Journal of Constructional Steel Research 216 (2024) 108556

Contents lists available at ScienceDirect

Journal of Constructional Steel Research

journal homepage: www.elsevier.com/locate/jcsr

Seismic performance study of a novel steel truss energy dissipation

coupling beam
Ye Liu a, Zhibei Ma a, Xinlei Lv a, b, *
a
Key Laboratory of Deep Underground Science and Engineering (Ministry of Education), College of Architecture and Environment, Sichuan University, Chengdu 610065,
China
b
Key Laboratory of Concrete and Prestressed Concrete Structures of the Ministry of Education, Southeast University, Nanjing 210096, China

A R T I C L E I N F O A B S T R A C T

Keywords: To reduce the inter-story drift ratio and floor acceleration of coupled shear walls, a novel replaceable steel truss
Coupling beams energy dissipation coupling beam was proposed. The coupling beam comprises two buckling-restrained dampers
Buckling-restrained damper with additional friction energy dissipation (BFDs) as diagonal web braces, two chords, and connectors, which is
Friction energy dissipation
able to transform the shear deformation of the coupling beam into the axial deformation of diagonal web braces.
Theoretical analysis
A series of quasistatic loading tests were carried out on coupled steel truss energy dissipation beams to discuss the
effects of the coupling beam height, damper type, section height of the large yield segment, bolt torque, and slot
hole length on the seismic performance of the coupling beam. The experimental results indicate that specimens
S2, S3, and S4 with the same yield segment height exhibited similar hysteresis performance. Although the
friction energy dissipation device was applied to specimen S4, the hysteresis performance did not improve due to
the dampers’ end rotation failure. Increasing the large yield segment height from 5 mm to 8 mm resulted in a
significant increase in the maximum shear force, energy dissipation, and fatigue life. Increasing the slot hole
length led to a degradation of the aforementioned properties. Furthermore, a theoretical analysis of the forces
acting on the end of the hinged dampers was conducted based on the end-rotation failure characteristic of the
dampers, and a verification formula for the fracture of the outer tube at the end was proposed.

1. Introduction diagonal reinforcement concrete coupling beams with appropriate

reinforcement had better seismic performance than traditional RC
In high-rise structures, shear walls find extensive use as systems for coupling beams. Although diagonal reinforcement concrete coupling
resisting lateral forces caused by seismic loads. In shear walls, coupling beams improve the ductility and energy dissipation performance to
beams play a role in transferring forces and dissipating seismic energy some extent, they also have drawbacks such as complex detailing,
[1,2]. During earthquakes, coupling beams experience significant cyclic inconvenient construction, and potential issues with diagonal rein­
shear deformation due to seismic forces. Therefore, the ductility of forcement buckling in cross-reinforced coupling beams.
coupling beams is a critical parameter that affects the failure mode of The concept of “resilient cities” has gained widespread recognition,
coupling beams. Enhancing the ductility and energy dissipation capacity leading to a shift from collapse-resistant design to repairable design for
of coupling beams while preventing brittle failure has become a research high-rise buildings [7]. Substantial costs may be incurred for the repair
focus. and replacement of coupling beams damaged during earthquakes. To
However, reinforced concrete (RC) coupling beams have poor enhance the seismic recovery capability of buildings, innovative con­
ductility [3]. To enhance the seismic performance of RC coupling beams, nectors that are easy to repair or replace need to be developed. Steel
researchers [4] have proposed diagonal reinforcement concrete structures have attracted widespread attention as a novel form of
coupling beams as alternatives to traditional RC coupling beams. Naish coupling beam due to their outstanding ductility and superior energy
et al. [5] and Lim et al. [6] conducted experimental research on diagonal dissipation performance. In recent years, scholars have proposed various
reinforcement concrete coupling beams, and the results showed that forms of replaceable energy dissipation coupling beams. Fortney et al.

* Corresponding author at: Key Laboratory of Deep Underground Science and Engineering (Ministry of Education), College of Architecture and Environment,
Sichuan University, Chengdu 610065, China.
E-mail address: [email protected] (Y. Liu).

https://doi.org/10.1016/j.jcsr.2024.108556
Received 30 December 2023; Received in revised form 13 February 2024; Accepted 15 February 2024
Available online 2 March 2024
0143-974X/© 2024 Elsevier Ltd. All rights reserved.
Y. Liu et al. Journal of Constructional Steel Research 216 (2024) 108556

[8] introduced the notion of a replaceable steel coupling beam, which effectively reducing the damage to the RC coupling beams and exhib­
comprises energy-dissipating beam segments made of I-shaped steel at iting self-centring behavior.
the mid-span and nonenergy-dissipating beam segments at the ends, A dual stiffness and stable energy dissipation capacity sequence is
with detachable connections between the energy-dissipating and achieved in the BFD through the use of a stopper device and the fabri­
nonenergy-dissipating segments to facilitate postearthquake repair and cation of a core composed of a small yield segment and a large yield
replacement of the energy-dissipating segments. Ji et al. [9] furthered segment that are linked in series [19]. Hence, in this study, a novel
the research on replaceable steel connections, suggesting a hybrid cross- replaceable steel truss energy dissipation coupling beam, consisting of
section of a belly plate with low-yield-strength steel for energy- two buckling-restrained dampers with additional friction energy dissi­
dissipating beam segments, proposing connection forms between the pation (BFDs) acting as diagonal web members, upper chord, lower
energy-dissipating and nonenergy-dissipating segments, as well as low- chord, and connectors, was proposed to transform the shear deformation
damage steel-reinforced concrete floor forms, and establishing a seismic of the coupling beam into the axial deformation of diagonal web braces.
design method for replaceable steel connections. The damper and the chords are connected through pins, which facilitates
In addition to utilizing the shear deformation of steel coupling beams the replacement of the damaged damper after an earthquake and greatly
for energy dissipation, various methods have been proposed to protect improves the post-earthquake recovery ability of the coupling beam.
coupling beams from severe earthquake damage, such as using different Additionally, only the damper components being damaged do not affect
types of energy dissipation dampers as “structural fuses” in coupling the main structural integrity of the coupling beam, thereby reducing the
beam structures [10]. Among the existing metallic dampers, buckling restoration costs. Quasistatic loading tests were conducted on six steel
restrained dampers (BRDs) have been extensively researched in the field truss energy dissipation coupling beams to investigate their
of seismic resistance and have demonstrated stable hysteretic behavior shear–displacement response and analyse their hysteretic behavior and
under cyclic loading, as reported by Kimura et al. [11], Mochizuki et al. failure modes. Additionally, a theoretical analysis of the forces at the
[12], and research teams in Japan [13–15]. Shi et al. [16] proposed a ends of the hinged dampers was carried out based on the failure mode of
steel coupling beam with friction dampers and conducted cyclic loading the dampers, and a a verification formula for the fracture of the outer
tests. The research results showed that the friction dampers exhibited tube at the ends of the hinged dampers was proposed.
stable and excellent energy dissipation capacity, with relatively stable
friction force. Christopoulos et al. [17] applied viscoelastic dampers in 2. Configuration
coupling beams and conducted full-scale tests under cyclic loads. The
results showed that coupling beams with viscoelastic dampers could To reduce the inter-story drift ratio and floor acceleration of coupled
provide distributed damping in high-rise buildings. Compared to other shear walls, a novel replaceable steel truss energy dissipation coupling
systems currently used in high-rise buildings, distributed damping sys­ beam was proposed for arrangement between two shear walls, as shown
tems can reliably reduce dynamic wind loads and seismic loads. Mao in Fig. 1. The coupling beam is used only on the exterior axes of a
et al. [18] conducted quasistatic tests on RC coupling beams with shape building. A steel truss energy dissipation coupling beam is composed of
memory alloy (SMA) dampers, and the findings indicated that the two diagonal web members, an upper chord, a lower chord, and con­
deformation of the RC coupling beams was focused on the SMA dampers, nectors, as shown in Fig. 1. The upper and lower chords adopt H-shaped

Fig. 1. The configuration of steel truss energy dissipation coupling beam.

2
Y. Liu et al. Journal of Constructional Steel Research 216 (2024) 108556

steel and set transverse stiffening ribs on their web plates to avoid a stopper device, and (3) a friction energy dissipation device, as shown
buckling of the chords under the concentrated load of the diagonal web. in Fig. 3. The buckling-restrained damper is constructed of a core, an
The chords are hinged to the wall by connectors at both ends. The di­ outer tube, and infill panels. The core of the buckling-restrained damper
agonal web members are buckling-restrained dampers with hinged is composed of two connecting segments, two elastic segments A and B,
joints at both ends. Two diagonal web members cross each other without and two yield segments (large and small yield segments), as illustrated in
contact (refer to the cross-sectional view A-A in Fig. 1) and are hinged to Fig. 3(a). The cross-sections A-A and B–B in Fig. 3(a) exhibit two yield
the upper and lower chords. The buckling-restrained damper undergoes segments with quasirectangular shapes, which consist of two straight
tension-compression deformation and dissipates energy under the edges and two arc edges. Specifically, the yield segment with a larger
earthquake. Only the inner core of the buckling-restrained damper en­ height (Hyl) is referred to as the large yield segment, while the one with a
ters the plastic state, making the energy dissipation and damage smaller height (Hys) is called the small yield segment. To prevent stress
concentrated on the damper, which facilitates replacement after the concentration at the junction of the elastic and yield segments, a
earthquake. Both ends of the damper in the coupling beam were hinged rounded transition is utilized. As shown in Fig. 3 (b), the outer tube has a
connections. To avoid end bending of the core of the damper, an end- square-shaped cross-section. Circular and slotted holes are positioned on
constrained sleeve was used to constrain the end rotation of the the outer tube, designed to facilitate both the connection of the outer
damper’s core. The hinge connection plate was used to hinge the chord tube and the placement of the friction energy dissipation device.
and diagonal web. The stopper device is used to restrict the axial deformation of the
Two types of buckling-restrained dampers, working as diagonal web small yield segment. When the stopper device is activated, the large
members, were used in this study: (1) conventional buckling-restrained yield segment can provide higher strength and stiffness and enable the
dampers (CBDs), as shown in Fig. 2, and (2) buckling-restrained BFD to continue working after the failure of the small yield segment. The
dampers with additional friction energy dissipation (BFDs) [19], as stopper device comprises an outer tube and bolts. A group of M6 bolts
shown in Fig. 3. The CBD consists of a core, an outer tube, and infill are inserted into the circular holes of the outer tube and elastic segment
plates, as shown in Fig. 2(a). The core is composed of connecting, A to assemble the outer tube and core. Another group of M6 bolts goes
yielding, and elastic segments (see Fig. 2(b)). The connecting segment is through the slot holes of the tube and the circular holes of the elastic
used to connect the CBD to the hinged joint, thereby allowing the CBD to segment B. The bolt is initially set at the middle of the slot hole, and the
join with the upper and lower chord members. An elastic segment is small yield segment can deform axially. When the bolt touches the wall
inserted between the two yield segments, and a locating hole is situated of the slot hole, the stopper device is triggered, which restricts the axial
on the elastic segment. Similarly, locating holes are placed at the center deformation of the small yield segment to a certain value defined by the
of the outer tube to connect the core using a positioning stud. The infill length of the slot hole Ls.
plate is implemented to reduce the lateral deformation of the yield To enhance the energy dissipation capacity of the BFD, a friction
segment and prevent it from buckling. In addition, the infill panel was energy dissipation device is added. The friction energy dissipation de­
rounded on both sides to ensure smooth contact with the rounded cor­ vice provides an additional energy dissipation source for the BFD. As
ners of the yield segment. shown in Fig. 3(c), a brass square friction plate with a set of circular
The BFD comprises three parts: (1) a buckling-restrained damper, (2) holes is mounted and has the same hole size as elastic segment B. The

Fig. 2. Configuration of the CBD (Units: mm).

3
Y. Liu et al. Journal of Constructional Steel Research 216 (2024) 108556

Fig. 3. Configuration of the BFD (Units: mm).

friction plate, outer tube, and elastic segment B are interconnected with
high-strength bolts. After applying torque to the bolt, the bolt moves
along the length direction of the slotted hole, driving the friction plate to
generate relative motion with the outer tube and activating the friction
energy dissipation mechanism. More details of the BFD can be referred
to Ref. [19].

3. Mechanics

3.1. Steel truss energy dissipation coupling beam mechanics

As the shear walls on both sides of the coupling beam rotate under
earthquake actions, the upper and lower chord members undergo rigid
body rotation with the wall. As the chord members rotate, the two
connected buckling-restrained dampers also undergo axial deformation,
Fig. 4. Mechanism of steel truss energy dissipation coupling beam.
with one damper in a state of compression and the other in tension, as
shown in Fig. 4. The coupling beam dissipates energy during earth­
structural mechanics. When the shear deformation of the coupling beam
quakes through the axial deformation of buckling-restrained dampers.
is Δb, the calculation of the horizontal reaction force and vertical reac­
The force analysis diagram of the coupled steel truss energy dissi­
tion force at the chord hinge joint is shown in Eqs. (1) and (2),
pation beam is shown in Fig. 5. The assumption is made that the
respectively.
structure can be simplified as a system of rods using the principles of

4
Y. Liu et al. Journal of Constructional Steel Research 216 (2024) 108556

2Fcy 2Fd 2EAy

Kb = = (Ldx sinθ + 2bcosθ)2 = (Ldx sinθ + 2bcosθ)2 (4)
Δb Δd B2e Ly B2e

where Ay is the cross-sectional area of the yield segment in the damper.

When the damper is BFD, take the cross-sectional area of the small yield
segment; Ly is the total length of the yield segment in the damper. When
the damper is BFD, the total length is the sum of the lengths of the large
and small yield segments.

3.2. BFD mechanics

As shown in Fig. 6, the BFD can be simplified as a configuration

consisting of a small plastic energy dissipation segment in parallel with a
frictional energy dissipation device, followed by a large plastic energy
dissipation segment in series [19]. Therefore, there are two failure
modes for the BFD: (1) Large yield segment failure. (2) The small yield
segment fails first, followed by the large plastic segment. In the former
case, the BFD will directly stop working when the large yield segment
Fig. 5. Force analysis diagram of the steel truss energy dissipation
fails first. In the latter case, the BFD degrades into a damper consisting of
coupling beam.
a friction energy dissipation device in series with the large yield segment
after the failure of the small yield segment. Then, the large yield segment
Fcx = Fd cosθ (1) also fails. Thus, the mechanisms of the BFD before and after the failure of
the small yield segment are discussed separately below.
Fd Ldx sinθ Fd
Fcy = = (Ldx sinθ + 2bcosθ) (2) Before the failure of the small yield segment, the mechanism of BFD
Be Be
consists of two stages: pre- and postactivation of the stopper device, as
where Fcx and Fcy are the horizontal reaction force and vertical reaction shown in Fig. 7. The initial state of the BFD is illustrated in Fig. 7(a),
force at the chord hinge joint, respectively. a is the horizontal distance where the bolt is centrally positioned within the slot. Fig. 7(b) shows the
from the hinge center of the end of the chord to the hinge center of the mechanism of BFD in the tensile state, where u represents the loading
damper; b is the distance from the hinge center of the damper to the displacement and u1 represents the displacement when the stopper de­
central axis of the coupling beam; Ldx is the horizontal projection length vice is activated. As u increases, both yield segments undergo axial
of the damper; Be is the hinge center distance at both ends of the chord; deformation. The bolt moves along the length direction of the slot,
Fd is the axial force of the damper when the shear deformation of the causing relative motion between the friction plate and the outer tube,
coupling beam is Δb; θ is the inclination angle of the damper. and the friction energy dissipation device is activated. The small yield
The axial deformation Δd of the damper and the shear deformation Δb segment enters the yielding state before the large yield segment. The
of the coupling beam can be calculated through the geometric rela­ axial force of the large yield segment is equal to the sum of the axial
tionship, as shown in Eq. (3). force of the small yield segment and the total frictional force between
the friction plate and the outer tube. When the bolt contacts the slot hole
Δd =
Ldx Δb
sinθ +
2bΔb
cosθ =
Δb
(Ldx sinθ + 2bcosθ) (3) wall (u = u1), the axial deformation of the small yield segment is
Be Be Be restricted, and the friction energy absorption device stops working. As u
Through the axial yield deformation and yield force of the damper, continues to increase, only the large yield segment can be elongated,
the shear elastic stiffness Kb of the steel truss energy-dissipating coupling providing additional strength and energy absorption capability for the
beam is calculated, as shown in Eq. (4). BFD. Additionally, as shown in Fig. 7(c), the mechanism of the BFD in
the compressive state is similar to that in the tensile state.
After the failure of the small yield segment, the mechanism of BFD
also consists of two stages: pre- and postactivation of the stopper device,

Fig. 6. The failure modes of BFD.

5
Y. Liu et al. Journal of Constructional Steel Research 216 (2024) 108556

Fig. 7. The mechanism of BFD before the failure of small yield segment [19].

Fig. 8. The mechanism of BFD afore the failure of small yield segment [19].

6
Y. Liu et al. Journal of Constructional Steel Research 216 (2024) 108556

as shown in Fig. 8. As the maximum friction force is less than the yield Fig. 9 shows the construction and dimensions of the end-constraint
force of the large yield segment, the large yield segments remain in the sleeve. The end-constraint sleeve was a Q345 steel cylindrical body
elastic state. The friction energy dissipation device is the sole operative with an outer diameter of 70 mm. It had an M20 bolt hole that was 45
component in the BFD prior to the activation of the stopper device. Once mm long at one end, which was used to connect the damper, while the
the stopper device is engaged, the large yield segment continues to other end was a hollow circular tube with an inner diameter of 43 mm.
deform, contributing supplementary strength and stiffness to the BFD. The ear plate, end plate, and end-constraint sleeve were welded together
As the deformation further increases, the large yield segment enters the to form the hinge joint of the damper.
plastic state, enhancing the energy dissipation capacity of the BFD. The chord was made of Q345 steel and had an H-shaped cross-
section. The flanges were characterized by a width of 150 mm and a
4. Test program thickness of 10 mm, while the web possessed a height of 50 mm and a
thickness of 10 mm. The length of the chord was 758 mm, as shown in
4.1. Specimen details and test matrix Fig. 10. To prevent localized buckling caused by the concentrated forces
of the damper, the web of the H-shaped chord was enhanced by the
Table 1 shows the test matrix and geometric dimensions. Each addition of a transverse stiffener, which had a thickness of 10 mm. To
specimen includes two identical dampers, which serve as the diagonal connect with the diagonal web using a hinge joint, one side of the
web member of the steel truss energy dissipation coupling beam. CBDs chord’s flange had slotted holes measuring 38 mm in length and 18 mm
with different lengths of elastic segments were used to study the effect of in width. Two hinge connection plates made of Q345 steel were installed
changing the height of the coupling beam (the centre distance of the at the slot on the flange. The construction and dimensions of the hinge
upper and lower chords Hb: 600 mm and 750 mm); four sets of BFDs connection plates are shown in Fig. 11. The hinge connection plate
were used to investigate the influence of the section height of the large employed can be either embedded within the concrete shear wall or
yield segment Hyl (5 mm and 8 mm), bolt torque Tb (0 N⋅m and 10 N⋅m), securely attached to it using connectors. To achieve effective connec­
and slot hole length Ls (10 mm and 12 mm) on the seismic performance tion, the chords and shear walls were connected by hinge connection
of the coupling beam. plates in this study. The hinge connection plate was made up of an end
Both plastic yield segments of the CBD were 100 mm long, and the plate with multiple circular holes and two ear plates that were welded
cross-sectional height of the plastic yield segment Hy was 5 mm. When together. The center of the circular holes aligned with the center of the
Hb was 600 mm, Lic and Lot were 410 mm and 310 mm, respectively. slot on the H-shaped steel flange. The hinge connection plate was con­
When the Hb was 750 mm, Lic and Lot were 525 mm and 425 mm, nected using high-strength friction bolts, which allowed it to move
respectively. The middle elastic segment was 20 mm long and had a 2 slightly along the length of the slot to adjust for machining errors during
mm diameter positioning hole on the elastic segment (see Fig. 2 (b)). the installation process.
Likewise, a positioning hole with a diameter of 5 mm was opened at the To ensure the safety of the connection between the coupling beam
center of the outer tube (see Fig. 2 (c)), and the core was connected to and the shear wall, this study proposed essential strength design re­
the outer tube by a positioning stud. quirements. The size of the ear plate meets the following requirements
The large and small yield segments of BFD were both 100 mm in for strength checking per GB 50017–2017 [22], as shown in Eqs. (5) to
length. The cross-sectional height of the small yield segment of the BFD (7). The tensile strength of the net cross-sectional area of the ear plate
was 5 mm. The Lic and Lot were 525 mm and 425 mm, respectively. The hole was calculated as follows:
center of the slot hole is aligned with the center of the circular hole on
Fdmax
elastic segment B. The width of the slot hole is 6.5 mm, and its length Ls σ1 = ≤f (5)
2t⋅(min(2t + 16, le1 − d0 /3) )
is a variable in the experiment with two designed sizes, 10 mm and 12
mm. The width of the cross-section for the infill panel is 20 mm. The The tensile (splitting) strength of the cross-sectional area of the ear
length, Lf, is designed to be 0.9 times the length of the yield segment, plate end was calculated as follows:
with space left to allow for free deformation in the length direction when
Fdmax
the yield segment is under compression. The infill panel thicknesses, Tf, σ2 = ( )≤f (6)
for the large and small yield segments were 6.75 mm and 8 mm, 2t le1 − 2d0
3
respectively. This was done to minimize the gap between the core and
the outer tube to 1 mm. The shear strength of the ear plate was calculated as follows:
The outer tubes of the CBD and BFD were made from Q235b seamless
Fdmax
square steel tubes, while their cores were made from Q235b round steel τ= √̅̅̅̅̅̅̅̅̅̅̅̅̅̅̅̅̅̅̅̅̅̅̅̅̅̅̅̅̅̅̅̅̅̅̅̅̅̅̅̅̅̅̅̅̅̅ ≤ fv (7)
bars. The yield segment was obtained through symmetrical weakening.
2 2
2t (le1 + d0 /2) − (d0 /2)
According to GB/T 228.1–2010 [20], the elastic modulus, yield stress,
and yield strain of Q235b were 202.8 GPa, 324.4 MPa, and 0.16%, where Fdmax is the maximum axial force of the damper, which is
respectively. The material of the friction plate is an H62 brass plate, and conservatively designed to be 100 kN.
its surface quality meets the requirements of GB/T 2040–2008 [21]. The The pin hole diameter, d0, is 16 mm. The length from the ear plate
bolts are 10.9 grade high-strength semithreaded bolts, and the manu­ hole wall to the top of the ear plate, le1, is 17 mm, and the length from
facturer provided their shear strength and yield stress were 310 MPa and the ear plate hole wall to the side of the ear plate, le2, is 18 mm. The ear
900 MPa, respectively. plate thickness, t, is 15 mm, and the ear plate’s tensile strength, f, is 305
MPa per GB/T 228.1–2010 [20].

Table 1 4.2. Test setup and loading protocol

Test matrix.
Specimen Damper type Hyl (mm) Ls (mm) Tb (N⸱mm) Hb (mm) The quasistatic test setup consisted of a hydraulic actuator and a self-
balancing reaction frame, as shown in Fig. 12. This study focused on
S1 CBD – – – 600
S2 CBD – – – 750 exploring the seismic performance of the coupling beam itself and
S3 BFD 5 10 0 750 determined the simplified equivalence of its boundary conditions during
S4 BFD 5 10 10 750 tests. As shown in Fig. 12, this is the loading test setup after boundary
S5 BFD 8 10 10 750 simplification. Application of the lateral cyclic load for each specimen
S6 BFD 8 12 10 750
was accomplished through a hydraulic actuator with a load capacity of

7
Y. Liu et al. Journal of Constructional Steel Research 216 (2024) 108556

Fig. 9. The construction and dimension of the end-constraint sleeve (Units: mm).

Fig. 10. The dimension of the chord (Units: mm).

Fig. 11. The dimension of the hinge connection plate (Units: mm).

shown in Fig. 13, which consisted of two stages. In the first stage, a
±1000 kN and a displacement capacity of ±250 mm. The hydraulic
variable amplitude loading was applied to the coupling beam. The
actuator was positioned horizontally at the height of the top beam’s
rotation angle of the coupling beam was monotonically increased from
centerline. The self-balancing reaction frame was hinged by a top beam,
0.5%, 1%, 1.5%, and 2% to 2.5% for reciprocating loading, with two
two crank-connecting rods, and a grade beam in pairs. The steel truss
loading cycles per level. During the second stage, a constant amplitude
energy dissipation coupling beam was installed between the top beam
loading was applied to the coupling beam, with a rotation angle of 3%.
and grade beam, and the steel pad beam was used to supplement the
This loading continued until both dampers in the coupling beam were
height of the coupling beam. The hydraulic actuator was linked to the
damaged. The value ni represents the number of cycles for constant
connecting rod of the upper beam at one extremity, while the opposite
amplitude loading. The loading speed was 0.2 mm/s. Through this
end was secured to the reaction wall. The load was collected by force
loading method, the strength reduction and stiffness degradation laws of
transducers attached to the hydraulic actuator. Displacement transducer
the structure can be effectively studied. This has significant implications
1 was used to measure the horizontal displacement generated by the
for gaining insights into the behavior of structures during earthquakes.
coupling beam during loading to draw the load–displacement hysteresis
Related references also adopted the loading method of this study
curve. Displacement transducer 2 was used to monitor the horizontal
[19,23].
displacement generated by the grade beam.
The test used a mixed loading protocol with displacement control, as

8
Y. Liu et al. Journal of Constructional Steel Research 216 (2024) 108556

Fig. 12. Test setup and instrumentation.

5. Experimental results and discussion loading cycle is illustrated in Fig. 16 for all the specimens, and the
calculation formula of ξeq,i is shown in Eq. (8). The equivalent residual
5.1. Hysteretic response deformation is shown in Fig. 17. Photographs of damaged specimen S4
are shown in Fig. 18.
The hysteresis responses of specimens S1-S6 are shown in Fig. 14,
1 ED,i
where the two-stage performance of the specimen (before and after the ξeq,i = (8)
4π ESD,i
failure of the 1st damper) is also displayed. The test results for Vmax
+
, Vmax
–

and Etot are shown in Fig. 14. Vmax and Vmax are the maximum positive
+ –
where ED,i and ESD,i are the total energy dissipation and the area of the
and negative shear forces of the coupling beam, respectively. Etot is the elastic strain energy per cycle, respectively.
cumulative energy dissipation of the coupling beam, which is equal to The hysteresis curves were relatively plump, with good symmetry, as
the sum of the areas surrounded by hysteresis loops. Fig. 15 shows the shown in Fig. 14, but there was a significant slippage phenomenon (i.e.,
representative hysteresis curves of specimen S1 before and after the increased displacement with almost constant force) and lower stiffness
failure of the 1st damper. The equivalent damping ratio ξeq is one of the values. Slippage mainly occurred as a result of gaps at hinges in the
key parameters for evaluating the hysteresis performance and energy coupling beams. Additionally, the end rotation of dampers can also lead
dissipation capability of specimens. The variation in ξeq,i with each

9
Y. Liu et al. Journal of Constructional Steel Research 216 (2024) 108556

Fig. 13. Loading protocols.

to slippage. The activation of the stopper device resulted in a noticeable end restraint sleeve and the outer tube of the damper (see Fig. 16(c)).
increase in the stiffness of the specimen, as depicted in Fig. 15 (a). This The deformation of the damper was primarily characterized by rota­
could be attributed to the limitation of deformation in the small yield tional deformation at the end in this situation, rather than axial defor­
segment, while the large yield segment offered supplementary stiffness mation, resulting in a hysteresis curve showing slippage behavior. The
and strength. Moreover, Fig. 15 (b) exhibits a decrease in the specimen’s single-loop energy dissipation at a rotation angle of 2.5% in specimen
shear force and energy absorption capabilities after the 1st damper S4 was 2.06 kJ, which was 76.24% higher than the corresponding value
failure until the 2nd damper’s destruction. of 1.16 kJ in specimen S3, indicating that the additional friction energy
A comparison of specimens S1 (Hb = 600 mm) and S2 (Hb = 750 dissipation device could improve the energy dissipation capacity of the
mm), with different heights of the coupling beam, reveals that as the damper. Fig. 16 (d) illustrates the friction marks of the friction plate in
height of the coupling beam increased, the positive and negative the BFD.
maximum shear forces (Vmax +
and Vmax
–
) of the coupling beam showed no With a maximum shear force of 85.48 kN, specimen S5 (Hyl = 8 mm)
significant change, but the cumulative energy dissipation Etot decreased demonstrated a significant improvement over specimen S4 (Hyl = 5
significantly, from 9.76 kJ to 6.64 kJ. Two dampers in specimen S1 mm), which registered a shear force of 52.08 kN. Moreover, a phe­
failed during the first and third cycles of constant amplitude loading nomenon of increased strength and stiffness of the coupling beam could
with a rotation angle of 3%. In contrast, two dampers in specimen S2 be clearly observed after the activation of the BFD stopper device in the
failed during the first cycle of variable amplitude loading with a rotation hysteresis curve of specimen S5, as shown in Fig. 14(e). The energy
angle of 2.5% and the first cycle of constant amplitude loading with a dissipation capacity of specimen S5 significantly improved compared to
rotation angle of 3%. This is mainly because as the height of the coupling that of specimen S4, with a 379.35% increase in cumulative energy
beam increased, the angle between the damper and the chord also dissipation. Additionally, specimen S5 had a substantially higher fatigue
increased when the horizontal projection length of the damper remained life, with 23 cycles of constant amplitude fatigue loading, in contrast to
constant. This results in larger axial deformations of the dampers in specimens S3 and S4, as well as specimens S1 and S2.
specimen S2 at the same rotation angle, causing earlier failure. This also The positive maximum shear forces of specimens S5 and S6 were
explains the lower energy dissipation capacity of S2 compared to S1. 82.53 kN and 63.31 kN, respectively. As the slot hole length Ls increased
Specimen S3 exhibited smaller increases in Vmax +
and Vmax
–
, with dif­ from 10 mm to 12 mm, the positive maximum shear force decreased by
ferences of 4.93% and 1.46% compared to S2, and showed a 25.3% 23.3%. The lower energy dissipation and fewer constant amplitude
increase in energy dissipation capacity. The first damper failure loading hysteresis cycles of specimen S6 compared to specimen S5 could
occurred at a hysteresis loop with a rotation angle of 2.5% in both be attributed to the longer slot hole length Ls of specimen S6. This results
specimens S2 and S3, and the second damper failure occurred during in a greater deformation of the small yield segment and limits the ability
constant amplitude loading with a rotation angle of 3%. However, the to fully utilize the performance of the large yield segment for specimen
second damper in specimen S3 failed at the 7th cycle of cyclic loading. S6, leading to earlier failure.
The torque Tb in Specimen S3 was 0, and the friction device did not As shown in Fig. 16, overall, due to the hardening of the yielding
affect energy consumption. Additionally, the equal height of yield segment and the initiation of friction devices, ξeq gradually increased
segment Hy in specimen S2 and the small yield segment Hys in specimen with increasing initial loading cycles. However, when the first damper in
S3 were the primary contributors to the negligible performance the coupling beam failed, the value of ξeq also decreased sharply. This
discrepancy between specimens S2 and S3. indicates that before the first damper failed, the coupling beam exhibi­
The maximum shear forces of specimens S3 (with bolt torque of 0 ted better damping performance; however, after the failure of the first
N⸱m) and S4 (with bolt torque of 10 N⸱m) were close, but the energy damper, the damping performance of the coupling beam declined
consumption of specimen S4 was 7.17 kJ, which was 13.8% lower than sharply. The residual displacements of specimens S5 and S6, ranging
that of specimen S3 (8.32 kJ). The first damper failure occurred at a from 25 mm to 35 mm, as shown in Fig. 17, were significantly greater
hysteretic loop with a rotation angle of 2.5% in both specimens S2 and than those of the other specimens. This is because specimens S5 and S6
S3. However, the second damper failure of specimen S4 also occurred at had larger yield segment heights Hyl, resulting in increased residual
a hysteretic loop with a rotation angle of 2.5%, earlier than specimen S3, deformation when the first damper failed and only a single damper
which failed at 3%. This phenomenon could be primarily attributed to remained activated.
the fact that when the rotation angle reached 2.5%, the end of the sec­
ond damper in specimen S4 experienced significant rotation under
compression (see Fig. 16(a)), with evident pressure marks between the

10
Y. Liu et al. Journal of Constructional Steel Research 216 (2024) 108556

Fig. 14. Hysteretic responses.

11
Y. Liu et al. Journal of Constructional Steel Research 216 (2024) 108556

Fig. 15. Representative hysteresis curves of specimen S1.

Fig. 16. Equivalent damping ratio of specimens.

Fig. 17. Equivalent residual deformation of specimens.
5.2. Failure modes
compression. Fractures mainly occurred at the peaks or valleys of the
The upper and lower chords, as well as the ear plates, in the steel yield segment due to the large strains at these locations.
truss energy dissipating connection beam exhibited no noticeable plastic However, two additional failure modes emerged in addition to the
deformation. Failure was concentrated in the dampers, and the failure yield segment fracture of the core: connecting segment fracture and
modes of the dampers can be seen in Fig. 19, with the two dampers in outer tube fracture. Table 2 presents a comprehensive summary of the
specimen S(i) labelled Si-1 and Si-2. After the tests, due to significant various failure modes observed in the dampers. The end rotation of the
deformation of the core in some specimens or buckling of the outer tube, damper during compression resulted in contact between the end-
the cores of dampers S1–1, S5–1, S5–2, and S6–1 could not be extracted constraining sleeve and the outer tube of the damper, causing a
from the outer tube. Therefore, the outer tubes of these specimens were noticeable pressure mark on the end of the outer tube. In more severe
cut, as shown in Figs. 19 (a), (i), (j), and (k). cases, due to the combined action of the end-constraining sleeve and the
The majority of the deformation in the dampers of coupling beams elastic segment of the core, fracture occurred at the end of the outer tube
primarily occurred in the yield segment of the core. Due to the of dempers S1–2, S3–2, S5–1, and S6–2, as shown in Figs. 19 (b), (f), (i),
restraining effect of the outer tube and filler on the yield segment, the and (l). Due to the rotation of the demper at the end, the end of the
multiwave deformation mode occurred in the yield segment under connection section in the core of dempers S6–1 and S6–2 fractured, as

12
Y. Liu et al. Journal of Constructional Steel Research 216 (2024) 108556

Fig. 18. Damaged photographs of specimen S4.

shown in Figs. 19 (k) and (l). (1) Contact mode A: When d > cLeo/Le1, the elastic segment at the end
The wear condition of the pin is presented in Fig. 20. The pins that of the core comes into two-point contact with the outer tube,
connect the chord and hinge connection plate showed only minor sur­ resulting in a large bending moment at the end of the core (see
face wear with no significant deformation, as shown in Fig. 20(a). Fig. 21 (a)).
However, plastic deformation occurred in the pins that connect the (2) Contact mode B: When d < c < cLeo/Le1, as the damper end ro­
chord and damper (see Fig. 20(b)), which was primarily attributed to the tates, two-point contact occurs between the outer tube of the
slippage and low stiffness of the hysteretic curve of the coupling beam. damper and the end-constrained sleeve of the articulated joint. In
Additionally, the presence of clearances at the connections due to this case, the end-constrained sleeve and outer tube bear the
multiple connections in the coupling beam may also contribute to the bending moment and limit the range of damper end rotation,
sliding. To minimize sliding, it is suggested to reduce assembly con­ enabling the normal tensile and compressive deformation of the
nections in the coupling beam and opt for welding connections. damper core (see Fig. 21 (b)).
(3) Contact mode C: When c < d < cLeo/Le1, the contact mode is in­
6. Theoretical analyses termediate between the previous two situations (see Fig. 21 (c)).

6.1. End rotation analysis of hinged dampers For all test specimens, c was 1 mm, d was 0.5 mm, and Leo was 30
mm. In specimen S1, with Le1 measuring 40 mm, according to dimen­
When a hinged damper is subjected to compression, the yield sional considerations, the contact mode was determined to be B, as
segment of the core undergoes multiwave deformation. Due to the shown in Fig. 21(b). However, due to significant plastic deformation
hinged end, there is a tendency for the damper end to rotate. After the occurred at contact point B1 on the damper’s outer tube, as shown in
damper end rotates, there are three possible contact modes that may Fig. 19(b), the end of the damper was allowed to rotate further. This
occur between the core, outer tube, and end-constrained sleeve, as rotation, in turn, drove the elastic segment of the core to rotate, estab­
detailed in Fig. 21. c represents the clearance on one side between the lishing contact with the outer tube and forming contact mode C. As the
core and outer tube, d represents the clearance on one side between the end of the damper continued to rotate, resulting in the fracture of the
outer tube and end-constrained sleeve, Le1 represents the constrained outer tube, accompanied by noticeable pressure marks on the outer tube,
length of the elastic segment of the core in the outer tube, and Leo rep­ as shown in Fig. 19(b).
resents the constrained length of the outer tube in the end-constrained For specimens S2 to S6, Le1 was longer compared to specimen S1,
sleeve. The three contact modes are described as follows: measuring 100 mm. According to dimensional considerations, the

13
Y. Liu et al. Journal of Constructional Steel Research 216 (2024) 108556

Fig. 19. Photographs of damper failure.

contact mode was identified as A, as shown in Fig. 21(a). However, in compression from the elastic segment, leading to contact between the
specimens S3, S5, and S6, there was significant plastic deformation in damper’s outer tube and the end-constrained sleeve, transitioning into
the elastic segment of the core, as illustrated in Figs. 19(f), (i), (k), and contact mode C. As the end of the damper continued to rotate, it sub­
(l). Furthermore, plastic deformation occurred in the outer tube due to sequently led to the fracture of the outer tube, as shown in Figs. 19(f),

14
Y. Liu et al. Journal of Constructional Steel Research 216 (2024) 108556

Table 2 (i), (k), and (l).

The failure mode of the specimen.
Specimen Damper Damper Yield Connecting Outer 6.2. Force analysis at the end of the hinged dampers
type label segment segment tube
fracture fracture fracture
As shown in Table 2, in addition to the failure of the plastic segment,
S1 CBD S1–1 √
the end of the outer tube of the hinged damper also experienced a
S1–2 √ √
S2 CBD S2–1 √
fracture, which was one of the failure modes of the damper. A force
S2–2 √ analysis was conducted on the end of the damper to determine the cause
S3 BFD
S3–1 √ of failure. Subsequently, design requirements for the outer tube were
S3–2 √ √ proposed to prevent end fracture. Contact states A and B may transition
S4–1
into a pattern similar to contact state C due to the plastic deformation of
√
S4 BFD
S4–2 √
S5–1 √ √ the end-constrained sleeve, outer tube and core. Unlike contact state C,
S5 BFD
S5–2 √ plastic deformation causes the corner angle of the damper end to further
S6 BFD
S6–1 √ √ increase, which can lead to contact at the damper end not necessarily a
S6–2 √ √
two-point contact but possibly a three-point or four-point contact.
However, to ensure a conservative design for the outer tube fracture, a
two-point contact analysis was used, as shown in Fig. 22, which
concentrated the contact force on one side of the outer tube.
The end-constrained sleeve and the elastic segment of the core were
considered as a single entity for the purpose of force analysis, as shown

Fig. 20. Pin wear.

Fig. 21. Three contact modes for end rotation [24].

15
Y. Liu et al. Journal of Constructional Steel Research 216 (2024) 108556

Fig. 22. Force analysis of damper end.

in Fig. 22(a). The entity is acted upon by an external force F at the hinge {
2c 2d c + d
}
ear plate. The end-constrained sleeve and the elastic segment of the core θd1 = max , , (11)
Le1 Leo Leo
are subjected to contact force Q. As the bending moment of the yield
segment of the core can be ignored, the end of the elastic segment of the ur = Le1 tanθd1 (12)
core is only subjected to the axial force Fd of the damper. The equation
for calculating the contact force Q can be derived from the force balance (
L3e2 ks Le2
)
equation of the entire entity, which is shown in Eq. (9). ub = Fsinθd1 cosθd1 + (13)
3EI e GAe
Fd utotal
Q= (9) FL2e2 sinθd1
Leo θd2 = + θd1 (14)
2EI e
where utotal is the moment arm between the center of the ear plate hole
and the center of the outer tube under the action of external force F. where Le1 is the length of the elastic segment inside the outer tube. Le2 is
As shown in Fig. 22(b), the rigid body rotation angle θd1 occurs in the the length of the extended segment outside the elastic segment. Le3 is the
elastic segment of the damper, which causes a rigid displacement ur of distance from the end of the extended segment to the hinge point of the
the extended section outside the elastic segment. Under the action of ear plate. ks is the shear shape factor, which is 10/9 for circular cross-
external force F, the outer extended section of the elastic segment un­ sections. Ie and Ae are the moments of inertia and area of the elastic
dergoes plastic deformation ub, resulting in a rotation angle θd2 at the segment, respectively.
end of the extended section. The utotal can be calculated by adding up ud1, The contact force Q on the outer casing can be calculated using Eqs.
ud2 caused by the rigid body rotation and plastic deformation of the (9) and (10), and the end fracture of the outer casing of the damper can
outer extended section, and the rigid body displacement at the center of be verified. Two types of fractures, tensile failure and shear failure, may
the ear plate hole ud3, as shown in Eqs. (10) to (14). occur in the square outer tube under contact force Q, as shown in Fig. 23.
Considering that the tensile strength of steel is higher than its shear
utotal = ud1 + ud2 + ud3 = Le1 tanθd1 + (ur + ub ) + Le3 tanθd2 (10) strength, the shear failure mode is adopted for calculation. Given that

Fig. 23. Force analysis of the end of the outer tube.

16
Y. Liu et al. Journal of Constructional Steel Research 216 (2024) 108556

the shear stress on the shear surface is uneven, the shear stress is
simplified and assumed to have a triangular distribution along the
length direction of the outer tube. The verification formula for the end
fracture of the outer tube is as follows.
Q
τmax = ≤ fs,o (15)
Leo to

where τmax is the maximum shear stress. fs,o is the shear strength of the
outer tube.
The end fracture analysis of the outer tube was conducted for spec­
imens S1-S6 using Eq. (15), and the results are presented in Fig. 24. The
maximum shear stress at the edge of the outer tube in all specimens was
calculated based on a shear strength not lower than that of Q235 steel
[22]. The design was therefore unsafe, resulting in the occurrence of
fractures in some of the specimens’ outer tubes. According to Eq. (15),
raising the outer tube thickness to 7.2 mm guarantees that the maximum
shear stress at the outer tube edges for all specimens stays below the
shear strength of the Q235 steel, thereby ensuring design safety and
preventing outer tube fracture.

7. Conclusions Fig. 24. Check results of end fracture of outer tube.

A postearthquake replaceable steel truss energy dissipation coupling CRediT authorship contribution statement
beam was proposed in this study. The coupling beam was mainly
composed of two buckling-restrained dampers with additional friction Ye Liu: Writing – original draft, Supervision, Investigation, Formal
energy dissipation (BFDs) as diagonal web braces and two chords. The analysis. Zhibei Ma: Writing – review & editing. Xinlei Lv: Writing –
hysteresis behavior and failure modes of the coupling beams were review & editing, Investigation, Formal analysis.
studied through quasistatic loading tests, and the end rotational char­
acteristics of the hinged dampers in the coupling beams were also
Declaration of competing interest
studied. The main conclusions of this study are as follows:
None.
(1) The proposed novel steel truss energy dissipation coupling beam
conveniently transformed the shear deformation of the coupling
Data availability
beam into the axial deformation of diagonal web braces. The
dampers were the only components that experienced plastic
Data will be made available on request.
deformation, which led to localized damage and energy dissipa­
tion, making them easier to replace after an earthquake. Addi­
Acknowledgements
tionally, adjusting the height of the coupling beam and the yield
segment of the dampers enables effortless tuning of the stiffness
The authors would like to extend their sincere gratitude to the Na­
and strength of the novel steel truss energy dissipation coupling
tional Natural Science Foundation of China (52208496) and Sichuan
beam.
Science and Technology Program (No 2021ZDZX0011).
(2) The hysteresis curves exhibited good symmetry and plumpness,
but significant slippage was observed, which can be caused by
References
gaps at hinges in the coupling beams and end rotation of
dampers. Specimens S2, S3 and S4 with the same yield segment [1] M.S. Zareian, M.R. Esfahani, A. Hosseini, Experimental evaluation of self-centering
height exhibited similar hysteresis performance. Although torque hybrid coupled wall subassemblies with friction dampers, Eng. Struct. 214 (2020)
was applied to specimen S4, the hysteresis performance did not 110644.
[2] H. Hu, J. Liu, G. Cheng, J. Li, Y.F. Chen, Seismic behavior of hybrid coupled wall
improve due to the end rotation failure of the dampers. system with replaceable endplate-steel coupling beam, J. Constr. Steel Res. 187
(3) As the large yield segment height Hyl increased from 5 mm to 8 (2021) 106997.
mm, the maximum shear force, energy dissipation and fatigue life [3] Y. Li, Y. Liu, Z. Chen, Seismic response assessment of a hybrid coupled wall
structure with novel self-centering steel truss coupling beams, Bull. Earthq. Eng. 18
all increased significantly. However, as the slot hole length Ls (2020) 2657–2680.
increased, the above properties all degraded due to a greater [4] K.A. Harries, P.J. Fortney, B.M. Shahrooz, P.J. Brienen, Practical design of
deformation of the small yield segment and a limitation to fully diagonally reinforced concrete coupling beams-critical review of ACI 318
requirements, ACI Struct. J. 102 (2005) 876.
utilize the performance of the large yield segment for specimen [5] D. Naish, A. Fry, R. Klemencic, J. Wallace, Reinforced concrete coupling beams-
S6 with a larger slot hole length. part I: testing, ACI Struct. J. 110 (2013) 1057.
(4) The majority of failures were observed in the dampers of the [6] E. Lim, S.-J. Hwang, T.-W. Wang, Y.-H. Chang, An investigation on the seismic
behavior of deep reinforced concrete coupling beams, ACI Struct. J. 113 (2016)
coupling beams, with most of the deformation occurring in the
217.
yield segment of the core. Furthermore, two additional failure [7] D.R. Godschalk, Urban hazard mitigation: creating resilient cities, Nat. Hazards
modes were identified, namely, connecting segment fracture and Rev. 4 (2003) 136–143.
[8] P.J. Fortney, B.M. Shahrooz, G.A. Rassati, Large-scale testing of a replaceable
outer tube fracture, in addition to the previously observed yield
“fuse” steel coupling beam, J. Struct. Eng. 133 (2007) 1801–1807.
segment fracture of the core. [9] X. Ji, Y. Wang, J. Zhang, T. Okazaki, Seismic behavior and fragility curves of
(5) Based on the end-rotation failure of the dampers in the test, a replaceable steel coupling beams with slabs, Eng. Struct. 150 (2017) 622–635.
theoretical analysis of the hinge damper end force was carried [10] J. Ou, X. Long, Q. Li, Seismic response analysis of structures with velocity-
dependent dampers, J. Constr. Steel Res. 63 (2007) 628–638.
out, and a verification formula for the fracture of the outer tube at [11] K. Kimura, Y. Takeda, K. Yoshioka, N. Furuya, Y. Takemoto, An experimental study
the ends of the hinged dampers was proposed. on braces encased in steel tube and mortar, in: Annual Meeting of the Architectural
Institute of Japan (in Japanese), 1976.

17
Y. Liu et al. Journal of Constructional Steel Research 216 (2024) 108556

[12] N. Mochizuki, Y. Murata, N. Andou, S. Takahashi, An experimental study on [19] Y. Lu, Y. Liu, Y. Wang, J. Liu, X. Huang, Development of a novel buckling-
buckling of unbonded braces under centrally applied loads, in: Annual Meeting of restrained damper with additional friction energy dissipation: component tests and
the Architectural Institute of Japan (in Japanese), 1988. structural verification, Eng. Struct. 274 (2023) 115188.
[13] A. Watanabe, Y. Hitomi, E. Saeki, A. Wada, M. Fujimoto, Properties of brace [20] GB/T 228.1–2010, Metallic materials-Tensile testing-Part 1: Method of test at room
encased in buckling-restraining concrete and steel tube, in: Proceedings of Ninth temperature, in: General Administration of Quality Supervision, Inspection and
World Conference on Earthquake Engineering, 1988, pp. 719–724. Quarantine of the People’s Republic of China, Beijing, 2010.
[14] M. Fujimoto, A. Wada, E. Saeki, T. Takeuchi, A. Watanabe, Development of [21] GB/T 2040-2008 Sheet of copper and copper alloy, General Administration of
unbonded brace, Quart. Column. 115 (1990) 91–96. Quality Supervision, Inspection and Quarantine of the People’s Republic of China,
[15] A. Watanabe, H. Nakamura, Study on the behavior of buildings using steel with low Beijing, 2008.
yield point, in: Proc, 10th World Conf on Earthquake Engineering: Balkema, 1992, [22] GB 50017–2017, Standard for design of steel structures, in: Ministry of Urban Rural
pp. 4465–4468. and Housing Development of the People’s Republic of China, Beijing, 2017.
[16] S. Xiao, W. Yandong, Q. Zhe, X. Ji, Cyclic loading test of steel coupling beam with [23] Z. Qu, X. Ji, X. Shi, Y. Wang, H. Liu, Cyclic loading test of steel coupling beams
friction damper, Eng. Mech. 33 (2016) 156–160. with mid-span friction dampers and RC slabs, Eng. Struct. 203 (2020) 109876.
[17] C. Christopoulos, M. Montgomery, Viscoelastic coupling dampers (VCDs) for [24] J. Zhao, B. Wu, J. Ou, Effect of brace end rotation on the global buckling behavior
enhanced wind and seismic performance of high-rise buildings, Earthq. Eng. Struct. of pin-connected buckling-restrained braces with end collars, Eng. Struct. 40
Dyn. 42 (2013) 2217–2233. (2012) 240–253.
[18] M. Chenxi, W. Dalei, W. Tao, L. Hui, Pseudo-static tests of reinforced concrete
coupling beams equipped with SAM dampers, Earthq. Eng. Struct. Dyn. 34 (2014)
140–147.

18