Performance of Confined Boundary Regions of RC Walls

under Cyclic Reversal Loadings

Rafik Taleb, Masanori Tani , Susumu Kono
Journal of Advanced Concrete Technology, volume 14 ( 2016 ), pp. 108-124

Criterion for Preventing Formation of Story Mechanism in Vertically Irregular Wall Buildings
Thuat V. Dinh , Toshikatsu Ichinose
Journal of Advanced Concrete Technology, volume 2 ( 2004 ), pp. 385-394

Inventory survey of the 2003 Zemmouri Algeria earthquake: Case study of Dergana City
Hassane Ousalem , Hakim Bechtoula
Journal of Advanced Concrete Technology, volume 3 ( 2005 ), pp. 175-183

Investigation of a Hybrid Technique for Seismic Retrofitting of Bare Frames

Md. N. Rahman , TetsuoYamakawa
Journal of Advanced Concrete Technology, volume 5 ( 2007 ), pp. 209-222
Journal of Advanced Concrete Technology Vol. 14, 108-124 April 2016 / Copyright © 2016 Japan Concrete Institute 108

Scientific paper

Performance of Confined Boundary Regions of RC Walls under Cyclic

Reversal Loadings
Rafik Taleb1*, Masanori Tani2 and Susumu Kono3

Received 4 November 2015, accepted 17 March 2016 doi:10.3151/jact.14.108

Abstract
Observed damages in reinforced concrete wall buildings following some recent earthquakes raised concerns about the
seismic performance of rectangular RC walls. Damages in RC walls included spalling and crushing of concrete and
longitudinal reinforcement buckling at boundaries as well as global buckling. Preliminary studies attributed these dam-
ages to the lack of adequate confinement and detailing in wall boundary regions and high axial load level. Prism speci-
mens representing wall boundaries were tested to study the influence of reinforcement detailing, cross-section slenderness,
and loading type on the damages, failure modes, and compressive capacity of isolated confined boundary regions of RC
rectangular walls. It was found that the tensile strain prior to compressive strain affected the performance of thin wall
boundaries and may lead to different failure modes when subjected to cyclic loading. It was also found that dense
transverse reinforcement detailing in thin confined boundaries did not improve their compressive capacity. Design and
detailing rules to prevent global buckling and reinforcement bar buckling were also evaluated. A Numerical model that
takes into account buckling of reinforcement was proposed to simulate response curves of cyclically tested specimens.
The model showed the influence of reinforcement buckling behavior on reducing the compressive capacity for elements
with buckling of reinforcement failure.

1. Introduction 1992). A common practice in the design of structural

walls in Japan and some other countries is the use of
Reinforced concrete structural walls are commonly used barbell shape cross sections with confined boundary
as lateral-load resisting components in medium and columns that can carry a large amount of axial load.
high-rise buildings located in earthquake-prone regions. However, modern architecture and design practices
When properly designed and detailed, RC walls are promoted the use of slender rectangular walls with the
considered to perform well under earthquake loading due confidence that these planar walls with uniform wall
to their lateral stiffness and strength. By providing ade- thickness can be designed with adequate ductility.
quate strength and ductility, these walls also behave in a Following the 2010 Chile earthquake and the 2011
ductile flexural manner. To achieve this goal, lateral New Zealand earthquake, observed damages in RC wall
instability, fracture or buckling of longitudinal rein- buildings raised concerns about the seismic performance
forcement should be prevented (Paulay 1986). Observa- of rectangular RC walls. In these earthquakes, severe
tions following the 1985 Chile earthquake (Wood et al. damages happened to RC walls in numerous walled
1987) had shown that most RC buildings with wall sys- buildings leading to partial or total collapse (Kato et al.
tems exhibited satisfactory performances. Although de- 2010; Moehle et al. 2010). RC Wall damage included
tailing requirements for the Chilean wall system as of spalling and crushing of concrete at boundaries that often
1985 were less strict compared to those of the US or spread over the entire wall width, buckled of longitudinal
Japan. Moreover, RC buildings with a large number of reinforcement under compression and fracture under
rectangular walls, but without special detailing for the tension at boundaries (Westenenk et al. 2012; Wallace et
transverse reinforcement, have been reported to perform al. 2012). Global wall buckling was also observed in
well during major earthquakes (Wallace and Moehle some damaged buildings. Figure 1 shows typical wall
damages following the 2010 Chile earthquake. The
damaged wall regions were typically localized over a
1 very limited height of the wall thickness. It was reported
PhD Candidate, Department of Environmental Science
that lack of adequate confinement and detailing in
& Technology, Tokyo Institute of Technology, Tokyo,
boundary regions was one of the main causes of those
Japan.
damages. These observations raised questions about the
Lecturer, National School of Built and Ground Works
mechanisms that lead to reinforcing bars buckling, con-
Engineering, Algeria. *Corresponding author,
crete crushing, and global wall buckling, as well as the
E-mail: [email protected]
2 quantity and configuration of transverse reinforcement at
Associate Professor, Department of Architecture &
wall boundaries required to ensure good performance by
Architectural Engineering, Kyoto University, Japan.
3 maintaining a stable compressive region and ensuring
Professor, Structural Engineering Research Center,
large deformation capacity without drastic decrease in
Tokyo Institute of Technology, Tokyo, Japan.
R. Taleb, M. Tani and S. Kono / Journal of Advanced Concrete Technology Vol. 14, 108-124, 2016 109

(a) Out-of-plane buckling (b) Concrete crushing over wall width

(c) Reinforcement fracture (d) Concrete crushing and reinforcement buckling

Fig. 1 Typical walls damage in the 2010 Chile earthquake (Wallace et al. 2012). Reprinted with permission of the Earth-
quake Engineering Research Institute.

load carrying capacity. Preliminary studies (Wallace codes by detailing provisions for the spacing of trans-
2012; Talleen et al. 2012) indicate that greater amounts verse reinforcement by using ratios of transverse rein-
of transverse reinforcement may be required for thin forcement spacings (s) to longitudinal bar diameter (db)
walls and that tighter spacing of transverse reinforcement s/db. Large s/db ratios may result in limited confinement
may be required to suppress buckling of vertical rein- of concrete, and leave longitudinal reinforcement more
forcement. In Japan, although the use of walls with vulnerable to buckling instability.
boundary columns is still the common practice, the AIJ Moehle et al. (2011) tested several isolated boundary
Standard for Structural Calculations of RC Buildings elements, and comparisons were made between elements
(AIJ 2010) was revised in 2010 to allow the use of RC subjected to compression only and those subjected to 4%
walls with rectangular cross-sections. tension strain prior to compression, although latter
Literature review has shown that limited studies have loading type is considered as an extreme situation. This
been conducted on confined boundary regions, mainly comparison showed different failure modes, and con-
with focus on lateral instability due to out-of-plane firmed the vulnerability of wall boundaries to
buckling. Some researchers revealed that the potential of out-of-plane buckling and the reduction of compressive
out-of-plane buckling does not depend only on the capacity due to large tensile strains prior to compression.
compressive strains in wall boundaries, but also on the Chryanidis and Tegos (2012) tested 5 boundary elements
magnitude of the inelastic tensile strains imposed on wall with similar geometry and detailing under different
boundary prior to compressive strain when subjected to pre-tension strain prior to compression ranging from 0%
reversal cyclic loading (Paulay and Goodsir 1985). Chai to 5%. Test results showed that specimens subjected to
and Elayer (1999) conducted an experimental study to large pre-tension strains (3% and 5%) showed different
examine the out-of-plane stability of RC columns, rep- failure modes and a significant loss of compressive
resenting the confined boundary regions of a ductile strength of more than 65%. Massone et al. (2014) con-
rectangular RC wall, under large amplitude reversed ducted a test program on 24 specimens of boundary
cyclic tension and compression loading. The study con- elements with different confinement configuration and
firmed the critical influence of the maximum tensile slenderness under monotonic compression. Five speci-
strain on the lateral stability of these members. Design mens were subjected to pre-strain in tension of about 2%
recommendations for minimum required wall thickness prior to compression. Test results revealed the effect of
have been formulated (Paulay and Priestley 1993; Chai pre-tension to reduce compressive capacity by about
and Kunnath 2005). These studies demonstrated the 15% compared to capacity of specimens tested under
response of ductile boundary elements which resulted in monotonic compression. This study clarified the influ-
global buckling instead of buckling or fracture of the ence of reinforcement detailing, slenderness and loading
longitudinal bars. Issues related to buckling of longitu- type (Monotonic and Cyclic) on the compressive capac-
dinal reinforcing bars are usually addressed in the design ity, damage progress and failure modes of confined
R. Taleb, M. Tani and S. Kono / Journal of Advanced Concrete Technology Vol. 14, 108-124, 2016 110

Table 1 Element specimens cross sections and reinforcement details.

Section Layout Long. Transv.
Specimen s/db Loading type
b×l (mm) (Unit: mm) Reinf Reinf.
1B-M 220
Monotonic
55 55 55 55
3-D4@80
8
1B-C (ρt = 0.22%) Cyclic

90
238×108 220
2B-M 55 55 55 55
Monotonic
6-D4@80
8
2B-C (ρt = 0.42%) Cyclic
90
B1-type

10-D10
3B-M 55
220
55 55 55
(ρl = 2.63%) Monotonic
6-D6@80
8
3B-C (ρt = 0.95%) Cyclic
90

242×112 220
4B-M 55 55 55 55
Monotonic
6-D6@60
6
4B-C (ρt = 1.27%) Cyclic
90

214
5B-M 53 54 54 53
Monotonic
3-D4@80
238×108 5
5B-C (ρt = 0.22%) Cyclic
84
B2-type

10-D16
6B-M 53
214
54 54 53
(ρl = 7.33%) Monotonic
6-D6@60
242×112 3.75
6B-C (ρt = 1.27%) Cyclic
84

1C-M 69 69
345
69 69 69
Monotonic
4-D4@70
363×68 7
1C-C (ρt = 0.22%) Cyclic
C-type

50

12-D10
3C-M 69 69
345
69 69 69
(ρl = 3.24%) Monotonic
6-D6@40
367×72 4
3C-C (ρt = 1.29%) Cyclic
50

Note: ρl: is the longitudinal reinforcement ration ρl = Al /(b × l ) , ρt is the transverse reinforcement ratio ρt = At /(l × s ) , Al and At
are the longitudinal and transverse reinforcement area, respectively.

boundary regions of RC rectangular walls. The study loading type (Monotonic and cyclic) on their compres-
also aimed to assess whether global out-of-plane buck- sive capacity, damage process and failure modes. The
ling and buckling of vertical reinforcement could be behavior of boundary regions in a ductile RC wall sub-
expected at the boundaries of concrete walls, and to jected to lateral loading was studied by isolating the
explore the relationship between the phenomena of boundary regions of the wall as axially loaded RC col-
concrete crushing and reinforcing bar buckling and to umn. Although this approach lacks strain gradient effects
determine if current detailing practices are adequate to expected across the wall section and ignore the contri-
prevent bar buckling under extreme lateral loading. It is bution of the shear component, the idealization is useful
also important to predict the ultimate deformation ca- to provide an understanding of the behavior and to iden-
pacity by building numerical model which takes into tify critical parameters involved during lateral loading of
account these damage situations. RC walls, where confined boundaries are subjected to
large amplitude of tension and compression cycles. It is
2. Experimental program also important to contribute to make an experimental
database related to failures by buckling of reinforcement
An experimental program was conducted in order to and global buckling.
bring insight on the seismic performance of confined end
regions of RC rectangular walls. The objective was to 2.1 Description of the test specimens
investigate the influence of longitudinal and transverse A total of sixteen (16) rectangular elements with two
reinforcement detailing, cross-section slenderness and different sectional dimensions (B-type and C-type) hav-
R. Taleb, M. Tani and S. Kono / Journal of Advanced Concrete Technology Vol. 14, 108-124, 2016 111

ing approximately similar cross-sectional area were Table 2 Concrete mechanical properties.
constructed and tested. Table 1 shows the cross-sectional Compressive Young’s Splitting
configurations, the layouts and amounts of reinforcement. Strain at peak
strength modulus strength
(%)
The elements were built without cover concrete so that to (MPa) (GPa) (MPa)
avoid a sudden drop in the response curves following 24.5 0.18 26.3 2.3
spalling of cover concrete, since the objective was to
assess ultimate behavior and final failure modes. The Table 3 Reinforcing bars mechanical properties.
cross sections dimensions (b×l) for B-type and for C-type
Young’s Ultimate
specimens represent two levels of slenderness (h/b). The Yield strength
Bar modulus strength
dimensions are measured from the outside of the trans- (MPa)
(GPa) (MPa)
verse reinforcement. The shorter side length of the sec- D4 / 363 532
tion corresponds to boundary wall thickness. For B-type D6 192 365 516
specimens, specimens from 1B to 4B were constructed D10 190 347 484
with four levels of transverse reinforcement ratio ranging D16 188 325 462
from 0.22% to 1.27%, respectively. Specimens 5B and D25 192 381 567
6B were constructed with similar transverse reinforce-
ment ratios as for 1B and 4B specimens, respectively, but were bent 180-degrees at their ends and hanged to a D25
with larger longitudinal reinforcement ratio. For C-type (SD345) deformed reinforcing bars in the upper and
specimens (1C and 3C), two levels of transverse rein- lower stub to ensure good anchorage. D25 bars were also
forcement were set and were also similar to transverse used as longitudinal reinforcement for lower and upper
reinforcement ratios of 1B and 4B, respectively. For each stubs with D10 transverse reinforcement. The tested
of these eight configurations, two identical specimens elements had 600mm height (h) with fixed at both ends to
were built to produce sixteen specimens so that each the lower and upper stubs. This height represents the
configuration was tested under monotonic compressive lower portion of the confined boundary in a wall where
load and under cyclic tension and compression reversal likely compressive failure may occur. Observations from
load. The last characters in the specimens label stand for previous experimental studies indicate that the com-
loading type, M for monotonic and C for cyclic. D4 pressive failure region is quite limited within a height of
(SD295A) deformed reinforcing bars were used for about 2.5 times the wall thickness (Markeset and
transverse reinforcement for lightly confined specimens Hillerborg 1995; Takahashi et al. 2013). The elements
and D6 (SD295A) for densely confined specimens. All were cast vertically in two stages, the lower stub was cast
transverse reinforcement had 135-degree hooks. D10 first and then the element and the upper stub as one part
(SD295A) deformed reinforcing bars were used for lon- with intentionally roughened surface created at lower
gitudinal reinforcement for B1-type (ρl = 2.63%) and stub-element interface to insure adherence. Table 2 and
C-type (ρl = 3.24%) specimens, while D16 (SD295A) Table 3 show measured material properties for rein-
deformed reinforcing bars were used for B2-type (ρl = forcing bars and concrete, respectively. A concrete mix
7.33%). with 13mm of maximum aggregate size and 12cm for
Figure 2 shows vertical reinforcement layout of 6B slump test was used.
and 1C configurations. Longitudinal reinforcing bars

700 700
400
400

40
50

Upper & Lower Stub

Long. Rienf.：9-D25
Tranv. Reinf.：4-D10
Tested region
1400
1400

600
520
600
500

Long. Rienf. 10-D16 (6B)

12-D10 (1C)
Transv. Reinf. 6-D6@60 (6B)
4-D4@70 (1C)
c Strain gauges
40
50

400
400

Fig. 2 Vertical reinforcement layout of 6B-M/6B-C (left) and 1C-M/1C-C (right).

R. Taleb, M. Tani and S. Kono / Journal of Advanced Concrete Technology Vol. 14, 108-124, 2016 112

2.2 Loading Method and Measurement 40~560mm (Z2) and 560~600mm (Z3) for C-type ele-
Figure 3 shows test setup and loading protocol. A Uni- ments. This difference in insert positions between B-type
versal Testing Machine with a capacity of 1500kN was and C-type was due to transverse reinforcement position.
used to apply vertical load on the upper stub under the Two displacement transducers (Gauges i and j) were
condition of uniaxial tension and compression. Only also installed between upper and lower stub at both sides
vertical displacement is possible and the head of the to check any possible inclination during test. Although
testing machine have no freedom for rotation or lateral variation of configurations and confinement may affects
displacement. For monotonic tests, the compression load the degree of localization and measured strain, measured
was applied gradually until failure. For cyclic tests axial strains are compared in an average manner. The nominal
loading history was determined based on the average axial strain, εnom, was defined experimentally as the strain
strain at the lower part of previously tested RC structural corresponding to average displacement at both ends of
walls (Taleb et al. 2014) and previous tests on isolated the specimen over its total height h (600mm).
RC boundary elements (Chai and Elayer, 1999). A ratio
of tensile-to-compressive strain of 5 was used as the 1 ⎡ ( N1 + N 3 + N 5 ) ( S2 + S4 + S6 ) ⎤
ε nom = ⎢ + ⎥ (1)
loading protocol. Thus, the loading cycle consisted of an 2 ⎣ h h ⎦
initial half cycle of axial tensile strain followed by a
compression half cycle with a nominal target compres- where, N1, N2 and N3 are displacements corresponding to
sive strain 1/5 of the axial tensile strain. The test was north side transducers 1, 3 and 5, respectively. S1, S2 and
terminated when the resistance of the specimen de- S3 are displacements corresponding to south side trans-
creased significantly and the specimen exhibited insta- ducers 2, 4 and 6, respectively, and h is the specimen
bility. Thus, two cycles of loading were applied that height (600mm).
correspond to yielding tensile strain followed by tensile
strains of 0.5%, 1%, 1.5% 2%, 3% and 4%. A load cycle
is considered stable if the target compressive strain was
reached in two successive cycles without excessive de-
crease in compressive capacity. The specimens were
tested at low rate of monotonic loading, which was in the
5 6 5 6

4075
5075

order of 1mm/min for compression and 4mm/min for Z3 Z3

tension to insure that no strain rate effects were intro-
duced to damage process. Figure 4 shows displacement
Z2 Z2 4

520
500

7 8
transducers positions for B-type and C-type configura- 7 3 4 8 3

tions. Displacement transducers were mounted to both

ends of the longer side length of the prism section which Z1

75 40
Z1
7550

was divided into three measuring zones Z1 (Gauges c 1 2 1 2

and d), Z2 (Gauges e and f) and Z3 (Gauges g and

h). These zones were set at intervals of 0~50mm for Z1
at the bottom part of the element, 50~550mm for Z2 at
the middle, and 550~600mm for Z3 at the top part for (a) B-type (b) C-type
B-type elements, and at intervals of 0~40mm (Z1),
Fig. 4 Displacement gauges positions.

UTM
1500 kN Capacity

Upper stub

Lower stub

(a) Test setup (b) Loading protocol

Fig. 3 Test setup and loading protocol.
R. Taleb, M. Tani and S. Kono / Journal of Advanced Concrete Technology Vol. 14, 108-124, 2016 113

3. Experimental Results and Discussion and the compressive strength as sum of the concrete
uniaxial compressive strength and compressive yield
3.1 Axial load-Axial Nominal Strain Relation- stress of longitudinal reinforcement.
ships
Figure 5 shows axial load versus nominal axial strain N c = As f c′ + As f y (3)
relationships for all specimens. Each plot in the figure
represents response relations for both monotonic and where Ac and As are the cross-sectional area of concrete
cyclic loading. The upper and lower dashed lines indicate, gross section and longitudinal reinforcement, respec-
respectively, the calculated loads corresponding to the tively, and f’c and fy are the concrete compressive cyl-
yielding of longitudinal reinforcements. inder strength and the yield strength of longitudinal re-
inforcement.
N t = As f y (2) In C-type elements, vibration of concrete during con-
crete casting was conducted manually using steel rods

Fig. 5 Axial load-Nominal axial strain relationships.

R. Taleb, M. Tani and S. Kono / Journal of Advanced Concrete Technology Vol. 14, 108-124, 2016 114

since the use of vibrator was not possible due to the lack Onset of bar buckling for 1B-C element was observed at
of space. Some small honeycombs were observed after approximately -220kN that correspond to about 80% less
removing the formwork. Hence, the unconfined com- than the load level of bar buckling under monotonic
pressive strength was not fully reached. It should be also compression. Onset of bar buckling happened after
noted that excessively large compressive strain for 6B-M unloading from the first cycle of 2% tensile strain,
and 6B-C elements was due to an inclination of the ele- similarly to 5B-C element. Onset of bar buckling for
ments prior to extensive crushing. For all specimens, a 1C-C element started when unloading from the second
stable response was observed under low levels of axial cycle of 1.5% tensile strain and loading to the corre-
tensile strains for element tested under cyclic loading. sponding compressive strain, that is compressive strain
However, increasing the tensile strain level led to dif- corresponding to compressive peak load. The following
ferent response. These differences and the comparison cycle was marked by buckling of several longitudinal
monotonic and cyclic loading response are summarized bars and capacity drop. This demonstrate vulnerability of
in the following. slender elements to bar buckling. Response curves of
It was noted that specimens with thin boundaries elements which failed due to concrete crushing showed a
(C-type) were not able to fully develop the compressive smoother decrease of load carrying capacity compared to
strength. These configurations could not provide suffi- elements with failure mode by longitudinal bar buckling.
cient confinement although the transverse reinforcement This smooth decrease was more pronounced as the ratio
ratio was high for 3C configuration. The low confine- of transverse reinforcement was higher.
ment ratio and large difference between longitudi- Comparing densely and lightly confined specimens, it
nal-to-transverse bar diameters led also to a lower com- was shown that well confined specimens revealed capa-
pressive capacity. Comparison of load carrying capacity bility to sustain larger tensile strain in a stable manner.
between monotonic and cyclic loadings showed no sig- However, dense transverse reinforcement detailing
nificant difference for all tested elements. added little to the compressive capacity, especially for
Failure of 1B, 5B and 1C configurations as well as thin elements. Comparison of compressive capacity of
2B-C element was due to longitudinal reinforcing bar 4B (ρt =1.27%) to 1B (ρt =0.22%) configurations showed
buckling. 3B, 4B and 6B configurations as well as 2B-M an increased capacity of about 16%, while comparison
and 3C-M elements failed due to crushing of concrete, between 3C (ρt =1.29%) and 1C (ρt = 0.22%) configu-
while global buckling failure was observed for 3C-C rations display similar capacity even though the trans-
element. Failure mode due to buckling of reinforcement verse reinforcement ratio in 3C was set more than 5 times
is indicated when apparent longitudinal bar buckling is of that in 1C. These observations suggest that it may not
observed with slight damage in only concrete surround- be even possible to provide enough confinement in thin
ing longitudinal bars. Failure mode due to concrete sections by close transverse reinforcement spacing be-
crushing is indicated when extensive damage is observed cause the core concrete width is small and the pattern of
in compressive concrete without any apparent longitu- concrete crushing indicates that compression strain
dinal bars buckling. concentrates over a short height.
Comparison between monotonic and cyclic response Comparing the two levels of slenderness (B-type and
for elements with failure mode governed by buckling of C-type), it was shown that although they had similar
longitudinal reinforcement (1B, 5B, and 1C configura- confined area, the compressive load capacity of C-type
tions) showed that prior tensile strain affects considera- elements was in the range of 25% to 40% less than the
bly the load level at onset of bar buckling. Onset of bar capacity of B-type elements. This was due to the thin
buckling for elements tested under monotonic compres- core concrete in C-type elements where a similar con-
sion (1B-M, 5B-M and 1C-M) was noted around the peak fining effect to section with small aspect ratio cannot be
point, followed by a rapid drop of the load carrying ca- obtained and spread of concrete crushing by confined
pacity and revealing that their failure was related to lon-
gitudinal bar buckling. Following bar buckling, the core
concrete could not sustain the total axial load and exten-
sive concrete crushing happened at bar buckling region.
Response curves of elements that failed by buckling of
longitudinal bars showed a quick decrease of axial load
after the peak compressive load was reached. Figure 6
compares load-strain curve for B-type elements tested
under monotonic load. A sudden drop of the capacity was
observed and manifested by the rapid concrete crushing
at the region of reinforcement buckling.
On the other hand, elements tested under cyclic load-
ing (1B-C, 5B-C and 1C-C) showed onset of bar buck-
ling at lower compressive load before cracks completely
closed following unloading from peak tensile strain. Fig. 6 Comparison of load- strain curves for B-type ele-
ments under monotonic load.
R. Taleb, M. Tani and S. Kono / Journal of Advanced Concrete Technology Vol. 14, 108-124, 2016 115

Table 4 Numerical values for observed damage states.

Specimen Reinf. Comp. Yield. Peak Reinf. Buckl. Global Buckl. Failure Mode
Laod (kN) -839.3 -1053.2 -1045.3 -
1B-M Reinf. Buckling
Strain (%) -0.2135 -0.391 -0.4425 -
Laod (kN) -879.4 -1027.6 -290.2 -
1B-C Reinf. Buckling
Strain (%) -0.192 -0.378 -0.033 -
Laod (kN) -607.2 -876 - -
2B-M Concrete Crushing
Strain (%) -0.2115 -0.4655 - -
Laod (kN) -733.5 -978.8 -327.3 -
2B-C Reinf. Buckling
Strain (%) -0.1945 -0.4355 -0.0985 -
Laod (kN) -781.2 -1186.8 - -
3B-M Concrete Crushing
Strain (%) -0.199 -0.6185 - -
Laod (kN) -685.9 -1095.4 - -
3B-C Concrete Crushing
Strain (%) -0.188 -0.6415 - -
Laod (kN) -601 -1189.7 - -
4B-M Concrete Crushing
Strain (%) -0.18 -0.897 - -
Laod (kN) -763.7 -1201.5 - -
4B-C Concrete Crushing
Strain (%) -0.205 -0.6295 - -
Laod (kN) -834.2 -1045.3 -989.3 -
5B-M Reinf. Buckling
Strain (%) -0.2005 -0.377 -0.5055 -
Laod (kN) -894 -1120.8 -477.3 -
5B-C Reinf. Buckling
Strain (%) -0.201 -0.4045 1.032 -
Laod (kN) -565.2 -1237.2 - -
6B-M Concrete Crushing
Strain (%) -0.2105 -1.5775 - -
Laod (kN) -540.8 -1175 - -
6B-C Concrete Crushing
Strain (%) -0.176 -1.468 - -
Laod (kN) -664.9 -795 -517.7 -
1C-M Reinf. Buckling
Strain (%) -0.1995 -0.436 -0.628 -
Laod (kN) -662.1 -730 -256.8 -
1C-C Reinf. Buckling
Strain (%) -0.1885 -0.2985 -0.226 -
Laod (kN) -435.7 -719.4 - -
3C-M Concrete Crushing
Strain (%) -0.151 -0.618 - -
Laod (kN) -532.1 -838.6 - -140.5
3C-C Concrete Crushing
Strain (%) -0.1495 -0.5975 - 3.794

core concrete could not be ensured. Imposing a minimum reinforcing bars at the top region followed by the spalling
wall thickness would be an alternative means to suppress of surrounding concrete and buckling of multiple longi-
failures due to global buckling (Chai and Kunnath 2005) tudinal bars at this region with a buckling length corre-
and maintain a stable compression zone. sponding to one transverse reinforcement spacing. On
the other hand, 1B-C element started damage under
3.2 Damage process and failure modes compression by spalling of surface concrete at
Damage process is presented for each configuration mid-height and the start of buckling of two corner bars
under monotonic and cyclic loading conditions. All with one transverse reinforcement spacing for buckling
elements tested under monotonic compression exhibited length. Extensive spalling of concrete was shown in the
a stable behavior without apparent damages until peak following compressive cycle with large buckling of lon-
load. Following peak load point, different damage evo- gitudinal bars. Buckling length of corner and intermedi-
lutions and failures were observed. For cyclically tested ate supported bars corresponded to one transverse rein-
elements, horizontal cracks appeared at top and bottom forcement spacing while this buckling length for un-
element-stub interface. They also appeared uniformly at supported intermediate bars corresponded to more than
transverse reinforcement planes when loading in tension, two spacing. Crushing of concrete was not so severe at
indicating that these cracks were initiated by the trans- the buckling region following longitudinal reinforcement
verse reinforcement. Further tension loading led to buckling.
widely opened horizontal cracks. Table 4 gives the nu- For 2B-M element, damage started with the appear-
merical values for the observed damage states. Damage ance of multiple vertical cracks at mid-height that
evolution and failure modes are described in the fol- quickly led to large spalling of surface concrete followed
lowing. by crushing of concrete and buckling of several corner
and intermediate longitudinal bars over one transverse
3.2.1 B1-type specimens reinforcement spacing. 2B-C started to damage under
Figure 7 shows the final damage situation for B1-type compression by the spalling of concrete around some
specimens. For 1B-M element subjected to monotonic corner bars at the middle and then at the top regions
compression, first cracks appeared near corner vertical followed by buckling of longitudinal bars at those loca-
R. Taleb, M. Tani and S. Kono / Journal of Advanced Concrete Technology Vol. 14, 108-124, 2016 116

(a) 1B-M (b) 1B-C (c) 2B-M (d) 2B-C

(e) 3B-M (f) 3B-C (g) 4B-M (h) 4B-C

Fig. 7 Final damage situation for B1-type specimens.

tions. At further compressive strains, extensive spalling 3.2.2 B2-type specimens

of surface concrete between longitudinal reinforcing bars Figure 8 shows the final damage situation for B2-type
at mid-height region of the element occurred followed by and C-type specimens. For 5B-C, horizontal cracks
buckling of many longitudinal bars over one transverse opened widely and new horizontal cracks formed at
reinforcement spacing. A sudden concrete crushing at mid-spacing between transverse reinforcement at further
that region happened at final stage. A fracture of one tensile strains. At final stage, buckling of multiple lon-
longitudinal bar was observed when loading from 3% to gitudinal reinforcing bars happened simultaneously at
4% tensile strain. Failure modes of 2B configuration was mid-height region over three and four spacing of trans-
different depending on loading type. 2B-M failed due to verse reinforcement after the spalling of surface concrete
crushing of concrete, while failure of 2B-C was attrib- between longitudinal reinforcing bars. The transverse
uted to buckling of reinforcement that led to a sudden reinforcement (D4@80) was not able to contribute ef-
concrete crushing. Further large strains in tension and fectively in retaining larger longitudinal bars diameter
compression resulted in an increased number of regions (D16) and preventing them from buckling over large
were corner longitudinal bars buckled. This damage buckling length even though ratios of hoop spacing to
situation indicate that the pre-cracks at transverse rein- longitudinal bar diameter, s/db = 5, is within the limit of
forcement planes due to previous tension strains prior to ACI 318-14. This suggests that anti-buckling detailing
compressive strain facilitates their buckling compared to provisions should also be related to the ratio of longitu-
element tested under monotonic compression. dinal-to-transverse bar diameters. Buckling of unsup-
3B-M started damage with the appearance of multiple ported intermediate bars was more pronounced com-
vertical cracks at different locations around corner bars. pared to other bars. 5B-M reached maximum capacity
Further compressive strains led to spalling of surface without visible damage, followed by spalling of concrete
concrete and crushing of core concrete. For 3B-C, and buckling of longitudinal reinforcement over two and
spalling of concrete around a pair of corner bars occurred three transverse reinforcement spacing similarly to 5B-C
with core concrete crushing at final stage at top region of element.
the element. Damage evolution for 4B-M and 4B-C For 6B-C element, horizontal cracks appeared only at
elements was similar to 3B-M and 3B-C elements, re- transverse reinforcement planes under tension loading
spectively. and opened widely as tensile strain increased. At final
R. Taleb, M. Tani and S. Kono / Journal of Advanced Concrete Technology Vol. 14, 108-124, 2016 117

(i) 5B-M (j) 5B-C (k) 6B-M (l) 6B-C

(m) 1C-M (n) 1C-C (o) 3C-M (p) 3C-C

Fig. 8 - Final damage situation for B2-type and C-type specimens.

loading stage, both 6B-C and 6B-M failed by crushing of However, crushing of concrete in 3C-C was followed by
compressive concrete followed by localized buckling of global buckling of the element when unloading from the
the damaged region, but no buckling of longitudinal second cycle of 4% tensile strain indicating that global
reinforcement was observed. The damaged region was buckling was driven by prior induced large tensile strain.
located at the lower portion for 6B-M and at the top for This phenomena demonstrates the vulnerability of con-
6B-C. fined boundaries of slender walls to tensile strain excur-
sions prior to compressive strain. Concrete crushing was
3.2.3 C-type specimens very limited in height compared to B-type specimens.
Both 1C-C and 1C-M specimens failed by buckling of In specimens failing due to concrete crushing, con-
longitudinal reinforcement under compression. Buckling crete crushed over a height ranging approximately be-
length was observed over two and three transverse rein- tween 2 to 3 times element width. Fracture under tension
forcement spacing for 1C-M, while it extended in 1C-C of longitudinal reinforcing bars was not observed ex-
over more than four spacing of transverse reinforcement cluding one longitudinal bars for 2B-C at tensile strain
due to pre-cracks induced by tensile strain. Pre-cracking larger than 3%. Globally, no difference of the failure
condition facilitates the buckling of longitudinal rein- modes were shown when comparing failures under
forcement in addition to the very thin concrete core. monotonic and cyclic loading condition. Exception was
Similar to 5B configuration, buckling of unsupported noted for 2B and 3C configurations. 2B-M failed due to
intermediate bars was more pronounced than other bars, crushing of concrete, but failure of 2B-C was due to
suggesting that restraining unsupported intermediate bars buckling of reinforcement that led to a sudden concrete
in the confined boundary region should be considered, crushing. Also, 3C-M failed due to extensive crushing of
especially for slender walls. concrete, while 3C-C element showed a limited concrete
The final failure for 3C-M element was caused due to crushing region at the base followed by out-of-plane
extensive crushing of compressive concrete at the bottom buckling. Prior crushing assisted the global buckling
of element over a very limited height corresponding to over almost the total height of the element and resulted in
approximately two transverse reinforcement spacing. a large out-of-plane displacement.
Crushing of concrete for 3C-C was also concentrated at
the bottom within limited height, similarly to 3C-M.
R. Taleb, M. Tani and S. Kono / Journal of Advanced Concrete Technology Vol. 14, 108-124, 2016 118

4. Prediction of failure modes and damage ratio for the wall boundary element and the maximum
situations tensile strain prior to compressive load, ε sm , was pro-
posed as Eq. (4) (Paulay and Priestley 1993; Chai and
4.1 Potential of out-of-plane buckling Elayed 1999; Parra and Moehle 2014).
Figure 9 shows the final buckled shape of 3C-C. A ver-
tical line was drawn to highlight the transverse dis- bcr 1 ε sm − 0.005
= (4)
placement of the buckled element in the figure. Wide kh0 π βξ
cracks, which developed at transverse reinforcement
planes as a result of a large yield excursion, did not close where bcr is the critical wall boundary thickness, h0 is the
prior to full development of maximum compressive clear height, β is the effective depth parameter for lon-
strength due to residual tensile strain in the previously gitudinal reinforcement assumed to be 0.8 for two layers
yielded longitudinal reinforcement. This damage situa- of longitudinal bars and 0.5 when a single central layer of
tion caused a critical condition affecting the lateral sta- bars is used, and ξ is a parameter related to mechanical
bility of the wall (Paulay and Priestley 1993; Chai and reinforcement ratio that should satisfy:
Elayer 1999). However, crushing of concrete at the base
of 3C-C prior to global buckling contributed in a large ⎛ 2m ⎛ 2m ⎞
2
4m ⎞⎟
out-of-plane displacement when unloading from the ξ ≤ 0.5 ⎜1 + − ⎜ ⎟ + (5)
⎜ 0.85 ⎝ 0.85 ⎠ 0.85 ⎟⎠
second cycle of 4% tensile strain since the base acted as a ⎝
pin joint. This reveals that both large tensile strain prior
with m=ρlfy/f’c is the mechanical reinforcing ratio. For
to compressive strain and prior crushing affect the global
practical design, Parra and Moehle (2014) suggest that ξ
buckling failure mode for slender walls. Imposing a
=0.25. Eq.(4) becomes then:
minimum wall thickness would be an alternative means
to eliminate global buckling. kh0 1
Buckling may not be easily perceptible at the design = (6)
bcr 0.7 ε sm − 0.005
level because their mechanism is difficult to quantify
even with the current analysis capabilities. Tendency to
Figure 10 compares theoretical relation for wall in-
buckle in RC walls depend primarily on the wall slen-
stability given by Eq. (4) and elements test results. A
derness ratio and loading history or specifically the
value of β equal to 1.0 was used in Eq. (4) to consider the
maximum tensile strain in the boundary longitudinal
total thickness since the specimens were built without
reinforcement. Parra and Moehle (2014) suggested that
cover concrete. The equation may be used to judge the
buckling instability might be related to two damage
potential of global buckling.
situations. One is that tensile yielding softens the
boundary in one direction for subsequent loading in the
4.2 Potential of longitudinal bars buckling
opposite direction under compression, leading to global
Buckling of longitudinal reinforcing bars are usually
lateral instability of an intact wall. The second is that the
addressed by limiting the ratio of transverse reinforce-
wall crushes first, leaving an even smaller and irregular
ment spacing to longitudinal bar diameter, s/db. Large
cross section, leading to instability of the reduced cross
s/db ratios result in limited confinement of concrete, and
section as a secondary buckling failure. However, failure
leave longitudinal reinforcement more vulnerable to
of 3C-C showed a third damage situation, where prior
buckling instability. ACI 318-14 limits the ratio of s/db to
crushing at the bottom led to a global buckling rather
6. Rodriguez et al. (1999) introduced a criterion to assess
than a local buckling of the crushed region.
the onset of bar buckling based on monotonic and cyclic
Based on buckling theory for prismatic sections under
tests on isolated reinforcing bars with various s/db ratios.
cyclic loading, a relation between the critical slenderness
Tests indicated that bars subjected to cyclic loading were

Elemen

Fig. 10 Comparison with theoretical relation for global

Fig. 9 Final buckling shape of 3C-C. buckling instability.
R. Taleb, M. Tani and S. Kono / Journal of Advanced Concrete Technology Vol. 14, 108-124, 2016 119

more susceptible to buckling failures than bars subjected the ACI 318-14. This suggests that anti-buckling detail-
to monotonic loading. A strain parameter was introduced ing provisions should also be related to the ratio of lon-
as an indicator of the onset of bar buckling. This ap- gitudinal-to-transverse bar diameters. Figure 12 shows
proach is limited to s/db ratio equal to 8 and does not take relation between longitudinal-to-transverse reinforce-
into account buckling susceptibility over multiple ment bar diameters times transverse reinforcement ratio
transverse reinforcement spacing. as an index to measure the effectiveness of transverse
Based on quasi-static tests on RC columns subjected to reinforcement to prevent bar buckling. The index was
lateral loads and constant or varying axial load, Kato et al. also able to predict vulnerable specimens to bar buckling.
(1995) proposed a model to estimate the buckling length A limit of 1.2 is suggested and this index is considered as
and the onset of inelastic buckling of corner reinforcing complementary to previous rules for eliminating bar
bars. The buckling length is given as a function of the buckling.
number of transverse reinforcement spacing over where
the buckling of longitudinal reinforcement is likely to ρt db / dt ≥ 1.2 (8)
happen. Transverse reinforcement index, α, was also
with ρt is the transverse reinforcement ratio, db is the
proposed as a design rule to prevent buckling of longi-
longitudinal reinforcing bar diameter, and dt is the
tudinal bars given as:
transverse reinforcing bar diameter.
( Aw / Al )( f wy / f y )
α= ≥ 0.039 (7) 5. Analytical Prediction of Cyclic Load -
s / db
Strain Relations
where Aw and Al are the areas of transverse and longitu-
dinal reinforcement, respectively, fwy and fy are yield In order to simulate the hysteretic behavior of cyclically
stress of transverse and longitudinal reinforcement, re- tested elements, an accurate and reliable prediction of
spectively, and s/db is the ratio of transverse reinforce- experimentally observed response is proposed. The
ment spacing to longitudinal bar diameter. Figure 11 model addresses important issues such as the hysteretic
shows relation between transverse reinforcement index behavior in both cyclic compression and tension; the
and observed failure mode for cyclically tested element. progressive degradation of stiffness of the unloading and
Although the lower limit of 0.039 for the index of reloading curves for increasing values of strain; and the
transverse reinforcement seems to be conservative, the effects of confinement, tension stiffening, and gradual
index was able to predict the vulnerable specimens to crack closure. The model takes into account concrete
buckling of longitudinal reinforcement as those with the damage and hysteresis, while retaining computational
lower transverse reinforcement ratios. Configuration 5B efficiency. The monotonic envelope curve of the hyster-
with the lowest index, showed the most drastic failure etic model for concrete in compression follows the
due to buckling of longitudinal bars, while buckling of monotonic stress-strain relation of modified Kent and
longitudinal reinforcement for configuration 2B was Park model (Scott et al. 1982) offering a good balance
only observed for 2B-C tested under cyclic loading. between simplicity and accuracy (Fig. 13a) . The hys-
Among the tested configurations, 2B was considered as a teretic behavior of concrete in both compression and
limit between failure due to longitudinal bars buckling tension were modeled using hysteretic unloading and
and failure by concrete crushing. In 5B configuration, reloading rules proposed by Yassin (1994) as a set of
transverse reinforcement did not effectively retain lon- linear stress-strain relations. The model is able to simu-
gitudinal bars and prevent them from buckling over large late stiffness degradation for both unloading and re-
buckling length even though ratios of hoop spacing to loading. The model provides the flexibility to represent
longitudinal bar diameter, s/db = 5, is within the limit of the hysteretic behavior of confined and unconfined con-
crete in both cyclic compression and tension (Fig. 13b).

Fig. 11 Transverse reinforcement index for tested ele- Fig. 12 Longitudinal to transverse reinforcement index.
ments.
R. Taleb, M. Tani and S. Kono / Journal of Advanced Concrete Technology Vol. 14, 108-124, 2016 120

(a) Envelope curve (Scott et al. 1982) (b) Hysteretic rules (Yassin, 1994)
Fig. 13 Stress-strain relations for concrete.

Ii (a) Menegotto-Pinto model (b) Monotonic compressive buckling model

Fig. 14 Stress-strain relations for reinforcing bars.

The numerical model used for reinforcing steel was Boundary

based on Menegotto-Pinto model (Menegotto and Pinto
1973). It was extended by Filippou et al. (1983) to in-
clude isotropic strain hardening effects as shown in Fig.
14a. To include the effect of buckling of reinforcement, bcr ε − 0.005 yes Global
Dhakal and Maekawa (Dhakal and Maekawa 2002) ≥ sm Buckling
model was implemented. Based on a calibration with kh0 βξ
experimental data, Dhakal and Maekawa model explic- No
itly provides the compressive stress - strain response of
the rebar that is linear for pre-yielding branch and follow s/db ≤ 6
Eq. (9) and Eq. (10) for post-yield behavior (Fig. 14b). Yes No
ρ t d b d t ≥ 1.2
σ sc ⎛ f ⎞ ⎛ ε sc − ε y ⎞
= 1 − ⎜1 − i ⎟ ⎜ ⎟⎟ for ε y < ε sc ≤ ε i (9)
σt ⎝ fit ⎠ ⎜⎝ ε i − ε y ⎠ Reinforcement
Concrete
Crushing Buckling
σ sc = f i − 0.02 Es (ε sc − ε i ) ≥ 0.2 f y for ε sc > ε i (10)

where σt and fit are the stresses in the tension envelope

corresponding to εsc (current strain) and εi (strain at the sNmax sbNmax

intermediate point), respectively. fy and εy are yielding

stress and strain, respectively. The coordinates of inter- Fig. 15 Procedure to determine compressive strength.
mediate point correlated to √fy L/db are given by:
where α is a coefficient that takes into account strain
εi fy L hardening which is equal to 0.75 for elastic-perfectly
= 55 − 2.3 ≥7 (11) plastic bars, and 1.0 for bars with continues linear
εy 100 db
hardening. L/db is the ratio of the buckling length to the
longitudinal bar diameter. Buckling length of rein-
fi ⎛ f y L ⎞ fi
= α ⎜1.1 − 0.016 ⎟ ; ≥ 0.2 (12) forcement was evaluated based on Kato et al. (1995).
fit ⎜ 100 d ⎟ fy Buckling length were evaluated as 3 transverse rein-
⎝ b ⎠
forcement spacing for specimens failing due to buckling
R. Taleb, M. Tani and S. Kono / Journal of Advanced Concrete Technology Vol. 14, 108-124, 2016 121

Table 5 Comparison between experimental and simulated peak load and strain.
Experiment Numerical Simulation Comparison
Specimen eNmax eε@max Failure cNmax sNmax sbNmax sε@max eNmax eε@max
(kN) (%) mode (kN) (kN) (kN) (%) / sbNmax / sε@max
1B-M -1053 -0.391 RB -957 -957 -0.206 1.10 1.90
1B-C -1029 -0.378 RB -946 -815 -0.203 1.26 1.86
2B-M -876 -0.466 CC -977 -977 -0.208 0.90 2.24
2B-C -979 -0.435 RB -968 -878 -0.208 1.12 2.09
-912
3B-M -1187 -0.619 CC -1030 -1030 -0.225 1.15 2.75
3B-C -1095 -0.642 CC -1101 -1101 -0.218 0.99 2.94
4B-M -1195 -0.630 CC -1062 -1062 -0.236 1.13 2.67
4B-C -1202 -0.897 CC -1051 -1051 -0.240 1.14 3.74
5B-M -1045 -0.249 RB -1350 -1350 -0.202 0.77 1.23
5B-C -1121 -0.278 RB -1309 -1158 -0.201 0.97 1.38
-1310
6B-M -1237 -1.578 CC -1398 -1398 -0.228 0.88 /
6B-C -1175 -1.468 CC -1392 -1392 -0.281 0.84 /
1C-M -795 -0.436 RB -990 -987 -0.206 0.81 2.12
-944
1C-C -755 -0.299 RB -990 -789 -0.206 0.96 1.45
Note: eNmax and eε@max are experimental peak load and corresponding strain, respectively. Nc compressive load capacity based on
uniaxial concrete strength and longitudinal reinforcement yielding. sNmax and sbNmax are simulated peak load without and with re-
inforcement buckling effect, respectively, and sε@max is the corresponding peak strain. CC: Concrete Crushing, RB: Reinforcement
Buckling, GB: Global Buckling.

of reinforcement, and one transverse reinforcement loading. The following conclusions were drawn.
spacing for specimens failing due to concrete crushing of Three different failure modes were observed depend-
global buckling. ing on confinement and slenderness levels: crushing of
Figure 15 shows the procedure followed to determine compressive concrete, buckling of longitudinal rein-
the compressive strength. Global buckling vulnerability forcement, and global buckling of element. Although
is judge using Eq. (4), and reinforcement buckling is load carrying capacity between monotonic and cyclic
assessed based on the longitudinal-to-transverse rein- loadings showed no significant difference, loading type
forcement index and the ratio of transverse reinforcement may lead to different final failure mode.
spacing to longitudinal bar diameter, buckling model is Dense transverse reinforcement detailing in thin con-
considered to determine the compressive strength. sNmax fined boundaries did not improve the performance of
and sbNmax are simulated peak load without and with walls. Imposing a minimum wall thickness would be an
buckling effect, respectively. Figures 16 and 17 shows alternative means to suppress failures due to global
comparison between experimental and analytical axial buckling of thin walls and efficiently use the confine-
load - axial strain for specimens tested under cyclic and ment. It was also shown that failure due to global buck-
monotonic loading, respectively. Table 5 compares ling is affected by both large tensile strain prior to com-
compressive strength point and corresponding strain. pressive strain and prior crushing of compressive con-
The analytical model for cyclically tested specimens crete.
captures reasonably well the measured response with Large transverse reinforcement spacing may result in
hysteretic shape of the response and stiffness degradation. buckling of longitudinal reinforcement following even
The cyclic properties of the implemented analytical limited tensile strain excursions. Intermediate unsup-
stress-strain relations for steel and concrete produce good ported bars are more susceptible to buckling. Supporting
correlation for global response, and was also able to all intermediate bars at the wall confined edge should be
predict the compressive strength reduction for specimens considered. Comparison between monotonic and cyclic
failing due to reinforcing bar buckling. For specimens response for elements with failure mode governed by
tested under monotonic loading, although compressive buckling of longitudinal reinforcement showed that prior
strength was overestimated especially for specimens tensile strain reduced considerably the load level at the
failing due to concrete crushing and global buckling, the onset of bars buckling. Transverse reinforcement index
model well simulated the post peak branch for specimens could be used as a design method of to prevent buckling
failing due to reinforcement buckling. of longitudinal bar. A design index was also proposed
that take into account the case of large longitudinal to
6. Conclusions transverse reinforcing bar diameters.
The proposed longitudinal-to-transverse reinforce-
An experimental study was conducted on sixteen RC ment index along with the ratio of transverse reinforce-
rectangular columns that idealize confined boundaries of ment spacing to longitudinal bar diameter present a
RC rectangular walls to examine the effects of slender- simple but effective anti-buckling measures of rein-
ness, reinforcement detailing and loading type on their forcement.
performance under monotonic and cyclic reversed axial An analytical model that include bar buckling was
R. Taleb, M. Tani and S. Kono / Journal of Advanced Concrete Technology Vol. 14, 108-124, 2016 122

Fig. 16 Measured and predicted cyclic axial load - axial strain relations.

proposed to predict cyclic response of tested specimens. References

The analytical model captures reasonably well the AIJ, (2010). “Standard for structural calculation of
measured response and was also able to predict the reinforced concrete structures.” Tokyo: Architectural
compressive strength reduction for specimens failing due Institute of Japan.
to reinforcing bar buckling. ACI, (2014). “Building code requirements for structural
concrete (ACI 318-14) and commentary
Acknowledgement (ACI318R-14).” American Concrete Institute, MI,
The authors gratefully acknowledge the financial support USA.
of the Ministry of Land, Infrastructure, Transportation Chai, Y. H. and Elayer, D. T., (1999). “Lateral stability of
and Tourism (Japan). The first author acknowledge the reinforced concrete columns under axial reversed
financial support of the Ministry of Education, Culture, cyclic tension and compression.” ACI Structural
Sports, Science and Technology (Japan). The efforts of Journal, 96(5), 780-790.
Mr. E. Yuniarsyah and M. Ogura graduate students at Chai, Y. H. and Kunnath, S. K., (2005). “Minimum
Tokyo Institute of technology (Kono & Shinohara Re- thickness for ductile RC structural walls.” Engineering
search Group) who contributed in the experimental tests Structures, 27(7), 1052-1063.
are highly appreciated. Chrysanidis, T. A. and Tegos, I. A., (2012). “The
influence of tension strain of wall ends to their
R. Taleb, M. Tani and S. Kono / Journal of Advanced Concrete Technology Vol. 14, 108-124, 2016 123

Fig. 17 Measured and predicted monotonic axial load - axial strain relations.

resistance against lateral instability for low-reinforced Kato, H., Tajiri, S. and Mukai, T., (2010). “Preliminary
concrete walls.” Proceedings of the 15th World reconnaissance report of the chile earthquake 2010.”
Conference on Earthquake Engineering, Lisbon 24-28 Building Research Institute, Japan.
September 2012. Markeset, G. and Hillerborg, A., (1995). “Softening of
Dhakal, R. P. and Maekawa, K., (2002). “Path-dependent concrete in compression localization and size effects.”
cyclic stress-strain relationship of reinforcing bar Cement and Concrete Research, 25(4), 702-708.
including buckling.” Engineering Structures, 24(11), Massone, L. M., Polanco, P. and Herrera, P., (2014).
1383-1396. “Experimental and analytical response of RC wall
Filippou, F. C., Popov, E. G. and Bertero, V. V., (1983). boundary elements.” Proceedings of the 10th U.S.
“Effects of bond deterioration on hysteretic behavior National Conference on Earthquake Engineering,
of reinforced concrete joints.” EERC Report No. Anchorage 21-25 July 2014.
UCB/EERC-83/19, Earthquake Engineering Research Menegotto, M. and Pinto, E., (1973). “Method of
Center, University of California, Berkeley. analysis for cyclically loaded reinforced concrete
Kato, D., Kanaya, J., Wakatsuki, K. and Honda, Y., plane frames including changes in geometry and
(1995). “Design method to prevent buckling of main non-elastic behavior of elements under combined
bars in RC members.” Pacific Conference on normal force and bending.” Proceedings of IABSE
Earthquake Engineering, Melbourne 20-22 Symposium on Resistance and Ultimate Deformability
November. of Structures Acted on by Well-Defined Repeated
R. Taleb, M. Tani and S. Kono / Journal of Advanced Concrete Technology Vol. 14, 108-124, 2016 124

Loads, Lisbon, 1973, pp. 15-20. Telleen, K., Maffei, J., Heintz, J. and Dragovich, J.,
Moehle, J. P., Acevedo, C. and Creagh, A., (2011). (2012). “Practical lessons for concrete wall design,
“Exploratory tests of wall boundary elements based on studies of the 2010 Chile earthquake.”
subjected to alternating tensile and compressive Proceedings of the 15th World Conference on
loadings.” Poster and oral presentations at the 2010 Earthquake Engineering, Lisbon 24-28 September
PEER Annual Meeting. 2012.
Moehle, J., Berger, J., Bray, J., Dengler, L., Greene, M., Scott, B. D., Park, R. and Priestley, M. J. N., (1982).
Mitrani-Reiser, J. and Siembieda, W., (2010). “The 27 “Stress-strain behavior of concrete confined by
February 2010 central south Chile earthquake: overlapping hoops at low and high strain rates.”
emerging research needs and opportunities.” Journal of the American Concrete Institute, 79(1),
Earthquake Engineering Research Institute. 13-27.
Parra, P. F. and Moehle, J. P., (2014). “Lateral buckling in Wallace, J. W., (2012). “Behavior, design, and modeling
reinforced concrete walls.” Proceedings of the 10th U.S. of structural walls and coupling beams - lessons from
National Conference on Earthquake Engineering, recent laboratory tests and earthquakes.” International
Anchorage 21-25 July 2014. Journal of Concrete Structures and Materials, 6(1),
Paulay, T., (1986). “The design of ductile reinforced 3-18.
concrete shear walls for earthquake resistance.” Wallace, J. W., Massone, L. M., Bonelli, P., Dragovich, J.,
Earthquake Spectra, (4), 783-823. Lagos, R., Lüders, C. and Moehle, J., (2012).
Paulay, T. and Goodsir, W. J., (1985). “Ductility of “Damage and implications for seismic design of RC
structural walls.” Bulletin of the New Zealand structural wall buildings.” Earthquake Spectra, 28(S1),
National Society for Earthquake Engineering, 18(3), S281-S299.
250-269. Wallace, J. W. and Moehle, J. P., (1992). “Ductility and
Paulay, T. and Priestley, M. J. N., (1993). “Stability of detailing requirements of bearing wall buildings.”
ductile structural walls.” ACI Structural Journal, 90(4), ASCE Journal of Structural Engineering, 116(6),
385-392. 1625-1644.
Rodriguez, M. E, Botero, J. C. and Villa, J., (1999), Westenenk, B., Llera, J. C., Besa, J. J., Jünemann, R.,
“Cyclic stress-strain behavior of reinforcing steel Moehle, J., Lüders, C., Inaudi, J. A., Elwood, K. J. and
including effect of buckling.” ASCE Journal of Hwang, S. J., (2010). “Response of reinforced
Structural Engineering, 125(6), 605-612. concrete buildings in concepción during the Maule
Taleb, R., Kono S., Tani, M. and Sakashita, M., (2014). earthquake.” Earthquake Spectra, 28(S1), S257-S280.
“Effects of end regions confinement on seismic Wood, S. L., Wight, J. K. and Moehle, J. P., (1987). “The
performance of RC cantilever walls.” Proceedings of 1985 Chile earthquake observations on
the 10th U.S. National Conference on Earthquake earthquake-resistant construction in Vina del Mar.”
Engineering, Anchorage 21-25 July 2014. Structural Research Series No. 532, University of
Takahashi, S., Yoshida, K., Ichinose, T., Sanada, Y., Illinois.
Matsumoto, K., Fukuyama, H. and Suwada, H., (2013). Yassin, M. H. M., (1994). “Nonlinear analysis of
“Flexural drift capacity of reinforced concrete wall prestressed concrete structures under monotonic and
with limited confinement.” ACI Structural Journal, cyclic loads.” Thesis (PhD), University of California,
110(1), 95-104. Berkeley.