ACI STRUCTURAL JOURNAL
Title no. 89-544
Effects of Transverse Reinforcement on
Seismic Performance of Columns
by Atorod Azizinamini, W. Gene Corley, and L. S. Paul Johal
Columns in building frames are normally designed to prevent both
hinging and crushing during an earthquake. However, columns in
buildings subjected to severe earthquakes may sometimes be sub-
jected to forces that cause hinging. The possibility of yielding occur-
ring at column ends makes it important to insure that columns are
capable of behaving in a ductile manner. To investigate the effects of
different transverse reinforcement details on the behavior of rectan-
gular and circular reinforced concrete columns, and to further eval-
uate simplification of some of the detailing requirements, 12 full-scale
columns were tested. Variables included the level of axial load,
amount and type of transverse reinforcement, and details of trans-
verse reinforcement.
Results indicate that flexural capacity of a column increases with
axial load but ductility reduces substantially. Increase in the amount
of transverse reinforcement results in increased ductility. Details of
transverse reinforcement, including hook bends and hook extensions,
can be further simplified. In all tests, limiting strain was well above
the 0.003 generally assumed in design.
Keywords: columns (supports); earthquake-resistant structures; loads (forces);
moments; reinforced concrete; tests.
With the exception of essential facilities, economic
considerations demand that most buildings located in
regions of high seismicity be provided with a resistance
considerably smaller than that required to remain elas-
tic under seismic forces resulting from a major earth-
quake. In such cases, reliance is placed on the available
ductility of the building members to dissipate energy
and avoid collapse.
Seismic-resistant design of reinforced ductile frame
buildings 1•6 provides a relative condition of a strong
column and weak beam at any junction. The intent is
to encourage hinging in the beams rather than in col-
umns. However, building performance during major
earthquakes7- 10 and results of analysis 11 - 14 indicate that
hinging can form in columns. Hinging can be attrib-
uted to several factors, including the following:
1. Design generally based on first mode of vibration,
whereas influence of higher modes of vibrations could
significantly alter the moment distributions among
frame members
442
TECHNICAL PAPER
2. Strain hardening of the flexural reinforcement in
beams
3. Bidirectionality of seismic forces
4. Differences between actual behavior of structures
during seismic events and mathematical modeling of the
building in design
Therefore, the possibility of plastic hinge formation
at column ends demands that building columns in seis-
mic areas have significant curvature ductility.
Previous research has shown that ductility of col-
umns can be improved with suitable confinement. 15 - 17
Testing has been directed toward understanding the
functions of lateral ties as confinement for the column
core, restraint against buckling of longitudinal rein-
forcement, and shear reinforcement.
Sakai and Sheikh 18 have summarized major research
conducted on the subject of confinement in concrete
columns. A general finding of most reported experi-
mental studies is that both concrete stress and strain
capacity are increased by an increase in the total
amount of transverse reinforcement, decrease in spac-
ing of transverse reinforcement, and increase in the
number of longitudinal bars located around the col-
umn perimeter and tied in the corner of a hoop bar.
In general, research conducted to evaluate the effects
of the transverse reinforcement arrangement and detail
on column performance has been limited.
RESEARCH SIGNIFICANCE
This paper describes the results of a test program to
investigate effects of different transverse reinforcement
types and associated details on hinging reinforced con-
crete columns subjected to loads simulating earthquake
effects.
ACI Structural Journal, V. 89, No. 4, July-August 1992.
Received June 20 1991, and reviewed under Institute publication policies.
Copyright © 1992, American Concrete Institute. All rights reserved, includin~
the making of copies unless permission is obtained from the copyright propn-
etors. Pertinent discussion will be published in the May-June.1993 ACI Struc-
tural Journal if received by Jan. I, 1993.
ACI Structural Journal I July-August 1992
Atorod Addntlmlnl is an assistant professor of civil engineering at the Univer-
sity of Nebraska-Lincoln. He holds a BS from the University of Oklahoma, and
MS and PhD in civil engineering from the University of South Carolina. He was
formerly on the staff of Construction Technology Laboratories in Skokie, Illi-
nois. His research activities include seismic behavior of ultra-high-strength
concrete columns, bond of reinforcing bars to high-strength concrete, and
composite columns. He is a member of ACI Committee 408, Bond and Devel-
opment of Reinforcement; and joint ACI-ASCE Committee 441, Concrete
Columns.

W. Gene Corley is Vice President of Construction Technology Laboratories. A

registered structural engineer, Dr. Corley has authored numerous technical
publications related to design and behavior of reinforced and prestressed con-
crete structures. He is Chairman of ACI Committee 318, Standard Building
Code; a member of ACI Committee 341, Earthquake-Resistant Concrete
Bridges; and of joint ACI-ASCE Committee 343, Concrete Bridge Design. He
has received many awards, including the Wason Medal for Research and the
Turner A ward from ACJ. Beam Stub
I
L. S. Paul Johal is Rese11rch Director at Precast/Prestressed Concrete Institute 'i:
(PC/), Chicago, Illinois. He was formerly with Walker Parking Consultants, !

Kalamazoo, Michigan, and the Construction Technology Laboratories of the

Portland Cement Association, Skokie, Illinois. He has extensive research, de-
sign, and evaluation experience, and has authored or coauthored several pub-
lications related to the structural use of concrete and masonry. He was coreci-
pient of the Prestressed Concrete Institute's Martin P. Kom Award for a paper
on structural research in 1975.

Test program
Eleven full-scale column specimens with square cross
sections and one with a circular cross section were
tested. Fig. 1 shows the overall dimensions. Each com-
plete column had a cross section of 18 x 18 in. (457 x
457 mm) and a height of 10ft 6 in. (3.20 m). The up-
per column constituted the test portion of each speci-
men. The test portion represents the column extending
upward from the beam-column connection to approxi-
mately the point of inflection. One full-scale specimen,
designated NC-11, was identical to the other eleven ex-
cept that the upper column's cross section was circular,
with an 18-in. (457-mm) diameter. The beam stub also
provided a point for application of lateral load and
strengthening of the joint. region, so that any hinging
occurred in the column rather than the joint. The
height of the upper column was greater than the height
of the lower column to force hinging into the upper
column. Except in Specimen NC-1, moment capacity of
the lower column was increased by providing addi-
tional vertical reinforcement.
ff
Support
Fig. /-Overall view of test specimen
Applied vertical load P. for each specimen is listed in
Table 1. The level of vertical load ranged from 20 to 40
percent of the column axial load capacity P 0 • Column
axial load capacity Po was determined in accordance
with the requirements of the 1983 ACI Building Code.
Fig. 2 shows the reinforcement details of a represen-
tative test column. Design compressive strength of con-
crete was 6000 psi (41.4 MPa). Specified yield stress of
vertical and transverse reinforcement was 60 ksi (41.4
MPa). Vertical reinforcement consisted of eight No. ·8
bars providing a reinforcement ratio of 0.0195.

Table 1 - Details of test variables

Verticalload Transverse reinforcement Confined
Specimen length of
designation P,IP.* kips DetaiJf Bar size A.,, in.' Percent column, in.
NC-1 0.30 570 A No.4 0.68 2.19 22
NC-2 0.20 380 B No.4 0.68 2.19 22
NC-3 0.40 780 B No.4 0.68 2.19 22
NC-4 0.30 580 B No.3 0.38 1.26 22
NC-5 0.30 575 c No.4 0.68 2.19 22
NC-6 0.30 520 D No.4 0.40 1.29 22
NC-7 0.30 540 E No.4 0.40 1.29 22
NC-8 0.30 560 Bl No.4 0.68 2.19 22
NC-9 0.30 530 F' No.4 0.40 1.29 22
NC-10 0.30 550 F** No.3 0.22 1.29 22
NC-11 0.30 460 G Ml in. spiral - 1.29 34
NC-12 0.30 554 F No.3 0.30 1.29 22
• P. = 0.85 J; (A, - A.,) + A., + J,; P, = vertical column load.
tDetails of transverse reinforcements shown in Fig. 3.
1Detail B modified by staggering inner and peripheral hoops.
'Detail F with no cro~s tie and No. 4 continuous square helix at 4-in. (100-mm) pitch.
**Detail F with no cross tie and No. 3 continuous square helix at 2\4-in. (57-mm) pitch.
I kip = 4.45 kN; I in. = 25.4 mm.

ACI Structural Journal I July-August 1992 443

 NC-5, and NC-8, hook bends for the inner hoops were
reduced to 90 deg, as shown in Fig. 3(b). Specimen NC-
5 also used overlapping peripheral hoops, as shown in
Fig. 3(c). Specimen NC-6 used single peripheral hoops
with 135-deg hook bends and six bar diameter exten-
sions, as shown in Fig. 3(d).
+-+--+ Transverse reinforcement for Specimen NC-7 also
+-+---+ consisted of single peripheral hoops. Each of these

lfil
-+ ~M<.~:'ed
Section A-A Section A-A
hoops was formed with four identical ties, as shown in

1
SQUARE TEST CIRCULAR TEST
COLUIINS COLUIINS
•-<
D t±::±
13h.:':P: Transverse
Reinforcement
Fig. 3(e). Specimen NC-8 used an arrangement shown
in Fig. 3(b), except that the inner and peripheral hoops
+-+--+ Details Vary were staggered vertically to provide a 2-in. (50-mm)
+-t-+ center-to-center spacing between them.
+-t-+
4'-0" I I I
+-t-+ Transverse reinforcement for Specimen NC-9 con-
j_- 308"

}==-o==!~
Note 1:- Lower column had
additional vertical
reinforcement not
shown in this Fig.
sisted of a No. 4 continuous square helix at 4-in. (100-
mm) pitch. Specimen NC-10 used a No. 3 continuous

t-~
2:- Lower column had
an IB"x 18" square
cross section for
all test columns.
Fig. 2-Reinforcement details for square columns (1 in.
= 25.4 mm)
square helix at 2!14-in. (57-mm) pitch. As shown in Fig.
3(f), Specimen NC-12 used a No. 3 continuous square
helix at 3Yz-in. (89-mm) pitch with No. 3 crossties at
3 Yz in. (89 mm) on center.
Transverse reinforcement for Specimen NC-11 con-
sisted of %-in. (10-mm) diameter continuous spiral re-
inforcement at a pitch of 2!14 in. (57 mm).
The length of column confined by transverse rein-

D 90"Bend
6 bar-die..
Extension
forcement was kept constant at 22 in. (0.56 m). The
confined length for Specimen NC-11 was 34 in. (0.86
m). Except for Specimens NC-8, NC-9, NC-10, and
NC-12, hoops were spaced 4 in. (100 mm) on center in
the confined region. Transverse reinforcement in the
unconfined region of column was designed to carry
(a) Detail A (b) Detall B
maximum shear stress. Clear cover over vertical rein-
90" Bend 6 bar-dla.
Extension forcement was maintained at 1.5 in. (38 mm) in upper
Overlappina and lower columns.

•a
Peripheral
Hoop

(c) Detail C (d) Detail D TEST SETUP

Continuous The test setup and loading arrangement are shown in
90 Bend Square Helix
Hoop
6 bar-dla. Fig. 4 and 5. A !-million lb (450-t)-capacity testing ma-
Extension No. 3 Cross
64 5"Bend[J
bar-dla. Tie a 3.5 in. chine was used to apply the vertical compressive force.
Extension on Center
Lateral load was applied with hydraulic rams pushing
(e) Detall E
against reaction frames. Several types of instruments
(f) Detail F
were used to obtain load displacement, moment curva-
Clear Concrete Cover • 1.5"
Hoop Spacina; = 4"' ture, vertical bar strain profiles, confining hoop strains,
plastic hinge lengths, and maximum concrete compres-
U:etrtc Equivalents:
1 ln. m 24.5 mm
sive strains.
1 ft. = 305 mm
(g) Detail G
Test procedure
Fig. 3-Detai/s of transverse reinforcement Each test was started by applying vertical load to the
column. During a test, this load was kept constant at a
Reinforcement details predetermined level. Horizontal force was applied in
Fig. 3 shows the transverse reinforcement types used increments, alternately in one direction and then in the
in the test specimens. The type used in each test col- opposite direction. The specimen was loaded to initial
umn is listed in Table 1. Transverse reinforcement for yielding in about three increments of horizontal force.
test specimen NC-1, as shown in Fig. 3(a), was de- Subsequent to initial yielding, loading was controlled by
signed in accordance with the provisions of Section deflection increments.
A.4.4. of the 1983 ACI Building Code. This required Basic loading cycles were generally applied as fol-
135-deg hook bends with 10 bar diameter extensions for lows: two cycles before yield, one cycle at yield, one
both inner and peripheral confining hoops. For all cycle at displacement ductility between one and two,
other specimens, hook extensions were reduced to six two cycles each at displacement ductility of two, and at
bar diameters. In addition, in Specimens NC-2 through subsequent displacement ductilities. Tes.ting was
444 ACI Structural Journal I July-August 1992
 Loading
Frame
+l.
T

t
4"-o"

~~~~~~1
Metric Equivalents Elevation
lin = 25.4 mm
1ft = 305 mm

Section A-A

Fig. 5-Loading arrangement

Bar No.1 Bar No.3

Fig. 4-Photo of the test specimen

stopped at a stage when the specimen could not sustain

the vertical load under increasing lateral displacement.
The displacement ductility ratio at a specific loading
cycle was defined as the maximum horizontal displace-
ment of column during that cycle divided by the maxi-
mum horizontal displacement of column at the loading
cycle resulting in first yield.

TEST RESULTS
Discrete transverse reinforcement
Specimens NC-1 through NC-8 employed discrete
transverse reinforcement. The plastic hinge formed at
the desired location in the upper column in all speci-
mens except for Specimen NC-1, where the plastic
hinge formed in the lower column. In general, all spec-
imens exhibited good strength and energy-dissipation
capabilities up to the last loading cycle. Specimens
failed by buckling of the longitudinal bars in compres-
sion, which coincided with the loss of anchorage of the
transverse reinforcement at one or more levels. Except
for Specimen NC-8, the inner hoops did not sustain any
visible anchorage loss. In Specimen NC-8, the inner
hoop located 7 in. above the stub lost anchorage at the
time of failure.
Loss of anchorage of peripheral hoops influenced
buckling of longitudinal bars. Fig. 6 shows the first
four layers of peripheral hoops and the corner bars for
Specimen NC-2. Fig. 6 also shows anchorage locations
ACI Structural Journal I July-August 1992
- - - - After Test Conclusion
- - Original Position
Fig. 6-Sketch of corner longitudinal bars before and
after testing, Specimen NC-2
and their final shape at the conclusion of testing. The
location of the anchorage was alternated at each level.
The length of the longitudinal bar engaged in buckling
was influenced by the anchorage location, as is evident
from Fig. 6. As the anchorages for the peripheral hoops
were lost, the laterally unbraced length of the longitu-
dinal bars increased, thereby expediting buckling. For
example, referring to Fig. 6, as anchorage locations
numbered 2 and 3 were lost, the longitudinal bar lo-
cated at the northeast corner lost complete lateral sup-
port at Point 1 and partial support at Point 2 due to the
loss of anchorage number 3.
445

Fig. 7- Test specimen NC-2 after test conclusion
A photograph of the hinging region of one specimen
utilizing discrete transverse reinforcement is shown in
Fig. 7. Observed lengths of hinging regions varied from
approximately 10 to 16 in. (254 to 406 mm).
Continuous transverse reinforcement
The upper column cross section for Specimens NC-9
through NC-12 was square. These columns utilized
continuous square helix transverse reinforcement.
Specimen NC-11 was the only specimen with a circular
upper column. It utilized continuous spiral transverse
reinforcement.
Specimens NC-9 through NC-12 showed good
strength and energy-dissipation capabilities. The pre-
dominant behavior of these three specimens can be
summarized as follows:
1. Compared to specimens with discrete transverse
446
reinforcement\ spalling of the cover concrete occurred
early and extended to higher distances above the stub at
the test's conclusion.
2. In general,•layers of transverse reinforcement lo-
cated close to the stub were still covered by concrete at
the test's conclusion. Height of nonspalled concrete
above the stub was approximately 2, 3, and 4 in. (51,
76, and 102 mm) for Specimens NC-9, NC-10, and NC-
12, respectively.
3. Specimens NC-9 and NC-10 failed by buckling of
the longitudinal bars located away from the corners.
The addition of crossties in Specimen NC-12 resulted in
simultaneous buckling of the middle and corner longi-
tudinal bars.
4. The continuous nature of transverse reinforce-
ment for these three specimens provided better con-
finement for core concrete than those with discrete
transverse reinforcement, and prevented spalling of
core concrete, particularly at the location of maximum
lateral bending of the buckled longitudinal bars.
5. Large elongations of the transverse reinforcement
were observed at the level of maximum lateral bending
of the buckled longitudinal bars. However, this did not
result in any visible cracking or fracturing of transverse
reinforcement at these levels.
Specimen NC-11 was the only one that had an upper
column with a circular cross section. The intent was to
compare behavior of the circular and square columns
with similar design moment capacity. Although the re-
sults of Specimen NC-11 could not be compared on a
one-to-one basis to the results of other specimens, some
comparisons between the general behavior of Specimen
NC-11 and Specimens NC-9, NC-10, and NC-12 can be
made. All four specimens utilized the same material,
number and size of longitudinal reinforcing bars, and
some form of continuous transverse reinforcement.
Some remarks and observations with regard to Speci-
men NC-11 can be made:
1. Specimen NC-11 failed prior to completion of the
first cycle of displacement at a ductility ratio of 10.
This level of ductility is higher than that of NC-9, NC-
10, or NC-12. The additional ductility can be attrib-
uted to the fact that the square helix transverse rein-
forcement of Specimens NC-9, NC-10, and NC-12 was
subjected to bending and axial tension while the circu-
lar spiral reinforcement of Specimen NC-11 was essen-
tially subjected to tension. Spiral reinforcement was
more efficient in confining core concrete, thereby re-
sulting in higher ductility. Cover concrete for NC-11
spalled off to 20 in. (508 mm) above the stub compared
to 24 (0.61 m) and 40 in. (1.0 m) in Specimens NC-9
and NC-10, respectively.
2. As with Specimens NC-9, NC-10, and NC-12,
transverse reinforcement located at the level of maxi-
mum lateral bending of the buckled longitudinal bars
experienced large elongations. However, no visible
cracking or fracturing of circular spiral reinforcement
was observed.
3. In the case of Specimens NC-9, NC-10, and NC-
12, at least the first layer of transverse reiiiforcement
ACI Structural Journal I July-August 1992
 DISPLACEMENT DUCTILITY Mall'lent
<Kip-ln.) i'B90
250 I 2 3 4 7 a HODIF"IED
KE:NT AND PARK - - - - - ·
SHEIKH

Curva.turl' ("I\ l Tonths)

Mn
ACI

-i'B9B

HORIZONTAL DISPLACEMENT <IN.l

Fig. 10-Moment versus curvature, Specimen NC-2
Fig. 8-Horizontalload versus displacement, Specimen
NC-2
Mo•ent
(Kip-ln.) 7999
I'IODIF'IOJ
KENT AND PARK -----·

DISPLACEMENT DUCTILITY SHEIKH

250 4 5

HORIZONTAL
LOAD
<KIPS! -3

Mn

{-
ACI

P-6 -......_._._
EFTECT

-250
Fig. 11-Moment versus curvature, Specimen NC-5

HORIZONTAL DISPLACEMENT <IN.l

Table 2 - Maximum calculated and measured
Fig. 9-Horizontalload versus displacement, Specimen horizontal load
NC-4
Maximum calculated Maximum measured
Specimen designation horizontal load horizontal load
NC-2 185 210
was still embedded in concrete at the test's conclusion. NC-3 190 235
NC-4 198 225
This was not true for Specimen NC-11. NC-5 196 230
NC-6 180 140
Horizontal load-displacement characteristics NC-7 187 215
NC-8 192 200
Fig. 8 and 9 show representative samples of horizon- NC-9 181 215
NC-10 187 210
tal load versus horizontal displacement obtained for the NC-11 ISO 165
test specimens. In these figures, the solid line parallel to NC-12 188 195
the horizontal axis represents the theoretical ultimate
lateral load, calculated using ACI 318-83 and neglect-
ing the P-!l. effect. However, if the P-!l. effect is in- Moment-curvature characteristics
cluded, the calculated maximum lateral load decreases Representative plots of moment versus curvature for
as lateral displacement increases. The dashed lines rep- Specimens NC-2 and NC-5 are shown in Fig. 10 and 11.
resent the envelope of the points corresponding to These plots include total moment at the upper column-
maximum lateral loads at increasing lateral displace- stub-beam interface. Calculated moment capacity Mn,
ment levels. For each displacement level, the maximum indicated in these figures, was determined using provi-
lateral load was calculated using ACI 318-83 provi- sions of the 1983 ACI Building Code for nominal mo-
sions. Maximum horizontal displacement ranged from ment strtmgth. Capacity-reduction factor q, was taken
five to nine times yield displacement. as 1.0. The curvatures in these plots are those obtained
In general, there was a drop in horizontal load-car- for the gage length within the first 4 in. (100 mm) above
rying capacity of the specimens at the onset of cover the stub. Curvature ductility was defined in a manner
concrete spalling. Table 2 lists the horizontal loads cor- similar to displacement ductility.
responding to ultimate moment capacity of the section, Fig. 10 and 11 also show the theoretical moment-
neglecting strain hardening and maximum applied hor- curvature relationships calculated using modified Park,
izontal load. Priestley, and Gill 19 and Sheikh and Uzumeri20 stress-
ACI Structural Journal I July-August 1992 447
 Table 3 - Test results
Measured Flexural strength, kip-in. M,
Specimen displacement
designation ductility Calculated M, MeasuredM, M, E,
NC-1 6 5184 6102 1.18 0.04
NC-2 8 4814 5773 1.20 0.063
NC-3 5 5198 6514 1.25 0.036
NC-4 5 5268 6152 1.17 0.036
NC-5 8 5226 6365 1.22 0.047
NC-6 5 4800 3651 0.76 0.036
NC-7 5 4950 5873 1.19 0.036
NC-8 6 5100 5351 1.05 0.085
NC-9 5 4810 5591 1.16 0.076
NC-10 5 4970 5379 1.08 0.065
NC-ll 9 4000 4543 1.14 0.134
NC-12 5 5000 5389 1.08 0.074

ps fy (MPa) 3. The presence of a stub column provides additional

o m m m
0.14 . - - - - - - - - , - : - - - - , . - - - - , - - - - , . - - ,
confinement to the adjacent column cross section. This
additional confinement tends to increase the flexural
• Current Data
K = 15forksi capacity of the column cross section at this location.
K = 100 for MPa
4. In developing their confined model, Sheikh and
Uzumeri assumed that, from the onset of loading, the
portion of the effectively confined concrete is some
• percentage of the core area. The percentage is a func-
0.08 tion of such parameters as arrangement and spacing of
I the lateral reinforcement. However, it is possible in the
Lower Bound
Eu early stages of loading that the full concrete area is ef-
0.06
• • 0 fective, particularly before the first observed crushing
of the cover concrete. In this study, the maximum mo-
D ment for test specimens occurred generally during the
0.04 B • 0
0 first cycle at the displacement ductility ratio of two. At
this stage, the first crushing of cover concrete was also
Ji 0 .ll.
observed. Even after first crushing, the longitudinal
D bars and transverse reinforcement were completely
covered by concrete. Only at displacement ductility ra-
tios of four and higher had spalling of cover concrete
extended to a stage where transverse reinforcement was
visible. These experimental observations suggest that
the size of the effective concrete area is not the same
ps fy (ksi) from the onset of loading to point of failure but,
rather, changes as the applied load increases.
Fig. 12-Effect of transverse hoop reinforcement on Measured displacement ductility, calculated flexural
limiting concrete strain capacity strength, measured flexural strength, and maximum
core compressive strain values for all tests are listed in
strain models for confined concrete. Similar behavior Table 3. The enhanced moment capacities can be par-
was observed for other test columns. The analyses show tially attributed to increased concrete strain capacity
an early peak of bending-moment capacity followed by provided by confinement of the core concrete by trans-
decreasing moment for increasing curvatures. The verse reinforcement. Of the specimens having axial
analysis line should define the maximum point of mo- loads equivalent to 30 percent of the column's axial
ment versus curvature for each cycle of the repeated re- load capacity, this enhanced moment capacity was
versals of loading used in this study. Typically, the ex- largest for Specimen NC-5. The only specimen for
perimental data exceed the capacity predicted by the which the ratio of maximum applied moment to calcu-
two analyses at increasing curvatures. The differences lated moment M. was less than 1 was Specimen NC-6.
between the analytical and experimental results could Fig. 12 shows the maximum core compressive strain
be explained, in part, as follows: for all specimens. The curve in Fig. 12 represents a
1. The confined concrete stress-strain model of lower bound to the test data.
Sheikh and UzumerF0 was developed based on concrete
columns under concentric loading. Thus, the effect of
strain gradient is not included in the model. CONCLUSIONS AND RECOMMENDATIONS
2. The confined stress-strain models of Kent, Priest- The following conclusions are based on comparison
ley, and Gill, and Sheikh and Uzumeri do not include of the test results with regard to each variable investi-
the effect of cyclic loading. gated. ·
448 ACI Structural Journal I July-August 1992
Axial load
For a constant amount of confinement, flexural ca-
pacity of a column increases with axial load, but duc-
tility is reduced substantially.
Details of transverse reinforcement
A comparison of test results from Specimens NC-1
and NC-5 indicates that the flexural capacity and duc-
tility of Specimen NC-5 were not reduced by the use of
overlapping peripheral hoops.
Specimen NC-7 failed before completion of the sec-
ond cycle at displacement ductility of 5. However, test
results indicate that the special hoops used in Specimen
NC-7 are capable of providing good strength and en-
ergy-dissipation capabilities associated with the dis-
placement ductility ratio of 5.
Staggering the inner and peripheral hoops in Speci-
men NC-8 resulted in lower flexural capacity than that
of Specimen NC-1. However, Specimen NC-8 still pro-
vided a displacement ductility ratio of 6 and failed be-
fore the completion of the second cycle at displacement
ductility ratio of 7.
Specimens NC-9, NC-10, and NC-12 used continu-
ous square helix transverse reinforcement. Specimens
NC-9 and NC-10 failed by buckling of the middle lon-
gitudinal bars. Thus, in Specimen NC-12, crossties were
added to provide additional support for middle longi-
tudinal bars. Specimen NC-12 failed by simultaneous
buckling of the corner and middle longitudinal bars.
All three specimens used the same volumetric ratio of
transverse reinforcement. A comparison of test results
from NC-9 and NC-10 indicates that, for the same
amount of transverse reinforcement, reducing the pitch
has no significant influence on flexural capacity. How-
ever, smaller pitch provides a slightly higher ductility.
Considering that Specimens NC-9, NC-10, and NC-12
failed prior to completion of two cycles at the displace-
ment ductility ratio of 6, it is concluded that the con-
tinuous square helix transverse reinforcement is not as
effective as the circular spiral reinforcement. This is
demonstrated by qualitative comparison of test results
from Specimen NC-11 with Specimens NC-9 through
NC-12. Reduced effectiveness can be attributed to the
mechanism by which the circular spiral and square he-
lix transverse reinforcement provide confinement for
core concrete. Circular spiral reinforcment is primarily
subjected to uniform tension, whereas square helix re-
inforcement is subjected to both bending and axial ten-
sion. Consequently, spiral reinforcement provides uni-
form confinement for core concrete, whereas square
helix reinforcement provides only partial confinement.
A qualitative comparison of the behavior of Speci-
mens NC-9, NC-10, and NC-12 with other specimens
utilizing discrete-type transverse reinforcement indi-
cates that continuous square helix reinforcement was
able to enclOSt( the core concrete more efficiently at the
test's conclusion. In general, the use of square helix
transverse reinforcement resulted in extensive cover
concrete spalling. It is recommended that the ACI
Building Code be modified to permit square and rec-
ACI Structural Journal I July-August 1992
tangular helical reinforcement designed on the same
basis as discrete transverse reinforcement.
Area of transverse reinforcement
The area of lateral reinforcement provided for the
square columns ranged between 46 and 97 percent of
the ACI 318-83 requirement. Although the area of pro-
vided lateral reinforcement for all test specimens was
less than the ACI 318-83 requirement, resulting dis-
placement ductilities exceed those implied by the code.
Comparison of results from Specimen NC-1 with
those of NC-4 and NC-7 indicates that the use of al-
most 50 percent less transverse reinforcement in these
two specimens resulted in only slightly lower ductility.
Maintained strength was also generally lower at all load
stages.
Hook bends of inner hoops
All specimens except NC-1 used 90-deg hook bends
for inner hoops. All specimens using inner hoops be-
haved satisfactorily. This indicates that a standard 90-
deg hook on the inner hoops is sufficient.
Hook extensions
Test results indicate a 10 bar diameter extension as
required by Section A.1 of the 1983 ACI Building Code
is not needed. Six bar diameter extensions, as required
by the 1989 ACI Building Code, produced displace-
ment ductilities exceeding those generally assumed in
design.
Limiting strain
Test results indicate that the limiting strain is well
above the 0.003 generally assumed in design, and that
Corley's eq~ation21 gives a considerably "lower bound"
estimate of the limiting strain. Under the loading con-
ditions considered, use of a value of limiting strain
higher than 0.003 appears to be appropriate for con-
fined concrete.
ACKNOWLEDGMENTS
Work reported in this paper was based on research jointly sup-
ported by the National Science Foundation under Grant No. CEE-85-
44164 and the Portland Cement Association. All tests were per-
formed in the structural laboratory of the Construction Technology
Laboratories, Inc. The authors would like to thank Donald W. Mus-
ser and Norman W. Hanson for their contributions on this project.
Any opinions, findings, and conclusions or recommendations are
those of the authors and do not necessarily reflect the views of the
National Science Foundation.
REFERENCES
1. ACI Committee 318, "Building Code Requirements for Rein-
forced Concrete and Commentary (ACI 318-89/ACI 318R-89),"
American Concrete Institute, Detroit, 1989, 353 pp.
2. "Uniform Builidng Code," International Conference of Build-
ing Officials, Whittier, 1988.
3. "Tentative Provision for the Development of Seismic Regula-
tions for Buildings," Publication No. ACT-3-06, Applied Technol-
ogy Council, Palo Alto, June 1978.
4. "Recommended Lateral Force Requirements and Commen-
tary," Seismology Committee, Structural Engineers Association of
California, San Francisco, 1980.
449
 5. "Code of Practice for the Design of Concrete Structures,"
Standards Association of New Zealand, Wellington, 1982 (Parts 1
and 2).
6. "Code for the Design of Concrete Structures for Buildings,"
Canadian Standards Association, Rexdale, 1984.
7. "Reducing Earthquake Hazards: Lessons Learned from Earth-
quakes," Publication No. 68-02, Earthquake Engineering Research
Institute, El Cerrito, Nov. 1986.
8. Rosenblueth, E., "Mexico Earthquake: A First Hand Report,"
Civil Engineering, ASCE, V. 56, No. 1, Jan. 1986, pp. 38-40.
9. Kreger, M. E., and Sozen, M.A., "Seismic Response of Impe-
rial County Services Building in 1979," Journal of Structural Engi-
neering, ASCE, V. 115, No. 12, Dec. 1989, pp. 3095-3111.
10. Ghosh, S. K., and Corley, W. G., "Behavior of Reinforced
Concrete Framing Systems," Proceedings of International Confer-
ence on the Mexico Earthquake-1985, ASCE, Sept. 1986.
11. Sharp, R. D., "Seismic Response of Inelastic Structures," PhD
dissertation, University of Canterbury, New Zealand, 1974.
12. Park, R., "Accomplishments and Research and Development
Needs in New Zealand," Workshop on Earthquake Resistant Rein-
forced Concrete Building Construction, University of California,
Berkeley, July 1977.
13. Bertero, V. V., and Zagajeski, S. W., "Optimal Inelastic De-
sign of Seismic-Resistant Reinforced Concrete Framed Structures,"
Nonlinear Design of Concrete Structures, CSCE-ASCE-ACI-CEB
International Symposium, University of Waterloo, Aug. 1979.
14. Akira, W., and Hirose, K., "Building Frames Subjected to 2-D
Earthquake Motion," Proceedings of ASCE Structures Con-
450
gress- Seismic Engineering, May 1989.
15. Richart, F. F.; Brandtzaeg, A.; and Brown, R. L., "Failure of
Plain and Spirally Reinforced Concrete in Compression," University
of Illinois Engineering Experimental Station, Bulletin No. 190, 1929.
16. Chan, W. W. L'"., "Ultimate Strength and Deformation of
Plastic Hinges in Reinforced Concrete Frameworks," Magazine of
Concrete Research, V. 7, No. 21, Nov. 1955, pp. 121-132.
17. Roy, H. E. H., and Sozen, M.A., "Concrete Ductility, Flex-
ural Mechanics of Reinforced Concrete," Proceedings of Interna-
tional Symposium, Miami, V. 1, Nov. 1964.
18. Sakai, K., and Sheikh, S. A., "What Do We Know about
Confinement in Reinforced Concrete Columns? (A Critical Review of
Previous Work and Code Provisions)," ACI Structural Journal, V.
86, No.2, Mar.-Apr. 1989, pp. 192-201.
19. Park, R.; Priestley, M. J. N.; and Gill, W. D., "Ductility of
Square-Confined Columns," Proceedings, ASCE, 108, ST4, Apr.
1982, pp. 929-950.
20. Sheikh, S. A., and Uzumeri, S. M., "Analytical Model for
Concrete Confinement in Tied Columns," Proceedings, ASCE, 108,
ST12, Dec. 1982, pp. 2703-2722.
21. Corley, W. G., "Rotational Capacity of Reinforced Concrete
Beams," Journal of the Structural Division, ASCE, 92, ST5, Pro-
ceedings Paper 4939, Oct. 1966, pp. 121-146.
22. Azizinamini, A.; Johal, L. S.; Hanson, N. W.; Musser, D. W.;
and Corley, W. G., "Effects of Transverse-Reinforcements on Seis-
mic Performance of Columns-A Partial Parametric Investigation,"
final report to the National Science Foundation, Washington, D.C.,
Construction Technology Laboratories, Inc., Sept. 1988, 500 pp.
I
ACI Structural Journal I July-August 1992