Construction and Building Materials 102 (2016) 893–903

Contents lists available at ScienceDirect

Construction and Building Materials

journal homepage: www.elsevier.com/locate/conbuildmat

A fracture mechanics-based approach to modeling the confinement

effect in reinforced concrete columns
Nguyen Duc Tung ⇑, Nguyen Viet Tue
Institute for Structural Concrete, Graz University of Technology, A-8010 Graz, Austria

h i g h l i g h t s

The effective stress in the transverse reinforcement at peak stress is estimated.

A relation between the crack formation, the fracture energy in compression, and the lateral deformation is established.

The stress and strain behavior of the confined concrete core is mechanically determined.

The comparison between test results and predictions is made.

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

Article history: It is well-known that the failure of concrete in compression is characterized by the localization of
Received 6 August 2015 deformations. As a consequence, the failure of a reinforced concrete column localizes, in its turn in the
Accepted 12 November 2015 same localized fracture zone in which passive confinement provided by the transverse reinforcement
Available online 21 November 2015
can be locally activated due to the significant expansion of the concrete in the lateral direction. By this
means, fracture mechanics-based approaches may be considered as applicable for the modeling of this
Keywords: effect. This paper presents a new approach to modeling of the effect of confinement in reinforced concrete
Confined concrete
columns through observations of the crack formation in the presence of a confining stress. The effective
Lateral strain
Transverse reinforcement
stress in the transverse reinforcement at peak stress is estimated by taking the compatibility of deforma-
Effective confining stress tions between the confined concrete core and the transverse reinforcement into consideration.
Deformation compatibility Furthermore the strain behavior of the confined concrete core is determined by establishing a relation
Compressive strength between the crack formation, which is represented by the fracture energy in compression, and the lateral
Ductility deformation. The comparison between test results and predictions, by the means of the confined com-
pressive strength with corresponding strain and the total stress–strain curve, shows that the proposed
approach yields a reasonable level of accuracy.
Ó 2015 Elsevier Ltd. All rights reserved.

1. Introduction modeling of the triaxial compressive strength. Many complicated

Research on the behavior of concrete under an active confine-
ment originated in the early 20th Century in the work of Richart
et al. [1]. In this pioneering work a linear relation between the
lateral stress (rl) and the triaxial compressive strength (fcc) was
introduced as
f cc ¼ f c þ k  rl
where the constant k was determined from tests to be 4.1, fc is the
compressive strength of unconfined concrete.
Since that time enormous efforts have been made both to
comprehend this phenomenon and also to achieve more accurate
⇑ Corresponding author.
E-mail address: n.tung@tugraz.at (N.D. Tung).
formulations for compressive meridian of the failure surface have
been proposed in the modeling context during the course of the
research history and recognized to more accurately predict the
triaxial compressive strength. Nevertheless, the formulation pro-
posed by Richart et al. [1] given above is still considered as suitable
for representing the confining effect and has been adopted in some
recent research works by modification of the constant k, e.g. [2–4].
ð1Þ
It is also an attempt to explain the confining effect through observ-
ing the crack formation of cylinders in a triaxial test, e.g. [5].
Instead of unstable distributed cracking, crack formation is more
stable with finer micro-cracking, when the lateral stress increases.
This leads not only to an increase of the axial load-carrying capac-
ity and the corresponding axial strain, but also to higher lateral
strain at peak stress [6].

http://dx.doi.org/10.1016/j.conbuildmat.2015.11.031
0950-0618/Ó 2015 Elsevier Ltd. All rights reserved.
894 N.D. Tung, N.V. Tue / Construction and Building Materials 102 (2016) 893–903

Notations

Ac area of specimen cross-section Wpost post-peak portion of compressive fracture energy per
Acc area of core of section enclosed by the center lines of the unit volume in the initial fracture zone of unconfined
perimeter spiral or hoop concrete
Apre pre-peak portion of compressive fracture energy per W’post post-peak portion of compressive fracture energy per
unit specimen area unit volume in the fracture zone of confined concrete
Apost post-peak portion of compressive fracture energy per dinel post-peak inelastic deformation
unit specimen area ec longitudinal strain of concrete
Asx,y areas of transverse reinforcement in x- and y-directions ecc longitudinal strain of confined concrete at peak stress
cx,y widths of confined concrete core of a rectangular rein- (fcc)
forced concrete column (cx > cy) eccl lateral strain of concrete of confined concrete at peak
Cx,y widths of a rectangular reinforced concrete column stress
(Cx > Cy) ecl lateral strain of concrete
dc diameter of confined concrete core of a circular rein- eco strain at peak stress of unconfined concrete
forced concrete column ecol lateral strains at the compressive strength of unconfined
D diameter of a circular reinforced concrete column concrete
Ec modulus of elasticity of concrete ehcc strain in the transverse reinforcement at peak stress of
Es modulus of elasticity of reinforcing steel confined concrete
fc compressive strength of concrete einel,i post-peak inelastic strain of the initial fracture zone of
fcc compressive strength of confined concrete unconfined concrete
fco in-place compressive strength of unconfined concrete e0 inel post-peak inelastic strain of the fracture zone of con-
fcu axial stress of confined concrete core at complete failure fined concrete core
fyh yield limit of transverse reinforcement qs geometric reinforcement ratio of longitudinal reinforce-
Ie confinement index Ie = rl/fc ment
Ke geometric coefficient of confinement effectiveness qsh geometric reinforcement ratio of transverse reinforce-
Ld length of the localized fracture zone of a concrete spec- ment
imen rc concrete stress
Ld,i length of the initial fracture zone of a concrete specimen rhcc stress in the transverse reinforcement at peak stress of
s clear spacing of ties or spirals confined concrete
wi clear spacing between longitudinal reinforcing bars rl confining (lateral) stress in an active confinement
Wpre pre-peak portion of compressive fracture energy per rle effective confining stress provide by transverse rein-
unit volume in the initial fracture zone of unconfined forcement in columns
concrete h inclination angle of the shear failure plane
W0 pre pre-peak portion of compressive fracture energy per
unit volume in the fracture zone of confined concrete

The effect of confinement can also be obtained in reinforced simply from regression analyses [18–21]. This inconvenience is
concrete columns through the lateral pressure provided by the caused by the fact that the concrete lateral strain in these models
transverse reinforcement arranged in the form of ties or spirals, is formulated as a function of the axial strain, which is unknown at
as demonstrated in Fig. 1. In contrast to the behavior of a concrete the beginning of the calculation.
specimen in the triaxial compression test, the confinement in rein- The compressive strength of confined concrete can be well
forced concrete columns is a passive phenomenon. Apart from the estimated in most existing models with the effective stress in the
influence of the column geometry and reinforcement arrangement, transverse reinforcement [16–20,22]. Some efforts have been made
which can be well captured using a geometric coefficient of con- to describe the ductile behavior of confined concrete in reinforced
finement effectiveness proposed by Mander et al. [7], the passive concrete members on the basis of fracture mechanics in modeling
confining stress in reinforced concrete columns depends decisively
on the interaction between the amount of the transverse reinforce-
ment and deformation capacity of confined concrete in the lateral
direction. It has been experimentally observed for normal-strength
concrete columns confined with normal-strength steel ties that the
stress in transverse reinforcement can generally reach its yield
limit at the peak stress of confined concrete [8]. By contrast, the
yielding limit of the transverse reinforcement can only be reached
at peak stress of confined concrete for high-strength concrete col-
umns employing a relatively large amount of transverse reinforce-
ment [9–11]. It thus becomes clear that an assumption of the
confining reinforcement yielding at the peak stress of confined
concrete previously adopted in some previous research works
[7,12–14] as well as in fib MC 2010 [15] can lead to an unsafe
design, especially for high-strength concrete columns tied by
high-strength steel. In order to avoid this, efforts have been made
to determine the effective stress in the transverse reinforcement
using complex iterative procedures [16,17] or formulations derived Fig. 1. Spirals and ties as confining reinforcement in reinforced concrete columns.
 N.D. Tung, N.V. Tue / Construction and Building Materials 102 (2016) 893–903
the confined concrete ductility, in addition to the empirical
approaches that are commonly employed [16,18,19,23–27].
Akiyama et al. [21] experimentally examined the compressive frac-
ture energy and the length of the localized fracture zone of con-
fined concrete columns, in which local stress–strain curves of
different locations were measured by strain gages embedded at
acrylic resin bars. These parameters were used to model the
post-peak regime of the average stress–strain curve of columns,
while formulations obtained from regression analysis that very
similar to the formulations proposed by Cusson and Paultre [16]
were proposed for the confined compressive strength and the axial
strain at peak.
From the other starting point and having considered the stress–
strain relation of the localized fracture zone proposed by Markeset
[28] as the material model for unconfined concrete, Meyer [29]
and Sint [30] added the energy portion released by the transverse
reinforcement to explain the ductility of confined concrete. Because
this energy portion is relatively small, the effect of confinement on
the concrete ductility described in these models is low when com-
pared to the test data, especially for well-confined specimens. The
confinement effect by the means of favoring a more stable crack for-
mation, leading to higher load carrying capacity and ductility
observed in triaxial tests, cannot be captured in such an approach.
In addition, attention should be paid to the fact that the projection
on the specimen axis of an inclined shear band developed after the
complete failure of a concrete specimen was taken as the length of
the localized fracture zone by Markeset [28], and was considered
between 2.5 and 3 times the smaller width of a specimen. With
advanced measuring techniques, [31–33] it has been experimentally
observed that the localized fracture zone develops in a relatively
narrow region at peak stress and moreover that this zone continu-
ously extends during the softening process to reach its final length
that approximately equals the length observed after the complete
failure. Considering the fact that passive confinement is only locally
activated in the localized fracture zone developed at peak stress due
to the significant expansion of concrete in the lateral direction, it
may well be more suitable to take the stress–strain relation of the
localized fracture zone developed at peak stress into account as
the material model for unconfined concrete, rather than that of the
final fracture zone at complete failure. In this context, the extraction
of the stress–strain relation for the localized fracture zone devel-
oped at peak stress from the total stress–displacement curves of
specimens is necessary [34,35].
This paper presents a new perspective toward the modeling of
the confinement in reinforced concrete members based on obser-
vations of the concrete crack formation in presence of a confining
stress as well as the relation between the crack formation and
the lateral deformation. In the modeling the differences in the
crack formation and the lateral deformation between unconfined
and confined concrete are accounted by the corresponding fracture
energies in compression. Taking the deformation compatibility
between confined concrete core in lateral direction and the elonga-
tion of the transverse reinforcement, it should be possible to over-
come the difficulties referred above in determining the effective
stress in the transverse reinforcement. The recent development is
limited to the modeling of the effect of confinement in reinforced
columns subjected to concentric loading.
2. Material model for unconfined concrete
It is generally recognized that the failure of a concrete specimen
in compression is characterized by the localization of deformations
in the so-called localized fracture zone. Experimental investiga-
tions [31,33,36] have shown that this zone does not remain con-
stant, but gradually extends during the softening process to
895
Fig. 2. Relation between the length of the initial fracture zone and the specimen
size according to Lersrisakulrat et al. [31].
reach its final length at the complete failure. Tung and Tue
[34,35] named the localized fracture zone initiated at the peak
stress as the initial fracture zone. The length of the initial fracture
zone (Ld,i) is observed by Lersrisakulrat et al. [31] to be dependent
on the shape and size of the specimen, as shown in Fig. 2. For spec-
pffiffiffiffiffi
imens having the root of cross-sectional area Ac P 180 mm, Ld,i
pffiffiffiffiffi
can be taken at 0.6 Ac . The final length of the localized fracture
zone (Ld) at the complete failure is observed to be approximately
three times the smallest width of the specimen (Ld = 3D).
During the softening process, the displacement in the actual
localized fracture zone must continue to soften, while unloading
occurs outside this zone. When the cracks grow to a location,
which still lies outside the localized fracture zone formed at the
peak stress, the lateral strain at this location also increases due
to the crack opening. An increase of the longitudinal strain at this
location is expected, because of the natural relation existing
between the lateral and the longitudinal strains. Thus the strain
behavior of this location changes from unloading to softening.
The strain behavior in the post-peak regime of a point in a speci-
men depends on its location and is then determined by one of
the following laws: (a) softening; (b) elastic unloading; and (c) a
combination of unloading and softening. The stress–displacement
relation of the whole specimen is then the integral of the stress–
strain relations of all points in it.
Considering the deformation localization, the post-peak com-
pressive fracture energy per unit specimen area (Apost) can be
introduced. This energy corresponds to the area under the
stress–inelastic displacement curve, as illustrated in Fig. 3(a), and
can be calculated as
Z f co =3
f co max
Apost ¼ dinel  dr þ  dinel ð2Þ
f co 3
where fco is unconfined compressive strength; dinel is the post-peak
inelastic displacement.
Assuming that the softening law of all points in the fracture
zone is identical and can be described by a linear relation; the dif-
ference between specific points is determined by the beginning of
the change from elastic unloading to softening. Therefore, the
inelastic strain of the initial fracture zone (einel,i) can be extracted
from the total stress–displacement curve of the entire specimen
through its maximum value developed at the minimum stress
1/3 fco by solving Eq. (2) as
Apost
inel;i ¼
emax ð3Þ
ð18
5
Ld þ 18
7
Ld;i Þ  f co
From a regression analysis of experimental results of specimens
with L/D P 3 performed by Jansen and Shah [37], Rokugo and Koy-
anagi [38], Lersrisakulsat et al. [31], and Watanabe et al. [32], as
shown in Fig. 4, the following relation for Apost is obtained
896 N.D. Tung, N.V. Tue / Construction and Building Materials 102 (2016) 893–903

Fig. 3. Behavior of a concrete specimen in compression: (a) total stress–displacement curve; (b) stress–strain curves (axial and lateral strains) of the initial fracture zone.

1 This energy portion corresponds to the energy dissipated by

Jansen/Shah (1997) loading the specimen up to the peak stress and then unloading it
Rokugo/Koyanagi (1992)
completely. Depending on the compressive strength class, loading
0.8 Lertsrisakulrat et al. (2001)
Watanabe et al. (2004)
up to the stress level r1 of approx. (0.6  0.8) fc and then unloading
dissipates no significant energy. This stress level also corresponds
0.6 to the stress limit according to fib MC 2010 [15] where the lateral
Apost / fc

strain can be estimated directly from the axial strain through Pois-
0.4 son’ ratio mc, which can generally be taken to be 0.2. The stress level
r1 is lower for normal-strength concrete than for high-strength
y=8.474.x -0.7752 concrete. It is generally possible to select this stress as 0.7 fco.
0.2
R2=0.782 When the compressive stress is in excess of this stress level, sub-
stantial crack formation is considered to take place. This is also evi-
0 dent by the disproportional increase of the lateral strain with an
10 30 50 70 90
increase of the axial strain [41]. The pre-peak energy is, thus, con-
Compressive strength fc (N/mm 2)
sidered to be the necessary energy for crack formation that associ-
Fig. 4. Compressive fracture energy versus compressive strength. ates with an increase of the lateral strain. This energy, thus,
corresponds to the increase of concrete lateral strain when the con-
crete axial stress increases from the stress r1 to the concrete com-
Apost ¼ 8:5f co
0:225
ðN=mmÞ ð4Þ pressive strength fco, as can be demonstrated by the following
expression
Fig. 3(b) shows the proposed stress–strain curve for the initial
fracture zone, which can be considered as the material model for
W pre ¼ K  ðecol  e1l Þ  f co ð8Þ
unconfined concrete. The stress–strain relation proposed by Sargin where e1l is the lateral strain at stress r1, e1l = mc  e1. The corre-
et al. [39] followed by many other researchers and thus included in sponding axial strain e1 at stress r1 is calculated using the stress–
some codes of practice in [15,40] is adopted for the pre-peak part strain relation given in Eq. (5) and is found for different concrete
of the stress–strain curve strength classes to be approx. 0.5 eco.
  ecol is the lateral strains at the compressive strength and can be
k  g  g2
rc ¼  f co ð5Þ taken to be 0.5 eco; K is a proportional factor presenting a relation
1 þ ðk  2Þ  g between the dissipated fracture energy and the increase of con-
crete lateral strain. This factor is considered to be constant for both
where g = ec/eco; eco is the strain at peak stress; and k = Ec  eco/fco.
unconfined and confined concrete. It is not necessary to determine
The proposed values for the modulus of elasticity Ec and the strain
the value of K, since this temporary parameter will be disappeared
at the peak stress eco included in fib MC 2010 [15] can be also
when calculating the fracture energy of confined concrete.
adopted.
The post-peak fracture energy per unit volume of the initial frac-
The post-peak part of the stress–strain curve is characterized by
ture zone corresponds to the area under the stress–inelastic strain
the linear relation, whose slope is controlled by the maximum
curve and can be calculated as
inelastic strain in the initial fracture zone that can be written as
! 2
W post ¼ f  emax ð9Þ
ðec  eco Þ  Ec 3 co inel;i
rc ¼ 1  f co ð6Þ
3=2Ec  emax
inel;i  f co The stress–strain relation of the initial fracture zone together
with the fracture characteristics are considered as the material
A definition for the fracture energy per unit volume of the initial model for unconfined concrete by the modeling the effect of
fracture zone can be introduced based on the assumption made confinement.
above that the stress–strain law of all points in the fracture zone
is identical. The pre-peak energy per unit volume in the damage
3. Compatibility of lateral deformations in reinforced concrete
zone (Wpre) can be calculated using the following expression
columns
 
f co
W pre ¼ ac  eco   f co ð7Þ 3.1. Lateral strain of confined concrete
Ec

where ac is the solidity factor of the stress–strain curves described It is observed experimentally that a confinement can have a
in Eq. (5). positive influence on the lateral strain at peak stress by the means
 N.D. Tung, N.V. Tue / Construction and Building Materials 102 (2016) 893–903 897

of favoring the crack formation. Based on the experimental results Nagashima et al. (1992)
provided by Imran and Pantazopoulou [6], Montoya et al. [42] pro- Hong et al. (2006)
posed a simple parabolic relation between the lateral strain ecl and 1
the normalized longitudinal strain ec/ecc, as described by the fol-

(σhcc /f yh )exp
lowing expression
 2
ec
ecl ¼ ð0:0019 þ 0:0242Ie Þ  ð10Þ
ecc 0.5

It should be noted that in the specimens tested by Imran and

Pantazoloulou [6], the confinement index Ie = rl/fc was varied in a
broad range from 0 to 0.9. Since the effective confinement stress
provided by the transverse reinforcement in reinforced concrete 0
columns is generally not greater than approx. 0.3 fc, this formula- 0 0.5 1
tion should be re-verified with specimens confined by moderate (σhcc /fyh )pred
lateral stresses for possible application to structural members.
Fig. 6. Experimental versus predicted stress in transverse reinforcement at
For moderate lateral stresses, the lateral strain according to Eq. maximum loading.
(10) is found to overestimate the lateral strain of specimens made
of high-strength concrete, while it underestimates the lateral 3.2. Strain in transverse steel at peak stress of concrete core
strain of ones with normal-strength concrete [34]. This result
appears to reflect the fact that high-strength concrete exhibits less Taking into account the compatibility of deformations between
deformation capacity in the lateral direction than normal-strength the confined concrete core in the lateral direction and the trans-
concrete and suggests that the influence of the compressive verse reinforcement (eccl = ehcc), the strain in transverse steel at
strength on the lateral strain should be given sufficient the maximum load (ehcc) can be calculated using Eq. (11) by replac-
consideration. ing fc with fco, ec/ecc = 1 and Ie = rle/fco as
Taking the influence of the compressive strength into consider-
rle
ation, a modified relation is proposed for estimating the lateral ehcc ¼ eccl ¼ 0:00176 þ 0:77 1:85
ð12Þ
strain of confined concrete f co
 2 where the in-place compressive strength of unconfined concrete fco
ec
ecl ¼ ð0:00176 þ 0:77Ie  f c0:85 Þ  ð11Þ is taken to be 0.85 fc.
ecc Considering a geometric coefficient of confinement effective-
In the same manner as in the formulation proposed by Montoya ness Ke, the effective confinement stress in the concrete core at
et al. [42], the lateral strain of concrete at confined compressive maximum load (rle) can be calculated from the stress in the trans-
strength eccl according to Eq. (11) can be directly estimated from verse reinforcement (rhcc) and the geometric ratio of reinforce-
the uniaxial compressive strength fc and the confinement index ment qsh = (Asx + Asy)/(cx + cx)/s using the following expression
Ie = rl/fc, and is independent of the corresponding axial strain ecc, rle ¼ K e  qsh  rhcc ð13Þ
which is not defined at the beginning of the design. This formula-
tion, thus, has potential capability for modeling the passive con- Substituting Eq. (13) into Eq. (12), one obtains
finement in reinforced concrete members without an iterative rhcc
procedure. ehcc ¼ 0:00176 þ 0:77K e  qsh  1:85
ð14Þ
f co
The proposed formulation was evaluated using test results of 40
cylinders under triaxial compression tested by Imran and Panta- The widely accepted expression for the geometric confinement
zopoulou [6], Candappa et al. [3,43], and Dahl [44] The concrete effectiveness coefficient Ke proposed by Mander et al. [7], as given
compressive strength fc and the confinement index Ie were in the in Eq. (15) is adopted in this study
range of 28.6–108.8 MPa and 0–0.35, respectively.  
K e ¼ 1  2ds c =ð1  qs Þ for circular columns with spirals
A comparison between experimental and predicted lateral    
Rw2i
strains of concrete at confined compressive strength eccl using K e ¼ 1  6cx cy 1  2csx 1  2csy =ð1  qs Þ for rectangular columns with ties
the proposed formulation is shown in Fig. 5. It can be observed that ð15Þ
the predicted lateral strains are in good agreement with the test
data. where wi is the clear distance between adjacent laterally supported
longitudinal bars; s is the clear vertical spacing between ties; dc, cx
20 and cy are the core dimensions measured to the centerline of the
Imran/Pantazopoulou (1996) ties, as illustrated in Fig. 1; and qs is the ratio of the area of longi-
Candappa et al. (2001)
Dahl (1992) tudinal reinforcement to the area of the confined core.
15 Assuming that the stress–strain relation of confining reinforce-
ment is ideal elasto-plastic, the stress and strain of reinforcement
at maximum load can be calculated through solving Eq. (14) as
10
0:00176
ehcc ¼ ; rhcc ¼ Es  ehcc ð16Þ
1  0:77K e  qsh  f 1:85
Es
co
5
If the calculated strain according to Eq. (16) exceeds the yield-
ing limit of the confining reinforcement (rhcc P ryh), this equation
0 is simplified as
0 5 10 15 20
f yh
ehcc ¼ 0:00176 þ 0:77K e  qsh  1:85
; rhcc ¼ f yh ð17Þ
f co
Fig. 5. Comparison of lateral strain eccl between tests and prediction.
898 N.D. Tung, N.V. Tue / Construction and Building Materials 102 (2016) 893–903

It is interesting to note that the calculated strain in the trans- confinement in graph form. Since confinement in reinforced con-
verse reinforcement at maximum load according to Eqs. (16) and crete columns is a passive phenomenon, the lateral stress provided
(17) depends on the amount and yield limit of the transverse rein- by transverse reinforcement can only be substantially activated
forcement (qsh and fyh), the compressive strength of concrete fco, when the lateral strain starts to increase disproportionally to the
and the confinement effectiveness coefficient Ke. For members axial strain (above stress r1). With more stable crack formation,
made of high-strength concrete, larger amounts of transverse rein- more microcracks can be formed when the stress in the concrete
forcement are required than for members made of normal-strength reaches the confined compressive strength, leading to larger lateral
concrete to obtain an effective use of the transverse reinforcement. strain at the peak stress. Similarly to Eq. (8), the pre-peak energy
A comparison between experimental results and predicted dissipated by a unit volume of confined concrete in the damage
stress in transverse reinforcement at maximum stress of the con- zone W0 pre can also be represented by the increase of the lateral
fined concrete core using Eqs. (16) and (17) for 45 high-strength strain when the concrete axial stress increases from stress r1 to
concrete columns tied by high-strength steel (fyh from the peak stress fcc (from e1l to eccl).
807–1420 MPa) performed by Nagashima et al. [9] and Hong
W 0pre ¼ K  ðeccl  e1l Þ  f cc ð18Þ
et al. [45] is shown in Fig. 6.
It can be observed that the predicted stresses in the reinforce- Substituting K, Wpre from Eqs. (7), (8) into Eq. (18), yields
ment are in good agreement with the test data. Especially satisfac-  
f co eccl  e1l
tory predictions can also be obtained for columns with poorly W 0pre ¼ ac  eco   f ð19Þ
activated transverse reinforcement (the ratio rhcc/fyh in range of Ec ecol  e1l cc
0.2–0.3). This could imply that the use of the yield limit when Taking into consideration a solidity factor of the stress–strain
determining the confinement effect can lead to unsafe predictions curves of confined concrete acc, the pre-peak energy per unit vol-
for these columns. ume in the damage zone of confined concrete can be described as
 
f cc
4. Enhancement in strength and ductility W 0pre ¼ acc  ecc   f cc ð20Þ
Ec
With the lateral stress provided by transverse reinforcement, Inserting Eq. (19) into Eq. (20) and solving this equation after
crack formation of the confined concrete core is more stable than ecc, one obtains
that of unconfined concrete. The compressive strength and the  
ac f eccl  e1l f cc
corresponding axial strain of the confined concrete core are, thus, ecc ¼  eco  co  þ ð21Þ
enhanced. Taking into consideration the dissipated energy for the
acc Ec ecol  e1l Ec
crack formation and propagation for unconfined and confined
concrete, the strength enhancement and ductility of the concrete
core are then described in this Section.

4.1. Confined concrete strength

With the effective confinement stress applied to the concrete

core rle, the triaxial compressive strength fcc can be calculated
using the linear relation proposed by Richart et al. [1], as given
in Eq. (1) by replacing fc with fco, rl with rle. The constant k is taken
to be 4.0, as suggested by Lu and Hsu [4] for an active confinement
and by Sigrist [14] for a passive confinement. This value is a little
lower than the proposed value by Richart et al. (at 4.1).

4.2. Axial strain at peak stress

Fig. 7 presents the stress–strain relations for concrete in com-

Fig. 8. Enhancement in strength and ductility.
pression in both axial and lateral directions with and without a

Fig. 7. Typical stress–strain relations for unconfined and confined concrete.

 N.D. Tung, N.V. Tue / Construction and Building Materials 102 (2016) 893–903 899

The solidity factors of the stress–strain curves of concrete in

MC 2010[15]
uniaxial and triaxial compression depend on the concrete com-
pressive strength fc and the confinement index Ie and are in range
between 0.8 and 0.9 [34]. The ratio ac/acc varies for different con-

11.58
0.91
0.11
0.98

0.85
0.85
0.87

0.97
0.96

0.94
1.04
1.02

0.80
0.90
0.90
0.80
crete strengths and confining grades between 0.9 and 1.0. For the
sake of simplicity, the ratio ac/acc can be taken to be 1.0.

Légeron[19]
4.3. Post-peak behavior

10.66
0.86

0.11
0.97

0.95
0.88
1.10
1.09
1.07

1.04

1.05
1.08

1.08

1.02
1.07

1.00
As presented in previous studies on the behavior of concrete in
compression [35], the post-peak behavior is considered to be
characterized by the crack opening and propagation of unstable

Mander et al.[7]
macro-cracks during the softening. Since the number of cracks at
the peak stress of concrete is much greater for confined concrete
(that can be represented by the lateral strain increase eccl  e1l )

0.81
0.97

0.82
0.95

0.92
0.11
0.99

0.85
0.85
0.88
0.91
0.91

12.13
1.04
1.02

1.00
than for unconfined concrete (represented by ecol  e1l ), the energy
dissipated for crack opening and propagation in the post-peak
regime of confined concrete is then expected to be higher than that

Proposed method
of unconfined concrete. Assuming that the dissipated post-peak
energy per unit volume of confined concrete in the damage zone

fcc,exp/fcc,pred
(W0 post) is proportional to the lateral strain increase and compres-
sive strength, this parameter can be estimated from the corre-

1.16
1.14

1.11
0.89

0.11
0.94
0.97

0.99

1.08

1.03
1.09

1.06

1.02

1.08
0.90

11.06
sponding value of unconfined concrete by the following expression
eccl  e1l f cc
W 0post ¼  W ð22Þ
ecol  e1l f co post

0.25–1.13
0.21–2.81
0.31–1.19
0.68–1.52

0.71–1.85

0.42–2.63
0.43–1.58
0.73–2.48
0.58–2.81
0.49–2.29
0.21–1.53
0.82–2.55
0.30–1.26

0.30–1.22
The post-peak energy of confined concrete W0 post is considered
qsh (%)
as the dissipated energy from the peak stress fcc to the complete
failure. Through observing normal-strength concrete columns
Scott et al. [46] and Mander et al. [47] considered that the ultimate
failure is characterized by the first hoop fracture. By contrast,
eventual failure of high-strength concrete columns is observed
445–1318
445–1318

279–1420
255–1420
807–1368

400–1000
400–1000
fyh (MPa)

452–629
255–589
296–309

392–770
508–666
310–340

330–500
by Cusson and Paultre [16] and Hong et al. [45] by the means of
formation of an inclined shear failure plane at which a hoop frac-
ture may still not occur. Taking the later observation as represen-
tative for the complete failure of both normal- and high-strength
concrete columns, the axial stress in the concrete core correspond-
52.5–115.9

52.4–102.9

31.3–124.0
60.4–118.1

82.0–105.6
60.0–124.0
60.0–124.0
41.1–42.4
24.2–25.3

41.4–97.1
61.2–97.1
31.3–40.9

27.0–32.0

40.4–116
ing to the complete formation of the shear failure plane proposed
fc (MPa)

by Cusson and Paultre [16] is adopted in this study, as

sin h þ cos h f yh  Asx
f cu ¼  6 f cc ð23Þ
cos h  sin h cx  s  tan h
cx,y (dc) (mm)

where the inclination angle of the shear failure plane h is estimated

214–244
168–209
170–301

170–440

using experimental data of high-strength concrete columns tested

by Cusson and Paultre [10] at h ¼ 29 þ 56ðrle =f co Þ.
267

195
440

209
204

220
220
210
400

200
Column details

Considering a linear stress–strain relation in the post-peak

regime, the corresponding axial strain at complete formation of
Circular

Circular

Circular

Circular

Circular
Square
Square
Square
Square

Square

Square

Square

Square

the shear failure plane ecu is calculated with the help of the illustra-
Type

tion in Fig. 8 as
–

ecu ¼ ecc þ e0inel  f ccEfc cu

ð24Þ
No. of columns

W0
¼ ecc þ 2ðf ccpost
þf cu Þ
 f ccEfc cu

The concrete axial stresses at maximum load and at the com-

20

272
25
24
16
15

28
27
14
27
15
24
16
21

plete failure, as well as the corresponding strains are thus deter-

mined from Eqs. (1), (21), (23) and (24).
Coefficient of variation (%)

4.4. Stress–strain relation

Saatcioglu and Razvi [11]
Razvi and Saatcioglu [53]
Cusson and Paultre [16]
Sheikh and Uzumeri [8]

Nagashima et al. [9]

The stress–strain relation proposed by Popovics [48] that was

Standard deviation
Mander et al. [47]

Montgomeri [52]

employed by most existing analysis-oriented models is adopted

Hong et al. [45]
Scott et al. [46]

Han et al. [49]

Tokculu [51]

for the ascending branch of the proposed approach. The mathemat-

Column details.

Li et al. [50]
Li et al. [50]

All columns

ical expression for this relation is given below

Author

Mean

ðec =ecc Þ  r
Table 1

Ec
rc ¼ f ; r¼ ð25Þ
r  1 þ ðec =ecc Þr cc Ec  Esec
900 N.D. Tung, N.V. Tue / Construction and Building Materials 102 (2016) 893–903

200 200
(a) y = 1.001x
(b)
y = 0.987x
R² = 0.881 160 R² = 0.898
160

fcc,exp (MPa)

fcc,exp (MPa)
120 120

80 80

40 40

0 0
0 40 80 120 160 200 0 40 80 120 160 200

fcc,pred (MPa) fcc,pred (MPa)

Fig. 9. Comparison of the confined concrete strength fcc between tests and predictions for: (a) square columns; (b) circular columns.

25 Section 4.3. Thus, a linear relation originating from the peak and
ending with the formation of the inclined shear failure plane
y = 1.069x
demonstrated in Fig. 8 is used for the descending branch of the
20 R² = 0.564
stress–strain curve of confined concrete.
It should be noted that the strain at the peak stress ecc, the ulti-
15 mate strain ecu, as well as the stress–strain relation described in
Eqs. (21), (24) and (25) can only be applied for the localized frac-
10 ture zone in a reinforced concrete column. With the positive influ-
ence of the confining stress on the crack formation, the length of
5 the localized fracture zone in confined concrete columns (Ld) is
expected to greater than that of a pure concrete specimen Ld,i. From
a regression analysis of the test data performed by Hong et al. [45],
0
0 5 10 15 20 25 the following relation for Ld is obtained

   
rle rle pffiffiffiffiffiffiffi
Fig. 10. Comparison of the axial strain at peak stress ecc between tests and Ld ¼ Ld;i  1 þ 15 ¼ 0:6 1 þ 15  Acc ð26Þ
predictions. f co f co

where Acc is the area of the confined concrete core.

where Esec is the secant modulus of elasticity of confined concrete,
Esec = fcc/Ec.
It is widely recognized that the Popovics’s relationship yields
flatter descending branches than the corresponding ones obtained
from tests, also, it cannot quantitatively represent the effect of con-
finement on the crack propagation in the softening discussed in
The effective confinement index Ie = rle/fco in reinforced con-
crete columns varies in general between 0.01 and 0.1. The pro-
posed value for Ld in Eq. (26) is thus in the range of (0.7–1.5)D.
This proposal approximately lies in the core area of the values pro-
posed by Akiyama et al. [21], which are between (0.47–1.83)D and
also obtained from a regression analysis of their own test data.

Fig. 11. Stress–strain curves of some normal-strength concrete columns tested by Sheikh and Uzumeri [8].
 N.D. Tung, N.V. Tue / Construction and Building Materials 102 (2016) 893–903
Fig. 12. Stress–strain curves of some normal- and high-strength concrete columns tested by Nagashima et al. [9].
Fig. 13. Stress–strain curves of some high-strength concrete columns tested by Cusson and Paultre [16].
5. Comparison with experimental results
The proposed model was evaluated using test results of 184
reinforced rectangular and 88 circular columns subjected to con-
centric loading that were collected from the literature. The width
of the confined concrete core cx,y (or dc), concrete compressive
strength fc, reinforcement ratio qsh, and the yield limit of trans-
verse reinforcement fyh were in the broad range of 170–440 mm,
24.2–124.0 MPa, 0.2–2.8%, and 255–1420 MPa, respectively.
Details of the column parameters are summarized in Table 1.
Fig. 9 shows a comparison between experimental results and
predicted confined concrete strength fcc using Eq. (1) for 184
square columns performed by Sheikh and Uzumeri [8], Scott
et al. [46], Nagashima et al. [9], Cusson and Paultre [16], Hong
et al. [45], Han et al. [49], Saatcioglu and Razvi [11], Li et al. [50]
and 88 circular columns tested by Toklucu [51], Mander et al.
[47], Li et al. [50], Montgomery [52] and Razvi and Saatcioglu
[53]. Trend lines are plotted through the origin (0, 0) for each type
of column with the corresponding slopes and coefficients of deter-
mination of the regression. It can be observed that the predicted
confined concrete strengths are in good agreement with the test
data.
Details of the comparison for individual test programs are sum-
marized in Table 1. The proposed method yields for all specimens
the mean of 1.03 and the standard deviation of 0.11. To evaluate
the quality of the proposed approach in relation to existing models,
the predicted confined concrete strength fcc using the proposed
method is also compared with some models proposed by Mander
et al. [7], Légeron and Paultre [19] as well as design equations
recently included in fib MC 2010 [15]. The test results versus the
predicted confined concrete strength using these models are also
summarized in Table 1. As can be seen among the investigated
equations the model proposed by Mander et al. [7]. predicts very
well the confined concrete strength for normal-strength concrete
columns tied by normal-strength steel, such as test programs of
901
Sheikh and Uzumeri [8], Scott et al. [46], Mander et al. [47], while
it in general overestimates the confined concrete strength for high-
strength concrete columns tied by high-strength steel, such as test
programs of Nagashima et al. [9], Cusson and Paultre [16], Mont-
gomery [52]. The unsatisfactory predictions for these test pro-
grams is caused by the fact that in this historical model the yield
limit of the high-strength confining steel is used when determining
the effective confining stress, regardless whether the reinforce-
ment yields or not at the concrete peak stress, as discussed in
Section 1. Similar tendency can be also observed for the predictions
of the model included in fib MC 2010, in which the yield limit of the
confining steel is employed. Having taken into account the actual
stress in the confining reinforcement at the concrete peak stress
obtained from a regression analysis, the model proposed by
Légeron and Paultre [19] yields better predictions than the models
mentioned above, which are similar to the results of the proposed
method.
Similarly, test results versus predicted axial strain of the
confined concrete core at the peak stress ecc using Eq. (21) for
117 columns performed by Sheikh and Uzumeri [8], Nagashima
et al. [9]., Cusson and Paultre [16], Hong et al. [45], Han et al.
[49] are plotted in Fig. 10. It can be observed that the predicted
axial strains of the confined concrete core at peak stress are in good
agreement with the test data. The slope of the trend line equals
1.07, indicating that the predictions are generally conservative.
The predicted axial strains of the confined concrete core at the
peak stress are also in reasonable correlation with the experimen-
tal ones with the coefficient of determination of the regression line
of 0.56. The less agreement of the predicted axial strain when
compared to the predicted confined compressive strength might
imply the fact that the modeling of concrete ductility is more
complicated than the strength modeling. Although only specimens
having gage length that similar to the calculated length of the
localized fracture zone were selected for the comparison, it should
also be noted that the measured strain cannot always represent the
902 N.D. Tung, N.V. Tue / Construction and Building Materials 102 (2016) 893–903
strain of the localized fracture zone. The actual failure can occur
outside of the gage length, as can be seen elsewhere [46].
Sample comparisons selected from different research programs,
shown in Figs. 11–13, indicate good agreement between experi-
mental and analytical stress–strain relationships
6. Conclusions
This paper presents an approach on the basis of fracture
mechanics to modeling the confinement in reinforced columns
based on observations of the concrete crack formation in presence
of a confining stress as well as the relation between the crack for-
mation and the lateral deformation. Taking into consideration the
interaction between the concrete core and the transverse rein-
forcement together with the compatibility of deformations in the
lateral direction, the stress and strain in the transverse reinforce-
ment is estimated. The effective confinement stress provided by
transverse reinforcement is thus calculated. The relations between
the pre-peak and the post-peak energies and the concrete lateral
strains for both unconfined and confined concrete are established
through the observation of the crack formation and propagation
of concrete in compression. The ductility of the concrete core is,
thus, described.
The proposed method is extensively verified through compar-
isons with experimental results in terms of stress in the transverse
reinforcement at the concrete peak stress, the confined compres-
sive strength and the corresponding concrete axial strain and also
the stress–strain curve. The comparisons show that the predictions
are in good agreement with the experimental results.
By observations of lateral strains, the presented approach has
potential capabilities for accurate assessment the effectiveness of
the confinement effect in reinforced concrete columns by the
means of effective use of materials, as well as for assessment the
risk of early spalling of the concrete cover. The basic considerations
for the assessments have been presented in outline in another
study [54].
Acknowledgments
The theoretical development presented in this study was
initiated during the research stay of the first author at Leipzig
University, which was financed by the German Academic Exchange
Service (DAAD). This support is greatly appreciated. The authors
also would like to thank Dr. Jiabin Li for beneficial discussions
and encouragement to complete this manuscript.
References
[1] F.E. Richart, A. Brandtzaeg, R.L. Brown, A study of the failure of concrete under
combined compressive stresses, Engineering Experiment Station, University of
Illinois, 1929, p. 102. Bulletin No. 185.
[2] F. Ansari, Q. Li, High strength concrete subjected to triaxial compression, ACI
Mater. J. 95 (6) (1998) 747–755.
[3] D.C. Candappa, J.G. Sanjayan, S. Setunge, Complete triaxial stress–strain curves
of high-strength concrete, J. Mater. Civ. Eng. ASCE 13 (3) (2001) 209–215.
[4] X. Lu, C.T.T. Hsu, Behavior of high strength concrete with and without steel
fiber reinforcement in triaxial compression, Cem. Concr. Res. 36 (2006)
1679–1685.
[5] D. Sfer, I. Carol, R. Gettu, Study of behavior of concrete under triaxial
compression, J. Eng. Mech. ASCE 128 (2) (2002) 156–163.
[6] I. Imran, S. Pantazopoulou, Experimental study of plain concrete under triaxial
stress, ACI Mater. J. 93 (6) (1996) 589–601.
[7] J.B. Mander, M.J.N. Priestley, R. Park, Theoretical stress–strain model for
confined concrete, J. Struct. Eng. ASCE 114 (8) (1988) 1804–1826.
[8] S.A. Sheikh, S.M. Uzumeri, Strength and ductility of tied concrete columns, J.
Struct. Eng. ASCE 106 (5) (1980) 1079–1102.
[9] T. Nagashima, S. Sugano, H. Kimura, A. Ichikawa, Monotonic axial compression
test on ultra-high-strength concrete tied columns, in: Proceedings of the 10th
World Conference on Earthquake Engineering, s.n., Madrid, Spain, 1992, 2983–
2988.
[10] D. Cusson, P. Paultre, High-strength concrete columns confined by rectangular
ties, J. Struct. Eng. ASCE 120 (3) (1994) 783–804.
[11] M. Saatcioglu, S. Razvi, High-strength concrete columns with square sections
under concentric compression, J. Struct. Eng. ASCE 124 (2) (1998) 1438–1447.
[12] J.P. Moehle, T. Cavanagh, Confinement effectiveness of crossties in RC, J. Struct.
Eng. ASCE 111 (10) (1985) 2105–2120.
[13] Y.K. Yong, M.G. Nour, E.G. Nawy, Behavior of laterally confined high-strength
concrete under axial loads, J. Struct. Eng. ASCE 114 (2) (1988) 332–351.
[14] V. Sigrist, The compression chord model for reinforced concrete members (in
German: Das Druckgurtmodell für Stahlbetonbauteile), Beton- und
Stahlbetonbau 104 (8) (2009) 496–504.
[15] fib Model Code for Concrete Structures 2010, International Federation for
Structural Concrete, Lausanne, 2013, ISBN 978-3-433-03061-5.
[16] D. Cusson, P. Paultre, Stress-strain model for confined high-strength concrete,
J. Struct. Eng. ASCE 121 (3) (1995) 468–477.
[17] W.P. Lokuge, J.G. Sanjayan, S. Setunge, Stress–strain model for laterally
confined concrete, J. Mate. Civ. Eng. ASCE 17 (2005) 261–268.
[18] S. Razvi, M. Saatcioglu, Confinement model for high-strength concrete, J.
Struct. Eng. ASCE 125 (3) (1999) 281–289.
[19] F. Legeron, P. Paultre, Uniaxial confinement model for normal- and high-
strength concrete columns, J. Struct. Eng. ASCE 129 (2) (2003) 241–252.
[20] P. Paultre, F. Legeron, Confinement reinforcement design for reinforced
concrete columns, J. Struct. Eng. ASCE 134 (5) (2008) 738–749.
[21] M. Akiyama, M. Suzuki, D.M. Frangopol, Stress-averaged strain model for
confined high-strength concrete, ACI Struct. J. 107 (2) (2010) 179–188.
[22] H.O. Shin, Y.S. Yoon, W.D. Cook, D. Mitchell, Effect of confinement on the axial
load response of ultrahigh-strength concrete columns, J. Struct. Eng. ASCE
(2014), http://dx.doi.org/10.1061/(ASCE)ST.1943-541X.0001106.
[23] B. Bousalem, N. Chikh, Development of a confined model for rectangular
ordinary reinforced concrete columns, Mater. Struct., RILEM 40 (2007) 605–
613.
[24] Q.G. Xiao, J.G. Teng, T. Yu, Behavior and modeling of confined high-strength
concrete, J. Compos. Constr. ASCE 14 (3) (2010) 249–259.
[25] P. Paultre et al., Behavior of steel fiber-reinforced high-strength concrete
columns under uniaxial compression, J. Struct. Eng. ASCE 136 (10) (2010)
1225–1235.
[26] P.D. Kiousis, P.G. Papadopoulos, H. Xenidis, Truss modeling of concrete
columns in compression, J. Eng. Mech. ASCE 136 (8) (2010) 1006–1014.
[27] J.C. Lim, T. Ozbakkaloglu, Stress–strain model for normal- and light-weight
concretes under uniaxial and triaxial compression, Constr. Build. Mater. 71
(2014) 492–509.
[28] G. Markeset, Failure of concrete under compressive strain gradients (Ph.D.
thesis), NTH University of Trondheim, 1993.
[29] J. Meyer, A contribution to the investigation of the deformation capacity of
concrete members subjected to bending (in German: Ein Beitrag zur
Untersuchung der Verformungsfähigkeit von Bauteilen aus Beton unter
Biegedruckbeanspruchung) (Ph.D. Thesis), Universität Leipzig, s.n., Germany,
1998.
[30] A. Sint, Ductility of bent members by failure in the concrete compression zone
(in German: Duktilität von Biegebauteilen bei Versagen der Betondruckzone),
Beton- und Stahlbetonbau 95 (5) (2003) 285–292.
[31] T. Lersrisakulrat, K. Watanabe, M. Matsuo, J. Niwa, Experimental study of
localization parameters in concrete subjected to compression, Concr. Lib. JSCE
38 (Dec. 2001) 225–236.
[32] K. Watanabe, J. Niwa, H. Yokota, M. Iwanami, Experimental study on stress–
strain curves of concrete considering localised failure in compression, J. Adv.
Concr. Technol., JCI 2 (3) (2004) 395–407.
[33] S. Puri, J. Weiss, Assessment of localized damage in concrete under
compression using acoustic emission, J. Mater. Civ. Eng. ASCE 18 (3) (2006)
325–333.
[34] N.D. Tung, Influence of the deformation capacity of concrete and steel on the
effect of the confinement in reinforced concrete columns (in German: Zum
Einfluss des Verformungsvermögens von Stahl und Beton auf die
Umschnürungswirkung bei Stahlbetonstützen) (Ph.D. thesis). Universität
Leipzig, s.n., Germany, 2009.
[35] N.D. Tung, N.V. Tue, Post-peak behavior of concrete specimens undergoing
deformation localization in uniaxial compression, Constr. Build. Mater. 99
(2015) 109–117.
[36] K. Watanabe, J. Niwa, M. Iwanami, H. Yokota, Localized failure of concrete in
compression identified by AE method, Constr. Build. Mater. 18 (2004) 189–
196.
[37] M. Jansen, S. Shah, Effect of length on strain compressive strain softening of
concrete, J. Eng. Mech. ASCE 123 (1) (1997) 25–35.
[38] K. Rokugo, W. Koyanagi, Role of compressive fracture energy of concrete on
the failure behavior of reinforced concrete beams, in: A. Carpinteri (Ed.),
Applications of fracture mechanics to reinforced concrete, Elsevier Applied
Science, s.l., 1992, pp. 437–464.
[39] M. Sargin, S.K. Ghost, V.K. Handa, Effect of lateral reinforcement upon the
strength and deformation properties of concrete, Mag. Concr. Res. 23 (75–76)
(1971) 99–110.
[40] CEN, Eurocode 2: Design of Concrete Structures – Part 1–1: General Rules for
Buildings (EN 1992-1-1), European Committee for Standardization, Brussels,
2010.
[41] J.C. Lim, T. Ozbakkaloglu, Lateral strain-to-axial strain relationship of confined
concrete, J. Struct. Eng. ASCE 141 (5) (2015) 1–18.
 N.D. Tung, N.V. Tue / Construction and Building Materials 102 (2016) 893–903
[42] E. Montoya, F. Vecchio, S. Sheikh, Compression field modeling of confined
concrete: constitutive models, J. Mater. Civ. Eng. 18 (4) (2006) 510–517.
[43] D.C. Candappa, S. Setunge, J.G. Sanjayan, Stress versus strain relationship of
high-strength concrete under high lateral confinement, Cem. Concr. Res. 29
(1999) 1977–1982.
[44] K.K.B. Dahl, A Failure criterion for normal and high-strength concrete,
Department of Structural Engineering, Technical University of Denmark,
1992. Rep. R. 286.
[45] K. Hong, M. Akiyama, S. Yi, M. Suzuki, Stress-strain behaviour of high-strength
concrete columns confined by low-volumetric ratio rectangular ties, Mag.
Concr. Res. 58 (2) (2006) 101–115.
[46] B.D. Scott, R. Park, M.J.N. Priestley, Stress–strain behavior of concrete confined
by overlapping hoops at low and high strain rates, ACI J. 79 (1) (1982) 13–27.
[47] J.B. Mander, M.J.N. Priestley, R. Park, Observed stress–strain behavior of
confined concrete, J. Struct. Eng. ASCE 114 (8) (1988) 1827–1849.
[48] S. Popovics, Analytical approach to complete stress–strain curves, Cem. Concr.
Res. 3 (5) (1973) 583–599.
[49] B.S. Han, S.W. Sihn, B.Y. Bahn, A model of confined concrete in high-
strength reinforced concrete tied columns, Mag. Concr. Res. 55 (3) (2003)
203–214.
903
[50] B. Li, R. Park, H. Tanaka, Strength and ductility of reinforced concrete members
and frames constructed using high strength concrete, University of
Canterbury, Christchurch, New Zealand, 1994, p. 373. Technical Rep. 94–5.
[51] M.T. Toklucu, Behavior of reinforced concrete columns confined with circular
spirals and hoops (Master thesis), Dept. of Civ. Eng., University of Toronto, s.n.,
Canada, 1992.
[52] D.L. Montgomery, Behavior of spirally reinforced high-strength concrete
columns (Master thesis), Dept. of Civ. Eng., University of Toronto, s.n.,
Canada, 1996.
[53] S. Razvi, M. Saatcioglu, Tests of high-strength concrete columns under
concentric loading, Ottawa Carleton Earthquake Engineering Research
Centre, s.n., Ottawa, Canada, 1996, p. 147. Rep. No. OCEERC 96–03..
[54] N.V. Tue, N.D. Tung, Deformation-based approach for determination of the
efficiency of the confinement in R/C members, in: Proceedings of the
Thirteenth East Asia-Pacific Conference on Structural Engineering and
Construction (EASEC-13), s.n., Sapporo, Japan, 2013, pp. 1–8.