by other authors [Bousias et al., 1992; Qiu
et al., 2002], namely quadrangular, rhombus and circular load paths.
In this context, and for the purpose of the study herein addressed, two simi-
lar rectangular RC columns were built and cyclically tested in two phases, with
constant axial load. As above-mentioned, each column was first tested under 1D
bending in one horizontal direction and then a second 1D cyclic test was performed
in the same column (already partially damaged) but in the other horizontal direc-
tion. Both tests were made under displacement controlled conditions.
For each test on the column specimens, the following general designation
“PB0D-N##” was adopted, where:
• D takes the value “1” for test in the column strong direction (direction N-S,
Fig. 1) and the value “2” for the weak direction test (direction W-E, Fig. 1);
• ## represents the reference number of the column specimen.

2.2. Column specimens


The column specimens are 1.70 m high and were cast on a strong square RC foun-
dation block with 1.30 m2 × 1.30 m2 in plan dimensions and 0.50 m high. Four holes
are arranged at the foundation block in order to fix the specimen with prestressed
steel rods to the laboratory strong floor.
The longitudinal reinforcement of the columns is composed by six 12 mm diam-
eter bars and the transversal reinforcement is made of 6 mm diameter stirrups,
spaced at 150 mm; the columns were built with continuous longitudinal reinforce-
ment bars, i.e. without lap splices. For the foundation block 16 mm diameter bars
were used. Figure 2 shows the reinforcement detailing of the column and founda-
tion. The specimen construction was rigorously supervised in order to achieve the

1250026-3
2nd Reading
January 25, 2013 9:21 WSPC/S1793-4311/238-JET 1250026

H. Rodrigues et al.
J. Earthquake and Tsunami Downloaded from www.worldscientific.com
by WSPC on 01/24/13. For personal use only.

Fig. 1. General scheme of the column specimens and testing directions.

design specifications in terms of materials, structural geometry and reinforcement


detailing. The RC specimens were built on a construction site and both traditional
wood formwork and workmanship were adopted in order to convey the common
practice construction as close as possible. The concrete was provided from a ready-
mixed concrete plant and vibrated in the casting phase with a usual vibrating
device. The column casting was made in a single operation comprising first the
footing and then the column.

2.3. Materials properties


The materials considered at the specimen design phase consisted on regular con-
crete, class C35/45 (NP-EN 206-1 [2000]), and reinforcing steel of class A400NR-
SD. Tests on samples of the materials used in the construction (steel reinforcement
and concrete) have been carried out and the results obtained are presented next.
For the reinforcement steel, three samples of 12 mm diameter bars (longitudinal
steel) were tested according to the NP EN 10002-1 [2006]. The obtained results
are summarized in Table 1, in terms of Young modulus, yielding strength, ultimate
strength and ultimate strain.
As already stated, the concrete considered at the design stage was a C35/45,
with a slump test result between 100 and 150 mm, corresponding to a slump class S3,
and aggregate maximum dimension of D22 (in mm), according to the Portuguese
Standard NP-EN0206-1 [2006]. During the casting phase 150 mm cubic concrete
samples were produced and Table 2 summarizes the results of the corresponding
compression tests.

1250026-4
2nd Reading
January 25, 2013 9:21 WSPC/S1793-4311/238-JET 1250026

Behavior of RC Building Columns Under Cyclic Loading: Experimental Study


J. Earthquake and Tsunami Downloaded from www.worldscientific.com
by WSPC on 01/24/13. For personal use only.

Fig. 2. RC column specimen dimensions and reinforcement detailing.

Table 1. Results from the tensile tests on steel samples.

Young modulus Yielding strength Ultimate strength Ultimate strain


Sample E (GPa) fsy (MPa) fsu (MPa) εsu (%)
1 197.75 435.77 543.80 21.27
2 195.00 435.20 547.50 19.42
3 191.23 426.93 542.68 20.28
Average 194.66 432.63 544.66 20.32

Table 2. Results from compressive tests on concrete specimens.

Compressive cubic Young’s


ultimate strength (MPa) Modulus (GPa)
Specimen Testing date Sample 1 Sample 2 Sample 3 Average Average
1st 28 days after casting 44.89 51.08 49.08 48.35 39.4
2nd At the testing time 59.15 55.55 57.90 57.53 44.3

1250026-5
2nd Reading
January 25, 2013 9:21 WSPC/S1793-4311/238-JET 1250026

H. Rodrigues et al.

2.4. Testing setup and instrumentation


Figure 3 shows the setup adopted for the experimental testing campaign on
RC columns. The system includes two horizontal actuators to apply the lateral
loads, one with 500 kN with +/−150 mm stroke and the other with 200 kN with
+/−100 mm stroke, and a 700 kN actuator to apply the vertical axial load, for
J. Earthquake and Tsunami Downloaded from www.worldscientific.com
by WSPC on 01/24/13. For personal use only.

(a)

(b)
Fig. 3. Testing setup at LESE laboratory: (a) schematic layout, and (b) general view.

1250026-6
2nd Reading
January 25, 2013 9:21 WSPC/S1793-4311/238-JET 1250026

Behavior of RC Building Columns Under Cyclic Loading: Experimental Study

which two steel reaction frames (lateral and vertical) and one lateral reaction wall
are used. Both the specimens and the reactions frames were fixed to the laboratory
strong floor with prestressed steel bars in order to avoid sliding and overturning
during the tests.
For the column tests presented in this paper, a constant axial load was applied
while the lateral loading was cycled under displacement controlled conditions. Since
the axial load actuator remains in the same position while the column specimen
laterally deflects during the test, a special device consisting of two sliding steel
plates exists between the top column section and the actuator, in order to minimize
spurious friction force effects. However, in order to measure these small friction
forces load cells in the two horizontal directions are connected to the upper plate
J. Earthquake and Tsunami Downloaded from www.worldscientific.com

(that is expected to displace laterally) and the corresponding measured forces are
subtracted from the forces read by the load cells of the horizontal actuators.
The hydraulic system for the vertical actuator was designed to keep constant oil
pressure, intending to impose a constant axial force during each test. Nonetheless,
by WSPC on 01/24/13. For personal use only.

in order to control small variations that may occur, the axial force is also monitored
during the test.
The horizontal actuators’ control and the data acquisition are both performed
through PXI controller systems run by home developed control and acquisition
routines based on the Lab VIEW software platform (www.ni.com). Data acquisition
and signal conditioning cards provide direct readings from load cells, Linear Variable
Differential Transformers (LVDTs) and other types of amplified analogical or digital
sensors [Delgado et al., 2009].
Since the global deformation of these slender column specimens are typically due
to predominant flexural deformations, the required instrumentation scheme for the
tests had to take this fact into account. Figure 4 shows the adopted instrumentation
layout which allows measuring the lateral displacement at several height levels, plus
the local relative displacement in several points strategically selected in order to
calculate the contribution of each deformation component (bending, shear and slip)
to the total column deflection, as described in the next sections.

2.5. Cyclic loading — imposed horizontal displacements


at the top-column
For all the tested specimens, a constant axial load of 170 kN was imposed, which
corresponds to a normalized axial force of 0.04. The normalized axial force level
considered in the tests corresponds to typical values for current reinforced concrete
buildings with three to four storeys, particularly in corner columns where the biaxial
response is likely to be more conditioning. In order to characterize the response of
column specimens, cyclic lateral displacements were imposed at the top of the
column with smooth increasing demand levels. Three cycles were repeated for each
lateral deformation demand level. This procedure allows for easy understanding
of the columns’ behavior, the comparison between tests and the development and

1250026-7
2nd Reading
January 25, 2013 9:21 WSPC/S1793-4311/238-JET 1250026

H. Rodrigues et al.
J. Earthquake and Tsunami Downloaded from www.worldscientific.com
by WSPC on 01/24/13. For personal use only.

Fig. 4. Displacement measurement: Instrumentation scheme adopted in the tests.

calibration of numerical models. The repetition of cycles for each displacement


demand allows obtaining information to better understand the stiffness and strength
degradation of the column, which is relevant also for the calibration of numerical
models. The adopted load paths are summarized in Table 3.

Table 3. Peak value for each series of cyclic imposed horizontal displacements (mm) (in brackets
drift in %).

Column N01 Column N02


N-S E-W E-S N-S
Column testing direction test label PB01-N01 PB02-N01 PB02-N02 PB01-N02
3 (0.2%) 3 (0.2%) 1 3 (0.2%)
5 5 3 (0.2%) 5
10 10 5 10
4 4 10 4
12 12 4 12
15 (1%) 15 (1%) 12 15 (1%)
7 7 15 (1%) 7
20 20 7 20
Lateral drift history 30 (2%) 25 20 25
40 30 (2%) 25 30 (2%)
50 (3.3%) 35 30 (2%) 35
40 40
45 (3.0%) 45 (3.0%)
50 50
55 55
60 (4.0%) 60 (4.0%)
65 65
70 (4.7%) 70 (4.7%)

1250026-8
2nd Reading
January 25, 2013 9:21 WSPC/S1793-4311/238-JET 1250026

Behavior of RC Building Columns Under Cyclic Loading: Experimental Study

For each of the non-damaged rectangular columns, the first cyclic tests were
performed in one of their principal directions. The column N01 was first tested in
its strongest direction and the test was stopped at 3.3% imposed drift. By contrast,
column N02 was initially tested in the weakest direction and it went up to 2%
drift. The difference of maximum imposed drifts for each tested column is related
to their different mechanical properties in each direction, taking into account that
the adopted stop testing criterion should ensure a desired damage level, though not
very excessive in order allow performing a second test on each column along the
orthogonal direction.
J. Earthquake and Tsunami Downloaded from www.worldscientific.com

3. Expressions and Procedures Adopted


in the Analysis of Test Results
3.1. General considerations
by WSPC on 01/24/13. For personal use only.

This section presents analytical equations and assumptions adopted in the analysis
of the experimental results, namely concerning: (i) the procedure for the column
lateral deformation splitting in terms of bending, slip and shear components, (ii) the
numerical models and assumptions for evaluating the global flexural properties of
the column cross-section and, (iii) the simplified formulation used to estimate the
lateral force-displacement response of the columns.
Each column was instrumented with transducers to measure two types of relative
displacements: (i) total horizontal displacement of different column in-height section
levels and, (ii) local deformation at the column base where large deformations
are expected (local deformations). For each test, the recorded total deformation
is analyzed, by splitting it in the bending, slip and shear components, based on a
simplified formulation presented in the following sub-sections.
In order to identify the characteristic points of the RC column non-linear
response, the monotonic moment-curvature diagrams of the column cross-section
were calculated for each horizontal loading direction resorting to fiber modeling
[McKenna, 1997]. The so-obtained non-linear moment-curvature plot allowed iden-
tifying the cracking, yielding, as well as the nominal and ultimate strength charac-
teristic points for the RC cross-section flexural behavior.
With the flexural characteristic points at column cross-section level, the global
force-displacement response of the RC column is predicted, for each horizontal load-
ing direction, using the simplified methodology proposed by Lehman [Lehman et al.,
1998]. The results obtained with these empirical equations are compared with the
experimental results (Sec. 4).

3.2. Deformation components


The total lateral deformation of a RC column subjected to lateral loads is the sum
of contributions due to bending, shear and slip of the longitudinal bars at column

1250026-9
2nd Reading
January 25, 2013 9:21 WSPC/S1793-4311/238-JET 1250026

H. Rodrigues et al.

ends can be calculated according to Eq. (1).


∆total = ∆bending + ∆shear + ∆slip . (1)

3.2.1. Bending deformation


Based on the relative deformation recorded in different points with the instru-
mentation layout scheme adopted in the tests, the column bending deformation is
estimated with the vertical and horizontal transducers, located at different column
heights [Mirand-Guedes, 1997].

3.2.2. Slip deformation


J. Earthquake and Tsunami Downloaded from www.worldscientific.com

The lateral displacement component associated to the slip was computed based on
the deformations recorded by the vertical transducers located at the column-base,
which are mounted 10 cm above the foundation block, (transducers 35 and 36 in
by WSPC on 01/24/13. For personal use only.

Fig. 4). These deformations led to a concentrated rotatio θslip at the base essentially
due to reinforcement slippage. However, should these deformations (measured by
the transducers at this column short slice) include any contribution from flexural
deformation, still the response analysis for this slice assumes that the recorded
deformation mainly results from the strain penetration of the longitudinal steel
rebars into the foundation block. Therefore, the rotation of the top slice of that
section is considered a measure of θslip and can be calculated from the deformations
read at the referred transducers; the top column lateral displacement component
due to slip is then obtained according to Eq. (2) [Sezen, 2008], where Ltot is the
total column height.
∆slip = θslip ∗ Ltot . (2)

3.2.3. Shear deformation


Typically, the column deformations due to shear are relatively small before the
development of large inclined cracks. For the columns herein analyzed, it is expected
a reduced shear deformation contribution to the total lateral deformation, because
their slenderness leads to relatively small shear deformation when compared to the
flexural deformation component [Sezen, 2008] and, in fact, from the observed exper-
imental test results of the present study, shear cracks did not occur. Therefore, for
simplicity in the column lateral deformation calculations, the shear deformation
was computed assuming the uncracked elastic shear stiffness and considering uni-
form shear stress distribution over the column cross-section. Accordingly, Eq. (3)
was adopted
V.Ltot
∆shear = , (3)
G.Ag
where V is the shear force, Ltot was above defined, G stands for the elastic shear
modulus that can be calculated by the classical expression, G = E/(2(1 + ν)), in

1250026-10
2nd Reading
January 25, 2013 9:21 WSPC/S1793-4311/238-JET 1250026

Behavior of RC Building Columns Under Cyclic Loading: Experimental Study

which E is the Young modulus and ν is the Poisson’s ratio, and Ag is the cross-
section effective shear area that, for rectangular sections, can be assumed to be 5/6
of the total cross-section area.

3.2.4. Flexural behavior properties of the column cross-sections


The moment-curvature curve and the main flexural properties of the RC column
cross-section were calculated, in both loading directions, resorting to a procedure
available in the OpenSeeS computer code [McKenna, 1997], by considering the
section geometry, the steel reinforcement distribution and the material properties.
The constant axial force foreseen in the tests was also included in the calculations
as an input parameter.
J. Earthquake and Tsunami Downloaded from www.worldscientific.com

For each studied cross-section a very refined mesh was adopted, both for con-
fined and unconfined concrete zones, and each reinforcement steel bar was con-
sidered as a single fiber element. In the numerical simulations of the cross-section
by WSPC on 01/24/13. For personal use only.

mechanical behavior the following material models were considered: (i) Concrete02
[McKenna, 1997] for the concrete fibers (see Fig. 5(a)), taking into account the
stirrups’ confinement effect and the behavior law proposed by Hognestad [Hognes-
tad, 1951]; (ii) Steel02 [McKenna, 1997] for the steel fibers (see Fig. 5(b)) based
on the Giuffr–Pinto formulation as implemented by Menegotto [CEB, 1996]. In
the Concrete02 model, fcm and fcum represent, respectively, the maximum and
residual compressive strengths, fctm is the tensile strength, ε0 is the strain at max-
imum compressive strength and εu stands for the transition between the softening
and residual branches of the concrete model. As for the Steel02 model, fy repre-
sents the yielding strength, E is the initial elastic modulus (Young modulus) and
Ep is the post-yielding tangent elasticity modulus introduced in the model by the
Stress

fct

R0
Strain
Stress

(fcum, εu)

(fcm, ε0)
2 x fcm /ε0 Strain
(a) (b)
Fig. 5. Material models [McKenna, 1997]: (a) for concrete model and (b) for steel.

1250026-11
2nd Reading
January 25, 2013 9:21 WSPC/S1793-4311/238-JET 1250026

H. Rodrigues et al.

strain-hardening ratio b = E/Ep. Concerning the concrete, different stress–strain


relationships are used to differentiate between the confined concrete core of the
sections and the unconfined concrete cover. In the numerical analysis, 3 cm for the
concrete cover was considered, as employed in the construction of the specimens.
Therefore, the longitudinal steel reinforcement in beams and columns is centered at
about 3.6 cm from the external surface of the RC elements. Table 4 summarizes the
average relevant mechanical properties’ values of the concrete used in the numeri-
cal model (obtained from the concrete tests presented in Table 2). The mechanical
properties (mean values) for the steel reinforcement adopted in the model for the
numerical analyzes are shown in Table 5 (based on the experimental results, see
Table 1).
J. Earthquake and Tsunami Downloaded from www.worldscientific.com

Figure 6 shows the numerical results in terms of non-linear moment-curvature


diagrams of column sections, as well as the identification of the characteristic points
for cracking, yielding and nominal strength, for both loading directions. The crack-
ing point refers to the first concrete fiber reaching the maximum tensile strength
by WSPC on 01/24/13. For personal use only.

while the yielding point corresponds to the first yielding of any reinforcement steel
fiber; the nominal strength is determined when the maximum compressive fiber
strain reaches 0.3%. Table 6 includes the summary of the characteristic points, in
terms of curvature and moment, obtained for both horizontal loading directions.

3.3. Column lateral force-displacement response


This section includes a brief summary of the method adopted for calculating the
global force-displacement response of columns, as proposed by Lehman et al. [1998],
developed and calibrated for circular RC columns. The method is the applied to the
study of the RC column herein addressed in both orthogonal horizontal directions

Table 4. Concrete: Parameters to the numerical model.


Mechanical parameter Value
Maximum compressive strength, fcm confined 45.5 MPa
unconfined 46.9 MPa
Strain at maximum compressive strength, ε0 confined 0.18%
unconfined 0.22%
Strain at transition between the softening and residual branche, εu confined 0.16%
unconfined 0.35%
Residual compressive strength, fcum confined 9.20 MPa
unconfined 0.00 MPa
Tensile ultimate strength, ftm 4.60 MPa

Table 5. Steel: Parameters to the numerical model.


Mechanical parameters Value
Yield stress, fsy 432.6 MPa
Elastic Young’s modulus, Es 195 MPa
Hardening strain, b 3.17%

1250026-12
2nd Reading
January 25, 2013 9:21 WSPC/S1793-4311/238-JET 1250026

Behavior of RC Building Columns Under Cyclic Loading: Experimental Study


J. Earthquake and Tsunami Downloaded from www.worldscientific.com

(a)
by WSPC on 01/24/13. For personal use only.

(b)
Fig. 6. Calculated non-linear moment-curvature monotonic response for the column cross-section
and characteristic points for cracking, yielding and nominal flexural strength: (a) strong direction;
(b) weak direction.

Table 6. Characteristic points of the RC cross-section moment-curvature curve.

Cracking Yielding Nominal strength


Curvature Moment Curvaturee Moment Curvature Moment
Loading φcr Mcr φy My φn Mn
direction (/m) (kN.m) (/m) (kN.m) (/m) (kN.mm)
Strong 4.20 × 10−4 19.78 9.36 × 10−3 82.70 5.63 × 10−2 103.66
Weak 8.40 × 10−4 9.87 2.10 × 10−2 43.67 11.48 × 10−2 49.63

1250026-13
2nd Reading
January 25, 2013 9:21 WSPC/S1793-4311/238-JET 1250026

H. Rodrigues et al.

and the results are described; the moment-curvature curves obtained in the previous
section were adopted for the following column response calculations.

3.3.1. Cracking displacement


The column response up to cracking is assumed to be linear elastic corresponding
to a cantilever column behavior loaded by a concentrated lateral force at the top.
Based on that assumption, the cracking displacements obtained for each horizontal
direction are presented in Table 7.

3.3.2. Yielding displacement


J. Earthquake and Tsunami Downloaded from www.worldscientific.com

The yielding of an axially loaded reinforced concrete element is a progressive pro-


cess. For columns, the definition of the yielding point is often assumed at the first
yielding of any longitudinal reinforcement bar, but this assumption generally under-
by WSPC on 01/24/13. For personal use only.

estimates the global column stiffness reduction associated with the yielding process.
Lehman proposes the calculation of the effective yielding displacement as the sum
of three components, namely those associated to flexure (∆f y ), slip (∆sy ) and shear
(∆vy ). According to this methodology, the so-obtained theoretical yielding displace-
ments, for each horizontal loading direction of the tested column type, are listed in
Table 7.

3.3.3. Ultimate displacement


The ultimate displacement calculated with the Lehman method is based on the
plastic hinge length, the yielding displacement as well as the yielding and ultimate
curvatures. The later were estimated from the moment-curvature curve, for each
analyzed direction of the rectangular RC column, considering the tensile limit strain
of the longitudinal steel at 7.5%, as proposed in Eurocode 2, Annex C [EC2, 2004].
The plastic hinge length adopted for the calculations were 0.2 m for the column
loaded in its strong direction, and 0.1 m for its weak direction, corresponding to
0.5 h according to Paulay and Priestley [1992]. Furthermore, at the end of the test
PB01-N01, the plastic hinge length was measured, which extended up to a value
of 18 cm, approximately 0.5 h (20 cm). Table 7 includes the so-obtained ultimate
column displacements, for each horizontal loading direction of the column.

Table 7. Global column response estimative based on the Lehman method.

Cracking Yielding Ultimate


Force Displ. Force Displ. Force Displ.
Fcr ∆cr Fy ∆y Fu ∆u
Column direction (kN) (mm) (kN) (mm) (kN) (mm)
Strong 13.19 0.32 55.13 13.49 83.38 77.03
Weak 6.58 0.63 29.11 27.165 40.83 101.40

1250026-14
2nd Reading
January 25, 2013 9:21 WSPC/S1793-4311/238-JET 1250026

Behavior of RC Building Columns Under Cyclic Loading: Experimental Study


J. Earthquake and Tsunami Downloaded from www.worldscientific.com
by WSPC on 01/24/13. For personal use only.

Fig. 7. Calculated theoretical tri-linear approximation of global response diagrams for the strong
and weak directions of the column.

3.3.4. Theoretical global force-displacement monotonic response diagrams


Based on the values listed in Table 7 for the RC column under analysis, the char-
acteristic points for each horizontal loading direction in terms of cracking, yielding
and ultimate strength (force and displacement) obtained according to the method-
ology proposed by Lehman, allow plotting the corresponding theoretical trilinear
approximation of global response diagrams as depicted in Fig. 7.

4. Column Test Results and Comparative Studies


4.1. Introductory comments
The most relevant results obtained from the four cyclic tests on columns (PB01-
N01, PB02-N01, PB02-N02, and PB01-N02) are reported hereafter, first for each
test independently concerning: (i) the observed damage evolution (cracking opening,
concrete spalling and reinforcement steel buckling), (ii) the final damage pattern
and, (iii) the columns’ response in terms of shear vs. top-displacement. The observed
column response is compared with the results obtained as described in Sec. 3, both
in terms of moment-curvature diagrams and deformation components. The observed
response is represented by points in the shear-drift curve at the instants when
the corresponding damage was visually observed during the test. The maximum
strength corresponds to the maximum shear force applied to the column speci-
men. The ultimate strength corresponds to a 20% strength reduction, relatively

1250026-15
2nd Reading
January 25, 2013 9:21 WSPC/S1793-4311/238-JET 1250026

H. Rodrigues et al.

to the maximum strength point, according to the procedure adopted by [Park


et al., 1987].
Subsequently, one sub-section is dedicated to the comparisons between the
results observed for each test. In the last sub-section, the visually observed damage
evolutions are compared with those predicted by an analytical method for damage
classification.
As above-mentioned, in all the tests, an initial vertical axial load of 170 kN was
applied, which was intended to be kept constant during the tests. However, the axial
load value monitoring in the four tests exhibited small variations. The maximum
axial load variation along each test was 3.5% for PB01-N01, 2.2% for PB02-N01,
1.5% for PB02-N02 and 4.0% for PB01-N02. These variations are negligible for the
J. Earthquake and Tsunami Downloaded from www.worldscientific.com

comparative purposes of this work.

4.2. Test PB01-N01


by WSPC on 01/24/13. For personal use only.

4.2.1. Damage evolution


For the PB01-N01 test, the horizontal displacements were cyclically applied in the
strong direction of the column. The initial horizontal cracks were observed for the
5 mm displacement amplitude cycles in both, north and south faces, namely one
at the column base section and two others at 10–15 cm and 30 cm high. At 10 mm
displacement cycles, the initial cracks (started in previous displacement demands)
developed further, two new cracks appeared and the column base crack became
wider. All the cracks were of bending type (horizontal). Concrete spalling started
during the 30 mm cycles, beginning at the column corners and extending for the
40 mm lateral displacement demands in both faces (north and south). For the 50 mm
displacement cycles, the spalled region developed more up to an extension of about
15 to 20 cm. At the end of the test, the corner reinforcing steel bars were fully
visible and buckling was evident (see Fig. 8).
In summary, cracking started at 0.33% drift, concrete spalling initiated at 2%
and bar buckling occurred at 3.33% drift. Figures 8 and 9 show the final damaged
state at the column base for each of the nominal peak displacements which were
measured with the internal transducer of the actuator. The really achieved net
displacements were measured by other LVDTs attached to an external reference
frame and the corresponding peak values at the actuator level are also included in
brackets in the same figure.

4.2.2. Test results


Figure 10 represents the global hysteretic cyclic column response diagram in terms
of base-shear vs. drift at the column top section. The diagram also includes the
cracking, yielding and nominal strength thresholds estimated with the fiber model
as described in Sec. 3, for the strong column direction, as well as the column
shear-drift response predicted with the Lehman method. Table 8 provides values

1250026-16
2nd Reading
January 25, 2013 9:21 WSPC/S1793-4311/238-JET 1250026

Behavior of RC Building Columns Under Cyclic Loading: Experimental Study

Fig. 8. PB01-N01 test: Damage pattern at the 3rd cycle of each nominal peak displacement level
J. Earthquake and Tsunami Downloaded from www.worldscientific.com

(net displacements in brackets).


by WSPC on 01/24/13. For personal use only.

Fig. 9. PB01-N01 test: Final state damage.

corresponding to the cracking, yielding and maximum strength characteristic points


that were obtained from detailed observation of the actual global response plot
depicted in Fig. 10.
From the analysis of the results presented in Fig. 10 and Table 8, the following
conclusions can be drawn:
• As already stated from the visual damage observation (described in the previous
sub-section) cracking started within the nominal 5 mm top displacement cycle, for
observed values of 1.3 and 2.16 mm. By contrast, the Lehman method application
provides a cracking displacement estimate of about 0.32 mm, which is rather
lower that the measured values. This observation might be due to the fact that
the theoretical prediction of initial cracking is directly related with the strict

1250026-17
2nd Reading
January 25, 2013 9:21 WSPC/S1793-4311/238-JET 1250026

H. Rodrigues et al.
J. Earthquake and Tsunami Downloaded from www.worldscientific.com
by WSPC on 01/24/13. For personal use only.

Fig. 10. PB01-N01 test: Obtained shear-drift diagram, with predicted characteristic force thresh-
olds and theoretical Lehman method diagram proposal.

Table 8. Characteristic points of the observed shear-drift response for the PB01-N01 test.

Cracking Yielding
Force Displ. Drift Force Displ. Drift
(kN) (mm) (%) (kN) (mm) (%)
PB01-N01 + 38.00 2.16 0.15 59.60 7.75 0.52
− 36.65 1.30 0.09 59.43 7.60 0.51
Maximum strength Ultimate strength
Force Displ. Drift Force Displ. Drift
(kN) (mm) (%) (kN) (mm) (%)
PB01-N01 + 73.04 38.26 2.55 — — —
− 68.40 38.51 2.57 — — —

condition of reaching the tensile strength at the extreme fiber of the base section
and, in fact, this can start much before than the external cracks become visible.
Moreover, theoretical cracking point estimates are very sensitive to the tensile
strength of concrete which is known to exhibit differences between test samples
and built specimens; in addition, this parameter was not effectively measured,
thus relying upon an estimate based on the compressive strength.
• The estimated yielding strength is close to the experimentally observed value,
but the corresponding yielding displacement calculated with the Lehman method
does not match well with the recorded one. From direct analysis of the observed
experimental shear-drift curve, and by best-fitting, the yielding displacement
is proposed a value of 7.75 mm (corresponding to 0.52% drift), whilst with the
Lehman method the obtained yielding displacement is 13.49 mm. Thus, the result

1250026-18
2nd Reading
January 25, 2013 9:21 WSPC/S1793-4311/238-JET 1250026

Behavior of RC Building Columns Under Cyclic Loading: Experimental Study

provided by the Lehman proposal largely overestimates the yielding displacement,


due to the relevant contribution of the slip deformation component (flexural:
8.8 mm, shear: 0.1 mm and slip: 4.61 mm).
• As stated before, the concrete spalling initiated at 2% drift and bar buckling at
3.33% drift. As can be observed in Fig. 10, for 2% drift deformation levels, corre-
sponding to concrete spalling, the strength degradation does not occur. However,
for the last displacement cycles (drift larger than 2.5%), when reinforcement
bar buckling started, a global lateral strength reduction is evident which further
amplifies with increasing lateral deformation demands.

Figure 11 illustrates the contribution of the different deformation components for


J. Earthquake and Tsunami Downloaded from www.worldscientific.com

the total lateral deformation observed for the test PB01-N01, obtained with the
procedure described in Sec. 3.2. According to Fig. 11, the flexural relative defor-
mation is dominant during the whole cyclic test, representing 70% to 80% of the
total lateral column deformation. As expected, the shear deformation contribution
by WSPC on 01/24/13. For personal use only.

for the total value is insignificant. The slip deformation component increases during
the test, but, due to the formation of large horizontal cracks in bending, the relative
weight of this deformation component becomes significantly reduced for increasing
drift.
Figure 12 shows the moment-curvature evolution at the column base section
obtained from the cyclic experimental results and, simultaneously, with the fiber
based numerical model presented in Sec. 3 (monotonic loading). The basis length
adopted to compute the curvature corresponds to the span of the first level of
LVDTs (as represented in Fig. 4), that is 20 cm. A good agreement is found between

Fig. 11. PB01-N01 test: Bending, slip and shear deformation components contribution for the
top displacement.

1250026-19
2nd Reading
January 25, 2013 9:21 WSPC/S1793-4311/238-JET 1250026

H. Rodrigues et al.
J. Earthquake and Tsunami Downloaded from www.worldscientific.com
by WSPC on 01/24/13. For personal use only.

Fig. 12. PB01-N01 test: Moment-curvature response (experimental results and numerical
estimation).

curvatures estimated from experimental results and those obtained from the numer-
ical models. After the yielding moment, the numerical result shows an increasing
strength branch, while the experimental results evidence strength reduction. This
difference derives from two issues, namely the strength reduction associated to the
cyclic loading during tests (which is not present in the numerical results because
they refer to the monotonic response) and the inability of the used models to con-
sider the reinforcement steel bar buckling phenomena observed in the test.

4.3. Test PB02-N01


4.3.1. Damage evolution
In the PB02-N01 test, the horizontal cyclic displacements were applied in the weak
direction of column N01 after the PB01-N01 test. As expected, it was verified that
no new cracks developed in the element during this second test. The spalled regions
that were observed in the previous test correspond only to the north and south
faces of the column. In this second test spalling developed further towards the east
and west faces after the 55 mm amplitude displacement cycle. The height of the
spalled region observed in this test was the same as the observed in the previous
test, 15 to 20 cm. At the 65 mm displacement cycle all concrete cover in the critical
region was completely fallen off and the central reinforcing steel bar started buckling
(see Fig. 13). The final damaged state of column N01 after the PB02-N01 test is
shown in Fig. 14. Column failure is reached, assuming as failure criteria a strength
reduction of 20% relatively to the maximum strength [Park et al., 1987].

1250026-20
2nd Reading
January 25, 2013 9:21 WSPC/S1793-4311/238-JET 1250026

Behavior of RC Building Columns Under Cyclic Loading: Experimental Study


J. Earthquake and Tsunami Downloaded from www.worldscientific.com

Fig. 13. PB02-N01 test: Damage pattern at the 3rd cycle of each nominal peak displacement
level (net displacements in brackets).
by WSPC on 01/24/13. For personal use only.

Fig. 14. PB02-N01 test: Final damage.

4.3.2. Test results


Figure 15 shows the global hysteretic response observed in the cyclic test, in terms
of shear vs. drift. This figure includes also the plot of the cracking, yielding and
nominal strength limits (as for the previous cases) obtained for the column weak
direction with the fiber model described before. Since this test was performed on
the column already damaged (during previous P01-N01 test), the Lehman method
was not applied to estimate the column response in this test. Table 9 lists the
maximum and ultimate shear force and the correspondent displacements obtained

1250026-21
2nd Reading
January 25, 2013 9:21 WSPC/S1793-4311/238-JET 1250026

H. Rodrigues et al.
J. Earthquake and Tsunami Downloaded from www.worldscientific.com
by WSPC on 01/24/13. For personal use only.

Fig. 15. PB02-N01 test: Obtained shear-drift diagram, with predicted characteristic force
limits.

Table 9. Characteristic points of the observed shear-drift response for the PB02-N01 test.

Cracking Yielding
Force Displ. Drift Force Displ. Drift
(kN) (mm) (%) (kN) (mm) (%)
PB02-N01 + — — — — — —
− 23.70 1.82 0.12 33.11 17.82 1.19
Maximum strength Ultimate strength
Force Displ. Drift Force Displ. Drift
(kN) (mm) (%) (kN) (mm) (%)
PB02-N01 + 26.20 47.26 3.15 20.94 57.20 3.81
− 35.26 28.09 1.87 — — —

by observation of the shear vs. drift test response diagram. Cracking and yielding
had already occurred in the previous test and, therefore, they are not indicated in
the table.
From the analysis of the results expressed in Fig. 15 and Table 9 the following
can be concluded:

• As visually observed during the test, concrete spalling occurred at the 55 mm


displacement cycle (corresponding to 3.66% drift), which induced strength reduc-
tion. For the subsequent cycles, the observed concrete spalling associated with
buckling of the central longitudinal rebars justifies the strength reduction evolu-
tion until the end of the test.

1250026-22
2nd Reading
January 25, 2013 9:21 WSPC/S1793-4311/238-JET 1250026

Behavior of RC Building Columns Under Cyclic Loading: Experimental Study

• The maximum and ultimate strength values, both the calculated (see Table 7)
and those obtained in the tests (see Table 9), are in good agreement, particularly
for the direction corresponding to positive displacements.
• The non-symmetric response of the column is justified, on the one hand by the
non-perfectly symmetric distribution of the longitudinal reinforcement in the
column cross-section, as well as the concrete cover thickness (particularly for
the column test on its weak direction, 20 cm height). On the other hand, the first
incursion of each cycle occurs always in the same sense (pulling the column from
East towards West), therefore inducing also a slight non-symmetric response.

Figure 16 illustrates the contribution of the different deformation components


J. Earthquake and Tsunami Downloaded from www.worldscientific.com

to the total lateral deformation observed for the test PB02-N01, obtained as for the
previous case. According to Fig. 16 the flexural and slip relative deformations are
dominant during the test, each component representing approximately 50% of the
total lateral column deformation. As expected, the shear deformation contribution
by WSPC on 01/24/13. For personal use only.

for the total value is insignificant. The slip component is important, which is in
accordance with the damage induced in the reinforcement steel bars during the
previous test.
Figure 17 shows the moment-curvature evolution at the column base obtained
from the cyclic test results and, simultaneously, the monotonic one obtained with
the fiber model. The basis length adopted to compute the curvature corresponds
to the span of the first level of LVDTs (as represented in Fig. 4), which is 20 cm.
It is clear that, the damage induced by the previous test leads to a rather poor
correspondence between the experimental response and the theoretical prediction.

Fig. 16. PB02-N01 test: Bending, slip and shear deformation components contribution for the
top displacement.

1250026-23
2nd Reading
January 25, 2013 9:21 WSPC/S1793-4311/238-JET 1250026

H. Rodrigues et al.
J. Earthquake and Tsunami Downloaded from www.worldscientific.com
by WSPC on 01/24/13. For personal use only.

Fig. 17. PB02-N01 test: Moment-curvature response (experimental results and numerical
estimation).

However, the peak strength prediction still agrees reasonably well with the experi-
mental one, though slightly overestimated as above referred.

4.4. Test PB02-N02


4.4.1. Damage evolution
For the PB02-N02 test, horizontal displacement cycles were first applied in the weak
column direction. The initial horizontal cracking was observed during the 10 mm
nominal displacement cycle (0.66% drift) in all the four faces. Cracks were observed
at the column base and at 10, 20 and 30 cm of the base. For the 15 mm displacement
cycles (1% drift) one additional crack appeared but all the cracks were horizontally
oriented. The test was run until the 30 mm imposed horizontal displacement and
no concrete spalling was observed (see Fig. 18). The test was stopped at this point
in order to allow performing afterwards the cyclic test in the other direction with
limited column damage as illustrated in Fig. 19 corresponding to the end of the
test.

4.4.2. Test results


Figure 20 shows the global hysteretic response of the column, obtained from the test
in terms of shear vs. drift. As for the previous cases, the plot includes the calculated
cracking, yielding and nominal strength thresholds for the column weak direction, as
well as the global force-displacement response predicted with the Lehman method.
Accordingly, Table 10 summarizes the cracking, yielding and maximum force point

1250026-24
2nd Reading
January 25, 2013 9:21 WSPC/S1793-4311/238-JET 1250026

Behavior of RC Building Columns Under Cyclic Loading: Experimental Study


J. Earthquake and Tsunami Downloaded from www.worldscientific.com

Fig. 18. PB02-N02 test: Damage pattern at the 3rd cycle of each nominal peak displacement
level (net displacements in brackets).
by WSPC on 01/24/13. For personal use only.

Fig. 19. PB02-N02 test: Final damage state at 30 mm nominal displacement cycle.

values (and the corresponding displacement and drift values) chosen so as to fit


adequately the observed experimental shear vs. drift curve.
The results presented in Fig. 20 and Table 10, allow pointing out the following
conclusions:

• According to the damage evolution commented before, cracking starts at the


10 mm nominal displacement cycle, but under force and displacement values much
larger than those obtained by the Lehman method that yields 0.63 mm for the
cracking displacement. Since this test is essentially the same as for the PB01-N01

1250026-25
2nd Reading
January 25, 2013 9:21 WSPC/S1793-4311/238-JET 1250026

H. Rodrigues et al.
J. Earthquake and Tsunami Downloaded from www.worldscientific.com
by WSPC on 01/24/13. For personal use only.

Fig. 20. PB02-N02 test: Obtained shear-drift diagram, with predicted characteristic force thresh-
olds and theoretical Lehman method diagram proposal.

Table 10. Characteristic points of the observed shear-drift response for the PB02-N02 test.

Cracking Yielding
Force Displ. Drift Force Displ. Drift
(kN) (mm) (%) (kN) (mm) (%)
PB02 N02 + 23.38 1.56 0.10 31.87 13.87 0.92
− 23.70 1.82 0.12 33.11 17.82 1.19
Maximum strength Ultimate strength
Force Displ. Drift Force Displ. Drift
(kN) (mm) (%) (kN) (mm) (%)
PB02 N02 + 34.03 29.90 1.99 — — —
− 35.26 28.09 1.87 — — —

one, though performed along the weaker column direction, the same comments
apply as included in Sec. 4.2.2.
• The estimated yielding strength is close to the experimentally observed value,
but the associated yielding displacement obtained with the Lehman method is
again not a good estimate. By best-fitting adjustment to the experimental results,
the yielding displacement reads 13.9 and 17.8 mm (or 0.92/1.19% drift), for posi-
tive and negative sense, respectively, while according to the Lehman method the
obtained yielding displacement is 27.16 mm.
• The maximum nominal strength estimate is close to the experimentally observed
value. This fact is further stressed in the comparative study included in a latter
sub-section where the observed experimental results are compared with the pre-
dicted results for all the tests.

1250026-26
2nd Reading
January 25, 2013 9:21 WSPC/S1793-4311/238-JET 1250026

Behavior of RC Building Columns Under Cyclic Loading: Experimental Study


J. Earthquake and Tsunami Downloaded from www.worldscientific.com

Fig. 21. PB02-N02 test: Bending, slip and shear deformation components contribution for the
top displacement.
by WSPC on 01/24/13. For personal use only.

Figure 21 illustrates the contribution of the different deformation components


to the total lateral deformation observed for the test PB02-N02, obtained as for
the previous test cases. As shown in Fig. 21, and similarly to the obtained before,
the flexural deformation is dominant over the whole cyclic test, amounting for 60%
of the total horizontal top displacement, again with insignificant shear deformation
contribution. The slip deformation component has an important contribution to
the total value, accounting for approximately 30% of the total drift value.
Figure 22 shows the moment-curvature evolution at the column-base obtained
from the cyclic test as well as the fiber based numerical model results (again only

Fig. 22. PB02-N02 test: Moment-curvature response (experimental results and numerical
estimation).

1250026-27
2nd Reading
January 25, 2013 9:21 WSPC/S1793-4311/238-JET 1250026

H. Rodrigues et al.

for monotonic loading). As referred before, the basis length adopted to compute
the curvature corresponds to the span of the first level of LVDTs (as represented
in Fig. 4), that is 20 cm. In general, a good agreement was observed between the
experimental results and the numerical prediction. After yielding, for each moment
value, the curvature observed in the test is larger than the corresponding numeri-
cally obtained value.

4.5. Test PB01-N021


4.5.1. Damage evolution
For the PB01-N02 test, the horizontal displacement was applied in the column
J. Earthquake and Tsunami Downloaded from www.worldscientific.com

strong direction, after the previous test (PB02-N02) on the same column along the
weak direction.
As expected, only small cracks appeared during this test, encompassed by some
by WSPC on 01/24/13. For personal use only.

increase of the existent ones. The concrete spalling started in the column corners
during the 35 mm nominal displacement cycles and the spalled region extended
mostly in the north and south column faces for the 40 mm cycles. During the 50 mm
cycles, the spalled region developed up to an extension of 15 to 20 cm high. For
this imposed nominal displacement, the corner reinforcing steel bars were visible
and buckling was observed. For the 60 mm cycle tensile failure occurred for one
longitudinal rebar, developing further during the subsequent imposed displacement
series. At the end of the nominal 70 mm cycles, all the corner longitudinal bars
were broken. In summary, for this test the achieved crack pattern was (to a certain
extent) an additional development of that already observed in the previous test.
Concrete spalling started at 2.66% drift, longitudinal reinforcing bars buckled at
3.33% drift and rebar failed at 4% drift demands (see Fig. 23). As for the previous
cases, Fig. 24 shows the final damage state observed at the end of the test.

4.5.2. Test results


The global hysteretic test response is presented in Fig. 25 in terms of shear vs.
global drift, which also includes the numerical predictions for cracking, yielding and
nominal strength points as before. Similarly, Table 11 lists the values corresponding
to the cracking, yielding and maximum strength, as obtained by best fitting to the
experimental shear vs. drift results.
From the analysis of the results presented in Fig. 25 and Table 11, the following
issues can be highlighted:

• According to the damage evolution reported before, some cracking had been
already installed during the previous test. The estimated yielding strength seems
to be close to the experimentally obtained one. The maximum displacement expe-
rienced in the previous test did not damage significantly the column.

1250026-28
2nd Reading
January 25, 2013 9:21 WSPC/S1793-4311/238-JET 1250026

Behavior of RC Building Columns Under Cyclic Loading: Experimental Study


J. Earthquake and Tsunami Downloaded from www.worldscientific.com
by WSPC on 01/24/13. For personal use only.

Fig. 23. PB01-N02 test: Damage pattern at the 3rd cycle of each nominal peak displacement
level (net displacements in brackets).

• From the damage visual observation, concrete spalling started at 2.2% drift
and the bar buckling began at 3.33% drift. As exhibited in Fig. 25, the con-
crete spalling is associated with a significant strength reduction for deformation
demand levels of 3.33%. In the subsequent cycles, steel rebars further buckled and
some of them failed in tension, inducing a more pronounced strength reduction.

Figure 26 illustrates the contributions of the different deformation components to


the total column lateral deformation observed for the test PB01-N02, obtained as
described in Sec. 3. According to these results, the relative flexural component
deformation still persists as the dominant one during the whole cyclic test, again
representing nearly 70% of the total lateral column deformation with negligible
shear deformation contribution. The slip mechanism contribution for the column

1250026-29
2nd Reading
January 25, 2013 9:21 WSPC/S1793-4311/238-JET 1250026

H. Rodrigues et al.
J. Earthquake and Tsunami Downloaded from www.worldscientific.com
by WSPC on 01/24/13. For personal use only.

Fig. 24. PB01-N02 test: Final damage.

Fig. 25. PB01-N02 test: Obtained shear-drift diagram, with predicted characteristic force limits.

deformation is practically constant during the test, with approximately 30% of the
total deformation. As for the previous cases, Fig. 27 shows the moment-curvature
diagram at the column base obtained from the cyclic test results, including also
the numerical response curve from fiber modeling for monotonic loading. The basis
length adopted to compute the curvature corresponds to the span of the first level
of LVDTs (as represented in Fig. 4), that is 20 cm. Similar comments can be made

1250026-30
2nd Reading
January 25, 2013 9:21 WSPC/S1793-4311/238-JET 1250026

Behavior of RC Building Columns Under Cyclic Loading: Experimental Study

Table 11. Characteristic points of the observed shear-drift response for the PB01-N02 test.

Cracking Yielding
Force Displ. Drift Force Displ. Drift
(kN) (mm) (%) (kN) (mm) (%)
PB01-N02 + — — — — — —
− — — — — — 64.93
Maximum strength Ultimate strength
Force Displ. Drift Force Displ. Drift
(kN) (mm) (%) (kN) (mm) (%)
PB01-N02 + 76.05 33.82 2.25 58.50 43.47 2.89
− 34.11 2.27 51.80 43.80 2.92 —
J. Earthquake and Tsunami Downloaded from www.worldscientific.com
by WSPC on 01/24/13. For personal use only.

Fig. 26. PB01-N02 test: Bending, slip and shear deformation components contribution for the
top displacement.

as for the case PB02-N01, wherein the numerical simulation provides poor approx-
imation to the experimental evidence mainly concerning the deformations in the
pre-yielding phase. As for the strength, both the yielding and peak ones, the numer-
ical and experimental results show quite good agreement, although after the peak
force the numerical response strongly deviates from the experimental one for the
reasons already pointed out in previous sections.

4.6. Comparative analysis of the test results


In this section a comparative analyses is performed concerning the experimental
results obtained for the four tests on the two tested rectangular section columns
with identical characteristics. The efficiency of column response simulations with
the simplified expressions, as well as with the refined numerical models, is also
addressed herein. The observed characteristics points, in terms of cracking, yield-
ing and maximum displacements, are first compared. Then, the accuracy of the

1250026-31
2nd Reading
January 25, 2013 9:21 WSPC/S1793-4311/238-JET 1250026

H. Rodrigues et al.
J. Earthquake and Tsunami Downloaded from www.worldscientific.com
by WSPC on 01/24/13. For personal use only.

Fig. 27. PB01-N02 test: Moment-curvature response (experimental results and numerical
estimation).

simplified method proposed by Lehman is analyzed. Finally, the contributions of


each deformation component (shear, bending, and slip) for the global column defor-
mation observed on the four tests are also compared.
In column N01, the cyclic loading is applied in the strongest direction (maximum
drift of 3.3%) while in column N02 the maximum drift imposed in the first test
(weakest direction) was 2.0%. Thus, the imposed deformations have induced a very
different condition for the second test. However, comparing the damage evolution
for these two tests, for the same level of imposed drift (Fig. 8 for column N01 and
Fig. 18 for Column N02), it can be observed that cracking starts at a earlier stage
for column N01 (5 mm) when compared with column N02 (10 mm). The same trend
is identified for the drift corresponding to spalling, which started at 2.0% for column
N01, whilst no indications of spalling were observed in column N02 for the same drift
level. The loading direction has an important influence in the damage state induced
to the column, justified by the larger column flexibility in its weakest direction.
As expected, for the second test of both columns, the initial columns’ stiffness
was considerably reduced by the damage induced during the first test. By contrast
this did not affect significantly the maximum column strength in the perpendicular
direction, but once the peak strength was reached in the second test, the strength
degradation became larger which is clearly related with the biaxial column loading.
Table 12 includes representative values for the cracking point, in terms of
strength and drift, observed in the first test of each column, as well as the val-
ues estimated by the Lehman method. Similarly, the corresponding yielding force
values, both the observed in the tests and those calculated by the same method,
are presented in Table 13.

1250026-32
2nd Reading
January 25, 2013 9:21 WSPC/S1793-4311/238-JET 1250026

Behavior of RC Building Columns Under Cyclic Loading: Experimental Study

Table 12. Cracking points: Test results and calculated values (Lehman values).

Observed Calculated
Test Force (kN) Drift (%) Force (kN) Drift (%)
PB01 N01 + 38.00 0.15 13.9 0.02
− 36.65 0.09
PB02 N02 + 23.38 0.10 6.58 0.04
− 23.70 0.12

Table 13. Observed and calculated yielding force.

Observed Calculated
Test Force (kN) Force (kN)
PB01 N01 + 59.60 55.13
J. Earthquake and Tsunami Downloaded from www.worldscientific.com

− 59.43
PB02 N02 + 31.87 29.13
− 33.11
by WSPC on 01/24/13. For personal use only.

According to Table 12, the cracking point prediction with the empirical formu-
lation is not satisfactory, when compared against the experimental results. The dif-
ferences can be due to difficulties on applying a precise visual criterion definition to
establish the cracking point, which was associated to the change of initial stiffness.
In fact, this is thought to be the major reason for the misagreement between the
experimental results and the calculated cracking point with the Lehman method.
Figure 28 graphically summarizes the calculated shear force vs. drift response
diagrams, both those obtained by the Lehman model and the experimentally

Fig. 28. Global shear-drift response: Estimated with the Lehman method and experimental
results.

1250026-33
2nd Reading
January 25, 2013 9:21 WSPC/S1793-4311/238-JET 1250026

H. Rodrigues et al.

recorded ones for the first test of each column. In terms of yielding shear force,
the calculated and experimentally observed values exhibited a satisfactory agree-
ment (see Table 13), which supports the ability of the adopted numerical tool to
estimate the flexural strength of RC elements.
The yielding displacement estimates obtained with the Lehman method gives
larger values than those derived from experimental results. According to the
Lehman procedure, the yielding displacement is determined as the sum of three
components, namely flexure, slip and shear, as discussed in Sec. 3.4. For both
columns, the results obtained with analytical calculations indicate that the shear
deformation component is less than 1% of the observed yielding displacement, the
flexure contribution is about 65% and the slip contribution is around 35%. Com-
J. Earthquake and Tsunami Downloaded from www.worldscientific.com

paring the calculations based on the Lehman method with the experimental results
(see Fig. 11) for column N01, the relative slip contribution for the yielding dis-
placement obtained with the analytical method is larger than the experimentally
obtained value. For column N02, the relative contributions obtained with the exper-
by WSPC on 01/24/13. For personal use only.

imental results and with the analytical expressions are similar (see Fig. 21). Even if
a good agreement was found in terms of relative contributions for the yielding dis-
placement, the differences between analytical and experimental results, in terms of
absolute yielding displacement, can be justified by the fact that the Lehman method
was developed and calibrated for circular RC columns, where the yielding is pro-
gressive, while for the rectangular columns under analysis the yield displacement
was associated with the first yielding of a longitudinal reinforcement bar.
Regarding the contribution of each deformation component to the lateral col-
umn displacement, the summary of the average results for each test is included in
Table 14. As already reported, for all the tests the shear deformation is irrelevant,
contributing for less than 1% of the total top-displacement.
Concerning the slip and bending deformation components, the results are in
good agreement with the obtained by other authors [Lehman et al., 1998; Calderone
et al., 2001; Sezen, 2008]. The bar slippage deformation contribution to the total
member lateral displacement can be significant and, therefore, it should be consid-
ered in the analysis and modeling of reinforced concrete members under severe cyclic
loadings. The adopted instrumentation scheme for the tests is in agreement with
the strategy followed by other authors [Calderone et al., 2001; Marefat et al., 2009].

Table 14. Contribution of each behavioral mechanism to


the global lateral column deformation.

Contribution to the
top-displacement
Test Shear Slip Bending
PB01 N01 (1st test) <1% 6% 75%
PB02 N01 <1% 45% 50%
PB02 N02 (1st test) <1% 35% 55%
PB01 N02 <1% 30% 60%

1250026-34
2nd Reading
January 25, 2013 9:21 WSPC/S1793-4311/238-JET 1250026

Behavior of RC Building Columns Under Cyclic Loading: Experimental Study

The test PB01-N01 corresponds to the first cyclic test on the column N01 for its
strong direction. From this test, it was observed that the lateral displacement of the
column is controlled essentially by flexural deformation, while the slip deformation
has a minor contribution (around 6%). For the second test on the same column (in
its weakest direction), the involved bending moment values are lower. Furthermore,
and even more important, the crack at the column base (induced in the previous
test) led to a rocking-like mechanism that justifies the larger slip deformation con-
tribution for the total lateral deformation (45%). As for column N02, it was tested
first in the weakest direction, the crack at the column base section developed soon
and the bar slippage reached an important contribution to the lateral deformation
(35%). Subsequently, in the second test of that column (in its strongest direction)
J. Earthquake and Tsunami Downloaded from www.worldscientific.com

the flexural deformation component tended to assume larger importance, leading to


a decrease of the slip deformation component; however, due to the damage induced
during the first test the slip contribution still kept a large value (30%). In any case,
for the analyzed columns, the bending displacement contribution is the dominant
by WSPC on 01/24/13. For personal use only.

component. The sum of the deformation components obtained in the simplified pro-
cedure is not equal to 100% because, besides the errors associated to the simplified
methodology, the displacement contributions due to non-linear shear deformation
and to sliding of the column base relatively to the reaction floor are not consid-
ered. The later contribution was measured and very small relative displacement
values were obtained, but, even so these minor displacements were subtracted to
the displacements recorded by the LVDTs.

4.7. Damage index


The characterization and quantification of structural damage is essential for the
structural performance and damage assessment of buildings under earthquake load-
ing. The definition of a damage measure and its corresponding quantification gives
an important contribution for the interpretation of the non-linear response of struc-
tures [Romão, 2002]. In practical terms, a damage measure indicates how close a
specific structural member is from a given limit-state such as, for example, collapse.
In the last years, several damage models have been proposed for RC structures
in order to allow quantifying the damage induced by earthquakes [Banon et al.,
1981; Park and Ang, 1985; Cosenza and Manfredi, 1993; Williams and Sexsmith,
1995; Chai et al., 1995; Bento, 1996]. These damage models, used in the structural
response analysis and damage evaluation, are based on specific structural properties
and structural parameters that measure the structural degradation [Bento, 1996].
In general, damage assessment methods can be used either at the global level,
as for instance a frame structure and/or a building storey, or at the local level, for
a single structural member such as columns and beams [Varum, 2003]. When it is
used for damage evaluation in structural elements, as considered in this work, the
damage index is usually calculated for the end element zones, where the non-linear
response and damage is expected to be concentrated in under cyclic loadings. The

1250026-35
2nd Reading
January 25, 2013 9:21 WSPC/S1793-4311/238-JET 1250026

H. Rodrigues et al.

formulation adopted in the columns test results analysis is based on the well-known
damage index methodology proposed by Park and Ang [Park and Ang, 1985].

4.7.1. Park and Ang damage index


This damage index proposed [Park et al., 1987] is defined by a linear combination
of the normalized maximum deformation and the normalized dissipated hysteretic
energy resulting from cyclic loading. The damage index DI is therefore expressed
by Eq. (4):

φmax dE
DI = +β , (4)
φu My φu
J. Earthquake and Tsunami Downloaded from www.worldscientific.com

where φmax is the maximum


 curvature demand of the structural member, φu is the
ultimate curvature, dE is the dissipated hysteretic energy and My is the yielding
moment of the structural member; β is a degradation parameter which represents
the cyclic response influence in the column damage and can be estimated with
by WSPC on 01/24/13. For personal use only.

empirical expressions based on structural parameters. The damage index typically


ranges from zero to one, but it is formally upper bounded by the unit value; however,
when DI ≥ 1 it means that total damage (collapse) is reached.
Several empirical expressions, based on experimental results, can be found in
the literature to estimate the strength degradation parameter β, for which a typical
value of 0.05 is often adopted [Kunnath et al., 1990]. Equation (5) was proposed
by Kunnath et al. [1990] and is one of the most used expressions to estimate β as
adopted by Arêde [1997]. This expression was empirically formulated from regres-
sion analysis of a large number of cyclic test data on RC elements (about 260),
representing typical building beams and columns. Therefore, the expression applies
for the particular materials, dimensions and detailing of the RC columns studied
in the present work and reads as:
β = 0.9100ρw (0.37. max{ν; 0.05} + 0.5(ωt − 0.17)2 ), (5)
where ρw is the volumetric confinement ratio (volume of closed stirrups divided
by the volume of confined concrete core), ν is the normalised axial stress (taken
positive for compression) and ωt is the mechanical ratio of tensile longitudinal
reinforcement.
Concerning the ultimate deformation capacity, although some expressions have
been also proposed out of several experimental tests performed up to failure on
beams and columns [Park et al., 1987]), for the present study, and considering the
assumption made for the determination of the ultimate curvature as discussed in
Sec. 3.4.3, the ultimate curvature capacity of columns was determined from the
moment-curvature curves obtained by imposing a tensile strain limit of 0.75% in
the longitudinal reinforcement steel. This option is justified since the tested columns
did not reach failure of the longitudinal reinforcement bars in order to allow testing
the column in the other direction. Consequently, the ultimate curvature could not
be calculated from test results based on the ultimate steel strain.

1250026-36
2nd Reading
January 25, 2013 9:21 WSPC/S1793-4311/238-JET 1250026

Behavior of RC Building Columns Under Cyclic Loading: Experimental Study

Table 15. Damage index vs. observed damage.

Damage inspection
Degree of damage Physical appearance Calculated local damage index
Collapse Total or partial building collapse >1.00
Severe Extensive crushing of concrete. Disclosure 0.75–1.00
of buckled reinforcements
Moderate Extensive large cracks. Spalling of 0.35–0.75
concrete in weaker elements
Minor Minor cracks throughout building. Partial 0.10–0.35
crushing of concrete columns
Slight Sporadic occurrence of cracking 0.00–0.10

In order to establish the correspondence between the damage index values cal-
J. Earthquake and Tsunami Downloaded from www.worldscientific.com

culated in the next section with the recorded damage. Table 15 includes the damage
index boundaries, for each damage degree, as typically observed after earthquake
events.
by WSPC on 01/24/13. For personal use only.

4.7.2. Damage index evolution for the tested columns


The damage index was computed from the test results on the two RC columns
cyclically tested. Equation (5) was applied, for each direction of both RC columns
in order to estimate β and, from the moment-curvature diagrams, the maximum
curvatures were obtained. The corresponding results (for the two directions) are
summarized in Table 16.
The damage index, DI, was first calculated from the test results of the non-
damaged columns, PB01-N01 and PB02-N02. The obtained results are plotted in
Figs. 29 and 30 in terms of DI evolution during the cyclic tests, which include the
DI values obtained for each column at the end of each test.
According to the results obtained in terms of damage index evolution for the
two tested columns, and relating with the typical damage description associated to
the DI values, it can be concluded that these results agree well with the physical
damage evolutions observed in the columns during the tests. In the case of PB01-
N01 test, the cracking, concrete spalling and bar buckling events are in accordance
with the predictions obtained with the damage index calculations. For the case of
PB02-N02 test, the accuracy of the damage index results is not so clear, due the
minimum damage installed in the column; however a good approximation is quite
apparent.

Table 16. Estimated parameters for RC


column damage index calculations.

Colum direction β φu (/m)


Strong 0.030 0.238
Weak 0.029 0.536

1250026-37
2nd Reading
January 25, 2013 9:21 WSPC/S1793-4311/238-JET 1250026

H. Rodrigues et al.
J. Earthquake and Tsunami Downloaded from www.worldscientific.com
by WSPC on 01/24/13. For personal use only.

Fig. 29. Damage index evolution for the test PB01-N01.

Fig. 30. Damage index evolution for the test PB02-N02.

The damage index calculations were not applied to the other subsequent tests
(PB02-N01 and PB01-N02), because the method is not adapted and calibrated to
biaxial loading, for which further research work is still required on the calibration
of the involved parameters.
From the damage index results analysis, it can be concluded that the influence
of the energy dissipation contribution on the damage index is not negligible, but it
is of minor importance when compared to the component related to the maximum
deformation. In column PB01-N01 the dissipated energy is responsible for 30% of
the DI calculated at the end of the test and, in the PB02-N01 case, this portion

1250026-38
2nd Reading
January 25, 2013 9:21 WSPC/S1793-4311/238-JET 1250026

Behavior of RC Building Columns Under Cyclic Loading: Experimental Study

decreases to 15%. In both columns, the difference found in the dissipated energy
contribution for the DI is associated with the non-linear response. This agrees with
the fact that, in column PB01-N01, the non-linear incursion is larger than in PB02-
N01, which increases the contribution of the obtained energy dissipation.

5. Final Remarks
In the present study, the behavior of two typical reinforced concrete columns sub-
jected to horizontal cyclic loading was presented and discussed.
As usual and expected, the damage was concentrated at the bottom of the
column, in the plastic hinge zone, within a distance corresponding to half of the
J. Earthquake and Tsunami Downloaded from www.worldscientific.com

cross-section height. No shear cracks were observed. The evidenced crack patterns
suggest that bending and slip mechanisms controlled the column response. More-
over, the results obtained with the simplified procedure adopted for the estimation
of the deformation components (based on the experimental results) prove that shear
by WSPC on 01/24/13. For personal use only.

deformation for these columns is negligible.


The experimental results were compared with those obtained with a simplified
method for the prevision of the global force-displacement response of columns. The
simplified method was not suitable for the identification of the yielding displace-
ment.
Regarding the damage index calculated from the test results of the two columns,
the results obtained were in good agreement with the observed damage evolution
during the tests. From these results it should be highlighted also the relatively large
contribution of the energy dissipation component in the calculated damage index,
particularly for the one column which has shown a pronounced non-linear response
with significant energy dissipation.

Acknowledgments
The authors acknowledge Mr. Nuno Reis for the construction of the RC col-
umn specimens and to the LESE laboratory staff, particularly Mr. Valdemar Lus,
Mr. Andre Martins and Eng. Lus Noites, for all the support in the preparation and
implementation of the testing set-up. This paper refers research work made with
financial contribution of “FCT — Fundao para a Cincia e Tecnologia”, Portugal,
that is gratefully acknowledged.

References
Abrams, D. P. [1987a] “Scale relations for reinforced concrete beam-columnjoints,” ACI
Struct. J. 84(6), 502–512.
Abrams, D. P. [1987b] “Influence of axial force variation on flexural behavior of reinforced
concrete columns,” ACI Struct. J. 84, 246–254.
Ang, B. G., Priestley, M. N. J. and Paulay, T. [1989] “Seismic shear strength of circular
reinforced concrete column,” ACI Struct. J. 86, 45–58.

1250026-39
2nd Reading
January 25, 2013 9:21 WSPC/S1793-4311/238-JET 1250026

H. Rodrigues et al.

Arêde, A. [1997] Seismic Assessment of Reinforced Concrete Frame Structures with a New
Flexibility based Element, PhD Thesis, Faculdade de Engenharia, Porto, Universidade
do Porto.
Atalay, M. B. and Penzien, J. [1975] The Seismic Behaviour of Critical Regions of Rein-
forced Concrete Components as Influenced by Moment, Shear and Axial Force, Earthq.
Engrg. Res. Center, Rep. No. UCB/EERC 75-19, University of California, Berkeley,
Ca.
Banon, H., Biggs, J. and Irvine, H. [1981] “Seismic damage in reinforced concrete frames,”
ASCE, J. Struct. Div. 107, N9.
Bento, R. [1996] Assessment of the Seismic Behaviour of Reinforced Concrete Framed
Structures: An Approach Based on Damage Indices, PhD Thesis, IST, Lisbon.
Bousias, S. N., Verzelleti, G., Fardis, M. N. and Magonette, G. [1992] “RC columns in
cyclic biaxial bending and axial load,” 10th World Conf. on Earthq. Engrg, Madrid,
J. Earthquake and Tsunami Downloaded from www.worldscientific.com

pp. 3041–3046.
Chai, Y., Romstad, K. and Bird, S. [1995] “Energy-based linear damage model for high-
intensity seismic loading,” ASCE, J. Struct. Eng. 121, N5.
Calderone, A. J., Lehman, D. E. and Moehle, J. P. [2001] Behavior of Reinforced Concrete
by WSPC on 01/24/13. For personal use only.

Bridge Columns Having Varying Aspect Ratios and Varying Lengths of Confinement,
PEER Report 2000/08.
CEB [1996] “RC frames under earthquake loading,” Lausanne Bull. Report No. 231 (303
pages), ISBN: 978-0-7277-2085-6 more information on: http://www.fib-international.
org/rc-frames-under-earthquake-loading
CEN [2003] “Eurocode 8: Design of structures for earthquake resistance — Part 1–3:
Strengthening and repair of buildings — European prEN 1998-1-3,” B. European
Committee for Standardization, Belgium.
Cosenza, E. and Manfredi, G. [1993] “The use of damage functionals in earthquake engi-
neering: A comparison between different methods,” Earthquake Engng. Struct. Dyn.,
22, 855–868. Doi: 10.1002/eqe.4290221003.
Dhakal, R. P., Mander, J. B. and Mashiko, N. [2006] “Identification of critical ground
motions for seismic performance assessment of structures,” Earthq. Eng. Struct. Dyn.
35(8), 989–1008.
Delgado, P., Rocha, P., Rodrigues, V., Arêde, A., Pouca, N. V., Costa, A., Delgado, R.
and Santos, M. [2005] Experimental Tests on Seismic Retrofit of RC Columns. 1st
US–Portugal International Workshop. Grand Challenges in Earthquake Engineering.
250 years after the 1755 Lisbon Earthquake, Lamego, pp. 16.1–16.10.
Delgado, P., Rodrigues, V., Rocha, P., Santos, M., Arde, A., Vila Pouca, N., Costa, A. and
Delgado, R. [2006] “Experimental tests on seismic retrofit of RC Piers,” 8NCEE —
Eighth U.S. National Conf. on Earthquake Engineering, San Francisco, California.
Delgado, R., Delgado, P., Vila-Pouca, N., Arêde, A., Rocha, P. and Costa, A. [2009] “Shear
effects on hollow section piers under seismic actions: Experimental and numerical
analysis,” Bull. Earthq. Eng. 7, 377–389.
Eberhard, M. O., Baldridge, Steven, Marshall, Justin, Mooney, Walter and Rix, G.J.
[2010] The MW 7.0 Haiti earthquake of January 12, 2010, USGS/EERI Advance
Reconnaissance Team report, U.S. Geological Survey Open-File Report 2010-1048,
p. 58.
EC2 [2004] “Design of concrete structure, part 1-1: General rules and rules for buildings,”
European Standard EN 1992-1-1, European Committee for Standardization (CEN),
Brussels, 2004.
Hognestad, E. [1951] A Study of Combined Bending and Axial Load in Reinforced Concrete,
Bulletin Series 339, Univ. of Illinois Exp. Sta., Illinois, U.S.A.

1250026-40
2nd Reading
January 25, 2013 9:21 WSPC/S1793-4311/238-JET 1250026

Behavior of RC Building Columns Under Cyclic Loading: Experimental Study

Kunnath, S. K., Reinhorn, A. M. and Park, Y. J. [1990] “Analytical modelling of inelastic


seismic response of RC structures,” ASCE J. Struct. Eng. 116, N4.
Lehman, D. E., Calderone, A. J. and Moehle, J. P. [1998] “Behavior and design of slender
columns subjected to lateral loading,” Proc. Sixth U.S. National Conf. Earthquake
Engineering, Earthquake Engineering Research Institute Oakland, California.
Li, K. N., Aoyama, H. and Otani, S. [1988] “Reinforced concrete columns under varying
axial load and bidirectional horizontal load reversals,” Proc. 9th World Conf. Earth-
quake Engineering, Tokoyo, Japan.
Li, L., Mander, J. B. and Dhakal, R. P. [2008] “Bidirectional cyclic loading experiment
on a 3D beam-column joint designed for damage avoidance,” ASCE J. Struct. Eng.
134(11), 1733–1742.
Low, S. and Moehle, J. P. [1987] Experimental Study of Reinforced Concrete Columns Sub-
ject to Multiaxial Cyclic Loading, Earthq. Engrg. Res. Center, Rep. No UCB/EERC
J. Earthquake and Tsunami Downloaded from www.worldscientific.com

87-14, Univ. of California, Berkeley, Ca.


Lu, Y., Vintzileou, E., Zhang, G.-F. and Tassios, T. P. [1999] “Reinforced concrete
scolumns under cyclic actions, Soil Dyn. Earthq. Eng. 18, 151–167.
Mander, J. B., Priestley, M. J. N. and Park, R. [1983] “Behavior of ductile hollow reinforced
by WSPC on 01/24/13. For personal use only.

concrete columns,” Bull. New Zealand National Society for Earthq. Eng. 97(7), 1969–
1990.
Marefat, M., Hassanzadeh Shirazi, S. M., Rostamshirazi, R. and Khanmohammadi, M.
[2009] “Cyclic response of concrete beams reinforced by plain bars,” J. Earthq. Eng.
13(4), 463–481.
Mirand-Guedes, J. P. [1997] Seismic Behaviour of Reinforced Concrete Bridges. Modelling,
Numerical Analysis and Experimental Assessment,” PhD Thesis, Universidade do
Porto, Porto.
McKenna, F. [1997] Object-Oriented Finite Element Programming: Frameworks for Anal-
ysis, Algorithms and Parallel Computing, PhD Thesis, University of California,
Berkeley.
NP-EN206 [2000] “Concrete — Specification, performance, production and conformity
(Portuguese Vertion),” NP-EN206: European Committee for Standardization.
NP EN 10002-1 [2006] “Tensile testing of metallic materials — Part 1: Method of test at
ambient temperature,” NP-EN206: European Committee for Standardization.
Ozcebe, G., Ramirez, J., Wasti, S.T. and Yaku, A. [2004]. 1 May 2003 Bingöl Earth-
quake Engineering Report, Report No. 2004/1, Structural Engineering Research Unit,
Middle East Technical University, Ankara, Turkey. Available at http://www.seru.
metu.edu.tr
Park, Y. J. and Ang, H. S. [1985] “Seismic damage model for reinforced concrete,”
ASCE — J. Struct. Eng. 111(4), 722–739.
Park, Y. J., Ang, A. H. S. and Wen, Y. K. [1987] “Damage-limiting aseismic design of
buildings,” Earthquake Spectra 3(1), 1–26.
Park, R., Priestley, M. J. N. and Gill, W. D. [1982] “Ductility of square confined concrete
columns,” J. Struct. Div. ASCE 108(4), 929–990.
Paulay, T. and Priestley, M. J. N. [1992] Seismic Design of Reinforced Concrete and
Masonry Buildings (John Wiley and Sons, New York), p. 767.
Priestley, M. J. N., Park, R. and Potangaroa, R. T. [1981] “Ductility of sprirally-confined
concrete columns,” J. Struct. Div. Proc. ASCE 107(ST1), 181–202.
Qiu, F., Li, W., Pan, P. and Qian, J. [2002] “Experimental tests on RC columns under
biaxial quasi-static loading,” Eng. Struct. 24, 419–428.
Rodrigues, V. [2006] Reforo Ssmico de Pilares de Beto Armado — Anlise e Validao Exper-
imental, MSc Thesis, Departamento de Engenharia Civl, Faculdade de Engenharia da
Universidade do Porto, Porto.

1250026-41
2nd Reading
January 25, 2013 9:21 WSPC/S1793-4311/238-JET 1250026

H. Rodrigues et al.

Romão, X. [2002] New Models for the Seismic Design of Structures, MSc Thesis, Faculty
of Engineering of the University of Porto, Porto.
Rossetto, T., Peiris, N., Alarcon, J., So, E., Sargeant, S., Sword-Daniels, V., Libber-
ton, C., Verrucci, E., Del Re, D. and Free, M. [2009] The l’Aquila, Italy Earthquake
of 6 April 2009 a Preliminary Field Report by EEFIT, EEFIT Report. Available
at http://www.istructe.org/resources-centre/technical-topic-areas/eefit/eefit-reports
(accessed September 2012).
Saatcioglu, M. and Ozcebe, G. [1989] “Response of reinforced concrete columns to simu-
lated seismic loading,” ACI Struct. J. 86(S1), 3–12.
Sezen, H. [2008] “Shear deformation model for reinforced concrete columns,” Struct. Eng.
Mech. 28(1), 39–52.
Taucer, F., Spacone, E. and Filippou, F. [1991] A Fiber Beam-Column Element for Seismic
Response Analysis of Reinforce Concrete Structures, University of California, Berkeley
J. Earthquake and Tsunami Downloaded from www.worldscientific.com

UCB/EERC-91/17.
Tirasit, P. and Kawashima, K. [2007] “Seismic performance of square reinforced concrete
columns under combined cyclic flexural and torsional loadings,” J. Earthq. Eng. 11(3),
425–452.
by WSPC on 01/24/13. For personal use only.

Varum, H. [2003] Seismic Assessment, Strengthening and Repair of Existing Buildings,


PhD Thesis, Civil Engineering Department Aveiro, University of Aveiro.
Williams, M. and Sexsmith, R. [1995] “Seismic damage indices for concrete structures: A
state of the art review,” Earth. Spec. 11(2), 319–349.

1250026-42

You might also like