You are on page 1of 14

Journal of the Franklin Institute 348 (2011) 11631176

A new eccentricity-based simulation to generate


ultimate conned interaction diagrams for circular
concrete columns
Ahmed Abd El-Fattah, Hayder A. Rasheed

, Asad Esmaeily
Department of Civil Engineering, Kansas State University, Manhattan, KS 66506, USA
Received 31 May 2009; accepted 6 October 2009
Available online 25 October 2009
Abstract
The analysis of circular concrete columns using unconned concrete models is a well established
practice. However, there is a necessity to develop realistic analysis and design tools that predict the
extreme ultimate capacity of such columns since modern codes and standards like AASHTO LRFD
are introducing extreme load events. The increase in strength and ductility due to full axial
connement is not applicable to pure bending and bending plus axial load simply because the area of
effective conned concrete is reduced. The higher the eccentricity the smaller the compressed portion
of the conned core. Accordingly, the ultimate conned strength is gradually reduced from the fully
conned value f
cc
0
(at zero eccentricity) to the unconned value f
c
0
(at innite eccentricity) as a
function of eccentricity to diameter ratio. A numerical analysis algorithm is developed using the nite
layer procedure and the secant stiffness approach within a framework of incremental-iterative
moment of area computations. The resulting nonlinear section analysis requires radial loading in
which the eccentricity is kept constant or the axial load is proportional to the applied moment. The
results are compared with existing experimental data and the widely used Mander model to
benchmark the present predictions.
& 2009 The Franklin Institute. Published by Elsevier Ltd. All rights reserved.
Keywords: Conned concrete; Nonlinear analysis; Circular columns; Eccentric loading
www.elsevier.com/locate/jfranklin
0016-0032/$32.00 & 2009 The Franklin Institute. Published by Elsevier Ltd. All rights reserved.
doi:10.1016/j.jfranklin.2009.10.005

Corresponding author. Tel.: +1 785 532 1589; fax: +1 785 532 7717.
E-mail address: hayder@ksu.edu (H.A. Rasheed).
1. Introduction
The development of column interaction diagrams for unconned concrete is a standard
analysis procedure. However, the need to develop analysis tools for the conned ultimate
capacity of columns is evident. Modern codes and standards are introducing the need to
perform extreme event analysis [1]. In previous studies, various models were implemented
to assess the ultimate conned capacity of columns under concentric axial loads. On the
other hand, the effect of connement in case of the eccentric axial load is not investigated
in such models [2]. Therefore, it is pertinent to relate the strength and ductility of conned
concrete to the degree of connement utilization in a new model [2,3]. The more the
eccentricity, the less the connement engaged until the connement effect is reduced to a
negligible level at pure bending.
In this study, a new eccentricity based model is developed for gradual transition of the
ultimate strength and ultimate strain, of conned circular columns, in between full
connement effects based on the Mander model for concentric sections and negligible
connement effects for pure bending. The eccentricity model is incorporated in a numerical
procedure that combines radial loading, nite layer method, secant stiffness procedure and
moment of area concept to incrementally-iteratively generate the momentcurvature
response of the column up to failure using a spectrum of eccentricities that yield the
conned column interaction diagram.
2. Material model
Mander model [4] has been the model of choice in connement studies since it is well-
known, widely benchmarked and expressed in terms of a single equation that traces the
fully conned and unconned curves, Fig. 1. Accordingly, it lends itself to a relatively easy
adaptation to partial connement. The main equations describing the model are given
below:
f
c

f
cc
0
xr
r 1 x
r
1
where r=E
c
/(E
c
E
sec
), E
sec
f
cc
0
=e
cc
, x=e
c/
e
cc
, E
c
5000

f
c
0
p
where f
c
0
and E
c
are in
MPa. f
cc
0
is the strength of the fully conned concrete and e
cc
is the strain corresponding to
f
cc
0
which is dened as
e
cc
e
c
0
1 5
f
cc
0
f
c
0
1

2
cc cc c
cc
f
cc
f
c
f
cu cu


Fig. 1. Eccentricity-based stressstrain family of curves.
A. Abd El-Fattah et al. / Journal of the Franklin Institute 348 (2011) 11631176 1164
where f
c
0
is the strength of unconned concrete and e
c
0
is the strain corresponding to f
c
0
.
The strength f
cc
0
is dened by
f
cc
0
f
c
0
1:254 2:254

1
7:94f
l
0
f
c
0
s
2
f
l
0
f
c
0
!
3
It is evident from Eqs. (2) and (3) that f
cc
0
reduces to f
c
0
and e
cc
reduces to e
c
0
when the
conning pressure f
l
0
reduces to zero. The conning pressure f
l
0
is dened as
f
l
0

1
2
k
e
r
s
f
yh
4
where r
s
4A
spiral
=d
s
s, k
e
1 s
0
=2d
s
=1 r
cc
, f
yh
is the yield strength of the
hoop reinforcement. Now, the intermediate strength of the partially conned eccentric
column f
cc
is expressed in terms of the eccentricity to diameter ratio in the new eccentricity
model:
f
cc

1
1
e
H
f
cc
0

1
1
H
e
f
c
0
5
e
cc
e
c
0
1 5
f
cc
f
c
0
1

6
here e
cc
is the strain corresponding to f
cc
. The intermediate stressstrain equation is the
same as Eq. (1) replacing f
cc
0
with f
cc
and e
cc
with e
cc
, Fig. 1.
3. Assumption for analysis
The following assumptions are made in the analysis:
1. There is perfect bond between the longitudinal reinforcing steel bars and the concrete.
2. Strains along the depth of the column section are assumed to be distributed linearly.
3. Concrete in tension is neglected after cracking.
4. The conned concrete compressive stressstrain relationship is that given by Eqs. (1),
(5) and (6) as a function of the eccentricity.
5. The axial force and bending moment are always applied at the geometric centroid (mid-
depth) of the section and then transferred to the updated inelastic centroid by statics.
6. The steel stressstrain relationship is elasticperfectly plastic. However, strain
hardening is also allowed by assuming a bilinear model, Fig. 2.
su
f
sy

y
f
su

Without strain hardening


With strain hardening
s
E
Fig. 2. Stressstrain relationship of reinforcing steel.
A. Abd El-Fattah et al. / Journal of the Franklin Institute 348 (2011) 11631176 1165
7. The section is numerically divided into a nite number of thin layers each of which is
assumed to have a constant strain e
ci
and stress f
ci
within the layer. The layer also has a
variable width B
i
, Fig. 3.
4. Sectional properties
Considering the axial rigidity (EA) as the overall effective modulus of elasticity times the
area, the integrated form of this property is
EA
X
nlr
i1
E
ci
B
i
t
X
nbr
j1
E
sj
E
cj
A
sj
7
where E
ci
f
ci
=e
ci
, E
si
f
si
=e
si
, B
i
is the width of layer i and A
sj
is the area of rebar j. The
location of the inelastic centroid measured from the bottom of the section is based on the
ratio of the moment of axial rigidity (EAM) to the axial rigidity (EA):
Y
c

P
nlr
i1
E
ci
B
i
tY
i

P
nbr
j1
E
sj
E
cj
A
sj
Y
sj
EA
8
where Y
i
and Y
sj
are measured from the bottom of the section. The exural rigidity (EI) is
the overall effective modulus of elasticity of the section times the moment of inertia about
the inelastic centroid (Y
c
):
EI
X
nlr
i1
E
ci
B
i
tY
i
Y
c

X
nbr
j1
E
sj
E
cj
A
sj
Y
sj
Y
c

2
9
The internal axial force (F
x
) of the section and the internal bending moment about the
inelastic centroid (M
o
) are given by
F
x

X
nlr
i1
f
ci
B
i
t
X
nbr
j1
f
sj
f
cj
A
sj
10
M
o

X
nlr
i1
f
ci
B
i
tY
i
Y
c

X
nbr
j1
f
sj
f
cj
A
sj
Y
sj
Y
c
11
G
P
G
P
o

G
Y
c
Y

G
M
c
M
H
Fig. 3. Idealized section, strain distribution and applied forces.
A. Abd El-Fattah et al. / Journal of the Franklin Institute 348 (2011) 11631176 1166
5. Numerical analysis
The section properties corresponding to any applied load step of constant eccentricity (e)
(i.e. axial force P
G
and bending moment M
G
=P
G
e) are evaluated as follows:
1. The axial force is transferred to the current inelastic centroid and the effective moment
about this centroid is calculated:
M
c
M
G
P
G
Y
G
Y
c
M
G
P
G
H
2
Y
c

12
2. The section curvature and axial strain at inelastic centroid can fully dene the strain
distribution:
f
M
c
EI
; e
o

P
G
EA
13
3. For the new strain distribution, the section properties in Eqs. (7)(11) are re-calculated.
The internal moment about the geometric centroid is transferred back:
M
Go
M
o
F
x
Y
G
Y
c
14
4. Convergence to the new equilibrium position is satised when
Y
c
Y
c previous
Y
c

oY
Tol
P
G
F
x
P
G

oF
Tol
M
G
M
Go
M
G

oM
Tol
15
5. If Eq. (15) is not satised, the newly calculated Y
c
is used in Eq. (12) and steps (14) are
repeated. The rst version of the present method of analysis was presented by Rasheed
and Dinno [5] for analyzing rectangular sections with axial load and uniaxial bending.
6. Simulation results
The present simulation program is capable of generating column interaction diagrams
for standard unconned compression analysis, conned compression analysis with and
without clear cover removal as well as with and without accounting for strain hardening of
longitudinal reinforcement. For the sake of benchmarking and verifying the accuracy of
the present simulation software, the different analysis interaction diagrams are compared
with (i) experimental data, (ii) results from well-known software for unconned
compression and (iii) results from using the Mander model for the entire range of
eccentricities, which is expected to be un-conservative. The following cases of experimental
data are compared rst:
A. Abd El-Fattah et al. / Journal of the Franklin Institute 348 (2011) 11631176 1167
Case 1: Columns tested by Fatis and Shah [6] The rst case is two octagonal columns
examined by Fatis and Shah. The rst column has a diameter of 23.65 in, a clear cover of
0.8 in, longitudinal reinforcement of 16 # 8 bars. The concrete strength is 4.77 ksi and the
steel yield strength is 43.5 and 61.3 ksi for the longitudinal and spiral reinforcement,
respectively. The spiral diameter is 0.394 in and the spiral spacing is 2.76 in. Fig. 4 shows a
sketch of the cross section of the column.
The rst analysis graph shows three interaction diagrams as they compare with the
experimental data point. It is evident from Fig. 5 that the unconned curve is the
inner most one then the conned curve excluding the clear cover at a compression strain of
0.003 (reduced area curve) followed by the outer most one which retains the cover until the
end of the analysis (full area curve). The experimental data point proves in this case
that the three curves are accurate and conservative enough. On the other hand, if
strain hardening is considered with a secondary modulus after yielding of 5% of the
original modulus, the experimental point gets inside the conned curves with and without
concrete cover, which makes the conned analysis un-conservative in this case, Fig. 6. This
is a clear indication that the behavior prior to the failure of the column included very
minimal strain hardening effects in the longitudinal reinforcement, if any. This is
reasonable since the longitudinal reinforcement used is made of mild steel, see the material
properties above.
The second column of Fatis and Shah has the same parameters as those of the rst
column except for the slightly lower concrete strength (4.71 ksi), slightly larger spiral
spacing (2.95 in) and signicantly larger spiral diameter (0.63 in). The longitudinal steel
yield strength is the same while the spiral steel yield strength is signicantly lower (40.6 ksi).
It is important to note that the experimental point has about the same axial loadmoment
capacity as that of column 1 even though the yield strength of the spiral reinforcement is
quite different, Figs. 5 and 7. This is attributed to the location of the point on the
interaction diagram close to the balanced point that is slightly affected by this change. It is
evident from Fig. 7 that the conned curves are accurate and conservative enough with
respect to the experimental data point. On the other hand, if strain hardening is considered
with a secondary modulus of 5% of the original modulus, the experimental point gets in
between the conned curves with and without concrete cover spalling, which makes the
curve of the removed cover the realistic one to consider in analysis, Fig. 8.
Fig. 4. Fatis and Shah column cross section.
A. Abd El-Fattah et al. / Journal of the Franklin Institute 348 (2011) 11631176 1168
Case 2: Column tested by Mander et al. [7] Case 2 presents a circular column tested by
Mander et al. in pure axial compression. The column has a diameter of 19.68 in, a clear
cover of 0.98 in, longitudinal reinforcement of 8 # 9 bars. The concrete strength is 4.49 ksi
and the steel yield strength is 42.9 ksi for longitudinal reinforcement and 49.3 ksi for spiral
reinforcement. The spiral diameter is 0.472 in and the spiral spacing is 2.04 in. The three
interaction diagrams are shown in Fig. 9. The experimental data point matches very closely
the middle curve that accounts for the cover removal in the case of neglecting strain
0
500
1000
1500
2000
2500
3000
3500
4000
0
Moment (kip.inch)
F
o
r
c
e

(
k
i
p
.
)
0
2000
4000
6000
8000
10000
12000
14000
16000
0
Moment (kN)
F
o
r
c
e

(
k
N
)
Full Area
Reduced Area
Experimental Point
Unconfined Concrete
200 400 600 800 1000 1200
2000 4000 6000 8000 10000 12000
Fig. 5. Interaction diagrams of column 1 without strain hardening.
0
500
1000
1500
2000
2500
3000
3500
4000
4500
0
Moment (kip.inch)
F
o
r
c
e

(
k
i
p
.
)
0
5000
10000
15000
20000
0
Moment (kN.m)
F
o
r
c
e

(
k
N
)
Full Area
Reduced Area
Experimental Point
Unconfined Concrete
200 400 600 800 1000 1200 1400
2000 4000 6000 8000 10000 12000 14000
Fig. 6. Interaction diagrams of column 1 with strain hardening.
A. Abd El-Fattah et al. / Journal of the Franklin Institute 348 (2011) 11631176 1169
hardening. Fig. 10, on the other hand, shows that the three diagrams are spaced out a little
further, when accounting for strain hardening, while the experimental data point still very
closely matches the middle curve. This example shows once again that the cover removal
curve is the one to be used in conned analysis.
0
500
1000
1500
2000
2500
3000
3500
0
Moment (Kip.inch)
F
o
r
c
e

(
K
i
p
.
)
0
2000
4000
6000
8000
10000
12000
14000
0
Moment (KN.m)
F
o
r
c
e

(
k
N
)
Full Area
Reduced Area
Experimental Point
Unconfined Concrete
1000 2000 3000 4000 5000 6000 7000 8000 9000 10000
200 400 600 800 1000
Fig. 7. Interaction diagrams of column 2 without strain hardening.
0
500
1000
1500
2000
2500
3000
3500
0
Moment (kip.inch)
F
o
r
c
e

(
k
i
p
.
)
0
2000
4000
6000
8000
10000
12000
14000
Moment (kN.m)
F
o
r
c
e

(
k
N
)
Full Area
Reduced Area
Experimental Point
Unconfined Concrete
2000 4000 6000 8000 10000 12000
0 200 400 600 800 1000 1200
Fig. 8. Interaction diagrams of column 2 with strain hardening.
A. Abd El-Fattah et al. / Journal of the Franklin Institute 348 (2011) 11631176 1170
Case 3: Columns tested by Dodd and Cooke [8]: Case 3 illustrates circular columns
investigated by Dodd and Cooke, one almost in pure bending and two in combined
bending and axial load. The column has a diameter of 7.87 in, a clear cover of 0.43 in,
0
500
1000
1500
2000
2500
3000
0
Moment (kip.inch)
F
o
r
c
e

(
k
i
p
.
)
0
2000
4000
6000
8000
10000
12000
Moment (kN.m)
F
o
r
c
e

(
k
N
)
Full Area
Reduced Area
Experimental Point
Unconfined Concrete
1000 2000 3000 4000 5000 6000
0 100 200 300 400 500 600
Fig. 9. Interaction diagrams of case 2 without strain hardening.
0
500
1000
1500
2000
2500
3000
0
Moment (kip.inch)
F
o
r
c
e

(
k
i
p
.
)
0
2000
4000
6000
8000
10000
12000
0
Moment (kN.m)
F
o
r
c
e

(
k
N
)
Full Area
Reduced Area
Experimental Point
Unconfined Concrete
1000 2000 3000 4000 5000 6000 7000
100 200 300
400 500
600
700
Fig. 10. Interaction diagrams of case 2 with strain hardening.
A. Abd El-Fattah et al. / Journal of the Franklin Institute 348 (2011) 11631176 1171
longitudinal reinforcement of 18 bars of 0.236 in diameter. The concrete strength is 5.07 ksi
and the steel yield strength is 65.2 ksi for longitudinal reinforcement and 36.25 ksi for spiral
reinforcement. The spiral diameter is 0.157 in and the spiral spacing is 1.1 in. It is expected
that the connement effect is less pronounced than the previous cases since the spiral
diameter is signicantly smaller despite the tight spiral spacing. The three interaction
diagrams are shown, in Fig. 11, to be relatively close to each other and the experimental
data point matches very closely all curves in pure bending. The second and third points, on
the other hand, yield all curves accurate and conservative. Fig. 12 shows no signicant
difference, when accounting for strain hardening, with the experimental data points still
closely comparing to all three curves in a similar fashion.
To benchmark the unconned compression analysis against widely-used software, three
cases are compared against the corresponding results of CSI-Section Builder. The
parameters of the rst case are as follows: A column diameter of 36 in, a clear cover of
1.5 in, longitudinal reinforcement of 13 # 11 bars. The concrete strength is 4.0 ksi and the
steel yield strength is 60 ksi for longitudinal and spiral reinforcement. The spiral
reinforcement is #5 bars. The second case has a smaller column diameter with the
following parameters: A column diameter of 25 in, a clear cover of 1 in, longitudinal
reinforcement of 12 # 10 bars. The concrete strength is 4.0 ksi and the steel yield strength is
60 ksi for longitudinal and spiral reinforcement. The spiral reinforcement is #4 bars. The
third case has yet a smaller diameter of column with the following parameters: A column
diameter of 20 in, a clear cover of 1 in, longitudinal reinforcement of 10 # 8 bars. The
0
50
100
150
200
250
300
350
400
0
Moment (kip.inch)
F
o
r
c
e

(
k
i
p
.
)
0
200
400
600
800
1000
1200
1400
1600
0
Moment (kN.m)
F
o
r
c
e

(
k
N
)
Full Area
Reduced Area
Experimental Points
Unconfined Concrete
50 100 150 200 250 300 350 400 450 500
10 20 30 40 50
Fig. 11. Interaction diagrams of case 3 without strain hardening.
A. Abd El-Fattah et al. / Journal of the Franklin Institute 348 (2011) 11631176 1172
0
50
100
150
200
250
300
350
400
450
0
0
Moment (kip.Inch)
F
o
r
c
e

(
k
i
p
.
)
0
500
1000
1500
2000
Moment (kN.m)
F
o
r
c
e

(
k
N
)
Full Area
Reduced Area
Experimental Points
Unconfined Concrete
50
50
100
10
150 200
20
250 300
30
350 400
40
450 500
Fig. 12. Interaction diagrams of case 3 with strain hardening.
0
1000
2000
3000
4000
5000
6000
0
Moment (kip-in)
L
o
a
d

(
K
i
p
)
0
4448
8896
13344
17792
22240
26688
0
Moment (kN.m)
L
o
a
d

(
k
N
)
5000 10000 15000 20000 25000
564.9 1129.8 1694.7 2259.6 2824.5
case 1
CSI case 1
Case 2
CSI case 2
CSI case 3
Case 3
Fig. 13. Comparison of unconned analysis with CSI-Section Builder.
A. Abd El-Fattah et al. / Journal of the Franklin Institute 348 (2011) 11631176 1173
concrete strength is 4.0 ksi and the steel yield strength is 60 ksi for longitudinal and spiral
reinforcement. The spiral reinforcement is #4 bars. The comparison for the three
interaction diagrams of the three columns between the present program results and those
of Section Builder is excellent, as shown in Fig. 13.
To compare the predictions of the present analysis against the well-known and widely-
used Mander model [4], the complete interaction diagram is generated using the Mander
model, which is known to be valid for pure axial compression only, relative to that of the
present procedure. The parameters of the column examined are as follows: A column
diameter of 23.65 in, a clear cover of 0.8 in, longitudinal reinforcement of 16 # 8 bars. The
concrete strength is 3.857 ksi and the steel yield strength is 43.5 ksi for both spiral and
longitudinal reinforcement. The spiral diameter is 0.394 in and the spiral spacing is 1.97 in.
It is evident from this comparison, Fig. 14, that the present analysis is very accurate and
just conservative enough compared with the experimental point. On the other hand, the
Mander model is shown to be accurate as well but not conservative enough relative to the
experimental point.
To show this comparison when the connement is more pronounced, another example
column is analyzed by the two procedures. The parameters in this example are: A column
diameter of 12 in, a clear cover of 1 in, longitudinal reinforcement of 8 # 7 bars. The
concrete strength is 4.0 ksi and the steel yield strength is 60 ksi for longitudinal and spiral
reinforcement. The spiral reinforcement is #5 bars and the spiral spacing is 1 in.
It is evident from Fig. 15 that the present analysis is signicantly more conservative than
that of the Mander model which indicates that the eccentricity based simulation is
necessary especially in connement-critical columns.
0
500
1000
1500
2000
2500
3000
0
Moment (kip.inch)
F
o
r
c
e

(
k
i
p
.
)
0
2224
4448
6672
8896
11120
13344
0
Moment (kN.m)
F
o
r
c
e

(
k
N
)
Experimental Point
Mander Model
Eccentricity Model
1000 2000 3000 4000 5000 6000 7000 8000 9000 10000
112.97 225.94 338.91 451.88 564.85 677.82 790.79 903.76 1016.73 1129.7
Fig. 14. Interaction diagrams by the present and Mander model.
A. Abd El-Fattah et al. / Journal of the Franklin Institute 348 (2011) 11631176 1174
7. Conclusions
In this study, a new eccentricity-based model was developed to generate realistic
simulations for the behavior of conned concrete circular columns subjected to combined
axial force and uniaxial bending moment. The analysis takes into account the well-known
Mander model for the two extreme cases of pure axial compression (fully conned section)
and pure bending moment (fully unconned section) and generates a family of curves
for the eccentric cases in the transition in between. A nite layer numerical procedure is
extended to circular columns in which the secant stiffness approach is formulated within
a framework of incremental-iterative computations using the moment of area concept.
The conned analysis results are benchmarked against experimental data for a wide
spectrum of points ranging from pure axial compression to pure bending and passing
by the balanced point. The unconned analysis results were successfully compared with
the interaction diagrams generated by the widely-used software CSI-Section Builder.
Finally, the present conned analysis was compared against the widely-used Mander
model for the full range of eccentricities within the same framework of incremental-
iterative computations. It is concluded from this comparison that the present procedure is
accurate and more conservative than that of the Mander model. Accordingly, the present
procedure represents advancement to the state of the art in the area of conned concrete
analysis.
0
200
400
600
800
1000
1200
1400
0
Moment (kip.inch)
F
o
r
c
e

(
k
i
p
.
)
0
889.6
1779.2
2668.8
3558.4
4448
5337.6
6227.2
0
Moment (kN.m)
F
o
r
c
e

(
k
N
)
Mander Model
Eccentricity Model
200 400 600 800 1000 1200 1400 1600 1800 2000
22.75 45.5 68.25 91 113.75 136.5 159.25 182 204.75 227.5
Fig. 15. Comparison between the present and Mander model.
A. Abd El-Fattah et al. / Journal of the Franklin Institute 348 (2011) 11631176 1175
Acknowledgement
This work was developed under a research project KSU 07-21 funded by the Kansas
Department of Transportation. The encouragement and support of Kenneth Hurst and
John Jones is highly acknowledged.
References
[1] AASHTO AASHTO LRFD Bridge Design Specications, American Association of State Highway and
Transportation Ofcials, third ed., SI Units, 2004, ISBN 1-56051-251-2.
[2] A. Esmaeily, K. Lucio, Analytical performance of reinforced concrete columns using various analytical
models, in: Y. Xiao, S. Kunnath, W. Yi (Eds.), International Symposium of Conned Concrete, American
Concrete Institute, SP 238-6, 2002, pp. 95110.
[3] K. Fam, A. Flisak, B. Rizkalla, S. Experimental, Analytical modeling of concrete-lled ber reinforced
polymer tubes subjected to combined bending axial loads, ACI Structural Journal 100 (4) (2003) 529539.
[4] J.B. Mander, M.J.N. Priestley, R. Park, Theoretical stressstrain model for conned concrete, ASCE Journal
of Structural Engineering 114 (8) (1988) 18041826.
[5] H.A. Rasheed, K.S. Dinno, An efcient nonlinear analysis of R.C. sections, Computers and Structures,
International Journal 53 (3) (1994) 613623.
[6] A. Fatis, S.P. Shah, Predictions of ultimate behavior of conned columns subjected to large deformations,
ACI Journal 82 (4) (1985) 423433.
[7] J.B. Mander, M.J.N. Priestley, R. Park, Observed stressstrain behavior of conned concrete, ASCE Journal
of Structural Engineering 114 (8) (1988) 18271849.
[8] L.L. Dodd, N. Cooke, Capacity of circular bridge columns subjected to base excitation, ACI Journal 97 (2)
(2000) 297308.
A. Abd El-Fattah et al. / Journal of the Franklin Institute 348 (2011) 11631176 1176

You might also like