The Cyclic Behaviour of Reinforced Concrete Chimney

Sections with and without Openings

John L. Wilson*
Faculty of Engineering and Industrial Sciences, Swinburne University of Technology, Australia

(Received: 8 January 2008; Received revised form: 20 August 2008; Accepted: 9 December 2008)

Abstract: The most vulnerable feature of most existing tall reinforced concrete
chimneys when subject to severe earthquake excitation is the relative weakness of the
sections around the openings near the chimney base. This paper documents an
experimental study undertaken to investigate the cyclic behaviour of typical chimney
sections with openings orientated to be bending critical and shear critical and the
results are compared with previous tests with no openings. Significantly, no such
cyclic tests have been previously been completed. The experimental results presented
include; failure mode, over-strength factor, ultimate curvature, available ductility,
hysteretic behaviour and strain distribution and the results are compared with some
predictive analytical section models and design guidelines. The research indicated that
chimney sections with openings were not brittle, but were significantly less ductile
than sections without openings. The paper is relevant for both assessing existing
chimney stacks and for the design of new chimneys in the vicinity of openings.

Key words: chimney, stack, seismic, cyclic behaviour, experimental tests, reinforced concrete, thin walled sections.

1. INTRODUCTION studies, a series of code design recommendations were

Tall reinforced concrete chimneys have traditionally
been designed to resist earthquake excitation in the
elastic range in the belief that such structures,
characterised by a large diameter/thickness (D/t) ratio,
were inherently brittle with no redundancy. This has
resulted in reinforced concrete chimneys being
prohibitively expensive in regions of high seismicity. A
series of experimental tests which examined the cyclic
behaviour of chimney sections demonstrated that well
detailed chimneys were not brittle but possessed some
ductility (Wilson 2002, 2003). These tests
complimented some of the previous tests carried out on
circular hollow reinforced concrete sections where the
D/t ratio was much smaller than those used in typical
chimney sections (Omote 1975; Mokrin 1985; Regan
1981; Schober 1984; Whittaker 1987; Zhan 1990; Yeh
2001). The experimental tests were used to develop a
non-linear dynamic analysis procedure for evaluating
the inelastic response of tall chimneys. Based on these
developed and incorporated into the 2001 CICIND code
for the design of reinforced concrete chimneys (Wilson
2000). The recommendations included a reduction in the
elastic seismic loads by a factor of R=2 through
detailing for ductility and providing over-strength
around openings to prevent the formation of brittle
failure modes.
A literature review has indicated few failures of tall
reinforced concrete chimneys from earthquake
excitation. This is not unexpected given the
conservative design parameters traditionally specified
for the seismic design of chimney structures. The
Tupras chimney is one of the few examples of chimneys
that have failed under seismic loading (Huang 2004;
Kilic 2003). This refinery chimney stack was 115m tall
and failed in a brittle and catastrophic manner in the
vicinity of the openings at around one third height. A
number of researchers have studied the failure
theoretically using non-linear finite element techniques

*Corresponding author. Email address: jwilson@swin.edu.au; Fax: +61-3-9214-4882.

Advances in Structural Engineering Vol. 12 No. 3 2009 411

 The Cyclic Behaviour of Reinforced Concrete Chimney Sections with and without Openings

and concluded that the opening was lightly reinforced 1

and the failure was precipitated by the unzipping failure 4565

of the poorly detailed reinforcement splice connections 2450 2115

in a region where flexural yielding occurred. This paper
Steel anchor block
Steel tube R/C pipe
describes an experimental study commissioned to better
understand the complex behaviour around significant
chimney openings under extreme cyclic loading and the
results are compared with previous tests undertaken

1200
without openings. Analyses indicated that the section
was weaker with the openings orientated to be ‘bending’
critical, however there was considerable uncertainty
whether the section could successfully transfer the shear
forces with the large openings orientated to be ‘shear’ Steel strap connection
critical. Consequently, two tests were undertaken with
the openings orientated to be (i) bending critical and (ii) 250kN Hydraulic actuator R/C Pilecap
shear critical to study the behaviour.
Figure 1. Test arrangement
2. RESEARCH SIGNIFICANCE
The most vulnerable feature of most existing tall (a)

reinforced concrete chimneys when subject to severe

earthquake excitation is the relative weakness of the
sections around the openings near the chimney base. This
paper documents an experimental study undertaken to
investigate the cyclic behaviour of typical chimney
sections with openings orientated to be bending critical
and shear critical and the results are compared with
previous tests with no openings. Significantly, no such
cyclic tests have been previously been completed. The
experimental results presented include; failure mode,
over-strength factor, ultimate curvature, available
ductility, hysteretic behaviour and strain distribution and (b)
the results are compared with some predictive analytical
section models and design guidelines. The research is
relevant for both assessing existing chimney stacks and for
the design of new chimneys in the vicinity of openings.

3. EXPERIMENTAL TEST SET-UP

Circular hollow reinforced concrete specimens were
designed and constructed to investigate the inelastic
behaviour of chimney sections under severe cyclic
loading as shown in Figures 1 and 2. The reinforced
concrete pipes were assembled and configured as
horizontal cantilevers and tested by applying a cyclic
transverse load at the free end. The cantilever fixed end
support consisted of a 320mm thick reinforced concrete
block (which simulated a pilecap) rigidly connected to a Figure 2. Test set-up, (a) Test #1; (b)Test #3
steel anchor block which was securely fixed to the
laboratory strong floor. The outside diameter of the had 600mm wide openings, which is the largest opening
pipes was D = 1194mm with a thickness of t = 35–40mm size is permitted in the CICIND and ACI307 codes of
resulting in a diameter to thickness ratio of D/t = 30–35 practice (subtended angle of 60°). The 600mm long
which was representative of real chimneys. opening for Test #2 was located 300mm from the fixed
The control Test #1 had no openings (as shown in end and orientated to be bending critical (ie. the
Figure 2a) whilst Test #2 and Test #3 (both t = 40mm) openings were located in the region of maximum

412 Advances in Structural Engineering Vol. 12 No. 3 2009

John L. Wilson
Table 1. Summary of key parameters for Tests #1, 2 and 3
Parameter Test #1
Openings Nil
Length 4565 mm
External Diameter (D) 1194 mm
Thickness (t) 35 mm
D/t 35
Axial Force (N) 226 kN
Axial Stress (fc) 1.75 MPa
F′c 40 MPa
Rebar (pv) actual 0.94%
Rebar (pv) equiv. 400MPa 0.70%
Rebar stress fy 200 MPa
Rebar stress f3% 300 MPa
Rebar stress fu 400 MPa
bending stresses), whilst the 800mm long opening for
Test #3 was rotated 90° to be shear critical as shown in
Figure 2b. The 800mm long opening corresponded to 20
times the chimney thickness which is the maximum
allowable length under the ACI307 rules, for the design
of reinforced concrete chimneys.
An axial load of 226KN was applied using two 16mm
diameter prestressing wires placed symmetrically top and
bottom within the pipe void. The maximum length of pipe
which could be spun by the commercial pipe manufacturer
was 2.44m which was effectively reduced to 2.20m to
allow a development length in excess of 200mm for
casting the pipe reinforcement into the foundation. An
additional length of steel pipe was designed and fabricated
and the two pipes connected using twelve steel straps
which were bolted and epoxied to the steel and concrete
sections respectively as shown in Figures 1 and 2. The
4.60 metre length of hybrid pipe resulted in a shear span to
member diameter of 3.8, which was considered
representative of reinforced concrete chimneys.
The longitudinal steel for all tests consisted of 5.8mm
diameter deformed bars with properties and percentages
summarised in Table 1 and characteristic of typical
chimneys. The reinforcement was embedded 200mm in
the anchor block and no laps were used. The hoop steel
consisted of 4.8mm diameter bars placed in a helix at
80mm centres and corresponded to around ρh = 0.5%
which was greater than the 0.2% minimum typically
specified for chimneys. In addition, three extra rebars
were placed within a 100mm wide strip each side of the
opening in both the longitudinal and circumferential
directions in accordance with the CICIND and ACI307
recommendations. The key parameters for Tests #1, #2
and #3 are summarised in Table 1 for completeness.
The cyclic lateral load was applied to the cantilever
tube using a hydraulic actuator with a ±125mm travel
Advances in Structural Engineering Vol. 12 No. 3 2009
Test #2 Test #3
2 off 600 × 600mm 2 off 600 × 800mm
4600 mm 4345 mm
1194 mm 1194 mm
40 mm 40 mm
30 30
226 kN 226 kN
1.55 MPa 1.55 MPa
40 MPa 40 MPa
0.53% 0.53%
0.75% 0.75%
530 MPa 530 MPa
560 MPa 560 MPa
590 MPa 590 MPa
range. The behaviour of the reinforced concrete pipe
under pseudo static cyclic loading was monitored using
displacement transducers (LVDT), displacement dial
gauges, Demec gauges (Tests #2, 3), photogrammetry
(Test #3) and load cells. The pipe units were tested
using quasi-static cyclic load testing procedures and
cyclically displaced to increasing ductility levels of:
µ = ±0.75, ±1, ±2, ±3 etc, for Tests #1 and #2 and
incremented at a finer resolution of ±0.25% drift
increments for Test #3. The yield displacement
(displacement at µ = 1) was found by extrapolating the
measured secant stiffness at the lesser of either 0.75 of
the theoretical ultimate load or at the onset of
reinforcement yield (Priestley and Park 1987). From the
quasi-static cyclic load testing undertaken in this
research study on limited ductile structures, it is
recommended that the displacements be increased
directly using drift increments rather than ductility
increments to ensure a sufficiently fine resolution of tip
displacements for each cycle.
In constructing the reinforced concrete models, the
laws of similitude were followed so that the behaviour
of the reinforcement, concrete and overall model
reflected the characteristics of a full scale prototype
thus enabling the results to be directly scaled (Wilson
2002). In particular, deformed reinforcement and a
typical full scale concrete mix was used for the model
tests (with the restriction of a 10mm maximum
aggregate size) to avoid the unrealistic bond
characteristics associated with micro-concrete and to
ensure that the experimental results were representative
of equivalent full scale prototype tests. A full
description of the experimental investigation and test
results is provided in journal references Wilson (2002,
2003) (Test #1) and the CICIND reports by Wilson
(2003, 2006) (Tests 2 and #3).
413
 The Cyclic Behaviour of Reinforced Concrete Chimney Sections with and without Openings

Force (kN)
(a) 100 (b) 80.0

Force (kN)
80 60.0
60
40.0
40
20.0
20
0.0
0 −80.0 −60.0 −40.0 −20.0 0.0 20.0 40.0 60.0 80.0
−100 −80 −60 −40 −20 0 20 40 60 80 100 −20.0 Displacement (mm)
−20 Displacement (mm)
−40.0
−40
−60.0
−60

−80 −80.0

−100

(c)
Force (kN)
100
80
60
40
20
0
−100 −80 −60 −40 −20 0 20 40 60 80 100
−20 Displacement (mm)
−40
−60
−80
−100

Figure 3. Lateral force versus displacement, (a) Test #1; (b) Test #2, (c)Test #3

4. OVERALL CYCLIC BEHAVIOUR consistent with a member undergoing inelastic

4.1. Test #1
The chimney section with no openings (Test #1)
behaved as a tough and ‘ductile’ structure as distinct
from a ‘brittle’ structure as shown in Figure 3a which
plots the applied force versus tip displacement. The
specimen achieved a tip displacement of ±85mm and a
1.9% drift after 13 cycles. A series of cracks developed
along the length of the pipe, and opened and closed and
lengthened as the longitudinal strains increased on
subsequent cycles.
The hysteresis shape was stable with increasing
displacements associated with strain hardening of the
reinforcement and increasing bending moments. The
reduction in stiffness associated with an increase in drift
is characteristic of the closure of wide cracks, softening of
the concrete matrix and the softening of the reinforcement
due to the “Bauschinger” effect. The pinched shape of the
hysteresis loops is common for members with low axial
loads. The loading section of the hysteresis loop is
deformations with the tensile reinforcement yielding and
wide cracks developing. The member displays quite high
stiffness initially on unloading which is characteristic of
the concrete unloading elastically in the compression
zone. Once the concrete stress has reduced to zero the
cracks begin to re-open leaving a section with an
essentially continuously open crack around the
circumference. Consequently the stiffness is greater
reduced (characterised by the pinched hysteresis), since
the section stiffness is essentially provided by the
reinforcement only. The section stiffness increases again
on reloading as the cracks close and the concrete
contributes to the stiffness as the concrete compressive
stresses develop. The pinching effect is less severe on
members with a high axial load since the neutral axis
depth will be greater and a portion of concrete will remain
in compression at all times to carry the axial load.
This ductile behaviour was achieved through yielding
of the reinforcement in tension, rather than non linear

414 Advances in Structural Engineering Vol. 12 No. 3 2009

John L. Wilson
Force (kN)
100
80
60
40
20
−0.01
0
0.00 0.01 0.02 0.03 0.04 0.05 0.06
−20
−40
−60
−80
−100
Figure 4. Lateral force versus extreme fibre strain Test #1
Figure 5. Damage in vicinity of fixed end Test #1
compressive behaviour of the concrete. For example at
a drift of 1.9%, the average maximum tensile strain
measured over a gauge length of 200mm was 5% (with
crack widths in the order of 3mm) compared with a
compressive strain less than 0.3% as shown in Figure 4.
As the cyclic drift increased, the cover concrete
around the hoop and longitudinal reinforcement in
the vicinity of the crack began to steadily spall, as the
concrete was cycled back and fourth from extreme
tension to compression. The damage developed in a
zone of concrete approximately 350mm wide adjacent
to the opening and was characterised by circumferential
and diagonal cracking (refer Figure 5). The pipe was
deemed to fail at a tip displacement of 99mm and drift
of 2.2%) when two of the longitudinal steel bars which
had buckled on the previous cycle in compression
fractured in tension. The buckling of the reinforcement
bar occurred due to the loss of the concrete cover in the
vicinity of the opening combined with the reduced E
value from the ‘Bauschinger’ effect. The buckling of the
longitudinal steel resulted in further concrete spalling
Advances in Structural Engineering Vol. 12 No. 3 2009
and a sharp kink developing in the reinforcement. The
longitudinal steel then fractured at the kink when the
load direction was reversed and the reinforcement was
subjected to extreme tensile strains. Transverse
confinement steel at a spacing of 5db (where db is the
longitudinal steel diameter) would be required to
0.07
prevent the longitudinal reinforcement from buckling.
Strain
4.2. Test #2
The chimney section with openings orientated to be
‘bending’ critical performed in a similar manner to Test #1
except the behaviour could better be described as
‘limited ductile’ with an ultimate tip displacement of
±67mm and a 1.5% drift, after 5 cycles (refer Fig 3b).
The average maximum tensile strain measured over a
gauge length of 200mm was 1.2% compared with a
compressive strain less than 0.3%. The lateral force
versus longitudinal strain plot was similar to that of
Figure 4 except the maximum strain was 1.2% which
was considerably less than the 5.0% measured for Test #1.
The specimen was deemed to fail at a tip displacement
of 76mm and a drift of 1.6% with the buckling and
subsequent fracture of exposed reinforcement bars
adjacent to the opening in a similar manner to the failure
in Test #1 (refer Figure 6).
The local strains measured with Demec gauges were
generally symmetrical around the openings with
significant longitudinal strains adjacent to the edge of
the opening. The maximum longitudinal tensile strains
measured over a gauge length of 200mm were in the
order of 1.2% with localised strains varying widely in
the range 0.2%–4.0% (measured over a gauge length of
50mm) reflecting the crack pattern in the pipe. The
longitudinal strains in the pipe near the centre at each
end of the opening were negligible and indicated
significant local stress redistribution. The
circumferential strains and stresses around the openings
were minimal and generally less than the cracking
strength of the concrete resulting in minimal stresses
being transferred to the additional circumferential
reinforcement. As expected the diagonal strains at the
corners of the openings were significant with local
strains varying in the range 0.7%–4.0%.
Significant distortion of the section in the vicinity of
the openings was observed, with the edge of the opening
buckling on the compression face at drift levels of
around 1.1%. Figure 6(c) shows the formation of an
out-of-plane buckle, providing clear evidence that the
chimney section in the vicinity of the opening was
distorting locally and violating the assumption that
plane sections remain plane. As a result of the observed
distortions, the 3D photogrammetry technique was
developed and introduced in Test #3 to measure the
415
 The Cyclic Behaviour of Reinforced Concrete Chimney Sections with and without Openings

(a) (b)

(c)

Figure 6. (a) Overview of damage Test #2; (b) Compression failure damage Test #2; (c) Elastic buckling along free edge of opening,
Test #2

in-plane ovalling effects which could not be captured 4.3. Test #3

with the linear transducers used to measure the surface
strains. (The photogrammetry technique utilises a high
resolution digital SLR camera, scalebar, a network of
retro-reflective targets and sophisticated software to
interpret the data. Photographs of the targets are taken
before and after an event and the relative changes to
the 3D position of the targets is calculated using the
principles of triangulation from which average surface
strains can be estimated).
The chimney section with openings orientated to be
‘shear’ critical performed in a similar manner to Test #1
except the behaviour could better be described as
‘limited ductile’ with an ultimate tip displacement of
±78mm and a 1.8% drift after 8 cycles (refer Figure 3c).
The average maximum tensile strain measured over a
gauge length of 200mm was 1.2 % compared with a
compressive strain less than 0.3%. The lateral force
versus longitudinal strain plot was similar to that of

416 Advances in Structural Engineering Vol. 12 No. 3 2009

John L. Wilson
(a)
Deformed shape
Original shape
v2 v1
h1
h2
Figure 7. (a) Explosive compressive failure Test #3; (b) Ovalling
distortion of Test #3 pipe opening (Max distortion at 1.8% drift:
v2-v1 = 18mm and h2-h1 = 13mm)
Figure 4 except the maximum strain of 1.2% was
considerably less than the 5.0% measured for Test #1.
The specimen was deemed to fail at a tip
displacement of 87mm and a drift of 2.0% when the
concrete shell on the compression side (between the 2
openings) buckled as a wide column panel adjacent to
the fixed end in a sudden, brittle and explosive manner
after significant inelastic deformations (refer Figure 7a).
The shell buckling was precipitated by buckling of the
longitudinal steel bars which had become exposed
due to the deterioration of the concrete and bond in the
anchor block adjacent to the fixed end. Surprisingly, the
section did not fail in shear, although significant shear
distortion was observed around the openings resulting
from the 3D shell action. The shear distortion is also
demonstrated in Figure 7(b) which shows the elastic
ovalling deformation of the cross-section at the
openings from photogrammetry measurements at a drift
Advances in Structural Engineering Vol. 12 No. 3 2009
of around 1.8%. Further, the longitudinal strains
measured in the vicinity of the openings did not satisfy
the assumption of ‘plane sections remain plane’
providing further evidence of the local shear distortion.
5. EXPERIMENTAL AND THEORETICAL
PERFORMANCE OF CHIMNEY
5.1. First Cracking
The cracking strength was calculated based on a tensile
strength of concrete given by ft = α. F ′c where α =
0.5 - 0.75 and using simple elastic section analysis
theory. These values of α are considered conservatively
high values for concrete in direct tension and direct
flexure respectively. In this application the tensile
stresses were considered uniform across the wall
thickness of the pipe and consequently a value of α = 0.5
has been assumed, resulting in ft = 3.2MPa. Table 2
shows a comparison of the calculated and actual bending
moments at first cracking and show overall good
agreement, with some variation associated with the
uncertainty of the tensile cracking strength and some
local thickness variations associated with the fabrication
process of the pipe sections.
5.2. First Yield
The bending moment strength at first yield was
calculated based on the reinforcement bar yield stress
and using simple elastic section analysis theory ignoring
the tensile strength of the concrete. Table 3 shows a
comparison of the calculated and actual bending
moments at first yield and show overall good agreement
between the experimental and theoretical results.
For completeness, the measured yield curvatures and
yield tip displacements are presented in Table 4. The
measurements for Tests #2 and #3 indicate that the pipe
with openings was considerably more flexible and also
indicated that the yield stress of the reinforcement bar
was considerably higher than that used for Test #1.
5.3. Ultimate Bending Moment Strength
The ultimate moment capacity was calculated using
standard ultimate section analysis theory (assuming
plane sections remain plane and using the stress-strain
properties for concrete and steel reinforcement) and
using the design charts from the 2001 CICIND model
concrete code (similar method to that recommended in
ACI 307). The calculated moment values from the
design charts have been interpolated by using the
equivalent percentage of 400MPa reinforcement and by
multiplying the design values by 1.15 to account for the
partial safety factor incorporated in the charts.
A comparison of the experimental and theoretical
values is listed in Table 5 and indicates that the
417
 The Cyclic Behaviour of Reinforced Concrete Chimney Sections with and without Openings

Table 2. Cracking bending moment values for Tests #1, 2 and 3

Cracking moment Test #1 Test #2 Test #3

Experiment 237 kNm 108 kNm 161 kNm
Theory 210 kNm 110 kNm 186 kNm

Table 3. Yield bending moment values for Tests #1, 2 & 3

Yield moment Test #1 Test #2 Test #3

Experiment 224 kNm 194 kNm 239 kNm
Theory 220 kNm 215 kNm 310 kNm

Table 4. Experimental yield curvature and displacement values for Tests #1, 2 & 3

Parameter Test #1 Test #2 Test #3

Yield curvature 0.0012 m−1 0.0025 m−1 0.0022 m−1
Yield displacement 4.4mm 13mm 15mm

Table 5. Ultimate bending moment values for Tests #1, 2 and 3

Ultimate Moment Test #1 Test #2 Test #3

Experiment 369 kNm 292 kNm 391 kNm
Theory 310 kNm 300 kNm 400 kNm
CICIND Charts 326 kNm 250 kNm 250 kNm
Over-strength Factor 1.30 1.35 1.80

experimental strengths are significantly greater than the
theoretical values. The CICIND chart values for Tests #2
and #3 have been assumed to be the same and equal to
Mu = 250kNm, since the charts do not specify the
critical orientation of openings. The over-strength
values presented in Table 5 are based on the nominal
design values from the CICIND charts and have been
calculated using Test #2 as an example: Ω = 1.15 ×
292/250 = 1.15 × 1.17 = 1.35. The result is particularly
significant since it indicates that the section with
openings orientated to be ‘shear’ critical (Test #3) is
around 1.3 times stronger than the ‘bending’ critical
case (Test #2), and that the section is capable of
transmitting the shear forces around the wide opening
through 3D shell action. In addition, the experimental
results demonstrated that the usual ‘plane sections
remain plane’ assumption is not valid for calculating the
ultimate strength due to the significant stress
redistribution and shell action (this is the subject of a
separate investigation using non-linear finite element
analyses). The fact that the specimen with an opening
(Test #3) is stronger than the specimen with no opening
(Test #1) is a reflection of the different reinforcement
properties used in Test #1, and the fact that additional
reinforcement was placed around the openings.
5.4. Ultimate Curvature
The maximum ultimate curvatures and maximum
longitudinal tensile strains averaged over a 200mm
gauge length are listed in Table 6. The measurements
indicate that sections with openings develop maximum
strains and curvatures in the order of only one quarter
that of sections with no openings, showing the limited
strain capacity. Interestingly, the maximum drifts were
comparable for the 3 tests, indicating that the
considerable distortion around the openings contributed
significantly to the tip displacements for Tests #2 and 3.
The Test #1 curvature of 0.051 m–1 compares very well
with a theoretical value of 0.053 m–1calculated using
standard ultimate section analysis theory assuming
plane sections remain plane.
5.5. Ultimate Displacement
The observed damage and curvature ductility
distribution indicated that most of the inelastic curvature
occurred over a 350mm section (0.30 times the
diameter) near the base and in the vicinity of the openings
for all 3 tests. This concentration of curvature over a
defined length can be modelled as a plastic hinge with a
nominal plastic hinge length. Using the nominal plastic
hinge length and average curvature values the cantilever

418 Advances in Structural Engineering Vol. 12 No. 3 2009

John L. Wilson

Table 6. Ultimate curvature and maximum tensile strains for Tests #1, 2 and 3

Parameter Test #1 Test #2 Test #3

Ultimate curvature 0.051 m–1 0.012 m–1 0.012 m–1
Maximum strains 5.0% 1.2% 1.2%
Maximum drift 1.9% 1.5% 1.8%

Table 7. Plastic hinge lengths for Tests #1, 2 and 3

Plastic hinge length Test #1 Test #2 Test #3

Observed 350mm 350mm 350mm
Calculated 350mm 1700mm 2200mm

Table 8. Displacement contributions for Tests #1, 2 and 3

Displacement Test #1 Test #2 Test #3

Elastic 10mm 20mm 25mm
Shear ~0 30mm 42mm
Plastic 76mm 17mm 11mm
Total 86mm 67mm 78mm

tip displacement, ∆u, can be estimated assuming the
displacement consists of an elastic component equal to
the displacement at yield and a plastic component
representing deformations associated with the plastic
hinge as shown in Eqn 1:
∆u = F/Fy × ∆y + (φ – F/Fy φy) lp (l-0.5 lp)
F = Lateral force applied to the pipe system
Fy = Lateral force at yield
∆y = Yield deflection at Fy
φ = Average curvature in plastic hinge at F
φy = Yield curvature at Fy
lp = Nominal plastic hinge length
l = Length of cantilever to opening
This equation was applied to the test pipe data, and the
plastic hinge length back calculated so that the predicted
and actual tip displacements matched. The calculated
plastic hinge lengths are summarised in Table 7
and indicate an excellent correlation with the observed
damage region of Test #1 but totally unrealistic
correlations for Tests #2 and 3. Clearly this approach
based on ‘plane sections remaining plane’ and assuming
that significant inelastic behaviour is associated with
flexure only, does not provided a good match for chimney
sections with large openings where 3D distortions of
the section, shear effects and second order effects
contribute significantly to the overall ultimate displacement.
The approximate contribution of elastic, shear and
plastic deformations to the overall tip displacement for
the 3 tests is summarised in Table 8. The elastic and
plastic contributions were calculated from Eqn 1 using a
plastic hinge length of 350mm, whilst the shear
contribution was calculated as the additional
displacement needed to equal the total tip displacement.
6. CONCLUSIONS AND SUMMARY
1. Specimen #1 with no opening displayed ‘ductile’
(1)
behaviour, whilst specimens #2 and #3 with
openings displayed ‘limited ductile’ behaviour,
but importantly not ‘brittle’ behaviour’. All
specimens achieved a drift in the range of
1.5%–1.9%. The failure mode of the specimen
with openings orientated to be ‘shear’ critical
(Test #3), was a sudden and explosive buckling
of the compression panel adjacent to the opening
after considerable inelastic excursions, and
surprisingly not due to direct shear failure.
2. The limited ductile behaviour exhibited by all
the pipe specimens is a result of small
compressive strains in the unconfined concrete
and larger tensile strains in the ductile
longitudinal reinforcement. All specimens
developed a damage zone adjacent to the fixed
end or opening, consisting of a 350mm zone
(0.3D) of significantly cracked concrete which
opened and closed and slowly degraded as the
displacements increased. The ultimate
longitudinal strains and curvatures in
the specimens with openings were only 1/4 of
the values of the specimens with no openings.
3. The ultimate strength of the chimney section
with openings was around 1.35 and 1.8 times

Advances in Structural Engineering Vol. 12 No. 3 2009 419

 The Cyclic Behaviour of Reinforced Concrete Chimney Sections with and without Openings
stronger than the nominal strength predicted
from the CICIND (or ACI 307) design charts for
the openings orientated to be ‘bending’ and
‘shear’ critical respectively. The result highlights
that the section with openings orientated to be
‘shear’ critical has considerable overstrength
available and is around one third stronger
compared with the ‘bending’ critical orientation.
4. Significant distortion was experienced in the
section around the openings demonstrating that
the assumption of plane sections remaining
plane was violated. Consequently the tip
displacements could not be predicted from
simple models using the concept of a plastic
hinge due to the extensive deformation from
shear and 3D ovalling distortions. Further, clear
evidence of elastic buckling of the free edge
around the opening was observed at drift levels
in the order of ±1.1% for Test #2.
5. The experimental research has improved the
understanding of the inelastic cyclic behaviour
and actual strength of chimney sections with
large openings which is relevant for assessing
existing chimney stacks and for the design of
new chimneys. A further study is currently in
progress to compare the experimental results
with a detailed non-linear and inelastic finite
element analysis to model the complex 3D
behaviour in the vicinity of the openings where
the standard assumption of ‘plane sections
remain plane’ is violated.
ACKNOWLEDGEMENTS
Acknowledgements and appreciation are extended to
the CICIND organisation and University of Melbourne
for the financial assistance and particularly Grant Rivett
for the technical support with the experimental program.
Additional thanks to PhD students Kittipoon Rodsin and
David Heath for assistance with the experimental data
analysis and photogrammetry study for Test #3.
REFERENCES
ACI Committee 307. (1998). Standard Practices for the Design and
Construction of Cast-in-Place Reinforced Concrete Chimneys,
American Concrete Institute, Farmington Hills, Michigan, USA.
CICIND. (2001). Model Code for Concrete Chimneys - Part A: The
Shell (commentary), International Committee on Industrial
Chimneys, Switzerland.
420
Huang, W., Gould, P.L., Martinez, R. and Johnson, G.S. (2004).
“Non-linear analysis of a collapsed reinforced concrete
chimney”, Earthquake Engineering and Structural Dynamics,
Vol. 33, pp. 485–498.
Kilic, S.A. and Sozen, M.A. (2003). “Evaluation of effects of August
17, 1999, Marmara earthquake on two tall reinforced concrete
chimneys”, ACI Structural Journal, Vol. 100, pp. 357–364.
Mokrin, Z.A.R. and Rumman, W.S. (1985). “Ultimate capacity of
reinforced concrete members of hollow circular sections
subjected on monotonic and cyclic bending”, ACI Journal,
Vol. 82, No. 5, pp. 653–656.
Omote, Y. and Takeda, T. (1975). Experimental and Analytical
Study on Reinforced Concrete Chimneys, Japan Earthquake
Engineering Promotion Society, Tokyo.
Priestley, M.J.N. and Park, R. (1987). “Strength and ductility of
concrete bridge columns under seismic loading”, ACI Structural
Journal, Vol. 84 No. 2, pp. 61–76.
Regan, P.E. and Hamadi, Y.D. (1981). “Behaviour of concrete
caisson and tower members”, in Concrete in the Oceans,
Technical Report No. 4, CIRIA/UEG, Cement and Concrete
Association, Department of Energy, United Kingdom.
Schober, H. and Schlaich, J. (1984). “Ultimate strength of reinforced
concrete chimneys”, Proceedings of the CICIND 5th
International Conference, Essex UK, pp 37–42.
Whittaker, D. (1988). Seismic Performance of Offshore Concrete
Platforms, Report No. 88-1, University of Canterbury,
Department of Civil Engineering.
Wilson, J.L. (2000). “Code recommendations for the aseismic design
of tall reinforced concrete chimneys”, CICIND Report, Vol. 16,
No. 2, pp. 8–12.
Wilson, J.L. (2002). “Aseismic design of tall reinforced concrete
chimneys”, ACI Structural Journal, Vol. 99 No. 5,
pp. 622–630.
Wilson, J.L. (2003). “Earthquake response of tall reinforced concrete
chimneys”, Engineering Structures, Vol. 25, No. 1, pp. 11–24.
Wilson, J.L. (2003). “Investigation of the cyclic behaviour of
reinforced concrete chimney sections with openings: Part 1 –
Experimental study”, CICIND Report, Vol. 19 No. 2, pp. 9–15.
Wilson, J.L. (2006). “Investigation of the cyclic behaviour of
reinforced concrete chimney sections with openings orientated to
be shear critical”, CICIND Report, Vol. 21 No. 2, pp. 43–49
Yeh ,Y.K., Mo, Y.L. and Yang, C.Y. (2001). “Seismic performance
of hollow circular bridge piers”, ACI Journal, Vol. 98, No. 6,
pp. 862–871
Zhan, F.A., Park, R. and Priestley, M.J.N. (1990). “Flexural strength
and ductility of circular hollow reinforced columns without
confinement on inside face”, ACI Structural Journal, Vol. 87,
No. 2, pp. 156–166.
Advances in Structural Engineering Vol. 12 No. 3 2009