See discussions, stats, and author profiles for this publication at: https://www.researchgate.

net/publication/279669762

Role of hoops on shear strength of reinforced concrete beam-column joints

Article  in  Aci Structural Journal · May 2005

CITATIONS READS
69 1,954

5 authors, including:

Hung-Jen Lee
National Yunlin University of Science and Technology
24 PUBLICATIONS   600 CITATIONS   

SEE PROFILE

Some of the authors of this publication are also working on these related projects:

Behavior and Modeling of Damage and Collapse in Reinforced Concrete Frame Joints Subjected to Near-Fault Earthquakes View project

All content following this page was uploaded by Hung-Jen Lee on 30 November 2015.

The user has requested enhancement of the downloaded file.

 ACI STRUCTURAL JOURNAL
Title no. 102-S45
Role of Hoops on Shear Strength of Reinforced Concrete
Beam-Column Joints
by Shyh-Jiann Hwang, Hung-Jen Lee, Ti-Fa Liao, Kuo-Chou Wang, and Hsin-Hung Tsai
This study investigated the effect of joint hoops on the shear
strength of exterior reinforced concrete beam-column joints subjected
to earthquake-type loading. Nine exterior reinforced concrete
beam-column subassemblages were tested under reverse cyclic
loading. All test specimens were designed to have adequate shear
strength of joints, determined by the softened strut-and-tie model.
The parameters considered include the amount and detail of joint
hoops. Test results indicated that the major function of the joint
hoop is to carry shear as a tension tie and to constrain the width of
crack. A lesser amount of hoop reinforcement with a wider spacing
could be used without significantly affecting the performance of
joints. Test data also demonstrated that a beam-column joint with-
out hoop bars can exhibit satisfactory seismic behavior, as long as
the joint is provided with adequate shear strength.
Keywords: beam-column; joint; hoop; reinforced concrete; shear; strut; tie.
INTRODUCTION
The role of transverse reinforcement and mechanism of
shear transfer in joint for seismic resistance are subjects of
much debate. Paulay, Park, and Priestley1 proposed shear
transfer mechanisms of a joint, as shown in Fig. 1, referred
to as a diagonal strut mechanism and a truss mechanism.
Both mechanisms are incorporated in the New Zealand NZS
3101 Standard2 which assumes adequate bond of reinforcing
bar in joint so that shear forces are transmitted to the joint
core. Considerable amounts of transverse reinforcement are
required to transfer tensile force and to necessitate a truss
mechanism in joint. In the U.S., the design method for beam-
column joints is given in the ACI 318 Building Code3 as well
as its companion document’s (ACI 352R) recommendations.4
The ACI method,3,4 however, assumes severe bond deterio-
ration of the reinforcing bar in joint and internal forces are
only transfer in a diagonal compression strut of concrete.
The role of hoop bars in the diagonal strut model is to
confine the core concrete. These conflicting opinions about
the function of transverse reinforcement lead to different
demands for hoop bars, as well as the disparity in detailing
criteria. Real behavior may be a combination of the diagonal
strut and the truss mechanism with the bond deterioration to
a certain degree of longitudinal reinforcement during cyclic
loading. Significant research progress has been made in
recent years,5 yet these design methods still differ from each
other significantly, indicating the complex behavior of beam-
column joints and the need for further research.
Currently, the ACI provisions3,4 emphasize the importance
of the confinement of the joint core. Hence, the closely
spaced transverse reinforcement in the end regions of laterally
loaded columns must be detailed within joints unless the
surrounding beams provide a suitable confinement. The use
of crossties is inevitable because the maximum spacing
between the legs of hoops is limited to 350 mm on center.
ACI Structural Journal/May-June 2005
TECHNICAL PAPER
These ACI requirements for adequate concrete confinement
lead to congested joints, which are very difficult to construct.
Moreover, the use of high-strength concrete would pose an
even worse situation for construction because the amount of
confining reinforcement is proportional to the concrete
strength.3,4 It seems necessary to declare more concisely
whether joint hoops serve to confine concrete, to resist shear,
or both.
A softened strut-and-tie (SST) model, satisfying the equi-
librium, compatibility, and constitutive laws of cracked rein-
forced concrete, has been developed to evaluate the shear
strength of beam-column joints.6,7 Joint hoops have two
roles in the shear-resisting mechanisms as postulated by the
SST model. The first is to form a tension tie and provide an
additional shear-transferring path beside the main diagonal
strut. The second is to control the widths of cracks and retard
the softening process of the cracked concrete. The SST
model suggests that transverse reinforcement is more important
in resisting shear than in confining concrete, so the required
amount and spacing limits of the joint hoops can be reduced.
The main objective of this study was to investigate the effect
of joint hoops on the shear strength of exterior reinforced
concrete beam-to-column connections subjected to earth-
quake-type loading. The other objective was to evaluate the
performance of exterior joints that conform to the SST design
philosophy. Nine exterior beam-to-column subassemblages
were tested under reverse cyclic loading to achieve these
objectives.
RESEARCH SIGNIFICANCE
This study is intended to clarify the effect of joint hoops
on the shear-resisting behavior of exterior beam-column
Fig. 1—Force-transfer mechanism in exterior joint.
ACI Structural Journal, V. 102, No. 3, May-June 2005.
MS No. 04-057 received April 2, 2004, and reviewed under Institute publication
policies. Copyright © 2005, American Concrete Institute. All rights reserved, including
the making of copies unless permission is obtained from the copyright proprietors.
Pertinent discussion including author’s closure, if any, will be published in the March-
April 2006 ACI Structural Journal if the discussion is received by November 1, 2005.
445

ACI member Shyh-Jiann Hwang is a professor in the Department of Construction
Engineering, National Taiwan University of Science and Technology, Taipei, Taiwan.
He received his PhD from the University of California, Berkeley, Calif., in 1989. His
research interests include seismic behavior of beam-column joints, shear strength of
reinforced concrete members, and seismic retrofitting of concrete structures.
ACI member Hung-Jen Lee is an assistant professor in the Department of Construction
Engineering, National Yunlin University of Science and Technology, Yunlin, Taiwan.
He received his PhD from the National Taiwan University of Science and Technology
in 2000. His research interests include behavior of beam-column joints and strut-and-
tie modeling.
Ti-Fa Liao is a structural engineer at Sino-Tech Engineering Consultants, Ltd.,
Taipei, Taiwan. He received his MS degree from the National Taiwan University of
Science and Technology in 1999.
Kuo-Chou Wang is a structural engineer at Evergreen Consulting Engineering, Inc.,
Taipei, Taiwan. He received his MS degree from the National Taiwan University of
Science and Technology in 2002.
Hsin-Hung Tsai is a professional engineer in Kaohsiung, Taiwan. He received his
MS degree from the National Taiwan University of Science and Technology in 2000.
connections under large load reversals. An attempt was
made to relieve the reinforcement congestion in joints. The
test results indicate that fewer joint hoops with wider spacing
could be used without significantly affecting the perfor-
mance of the beam-column connections if the joints are
provided with adequate shear strength according to the SST
model.
DESIGN OF BEAM-COLUMN JOINT
ACI Code method
There are four major requirements for the seismic design
of beam-column joints in the ACI 318 Code:3
1. The flexural strength ratio MR should satisfy
ΣM
M R = -----------c ≥ 6
--- (1)
ΣM b 5
where ΣMc and ΣMb are the sum of nominal moment
capacity of columns and beams calculated at the joint faces,
respectively. This requirement, referred to as the strong-
column weak-beam philosophy, tends to avoid severe
yielding in columns of the lateral force resisting system.
2. For a hooked bar anchorage in the exterior joint, the
development length l dh of the beam bar shall not be less than
the largest of 8db, 150 mm, and the length given by
fy db
l dh = ----------------
- (2)
5.4 f c′
where fy and db is the yield strength and diameter of the beam
bar, respectively; and fc′ is the concrete strength (in MPa).
3. The design shear forces acting on joints shall not exceed
specified limits which are classified according to geometry and
confinement of joints.3 For the case of exterior beam-column
joints subjected to earthquake-introduced loading, as shown
in Fig. 1, the horizontal joint shear force Vjh is estimated as
V jh = T – V col (3)
where T is the tensile force in the beam reinforcement, and
Vcol is the horizontal column shear above the joint. An over-
446
strength factor of 1.25 for the beam reinforcement should be
included in estimating the design joint shear force Vjh,u.3,4
The nominal joint shear strength given in the ACI Code3
is calculated by
ACI
V jh, n = γ f c′ A j (4)
where γ is equal to 1.7, 1.25, or 1.0, depending on the geom-
etry and the level of confinement exerted by the beams
framing into the joint, and Aj is the effective joint area.3
4. The specified transverse hoop reinforcement shall be
provided within the joint for earthquake loading. The required
minimum ratio of the confining reinforcement area is ρhACI
given for rectangular sections by the following relationship3
ACI A sh f c′  A g  f c′
ρh = ------------ = 0.3 ------  --------- – 1 and not less than 0.09 ------ (5)
s h h″ f yh  A ch  f yh
where Ash is the total area of confining transverse reinforcement
within vertical spacing sh and is in the direction perpendicular
to the column core width h″, which measured center-to-
center of the hoops; fyh is the yield strength of the hoop
reinforcement; Ag is the gross column section area; and Ach
is the section area of column core measured out-to-out of
hoop reinforcement.
Softened strut-and-tie (SST) method
The strong-column weak-beam philosophy for a ductile
reinforced concrete frame (Eq. (1)), the development length
of hooked bars (Eq. (2)), and the equilibrium of forces acting
on a joint (Eq. (3)) are also adopted in this study. Further, the
beam bars must be bent into the joint as close to the far face
of the column as possible, as shown in Fig. 1(a), where a
diagonal compression Cd will develop from the bend of the
hooked bars. Based on the simplified version of the SST
model,8 the nominal horizontal shear strength of joints can
be defined as
SST
V jh, n = C d, n cos θ = ( K h + K v – 1 )ζ f c′ A str cos θ (6)
where Cd,n is the nominal diagonal compression strength of
the joint (Fig. 1(a)); θ is the angle of inclination between
diagonal strut and the horizontal axis; Kh and Kv are the
indexes of the horizontal and vertical ties, respectively; ζ
is the softening coefficient and is given approximately by
ζ ≈ 3.35/ f c′ ≤ 0.52; and Astr is the effective area of the
diagonal strut.
The horizontal tie index Kh is expressed as
A th f yh
K h = 1 + ( K h – 1 ) -------------
- ≤ Kh (7)
Fh
where Ath is the area of the horizontal tie in the direction of
loading, and Kh is the horizontal tie index with horizontal
reinforcement remaining elastic and can be estimated as
1
K h ≈ ------------------------------------
- (8)
1 – 0.2 ( γ h + γ 2h )
ACI Structural Journal/May-June 2005
where γh is the fraction of horizontal joint shear carried by
the horizontal tie in the absence of the vertical tie, and is
defined as9
2 tan θ – 1
γ h = ----------------------
- for 0 ≤ γ h ≤ 1 (9)
3
The balanced amount of the horizontal tie force Fh , which
indicates the simultaneous occurrence of the yielding of
horizontal tie and the crushing of diagonal strut, is given by
F h = γ h × ( K h ζ f c′ A str ) × cos θ (10)
The related equations associated with the vertical tie are
the same as Eq. (7) to (10), except that all the h subscripts are
replaced by v, and θ is replaced by (90 degrees – θ).
To prevent joint shear failure after beam yielding, it is
recommended to design the hoop reinforcement in joints
remaining elastic. Accordingly, the required horizontal tie
force in the design of joint is given by
actuator using displacement-controlled loading history as
shown in Fig. 4. Sufficient shear reinforcement was provided
in the beam and columns to prevent shear failure outside of
the joint. All test specimens were designed with sufficient
shear strength according to the SST model. The examined
parameters include the amount and the detailing of the joint
hoops. Figure 5 presents the dimensions and the reinforcement
details of joints.
All specimens were cast flat rather than vertically as in
actual construction. Table 1 presents the test-day compressive
strength of the concrete cylinders and the average yield
stress of the reinforcing steel. The design values of the test
parameters and the actual values for each specimen are listed
in Table 2.
Specimen 0T010 is a lower-bound unit, which had barely
sufficient joint strength (Vjh,u/ V SST
jh, n ≈ 1 ) and no joint hoop
(Table 2). From the perspective of the SST model, ACI 318-023

A th f yh ≥ γ h V jh, u (11)

Comparison of SST method and ACI Code

The nominal joint shear strength and the transverse rein-
forcement of the proposed SST model are compared with
those of the ACI Code.3 To demonstrate the difference
between the two methods, a case study of selected exterior
joints with various parameters is shown in Fig. 2(a). The
studied parameters include the concrete strength and the
angle of diagonal compression strut. It should be noted that
the proposed equations are functions of these two parame-
ters, while the angle of diagonal compression strut is not
considered in the ACI Code equations.
Figure 2(b) illustrates contour of the ratio of Eq. (6) to
Eq. (4) under the variations of fc′ and θ. The shear strength
of proposed method is relatively lower than that of the ACI
Code3 for the selected joints, especially for the joints with
lower fc′ and higher θ. This phenomenon is attributed to the Fig. 2—Comparison of softened strut-and-tie method with
softened concrete strength ζ fc′ and horizontal projection of ACI Code for design of exterior joints.
cosθ in Eq. (6).
A comparison of requirements on transverse reinforcement
is shown in Fig. 2(c). The transverse reinforcement area of
the proposed Eq. (11) is converted to the confining reinforce-
ment ratio and compared with Eq. (5) in the direction of
loading. As shown in Fig. 2(c), the joints with a smaller θ of
diagonal strut can resist the horizontal shears more
efficiently8 and reduce the demand of hoop reinforcement. In
general, the proposed equations require less hoop reinforcement
and predict more conservative shear strength for the exterior
joints. For the case of fc′ = 70 MPa and θ = 45 degrees marked
in Fig. 2, the ratio of the proposed design to the ACI require-
ments is 0.68 on joint shear strength and 0.44 on hoop rein-
forcement. This case of exterior joints is close to the
specimens presented in the following.

TEST PROGRAM
Nine exterior reinforced concrete beam-column connections,
using concrete with design compressive strength of 70 MPa
and Grade 60 steel, were tested.10-12 Figure 3 depicts a
typical beam-column connection tested in this study. A
horizontal load was applied to the end of beam with an Fig. 3—Specimen configuration and test setup.

ACI Structural Journal/May-June 2005 447

overestimates the shear strength of the exterior joints. Specimen 3T44 was as high as 2.44% and its spacing was
Notably, the design joint shear force is approximately 2/3 of 97 mm (Table 2).
ACI
the ACI value (Vjh,u / V jh, n ≈ 2 ⁄ 3 ; Table 2). The flexural Series 210 was intended to test the effectiveness of the
strength ratio MR is approximately 3 in this study (Table 2). hoop detailing based on a concept of shear resistance. The
To provide confinement through the joint, Specimen 3T4410 specimens of Series 2 had the same dimensions as those of
was detailed with the joint hoops as per the ACI require- Series 1 (Fig. 5), but the joint hoop ratios of Series 2 were
ments.3, 4 As shown in Fig. 5(b), the joint of Specimen 3T44 only approximately 1/4 of the ACI requirement (Column (8)
was reinforced with three layers of double No. 4 hoops with of Table 2). The hoop details of Specimen 3T3 conformed to
crossties. The ratio of transverse reinforcement ρh for ACI 318-02.3 Specimen 3T3 was constructed using three
layers of No. 3 hoops with crossties at the joint (Fig. 5(d)).
Table 1—Material properties However, the strict detailing rules of joint hoops with the
purpose of confinement were reduced for Specimens 2T4
Reinforcement
and 1T44. The joint of Specimen 2T4 was built using two
Beam bars Column bars
Concrete No. 8 No. 10 Joint hoops layers of No. 4 hoops without crossties (Fig. 5(e)). Further-
fc′, fy , fu , fy , fu , fy , more, two sets of No. 4 hoops, functioning as a tension tie,
Specimen MPa MPa MPa MPa MPa Size MPa
(1) (2) (3) (4) (5) (6) (7) (8) (9)
0T010 67.3 430 605 421 659 — — Table 2—Design parameters
Series 1 3T4410 76.8 430 605 421 659 No. 4 498 Joint strength Joint hoops
12 61.8 435 661 430 648 — — V jh, u V jh, u ρh A th f yh
1B8 ---------- ---------- Spacing, ---------
- ---------------
SST ACI ACI
10 Specimen MR V jh, n V jh, n Amount mm ρ h γ h V jh, u
3T3 69.0 430 605 421 659 No. 3 471
(1) (2) (3) (4) (5) (6) (7) (8) (9)
Series 2 2T410 71.0 430 605 421 659 No. 4 498
0T0 2.9 1.00 0.62 None — 0 0
1T4410 72.8 430 605 421 659 No. 4 498 (3.0) (1.08) (0.67)
3T411 75.2 491 715 458 690 No. 4 436 Series 2.9 0.88
3T44 (3.0) 0.62 3 layers 97 0.91 1.63
1 (0.90) (0.63) 2-No. 4 (1.00) (1.85)
Series 3 2T5 11 76.6 491 715 458 690 No. 5 469
1B8 2.9 1.00 0.62 None — 0 0
1T5511 69.7 491 715 458 690 No. 5 469 (2.9) (1.14) (0.71)
Note: 1 MPa = 145 psi. 3T3 2.9 0.95 0.62 3 layers 97 0.25 0.45
(3.0) (1.01) (0.66) 1-No. 3 (0.29) (0.48)
Series 2T4 2.9 0.94 0.62 2 layers 146 0.20 0.54
2 (3.0) (0.98) (0.65) 1-No. 4 (0.24) (0.62)

1T44 2.9 0.94 0.62 1 layer 293 0.20 0.54

(2.9) (0.97) (0.65) 2-No. 4 (0.23) (0.62)
3T4 3.1 0.76 0.54 3 layers 97 0.46 1.01
(3.0) (0.89) (0.63) 1-No. 4 (0.45) (0.88)
Series 2T5 3.1 0.76 0.54 2 layers 146 0.31 1.05
3 (3.0) (0.88) (0.63) 1-No. 5 (0.33) (0.98)
3.1 0.76
1T55 (3.0) 0.54 1 layer 293 0.31 1.05
(0.92) (0.66) 2-No. 5 (0.36) (0.98)
Note: 1 mm = 0.0394 in.; numbers outside parentheses are design values; numbers
Fig. 4—Loading history. inside parentheses are actual values.

Fig. 5—Beam-column joint details.

448 ACI Structural Journal/May-June 2005

were grouped into the middle of the joint of Specimen 1T44
(Fig. 5(f)).
The details of the specimens of Series 311 were quite
similar to those of Series 2 (Fig. 5), but the number of joint
hoops was increased to approximately 1/3 of the ACI
requirement in Series 3 (Column (8) of Table 2). Ties remaining
elastic are commonly believed to maintain the integrity of
the members under severe cyclic loading caused by earth-
quakes. Therefore, the number of joint hoops in Series 3 was
designed to remain in the elastic range using Eq. (11). The
specimens of Series 3 were provided with barely enough
joint hoops, that is, Ath fyh/γhVjh,u ≈ 1 (Table 2).
Specimen 1B812 was designed to check whether the joint
hoop reinforcement can be replaced with the beam inter-
mediate bars passing through the joint. One U-shaped No. 8
bar was detailed in the middle of joint as shown in Fig. 5(c).
DISCUSSION OF TEST RESULTS
The relationships of applied load P versus the drift ratio D
for all specimens are illustrated in Fig. 6, where the drift ratio
D is defined as the deflection ∆ of the load point divided by
the distance L between the load point and the centerline of
the column. The curves of the joint force Vjh versus the joint
shear deformation γj are also plotted in Fig. 6, where Vjh was
calculated using Eq. (3) and γj was measured by gauges.
Fig. 6—Load-versus-deflection response of specimens.
ACI Structural Journal/May-June 2005
Table 3 presents the test results concerning strength and
ductility. The measured joint strength Vjh,test was obtained from
the peak applied load Pmax using Eq. (3). The displacement
ductility µ is defined as ∆max /∆y, where ∆max is the hori-
zontal displacement at the free end of the beam corre-
sponding to Pmax. The yield displacements ∆y for all test
units were determined by extrapolating the measured
horizontal displacement at 0.75Mn, linearly to Mn, which is
the nominal flexural strength of the beam.
The failure modes of the test specimens were classified as
BF or BF-JS (Table 3). The term BF refers to the beam
flexural failures due to the buckling of the beam bars. The
classification of BF-JS means the joint shear failure after the
development of the beam flexural hinge. The occurrence of
the beam bar buckling and the distortion of the joint shear
deformation are shown in Fig. 6. Photographs of these two
different modes of the damaged specimens at the conclusion
of the tests are presented in Fig. 7.
Importance of shear strength
Specimen 0T0 was detailed without any horizontal rein-
forcement in the joint (Fig. 5(a)). Sufficient shear strength,
however, was provided to the joint of Specimen 0T0 according
to the SST model. Although failed in joint, Specimen 0T0
exhibited a satisfactory hysteretic response (Fig. 6(a)), and
449
its displacement ductility µ can reach a value of 5.2 (Table 3).
This observation indicates that designing the joint with
sufficient shear strength ensures a certain level of performance
even without joint reinforcement. Notably, the axial
compressive load on the columns during the tests was very
small (Fig. 3). A high axial load increases the probability of
buckling of the longitudinal column bars.
As shown in Table 3, the BF-JS failure is associated with
the strength ratios Vjh,test / V SST
jh, n larger than unity, and the BF
failure is associated with Vjh,test/ V SST jh, n under unity. This
represents the SST model can accurately predict the joint
shear strength which is critical in determining the seismic
behavior of beam-column connections. On the other hand,
the strength ratios Vjh,test / V ACI jh, n are always under unity
despite the occurrence of joint failure (Table 3). This result
indicates that ACI 318-023 overestimates the shear strength of
the exterior beam-column joints. Similar observations were
also reported by Ehsani and Wight.13
Effect of confinement
The test units of Series 2, Specimens 3T3, 2T4, and 1T44,
shared the same properties except for the joint hoops. The
detailing of the joint hoops for Specimen 3T3 followed the
strict requirements of ACI provisions. In Specimen 1T44,
however, the crosstie was removed and the vertical hoop
spacing was increased to 300 mm in the joint (Fig. 5). In
disregard of the breakdown on confinement effect, the Series 2
specimens exhibited similar seismic behavior. This can be
examined through their hysteretic curves in Fig. 6 and other
indexes, such as strength, ductility, energy dissipation and
failure mode, as shown in Table 3. The similarities in seismic
behavior among the Series 2 units imply that the joint hoop
plays no role in confining concrete core. The tests of Series 3
repeated the same experience with different failure mode.
The beneficial effect of confinement by transverse rein-
forcement on the seismic behavior of beam-column joint was
not found in this study. The increase in the strength and
ductility of concrete confined by reinforcement is observed
for the axially loaded columns or the beams under flexure, in
which the concrete deforms under the plane-remaining-plane
restriction. The beam-column joint, however, is a region of
high shear, thus the joint deformation is governed by the
Mohr’s compatibility instead of the Bernoulli’s compatibility.
Consequently, the concrete strength within the joint should
be described by the softening phenomenon.14,15 The joint
hoop should be considered as crack-controlled steel and not
as concrete-confining reinforcement. Therefore, joint hoops
are added to retard the strength deterioration of concrete, and
not to enhance the strength of concrete. The crack-controlled
reinforcement requires less steel and wider spacing for detailing.
It is worthwhile to point out that Specimen 3T4 had the
same failure mode of beam hinging as Specimen 3T44 (Fig. 6).
The number of joint hoops in Specimen 3T4, however, are
only 1/2 of those within Specimen 3T44. The joint hoop
ratios ρh/ρhACI of Specimens 3T4 and 3T44 are 0.46 and
0.91, respectively (Table 2). This reveals that the ACI
requirement on the number of joint hoops is unnecessary.

Effect of tension tie

Fig. 7—Damaged specimens at conclusion of tests.
Although the test units of Series 2 and 3 had various joint
hoops, they exhibited similar hysteretic behavior, as shown
in Fig. 6. Among those changes in joint hoops, the effects of
tension tie were carefully maintained in the perspective of
the SST model, shown as Ath fyh/γhVjh,u in Table 2. Changing
the joint hoops without changing the effect of the tension tie
can preserve the seismic behavior of beam-column joints,
which means the effect of tension tie is the key role of joint
hoops in seismic resistance.

Table 3—Test results

Test results Comparisons
Strength Ductility Hoop Shear strength
Dissipated A th f yh V jh, test V jh, test
Pmax , Vjh,test , Pn,calc , - --------------
energy to Failure ------------------- SST
- --------------
ACI
-
Specimen kN kN kN ∆y, mm µ γ V
8% drift, kJ mode h jh, test V jh, n V jh, n
(1) (2) (3) (4) (5) (6) (7) (8) (9) (10) (11) (12)
0T0 192 997 167 24 5.2 146 BF-JS 0 1.11 0.69
Series 1 3T44 205 1065 169 24 8.7 267 BF 1.69 0.99 0.69
1B8 242 1257 208 26 4.9 146 BF-JS 0 1.45 0.91
3T3 218 1132 168 21 8.2 240 BF-JS 0.41 1.17 0.77
Series 2 2T4 208 1080 169 22 7.5 222 BF-JS 0.56 1.09 0.73
1T44 200 1039 170 22 7.8 222 BF-JS 0.58 1.03 0.69
3T4 214 1110 190 24 7.1 269 BF 0.88 0.90 0.63
Series 3 2T5 224 1162 190 24 7.1 255 BF 0.94 0.92 0.66
1T55 217 1126 189 23 7.3 261 BF 0.97 0.93 0.67
Note: 1 kN = 0.2248 kip; 1 mm = 0.0394 in.; failure mode BF means flexural failure of beam hinging; and BF-JS means joint
shear failure after beam hinging.

450 ACI Structural Journal/May-June 2005

 Figure 8 presents the strains measured in the joint hoops Although the crack-controlling ability of Specimen 1T55
for Specimens 3T3 and 3T44, in which Gauges 1 and 2 was inferior to that of Specimen 3T4, the crack-controlling
measured the strains due to shear-resisting tension and ability of Specimen 1T55 was superior to that of Specimen 3T3
Gauges 3 and 4 measured the strains of the crack-controlling (Fig. 10). This finding indicates that the number of joint
effect only. In general, the strain readings of Gauges 1 and 2 hoops is more important than the joint hoop spacing if the
are larger than those of Gauges 3 and 4. As shown in Fig. 8, vertical spacing is under 300 mm. A comparison of their
the readings of Gauges 1 and 2 increase rapidly and in strength, ductility, energy dissipation, and failure modes
proportion to the growth of joint shear Vjh for drift less than (Table 3) indicates that Specimen 1T55 is as effective as
1%. Afterward, Gauges 1 and 2 gained strains slowly due to Specimen 3T4 in seismic design. In the light of the test
joint deformation and a mild increase in Vjh until yielding results, it is thus concluded that the vertical spacing of joint
occurred. The readings of Gauges 3 and 4 increased gradually hoop up to 300 mm is an acceptable range.
with the expansion of joints but grew at a steeper rate once
the legs of hoops in the perpendicular direction were yielded
(Fig. 8(b)).
The specimen designs in Series 2 and 3 were quite similar,
except that the number of joint hoops in Series 3 was double
that of Series 2 (Fig. 5). The joint hoops in Series 3 were
designed to remain in the elastic range for shear-transferring
mechanism in the SST model. Figure 9 demonstrates the
strains of the joint hoops of Series 3 were in the elastic range
(Fig. 9(d) to (f)), and those in Series 2 yielded at large drifts
(Fig. 9(a) to (c)). It should be noted that comparable strain
readings were observed among specimens with different
detailing for Series 2 and 3 in Fig. 9. This observation
implies that the major characteristic of joint hoops can be
captured by considering them to be tension tie. It is also
noted that the SST model can predict the required strength of
tension tie reasonably well using Eq. (11), which corresponds
to 1/3 of the ACI requirement for the Series 3 units (Table 2).
Fig. 8—Measured strain in joint hoops for Specimens 3T3
and 3T44.
Effect of crack control
Series 2 specimens failed by the crushing of the concrete
in the beam-column joints, and Series 3 specimens failed by
the buckling of compression bars in the beam hinges (Fig. 7).
The joint hoops in the Series 2 units yielded in the middle of
the displacement reversals (Fig. 6(d) to (f)), and the yielding
of Series 3 joint hoops was effectively delayed by adding a
sufficient amount of transverse steel (Fig. 6(g) to (i)). Joint
hoop yielding has a decisive influence on the failure modes
of the beam-column joints. The early yielding of joint hoops
is believed to provide no reliable mechanism to restrain the
deterioration of the joint concrete and leads to the joint
failure at larger displacement levels. Designing the joints
with the joint hoops remaining elastic is appealing for the
special moment-resisting frames, because it maintains the
integrity of the members under severe cyclic loading caused
by earthquakes. Fig. 9—Measured strain in joint hoops for Series 2 and 3.
The different detailing in joint hoops for Series 2 and 3
preserved the equivalent tie action but slightly altered the
effect of crack control among specimens. The crack-controlled
reinforcement is more effective with smaller spacing. Among
the three specimens in Series 3, Specimen 3T4 had the best
detailing for crack control, followed by Specimen 2T5;
Specimen 1T55 had the poorest detailing. With reference to
the Vjh-versus-γj curves in Fig. 6(g) to (i), Specimen 3T4
exhibited a straight linear relationship, while Specimen 1T55
displayed a wider-banded linearity. As shown in Fig. 9(d) to
(f), Specimen 3T4 experienced the fewest strains in the joint
hoops. Figure 10 presents the measured crack widths in the
joint regions and confirms the aforementioned argument. As
can be seen, Specimen 3T4 yields the smallest crack width,
followed by Specimen 2T5; Specimen 1T55 had the largest
crack width. Fig. 10—Comparison of joint crack width.

ACI Structural Journal/May-June 2005 451

Beam or column intermediate bars
A U-shaped beam intermediate bar was expected to
perform as a tension tie in the joint of Specimen 1B8. As can
be seen in Fig. 11, the beam intermediate bar had a tie action
at the 1% drift. At the 2% drift or above, the yielding of beam
intermediate bar had occurred at the beam-column interface
and penetrated into the joint. The yielded beam intermediate
bar would have no spare tension capacity to act as shear-
resisting reinforcement due to the participation in beam
flexure. Figure 6(a) and (c) show Specimen 1B8 had similar
hysteretic behavior with respect to that of Specimen 0T0.
The beam intermediate bar of Specimen 1B8, however,
introduced more internal forces acting on the joint, generated
a larger value of strength, and led to a earlier shear failure of
the joint (Table 3). It is evident that beam intermediate bars
passing through the joint cannot be used as a replacement for
conventional horizontal joint core hoops as shear rein-
forcement. Similar findings had been reported by Wong,
Priestley, and Park.16
Fig. 11—Measured strain of intermediate beam bar in
Specimen 1B8.
Fig. 12—Vertical-tie action of intermediate column bar.
452
In contrast to the beam intermediate bar, the column inter-
mediate bars would not be highly stressed in tension because
the columns are expected to remain elastic. Thus they can be
expected to contribute to the vertical tie mechanism of the
joint shear resistance. As shown in Fig. 12(a), the measured
strains of the column intermediate bar in Specimen 2T55
were all kept in elastic range throughout the tests. The
measured strains of Gauge 6 were greater than those of
Gauges 5 and 7 on a column intermediate bar passing
through the joint of Specimen 2T55. This shows column
intermediate bars are able to transfer vertical tie force Fv,
which was shown in Fig. 12(b). The other aspect on the bond
capacity should be carefully examined.
Because the column intermediate bars pass through the
tension zones of column, the tie force Fv, shown in Fig. 12(b),
can be transferred to the concrete only by the bond stresses
along the embedment length. According to ACI 318-02,3 the
development length ld of the uncoated column bar in the
normalweight concrete is
l 9 f y  ---------------
db 
----d- = ------ --------- - (12)
db 10 f ′  c + K tr
c
where db is the nominal diameter of the bar to be developed;
fy is the yield strength of the bar; c is the spacing or cover
dimension; and Ktr is the transverse reinforcement index. By
taking the terms db/(c + Ktr) = 0.4 for the joint region and
ld = hb /2 (hb is the beam depth), the tie-force Fv in Fig. 12(b)
can be roughly estimated as
F T1 – T2 h A tv f yv – T 2
- ≈ 1.4 f c′ ----b- ≤ ------------------------
------v- = ---------------- - (13)
A tv A tv db A tv
where Atv is the area of the vertical tie; T1 is the tension force
of the vertical tie at the mid-depth of the joint; T2 is the
tensile force of the vertical tie at column face as a result of
its participation in column flexure; and fyv is the yield
strength of the column intermediate bars.
CONCLUSIONS
Based on the cyclic load tests of nine exterior beam-
column joints, and within the limitations of the test data, the
following conclusions are made:
1. The joint hoops are found to act as a tension tie as well
as to constrain the crack width. The current ACI requirements,
viewing the joint hoop as confining the concrete core, are
unnecessary and very difficult for construction. The test
results indicate that less hoop reinforcement with wider
vertical spacing up to 300 mm could be used without signif-
icantly affecting the performance of joints;
2. Without a high axial load in column, a beam-column
joint without any hoop can exhibit satisfactory seismic
behavior as long as the joint is provided with adequate shear
strength according to the SST model;
3. The deterioration of beam-column joint under
displacement reversals could be effectively restrained by the
elastic joint hoops. For beam-column joints, where there is a
need for sustained strength under deformation reversals, it is
recommended to design the joints with adequate shear
strength and with a sufficient number of hoops to remain in
elastic range under earthquake loading. The required number
ACI Structural Journal/May-June 2005
 of hoops can be determined by viewing joint hoops as
tension tie and using the equation of SST model;
4. Beam intermediate bars passing through the joint cannot
be used to replace horizontal joint hoops as shear reinforcement
due to their participation in beam flexure; and
5. To function as vertical tie for beam-column joint, the
available tensile capacity as well as the bond transfer ability
of the column intermediate bars should be carefully checked.
ACKNOWLEDGMENTS
The authors would like to thank the National Science Council of the
Republic of China for financially supporting this research under Projects
NSC 88-2211-E-011-011, NSC 89-2211-E-011-011, NSC 90-2211-E-011-
055, and NSC 91-2811-E-011-003.
REFERENCES
1. Paulay, T.; Park. R.; and Priestley, M. J. N., “Reinforced Concrete
Beam -Column Joints under Seismic Actions,” ACI JOURNAL, Proceedings
V. 75, No. 11, Nov. 1978, pp. 585-593.
2. NZS 3101:1995, “Concrete Structures Standard, NZS 3101,” Standards
Association of New Zealand, Wellington, 1995, 256 pp.
3. ACI Committee 318, “Building Code Requirements for Structural
Concrete (ACI 318-02) and Commentary (318R-02),” American Concrete
Institute, Farmington Hills, Mich., 2002, 443 pp.
4. Joint ACI-ASCE Committee 352, “Recommendations for Design of
Beam-Column Joints in Monolithic Reinforced Concrete Structures
(ACI 352R-02),” American Concrete Institute, Farmington Hills, Mich.,
2002, 40 pp.
5. Jirsa, J. O., ed., Design of Beam-Column Joints for Seismic Resistance,
SP-123, American Concrete Institute, Farmington Hills, Mich., 1991,
518 pp.
6. Hwang, S. J., and Lee, H. J., “Analytical Model for Predicting Shear
ACI Structural Journal/May-June 2005
View publication stats
Strengths of Exterior Reinforced Concrete Beam-Column Joints for
Seismic Resistance,” ACI Structural Journal, V. 96, No. 5, Sept.-Oct. 1999,
pp. 846-857.
7. Hwang, S. J., and Lee, H. J., “Analytical Model for Predicting Shear
Strengths of Interior Reinforced Concrete Beam-Column Joints for Seismic
Resistance,” ACI Structural Journal, V. 97, No. 1, Jan.-Feb. 2000, pp. 35-44.
8. Hwang, S. J., and Lee, H. J., “Strength Prediction for Discontinuity
Regions by Softened Strut-and-Tie Model,” Journal of Structural Engineering,
ASCE, V. 128, No. 12, Dec. 2002, pp. 1519-1526.
9. “CEB-FIP Model Code 1990,” Thomas Telford Services, Ltd., London,
for Comité Euro-International du Béton, Lausanne, Switzerland, 1993, 437 pp.
10. Liao, T. F., “Shear Strength of Reinforced Concrete Beam-Column
Joints for Seismic Resistance,” master’s thesis, Department of Construction
Engineering, National Taiwan University of Science and Technology,
Taipei, Taiwan, 1999, 194 pp. (in Chinese)
11. Wang, K. C., “Seismic Design and Detailing of Reinforced Concrete
Exterior Beam-Column Joints,” master’s thesis, Department of Construction
Engineering, National Taiwan University of Science and Technology,
Taipei, Taiwan, 2002, 301 pp. (in Chinese)
12. Tsai, H. H., “A Study of Detailing of Shear Reinforcement of High
Strength Concrete Beam-Column Joints for Seismic Resistance,” master’s
thesis, Department of Construction Engineering, National Taiwan University
of Science and Technology, Taipei, Taiwan, 2000, 263 pp. (in Chinese)
13. Ehsani, M. R., and Wight, J. K., “Exterior Reinforced Concrete
Beam-to-Column Connections Subjected to Earthquake-Type Loading,”
ACI JOURNAL, Proceedings V. 82, No. 4, July-Aug. 1985, pp. 492-499.
14. Vecchio, F. J., and Collins, M. P., “Compression Response of
Cracked Reinforced Concrete,” Journal of Structural Engineering, ASCE,
V. 119, No. 12, Dec. 1993, pp. 3590-3610.
15. Zhang, L. X. B., and Hsu, T. T. C., “Behavior and Analysis of 100 MPa
Concrete Membrane Elements,” Journal of Structural Engineering, ASCE,
V. 124, No. 1, Jan. 1998, pp. 24-34.
16. Wong, P. K. C.; Priestley, M. J. N.; and Park, R., “Seismic Resistance
of Frames with Vertically Distributed Longitudinal Reinforcement in
Beams,” ACI Structural Journal, V. 87, No. 4, July-Aug. 1990, pp. 488-498.
453