Вы находитесь на странице: 1из 9

Confinement of Columns and Wall Boundary Elements

Using Headed Bars


D. Mitchell, M.ASCE 1; W. D. Cook 2; Y. Deng 3; and J. Liu 4

Abstract: Two series of full-scale specimens were constructed and tested under pure axial load to investigate the effectiveness of confine-
ment provided by headed bars. One series consisted of columns and the other series represented boundary elements in shear walls. Each
Downloaded from ascelibrary.org by University of Exeter on 07/15/15. Copyright ASCE. For personal use only; all rights reserved.

series consisted of five specimens that included a specimen without any confinement, a specimen with closed ties, a specimen with seismic
hoops and crossties, a specimen with closed hoops and headed bars replacing the crossties, and a specimen that had all of the transverse
reinforcement replaced by headed bars. It was concluded that headed bars provided a viable solution when used to replace crossties,
offering similar confinement. In these monotonically loaded tests, it was also found that headed bars could be used to replace entirely
the hoops and the crossties, offering similar confinement effects. DOI: 10.1061/(ASCE)ST.1943-541X.0000850. © 2013 American Society
of Civil Engineers.
Author keywords: Headed bars; Crossties; Transverse reinforcement; Concrete and masonry structures.

Introduction Ghali 2002) and walls (Dilger and Ghali 1997) containing headed
reinforcement.
Transverse reinforcement in reinforced concrete columns provides
confinement to the concrete in compression, prevents premature
buckling of the longitudinal compression bars, and serves as shear Test Program
reinforcement. Under seismic loading, spalling of the concrete
cover could be extensive and the strength and ductility of members Five large-scale columns (500 × 500 × 2,000 mm) and five wall
would depend greatly on the effectiveness of the confinement pro- boundary elements (530 × 400 × 2,000 mm) were tested under
vided to the concrete core. The use of headed bars as transverse concentric compressive loading. The clear covers for the column
reinforcement offers many advantages. Heads that are properly and the boundary element specimens are 40 and 20 mm, respec-
connected and adequately sized can enable the development of tively, measured to the transverse reinforcement or outermost sur-
the tensile strength of the bar at the extremities of the bar, unlike face of the heads of the headed bars. All column specimens were
hooked bars. Using headed reinforcing bars instead of hooked bars, constructed using 8–25M (db ¼ 25.2 mm, As ¼ 500 mm2 ) longi-
not only reduces reinforcement congestion, but also can make tudinal bars, while the boundary element specimens used 10–20M
construction of the reinforcing cages easier. In addition, the con- (db ¼ 19.5 mm, As ¼ 300 mm2 ) longitudinal bars. Details of the
tribution of the head size and the rigid head-to-bar connection transverse reinforcement and the corresponding ACI code clauses
can increase the reinforcement’s confining effect. The research re- (ACI 2011) are summarized in Fig. 1. Specimens C1 and BE1 were
ported in this paper is aimed at investigating the confining effect of constructed without any transverse reinforcement. Specimens C2
friction-welded headed bars serving as transverse reinforcement in and BE2 represent a column and a boundary element with poor
columns and boundary elements of walls subjected to monotonic confinement, such as would be found in an older structure.
axial compressive loading. Although there has been a significant Specimen C2 contained No. 5 (db ¼ 16 mm, As ¼ 200 mm2 )
number of experimental studies on columns containing hoops, ties, hoops with 90° bend anchorages at 400-mm spacing as required
and crossties (e.g., Sheikh and Uzumeri 1980; Mander et al. 1984, for nonductile columns. Specimen BE2 contained U-shaped
1988; Razvi and Saatcioglu 1994; Cusson and Paultre 1994, 1995; No. 4 (db ¼ 12.7 mm, As ¼ 129 mm2 ) ties at a spacing of
Légeron and Paultre 2003; Sharma et al. 2005), there are only a 500 mm. The ends of the U-bars had a 300-mm splice in the short
few experimental results reported on columns (Youakim and direction of the boundary element. It is noted that this detail does
not conform to the requirements of the 2011 ACI code. These
1
Professor, Dept. of Civil Engineering and Applied Mechanics, McGill hoops have an excessive spacing as well as poor end anchorage
Univ., Montreal, QC, Canada H3A 0C3 (corresponding author). E-mail: that will become ineffective once spalling of the concrete cover
denis.mitchell@mcgill.ca occurs. Column C3 was designed and detailed as a ductile column
2
Research Associate, Dept. of Civil Engineering and Applied in a special moment frame. The details of boundary element
Mechanics, McGill Univ., Montreal, QC, Canada H3A 0C3. BE3 conform to the ACI code requirements for special structural
3
Software Engineer, Microsoft, One Microsoft Way, Redmond, walls with a calculated maximum extreme fiber compressive
WA 98052-6399. stress less than 0.2 times the specified concrete compressive
4
Structural Engineer, Sargent & Lundy, LLC, 55 E. Monroe St.,
strength. The consecutive crossties were alternated end-for-end
Chicago, IL 60603.
Note. This manuscript was submitted on August 20, 2012; approved on
along the longitudinal reinforcement in both Specimens C3
April 8, 2013; published online on April 10, 2013. Discussion period open and BE3.
until April 20, 2014; separate discussions must be submitted for individual Specimens C4 and BE4 are companion specimens to Spec-
papers. This paper is part of the Journal of Structural Engineering, imens C3 and BE3 with friction-welded headed bars replacing
© ASCE, ISSN 0733-9445/04013080(9)/$25.00. the crossties. No. 5 and No. 4 headed bars had circular heads,

© ASCE 04013080-1 J. Struct. Eng.

J. Struct. Eng. 2014.140.


Column Specimens Boundary Element Specimens Confinement
(500 mm x 500 mm) (530 mm x 400 mm)

C1 BE1 No confinement

C1 & BE1: No
8-25M bars 10-20M bars transverse reinforcement
(typ.) (typ.)

56 mm cover 33 mm cover

C2 BE2 Poor confinement


Downloaded from ascelibrary.org by University of Exeter on 07/15/15. Copyright ASCE. For personal use only; all rights reserved.

No.5 ties C2: ties (ACI 7.10.5.2)


@400 mm No.4 ties BE2: nominal ties
with 90° 6db @500 mm
extensions

40 mm cover 20 mm cover

C3 BE3 Seismic confinement


No.5 ties No.4 ties
@125 mm with @120 mm with
C3: Seismic hoops &
90° and 135° 90° and 135°
crossties (ACI 21.6.4)
6db extensions 6db extensions
BE3: Seismic hoops &
crossties (ACI 21.9.6.3)
No.5 hoops No.4 hoops
@125 mm @120 mm

40 mm cover 20 mm cover

C4 BE4 Seismic confinement


No.5 headed No.4 headed
bars @ 125 mm bars @120 mm C4: Seismic hoops &
with circular with circular headed crossties
heads, φ=50 mm heads, φ=40 mm (ACI 21.6.4)
BE4: Seismic hoops &
No.5 hoops No.4 hoops headed crossties
@125 mm @120 mm (ACI 21.9.6.3)
40 mm cover 20 mm cover

C5 BE5 Seismic confinement


No.5 headed No.4 headed C5: headed bars
bars @125 mm bars @120 mm BE5: headed bars
with circular with circular
heads, φ=50 mm heads, φ=40 mm

40 mm cover 20 mm cover
over heads over heads

Fig. 1. Specimen details

having diameters of 50 and 40 mm for C4 and BE4, respectively. the four corners. Strains were measured using electrical resistance
For Specimens C5 and BE5, the transverse reinforcement consisted strain gauges glued to the hoops, crossties, and vertical bars near
of only friction-welded headed bars, including the peripheral the midheight of the specimens.
reinforcement. These specimens are companion specimens to Thin layers of high-strength capping compound were used to
C3 and BE3 that contained the conventional seismic hoops and cap the top and bottom of each specimen to ensure that the speci-
crossties. men was properly aligned and to provide uniform stress distribu-
The specimens were tested using an 11,400-kN capacity tions at the column ends. A loading rate of 4 kN per s was used up
computer-controlled universal testing machine. The load cell of to 4,000 kN and then displacement control at a rate of 0.0025 mm
the testing machine measures the axial loads applied to the spec- per s was used. The testing was continued until the specimen’s
imens. A pair of linear voltage differential transformers (LVDTs) resistance dropped to 50% of the peak load or the compressive
were placed on the front and on the back faces of each specimen, displacement reached 30 mm. Fig. 2 shows a column in the testing
over a gauge length of 1,500 mm, to measure the shortening at machine.

© ASCE 04013080-2 J. Struct. Eng.

J. Struct. Eng. 2014.140.


Material Properties 60,000 psi (414 MPa). The No. 5 and No. 4 headed bars had
two friction-welded heads having diameters of 50 and 40 mm
Concrete and thicknesses of 12.7 and 10.2 mm, respectively. The head sizes
were chosen such that the head provided an area of 10 times the bar
Two different batches of ready-mix concrete, both with a specified area, resulting in a net bearing area of the head of 9 times the
28-day compressive strength of 30 MPa, were used for the column bar area enabling the development of the tensile strength of the
and boundary element series. Standard cylinders and flexural bar at the head (Ghali and Dilger 1998; CSA 2004). It is noted that
beams cast with each series were given the same curing regime the head sizes used are larger than the minimum requirements of the
as the test specimens (moist cured for 5 days). The concrete proper- 2011 ACI code. The mechanical properties of the reinforcing steel
ties were determined during the middle of the testing period at an are summarized in Table 2.
age of 35 days. The compressive strength, fc0 , and compressive
stress-strain relationships were determined from three 150-mm
diameter by 300-mm cylinders for each batch. The modulus of rup-
ture, fr , was determined from three 150 × 150 × 600-mm long
Test Results
Downloaded from ascelibrary.org by University of Exeter on 07/15/15. Copyright ASCE. For personal use only; all rights reserved.

flexural beams subjected to third-point loading over a span of Fig. 3 shows the total load versus average strain responses for the
450 mm. The splitting tensile strength, f sp , was determined from column and boundary element specimens. Four vertical LVDTs
three 150 × 300-mm cylinders. Table 1 summarizes the concrete
were used to obtain the average longitudinal strains. All specimens
properties.
behaved in a similar manner in the ascending portion of the loading
response until spalling of the concrete cover occurred. Unconfined
Reinforcement Specimens C1 and BE1 experienced a sudden drop off of load once
The 25M and 20M longitudinal bars were hot-rolled, weldable
grade 400 (minimum yield of 400 MPa) steel. The transverse
reinforcement was constructed from weldable grade No. 4 and Table 2. Properties of Reinforcing Steel
No. 5 U.S. bars with a minimum specified yield stress of Diameter Area fy εy εsh f ult εult
Bar (mm) (mm2 ) (MPa) (%) (%) (MPa) (%)
25M 25.2 500 426 0.213 1.55 539 17.25
20M 19.5 300 456 0.228 2.43 580 23.17
Number 5 16.0 200 483 0.241 1.10 686 12.60
Number 4 12.7 129 472 0.236 1.42 669 21.85

(a)

Fig. 2. Test setup

Table 1. Concrete Properties


f c0 εc0 f sp fr
Specimens Variability (MPa) (%) (MPa) (MPa)
Column Average 38.1 0.21 3.01 4.56 (b)
Standard deviation 1.08 0.01 0.33 0.17
Boundary Average 31.0 0.21 2.45 3.56 Fig. 3. Load versus average strain relationships: (a) column series;
element Standard deviation 0.53 0.01 0.04 0.11
(b) boundary element series

© ASCE 04013080-3 J. Struct. Eng.

J. Struct. Eng. 2014.140.


the peak load was reached. Specimens C2 and BE2 have very lim- well confined by the heads and the vertical bars anchored by the
ited confinement and experienced brittle failures after reaching heads showed no signs of buckling. Strain gauges near the heads of
their peak loads. The cover spalling, bar buckling, and poor con- the crossties indicated that average strains of about 1.83 and 3.48
finement resulted in a sudden drop of load to 50% of their peak times the yield strain were reached in Specimens C4 and BE4, re-
loads. The anchorage of the transverse reinforcement was ineffec- spectively. The appearance of Specimens C4 and BE4 after testing
tive once spalling of the cover occurred. At the peak load level, the is shown in Figs. 5(b and e). The performance of the headed bars is
stress in the transverse reinforcement was typically about 50% of as good as or better than the performance of the crossties with alter-
the yield stress. For the specimens with ductile detailing, the strains nating 90° and 135° hooks. The significant increase of the postpeak
in the transverse reinforcement increased significantly after the resistance of Specimen BE4 is due to the beneficial effects of the
cover concrete spalled and these specimens experienced yielding heads and the improved end anchorage of the headed bars. Speci-
of the transverse reinforcement. The appearance of the two series men BE4 experienced the highest ductility of all the boundary
of tests after testing is shown in Fig. 4. element specimens [see Fig. 3(b)].
Specimens C3 and BE3 both contain typical seismic details with Specimens C5 and BE5 were confined entirely by headed bars.
hoops and crossties. The influence of this improved confinement is Specimen C5 displayed a similar ductility to Specimens C3 and C4
Downloaded from ascelibrary.org by University of Exeter on 07/15/15. Copyright ASCE. For personal use only; all rights reserved.

an increase in the peak axial load and a significant increase in the and had a gradual decay in load carrying capacity even at very large
deformability after the peak load. The improved postpeak behavior strains [see Fig. 3(a)]. The excellent postpeak behavior is due to
is due to the improved confinement of the core concrete and the beneficial effects provided by the interlocking headed bars.
the effectiveness of the closely spaced hoops and crossties in pre- Fig. 5(c) shows column Specimen C5 at the end of testing, indicat-
venting premature bar buckling. However, at significant axial com- ing that at these large deformations, the peripheral headed bars
pressive strains, the 90° bend hooks of the crossties opened up were still interlocked. Fig. 5(c) also shows the spreading of com-
indicating some loss of anchorage [see Figs. 5(a and d)]. Strain pressive stresses in a conelike fashion emanating from the 50-mm
diameter heads at the corners of the column. These localized com-
gauges near the 90° bend anchorages of the crossties indicated that
pressive stresses not only contribute significantly to confinement,
average strains of about 1.38 and 1.23 times the yield strain were
but also assist in the interlocking of the peripheral headed bars at
reached in Specimens C3 and BE3, respectively. Specimens C4 and
the corners. This interlocking prevents the headed bars from open-
BE4, with the crossties replaced by headed bars, displayed better
ing up at the corners allowing them to act as peripheral hoop
postspalling response compared with Specimens C3 and BE3, re-
reinforcement. Strain gauges near the heads of the crossties indi-
spectively. Both specimens exhibited more gradual strength decay.
cated that average strains of about 1.0 and 3.5 times the yield strain
After cover spalling, the heads of the headed bars in Specimens C4 were reached in Specimens C5 and BE5, respectively. The response
and BE4 were exposed, however, the core concrete was extremely of Specimen BE5 was not quite as good as that of Specimen BE4,
but was better than the boundary element specimen with typical
seismic hoops and crossties (BE3). At a large axial compressive
strain of 0.007, the headed bars started losing some interlocking
at the corners and some local buckling of the longitudinal bars oc-
curred [see Fig. 5(f)]. It is believed that the smaller diameter
(40 mm) of the heads used in the boundary element specimens
was the cause of this loss of interlocking. The peripheral headed
bars yielded but they did not develop strains well above yield.

Comparison of Column Specimens

Table 3 summarizes the experimental results obtained for all of the


specimens. For the column specimens, the maximum axial load,
Pmax , for each specimen, varied between 9,418 and 10,838 kN.
Column C2 has a peak load about the same as that for Column
C1. Column C3 had much better confinement and had a capacity
15% greater than Column C2. Columns C3, C4, and C5 all have the
same amount of transverse reinforcement and have similar
peak loads.
The maximum loads are compared with their corresponding pre-
dicted axial capacities, computed according to the ACI code (ACI
2011) as
Po ¼ 0.85fc0 ðAg − Ast Þ þ fy Ast ð1Þ

where Ag = the gross concrete area; and Ast = the area of the lon-
gitudinal reinforcement.
The ratio, Pmax =Po , ranges from 0.97 to 1.12. The lower
ratios are observed in Columns C1 and C2 that had little or no
confinement.
Table 3 also shows the maximum load, Pc , carried by the con-
crete that was determined by subtracting the forces in the longitu-
Fig. 4. Photographs after testing: (a) column series; (b) boundary
dinal bars from the total applied load. The measured average
element series
longitudinal strain, together with the elastoplastic stress-strain

© ASCE 04013080-4 J. Struct. Eng.

J. Struct. Eng. 2014.140.


Downloaded from ascelibrary.org by University of Exeter on 07/15/15. Copyright ASCE. For personal use only; all rights reserved.

Fig. 5. Photographs of failure regions: (a) C3; (b) C4; (c) C5; (d) BE3; (e) BE4; (f) BE5

Table 3. Summary of Experimental Results


Axial loads Axial strains
Specimens Pmax (kN) Pmax =Po Pc (kN) Pc =Poc Pcc (kN) Pcc =Pocc εcP max (%) εcP max =εc0 εcc (%) 0
εcc =εc0
C1 9,812 1.01 8,179 1.01 — — 0.206 0.98 — —
C2 9,418 0.97 7,853 0.97 — — 0.197 0.94 — —
C3 10,838 1.12 9,134 1.13 6,895 1.34 0.280 1.33 0.590 2.81
C4 10,226 1.06 8,522 1.05 6,817 1.32 0.248 1.18 0.586 2.79
C5 10,595 1.10 8,891 1.12 7,203 1.60 0.268 1.27 0.552 2.63
BE1 6,694 0.95 5,551 0.96 — — 0.194 0.92 — —
BE2 6,848 0.97 5,589 0.97 — — 0.212 1.01 — —
BE3 7,620 1.08 6,252 1.08 5,328 1.20 0.297 1.41 0.589 2.80
BE4 7,647 1.08 6,279 1.09 5,831 1.31 0.340 1.62 0.560 2.67
BE5 7,527 1.06 6,159 1.08 5,744 1.48 0.335 1.60 0.602 2.87

0
relationship for the steel, was used to determine the force carried by load on the confined core, Pcc and the ratios of εcc =εc0 . These ratios,
the steel. The maximum force in the concrete, Pc , is compared with which vary from 2.63 to 2.81, indicate that there is a significant
the code expression for the maximum load carried by only the con- increase in the strain at the peak stress in the confined core.
crete, Poc ¼ 0.85fc0 ðAg − Ast Þ. It is noted that the ratios of Pc =Poc
are all above 1.0 for the confined column specimens. Following the
approach used by Sheikh and Uzumeri (1980), Cusson and Paultre Comparison of Boundary Element Specimens
(1994), and Sharma et al. (2005), the maximum load carried by the
confined concrete in the core, Pcc , was determined for the speci- For the boundary element specimens, the maximum axial load,
mens with seismic confinement. This load is compared with the Pmax , applied on each specimen during testing, varied between
load Pocc ¼ 0.85f c0 Acc where Acc = the confined concrete core area 6,694 and 7,647 kN (see Table 3). Specimen BE4 exhibited the
taken as the net area inside centerline of perimeter hoops or headed highest peak load, with Specimen BE3 having the next highest.
bars. The confined column specimens exhibited Pcc =Pocc ratios The ratio, Pmax =Po , ranges from 0.95 to 1.08. The lower ratios
varying from 1.32 to 1.60 indicating that the confined core concrete are for Specimens BE1 and BE2. It is noted that the ratios of
was able to develop strengths well above that of the unconfined Pc =Poc are all above 1.0 for the confined boundary element spec-
concrete. Table 3 gives the concrete strain, εcP max , corresponding imens. The confined boundary element specimens exhibited
to the maximum axial load. For the three confined columns, this Pcc =Pocc ratios ranging from 1.20 to 1.48, indicating that the con-
strain varies from 0.0025 to 0.0028. These values are somewhat fined core concrete was able to develop strengths well above that of
higher than the strain at maximum stress, εc0 , for the corresponding the unconfined concrete.
unconfined concrete cylinders. The ratio, εcP max =εc0 , ranges from The strain, εcP max , corresponding to the maximum axial load,
1.18 to 1.33 for Columns C3, C4, and C5, whereas the unconfined ranges from 0.0019 to 0.0034. The ratio, εcP max =εc0 , ranges from
Specimens C1 and C2 have ratios slightly less than 1.0. Table 3 also 0.92 to 1.62. The higher ratios are observed in the well-confined
0
gives the confined concrete strain, εcc , corresponding to the peak specimens, with the headed bar specimens having the highest

© ASCE 04013080-5 J. Struct. Eng.

J. Struct. Eng. 2014.140.


0
values. The ratios of εcc =εc0 range from 2.67 to 2.87, indicating the where wi = the clear distance between adjacent laterally supported
beneficial effects of confinement on the strain capacity of the longitudinal bars; s 0 = the clear vertical spacing between hoops; cx
confined core. and cy = the core dimensions measured to the centerline of the
hoops; and ρcc = the ratio of the area of longitudinal reinforcement
to the area of the confined core.
Analytical Model for Predicting Confinement Cusson and Paultre (1995) used the same expressions as that of
Mander et al. (1988) for determining the confinement effectiveness
Mander et al. (1984, 1988) proposed an approach that was modified coefficient, K e . Instead of calculating different effective confine-
from the analytical model proposed by Sheikh and Uzumeri (1982) ment pressures in the x and y directions for a rectangular column,
to predict the confinement effectiveness of rectangular hoops with Cusson and Paultre (1995) simply determine an overall effective
or without crossties. The effective lateral confining pressure, fle0 , is lateral confining pressure, f le0 , given as
expressed as
 
K e f h Ashx þ Ashy
fle0 ¼ K e fl ð2Þ f le0 ¼ K e ρse f h ¼ ð4Þ
cx þ cy
Downloaded from ascelibrary.org by University of Exeter on 07/15/15. Copyright ASCE. For personal use only; all rights reserved.

s
where fl = the lateral pressure on the confined concrete from the
transverse reinforcement, assumed to be uniformly distributed on where ρse = the effective transverse reinforcement ratio; fh = the
the concrete core. The confinement effectiveness coefficient, K e , stress in the hoops; s = the center-to-center spacing of the hoops;
is defined as the effective confined concrete area midway between and Ashx and Ashy = the areas of transverse reinforcement in the x
the transverse reinforcement divided by the net area of the con- and y directions.
crete measured to the centerline of the transverse reinforcement. Légeron and Paultre (2003) provided a method for determin-
Accounting for the arching of the confined concrete in both the ing the stress in the transverse reinforcement, f h , at the peak
horizontal and vertical directions, the confinement effectiveness co- of the confined concrete stress-strain response and, hence, ena-
efficient, K e , for a rectangular column can be expressed as bling the determination of fle0 . Cusson and Paultre introduced
Pn an effective confinement index at the peak confined concrete
w2i 0 0
ð1 − i¼1 6cx cy Þð1 − 2cs x Þð1 − 2cs y Þ stress as
Ke ¼ ð3Þ
ð1 − ρcc Þ

12000
C3
A
Concrete Axial Stress, fc (MPa)

f cc prediction
test
Confined concrete
8000
Load, kN

fc a confined
0.5f cc B core
C
Unconfined cover
0.5f c b 4000
concrete concrete
steel
Ec
c
ε c ε c50 ε cc ε c50c
0
Compressive Strain, ε c 0.0 0.4 0.8 1.2 1.6
Strain, %
Fig. 6. Proposed stress-strain curve for confined high-strength concrete
(data from Cusson and Paultre 1995) Fig. 7. Predicted load-deformation response of Specimen C3

Table 4. Confinement Parameters and Predicted Confined Concrete Properties


Specimen wi a (mm) Acc (mm2 ) s 0 (mm) fle0 (MPa) fle0 =f c0 0
fcc (MPa) 0
f cc =f c0 0
εcc (%) 0
εcc =εc0
C2 w1 ¼ 338 159,200 384 0.07 0.002 39.5 1.04 0.22 1.03
w2 ¼ 338
C3, C4 w1 ¼ w2 ¼ 156 159,200 109 3.53 0.093 55.4 1.45 0.63 3.01
w3 ¼ w4 ¼ 156
C5 w1 ¼ w2 ¼ 144 139,300 109 3.73 0.098 56.1 1.47 0.66 3.15
w3 ¼ w4 ¼ 144
BE2 w1 ¼ 426 162,800 487 0.02 0.001 32.6 1.02 0.21 1.01
w2 ¼ 296
BE3, BE4 w1 ¼ w2 ¼ w3 ¼ 129 162,800 107 2.42 0.076 44.7 1.39 0.54 2.58
w4 ¼ w5 ¼ 138
BE5 w1 ¼ w2 ¼ w3 ¼ 120 146,400 107 2.84 0.089 46.2 1.44 0.61 2.91
w4 ¼ w5 ¼ 125
a
wi values given for two adjacent sides (other two sides are identical).

© ASCE 04013080-6 J. Struct. Eng.

J. Struct. Eng. 2014.140.


12000 9000
C2 BE2
test test
prediction prediction

8000 6000
Load, kN

Load, kN
4000 3000

(a) 0 (b) 0
Downloaded from ascelibrary.org by University of Exeter on 07/15/15. Copyright ASCE. For personal use only; all rights reserved.

12000 9000
C3 BE3

8000 6000

Load, kN
Load, kN

4000 3000

(c) 0 (d) 0

12000 9000
C4 BE4

8000 6000
Load, kN

Load, kN

4000 3000

(e) 0 (f) 0

12000 9000
C5 BE5

8000 6000
Load, kN
Load, kN

4000 3000

0 0
0.0 0.4 0.8 1.2 1.6 0.0 0.4 0.8 1.2 1.6
(g) Strain, % (h) Strain, %

Fig. 8. Comparison of predicted and measured load-deformation responses

© ASCE 04013080-7 J. Struct. Eng.

J. Struct. Eng. 2014.140.


fle0 Predicted Behavior of Test Specimens
I e0 ¼ ð5Þ
fc0
Specimens C1 and BE1 do not contain any transverse reinforce-
Based on regression analyses, Légeron and Paultre (2003) pro- ment and, hence, the code design expression given in Eq. (1)
posed equations for the peak concrete confined stress and strain as was used to predict the strength. The predicted responses for Spec-
well as the fitting parameters for the postpeak confined concrete imens C2, C3, C4, C5, BE2, BE3, BE4, and BE5 were made using
stress-strain relationship. The proposed relationship between the the expressions based on the research of Mander et al. (1988) as
0 =f 0 , and the effective con-
strength gain of confined concrete, fcc modified by Légeron and Paultre (2003). This approach overesti-
c
finement index, I e0 , is given as mated the contribution of the cover concrete for this full-scale test
series. Hence, the stress-strain relationship for the concrete cover
0 =f 0 ¼ 1 þ 2.4ðI 0 Þ0.7
fcc ð6Þ was modified such that after reaching a strain in the unconfined
c e
concrete of 0.003, the stress dropped linearly to zero at a strain
and proposed the following equation for the concrete strain at the of 0.004.
0
peak compressive stress of the confined concrete, εcc : Table 4 gives the confinement parameters that were assumed for
Downloaded from ascelibrary.org by University of Exeter on 07/15/15. Copyright ASCE. For personal use only; all rights reserved.

0
the predictions of the concrete stress-strain responses including the
εcc =εc0 ¼ 1 þ 35ðI e0 Þ1.2 ð7Þ clear spacings between the laterally supported longitudinal bars, wi .
The spacing, s 0 , was assumed to be the clear distance between the
Mander et al. (1984) proposed a unified stress-strain relation- peripheral ties. The ratio f le0 =fc0 is an index of confinement effi-
ship for confined concrete, applicable to both circular and rectan- ciency (Cusson and Paultre 1995), which accounts for the arching
gular shaped transverse reinforcement, which is based on an action occurring in the concrete core and for the actual stress in the
equation suggested by Popovics (1973). Based on Mander et al.’s transverse reinforcement at peak strength of the confined concrete.
approach, Cusson and Paultre (1995) proposed a stress-strain curve Table 4 also indicates the predicted peak confined concrete stress
(see Fig. 6). and strain. The predicted ratios of f cc0 =f 0 and ε 0 =ε 0 (Table 4) com-
c cc c
This relationship, in general form, for the confined concrete pare reasonably well with the experimentally determined ratios of
stress, fc , as a function of the concrete longitudinal strain, εc , Pcc =Pocc and εcc 0 =ε 0 (Table 3), respectively, for each of the
0 c
for cases when εc ≤ εcc is given as specimens.
 0 Þ
 Fig. 7 illustrates the three different components contributing to
0 kðεc =εcc
f c ¼ f cc 0 k
ð8Þ the overall response for Specimen C3. It is noted that the force in
k − 1 þ ðεc =εcc Þ
the concrete from the predictions was multiplied by the factor 0.85
where k = a factor that controls the initial slope and the curvature of to obtain the predicted force deformation responses. This factor of
the ascending branch, given by 0.85 is the same factor assumed in the ACI code for predicting the
concrete contribution to the axial load resistance of a column.
Ec Fig. 8 compares the predicted load-strain responses with the ex-
k¼ 0 0 ð9Þ
Ec − ðf cc =εcc Þ perimental responses for the column and boundary element spec-
imens that contained transverse reinforcement. Specimens C2 and
and Ec = the tangent modulus of the unconfined concrete. BE2 were poorly confined by widely spaced transverse reinforce-
For the relationship for the descending part of the stress-strain ment anchored with 90° bend tie hooks and, hence, in the prediction
0
curve (when εc ≥ εcc ), Légeron and Paultre (2003) made some a limiting compressive strain of 0.004 was used to account for the
modifications to the relationship proposed by Fafitis and Shah complete loss of the concrete cover and severe bar buckling for
(1985). This expression is given as these two specimens. As shown in Fig. 8, the predictions agree rea-
sonably well with the experimental results.
0 0 k2
f c ¼ fcc exp½k1 ðεc − εcc Þ  ð10Þ

where Conclusions
ln 0.5 The following conclusions arise from the research on axially loaded
k1 ¼ 0 Þk2
ð11Þ
ðεcc50 − εcc columns and boundary elements confinement, including the use of
friction-welded headed bars with heads having a net bearing area of
nine times the bar area:
k2 ¼ 1 þ 25ðI e50 Þ2 ð12Þ
1. The combination of peripheral seismic hoops, together with
headed bars acting as seismic crossties, resulted in columns
The strain, εcc50 , is defined as the strain in the confined concrete
and boundary elements that performed as well as or better
when the stress drops to 50% of the peak strength of confined con-
than specimens constructed with seismic hoops and seismic
crete. This strain is given as
crossties.
εcc50 =εc50 ¼ 1 þ 60I e50 ð13Þ 2. The replacement of seismic crossties with headed bars resulted
in columns and boundary elements, which were easier to
where I e50 ¼ ρse f hy =fc0 . construct.
Cusson and Paultre suggested that εc50 could be assumed to be 3. The replacement of all of the transverse reinforcement by
0.004 if no experimental data was available and they also suggested headed bars, while giving similar performance as the speci-
that the proposed confinement model could be used for the predic- mens with hoops and crossties, can be difficult to place in col-
tion of the behavior of unconfined concrete columns using umns without a peripheral hoop to maintain alignment.
k2 ¼ 1.5. Sharma et al. (2005) compared a number of analytical 4. The friction-welded heads having a net bearing area of nine
models with several column test series and concluded that the times the bar area offer the following advantages:
model proposed by Légeron and Paultre (2003) gave the most con- a. Spreading of the compressive stresses from under the
sistent predictions of the actual test responses. heads provides confinement to the core concrete;

© ASCE 04013080-8 J. Struct. Eng.

J. Struct. Eng. 2014.140.


b. The heads of these bars enable the development of high- Canadian Standards Association (CSA). (2004). “Design of concrete struc-
tensile strains close to the heads, even after concrete cover tures.” CSA A23.3-04, Rexdale, ON, Canada.
spalling has occurred; and Cusson, D., and Paultre, P. (1994). “High-strength concrete columns con-
c. The presence of the stiff heads reduces the effective spa- fined by rectangular ties.” J. Struct. Eng., 10.1061/(ASCE)0733-9445
cing between the transverse reinforcing elements, hence, (1994)120:3(783), 783–803.
leading to better control of longitudinal bar buckling. Cusson, D., and Paultre, P. (1995). “Stress-strain model for confined high-
strength concrete.” J. Struct. Eng., 10.1061/(ASCE)0733-9445(1995)
5. In predicting the responses, it was found that the stress-strain
121:3(468), 468–477.
relationship for the concrete cover had to be adjusted to ac-
Dilger, W. H., and Ghali, A. (1997). “Double-head studs as ties in concrete
count for the more brittle nature of cover spalling in these walls and columns.” Concr. Int., 19(6), 59–66.
full-scale test specimens. Fafitis, A., and Shah, S. P. (1985). “Lateral reinforcement for high-strength
6. Reasonably accurate predictions can be obtained for the stress- concrete columns.” SP-87, American Concrete Institute, Detroit,
strain relationship of the concrete confined by headed bars by 213–232.
using the method described in this paper. Ghali, A., and Dilger, W. H. (1998). “Anchoring with double-head studs.”
This research demonstrates that headed bars offer a practical Concr. Int., 20(11), 21–24.
Downloaded from ascelibrary.org by University of Exeter on 07/15/15. Copyright ASCE. For personal use only; all rights reserved.

method of providing confinement in columns and wall boundary Légeron, F., and Paultre, P. (2003). “Uniaxial confinement model for
elements subjected to axial loading. Further research on the re- normal- and high-strength concrete columns.” J. Struct. Eng., 10.1061/
versed cyclic loading behavior of columns and walls containing (ASCE)0733-9445(2003)129:2(241), 241–252.
headed bars is needed. Mander, J. B., Priestley, M. J. N., and Park, R. (1984). “Seismic design of
bridge piers.” Res. Rep. 84–2, Univ. of Canterbury, New Zealand.
Mander, J. B., Priestley, M. J. N., and Park, R. (1988). “Theoretical stress-
Acknowledgments strain model for confined concrete.” J. Struct. Eng., 10.1061/(ASCE)
0733-9445(1988)114:8(1804), 1804–1823.
The donation of materials by Headed Reinforcement Corporation Popovic, S. (1973). “A numerical approach to the complete stress-strain
(HRC) in Fountain Valley, CA, is greatly appreciated. Kjell L. Dahl curve of concrete.” Cem. Concr. Res., 3(5), 583–899.
and Christian Dahl of HRC provided practical guidance on the use Razvi, S. R., and Saatcioglu, M. (1994). “Strength and deformability of
of headed bars. David Mar and Leo Panian of Tipping Mar and confined high strength concrete columns.” ACI Struct. J., 91(6),
Associates provided valuable discussions and examples of applica- 678–687.
tions of headed reinforcement. The financial support of the Natural Sharma, U. K., Bhargava, P., and Kaushik, S. K. (2005). “Behavior of con-
Sciences and Engineering Research Council of Canada (NSERC) is fined high strength concrete columns under axial compression.” J. Adv.
gratefully acknowledged. Concr. Technol., 3(2), 267–281.
Sheikh, S. A., and Uzumeri, S. M. (1980). “Strength and ductility of tied
concrete columns.” J. Struct. Eng., 1079–1102.
References Sheikh, S. A., and Uzumeri, S. M. (1982). “Analytical model for concrete
confinement in tied columns.” J. Struct. Eng., 2073–2722.
American Concrete Institute (ACI). (2011). “Building code requirements Youakim, S. A., and Ghali, A. (2002). “Ductility of concrete columns with
for structural concrete and commentary.” ACI 318-11, Detroit. double-head studs.” ACI Struct. J., 99(4), 480–487.

© ASCE 04013080-9 J. Struct. Eng.

J. Struct. Eng. 2014.140.

Вам также может понравиться