Civil Engineering Dimension, Vol. 11, No.

2, September 2009, 78-88

ISSN 1410-9530 print / ISSN 1979-570X online

Effects of Confinement on Interaction Diagrams of Square

Reinforced Concrete Columns

Tavio, T.1, Wimbadi, I.1, Kusuma Negara, A.1, Tirtajaya, R.1

Abstract: To prevent brittle failure, the design of a structural column in a seismic-resistant

building is of important consideration, particularly in terms of confinement. In the recent
building code, the need of closely-spaced stirrups in a structural member, such as column
becomes compulsory due to the ductility and strength considerations. However, the design is
based on the simplified block stress of unconfined concrete, and does not account for the strength
gain due to the presence of confinement. To investigate the effects of lateral confinement on the
column capacity, an analytical study is carried out. Both the strength gain in concrete core and
the loss of strength in the cover are considered in the analytical models to exhibit the remaining
strength gain after the mobilization of strength gain in the core concrete to compensate the loss
of strength in the concrete cover. There are six key parameters primarily influence the
effectiveness of lateral confinement. The most influencing parameter is found to be the spacing of
transverse steel. The presence of closely-spaced lateral confinement significantly increases the
magnitude of stress-strain curve of concrete. This increase expands the interaction diagram of
the column particularly when it is in the compression-controlled region (for lower-story columns
when axial load dominates the behavior).

Keywords: column capacity, confinement effects, interaction diagram, lateral reinforcement..

Introduction since 1956 edition [3], and it has been in the code
since then. The concept was also adopted in the
The effects of confinement on a structural column in national building code [4] for flexural design. In SNI
a building are mainly due to the presence of lateral 03-2847-2002 [5], it remains applicable for flexural
reinforcement provided over the column height. It design of reinforced concrete members. The existing
results in higher capacity and ductility of a column interaction diagrams developed for the column
that help to prevent the column from brittle failure. capacity are also based on this assumption that does
Laterally-confined columns have higher capacity not account for the strength gain from the presence
than the unconfined ones since the concrete core of of confinement. Even though the block stress concept
the columns gains the strength from the mobilization has long been adopted as a reasonable approach, the
of lateral confinement. Recent development in research indicated that the presence of confinement
research and design engineering, particularly in in a concrete column would affect the actual
reinforced concrete structures often requires higher compressive stress-strain curve of concrete. This
capacity and ductility of structural members. To effort gives a more accurate prediction on the
prevent a building structure from brittle failure, the compressive force of concrete in a column, and thus,
design of a structural column in a seismic-resistant resulting further in a more efficient column cross-
building is of important consideration, particularly in section [6]. With advancement of computer
terms of confinement [1]. programming and technology, the computational
effort can be much accelerated by implementing the
Up to present, the design of a structural column is numerical procedure to solve the stress-strain
based on the simplified block stress of unconfined curves.
concrete proposed by Whitney [2]. This proposed
block stress was adopted by the ACI Building Code To investigate the effects of lateral confinement on
the column capacity, an analytical study is carried
out. Both the strength gain in concrete core and the
1 Department of Civil Engineering, Sepuluh Nopember loss of strength in the cover are considered in the
Institute of Technology (ITS), Surabaya, Indonesia
E-mail: [email protected] analytical models to exhibit the remaining strength.
So far, this strength gain in the confined core is used
Note: Discussion is expected before November, 1st 2009, and will only for the compensation of the possible strength
be published in the “Civil Engineering Dimension” volume 12,
number 1, March 2010. loss due to the spalling of concrete cover (which is
unconfined). Recent codes of practice still disregard
Received 11 October 2008; revised 6 January 2009; accepted 16
April 2009. this effect for the design purposes and, thus in the

78
 Tavio, T., et al.. / Effects of Confinement on Interaction Diagrams / CED, Vol. 11, No. 2, September 2009, pp. 78–88

conservative side. In this numerical study of confined For descending branch, εc > ε co :
concrete columns, the behavior of concrete core is
modeled by the stress-strain relationship of confined f c = f c′[1 − Z 0 (ε c − ε co )]
concrete, whereas for the cover, as unconfined
concrete. Several stress-strain relationships of where:
confined concrete available in literature are adopted 0.5
Z0 =
in the study, namely Kent-Park [7], Sheikh-Uzumeri ε 50 u − ε co
[8], Mander et al. [9], Yong-Nawy [10], Cusson-
Paultre [11], Diniz-Frangopol [12], Kappos- 3 + 0.002 f c′
ε 50u =
Konstantinidis [13], Hong-Han [14], and Kusuma- f c′ − 1000
Tavio [15].
in which, ε 50u is the unconfined concrete strain when
From the study, it can be concluded that there are the stress reaches 50 percent of peak stress, ε co the
six key parameters primarily influence the unconfined concrete strain at the peak stress, and
effectiveness of lateral confinement. The most
influencing parameter is found to be the spacing of f c′ the compressive strength of unconfined concrete
transverse steel. Even though, the codes ignore the (in psi, 1 psi = 0.006895 MPa).
effect of confinement on the strength gain due to the
conservative consideration for the design purposes, Popovics Model [16]
the authors still intend to discover the actual
possible remaining gain of strength due to the For entire stress-strain curve of concrete, the stress
presence of confinement. is calculated using the following equation:
⎛ε ⎞ n
Unconfined Concrete Models f c = f c′⎜⎜ c ⎟⎟ (5)
⎝ ε co ⎠ ⎡ ⎛ εc ⎞ ⎤
n

The unconfined concrete models adopted in the study ⎢n − 1 + ⎜⎜ ⎟⎟ ⎥

are Whitney’s block stress [2], Kent-Park [7], ⎢⎣ ⎝ ε co ⎠ ⎥⎦
Popovics [16], and Thorensfeldt et al. [17] models. where:
Brief summary of these models are described
subsequently.
ε co = 0.0005 ( f c′) 0.4
f c′
Whitney’s Block Stress [2] n = 0.8 +
17
This model is only used at the ultimate limit state. All units are in continental system, where 1 psi =
The compressive stress of concrete is asummed to be 0.006895 MPa.
constant as a block stress at the following value:
f c = 0.85 f c′ (1) Thorenfeldt et al. Model [17]

a = β1 c (2) For entire stress-strain curve of concrete, the stress

where: is calculated using the following equation:
⎧ 0.85 if f c′ ≤ 30 MPa ⎛ε ⎞
β1 = ⎪0.85 − ( f ′ − 30) 0.05 if 30 MPa < f ′ ≤ 58 MPa n
⎨ f c = f c′⎜⎜ c ⎟⎟ (6)
⎝ ε co ⎠ ⎡ ⎛ εc ⎞ ⎤
c c nk
⎪ 7
if f ′ > 58 MPa ⎢n − 1 + ⎜⎜ ⎟⎟ ⎥
⎩ 0.65 c

ε cu = 0.003 ⎢⎣ ⎝ ε co ⎠ ⎥⎦
in which, c is the distance of neutral axis from extreme where:
compressive fiber of concrete, β1 the conversion ε
⎧
factor from parabolic to rectangular shape as a ⎪⎪ 1 if c ≤ 1
k= ⎨ ε co
function of concrete compressive strength, and ε cu f c′ ε
⎪ 0.67 + if c > 1
the ultimate strain of unconfined concrete. ε co
⎩⎪ 62
Kent-Park Model [7] f c′
n = 0.8 +
17
For ascending branch, εc ≤ ε co ( ε co = 0.002):
Ec = 3320 f co′ + 6900 (in MPa)
⎡ 2ε ⎛ ε ⎞ 2 ⎤
′
f c = f c′⎢ c − ⎜⎜ c ⎟⎟ ⎥ (3)
ε co = f c ⎛⎜ n ⎞⎟
⎢⎣ ε co ⎝ ε co ⎠ ⎥⎦ Ec ⎝ n − 1 ⎠

79
 Tavio, T., et al.. / Effects of Confinement on Interaction Diagrams / CED, Vol. 11, No. 2, September 2009, pp. 78–88

Confined Concrete Models 0.5

Z=
3 b
The confined concrete models adopted in the study ρs c
are Kent-Park [7], Sheikh-Uzumeri [8], Mander et 4 s
al. [9], Yong-Nawy [10], Cusson-Paultre [11], Diniz- ′ ′
f cc = K s f co
Frangopol [12], Kappos-Konstantinidis [13], Hong-
f co′ = 0.85 f c′
Han [14], and Kusuma-Tavio [15] models. The most
obvious difference of all these confined stress-strain ε cc1 = 80 K s f c′ × 10 −6
models is particularly in term of ductility along the
⎡ 248 ⎡ ⎛ s ⎞ ⎤ ρ s f s′ ⎤
2
descending branch [18]. Brief summary of these
ε cc 2 = ε co ⎢1 + ⎢1 − 5.0⎜⎜ ⎟⎟ ⎥ ⎥
models are described subsequently. ⎢ C ⎢ ⎝ bc ⎠ ⎥ f c′ ⎥
⎣ ⎣ ⎦ ⎦
2 ⎡
Kent-Park Model [7] (bc ) ⎛ s ⎞ ⎤
2
nC ⎞ ⎛
2
K s = 1.0 + ⎢⎜1 − ⎟ ⎜1 − ⎟ ⎥ ρ s f s′
140 Pocc ⎢⎜⎝ 5.5bc2 ⎟⎠ ⎜⎝ 2bc ⎟⎠ ⎥
For ascending branch, εc ≤ 0.002: ⎣ ⎦
⎡ 2ε c ⎛ ε c ⎞ 2 ⎤ Pocc = f co′ ( Acc )
f c = f c′⎢ −⎜ ⎟ ⎥ (7)
⎢⎣ 0.002 ⎝ 0.002 ⎠ ⎥⎦ ε co = 0.0022
bc
For descending branch, 0.002 ≤ εc ≤ ε20c: ε cc85 = 0.225ρ s + ε cc 2
s
f c = f c′[1 − Z (ε c − 0 . 002 )] (8) in which, Acc is the area of confined concrete core,
For horizontal branch, εc ≥ ε20c: bc the width of confined concrete core measured
f c = 0.2 f c′ (9) center-to-center of lateral reinforcement, C the
distance between longitudinal reinforcement
where:
confined laterally by lateral reinforcement, K s the
0.5
Z= magnification factor, f s′ the stress in lateral
ε 50u + ε 50 h − 0.002
reinforcement at the maximum stress of confined
3 + 0.002 f c′ concrete (assume f s′ = f yh at the peak stress), and
ε 50u =
f c′ − 1000 n the number of ineffective parabolic area in concrete
3 b′′ core, or the number of longitudinal reinforcement
ε 50 h = ρs confined laterally by lateral reinforcement.
4 sh
where ρs is the volumetric ratio of lateral Mander et al. Model [9]
reinforcement to the confined concrete core For entire stress-strain curve of concrete, the stress
measured outer-to-outer of lateral reinforcement, b′′ is calculated using the following equation:
the width of confined concrete core measured outer- f cc′ xr
to-outer of lateral reinforcement, and sh the spacing fc = (13)
r − 1 + xr
of lateral reinforcement. where:
εc
Sheikh-Uzumeri Model [8] x=
ε cc
For ascending branch, ε c ≤ ε cc1 : Ec
r=
⎡ ⎛ε ⎞ ⎛ε ⎞ 2
⎤ Ec − Esec
f c = f cc′ ⎢2⎜⎜ c ⎟⎟ − ⎜⎜ c ⎟⎟ ⎥ (10)
⎢⎣ ⎝ ε cc ⎠ ⎝ ε cc ⎠ ⎥⎦ Ec = 5000 f c′ MPa

For horizontal branch, ε cc1 < ε c ≤ ε cc 2 : f cc′

Esec =
ε cc
f c = K s f co′ (11)
For descending branch, ε cc 2 < ε c ≤ ε cc 30 : ⎡ ⎛ f cc′ ⎞⎤
ε cc = ε co ⎢1 + 5⎜⎜ − 1⎟⎟⎥
f c = f cc′ [1 − Z (ε c − ε cc )] (12) ⎣ ⎝ f c′ ⎠⎦
where: ε co = 0.002

80
 Tavio, T., et al.. / Effects of Confinement on Interaction Diagrams / CED, Vol. 11, No. 2, September 2009, pp. 78–88

⎛ 7.94 f l′ f ′⎞ ⎡ Ei 4 E2i ⎤
f cc′ = f c′ ⎜⎜ − 1.254 + 2.254 1 + − 2 l ⎟⎟ D = (ε i − ε 2i )⎢ − ⎥
⎝ f c′ f c′ ⎠ ⎣ ( f cc′ − fi ) ( f cc′ − f 2i ) ⎦
A f
Ke = e Ei = i
Acc εi
f l′x = ke ρ x f yh (x-direction) f 2i
E2 i =
f l′y = ke ρ y f yh (y-direction) ε 2i
⎛ n
(w′ )2 ⎞ ⎛ s′ ⎞⎟ ⎛⎜1 − s′ ⎞⎟
⎜1 − ∑ i ⎟ ⎜1 −
⎡ ⎛ 0.254 s ⎞ ⎤
⎜ i=1 6b d ⎟ ⎜ 2b ⎟ ⎜ 2d ⎟ ⎢1 + 0.11⎜1 − h ′′ ⎟ ⎥
c c ⎠⎝ c ⎠⎝ c ⎠ ⎝ ⎠
Ke = ⎝ f cc′ = K s f c′ = ⎢ ⎥ f′
(1 − ρ cc ) ⎢⎛
⎢⎜⎜ ρ s +
nφs
ρ⎟
⎞ f yh ⎥ c
⎥
ε cu = 0.004 + 1.4 ρ s f yhε sm f cc′ ⎣⎢⎝ 0.31496 sφ l ⎟⎠ f c′ ⎥⎦
in which, bc , d c is the cross-sectional dimension of
⎟(145ρ s f yh )
⎛ 0.734 s ⎞
0.0035⎜1 −
23
confined concrete core measured center-to-center of
⎝ ′′
h ⎠
lateral reinforcement in the x and y directions, ε cc = 0.00265 +
respectively, s′ the clear spacing of lateral 145 f c′
reinforcement, Ae the effective area of confined
⎡ ⎛ f′⎞ ⎤
concrete core, wi′ the ith clear spacing from two f i = f cc′ ⎢0.25⎜⎜ c ⎟⎟ + 0.4⎥
⎣ ⎝ f cc′ ⎠ ⎦
adjacent longitudinal reinforcement, ρcc the ratio of
cross-sectional area of longitudinal reinforcement to ⎡ ⎛ ε cc ⎞ ⎤
ε i = K s ⎢1.4⎜⎜ ⎟⎟ + 0.0003⎥
area of confined concrete core, and ε sm the strain of ⎣ ⎝ Ks ⎠ ⎦
reinforcing steel at maximum tensile stress.
⎡ ⎛ f′ ⎞ ⎤
f 2i = f cc′ ⎢0.025⎜ c ⎟ − 0.065⎥ ≥ 0.3 f cc′
Yong et al. Model [10] ⎣ ⎝ 1000 ⎠ ⎦

For ascending branch, ε c ≤ ε cc : ε 2i = 2ε i − ε cc

AX + BX 2 in which, h′′ is the width of confined concrete core
Y= (14)
1 + ( A − 2)X + (B + 1)X 2 measured inner-to-inner of lateral reinforcement, n
the number of longitudinal reinforcement, φs the
For descending branch, ε c ≥ ε cc :
nominal diameter of lateral reinforcement, φl the
CX + DX 2
Y= (15) nominal diameter of longitudinal reinforcement, ρ
1 + (C − 2)X + (D + 1)X 2
the ratio of cross-sectional area of longitudinal
where:
reinforcement to gross area of concrete ( As Ag ),
εc
X = and wc the concrete density in kN/m3. All units are
ε cc
in continental system (1 psi = 0.006895 MPa).
f
Y= c
f cc′ Cusson-Paultre Model [11]
ε cc
A = Ec For ascending branch, ε c ≤ ε cc :
f cc′
⎡ k (ε c ε cc ) ⎤
⎡ ( A − 1)2 ⎤ f c = f cc′ ⎢ k ⎥
(16)
B=⎢
0 . 55
⎥ −1 ⎣ k − 1 + (ε c ε cc ) ⎦
⎣ ⎦
For descending branch, ε c ≥ ε cc :
Ec = 36.78wc1.5 f c′
(ε − ε i ) ⎡ ε 2i Ei − 4ε i E2i ⎤ [
f c = f cc′ exp k1 (ε c − ε cc ) 2
k
] (17)
C = 2i
ε cc ⎢⎣ ( f cc′ − f i ) ( f cc′ − f 2i ) ⎥⎦ where:

81
 Tavio, T., et al.. / Effects of Confinement on Interaction Diagrams / CED, Vol. 11, No. 2, September 2009, pp. 78–88

Ec ⎡ ⎛ ε ⎞ ⎤
A
k= f c = f cc′ ⎢1 − ⎜⎜1 − c ⎟⎟ ⎥ (18)
Ec − ( f cc′ ε cc ) ⎢⎣ ⎝ ε cc ⎠ ⎥⎦
Ec = 3320 f c′ + 6900 For descending branch, ε c ≥ ε cc :
k1 =
ln 0.5
[
f c = f cc′ exp − k (ε c − ε cc )
1.15
] (19)
(ε cc50 − ε cc ) k2
where:
1.4
⎛f ⎞ Ash f yh
k2 = 0.58 + 16⎜⎜ le ⎟⎟ fl =
⎝ f co′ ⎠ de s
Ash = λ Ast
0.7
f cc′ ⎛f ⎞
= 1.0 + 2.1⎜⎜ le ⎟⎟
f co′ ⎝ f co′ ⎠ f le = C f f l
1.7
⎛ f le ⎞ s
ε cc = ε co + 0.21⎜⎜ ⎟⎟ Cf =1−
⎝ f co′ ⎠ de
A = Ec .ε cc f cc′
1.1
⎛f ⎞
ε cc50= ε o50 + 0.15⎜⎜ le ⎟⎟
⎝ f co′ ⎠ Ec = 33 wc1.5 f c′
f hcc = f yh
k = 0.17 f c′ exp (− 0.01 f le λ1 )
ε o 50 = 0.004
K f ⎛ A + Ashy ⎞⎟ λ1 = 1 + 25
f le
f c′
[1 − exp ( f c′ 44.79 )
9
]
f le = K e f l = e hcc ⎜ shx
s ⎜⎝ bcx + bcy ⎟⎠
⎛ 21 ⎞
⎡ n
(wi )2 ⎤ ⎛⎜1 − 0.5 s′ ⎞⎟ ⎛⎜1 − 0.5 s′ ⎞⎟ f cc′ = f c′ + ⎜⎜1.15 + ⎟⎟ f le
⎢ ∑
1 − ⎥⎜ ⎝ f c′ ⎠
⎢ i =1 6bcxbcy ⎦⎥ ⎝
⎣ bcx ⎟⎠ ⎜⎝ bcy ⎟⎠
Ke = ε cc = 1.027 × 10 −7 f c′ + 0.0296
f le
+ 0.00195
1 − ρt f c′
Efective confinement index = IPe = f le f co′
in which, d e is the equivalent diameter of lateral
ε hcc = 0.5ε cc [1 − ( f le f cc′ )] reinforcement, Ash the total cross-sectional area of
in which, Ashx is the area of lateral reinforcement at lateral reinforcement in a section including crossties,
cross section perpendicular to x-axis, Ashy the area Ast the cross-sectional area of lateral reinforcement,
of lateral reinforcement at cross section C f the corrective factor for confinement, and λ a
perpendicular to y-axis, f l the stress of nominal
factor depending on the configuration type of lateral
lateral reinforcement acting in concrete core, f le the reinforcement. All units are in SI system.
effective confining stress acting in concrete core,
Kappos-Konstantinidis Model [13]
f hcc the stress of lateral reinforcement at peak
stress of confined concrete, k the coefficient affecting For ascending branch, 0 < ε c ≤ ε cc :
the slope of ascending branch of stress-strain curve,
k1 the coefficient affecting the slope of descending ⎛ε ⎞⎛ E c ⎞
f cc′ ⎜⎜ c ⎟⎟⎜ ⎟
⎜ ⎟
branch of stress-strain curve, k 2 the coeffient ⎝ ε cc ⎠⎝ E c − E p ⎠
affecting the curvature of descending branch of fc = Ec
(20)
⎛ Ec ⎞
stress-strain curve, and ε hcc the strain of lateral
⎟ −1+ ⎜ ε c
⎛ ⎞ Ec − E p
⎜ ⎟⎟
reinforcement when the stress reaches f hcc . All ⎜E −E ⎟ ⎜ε
⎝ c p ⎠ ⎝ cc ⎠
units are in SI system.
For descending branch, ε c > ε cc :
Diniz-Frangopol Model [12]
⎡ ε − ε cc ⎤
f c = f cc′ ⎢1 − 0.5 c ⎥ ≥ 0.3 f cc′ (21)
For ascending branch, ε c ≤ ε cc : ⎣ ε cc 50 − ε cc ⎦

82
 Tavio, T., et al.. / Effects of Confinement on Interaction Diagrams / CED, Vol. 11, No. 2, September 2009, pp. 78–88

⎛ f ′⎞
0.3 0.56
(in MPa) ⎛ f le ⎞
Ec = 22,000⎜ c ⎟ ε cc = ε co + 0.015⎜⎜ ⎟⎟
⎝ 10 ⎠ ⎝ f co′ ⎠
f cc′ f le = K e ρ s f hcc
Ep =
ε cc
⎡ (wi )2 ⎤ ⎛⎜1 − 0.5 s′ ⎞⎟ ⎛⎜1 − 0.5 s′ ⎞⎟
f cc′ = f co′ + 10.3(αρ s f yh )
n
⎢ ∑− ⎥⎜
0.4
1
⎢ i=1 6bcxbcy ⎦⎥ ⎝
⎣ bcx ⎟⎠ ⎜⎝ bcy ⎟⎠
Ke =
f co′ = 0.85 f c′ 1 − ρt
[
ε cc = 1 + 32.83(αω w )1.9 ε co ] ⎧⎪ ⎛ K e ρ s ⎞ ⎫⎪
0.70

f hcc = Es ⎨0.45ε co + 0.73⎜

⎜ ⎟
⎟ ⎬ ≤ f yh
0.70( f c′ ) ⎝ f co′ ⎠ ⎪⎭
0.31

ε co = ⎪⎩
1,000
f co′ = 0.85 f c′
ρ s f yh
ωw = ε co = 0.0028 − 0.0008 k 3
f c′
⎛ ∑ (bi )2 ⎞⎛ k3 = 40 f co′ ≤ 1.0
α = ⎜1 − ⎟⎜1 − s ⎞⎟⎛⎜1 − s ⎞⎟
6bc d c ⎟⎠⎜⎝ 2bc ⎟⎠⎜⎝ 2d c ⎟⎠
⎜ in which, E s is the modulus of elasticity of lateral
⎝ reinforcement. All units are in SI system.
ε cc 50 = ε co + 0.0911(αω w ) 0. 8

Kusuma-Tavio Model [15]

in which, α is a factor accounting for the
effectiveness of confinement, ω w the mechanical For ascending branch, ε c ≤ ε cc :
volumetric ratio of lateral reinforcement, bi the K bε b − ε b2
f c = f cc′ (24)
distance between two adjacent longitudinal
1 + (K b − 2)ε b
reinforcement measured center-to-center of
reinforcement, αω w the effective capacity of lateral For descending branch, ε c > ε cc :
reinforcement, and E p the secant modulus of
f c = f cc′ − Edes (ε c − ε cc ) (25)
elasticity of concrete at peak stress.
where:
Hong-Han Model [14]
Ecε cc
Kb =
For ascending branch, 0 < ε c ≤ ε cc : f cc′
⎧⎪ ⎛ ε ⎫⎪ ⎞
α
εc
f c = f cc′ ⎨1 − ⎜⎜1 − c ⎟⎟ εb =
⎬ (22) ε cc
⎪⎩ ⎝ ε cc ⎪⎭ ⎠
For descending branch, ε c > ε cc : Ec = 0.043wc1.5 f c′ (in MPa)

f c = f cc′ − Edes (ε c − ε cc ) (23) f le = 0.5 ke ρ s f yh

where:
⎛ ∑ bi2 ⎞⎛
2
⎞
f co′
3
ke = ⎜⎜1 − ⎟⎜1 − s ⎟
Edes = 0.026 ⎟⎜ ⎟
f le
0.4 ⎝ 6 bc d c ⎠⎝ bc ⎠
ε cc Edes =
12.2
α = Ec ρ s f yh ( f c′)2
f cc′
Ec = 3320 f co′ + 6900 f cc′
ε cu = ε cc +
0.70 2 Edes
f cc′ ⎛f ⎞
= 1.0 + 4.1⎜⎜ le ⎟⎟
f co′ ⎝ f co′ ⎠

83
 Tavio, T., et al.. / Effects of Confinement on Interaction Diagrams / CED, Vol. 11, No. 2, September 2009, pp. 78–88

in which, Edes is the strength reduction factor, ke a Effects of Confinement on Interaction

factor accounting for effectiveness of confinement, Diagram of Concrete Columns
and s the spacing of lateral reinforcement measured
center-to-center of reinforcement. To investigate the amount of capacity gain in axial
load and bending moment due to the confinement
effects, an analytical study is conducted on a column
Effects of Confinement on Stress-Strain
model with the following data: (a) unconfined
Curves of Concrete
concrete compressive strength, fc′: 30 and 60 MPa, (b)
cross section: width (B) and depth (H), 400 mm, (c)
The effects of confinement on stress-strain curves of
longitudinal reinforcement: 8, 19 mm diameter bars
concrete are investigated using a program developed
(ρt = 1.43 percent), (d) lateral reinforcement:
by the authors, namely ConfinedCOL v.1 [19], and
diameter 10 mm, (e) concrete cover 40 mm, (f)
the results are shown in Fig. 1. The presence of
spacing of lateral reinforcement 100 mm, (g) yield
closely-spaced lateral confinement significantly
strength of lateral reinforcement, fyh: 240 MPa, and
increases the magnitude of stress-strain curve of
(h) yield strength of longitudinal reinforcement, fyl:
concrete. Summary of the effects of confinement
240 MPa. The interaction diagram is also
parameters on the stress-strain curves of concrete
constructed using ConfinedCOL v.1 [19]. All models
according to several models proposed earlier are
discussed in the foregoing section are used to observe
given in Table 1. The most influencing parameter is
the effects of confinement of each model on the
found to be the spacing of transverse steel.
capacity gain of a column. From the results of the
analysis, it indicates that there is a capacity gain in
Effects of Confinement on Column Capacity axial load and bending moment of confined concrete
column compared to that of the unconfined one,
The effects of confinement directly influence the particularly in the compression-controlled region
shape and magnitude of stress-strain curve of shown in Fig. 2. For instance, the capacity gain is
concrete. This in turn will affect the compresive force shown in hatched region in Fig. 3 according to the
per unit width of concrete, cc. This gain further latest model [14]. This is due to the expansion of the
increases the compressive force of concrete, Cc, as area of compressive concrete in a column section
follows: with the presence of confinement produced by lateral
Cc = ccb (26) reinforcement. Note that for confined concrete
where cc is the compressive force of concrete per columns, the behavior of concrete core is modeled by
the stress-strain relationship of confined concrete,
unit width (N/mm), and b the width of compressive whereas for the cover, it is assumed as unconfined
section (mm). The increase of the compressive force concrete. The relevant stress-strain models are
of concrete (Cc) will automatically improve the adopted in the analysis to properly accommodate
nominal capacity of a column subjected to axial load both regions of concrete cross-section.
(P) and bending moment (M), or in other words, the
interaction diagram of the column is enlarged. Recent building code requires closely-spaced lateral
reinforcement in a reinforced concrete column to
The effects of confinement on the strength gain due satisfy the ductility and strength requirements of a
to the presence of confinement through the seismic-resistant building [20]. Even though the
requirement of minimum lateral reinforcement have codes ignore the effect of confinement on the
already been considered in the building code. strength gain due to the conservative consideration
However, this strength gain is used only for the for the design purposes, with the capacity gain due to
compensation for the possible strength loss due to confinement effects shown in the analysis, the
the spalling of concrete cover (which is unconfined). authors still expect that a reinforced concrete column
Recent codes of practice still disregard this effect for could resist higher axial load and bending moment
the design purposes and, thus in the conservative in the future design. Tables 2 and 3 show the
side. In the proposed analytical models, both the substantial capacity gains of confined concrete
strength gain in concrete core and the loss of columns compared to the unconfined one in terms of
strength in the cover are considered to exhibit the axial load and bending moment using the adopted
remaining strength gain after the mobilization of models after the mobilization of strength gain in the
strength gain in the core concrete to compensate the core concrete to compensate the loss of strength in
loss of strength in the concrete cover. For confined the concrete cover. The use of several stress-strain
concrete columns, the behavior of concrete core is model of unconfined concrete does not demonstrate
modeled by the stress-strain relationship of confined significant difference in terms of strength at this
concrete, whereas for the cover, it is assumed as state.
unconfined concrete.

84
 Tavio, T., et al.. / Effects of Confinement on Interaction Diagrams / CED, Vol. 11, No. 2, September 2009, pp. 78–88

Figure 1. Effects of confinement on stress-strain curves of concrete

Table 1. Summary of effects of confinement parameters on stress-strain curves of concrete

Confinement Parameters
Confinement Diameter of Spacing of Yield Strength Configuration of Number of Configuration of
Parameters
Model Lateral Lateral of Lateral Lateral Longitudinal Longitudinal
Reinforcement Reinforcement Reinforcement Reinforcement Reinforcement Reinforcement
Peak Stress – – – – – –
Kent-Park Peak Strain – – – – – –
Ultimate Strain + + – + – –
Peak Stress + + + + + +
Sheikh-Uzumeri Peak Strain + + + + + +
Ultimate Strain + + + + + +
Peak Stress + + + + + +
Mander et al. Peak Strain + + + + + +
Ultimate Strain + + + + + +
Peak Stress + + + + + –
Yong et al Peak Strain + + + + – –
Ultimate Strain + + + + + –
Peak Stress + + + + + +
Cusson-Paultre Peak Strain + + + + + +
Ultimate Strain + + + + + +
Peak Stress + + + + – –
Diniz-Frangopol Peak Strain + + + + – –
Ultimate Strain + + + + – –
Peak Stress + + + + + +
Kappos-
Peak Strain + + + + + +
Konstantinidis
Ultimate Strain + + + + + +
Peak Stress + + + + + +
Hong-Han Peak Strain + + + + + +
Ultimate Strain + + + + + +
Peak Stress + + + + + +
Kusuma-Tavio Peak Strain + + + + + +
Ultimate Strain + + + + + +
Notes: + = affecting, ─ = not affecting.

85
 Tavio, T., et al.. / Effects of Confinement on Interaction Diagrams / CED, Vol. 11, No. 2, September 2009, pp. 78–88

(a) f c′ = 30 MPa

(b) f c′ = 60 MPa

Figure 2. Comparisons between unconfined and confined interaction diagrams of concrete columns

86
 Tavio, T., et al.. / Effects of Confinement on Interaction Diagrams / CED, Vol. 11, No. 2, September 2009, pp. 78–88

Figure 3. Capacity gain of confined concrete columns in compression-controlled region

Conclusions axial load and bending moment by maintaining

its size without any enlargement, particularly for
From the study, the following conclusions can be lower-story columns which are dominated by the
drawn: axial load rather than flexure.
1. Three key parameters affecting the shape and 5. Further study needs to be carried out in the
magnitude of stress-strain curve of concrete are future, particularly in three dimensional models
the peak stress, the strain at peak stress, and the to capture the three dimensional cracking/
ultimate strain. fracturing behavior of concrete to confirm that
2. It can be concluded that there are six key the capacity gain of a column could be accounted
parameters primarily influence the effectiveness for in the future design codes with confidence.
of lateral confinement. The most influencing
parameter is found to be the spacing of transverse Acknowledgment
steel.
3. There is still a possible remaining capacity gain The authors would like to express their sincere
in axial load and bending moment of confined gratitude for the generous supports in terms of
concrete column compared to that of the facilities provided by Laboratory of Concrete and
unconfined one, particularly in the compression- Building Materials, Department of Civil Engineering,
controlled region, after the mobilization of Sepuluh Nopember Institute of Technology (ITS),
strength gain in the core concrete to compensate Surabaya, Indonesia.
the loss of strength in the concrete cover.
4. Even though, the codes ignore the effect of
confinement on the strength gain due to the
References
conservative consideration for the design purposes, 1. Villaverde, R., Methods to Assess the Seismic
with the remaining capacity gain found due to Collapse Capacity of Building Structures: State
confinement effects, the authors still hope that in of the Art, Journal of Structural Engineering,
the future design a more economical reinforced ASCE, V.133, No. 1, January 2007, pp. 57-66.
concrete column can be expected to resist higher

87
 Tavio, T., et al.. / Effects of Confinement on Interaction Diagrams / CED, Vol. 11, No. 2, September 2009, pp. 78–88

2. Whitney, C. S., Design of Reinforced Concrete 12. Diniz, S. M. C., and Frangopol, D. M., Strength
Members under Flexure or Combined Flexure and Ductility Simulation of High-Strength
and Direct Compression, ACI Journal, March Concrete Columns, Journal of Structural
1937, V. 33, No. 3, pp. 483-498. Engineering, ASCE, V. 123, No. 10, October
1997, pp. 1365-1374.
3. ACI Committee 318, Building Code
Requirements for Structural Concrete (ACI 318- 13. Kappos, A. J., and Konstantinidis, D.,
56), American Concrete Institute, Detroit, Statistical Analysis of Confined High-Strength
Michigan, 1956. Concrete Columns, Material and Structures, V.
32, Dec. 1992, pp. 734-748.
4. Departemen PU, Tata Cara Perhitungan
Struktur Beton untuk Bangunan Gedung (SK 14. Hong, K. N., and Han, S. H., Stress-Strain
SNI T-15-1991-03), Yayasan LPMB, Bandung, Model of High-Strength Concrete Confined by
1991. Rectangular Ties, Journal of Structural
Engineering, KSCE, V. 9, No. 3, 2005, pp. 225-
5. Purwono, R., Tavio; Imran, I., and Raka, I G. P., 232.
Tata Cara Perhitungan Struktur Beton untuk
Bangunan Gedung (SNI 03-2847-2002) 15. Kusuma, B., and Tavio, Unified Stress-Strain
Dilengkapi Penjelasan (S-2002), ITS Press, Model for Confined Columns of Any Concrete
Surabaya, March 2007. and Steel Strengths, Proceeding of the
International Conference on Earthquake
6. Fanella, D. A., Munshi, J. A., and Rabbat, B. G., Engineering and Disaster Mitigation, 14-15
Notes on ACI 318-99 Building Code Apr. 2008, Jakarta, Indonesia, pp. 502-509.
Requirements for Structural Concrete with
16. Popovics, S., A Numerical Approach to the
Design Applications, Portland Cement
Complete Stress-Strain Curve for Concrete,
Association, Skokie, Illinois, 1999.
Cement and Concrete Research, V. 3, No. 5,
7. Kent, D. C., and Park, R., Flexural Members 1973, pp. 583-599.
with Confined Concrete, Journal of Structural
17. Thorensfeldt, E., Tomaszewicz, A., and Jensen,
Division, ASCE, V. 97, No. ST7, July 1971, pp. J. J., Mechanical Properties of High-Strength
1969-1990. Concrete and Application in Design,
8. Sheikh, S. A., and Uzumeri, S. M., Analytical Proceedings of the Symposium Utilization of
Model for Concrete Confinement in Tied High Strength Concrete, Tapir, Trondheim,
Columns, Journal of the Structural Division, 1987, pp. 149-159.
ASCE, V. 108, No. ST12, Dec. 1982, pp. 2703- 18. Tavio, Budiantara, I N., and Kusuma, B.,
2722. Spline Nonparametric Regression Analysis of
9. Mander, J. B., Priestley, M. J. N., and Park, R., Stress-Strain Curve of Confined Concrete, Civil
Engineering Dimension, Journal of Civil
Theoretical Stress-Strain Model for Confined
Engineering Science and Application, V. 10, No.
Concrete, Journal of the Structural Division,
1, Department of Civil Engineering, Petra
ASCE, V. 114, No. ST8, Aug. 1988, pp. 1804-
Christian University, Surabaya, March 2008,
1825.
pp. 14-27.
10. Yong, Y. K., Nour, M. G., and Nawy, E. G.,
19. Tirtajaya, R., Analisis Penampang Kolom Beton
Behavior of Laterally Confined High-Strength Bertulang Menggunakan Visual Basic 6.0
Concrete under Axial Loads, Journal of the dengan Memperhitungkan Efek Pengekangan,
Structural Division, ASCE, V. 114, No. ST2, Final Project, Department of Civil Engineering,
February 1988, pp. 332-351. Sepuluh Nopember Institute of Technology
11. Cusson, D., and Paultre, P., Stress-Strain (ITS), Surabaya, August 2008.
Model for Confined High-Strength Concrete, 20. Purwono, R., and Tavio, Evaluasi Cepat Sistem
Journal of Structural Engineering, ASCE, V. Rangka Pemikul Momen Tahan Gempa, ITS
121, No. 3, March 1995, pp. 468-477. Press, Surabaya, September 2007.

88