,
Steel, and SteelConcrete Coupling Beams
Kent A. Harries
1
, M.EERI, Bingnian Gong
2
, and Bahram M. Shahrooz
3
Corresponding (first) author: Kent A. Harries
Mailing address: USCCivil Engineering, 300 Main Street, Columbia, SC 29208
Phone: 803 777 0671
Fax: 803 777 0670
Email address: harries@engr.sc.edu
Submission date for review copies: February 7, 2000
Submission date for cameraready copy: June 21, 2000
(KAH) Dept. of Civil and Environmental Engineering, University of South Carolina, Columbia, SC 29208
(BG) CBM Engineers, Inc., 1700 W. Loop So., Suite 830, Houston, TX 77027
(BMS) Dept. of Civil and Environmental Engineering, University of Cincinnati, Cincinnati, OH 452210071
Behavior and Design of Reinforced Concrete,
Steel, and SteelConcrete Coupling Beams
Kent A. Harries
4
, M.EERI, Bingnian Gong
5
, and Bahram M. Shahrooz
6
The efficiency of coupled wall systems to resist lateral loads is well
known. In order for the desired behavior of the coupled wall system to be
attained, the coupling beams must be sufficiently strong and stiff. The
coupling beams, however, must also yield before the wall piers, behave in a
ductile manner, and exhibit significant energy absorbing characteristics. This
paper reviews the current state of the art for the design of conventional
reinforced concrete, diagonally reinforced concrete, steel and composite steel
concrete coupling beams. Although not exhaustive, critical aspects of the
design of these systems are presented.
INTRODUCTION
An efficient structural system for seismic resistance can be achieved if openings in
structural walls can be arranged in a regular pattern as shown in Figure 1. In this manner, a
number of individual wall piers are coupled together to produce a system having large lateral
stiffness and strength. By coupling individual flexural walls, the lateral load resisting
behavior changes to one where overturning moments are resisted partially by an axial
compressiontension couple across the wall system rather than by the individual flexural
action of the walls. As a result, a significantly stiffer lateral load resisting system is achieved.
The beams that connect individual wall piers are referred to as coupling beams. Coupling
beams are analogous to and serve the same structural role as link beams in eccentrically
braced frames. Wellproportioned coupling beams above the second floor generally develop
plastic hinges simultaneously, and are subjected to similar end rotations over the height of the
structure (Aktan and Bertero 1981, AristizabalOchoa 1982, Shiu et al. 1981). Hence,
dissipation of input energy can be distributed over the height of the building in the coupling
beams rather than concentrating it predominantly in the firststory wall piers (Aktan and
Bertero 1984, Fintel and Ghosh 1982, Park and Paulay 1975). For optimum performance, the
energy dissipating mechanism should involve formation of plastic hinges in most of the
coupling beams and at the base of each wall. This mechanism is similar to the strong
columnweak girder design philosophy for ductile moment resisting frames (Paulay 1971,
Park and Paulay 1975). In order for the desired behavior of the coupled wall system to be
attained, the coupling beams must be sufficiently strong and stiff. The coupling beams,
however, must also yield before the wall piers, behave in a ductile manner, and exhibit
significant energy absorbing characteristics.
(KAH) Dept. of Civil and Environmental Engineering, University of South Carolina, Columbia, SC 29208
(BG) CBM Engineers, Inc., 1700 W. Loop So., Suite 830, Houston, TX 77027
(BMS) Dept. of Civil and Environmental Engineering, University of Cincinnati, Cincinnati, OH 452210071
Coupling
Beams
(a) (b) (c)
Figure 1. Examples of coupled wall geometry.
The degree of coupling of a coupled wall system is defined as the ratio of the total
overturning moment resisted by the coupling action (i.e., the axial load couple formed in the
walls) to the total overturning moment. The coupling action between wall piers is developed
through shears in the coupling beams. The hysteretic characteristics of coupling beams,
therefore, may substantially affect the overall response of a coupled wall system particularly
for structures having a high degree of coupling (Aktan and Bertero 1981, Harries 1999,
Paulay 1986, Paulay and Taylor 1981, Paulay 1980). Coupling beams should be designed to
avoid over coupling, which causes the system to act as a single pierced wall with little frame
action. Similarly, light coupling should also be avoided as it causes the system to behave like
two isolated walls (Aktan and Bertero 1987, 1984, 1981, AristizabalOchoa 1987, 1982,
AristizabalOchoa et al. 1979, Lybas and Sozen 1977, Shiu et al. 1981, 1984).
Reinforced concrete coupling beams may have conventional longitudinal reinforcement,
diagonal reinforcement or a hybrid reinforcing steel arrangement. Coupling beams may also
be fabricated from rolled or builtup steel shapes, or have a composite steelconcrete design.
Additionally, recent investigations have proposed alternative hybrid designs.
REINFORCED CONCRETE COUPLING BEAMS
BEHAVIOR
It has been shown that the ductility capacity of a coupled wall system increases with
increasing degrees of coupling (Harries 1998). The degree of coupling is a function of the
relative stiffness and strength of the beams and walls. Saatcioglu et al. (1987) showed that
coupling beams must be capable of developing displacement ductilities in excess of 6 for
system ductilities of about 4 to be achieved.
Several researchers have investigated novel approaches for improving the ductility and
energy absorption capacity of reinforced concrete coupling beams. For spantodepth ratios
less than about 2, a system using specially detailed diagonal reinforcement, developed by
Paulay (e.g., Paulay and Binney 1974), has been shown to significantly improve the reversed
cyclic loading response. Shiu et al. (1978) confirmed the improved behavior of diagonally
reinforced beams over conventionally reinforced beams. However, these tests also
demonstrated that for larger span to depth ratios (2.5 and 5) the diagonal reinforcement was
not as efficient due to its lower angle of inclination and hence its reduced contribution in
resisting shear.
The stiffness required of "efficient" coupling beams usually results in moment arm to
effective depth ratios (a/d ratio) of less than 2 (Aktan and Bertero 1981). Indeed, a/d ratios of
1 or less are common (Aktan and Bertero 1981, Bertero 1980, Bertero et al. 1974, Paulay
1977, Paulay and Binney 1974, Paulay 1971). As the a/d ratio decreases, arching action
contributes more significantly to shear strength than flexural behavior (Aktan and Bertero
1981, Paulay 1977, Paulay and Binney 1974, Paulay 1971). In conventionally reinforced
coupling beams, both top and bottom reinforcement may simultaneously undergo tension as a
result of diagonal cracking. Thus, it is recommended that the increased ductility due to
compression bars not be accounted for (Park and Paulay 1975, Paulay 1971). Coupling
beams require special detailing particularly if ' f 5 . 0 v
c u
> ( ' f 6 v
c u
> in psi units) and
a/d<3.0.
Figure 2. Examples of reinforced concrete coupling beams (Harries et al. 1998).
Conventional reinforcement, i.e., longitudinal bars with transverse steel (see Figure 2a),
performs satisfactorily at nominal shear stresses of ' f 25 . 0
c
( ' f 3
c
) or less when the
flexural mode controls the failure mechanism (Aktan and Bertero 1981, AristizabalOchoa
1987). Sliding shear at the face of the wall begins to affect the response of the beams in the
range of ' f 3 . 0
c
( ' f 5 . 3
c
) to ' f 5 . 0
c
( ' f 6
c
). By providing intermediate midheight
longitudinal bars, the hysteretic response is improved and strength deterioration due to shear
is delayed (Aktan and Bertero 1984, Scriber and Wight 1978, Tassios et al. 1996). Beams
with intermediate bars do not perform well when the shear stress is greater than ' f 5 . 0
c
( ' f 6
c
). Providing cranked diagonal reinforcement near the beam ends has been shown to
improve the hysteretic behavior by preventing sliding shear, and by spreading the hinging
regions away from the wall face (Paparoni 1972, AristizabalOchoa 1987). This detail,
however, poses construction difficulties and results in extra cost.
8  No 25
8  No 25
2  No 15
midside
7
0
0
m
m
400 mm
No 10 hoops
@ 90 mm
4  No 30 in
No 10 hoops
@ 100 mm
4  No 30 in
No 10 hoops
@ 100 mm
7
0
0
m
m
400 mm
No 10 hoops
@ 300 mm
(a) Conventionally reinforced coupling beam
(b) Diagonally reinforced coupling beam
3000 mm 1.3l =
1200 mm
d
1200 mm
1.3l = 1450 mm
d
wall steel
not shown
wall steel
not shown
Fulllength diagonal reinforcement, as shown in Figure 2b, is a relatively simpler
detailing alternative, and has been found to maintain stable hysteretic behavior provided the
bars are adequately confined by spirals or hoops, and sufficient concrete cover is provided
(Aktan and Bertero 1981, Paulay and Binney 1974). This confinement provides resistance
against buckling to the individual bars and the bundled diagonal bar group. Such diagonal
reinforcement is perhaps the only successful solution for reducing the potential for sliding
shear, and enhancing the hysteretic characteristics of coupling beams having a/d ratios as
large as 3.33 (Aktan and Bertero 1981).
Considering that higher modes contribute to the response of coupling beams more than to
walls (Aktan and Bertero 1981, Mahin and Bertero 1975), these elements are subjected to
large inelastic end rotations and reversals. As a result, adequate anchorage of the coupling
beam reinforcement in the walls is necessary. The possibility of buckling of compression bars
at the beamwall interface (where large cracks are anticipated) is also minimized by
providing effective confinement (Aktan and Bertero 1981, Barney et al. 1978, Paulay 1980,
Paulay and Santhakumar 1976, Paulay and Binney 1974). It is noted that anchorage and
confinement requirements often make these diagonally reinforced coupling beams very
difficult to construct due to congestion at the center of the beam and at the wall faces. (Recall
that there will often be concentrated vertical wall steel at these interfaces.) Figure 3 shows a
typical example of a diagonally reinforced coupling beam and its wall anchorage.
Figure 3. Diagonally reinforced concrete coupling beam (photo courtesy of Reinforcing Steel
Institute of Canada).
Table 1 summarizes the guidelines for the selection and detailing of reinforced concrete
coupling beams for seismic design given by the ACI 31895 and ACI 31899 Building Code
Requirements for Structural Concrete (ACI 1995, 1999), CSA A23.394 Design of Concrete
Structures (CSA 1994), and NZS 3101 Concrete Structures Standard (NZS 1995). When no
seismic design is required, coupling beams may be detailed as regular reinforced concrete
beams. It is noted, nonetheless, that these beams do typically exhibit very high shear to
moment ratios requiring special attention.
concentrated vertical
reinforcing at face
of wall
diagonal
primary
reinforcement
Table 1. Guidelines for selecting reinforced concrete coupling beams
Governing Limits
L/d shear stress, v
(psi)
Type of
Coupling
Beam
Design Criteria
ACI 31895 Clause 21.6
2 ≥ L/d
,
`
.

+ ≤
y
f
n
? '
c
f v 167 . 0
2 ≤ L/d
,
`
.

+ ≤
y
f
n
? '
c
f
c
a v
where 0.167 < α
c
≤ 0.25
ACI 31895 does not have provisions for diagonally
reinforced concrete members
ACI 31899 Clause 21.6.7
4 / ≥ d L
,
`
.

+ ≤
y
f
n
? '
c
f
c
a v
where 0.167 < α
c
≤ 0.25
conventionally
reinforced
beams may be designed as flexural
members according to Clause 21.3,
however the commentary appears to
invoke Clause 21.6.4
4 / ≤ d L
' 83 . 0
c
f v ≤
diagonally or
conventionally
reinforced
see above or below
2 / ≤ d L
' 83 . 0
' 33 . 0
c
f v
c
f v
≤
≥
diagonally
reinforced
diagonal reinforcement to carry shear
(A
vd
f
y
≥ V
n
/sinα)
CSA A23.394 Clauses 21.3 and 21.5
4 / ≥ d L
' ) / ( 1 . 0
c
f d L v ≤
conventionally
reinforced
i) beams capable of significant flexural
hinging (V
r
≥ 2M
p
/L)
ii) provide confinement and antibuckling
reinforcement
4 / ≤ d L
'
c
f v ≤
diagonally
reinforced
i) diagonal reinforcement to carry shear
(φ
s
A
s
f
y
≥ V
u
/sinα)
ii) provide adequate confinement and
anchorage for diagonal bars
NZS 31011995 Clauses 12.4.6.2 and 9.4.4.2
' ) / ( 1 . 0
c
f d L v ≤
conventionally
reinforced
i) beams capable of significant flexural
hinging (V
r
≥ 2M
p
/L)
ii) provide confinement and antibuckling
reinforcement
not
specified
' 25 . 0
c
f v ≥
some diagonal
reinforcement
required
i) diagonal reinforcement to carry shear
(φ
s
A
s
f
y
≥ V
u
/sinα)
ii) provide adequate confinement and
anchorage for diagonal bars
The ACI 31895 seismic provisions do not specifically address details for coupling
beams. Clause 21.6.5.7 limits the nominal shear strength of a “horizontal wall segment” to
'
c cp
f 0.83A (
'
c cp
f 10A ). The accompanying commentary states that the coupling beam is
“in effect, a pier rotated through 90 deg.” Applying this interpretation allows the nominal
shear strength of a coupling beam to be determined from Clauses 21.6.5.2 or 21.6.5.3,
depending on the spantodepth ratio of the beam. If this interpretation is applied, Clause
21.6.5.3 permits the shear stress in a very short coupling beam (L/d < 2.0) to be greater than
that in a longer beam. From the previous discussion of coupling beam behavior, it is clear
that this interpretation runs counter to observed beam behavior. ACI 31895 does not address
diagonally reinforced concrete beams in any way.
ACI 31899 has adopted specific provisions for coupling beams. For beams having span
todepth ratios, L/d, greater than 4.0, coupling beams are designed as flexural members
satisfying the requirements of Clause 21.3. It is noted, however, that Clause 21.6.4.5
(formerly 21.6.5.7 in 31895) and the accompanying commentary continues to suggest that
“horizontal wall segments” may be treated as “pier[s] rotated through 90 deg.” as discussed
above. For coupling beams having spantodepth ratios between 2 and 4, Clause 21.6.7.2
allows the use of diagonal or conventional reinforcement. For beams having spantodepth
ratios less than 2 and a factored shear force exceeding
'
c
f 0.33 (
'
c
f 4 ), diagonal
reinforcement is required.
In Table 1, it is noted that the Canadian standard CSA A23.3 provides the clearest
direction as to the use of diagonal reinforcement: diagonal reinforcement requirements are
triggered by both spantodepth ratios and shear stresses in the beam. The New Zealand
standard NZS 3101, on the other hand, combines these triggers into one: shear stress,
normalized by the spantodepth ratio. NZS 3101 does, however, require diagonal
reinforcement if the shear stress exceeds an absolute threshold value of
'
c
f 0.25 (
'
c
f 3 ).
Furthermore, NZS 3101 Commentary Clause C9.4.4.2 appears to permit shear to be resisted
by a combination of diagonal and conventional reinforcement providing that the contribution
of the conventional reinforcement remains below the threshold value. For seismic design, the
New Zealand Standard does recommend, however, that 100% of the shear forces be carried
by diagonal reinforcement in any case where diagonal reinforcement is required. The design
and detailing requirements for diagonally reinforced concrete coupling beams are essentially
identical in both the Canadian and New Zealand standards.
ANALYSIS OF WALLS COUPLED WITH REINFORCED CONCRETE COUPLING
BEAMS
Geometric Modeling
A simple twodimensional equivalent frame model suitable for elastic analysis of coupled
wall systems is illustrated in Figure 4. Such a model can be used to compute beam and wall
design forces. This model has several drawbacks; the most significant of which is that it does
not simulate the shift of the neutral axis in the wall piers beyond elastic limit. Other models
appropriate for inelastic analysis are available (Kunnath and Reinhorn 1994, Kabeyasawa et
al. 1983, Teshigawara et al. 1996, Milev and Kabeyasawa 1996, Vulcano and Bertero 1987).
Multiple pier walls may be similarly modeled or may be reduced to a simpler twopier model
using further analytic tools (StaffordSmith and Coull 1991).
Rigid
Elements
Frame
Elements
L
Figure 4. Two dimensional equivalent frame model of coupled wall system.
Modeling Beam and Wall Properties
Elastic analysis is sufficient to determine a theoretical degree of coupling for a coupled
wall structure and thus the distribution of internal forces in the walls and beams. Local
inelasticities at the beamwall interfaces, redistribution of forces between coupling beams and
from the tension wall to the compression wall and strain hardening effects all contribute to
reducing the theoretical degrees of coupling. Reduced section properties, accounting for
cracking and loss of stiffness due to cycling can be used to determine a more appropriate
value for the theoretical degree of coupling. Table 2 shows the reduced stiffness values
suggested by the current ACI, CSA and NZS standards. ACI 31899 recommends the use of
“any set of reasonable assumptions…[which are] consistent throughout the analysis” and
recommends using one half of the gross section properties for flexural members. It is
interesting to note that, using the reduction factors recommended by the New Zealand
Standard, one computes effectively stiffer coupling beams and more flexible wall piers,
resulting in a larger degree of coupling, and hence greater beam forces, than one computes
with the ACI or Canadian recommendations.
Table 2. Recommended reduced member stiffnesses
Member ACI 31899 CSA A23.394 NZS 31011995
1
compression wall in flexure 0.70EI (uncracked) 0.80EI 0.45EI
tension wall in flexure 0.35EI (cracked) 0.50EI 0.25EI
compression wall axial 1.00EA 1.00EA 0.80EA
tension wall axial 0.35EI (inferred) 0.50EA 0.50EA
average wall axial
2
not applicable 0.75EA
not applicable
conventionally reinforced
coupling beam
0.35EI
2
) / ( 3 1
20 . 0
L h
EI
+
2
) / ( 8 1
40 . 0
L h
EI
+
diagonally reinforced
coupling beam
not applicable
2
) / ( 3 1
40 . 0
L h
EI
+
2
) / ( 7 . 2 7 . 1
40 . 0
L h
EI
+
1. NZS 3101 has different recommendations for different limit states. The values corresponding to the most
critical limit state are given here.
2. CSA A23.3 suggests that an average axial wall stiffness is appropriate to simplify analyses.
Design
As shown in Table 1, depending on the shear span, coupling beams are designed
differently.
Flexural Failure Mode
Coupling beams are designed and detailed as standard beams if both L/d ≥ 4 and
d b ' f 0.33 V
w c u
≤ ( d b ' f 4 V
w c u
≤ ). If the structure is located in high seismic regions (UBC
Zone 4, NEHRP Seismic Design Category D or E), provisions from ACI 31899 Section 21.3
are used. ACI 31899 Section 21.9 is followed for structures in moderate seismic regions
(UBC Zone 3, NEHRP Seismic Design Category C). It is noted that ACI 31899 provisions
for minimum beam width (Clauses 21.3.1.3 and 21.3.1.4) may be waived if rational analysis
shows that lateral stability of the beam is adequate, or lateral stability is provided by
alternative means. Often the provision of an integral slab will be sufficient to provide lateral
stability.
The longitudinal flexural bars in the beam should be anchored adequately in the wall
piers. For high seismic regions (UBC Zone 4, NEHRP Seismic Design Category D or E),
development length requirements in ACI 31899 Clause 21.5.4 must be met in addition to
those in ACI 31899 Chapter 12. For moderate seismic regions (UBC Zone 3, NEHRP
Seismic Design Category C), development lengths for the longitudinal bars based on ACI
31899 Chapter 12 are considered sufficient.
Shear Failure Mode
If L/d < 4 and d b ' f 0.33 V
w c u
> ( d b ' f 4 V
w c u
> ), two intersecting groups of diagonal
bars are to be used to reinforce coupling beams. Although it is preferable for each
intersecting group to contain the same number and size of reinforcing bar, this is often not
practical. Where this is the case, it is necessary to ensure that the amount of steel provided in
each diagonal is the same. If L/d < 4 and d b ' f 0.33 V
w c u
< ( d b ' f 4 V
w c u
< ), diagonal
reinforcement may also be used. Each group of diagonal bars must have a minimum of four
bars placed in a core with a minimum depth and width equal to the greater of b
w
/5 in the
plane of the diagonal bars or b
w
/2 in the plane of the beam (ACI 31899 Clause 21.6.7.4 (a)).
The diagonal bars must be confined according to ACI 31899 clauses 21.4.4.1 through
21.4.4.3. The required area of one leg of the diagonal reinforcement, A
vd
is (ACI 31899
Clause 21.6.7.4):
d b f f A V
w c y vd n
' 83 . 0 sin 2 ≤ · α (1)
Although the diagonal reinforcement is intended to prevent sliding shear, the value of
coupling beam shear is limited to d b f
w c
' 83 . 0 ( d b f
w c
' 10 ). Minimum horizontal and
vertical reinforcement requirements for deep beams (ACI 31899 clauses 11.8.9 and 11.8.10)
must also be respected.
Considering the importance of adequate anchorage for the diagonal bars, it has been
recommended that the development length, calculated based on Chapter 21 of ACI 31899,
inside the wall piers be increased by as much as 50%, if possible (Paulay 1986). Although
there are no provisions in the Standards for increasing the development length, it is
recommended that these bars be considered as “top cast” bars and that appropriate factors be
applied as shown in Figure 2 (Harries et al. 1998). Often geometric constraints within the
wall will require that these bars be provided with bent or hooked anchorages (see Figure 3).
As previously stated, this detail may be very difficult to construct.
STEEL OR STEELCONCRETE COMPOSITE COUPLING BEAMS
Structural steel members provide a viable alternative to reinforced concrete coupling
beams. Furthermore, a steel member may be encased with varying levels of longitudinal and
transverse reinforcement. The resulting steelconcrete coupling beams are referred to as
composite coupling beams herein. The advantages of steel or composite coupling beams
become particularly apparent in cases where height restrictions do not permit the use of deep
reinforced concrete beams, or where the required capacities and stiffness cannot be
developed economically by a concrete beam.
Similar to reinforced concrete coupling beams, energy dissipation characteristics of steel
or composite coupling beams play an important role in the overall performance. Depending
on the coupling beam length, steel coupling beams may be detailed to dissipate a major
portion of the input energy by flexure or by shear. For most coupling beams, however, it is
more advantageous to design the coupling beams as “shear critical,” or shear yielding
members since such members exhibit a more desirable mode of energy dissipation. Such a
choice is not possible for reinforced concrete members.
The coupling beamwall connection depends on whether the wall boundary element is
reinforced conventionally or contains embedded structural steel columns. In the latter case,
the connection is effectively a joint between a steel beam and column, as shown in Figure 5.
Standard design methods for steel structures can be used for this connection (AISC 1995,
1997); however, proper measures must be taken to prevent failures similar to those observed
in steel momentresisting frames during recent earthquakes. Furthermore, such a composite
wall system will not require significant amounts of boundary element reinforcement.
steel boundary column
shear studs
moment connection
face of wall
steel
coupling beam
holes in beam
wall steel not shown
Figure 5. Coupling beam connection to embedded steel column.
If the wall boundary element is reinforced with longitudinal and transverse reinforcing
bars, a typical connection involves embedding the coupling beam into the wall and
interfacing it with the boundary element, as shown in Figure 6. In addition to boundary
element reinforcing, embedded steel members may also be provided with welded vertical
reinforcing passing through their flanges or shear studs to resist horizontal loads (PCI 1992).
The vertical bars may be attached to the flanges through mechanical half couplers which are
welded to the flanges. These bars must have adequate development length, and must be large
enough to provide enough stiffness to the coupling beam. Shahrooz et al. (1992, 1993) have
found 25.4 mm diameter (#8) bars to be sufficient for this purpose. Such details result in
greater congestion in the boundary element. It has been shown that these additional welded
appendages are not necessary provided that the embedment region is sufficiently long and
that there is adequate boundary element reinforcing to control cracking at the beamwall
interface (Harries 1995, Harries et al. 1997, Shahrooz et al. 1993). Harries et al. recommend
that twothirds of the required vertical boundary element steel be located within a distance of
one half the embedment length from the face of the wall. Furthermore, the width of the
boundary element steel should not exceed 2.5 times the coupling beam flange width.
Satisfying these requirements will provide adequate control of the gap that opens at the beam
flanges upon cycling. With typical boundary element designs, these requirements are easily
met.
Figure 6. Steel coupling beam embedment details (Harries et al. 1998).
No. 10 hoops and hairpins
No. 10 hoops and hairpins
SECTION AA
SECTION BB
SECTION AA
SECTION BB
A A
B B
A A
B B
(a) flexure critical steel coupling beam (b) shear critical steel coupling beam
900
900
1000
14  No. 35
No. 10 hoops @ 200
12  No. 25
No. 10 hoops @ 150
No. 10 @ 150 vertical e.f.
No. 10 @ 100 horizontal e.f.
No. 10 @ 150 vertical e.f.
No. 10 @ 100 horizontal e.f.
As is illustrated in Figure 6a, lighter “flexure critical” coupling beams require more
boundary element steel in the wall since a greater proportion of the overturning moment is
carried by the wall. This may also result in the need to provide “barbells” at the toes of the
walls, further complicating construction.
It is not necessary, nor is it practical, to pass boundary element reinforcing through the
web of the embedded coupling beam. It is however necessary to provide good confinement in
the region of the embedded web. This may be accomplished using hairpins and cross ties
parallel to the web as shown in Figure 6 (section AA). Additionally, vertical boundary
element reinforcement in this region may also be relied upon to provide significant
confinement to the embedded region (Harries et al. 1997).
STEELCONCRETE COMPOSITE COUPLING BEAMS
The selection of the embedded steel beam in a composite coupling beam typically
neglects the contribution of the encasing concrete. The contribution of the encasing concrete,
however, must be accounted for in subsequent stiffness and embedment length calculations.
The presence of the encasing concrete does permit the detailing requirements of the
embedded member to be relaxed. Gong and Shahrooz (1998) have shown that stiffeners are
not required in the case of a composite coupling beam. The surrounding concrete is sufficient
to provide adequate stability and thus ductility to the embedded steel beam.
ANALYSIS OF WALLS COUPLED WITH STEEL OR STEELCONCRETE COMPOSITE
COUPLING BEAMS
Stiffness
For steel coupling beams, standard methods are used to calculate the stiffness. The
stiffness of steelconcrete composite coupling beams needs to account for the increased
stiffness due to encasement. Stiffness based on gross transformed sections should be used to
calculate the upper bound values of demands in the walls, most notably wall axial force. The
cracked transformed section moment of inertia may be used when deflection limits or
coupling beam shear angles are checked. A previous study indicates that the additional
stiffness due to the presence of a composite floor slab is lost shortly after coupling beams
undergo small deformations (Shahrooz and Gong 1998). Until additional experimental data
become available, it is recommended to include the participation of the floor slab when the
stiffness of steelconcrete composite coupling beams are computed. Effective flange width
for T beams as specified in ACI 8.10.2 may be used for this purpose. Deflection limits should
be checked neglecting the contribution of the floor slab.
Geometric Modeling
Previous studies (Shahrooz et al. 1992, 1993, Gong et al. 1997, 1998, Harries et al. 1997)
suggest that steel or steelconcrete composite coupling beams are not fixed at the face of the
wall. As part of design calculations, the additional flexibility needs to be taken into account
to ensure that wall forces and lateral deflections are computed with reasonable accuracy.
Based on experimental data (Shahrooz et al. 1992, 1993, Gong et al. 1997, 1998), the
“effective fixed point” of steel or steelconcrete composite coupling beams may be taken as
approximately onethird of the embedment length from the face of the wall. Thus, the
effective clear span, L of the frame element representing the coupling beam in the model
shown in Figure 4 is:
e clear
L L L 6 . 0 + · (2)
The aforementioned method assumes that the walls have been modeled by column
elements located at the centroid of wall piers as shown in Figure 4. Finite element modeling
of structures have become more common in recent years where shell elements are used to
model the wall piers. These elements are typically connected to nodes along fictitious
column lines which are placed at the wall boundaries. One may attempt to account for the
additional flexibility of steel/steelconcrete composite coupling beams by shifting the column
lines, which define the walls, to a distance equal to 1/3 of the embedment length from the
face of the wall. However, this approach will alter the wall properties, and other methods are
necessary. A possible approach is to obtain equivalent values for the coupling beam axial
area and moment of inertia in order to incorporate the additional beam flexibility. Equations
to obtain the lower and upper bound values of the equivalent area and moment of inertia have
been derived (Tunc et al. 2000). Another alternative is to use an effective stiffness.
Approaching the same decrease in stiffness from another point of view, Harries et al.
(1998) propose the use of an effective stiffness of the steel coupling beam of 0.60kEI. The
factor k represents the reduction in flexural stiffness due to shear deformations:
1
2
12
1
−
]
]
]
+ ·
w
GA L
EI
k
λ
(3)
If the modeling technique used includes shear deformations, k = 1.
With this approach, the effective length of the beam in increased only to account for
spalling at the face of the wall:
c L L
clear
2 + · (4)
This approach does not require a preliminary design to determine the coupling beam
embedment length. The effective stiffness of 0.60kEI has been observed to agree very well
with experimental observations of beams having significantly different embedment details
(Harries 1995)
DESIGN
Steel Coupling Beams
Well established guidelines for shear links in eccentricallybraced frames (AISC 1997)
may be used to design and detail steel coupling beams. The expected coupling beam rotation
angle plays an important role in the required beam details such as the provision of stiffeners.
This angle is computed with reference to the collapse mechanism shown in Figure 7 which
corresponds to the expected behavior of coupled wall systems, i.e., plastic hinges at the base
of walls and at the ends of coupling beams. The value of θ
p
is taken as C
d
θ
e
, where the
elastic interstory drift angle, θ
e
, is computed under code level lateral loads (e.g., ICBO 1994;
NEHRP 1997). Knowing the value of θ
p
, shear angle, γ
p
is calculated as:
L
L
wall p
p
θ
γ · (5)
Note that in this equation the additional flexibility of the steel or composite coupling
beams may be taken into account by increasing the length of coupling beam as indicated in
Equations 2 or 4. This approach provides a reasonable approximation even if shell elements
have been used to model wall piers.
SteelConcrete Composite Coupling Beams
Previous research on steelconcrete composite coupling beams (Shahrooz et al. 1992,
1993, Gong et al. 1997, 1998) indicates that nominal encasement around steel beams
provides adequate resistance against flange and web buckling. Therefore, composite coupling
beams may be detailed without web stiffeners. Due to inadequate data regarding the influence
of encasement on local buckling, minimum flange and web thicknesses similar to steel
coupling beams should still be used.
Figure 7. Determination of coupling beam angle of rotation.
Coupling Beam Embedment
The coupling beam is embedded in the wall such that its capacity can be developed. A
number of methods may be used to calculate the necessary embedment length. Either method
can be used although the equations proposed by Marcakis and Mitchell (1980) generally
result in slightly longer embedment lengths.
Mattock and Gaafar (1982)
This model is based on mobilizing an internal moment arm between bearing forces C
f
and
C
b
as shown in Figure 8a. A parabolic distribution of bearing stresses is assumed for C
b
, and
C
f
is computed by a uniform stress equal to 0.85f'
c
. The bearing stresses are distributed over
the beam flange width, b
f
. Following these assumptions and calibrating against experimental
data, the required embedment length, L
e
, may be determined from:
]
]
]
]
]
+
−
,
`
.

·
e
e f
f
wall
c u
L
a
L b
b
t
f V
88 . 0
22 . 0 58 . 0
' 05 . 4
1
1
66 . 0
β
β (6)
L
L
wa ll
γ
p
θ
p
γ θ
p p
=
L
L
wall
plastic hinges
Marcakis and Mitchell (1980)
Using slightly different assumed stress distributions shown in Figure 8b, and assuming a
rigidbody motion of the embedded element, the following expression may be used for
determining the required embedment length, L
e
. This expression has also been calibrated
against experimental data.
) (
6 . 3
1
) ( ' ' 85 . 0
c L
e
c L b f
V
e
e c c
u
−
+
−
·
φ
(7)
Where “e” is the eccentricity from the midspan of the coupling beam to the center of the
effective embedment as shown in Figure 8b.
In either method, the value of the shear span, “a” is taken as one half of coupling beam
length if the inflection point is assumed to be at the midspan. It is suggested that the shear
span be increased to account for spalling of the walls as calculated in Equation 4.
The significant difference between these methods is that Marcakis and Mitchell account
for spreading of the compression forces within the wall. As a result, the width of the effective
compression zone, b’, may be greater than the width of the beam flange, b
f
.
Figure 8. Methods for determining embedment capacity.
c c L/2 = a L/2 = a
e
V
u
V
u
spalled cover
concrete
C C
L L
L  c
2
e
L  c
e
L
e
L
e
C
b
C
b
C
f
C
f
ε
f
= 0.003 ε
f
= 0.003
ε
b
ε
b
x
f
β
1
x
f
1/3(L c)
e
x
b
βx
b
0.85f
c
0.85f
c
αf
c
b < 2.5b
f
b
f
C
b
3/4(L c)
e
assumed
strain
distribution
assumed
stress
distribution
(a) Mattock and Gaafar (1982) (b) Marcakis and Mitchell (1980)
spalled cover
concrete
load spreading
Calculation of Embedment Length
The embedment length, L
e
, required to develop the coupling beam shear, V
u
, may be
calculated by either of the two aforementioned methods.
For steel coupling beams, V
u
is taken as 1.5 times the plastic shear capacity of the steel
member, V
p
:
w f y p u
t t h F x V V ) 2 ( 6 . 0 5 . 1 5 . 1 − · · (8)
For “flexure critical” steel beams, the shear capacity may be taken as the shear
corresponding to the development of the plastic moment capacity:
L
M
L
M
V
r
s
r
u
2
5 . 1
2
35 . 1 · ·
φ
(9)
The contribution of concrete encasement to the shear capacity of composite coupling
beams needs to be taken into account when the embedment length is calculated (Gong et al.,
1997 and 1998):
( )
]
]
]
,
`
.

+ + − · + ·
s
d f A
d b f t t h F V V V
y s
w c w f y RC p u
' 167 . 0 ) 2 ( 6 . 0 56 . 1 ) ( 56 . 1 (10)
Additional “transfer bars” attached to the beam flanges (as discussed above) can
contribute to the capacity of the embedment. The embedment length computed by Mattock
and Gaafar (1982) or Marcakis and Mitchell (1980) can be modified to account for the
additional strength. To ensure that the calculated embedment length is sufficiently large to
avoid excessive inelastic damage in the connection region, it is recommended that the
contribution of “transfer bars” be neglected.
DESIGN EXAMPLES
Examples included in Appendix A illustrate the designs of coupling beams discussed in
this paper. Details of the beam and wall designs contained in Appendix A may be found in
Harries (1996) and Harries et al. (1998). These examples are intended for illustrative
purposes.
CONCLUSIONS
The efficiency of coupled wall systems to resist lateral loads is well known. In order for
the desired behavior of the coupled wall system to be attained, the coupling beams must be
sufficiently strong and stiff. The coupling beams, however, must also yield before the wall
piers, behave in a ductile manner and exhibit significant energy absorbing characteristics.
This paper reviews the current state of the art for the design of conventional reinforced
concrete, diagonally reinforced concrete, steel and composite steelconcrete coupling beams.
Although not exhaustive, critical aspects of the design of these systems are presented.
In the case of diagonally reinforced concrete coupling beams, only included in ACI
provisions for the first time in ACI 31899, considerable additional guidance in design and
detailing may be found in following Canadian (CSA) or New Zealand Standards (NZS).
Steel coupling beams may be efficiently designed and detailed as link beams in
eccentrically braced frames. When encased in concrete, considerable stability is imparted to
the beam allowing simplified details to be used.
The proper design of any coupled system must ensure the desired hierarchy of hinging. In
addition it is critical that hinges in the coupling beam remain in their clear span and do not
propagate into the walls. As such proper design of the embedment region is critical. There are
a number of methods of determining the required embedment length and details of steel or
composite coupling beams. All of these methods are based on an assumed rigid body rotation
of the embedment region of the coupling beam.
APPENDIX A – COUPLING BEAM DESIGN EXAMPLES
Salient points of the design of four types of coupling beam are presented. The beams are
located in eighteen storey prototype structures located in Vancouver, Canada. The plan of
these prototype structures is shown in Figure A1. Seismic design forces were determined
using the 1995 National Building Code of Canada equivalent lateral force method. Base
shears were determined to be in the range of 0.036W to 0.041W depending on the prototype
structure. These same base shears would be found by applying the NEHRP equivalent lateral
force method using R=5.8 to 6.7. NEHRP recommends a value of R=6 for either prototype
structure.
Figure A1. Plan of prototype structures.
Each of the 3000 mm long prototype beams are designed as either conventionally
reinforced concrete coupling beams or “flexure critical” steel coupling beams. Similarly, the
1200 mm beams are designed as both diagonally reinforced concrete coupling beams and
“shear critical” steel beams.
CONVENTIONALLY REINFORCED CONCRETE COUPLING BEAMS
The final details of the 3000 mm long prototype conventionally reinforced concrete
coupling beam are shown in Figure 2a. The design forces on these beams were determined to
be 467 kN in shear and 700 kNm in flexure. The spantodepth ratio, L/d, for these beams is
4.3 and the shear stress is approximately ' f 42 . 0
c
, thus diagonal reinforcement is required.
The design progressed according to standard beam design methods of ACI 31899 or CSA
A23.394.
Typical floor plan
9000
14 000
1200
6400 6400
500
Fully (heavily) coupled
base shear = 0.036W
Prototype cores
Note: wall thickness = 500 mm (storys 1 to 5)
= 400 mm (storys 6 to 18)
9000
14 000
3000
5500 5500
500
Partially (lightly) coupled
base shear = 0.041W
900 mm x 900 mm "barbells"
(storys 1 to 5)
800 mm x 800 mm "barbells"
(storys 6 to 18)
42 000
6000 (typ.)
3
0
0
0
0
6
0
0
0
(
t
y
p
.
)
550 x 550 column (typ.)
200 mm thick flat plate
N
5500
end bays
DIAGONALLY REINFORCED CONCRETE COUPLING BEAMS
The final details of the 1200 mm long prototype diagonally reinforced concrete coupling
beam are shown in Figure 2b. The design shear force for these beams was determined to be
770 kN. The spantodepth ratio is 1.7 and the shear stress is approximately ' f 25 . 0
c
, thus
conventional reinforcement is permitted. The design progressed according to standard beam
design methods of ACI 31899 or CSA A23.394:
Determine required dimensions in diagonally reinforced beam
Assume diagonal bar bundles consisting of 2 layers of No. 30 bars. Bar bundles,
including ties are 100 mm deep and 200 mm wide (see Fig. 2b).
d = 700  40  100/2 = 610 mm d’ = 90 mm
tan α = (h  2d’)/L
u
= 700  (2 x 90)/1200 therefore: α = 24
o
Selection of main diagonal reinforcement
Diagonal reinforcement is determined from Equation A1:
2
2783
24 sin 400 85 . 0 2
770000
sin 2
mm
x x x f
V
A
y s
n
s
·
°
· ·
α φ
(A1)
therefore use 4  No. 30 bars in each diagonal (A
s
= 2800 mm
2
)
Embedment and confinement of diagonal reinforcement
Recommended embedment length = 1.5L
d
= 1.5 x 1480 = 2220 mm
Seismic hoops are required over entire length of diagonal with spacing not exceeding:
6d
b
= 6 x 30 = 180 mm
24d
b
(hoops) = 24 x 10 = 240 mm
100 mm
Thus, provide No. 10 closed hoops at a spacing of 100 mm over the length of the
diagonals.
“FLEXURE CRITICAL” STEEL COUPLING BEAMS
Ideally, the selection of a steel coupling beam must provide adequate stiffness for the
desired degree of coupling, while limiting overstrength. Furthermore, the steel section must
be narrow enough to fit within the concentrated wall reinforcement. For longer, “flexure
critical” coupling beams, these criteria can often be satisfied with standard rolled shapes. For
shorter spans, however, builtup sections may be required.
Details of the “flexure critical” coupling beam are shown in Figure A2a. For the 3000
mm prototype steel coupling beam the design shear was determined to be 597 kN,
corresponding to a moment of 895 kNm. For these forces, a W610 x 140 (W24 x 94; inch
pound units) section with a 915 mm embedment region was designed.
At this point, it is important to verify that the section will indeed behave as expected. In
this case, that it is “flexure critical.” The factored shear capacity is determined to be 1340 kN
while the shear corresponding to the unfactored probable moment capacity is:
kN
x
x x x
c L
Z F x
c L
M
y p
734
40 2 3000
2510 300 5 . 1 2
2
5 . 1 2
2
2
·
+
·
+
·
+
(A2)
Thus the beam is “flexure critical.”
Required embedment length
The required embedment length is determined from the following relation, (based on
Marcakis and Mitchell, 1980), where the factored shear resistance of the embedment, V
u
, is
taken as 1.35M
r
/2Lφ
s
= 1.35 x 727 / 0.9 = 1090 kN (see Figure 8b).
) (
6 . 3
1
) ( ' ' 85 . 0
c L
e
c L b f
V
e
e c c
u
−
+
−
·
φ
(A3)
Because of the presence of a “barbell” at the wall toe, b’ may be taken as 2.5b =
2.5 x 230 = 575 mm. Taking b’ = 575 mm, e = (3000 + L
e
+ 40) / 2, and c = 40 mm. Solving
Equation A3 yields: L
e
= 915 mm.
It is necessary to ensure that there is adequate concentrated boundary element
reinforcement, A
sc
, to develop V
e
:
2
3206
400 85 . 0
1090000
mm
x f
V
A
y s
e
sc
· · ·
φ
(A4)
This requirement is essentially satisfied, as this steel represents a small portion of what is
required in the region of concentrated reinforcement, and there will typically be some
overstrength provided. Details of the embedment region of this prototype are shown in Figure
6a.
At the beamwall interface, doublesided stiffeners are provided to simplify the formwork
and to aid in the force transfer between the steel beam and the concrete wall. Due to the high
moment to shear ratio and the need to ensure lateral stability, singlesided stiffeners are
located 335 mm from the face of the wall.
The required embedment length may also be determined using the equation proposed by
Mattock and Gaafar (1982):
]
]
]
]
]
+
−
,
`
.

·
e
e f
f
wall
c u
L
a
L b
b
t
f V
88 . 0
22 . 0 58 . 0
' 05 . 4
1
1
66 . 0
β
β (A5)
If the bearing area width, t
wall
, is assumed to be equal to the entire 900 mm width of the
“barbell,” the required embedment length is found to be 725 mm. It is noted that if the width
of the effective bearing area is limited to 2.5b
f
(Marcakis and Mitchell 1980), the required
embedment length is increased to 870 mm.
Figure A2. Prototype steel coupling beams.
“SHEAR CRITICAL” STEEL COUPLING BEAM
For the prototype structure having the 1200 mm long steel coupling beam and a high
degree of coupling, the coupling beam was designed to yield in shear while remaining elastic
in flexure. A 600 mm deep builtup section having an embedment length of 1000 mm was
selected. The high shear demands on this beam necessitated the provision of singlesided
stiffeners spaced at 300 mm over the entire span, these were continued into the embedment to
provide stability in the event of excessive spalling. Figure A2b shows details for this “shear
critical” steel coupling beam. Details of the embedment region are shown in Figure 6b.
The design forces for this beam were 764 kN and 458 kNm for shear and moment,
respectively. The design criteria called for a depth of no more than 600 mm (similar to the
“flexure critical” and concrete beams). The resulting builtup section has dimensions shown
in Figure A2.
The required embedment length is determined from Equation A3 to be 1000 mm, similar
to that required for the longer “flexure critical” beam. Similarly, using Equation A5, the
required embedment length is determined to be 780 mm.
3000 mm
335 mm
915 mm
1200 mm
50 mm
1000 mm
300 mm typ.
(a) “flexure critical” prototype
(b) “shear critical” prototype
W610 x 140
builtup section
d = 617 mm
b = 230 mm
t = 22 mm
t = 13 mm
f
w
d = 600 mm
b = 225 mm
t = 25 mm
t = 9 mm
f
w
COMPOSITE COUPLING BEAMS
The previously designed steel coupling beams are assumed to be encased as shown in
Figure A3. The longitudinal reinforcement consists of four 10M (diameter = 11.3 mm) corner
bars and two 10M midside bars. The transverse reinforcement consists of 10M closed hoops
spaced at 390 mm. For simplicity, it is assumed that the steel beams are identical to those
used previously, i.e., the contribution of concrete is ignored when the steel coupling beams
are sized. This would often be the case for composite coupling beam design. The dimensions
of the steel beams are given in Figure A2.
Figure A3. Cross section of composite coupling beam.
The overall dimensions for both the 3000 mm and 1200 mm long coupling beams are h =
780 mm and b
w
= 390 mm. These dimensions were obtained based on assuming that the
longitudinal bars would be at about 25 mm from the corners of the flanges, and by providing
40 mm of cover to the transverse reinforcement. The concrete compressive strength is taken
as 35 MPa, and the yield strength of the longitudinal bars is taken as 420 MPa. The required
embedment length for the flexural and shear critical beam is calculated in the following.
Flexural Critical Beam
The shear capacity of the coupling beam is determined as the sum the contribution from
the steel beam and that of the of the concrete encasement (Equation 9):
( )
]
]
]
,
`
.

+ + − · + ·
s
d f A
d b f t t h F V V V
y s
w c w f y RC p u
' 167 . 0 ) 2 ( 6 . 0 56 . 1 ) ( 56 . 1 (A6)
Using an effective depth, d = 780 – 81.5 = 698.5 mm, the ultimate shear capacity of the
coupling beam is found to be 2747 kN.
The required embedment length may be calculated using Equation A5 (Mattock and
Gaafar 1982). Similar to the previous example, t
wall
is taken as 900 mm. The value of b
f
is
taken as the flange width of the steel coupling beam. The value of “a” is taken as half of the
3000 mm span. By applying Equation A5, the required embedment length, L
e
, is found to be
1285 mm.
The required embedment length determined using the method proposed by Marcakis and
Mitchell (1980), given in Equation A3 is found to be 1690 mm. Similar to the previous
example, this assumes that b’ is taken as 2.5b = 2.5 (230) = 575 and e = 0.5L + c+ 0.5(L
e
c) =
1520 + 0.5L
e
(see Figure 8).
760 mm
No. 10 corner
and midside bars
No. 10 closed hoops
or hairpins @ 390 mm
390 mm
It is suggested that since the beam is “flexure critical” the value of V
p
may be replaced
with the value of the shear required to achieve the ultimate flexural capacity of the beam,
shown above to be equal to 1.35 x 2M
r
/Lφ
s
= 1.35 x 727 / 0.9 = 1090 kN. In this case, the
required shear capacity of the embedment becomes:
,
`
.

+ + · + ·
s
d f A
d b f
L
M
V
L
M
V
y s
w c
r
RC
s
r
u
' 167 . 0 56 . 1
2
5 . 1 56 . 1
2
35 . 1
φ
(A7)
For the “flexure critical” beam designed above, the required shear capacity of the
embedment region is V
u
= 1090 + 1.56(420) = 1745 kN. The require embedment length is
found to be 970 mm (Equation A5) or 1280 mm (Equation A3).
Shear Critical Beam
For the “shear critical” beam, the method is identical to that used for the “flexural
critical” composite beam although only Equation A6 is considered. The design shear force
for the embedment, V
u
is 2,026 kN. Based on Mattock and Gaafar’s model (Equation A5),
setting t
wall
= 500 mm, the necessary embedment length is 1025 mm. When using the model
proposed by Marcakis and Mitchell (Equation A3), the value of b’ is approximately 400 mm,
limited by the width of the confining steel (see Figure 8b), and e = 620 + 0.5L
e
. Based on this
method, the value of L
e
is 1350 mm.
The preceding calculations demonstrate the effect of including the capacity of the
concrete encasement. Although the embedments are longer, it has been shown that encased
steel coupling beams do not require the significant stiffener details (see Figure A2) in order
to ensure adequate ductility and stability.
DESIGN OF WALL PIERS
The design of the wall piers is then carried out for flexure using axial loads resulting from
full plastification of all coupling beams. Figure 6 gives some indication as to the resulting
boundary element reinforcement required for these prototype structures.
ACKNOWLEDGEMENTS
The authors would like to gratefully acknowledge the input and support of members of
American Concrete Institute (ACI) Committee 335 – Composite and Hybrid Structures and
Professor Denis Mitchell of McGill University. The design methods and recommendations
are based on several projects funded by the U.S. National Science Foundation (Grants No.
BCS8922777 and BCS9319838 with Dr. Shih Chi Liu as the program director) and the
Natural Science and Engineering Research Council of Canada. Any opinions, findings, and
conclusions or recommendations expressed in this paper are those of the authors and do not
necessarily reflect the views of the sponsors.
NOTATION
a = shear span of coupling beam, typically taken as L/2
A = cross sectional area of wall pier
A
s
= area of transverse reinforcement in concrete encasement
A
sc
= area of concentrated reinforcement in boundary element of wall pier
A
vd
= area of reinforcement in each group of diagonal bars
A
w
= area of web of steel beam
b
'
= effective width of the concrete compression block
b
f
= coupling beam flange width, also the bearing width of the embedment
b
w
= web width of concrete beam or concrete encasement
c = depth of concrete cover
C
d
= deflection amplification factor defined by NEHRP (BSSC 1997)
C
f
= resultant bearing (compressive) force in embedment
d = distance from the extreme compression fiber to centroid of longitudinal tension
reinforcement in concrete section
e = eccentricity from midspan of beam to center of embedment
E = Young’s modulus of coupling beam or wall pier
f
c
’ = specified concrete compressive strength
f
y
= yield strength of reinforcing steel
F
y
= yield strength of steel beam
G = shear modulus
h = overall depth of coupling beam
I = moment of inertia of coupling beam or wall pier
k = stiffness factor accounting for shear deformations in flexural members
L = effective length of coupling beam
L
clear
= length of coupling beam measured from face of wall piers
L
e
= embedment length of the steel coupling beams inside the wall piers
L
wall
= distance between centroidal axes of the wall piers
M
r
= moment capacity of steel beam
s = spacing of transverse reinforcement
t
f
= flange thickness of steel beam
t
w
= web thickness of steel beam
t
wall
= thickness of wall pier
v = shear stress
v
u
= factored shear stress
V
c
= factored shear force resisted by concrete
V
n
= nominal shear capacity of coupling beam
V
p
= plastic shear capacity of coupling beam
V
r
= factored shear capacity of coupling beam
V
RC
= shear force resisted by reinforced concrete encasement
V
u
= factored shear force on section
x
b
= length of compression block at back of embedment
x
f
= length of compression block at front of embedment
α = angle of inclination between diagonal reinforcement and longitudinal reinforcement
α
c
= coefficient defining relative contribution of concrete strength to wall strength
β or β
1
= ratio of the average concrete compressive strength to the maximum stress
γ
p
= shear angle
λ = shape factor
θ
p
= plastic interstory drift angle
θ
e
= elastic interstory drift angle computed under code level lateral loads
ρ
n
= ratio of area of distributed reinforcement parallel to plane of shear resistance to the
concrete area perpendicular to that plane
φ
c
= material resistance factor for concrete = 0.60
φ
s
= strength reduction factor for steel = 0.90
REFERENCES CITED
American Concrete Institute (ACI) Committee 318, 1995, Building Code Requirements for
Reinforced Concrete and Commentary (ACI 31895/ACI 318R95), Farmington Hills, MI, 353 pp.
American Concrete Institute (ACI) Committee 318, 1999, Building Code Requirements for
Reinforced Concrete and Commentary (ACI 31899/ACI 318R99), Farmington Hills, MI.
American Institute of Steel Construction (AISC), 1995, Manual of Steel Construction, Load and
Resistance Factor Design, Chicago, IL.
American Institute of Steel Construction (AISC), 1997, Seismic Provisions for Structural Steel
Buildings, Chicago, IL.
Aktan, A. E. and Bertero, V. V., 1981, The seismic resistant design of R/C coupled structural walls,
Report No. UCB/EERC81/07, Earthquake Engineering Research Center, University of
California, Berkeley.
Aktan, A. E. and Bertero, V. V., 1984, Seismic response of R/C framewall structures, ASCE Journal
of the Structural Division, 110 (ST 8), 18031821.
Aktan, A. E. and Bertero, V. V., 1987, Evaluation of seismic response of RC building loaded to
failure, ASCE Journal of the Structural Division, 113 (ST 5), 1092 1108.
AristizabalOchoa, J. D., 1982, Dynamic response of coupled wall systems, ASCE Journal of the
Structural Division, 108 (ST 8), 18461857.
AristizabalOchoa, J. D., 1987, Seismic behavior of slender coupled wall systems, ASCE Journal of
the Structural Division, 113 (ST 10), 22212234.
Barney, G. B. et al., 1978, Earthquake Resistant Structural Walls  Test of Coupling Beams, Report to
NSF, submitted by Portland Cement Association, Research and Development, Skokie, IL.
Bertero, V. V., 1980, Seismic behavior of R/C structural walls, Proceedings of the Seventh World
Conference of Earthquake Engineering, Istanbul, Turkey, 1980.
Bertero, V. V., et al., 1974, Hysteretic behavior of reinforced concrete flexural members with web
reinforcement, Report No. UCB/EERC 749, Earthquake Engineering Research Center,
University of California, Berkeley.
Building Seismic Safety Council (BSSC) 1998, NEHRP Recommended Provisions for Seismic
Regulations for New Buildings and Other Structures (FEMA 302) and Commentary (FEMA 303),
1997 edition, Washington, DC.
Canadian Standards Association (CSA), 1994, CSA A23.394 Design of Concrete Structures, Rexdale,
Canada.
Fintel, M. and Ghosh, S. K., 1982, Case study of aseismic design of a 16story coupled wall using
inelastic dynamic analysis, ACI Journal, 79 (3), 171179.
Gong B., Shahrooz B. M., 1998, Seismic behavior and design of composite coupled wall systems,
Report No. UCCII 98/01, Cincinnati Infrastructure Institute, Cincinnati, OH.
Gong B. Shahrooz B. M., and Gillum A. J., 1997, Seismic behavior and design of composite coupling
beams, Composite Construction in Steel and Concrete III, ASCE, New York, 258271.
Gong B., Shahrooz B. M., and Gillum A. J., 1998, Cyclic response of composite coupling beams, ACI
Special Publication 174 – Hybrid and Composite Structures, Farmington Hills, MI, 89112.
Harries, K. A., 1995, Seismic Design and Retrofit of Coupled Walls Using Structural Steel, Ph.D.
Thesis, McGill University, 229 pp.
Harries, K. A., 1996, Design and analysis of four prototype eighteen storey coupled wall structures,
McGill University Department of Civil Engineering and Applied Mechanics Structural
Engineering Series Report No. 9601, 122 pp.
Harries, K. A., 1998, Ductility and deformability of coupling beams in reinforced concrete coupled
walls, Proceedings of the Eighth Canadian Conference on Earthquake Engineering, Vancouver,
June 1999, 475481.
Harries, K. A., Mitchell, D., Redwood, R. G. and Cook, W. D., 1997, Seismic design of coupling
beams  A case for mixed construction, Canadian Journal of Civil Engineering, 24(3), 448459.
Harries, K. A., Mitchell, D., Redwood, R. G. and Cook, W. D., 1998, Seismic design and analysis of
prototype coupled wall structures, Canadian Journal of Civil Engineering, 25 (5), 808818.
International Council of Building Officials (ICBO), 1994, Uniform Building Code, Volume 2,
Structural Engineering Design Provisions, Whitier, CA.
Kabeyasawa T., Shiohara H., Otani S., and Aoyama H., 1983, Analysis of the fullscale sevenstory
reinforced concrete test structure, Journal of the Faculty of Engineering, University of Tokyo,
XXXVII (2), 431478.
Kunnath, S. K. and Reinhorn, A. M., 1994, IDARC2D Version 3.1: Inelastic Damage Analysis of
Reinforced Concrete Building Structures, State University of New York at Buffalo.
Lybas, J. M. and Sozen, M. A., 1977, Effect of beam strength and stiffness on dynamic behavior of
reinforced concrete walls, SRS No. 444, University of Illinois, UrbanaChampaign.
Mahin, S. A. and Bertero, V. V., 1975, An evaluation of some methods for predicting seismic
behavior of reinforced concrete buildings, Report No. UCB/EERC 7515, Earthquake Engineering
Research Center, University of California, Berkeley.
Marcakis K. and Mitchell D., 1980, Precast concrete connections with embedded steel members.
Prestressed Concrete Institute Journal, 25 (4), 88116.
Mattock A. H. and Gaafar G. H., 1982, Strength of embedded steel sections as brackets. ACI Journal,
79 (2).
Milev J. I. and Kabeyasawa T., 1996, Modeling of reinforced concrete shear walls in RC framewall
structures, personal communication, Tokyo.
New Zealand Standards Association (NZS), 1995, NZS 3101:1995 Concrete Structures Standard.
Paparoni, M., 1972, Model studies of coupling beams, Proceedings of the International Conference
on Tall Concrete and Masonry Buildings, 3, 671681.
Park, R. and Paulay, T., 1975 Reinforced Concrete Structures, John Wiley and Sons, Inc., New York.
Paulay, T., 1971, Coupling beams of reinforced concrete shear walls, ASCE Journal of the Structural
Division, 97 (ST 3), 843862.
Paulay, T., 1977, Coupling beams of reinforced concrete shear walls, Proceeding of a Workshop on
EarthquakeResistant Reinforced Concrete Building Construction, University of California,
Berkeley, July 1115, 1977, 14521462.
Paulay, T., 1980, Earthquake resisting shear walls  New Zealand design trends, ACI Journal, 77 (3),
144152.
Paulay, T., 1986, The design of ductile reinforced concrete structural walls for earthquake resistance,
Earthquake Spectra, 2 (4), 783823.
Paulay, T. and Binney, J. R., 1974, Diagonally reinforced concrete beams for shear walls, ACI Special
Publication SP 42  Shear in Reinforced Concrete, 579598.
Paulay, T. and Santhakumar, A. R., 1976 Ductile behavior of coupled shear walls, ASCE Journal of
the Structural Division, 102 (ST 1), 93108.
Paulay, T. and Taylor, R. G., 1981, Slab coupling of earthquakeresisting shear walls, ACI Journal,
78 (2), 130140.
Precast/Prestressed Concrete Institute (PCI), 1992, PCI Design Handbook, fourth edition.
Saatcioglu, M., Derecho, A. T. and Corley, W. G., 1987, Parametric study of earthquakeresistant
coupled walls, ASCE Journal of the Structural Division, 113 (1), 141157.
Scriber, C. F. and Wight, J. K., 1978, Delaying shear strength decay in reinforced concrete flexural
members under large load reversals, Report UMEE 78 R2, Department of Civil Engineering,
University of Michigan.
Shahrooz B. M. and Gong, B., 1998, Steelconcrete coupling beams: a critical overview of design
guidelines, Structural Engineers World Wide 1998, paper T1863.
Shahrooz B. M., Remmetter M. A. and Qin F., 1992, Seismic response of composite coupled walls,
Composite Construction in Steel and Concrete II, ASCE, 429441.
Shahrooz B. M., Remmetter M. A. and Qin F., 1993, Seismic design and performance of composite
coupled walls, ASCE Journal of the Structural Division, 119 (11), 32913309.
Shiu, N. K., Barney, G. B., Fiorato, A. E. and Corley, W. G., 1978, Reversed load tests of reinforced
concrete coupling beams, Proceedings of the Central American Conference on Earthquake
Engineering, El Salvador, 239249.
Shiu, N. K., Barney, G. B., Fiorato, A. E. and Corley, W. G., 1981, Earthquake Resistant Walls 
Coupled Wall Test, Report to NSF submitted by Portland Cement Association, Research and
Development, Skokie, Illinois.
Shiu, N. K., Takayangi, T. and Corley, W.G., 1984, Seismic behavior of coupled wall systems, ASCE
Journal of the Structural Division, 110 (ST 5), 10511066.
StaffordSmith, B. and Coull, A., 1991, Tall Building Structures, Wiley Interscience.
Tassios, T. P., Moretti, M., and Bezas, A., 1996, On the coupling behavior and ductility of reinforced
concrete coupling beams of shear walls, ACI Structural Journal 93 (6), 711720.
Teshigawara M., Kato M., Sugaya K.I., 1996, Seismic test on 12story Tshaped coupled shear wall,
Proceedings of the Third Joint Technical Coordinating Committee Meeting, Hong Kong,
December 1214, 1996.
Tunc G., Deason J., and Shahrooz B. M., 2000, Seismic response of R.C. core wallsteel frame hybrid
structures, Report No. UCCII 00/01, Cincinnati Infrastructure Institute, Cincinnati, OH.
Vulcano A. and Bertero V. V., 1987, Analytical models for predicting the lateral response of RC
shear walls, Report No. UCB/EERC87/19, Earthquake Engineering Research Center, University
of California, Berkeley.
Behavior and Design of Reinforced Concrete, Steel, and SteelConcrete Coupling Beams
Kent A. Harries , M.EERI, Bingnian Gong , and Bahram M. Shahrooz
4 5 6
The efficiency of coupled wall systems to resist lateral loads is well known. In order for the desired behavior of the coupled wall system to be attained, the coupling beams must be sufficiently strong and stiff. The coupling beams, however, must also yield before the wall piers, behave in a ductile manner, and exhibit significant energy absorbing characteristics. This paper reviews the current state of the art for the design of conventional reinforced concrete, diagonally reinforced concrete, steel and composite steelconcrete coupling beams. Although not exhaustive, critical aspects of the design of these systems are presented. INTRODUCTION An efficient structural system for seismic resistance can be achieved if openings in structural walls can be arranged in a regular pattern as shown in Figure 1. In this manner, a number of individual wall piers are coupled together to produce a system having large lateral stiffness and strength. By coupling individual flexural walls, the lateral load resisting behavior changes to one where overturning moments are resisted partially by an axial compressiontension couple across the wall system rather than by the individual flexural action of the walls. As a result, a significantly stiffer lateral load resisting system is achieved. The beams that connect individual wall piers are referred to as coupling beams. Coupling beams are analogous to and serve the same structural role as link beams in eccentrically braced frames. Wellproportioned coupling beams above the second floor generally develop plastic hinges simultaneously, and are subjected to similar end rotations over the height of the structure (Aktan and Bertero 1981, AristizabalOchoa 1982, Shiu et al. 1981). Hence, dissipation of input energy can be distributed over the height of the building in the coupling beams rather than concentrating it predominantly in the firststory wall piers (Aktan and Bertero 1984, Fintel and Ghosh 1982, Park and Paulay 1975). For optimum performance, the energy dissipating mechanism should involve formation of plastic hinges in most of the coupling beams and at the base of each wall. This mechanism is similar to the strong columnweak girder design philosophy for ductile moment resisting frames (Paulay 1971, Park and Paulay 1975). In order for the desired behavior of the coupled wall system to be attained, the coupling beams must be sufficiently strong and stiff. The coupling beams, however, must also yield before the wall piers, behave in a ductile manner, and exhibit significant energy absorbing characteristics.
(KAH) Dept. of Civil and Environmental Engineering, University of South Carolina, Columbia, SC 29208 (BG) CBM Engineers, Inc., 1700 W. Loop So., Suite 830, Houston, TX 77027 (BMS) Dept. of Civil and Environmental Engineering, University of Cincinnati, Cincinnati, OH 452210071
For spantodepth ratios less than about 2. the axial load couple formed in the walls) to the total overturning moment. The degree of coupling is a function of the relative stiffness and strength of the beams and walls. a system using specially detailed diagonal reinforcement. 1984. AristizabalOchoa 1987. Saatcioglu et al. which causes the system to act as a single pierced wall with little frame action. 1981. REINFORCED CONCRETE COUPLING BEAMS BEHAVIOR It has been shown that the ductility capacity of a coupled wall system increases with increasing degrees of coupling (Harries 1998). Shiu et al. Paulay and Taylor 1981. The coupling action between wall piers is developed through shears in the coupling beams. developed by Paulay (e.e. 1981. (1978) confirmed the improved behavior of diagonally reinforced beams over conventionally reinforced beams. Several researchers have investigated novel approaches for improving the ductility and energy absorption capacity of reinforced concrete coupling beams.. 1979.Coupling Beams (a) (b) (c) Figure 1. Indeed. or have a composite steelconcrete design. Examples of coupled wall geometry. Coupling beams should be designed to avoid over coupling.5 and 5) the diagonal reinforcement was not as efficient due to its lower angle of inclination and hence its reduced contribution in resisting shear. The degree of coupling of a coupled wall system is defined as the ratio of the total overturning moment resisted by the coupling action (i. Shiu et al. light coupling should also be avoided as it causes the system to behave like two isolated walls (Aktan and Bertero 1987. Coupling beams may also be fabricated from rolled or builtup steel shapes. these tests also demonstrated that for larger span to depth ratios (2. therefore.. Lybas and Sozen 1977. Similarly.g. 1982. Harries 1999. The hysteretic characteristics of coupling beams. has been shown to significantly improve the reversed cyclic loading response. (1987) showed that coupling beams must be capable of developing displacement ductilities in excess of 6 for system ductilities of about 4 to be achieved. Additionally. Paulay 1980). may substantially affect the overall response of a coupled wall system particularly for structures having a high degree of coupling (Aktan and Bertero 1981. recent investigations have proposed alternative hybrid designs. a/d ratios of . 1984). The stiffness required of "efficient" coupling beams usually results in moment arm to effective depth ratios (a/d ratio) of less than 2 (Aktan and Bertero 1981). However. Paulay 1986. Paulay and Binney 1974). diagonal reinforcement or a hybrid reinforcing steel arrangement. Reinforced concrete coupling beams may have conventional longitudinal reinforcement. AristizabalOchoa et al.
No 25 400 mm wall steel not shown 3000 mm 1. Paulay and Binney 1974. Sliding shear at the face of the wall begins to affect the response of the beams in the range of 0. Providing cranked diagonal reinforcement near the beam ends has been shown to improve the hysteretic behavior by preventing sliding shear. it is recommended that the increased ductility due to compression bars not be accounted for (Park and Paulay 1975. Paulay 1977. AristizabalOchoa 1987). 1974. 8 . Paulay and Binney 1974. Beams with intermediate bars do not perform well when the shear stress is greater than 0. however. both top and bottom reinforcement may simultaneously undergo tension as a result of diagonal cracking.3ld = 1450 mm wall steel not shown 1200 mm (b) Diagonally reinforced coupling beam Figure 2. Paulay 1971). This detail.No 15 midside 8 . Examples of reinforced concrete coupling beams (Harries et al. Tassios et al.5 fc ' ( 6 f c ' ).1 or less are common (Aktan and Bertero 1981. As the a/d ratio decreases. i. Conventional reinforcement. In conventionally reinforced coupling beams. . 1998).25 fc ' ( 3 f c ' ) or less when the flexural mode controls the failure mechanism (Aktan and Bertero 1981. Coupling beams require special detailing particularly if v u > 0.5 fc ' ( 6 f c ' ). Bertero 1980.No 30 in No 10 hoops @ 100 mm 400 mm 700 mm 1.0.No 30 in No 10 hoops @ 100 mm No 10 hoops @ 300 mm 4 . Paulay 1977. 1996).No 25 No 10 hoops @ 90 mm 2 .3ld = 1200 mm 700 mm (a) Conventionally reinforced coupling beam 4 . and by spreading the hinging regions away from the wall face (Paparoni 1972. the hysteretic response is improved and strength deterioration due to shear is delayed (Aktan and Bertero 1984. poses construction difficulties and results in extra cost. Paulay 1971). AristizabalOchoa 1987).5 fc ' ) to 0. Thus. By providing intermediate midheight longitudinal bars. performs satisfactorily at nominal shear stresses of 0.e.5 f c ' ( v u > 6 f c ' in psi units) and a/d<3. Bertero et al. longitudinal bars with transverse steel (see Figure 2a). arching action contributes more significantly to shear strength than flexural behavior (Aktan and Bertero 1981.3 f c ' ( 3. Scriber and Wight 1978.. Paulay 1971).
When no seismic design is required. As a result. CSA A23. adequate anchorage of the coupling beam reinforcement in the walls is necessary. Mahin and Bertero 1975). that these beams do typically exhibit very high shear to moment ratios requiring special attention.394 Design of Concrete Structures (CSA 1994). The possibility of buckling of compression bars at the beamwall interface (where large cracks are anticipated) is also minimized by providing effective confinement (Aktan and Bertero 1981. Paulay and Binney 1974). and NZS 3101 Concrete Structures Standard (NZS 1995). Paulay and Santhakumar 1976. nonetheless. Barney et al. Table 1 summarizes the guidelines for the selection and detailing of reinforced concrete coupling beams for seismic design given by the ACI 31895 and ACI 31899 Building Code Requirements for Structural Concrete (ACI 1995. and sufficient concrete cover is provided (Aktan and Bertero 1981. and enhancing the hysteretic characteristics of coupling beams having a/d ratios as large as 3. (Recall that there will often be concentrated vertical wall steel at these interfaces. Considering that higher modes contribute to the response of coupling beams more than to walls (Aktan and Bertero 1981. . This confinement provides resistance against buckling to the individual bars and the bundled diagonal bar group.33 (Aktan and Bertero 1981). Paulay 1980. 1978.Fulllength diagonal reinforcement. Such diagonal reinforcement is perhaps the only successful solution for reducing the potential for sliding shear. Diagonally reinforced concrete coupling beam (photo courtesy of Reinforcing Steel Institute of Canada). as shown in Figure 2b. Paulay and Binney 1974). and has been found to maintain stable hysteretic behavior provided the bars are adequately confined by spirals or hoops. It is noted. 1999). coupling beams may be detailed as regular reinforced concrete beams. is a relatively simpler detailing alternative. these elements are subjected to large inelastic end rotations and reversals. concentrated vertical reinforcing at face of wall diagonal primary reinforcement Figure 3. It is noted that anchorage and confinement requirements often make these diagonally reinforced coupling beams very difficult to construct due to congestion at the center of the beam and at the wall faces.) Figure 3 shows a typical example of a diagonally reinforced coupling beam and its wall anchorage.
4.25 beams may be designed as flexural members according to Clause 21.6.6 Design Criteria L/d ≥ 2 L/d ≤ 2 ACI 31895 does not have provisions for diagonally reinforced concrete members where 0. Guidelines for selecting reinforced concrete coupling beams Governing Limits L/d shear stress.6.83 f c ' L/d ≤4 diagonally or conventionally reinforced diagonally reinforced see above or below diagonal reinforcement to carry shear (Avdfy ≥ Vn/sinα) L/d ≤ 2 v ≥ 0.33 f ' c v ≤ 0.167 < αc ≤ 0.7 L/d ≥ 4 v ≤ a c f '+? f c n y conventionally reinforced where 0. however the commentary appears to invoke Clause 21.3 and 21.167 f ' + ? f c n y v ≤ a c f '+? f c n y Type of Coupling Beam ACI 31895 Clause 21.2 and 9.3.5 L/d ≥4 v ≤ 0.167 < αc ≤ 0.1( L / d ) f ' c conventionally reinforced i) beams capable of significant flexural hinging (Vr ≥ 2Mp /L) ii) provide confinement and antibuckling reinforcement L/d ≤4 v≤ fc ' diagonally reinforced i) diagonal reinforcement to carry shear (φsAsfy ≥ Vu/sinα) ii) provide adequate confinement and anchorage for diagonal bars NZS 31011995 Clauses 12.2 conventionally reinforced i) beams capable of significant flexural hinging (Vr ≥ 2Mp /L) ii) provide confinement and antibuckling reinforcement not specified v ≤ 0.83 f ' c CSA A23.Table 1.25 f ' c some diagonal reinforcement required i) diagonal reinforcement to carry shear (φsAsfy ≥ Vu/sinα) ii) provide adequate confinement and anchorage for diagonal bars .4 v ≤ 0.394 Clauses 21. v (psi) v ≤ 0.4.4.6.25 ACI 31899 Clause 21.1( L / d ) f ' c v ≥ 0.
The accompanying commentary states that the coupling beam is “in effect. Clause 21. For seismic design. it is clear that this interpretation runs counter to observed beam behavior. however.33 f c' ( 4 f c' ). however.4. Vulcano and Bertero 1987). the New Zealand Standard does recommend. The New Zealand standard NZS 3101. NZS 3101 does.5. Furthermore. however.6. normalized by the spantodepth ratio. L/d.7. that 100% of the shear forces be carried by diagonal reinforcement in any case where diagonal reinforcement is required.6.4.5.4.3 permits the shear stress in a very short coupling beam (L/d < 2. that Clause 21. In Table 1. For coupling beams having spantodepth ratios between 2 and 4. combines these triggers into one: shear stress. greater than 4.0) to be greater than that in a longer beam. For beams having spantodepth ratios.5.3. From the previous discussion of coupling beam behavior.” as discussed above. 1983.6. Teshigawara et al.5.2 allows the use of diagonal or conventional reinforcement.7 in 31895) and the accompanying commentary continues to suggest that “horizontal wall segments” may be treated as “pier[s] rotated through 90 deg.6. Milev and Kabeyasawa 1996.6. the most significant of which is that it does not simulate the shift of the neutral axis in the wall piers beyond elastic limit. Clause 21. diagonal reinforcement is required. NZS 3101 Commentary Clause C9. It is noted.0. it is noted that the Canadian standard CSA A23. If this interpretation is applied. The design and detailing requirements for diagonally reinforced concrete coupling beams are essentially identical in both the Canadian and New Zealand standards.3. Multiple pier walls may be similarly modeled or may be reduced to a simpler twopier model using further analytic tools (StaffordSmith and Coull 1991).7 limits the nominal shear strength of a “horizontal wall segment” to 0. Other models appropriate for inelastic analysis are available (Kunnath and Reinhorn 1994.2 or 21. Kabeyasawa et al.3 provides the clearest direction as to the use of diagonal reinforcement: diagonal reinforcement requirements are triggered by both spantodepth ratios and shear stresses in the beam. Clause 21.6.2 appears to permit shear to be resisted by a combination of diagonal and conventional reinforcement providing that the contribution of the conventional reinforcement remains below the threshold value. a pier rotated through 90 deg. require diagonal reinforcement if the shear stress exceeds an absolute threshold value of 0. For beams having spantodepth ratios less than 2 and a factored shear force exceeding 0.6.83A cp fc' ( 10A cp f c' ). ACI 31895 does not address diagonally reinforced concrete beams in any way. 1996.The ACI 31895 seismic provisions do not specifically address details for coupling beams.” Applying this interpretation allows the nominal shear strength of a coupling beam to be determined from Clauses 21.5 (formerly 21.5. coupling beams are designed as flexural members satisfying the requirements of Clause 21. Such a model can be used to compute beam and wall design forces. depending on the spantodepth ratio of the beam. ANALYSIS OF WALLS COUPLED WITH REINFORCED CONCRETE COUPLING BEAMS Geometric Modeling A simple twodimensional equivalent frame model suitable for elastic analysis of coupled wall systems is illustrated in Figure 4.25 f c' ( 3 f c' ). This model has several drawbacks. on the other hand. ACI 31899 has adopted specific provisions for coupling beams. .
one computes effectively stiffer coupling beams and more flexible wall piers.00EA 0.35EI not applicable CSA A23.394 0. CSA and NZS standards.Frame Elements Rigid Elements L Figure 4.7 + 2.45EI 0. Two dimensional equivalent frame model of coupled wall system. 2. resulting in a larger degree of coupling.25EI 0.40 EI 1 + 8(h / L) 2 0.75EA 0. Local inelasticities at the beamwall interfaces. It is interesting to note that.00EA 0.80EA 0.7(h / L) 2 1. The values corresponding to the most critical limit state are given here.50EA 0. than one computes with the ACI or Canadian recommendations. CSA A23. accounting for cracking and loss of stiffness due to cycling can be used to determine a more appropriate value for the theoretical degree of coupling. Recommended reduced member stiffnesses Member compression wall in flexure tension wall in flexure compression wall axial tension wall axial average wall axial2 conventionally reinforced coupling beam diagonally reinforced coupling beam ACI 31899 0. .50EA not applicable 0. ACI 31899 recommends the use of “any set of reasonable assumptions…[which are] consistent throughout the analysis” and recommends using one half of the gross section properties for flexural members. NZS 3101 has different recommendations for different limit states.3 suggests that an average axial wall stiffness is appropriate to simplify analyses.40EI 1.35EI (inferred) not applicable 0.70EI (uncracked) 0. Modeling Beam and Wall Properties Elastic analysis is sufficient to determine a theoretical degree of coupling for a coupled wall structure and thus the distribution of internal forces in the walls and beams.80EI 0. using the reduction factors recommended by the New Zealand Standard. Reduced section properties.35EI (cracked) 1.50EI 1. and hence greater beam forces. Table 2 shows the reduced stiffness values suggested by the current ACI. redistribution of forces between coupling beams and from the tension wall to the compression wall and strain hardening effects all contribute to reducing the theoretical degrees of coupling.20 EI 1 + 3(h / L ) 2 0. Table 2.40 EI 1 + 3(h / L) 2 NZS 310119951 0.
The required area of one leg of the diagonal reinforcement.1. depending on the shear span. the value of coupling beam shear is limited to 0. For high seismic regions (UBC Zone 4. two intersecting groups of diagonal bars are to be used to reinforce coupling beams.7.3. calculated based on Chapter 21 of ACI 31899.8.7.4): Vn = 2 Avd f y sin α ≤ 0. NEHRP Seismic Design Category D or E).6. coupling beams are designed differently. Each group of diagonal bars must have a minimum of four bars placed in a core with a minimum depth and width equal to the greater of bw/5 in the plane of the diagonal bars or bw/2 in the plane of the beam (ACI 31899 Clause 21.1 through 21.4. Flexural Failure Mode Coupling beams are designed and detailed as standard beams if both L/d ≥ 4 and Vu ≤ 0. The diagonal bars must be confined according to ACI 31899 clauses 21.33 f c ' b w d ( Vu < 4 fc 'b w d ).4 must be met in addition to those in ACI 31899 Chapter 12.83 f c 'bwd (1) Although the diagonal reinforcement is intended to prevent sliding shear. Where this is the case. inside the wall piers be increased by as much as 50%.Design As shown in Table 1. Although it is preferable for each intersecting group to contain the same number and size of reinforcing bar. it is necessary to ensure that the amount of steel provided in each diagonal is the same. NEHRP Seismic Design Category C). Often the provision of an integral slab will be sufficient to provide lateral stability. It is noted that ACI 31899 provisions for minimum beam width (Clauses 21.9 and 11.4. it has been recommended that the development length. If the structure is located in high seismic regions (UBC Zone 4. it is recommended that these bars be considered as “top cast” bars and that appropriate factors be .9 is followed for structures in moderate seismic regions (UBC Zone 3.83 f c 'bw d ( 10 f c 'bw d ).33 f c 'b w d ( Vu ≤ 4 f c 'b w d ).3 are used.6. Considering the importance of adequate anchorage for the diagonal bars.10) must also be respected. ACI 31899 Section 21. if possible (Paulay 1986).4) may be waived if rational analysis shows that lateral stability of the beam is adequate.4.3. diagonal reinforcement may also be used. A is (ACI 31899 vd Clause 21. NEHRP Seismic Design Category C).4.3 and 21. provisions from ACI 31899 Section 21.3. Although there are no provisions in the Standards for increasing the development length. or lateral stability is provided by alternative means.33 fc 'b wd ( Vu > 4 f c 'b w d ). this is often not practical.5.1. Shear Failure Mode If L/d < 4 and Vu > 0. The longitudinal flexural bars in the beam should be anchored adequately in the wall piers. development length requirements in ACI 31899 Clause 21. NEHRP Seismic Design Category D or E).4 (a)).8. For moderate seismic regions (UBC Zone 3. Minimum horizontal and vertical reinforcement requirements for deep beams (ACI 31899 clauses 11. If L/d < 4 and Vu < 0. development lengths for the longitudinal bars based on ACI 31899 Chapter 12 are considered sufficient.
a typical connection involves embedding the coupling beam into the wall and . however. The coupling beamwall connection depends on whether the wall boundary element is reinforced conventionally or contains embedded structural steel columns. For most coupling beams. Furthermore. steel boundary column wall steel not shown shear studs steel coupling beam holes in beam moment connection face of wall Figure 5. Coupling beam connection to embedded steel column. Often geometric constraints within the wall will require that these bars be provided with bent or hooked anchorages (see Figure 3). If the wall boundary element is reinforced with longitudinal and transverse reinforcing bars. The resulting steelconcrete coupling beams are referred to as composite coupling beams herein. or where the required capacities and stiffness cannot be developed economically by a concrete beam. as shown in Figure 5. STEEL OR STEELCONCRETE COMPOSITE COUPLING BEAMS Structural steel members provide a viable alternative to reinforced concrete coupling beams. it is more advantageous to design the coupling beams as “shear critical. this detail may be very difficult to construct.” or shear yielding members since such members exhibit a more desirable mode of energy dissipation. As previously stated. 1997). the connection is effectively a joint between a steel beam and column. Such a choice is not possible for reinforced concrete members. steel coupling beams may be detailed to dissipate a major portion of the input energy by flexure or by shear. Depending on the coupling beam length. a steel member may be encased with varying levels of longitudinal and transverse reinforcement. such a composite wall system will not require significant amounts of boundary element reinforcement.applied as shown in Figure 2 (Harries et al. however. proper measures must be taken to prevent failures similar to those observed in steel momentresisting frames during recent earthquakes. Similar to reinforced concrete coupling beams. energy dissipation characteristics of steel or composite coupling beams play an important role in the overall performance. In the latter case. Standard design methods for steel structures can be used for this connection (AISC 1995. The advantages of steel or composite coupling beams become particularly apparent in cases where height restrictions do not permit the use of deep reinforced concrete beams. 1998). Furthermore.
No. these requirements are easily met. 10 hoops @ 200 No. 1997. Satisfying these requirements will provide adequate control of the gap that opens at the beam flanges upon cycling. the width of the boundary element steel should not exceed 2. 1998). recommend that twothirds of the required vertical boundary element steel be located within a distance of one half the embedment length from the face of the wall. . A A A A B B B B No. (1992. 35 No. 1993). 10 hoops and hairpins No. With typical boundary element designs. 10 @ 100 horizontal e. Harries et al. Such details result in greater congestion in the boundary element. as shown in Figure 6.f.interfacing it with the boundary element.No. Shahrooz et al. SECTION BB (a) flexure critical steel coupling beam Figure 6.f. 1993) have found 25. Furthermore. The vertical bars may be attached to the flanges through mechanical half couplers which are welded to the flanges. 10 hoops and hairpins SECTION AA 900 1000 900 12 . 10 @ 150 vertical e. and must be large enough to provide enough stiffness to the coupling beam. These bars must have adequate development length. embedded steel members may also be provided with welded vertical reinforcing passing through their flanges or shear studs to resist horizontal loads (PCI 1992). 10 hoops @ 150 No.5 times the coupling beam flange width. In addition to boundary element reinforcing. It has been shown that these additional welded appendages are not necessary provided that the embedment region is sufficiently long and that there is adequate boundary element reinforcing to control cracking at the beamwall interface (Harries 1995. 10 @ 150 vertical e.f. Harries et al. 25 No.4 mm diameter (#8) bars to be sufficient for this purpose. Steel coupling beam embedment details (Harries et al. No. SECTION BB (b) shear critical steel coupling beam SECTION AA 14 . 10 @ 100 horizontal e. Shahrooz et al.f.No.
It is however necessary to provide good confinement in the region of the embedded web. 1992.2 may be used for this purpose. 1992. nor is it practical.10. Thus. L of the frame element representing the coupling beam in the model shown in Figure 4 is: . the effective clear span. Until additional experimental data become available. the “effective fixed point” of steel or steelconcrete composite coupling beams may be taken as approximately onethird of the embedment length from the face of the wall. The stiffness of steelconcrete composite coupling beams needs to account for the increased stiffness due to encasement. The surrounding concrete is sufficient to provide adequate stability and thus ductility to the embedded steel beam. 1997). 1997. STEELCONCRETE COMPOSITE COUPLING BEAMS The selection of the embedded steel beam in a composite coupling beam typically neglects the contribution of the encasing concrete. As part of design calculations. Gong et al. A previous study indicates that the additional stiffness due to the presence of a composite floor slab is lost shortly after coupling beams undergo small deformations (Shahrooz and Gong 1998). Stiffness based on gross transformed sections should be used to calculate the upper bound values of demands in the walls.As is illustrated in Figure 6a. Gong and Shahrooz (1998) have shown that stiffeners are not required in the case of a composite coupling beam. lighter “flexure critical” coupling beams require more boundary element steel in the wall since a greater proportion of the overturning moment is carried by the wall. the additional flexibility needs to be taken into account to ensure that wall forces and lateral deflections are computed with reasonable accuracy. Gong et al. The cracked transformed section moment of inertia may be used when deflection limits or coupling beam shear angles are checked. Geometric Modeling Previous studies (Shahrooz et al. vertical boundary element reinforcement in this region may also be relied upon to provide significant confinement to the embedded region (Harries et al. however. further complicating construction. must be accounted for in subsequent stiffness and embedment length calculations. 1993. Based on experimental data (Shahrooz et al. it is recommended to include the participation of the floor slab when the stiffness of steelconcrete composite coupling beams are computed. This may also result in the need to provide “barbells” at the toes of the walls. Effective flange width for T beams as specified in ACI 8. Additionally. to pass boundary element reinforcing through the web of the embedded coupling beam. 1997) suggest that steel or steelconcrete composite coupling beams are not fixed at the face of the wall. Harries et al. Deflection limits should be checked neglecting the contribution of the floor slab. This may be accomplished using hairpins and cross ties parallel to the web as shown in Figure 6 (section AA). 1993. standard methods are used to calculate the stiffness. The presence of the encasing concrete does permit the detailing requirements of the embedded member to be relaxed. ANALYSIS OF WALLS COUPLED WITH STEEL OR STEELCONCRETE COMPOSITE COUPLING BEAMS Stiffness For steel coupling beams. most notably wall axial force. 1998. It is not necessary. 1997. The contribution of the encasing concrete. 1998).
NEHRP 1997). θe. (1998) propose the use of an effective stiffness of the steel coupling beam of 0. Finite element modeling of structures have become more common in recent years where shell elements are used to model the wall piers. where the elastic interstory drift angle.6 Le (2) The aforementioned method assumes that the walls have been modeled by column elements located at the centroid of wall piers as shown in Figure 4. 2000). the effective length of the beam in increased only to account for spalling at the face of the wall: L = L clear + 2 c (4) This approach does not require a preliminary design to determine the coupling beam embedment length. Knowing the value of θp. A possible approach is to obtain equivalent values for the coupling beam axial area and moment of inertia in order to incorporate the additional beam flexibility.e. This angle is computed with reference to the collapse mechanism shown in Figure 7 which corresponds to the expected behavior of coupled wall systems. is computed under code level lateral loads (e. The effective stiffness of 0.60kEI has been observed to agree very well with experimental observations of beams having significantly different embedment details (Harries 1995) DESIGN Steel Coupling Beams Well established guidelines for shear links in eccentricallybraced frames (AISC 1997) may be used to design and detail steel coupling beams.60kEI. γp is calculated as: γp = θ p Lwall L (5) Note that in this equation the additional flexibility of the steel or composite coupling beams may be taken into account by increasing the length of coupling beam as indicated in .. which define the walls. One may attempt to account for the additional flexibility of steel/steelconcrete composite coupling beams by shifting the column lines. plastic hinges at the base of walls and at the ends of coupling beams. The factor k represents the reduction in flexural stiffness due to shear deformations: 12 EIλ k = 1 + 2 L GAw −1 (3) If the modeling technique used includes shear deformations. Approaching the same decrease in stiffness from another point of view. These elements are typically connected to nodes along fictitious column lines which are placed at the wall boundaries. and other methods are necessary. However. Another alternative is to use an effective stiffness. this approach will alter the wall properties.g.L = Lclear + 0. Equations to obtain the lower and upper bound values of the equivalent area and moment of inertia have been derived (Tunc et al. k = 1. The value of θp is taken as Cd θe. With this approach. i. shear angle. The expected coupling beam rotation angle plays an important role in the required beam details such as the provision of stiffeners. to a distance equal to 1/3 of the embedment length from the face of the wall. ICBO 1994.. Harries et al.
85f' c. This approach provides a reasonable approximation even if shell elements have been used to model wall piers. 1998) indicates that nominal encasement around steel beams provides adequate resistance against flange and web buckling. 1997. Coupling Beam Embedment The coupling beam is embedded in the wall such that its capacity can be developed.58 − 0. Mattock and Gaafar (1982) This model is based on mobilizing an internal moment arm between bearing forces Cf and Cb as shown in Figure 8a. bf. Following these assumptions and calibrating against experimental data.Equations 2 or 4. Gong et al. minimum flange and web thicknesses similar to steel coupling beams should still be used. and Cf is computed by a uniform stress equal to 0. plastic hinges θp γp Lwall γp = θp L L Lwa ll Figure 7.05 f c ' wall bf 0. composite coupling beams may be detailed without web stiffeners. 1992. A parabolic distribution of bearing stresses is assumed for Cb. SteelConcrete Composite Coupling Beams Previous research on steelconcrete composite coupling beams (Shahrooz et al. Therefore. Due to inadequate data regarding the influence of encasement on local buckling. Le.66 0.22 β1 β1b f Le 0. the required embedment length. Either method can be used although the equations proposed by Marcakis and Mitchell (1980) generally result in slightly longer embedment lengths. A number of methods may be used to calculate the necessary embedment length.88 + a Le (6) . The bearing stresses are distributed over the beam flange width. may be determined from: t Vu = 4. 1993. Determination of coupling beam angle of rotation.
. bf. e Le .6e 1+ ( Le − c) (7) Where “e” is the eccentricity from the midspan of the coupling beam to the center of the effective embedment as shown in Figure 8b.85fc (b) Marcakis and Mitchell (1980) Figure 8.003 C L b < 2. and assuming a rigidbody motion of the embedded element. b’. the width of the effective compression zone. It is suggested that the shear span be increased to account for spalling of the walls as calculated in Equation 4. This expression has also been calibrated against experimental data.5bf Cb 1/3(L ec) 0. Vu = 0.003 distribution xb Cf Le xf εf = 0. the following expression may be used for determining the required embedment length. “a” is taken as one half of coupling beam length if the inflection point is assumed to be at the midspan. Le.Marcakis and Mitchell (1980) Using slightly different assumed stress distributions shown in Figure 8b.85φc f c ' b' ( Le − c ) 3. the value of the shear span. may be greater than the width of the beam flange. Methods for determining embedment capacity. In either method.c 2 Cb c L/2 = a spalled cover load spreading concrete bf V u Cf Le εb C L εb assumed strain ε f = 0. As a result.85f c 3/4(L e c) (a) Mattock and Gaafar (1982) assumed αfc stress distribution β1x f βxb 0. The significant difference between these methods is that Marcakis and Mitchell account for spreading of the compression forces within the wall.c Cb c L/2 = a spalled cover concrete Vu Le .
Although not exhaustive.56(V p + VRC ) = 1.. (1998). considerable additional guidance in design and detailing may be found in following Canadian (CSA) or New Zealand Standards (NZS). The embedment length computed by Mattock and Gaafar (1982) or Marcakis and Mitchell (1980) can be modified to account for the additional strength. only included in ACI provisions for the first time in ACI 31899. steel and composite steelconcrete coupling beams. Vp: Vu = 1 . Vu is taken as 1. The coupling beams. V . diagonally reinforced concrete. however.5V p = 1. may be u calculated by either of the two aforementioned methods. the coupling beams must be sufficiently strong and stiff. To ensure that the calculated embedment length is sufficiently large to avoid excessive inelastic damage in the connection region.Calculation of Embedment Length The embedment length. 1997 and 1998): A f d Vu = 1.35 2M r 2M r = 1.167 fc 'bw d + s y s (10) Additional “transfer bars” attached to the beam flanges (as discussed above) can contribute to the capacity of the embedment. . For steel coupling beams. required to develop the coupling beam shear. DESIGN EXAMPLES Examples included in Appendix A illustrate the designs of coupling beams discussed in this paper. Details of the beam and wall designs contained in Appendix A may be found in Harries (1996) and Harries et al.5 x 0. the shear capacity may be taken as the shear corresponding to the development of the plastic moment capacity: Vu = 1. Le. This paper reviews the current state of the art for the design of conventional reinforced concrete. critical aspects of the design of these systems are presented.5 times the plastic shear capacity of the steel member.6 Fy ( h − 2t f )tw ) + 0.5 Lφs L (9) The contribution of concrete encasement to the shear capacity of composite coupling beams needs to be taken into account when the embedment length is calculated (Gong et al. it is recommended that the contribution of “transfer bars” be neglected.6F y ( h − 2t f )tw (8) For “flexure critical” steel beams. In the case of diagonally reinforced concrete coupling beams. must also yield before the wall piers. behave in a ductile manner and exhibit significant energy absorbing characteristics. These examples are intended for illustrative purposes. CONCLUSIONS The efficiency of coupled wall systems to resist lateral loads is well known. In order for the desired behavior of the coupled wall system to be attained.56(0.
.Steel coupling beams may be efficiently designed and detailed as link beams in eccentrically braced frames. There are a number of methods of determining the required embedment length and details of steel or composite coupling beams. When encased in concrete. In addition it is critical that hinges in the coupling beam remain in their clear span and do not propagate into the walls. The proper design of any coupled system must ensure the desired hierarchy of hinging. As such proper design of the embedment region is critical. All of these methods are based on an assumed rigid body rotation of the embedment region of the coupling beam. considerable stability is imparted to the beam allowing simplified details to be used.
30 000 . Base shears were determined to be in the range of 0. CONVENTIONALLY REINFORCED CONCRETE COUPLING BEAMS The final details of the 3000 mm long prototype conventionally reinforced concrete coupling beam are shown in Figure 2a. The beams are located in eighteen storey prototype structures located in Vancouver. Similarly. Canada. Plan of prototype structures. The design forces on these beams were determined to be 467 kN in shear and 700 kNm in flexure.7.APPENDIX A – COUPLING BEAM DESIGN EXAMPLES Salient points of the design of four types of coupling beam are presented. These same base shears would be found by applying the NEHRP equivalent lateral force method using R=5. thus diagonal reinforcement is required.394. L/d. 900 mm x 900 mm "barbells" (storys 1 to 5) 800 mm x 800 mm "barbells" (storys 6 to 18) Note: wall thickness = 500 mm (storys 1 to 5) = 400 mm (storys 6 to 18) 550 x 550 column (typ.041W depending on the prototype structure. Each of the 3000 mm long prototype beams are designed as either conventionally reinforced concrete coupling beams or “flexure critical” steel coupling beams.) 42 000 5500 end bays 6400 9000 6400 1200 14 000 Typical floor plan Fully (heavily) coupled base shear = 0.036W Prototype cores Figure A1.036W to 0. the 1200 mm beams are designed as both diagonally reinforced concrete coupling beams and “shear critical” steel beams. The spantodepth ratio.42 f c ' . for these beams is 4. NEHRP recommends a value of R=6 for either prototype structure.) 6000 (typ. The design progressed according to standard beam design methods of ACI 31899 or CSA A23. The plan of these prototype structures is shown in Figure A1. Seismic design forces were determined using the 1995 National Building Code of Canada equivalent lateral force method.) 500 9000 N 5500 3000 5500 14 000 Partially (lightly) coupled base shear = 0.041W 200 mm thick flat plate 500 6000 (typ.8 to 6.3 and the shear stress is approximately 0.
while limiting overstrength. “FLEXURE CRITICAL” STEEL COUPLING BEAMS Ideally. Furthermore. these criteria can often be satisfied with standard rolled shapes.DIAGONALLY REINFORCED CONCRETE COUPLING BEAMS The final details of the 1200 mm long prototype diagonally reinforced concrete coupling beam are shown in Figure 2b. d = 700 . 10 closed hoops at a spacing of 100 mm over the length of the diagonals.40 .2d’)/Lu = 700 . .25 fc ' .5Ld = 1. For the 3000 mm prototype steel coupling beam the design shear was determined to be 597 kN. For these forces. 2b).No. builtup sections may be required. a W610 x 140 (W24 x 94. including ties are 100 mm deep and 200 mm wide (see Fig. The spantodepth ratio is 1. For longer.5 x 1480 = 2220 mm Seismic hoops are required over entire length of diagonal with spacing not exceeding: 6db = 6 x 30 = 180 mm 24db (hoops) = 24 x 10 = 240 mm 100 mm Thus. The design progressed according to standard beam design methods of ACI 31899 or CSA A23.394: Determine required dimensions in diagonally reinforced beam Assume diagonal bar bundles consisting of 2 layers of No. 30 bars. however. corresponding to a moment of 895 kNm. thus conventional reinforcement is permitted. For shorter spans.7 and the shear stress is approximately 0.100/2 = 610 mm d’ = 90 mm tan α = (h . provide No. The design shear force for these beams was determined to be 770 kN. Details of the “flexure critical” coupling beam are shown in Figure A2a.(2 x 90)/1200 therefore: α = 24o Selection of main diagonal reinforcement Diagonal reinforcement is determined from Equation A1: As = Vn 770000 = = 2783mm 2 2φ s f y sin α 2 x 0. the steel section must be narrow enough to fit within the concentrated wall reinforcement.85 x 400 x sin 24° (A1) therefore use 4 . “flexure critical” coupling beams. the selection of a steel coupling beam must provide adequate stiffness for the desired degree of coupling. 30 bars in each diagonal (As = 2800 mm2) Embedment and confinement of diagonal reinforcement Recommended embedment length = 1. Bar bundles. inchpound units) section with a 915 mm embedment region was designed.
At the beamwall interface.6 e 1+ ( Le − c ) (A3) Because of the presence of a “barbell” at the wall toe. to develop Ve: Asc = Ve 1090000 = = 3206 mm 2 φ s f y 0. that it is “flexure critical. and c = 40 mm.35Mr/2Lφs = 1. Asc. In this case. The required embedment length may also be determined using the equation proposed by Mattock and Gaafar (1982): t Vu = 4.85φc f c ' b' ( Le − c ) 3. it is important to verify that the section will indeed behave as expected.66 0.05 f c ' wall b f 0.58 − 0. as this steel represents a small portion of what is required in the region of concentrated reinforcement. and there will typically be some overstrength provided. b’ may be taken as 2.At this point. singlesided stiffeners are located 335 mm from the face of the wall.5F y Z L + 2c = 2 x1.” Required embedment length The required embedment length is determined from the following relation.9 = 1090 kN (see Figure 8b). Details of the embedment region of this prototype are shown in Figure 6a. 1980).5 x 230 = 575 mm. Due to the high moment to shear ratio and the need to ensure lateral stability. Taking b’ = 575 mm. Solving Equation A3 yields: Le = 915 mm.5b = 2. e = (3000 + Le + 40) / 2. Vu = 0. where the factored shear resistance of the embedment.35 x 727 / 0.88 + a Le (A5) . Vu. is taken as 1. It is necessary to ensure that there is adequate concentrated boundary element reinforcement.85 x 400 (A4) This requirement is essentially satisfied. (based on Marcakis and Mitchell.22 β 1 β 1b f Le 0. doublesided stiffeners are provided to simplify the formwork and to aid in the force transfer between the steel beam and the concrete wall.5 x300 x2510 = 734kN 3000 + 2 x 40 (A2) Thus the beam is “flexure critical.” The factored shear capacity is determined to be 1340 kN while the shear corresponding to the unfactored probable moment capacity is: 2M p L + 2c = 2 x1.
The resulting builtup section has dimensions shown in Figure A2. Prototype steel coupling beams. Similarly. similar to that required for the longer “flexure critical” beam. Details of the embedment region are shown in Figure 6b. these were continued into the embedment to provide stability in the event of excessive spalling.” the required embedment length is found to be 725 mm. is assumed to be equal to the entire 900 mm width of the “barbell. (b) “shear critical” prototype Figure A2.If the bearing area width. . The design criteria called for a depth of no more than 600 mm (similar to the “flexure critical” and concrete beams).5bf (Marcakis and Mitchell 1980). Figure A2b shows details for this “shear critical” steel coupling beam. It is noted that if the width of the effective bearing area is limited to 2. using Equation A5. the coupling beam was designed to yield in shear while remaining elastic in flexure. the required embedment length is increased to 870 mm. The design forces for this beam were 764 kN and 458 kNm for shear and moment. “SHEAR CRITICAL” STEEL COUPLING BEAM For the prototype structure having the 1200 mm long steel coupling beam and a high degree of coupling. The required embedment length is determined from Equation A3 to be 1000 mm. twall. 3000 mm d = 617 mm b = 230 mm t f = 22 mm t w = 13 mm 915 mm 335 mm W610 x 140 (a) “flexure critical” prototype d = 600 mm b = 225 mm t f = 25 mm t w = 9 mm builtup section 50 mm 1200 mm 1000 mm 300 mm typ. the required embedment length is determined to be 780 mm. A 600 mm deep builtup section having an embedment length of 1000 mm was selected. respectively. The high shear demands on this beam necessitated the provision of singlesided stiffeners spaced at 300 mm over the entire span.
COMPOSITE COUPLING BEAMS The previously designed steel coupling beams are assumed to be encased as shown in Figure A3.6 Fy ( h − 2t f )tw ) + 0. By applying Equation A5.e. The value of “a” is taken as half of the 3000 mm span. Cross section of composite coupling beam.5L + c+ 0.5(Lec) = 1520 + 0. Le.. t all is taken as 900 mm.167 fc 'bw d + s y s (A6) Using an effective depth. is found to be 1285 mm. The required embedment length determined using the method proposed by Marcakis and Mitchell (1980). The required embedment length may be calculated using Equation A5 (Mattock and Gaafar 1982). the contribution of concrete is ignored when the steel coupling beams are sized. The transverse reinforcement consists of 10M closed hoops spaced at 390 mm. Similar to the previous example.5b = 2. The value of bf is w taken as the flange width of the steel coupling beam. No. The longitudinal reinforcement consists of four 10M (diameter = 11. This would often be the case for composite coupling beam design. The required embedment length for the flexural and shear critical beam is calculated in the following. 10 closed hoops or hairpins @ 390 mm 390 mm Figure A3. d = 780 – 81. Similar to the previous example.5 (230) = 575 and e = 0. this assumes that b’ is taken as 2.56(0. For simplicity. . and by providing 40 mm of cover to the transverse reinforcement. The dimensions of the steel beams are given in Figure A2. The concrete compressive strength is taken as 35 MPa.5Le (see Figure 8). and the yield strength of the longitudinal bars is taken as 420 MPa. The overall dimensions for both the 3000 mm and 1200 mm long coupling beams are h = 780 mm and bw = 390 mm.5 mm. the required embedment length. i. Flexural Critical Beam The shear capacity of the coupling beam is determined as the sum the contribution from the steel beam and that of the of the concrete encasement (Equation 9): A f d Vu = 1. 10 corner and midside bars 760 mm No. given in Equation A3 is found to be 1690 mm.3 mm) corner bars and two 10M midside bars.56(V p + VRC ) = 1.5 = 698. the ultimate shear capacity of the coupling beam is found to be 2747 kN. it is assumed that the steel beams are identical to those used previously. These dimensions were obtained based on assuming that the longitudinal bars would be at about 25 mm from the corners of the flanges.
56VRC = 1. Based on this method.9 = 1090 kN.35 As f y d 2M r 2M r + 1. the required shear capacity of the embedment region is V = 1090 + 1. Shih Chi Liu as the program director) and the Natural Science and Engineering Research Council of Canada.It is suggested that since the beam is “flexure critical” the value of Vp may be replaced with the value of the shear required to achieve the ultimate flexural capacity of the beam. the method is identical to that used for the “flexural critical” composite beam although only Equation A6 is considered. Shear Critical Beam For the “shear critical” beam. the value of Le is 1350 mm. The design methods and recommendations are based on several projects funded by the U.5Le. DESIGN OF WALL PIERS The design of the wall piers is then carried out for flexure using axial loads resulting from full plastification of all coupling beams. BCS8922777 and BCS9319838 with Dr. shown above to be equal to 1. National Science Foundation (Grants No.35 x 2Mr/Lφs = 1. Based on Mattock and Gaafar’s model (Equation A5). u setting twall = 500 mm. the necessary embedment length is 1025 mm.56(420) = 1745 kN. findings. limited by the width of the confining steel (see Figure 8b). it has been shown that encased steel coupling beams do not require the significant stiffener details (see Figure A2) in order to ensure adequate ductility and stability. V is 2. NOTATION a A As Asc Avd = = = = = shear span of coupling beam. the required shear capacity of the embedment becomes: Vu = 1. typically taken as L/2 cross sectional area of wall pier area of transverse reinforcement in concrete encasement area of concentrated reinforcement in boundary element of wall pier area of reinforcement in each group of diagonal bars .56 0.5 + 1. The require embedment length is u found to be 970 mm (Equation A5) or 1280 mm (Equation A3).167 f c 'bw d + Lφ s L s (A7) For the “flexure critical” beam designed above. The design shear force for the embedment. Figure 6 gives some indication as to the resulting boundary element reinforcement required for these prototype structures. In this case.026 kN. and e = 620 + 0.35 x 727 / 0. When using the model proposed by Marcakis and Mitchell (Equation A3). The preceding calculations demonstrate the effect of including the capacity of the concrete encasement. Although the embedments are longer.S. Any opinions. ACKNOWLEDGEMENTS The authors would like to gratefully acknowledge the input and support of members of American Concrete Institute (ACI) Committee 335 – Composite and Hybrid Structures and Professor Denis Mitchell of McGill University. the value of b’ is approximately 400 mm. and conclusions or recommendations expressed in this paper are those of the authors and do not necessarily reflect the views of the sponsors.
60 strength reduction factor for steel = 0. also the bearing width of the embedment web width of concrete beam or concrete encasement depth of concrete cover deflection amplification factor defined by NEHRP (BSSC 1997) resultant bearing (compressive) force in embedment distance from the extreme compression fiber to centroid of longitudinal tension reinforcement in concrete section eccentricity from midspan of beam to center of embedment Young’s modulus of coupling beam or wall pier specified concrete compressive strength yield strength of reinforcing steel yield strength of steel beam shear modulus overall depth of coupling beam moment of inertia of coupling beam or wall pier stiffness factor accounting for shear deformations in flexural members effective length of coupling beam length of coupling beam measured from face of wall piers embedment length of the steel coupling beams inside the wall piers distance between centroidal axes of the wall piers moment capacity of steel beam spacing of transverse reinforcement flange thickness of steel beam web thickness of steel beam thickness of wall pier shear stress factored shear stress factored shear force resisted by concrete nominal shear capacity of coupling beam plastic shear capacity of coupling beam factored shear capacity of coupling beam shear force resisted by reinforced concrete encasement factored shear force on section length of compression block at back of embedment length of compression block at front of embedment angle of inclination between diagonal reinforcement and longitudinal reinforcement coefficient defining relative contribution of concrete strength to wall strength ratio of the average concrete compressive strength to the maximum stress shear angle shape factor plastic interstory drift angle elastic interstory drift angle computed under code level lateral loads ratio of area of distributed reinforcement parallel to plane of shear resistance to the concrete area perpendicular to that plane material resistance factor for concrete = 0.Aw b' bf bw c Cd Cf d = = = = = = = = e = E = fc’ = fy = Fy = G = h = I = k = L = Lclear = Le = Lwall = Mr = s = tf = tw = twall = v = vu = Vc = Vn = Vp = Vr = VRC = Vu = xb = xf = α = αc = β or β1= γp = λ = θp = θe = ρn = φc φs = = area of web of steel beam effective width of the concrete compression block coupling beam flange width.90 .
. 122 pp. and Bertero. and Ghosh. K. 1984. Hysteretic behavior of reinforced concrete flexural members with web reinforcement. Seismic response of R/C framewall structures. Cincinnati. J. ASCE Journal of the Structural Division. Seismic behavior of slender coupled wall systems. Proceedings of the Seventh World Conference of Earthquake Engineering. Washington. Shahrooz B. A. American Institute of Steel Construction (AISC). 171179. Report No. 108 (ST 8). CSA A23. et al.. Bertero.REFERENCES CITED American Concrete Institute (ACI) Committee 318. Seismic behavior of R/C structural walls. Dynamic response of coupled wall systems. Farmington Hills. 89112. 1999. 258271. Chicago. ACI Journal. submitted by Portland Cement Association. Case study of aseismic design of a 16story coupled wall using inelastic dynamic analysis. A. Istanbul. M. 1998. V.. Manual of Steel Construction. 1981. V.. ASCE Journal of the Structural Division. Report No. McGill University. D. New York. Gong B. UCB/EERC 749. G. 1987. 9601. MI. UCCII 98/01. Aktan. ACI Special Publication 174 – Hybrid and Composite Structures.D.. V.. Earthquake Resistant Structural Walls . V. Load and Resistance Factor Design. et al. 229 pp. Berkeley. Turkey. Canadian Standards Association (CSA). A.. Harries. B. UCB/EERC81/07.... Bertero. Farmington Hills. 1974. J. Fintel. McGill University Department of Civil Engineering and Applied Mechanics Structural Engineering Series Report No. and Bertero.. 18031821. Composite Construction in Steel and Concrete III. Building Seismic Safety Council (BSSC) 1998. V. V. 1995. 1995. University of California. V. Evaluation of seismic response of RC building loaded to failure.. 113 (ST 5). MI. 1982. Gong B.. V. 1092. ASCE. A. ASCE Journal of the Structural Division. NEHRP Recommended Provisions for Seismic Regulations for New Buildings and Other Structures (FEMA 302) and Commentary (FEMA 303). Aktan. IL. ASCE Journal of the Structural Division. Berkeley. D. Report to NSF. Farmington Hills. Design and analysis of four prototype eighteen storey coupled wall structures.. Seismic behavior and design of composite coupled wall systems. 1980. Seismic Provisions for Structural Steel Buildings. Cincinnati Infrastructure Institute. AristizabalOchoa. E.Test of Coupling Beams. 113 (ST 10). Building Code Requirements for Reinforced Concrete and Commentary (ACI 31899/ACI 318R99). K. 1997. 1987. 1980. Seismic behavior and design of composite coupling beams. The seismic resistant design of R/C coupled structural walls. A.. Report No.. J. American Concrete Institute (ACI) Committee 318. and Gillum A. V. University of California. Aktan. Building Code Requirements for Reinforced Concrete and Commentary (ACI 31895/ACI 318R95). 1994. Seismic Design and Retrofit of Coupled Walls Using Structural Steel.. Shahrooz B. 1997 edition. Shahrooz B. 1982. S. Earthquake Engineering Research Center. IL. OH. Cyclic response of composite coupling beams. Earthquake Engineering Research Center. Barney. Thesis. 1998. and Bertero.1108. 22212234. J. E.. Rexdale. 79 (3). and Gillum A. M. Canada.. MI. Chicago. 110 (ST 8). AristizabalOchoa. V. 18461857. 353 pp. Ph. 1995.394 Design of Concrete Structures. 1978. Gong B. 1996. K. IL. DC. Harries. 1997. E. M.. American Institute of Steel Construction (AISC). Skokie. M. Research and Development.
W. 1982. 130140.. 102 (ST 1). 808818. and Reinhorn. R. J... W. 1997. Paulay. Paparoni. 1977. University of Illinois. H.. R. G. 671681. 1995. 1974. R. Seismic design of coupling beams . K. 77 (3). IDARC2D Version 3. International Council of Building Officials (ICBO). 448459. Proceedings of the International Conference on Tall Concrete and Masonry Buildings. Mahin. 79 (2). 1972. Ductility and deformability of coupling beams in reinforced concrete coupled walls. Kabeyasawa T. 783823. Report No. 843862.New Zealand design trends. July 1115. H. 88116. Milev J.. Mitchell. New Zealand Standards Association (NZS). and Gaafar G. Paulay. University of Tokyo. Canadian Journal of Civil Engineering. 1992.. 1980. A. Shiohara H. Earthquake Spectra. T. Earthquake Engineering Research Center.. Kunnath. 2 (4). and Sozen. 1998. and Mitchell D. 1994. Effect of beam strength and stiffness on dynamic behavior of reinforced concrete walls. and Bertero. K. T. 1975. NZS 3101:1995 Concrete Structures Standard. A. M.. S. 93108. CA.. . Structural Engineering Design Provisions.Shear in Reinforced Concrete. Lybas. 113 (1). 24(3). A. Saatcioglu. R. Precast/Prestressed Concrete Institute (PCI). Mitchell. Marcakis K. Seismic design and analysis of prototype coupled wall structures.. University of California. and Corley. A. and Cook. I. 141157. 25 (5). Park. and Aoyama H. A.. fourth edition. Mattock A. T. Analysis of the fullscale sevenstory reinforced concrete test structure. UrbanaChampaign. Journal of the Faculty of Engineering. PCI Design Handbook. V. Vancouver. Berkeley. R. Tokyo. Paulay. M. T. Proceeding of a Workshop on EarthquakeResistant Reinforced Concrete Building Construction.. 1981. 444. V. Proceedings of the Eighth Canadian Conference on Earthquake Engineering. A. UCB/EERC 7515. A. Strength of embedded steel sections as brackets. Inc. W. State University of New York at Buffalo.A case for mixed construction. D. 475481. Coupling beams of reinforced concrete shear walls. ASCE Journal of the Structural Division. S. 3. 78 (2). Model studies of coupling beams. 97 (ST 3). John Wiley and Sons. T. 144152. 1977. 25 (4). 1971. Diagonally reinforced concrete beams for shear walls. Paulay.. M. Coupling beams of reinforced concrete shear walls. ACI Journal. M. G. 1986. K. R. 1980.. Paulay. Parametric study of earthquakeresistant coupled walls. Otani S.. An evaluation of some methods for predicting seismic behavior of reinforced concrete buildings. The design of ductile reinforced concrete structural walls for earthquake resistance. 1994. Uniform Building Code.. Paulay. T.... G.. 431478. D. D. SRS No. Whitier.. ACI Journal. 1976 Ductile behavior of coupled shear walls. ASCE Journal of the Structural Division. 1998. personal communication. and Kabeyasawa T. and Cook. Modeling of reinforced concrete shear walls in RC framewall structures.. Earthquake resisting shear walls . T. and Binney. G.. K. New York. Slab coupling of earthquakeresisting shear walls. XXXVII (2). Harries. Derecho. J. D. June 1999.Harries. Redwood.1: Inelastic Damage Analysis of Reinforced Concrete Building Structures. 1987.. Canadian Journal of Civil Engineering. Harries. and Taylor. 579598. 1975 Reinforced Concrete Structures. 1977. 14521462. Precast concrete connections with embedded steel members. Paulay... ACI Journal.. and Paulay. A. 1983.. 1996. and Santhakumar. Volume 2. Redwood. T. Berkeley. ASCE Journal of the Structural Division. Prestressed Concrete Institute Journal. ACI Special Publication SP 42 .. M. University of California. T.
C. 1993. Structural Engineers World Wide 1998. Tunc G. Shahrooz B... G.. University of California. Skokie... Earthquake Resistant Walls Coupled Wall Test. E. C. Proceedings of the Central American Conference on Earthquake Engineering. and Coull. K.. V. W. UCB/EERC87/19. Report No.. 711720. A. and Qin F. 1978. M. Proceedings of the Third Joint Technical Coordinating Committee Meeting. 2000. OH. Seismic response of composite coupled walls. and Bertero V. Seismic design and performance of composite coupled walls. Shahrooz B..Scriber. Kato M. N. K. Reversed load tests of reinforced concrete coupling beams. 1996. Deason J. 1991. Hong Kong. B. Composite Construction in Steel and Concrete II. Wiley Interscience.. Barney.. and Corley.. P. Fiorato. and Qin F.. and Corley. F. UCCII 00/01. J. Berkeley. . K. N.I. Vulcano A. Remmetter M. G. A. 119 (11). A. El Salvador. Shiu. and Wight. 32913309. StaffordSmith. Tall Building Structures. Report No. 1998. 1996. December 1214. Shahrooz B. 1987. 1992. On the coupling behavior and ductility of reinforced concrete coupling beams of shear walls. M. 1996. W.. ASCE... N. Seismic behavior of coupled wall systems. G. Seismic response of R. Cincinnati.G. Barney. 239249.. Cincinnati Infrastructure Institute. A. T. Steelconcrete coupling beams: a critical overview of design guidelines. Sugaya K. B.. ACI Structural Journal 93 (6). M. and Shahrooz B. G. 1981.. University of Michigan. M. Teshigawara M. Remmetter M... Illinois. W. B. A. Moretti. M. and Corley. Research and Development. Earthquake Engineering Research Center. 1984. and Bezas. 429441. 1978. K. Delaying shear strength decay in reinforced concrete flexural members under large load reversals. A. B... E. core wallsteel frame hybrid structures. Takayangi. paper T1863. Fiorato.. Tassios. 10511066. ASCE Journal of the Structural Division. Department of Civil Engineering.. 110 (ST 5). Shiu. Report UMEE 78 R2.. and Gong. T. ASCE Journal of the Structural Division. Report to NSF submitted by Portland Cement Association. Shiu. Analytical models for predicting the lateral response of RC shear walls. Seismic test on 12story Tshaped coupled shear wall.