Engineering Structures 247 (2021) 113232

Contents lists available at ScienceDirect

Engineering Structures
journal homepage: www.elsevier.com/locate/engstruct

Shake-table testing of 2-story steel framed building with self-centering

modular panels and slit steel plate walls
Gongling Chu a, b, Wei Wang a, b, *, Yunfeng Zhang c, *
a
State Key Laboratory of Disaster Reduction in Civil Engineering, Tongji University, Shanghai 200092, China
b
Department of Structural Engineering, Tongji University, Shanghai 200092, China
c
Department of Civil & Environmental Engineering, University of Maryland, College Park, MD 20742, USA

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

Keywords: Recent numerical study and cyclic testing results have suggested that the self-centering modular panels (SCMP)
Shaking table tests offer a promising prefabricated structural panel technology for enhanced seismic performance of beam-through
Seismic behaviour steel frames (BTSFs). Several types of seismic fuse devices, including tension-only braces and slit steel plate shear
Self-centering structures
walls (SWs) for energy dissipation, have been studied for their viability in adoption by SCMP. To further
Beam-through steel frames
Damage control
investigate the dynamic response and resilience behavior of this seismic-force-resisting system, shaking table
Quick repairing tests have been conducted on a 2-story SCMP-BTSF building model with SWs as seismic fuse device. Two far-field
ground motion records with various intensity levels were used in the shake table tests. The presented experi­
mental testing study is the first shaking table test of SCMP-BTSF, which is a newly proposed pre-fabricated
structure system with self-centering features. Nonlinear finite element (FE) model was also established to
perform time history analysis of the test structure and assist with specimen design. The shake test results show
that the structure response under frequent ground motions meets the inter-story drift limit given in Chinese code
- GB 50011–2010 (Code for Seismic Design of Buildings, abbreviated as CSDB hereafter). When the hazard level
was greater than that of frequent ground motions, the system exhibited flag-shaped hysteresis curves of the story
shear as expected. It is observed that the SWs dissipated seismic energy through yielding and buckling of steel
slats between slits, while other structural members and PTFs remained elastic. No cracks were seen in the
concrete floor slabs after all shake table tests. After replacing SWs, the test structure showed nearly identical
dynamic properties and seismic responses to the original test structure.

1. Introduction
While structures designed to modern seismic codes are anticipated to
survive severe earthquakes without collapse, they intend to experience
extensive structural and non-structural damage under design level
earthquakes [1]. For conventional seismic force resisting systems,
ductile damage is expected to be the source of seismic energy dissipation
during strong earthquakes. For example, steel moment resisting frames
(MRFs) and steel concentrically braced frames (CBFs) are designed to
dissipate energy through yielding near the beam ends and yielding of
braces, respectively [2–4]. In general, these energy dissipation mecha­
nisms are related to damage mechanisms of main structural members,
which may cause large residual drifts of buildings. In damage-controlled
structures, the damage would be concentrated in some specified
replaceable elements which yield prior to the other structural members
[5]. After strong earthquakes, only damaged structural fuse members
need to be repaired, which eases the repair process and thus cuts the
repair cost if replaceable fuse members are used. Recently, replaceable
seismic fuse members have drawn interests and metal hysteretic fuse
dampers have been investigated, such as steel angles [6,7], U-shaped
steel plates [8], shear links [9] and others.
Another key measure in lowering the seismic loss is to minimize the
residual inter-story drifts of the structure, since the residual inter-story
drift is regarded as a critical factor in economic feasibility of repairing
a damaged structure after earthquakes [10]. Considering the indirect
loss-of-function cost due to building closure, repairing a building with
large residual inter-story drifts may be much more expensive than
reconstruction. In recent years, several high-performance seismic
lateral-force-resisting systems, which can return to the pre-earthquake
position and confine damage to replaceable seismic fuse members,

* Corresponding authors.
E-mail addresses: weiwang@tongji.edu.cn (W. Wang), zyf@umd.edu (Y. Zhang).

https://doi.org/10.1016/j.engstruct.2021.113232
Received 26 September 2020; Received in revised form 30 April 2021; Accepted 15 September 2021
Available online 22 September 2021
0141-0296/© 2021 Elsevier Ltd. All rights reserved.
G. Chu et al.
Fig. 1. Schematic diagram of SCMP (PTF-SW system).
Fig. 2. Schematic diagram of a beam-through steel frame.
X Direction
Y Direction
SCMP
Tension-only brace
Fig. 3. Floor plan of the test structure.
have been investigated [6,11–18]. One of such self-centering seismic
systems is the self-centering modular panel (SCMP) [19,20], which is
comprised of two primary components - a post-tensioned frame (PTF)
and energy dissipation devices. The PTF is composed of horizontal
boundary elements (HBEs), vertical boundary elements (VBEs) and post-
2
Engineering Structures 247 (2021) 113232
tension (PT) cables. Fig. 1 illustrates a PTF-SW system. The SCMP can be
shop fabricated and then installed into a frame through bolted connec­
tions at the construction site. The PTF-SW exhibits a typical flag-shaped
hysteresis loop [21], since the bilinear elastic behavior of PTF [22] and
the pinching hysteretic behavior of slit walls [23,24] are combined. In
this study, the SCMP is installed in a beam-through steel frame (BTSF)
system (as shown in Fig. 2), which is a rapid-construction frame mainly
used for low-rise fabricated building [25–27]. In the system, the BTSF
functions as gravity load carrying frame composed of strong through-
beams and small-section columns, while the PTF-SWs act as seismic
lateral-force-resisting system. While several self-centering structure
systems have been investigated through experimental shake-table
testing [28–35], the current experimental investigation reported here
is the first shaking table tests conducted on SCMP-BTSF.
To investigate the nonlinear seismic response and its damage pattern
of the concerned seismic-force-resisting system with SCMPs subjected to
real ground motion records, shaking table test of two-story BTSF
buildings with SCMPs (PTF-SWs) was conducted and specific test results
are discussed here. The test specimen was conducted to subject to two
earthquake ground motion records selected from FEMA P695 far-field
ground motion records [36] with various intensity levels. Details of
the specimen and test setup are first presented. To probably predict the
structural response in testing, a FE model is developed in OpenSees [37],
and numerical simulation results of modal properties and nonlinear
responses are presented. The experimental test results presented here
include fundamental modal properties, typical acceleration and inter-
story drift response time histories, internal force response and descrip­
tion of observed behavior and damage patterns. The experimental re­
sults show that the structure can re-center after strong earthquakes and
damage is only concentrated to the fuse member (i.e., steel plate wall -
SWs in this study) while other structural members remained elastic.
Furthermore, minimal PT stress loss measured using a PT force sensor
was observed over multiple tests up to very strong intensity ground
motions. The shaking table test results further confirm the promising
seismic performance of the proposed self-centering system.
2. Test specimen, setup and instrumentation
2.1. Test structure
The test structure is designed referencing to the current Chinese
seismic code - Code for Seismic Design of Buildings (CSDB, 2010) [38].
The soil condition is assumed to be Group 2 and Type II with charac­
teristic period of 0.4 s. The peak ground acceleration (PGA) for frequent
ground motion (63% in 50 years probability of exceedance), which is
used in elastic design in CSDB [38], is 0.2 g. The corresponding
maximum value of seismic influence coefficient is 0.448 (equivalent to
the design basis earthquake’s response spectral value at short periods,
SDS, of 0.448 g). The PTFs of the SCMP specimens in this shake-table
testing were reused from previous quasi-static cyclic tests and their
hysteretic behavior can be found in reference [21]. To work within the
operation limit of the shaking table and existing SCMPs’ lateral force
resistance, the mass of test structure is set as 33.0 tons (7.5 tons for the
1st-floor and 25.5 tons for the 2nd-floor). This floor mass distribution
intends to match the experimental measured lateral force capacity of
two SCMPs installed in the 1st-floor and 2nd-floor to avoid potential soft
story issue that might arise if uniform floor mass distribution was
adopted.
The main frame (gravity load carrying frame) of the test structure is a
two-story two-bay BTSF in the test load direction. Fig. 3 shows the test
structure’s horizontal layout and has been designed to fit the steel
mounting setup on the shaking table available in the lab. The total
height of the test structure is 6,880 mm and its elevation views in the X
and Y direction are shown in Fig. 4 and Fig. 5 respectively. The steel
through-beams and columns in the BTSF are I-sections: H300 × 150 ×
10 × 12 and H125 × 125 × 7 × 9, respectively. Both beams and columns
G. Chu et al. Engineering Structures 247 (2021) 113232

Fig. 4. Elevation view of the test structure in the X direction: (a) interior frame equipped with SCMPs; (b) bare exterior frames.

Fig. 5. Elevation view of the test structure in the Y direction: (a) braced perimeter bay; (b) unbraced interior bay.

are made of Grade Q345 steel (nominal yield strength is 345 MPa in
(b) unbraced interior bay
China code). High-strength bolt slip-critical connection are used in the
beam-to-column joints. To eliminate slab composite effect, the rein­
forced concrete floor slab does not enclose the beam-to-column con­
nections through pre-arranged gap space around the connections. A
typical beam-column connection is illustrated in Fig. 6. Because the
structure is only loaded in the X direction (i.e., seismic excitation di­
rection), tension-only concentrically braces are installed in the perim­
eter bays in the Y direction to control Y-direction drift and accidental
torsional response of the test structure. All tension-only braces are made
of Grade Q235 steel strap with a rectangular cross section of 80 × 7.
The SCMPs are installed in the 3600-mm-wide bay of the interior
Fig. 6. Close-up view of PTF’s beam-to-column connection.

3
G. Chu et al. Engineering Structures 247 (2021) 113232

Table 1
Dimensions and section sizes of post-tensioned steel frames.
Story HBE (mm) VBE (mm) PT cables Gap
area opening
Section Length Section Length
(mm2) force (kN)

1 H350 × 2,000 H340 × 3,000 1,120 151 ( − 9 9)

150 × 20 150 × 30
× 40 × 35
2 H300 × 2,000 H340 × 3,000 840 114
150 × 20 150 × 30 ( − 1 1 7)
× 35 × 35

frame in the X direction. HBEs and VBEs of the PTFs are made of Grade
Q345 steel. PT cables use 1 × 7–15.2–1860 type, whose cross-section
area is 140 mm2 and ultimate tensile strength is 1,860 MPa, as
defined in Steel Strand for Prestressed Concrete (SSPC) [39]. Yield
strength of the PT cables is stipulated as 90 percent of the ultimate
tensile strength. The initial prestress of the 1st-story PTF’s PT cables is
measured as 484.8 MPa (at lower HBE) and 516.9 MPa (at upper HBE),
while that of the 2nd-story PTF’s PT cables is 603.2 MPa (at lower HBE)
and 421.6 MPa (at upper HBE). More details about the dimension of
PTFs, including the section of PTF elements and the total cross-section
area of PT cables, are given in Table 1. The SWs are installed through
high-strength bolt slip-critical connections to the PTFs. Fig. 7 shows
pictures of connections between SW and PTF. The main slit steel plates
are made of Grade Q235 steel, while the other parts of SWs, such as edge
stiffeners and bottom end plates, are made of Grade Q345 steel. The Fig. 8. Dimensions and configuration of slit steel plate walls: (a) 1st-story SW;
main slit steel plates are 4 mm thick plate and the slits are 10 mm wide. (b) 2nd-story SW.
More details about dimension of SWs are shown in Fig. 8.
Payload mass are secured to the reinforced concrete floor slabs in
which the steel reinforcements are Grade Q345 steel and the concrete
are Grade C30 (the design value of axial compressive strength is 14.3
MPa in China code, Code for Design of Concrete Structures [40]). The
two floor slabs are designed to be 40 mm and 120 mm thick, respec­
tively, in consideration of the floor mass requirement and corresponding
payload acting on them. The steel reinforcement of the 1st-floor is weld
steel mesh with a diameter of 4 mm for each plain steel rebar and
spacing of 40 mm between rebars. The 2nd-floor uses steel ribbed rebars
with a diameter of 8 mm in both X and Y direction and the distance
between rebars is 150 mm. In final, the actual seismic floor mass
(including payload mass blocks) of the test structure is 7,550 kg and
25,580 kg for the 1st-floor and the 2nd-floor respectively.

2.2. Design performance

In these shaking table tests, the SCMP-BTSF structure was designed

to meet the seismic code requirement of CSDB (2010) [38]. The ex­ North South
pected performance of the test model is summarized as follows:

Fig. 9. Photo of the test structure on the shake table.

Fig. 7. Photo of bolted connection between SW and PTF: (a) upper connection; (b) lower connection.

4
G. Chu et al.
Fig. 10. Schematics of FE model (not to scale): (a) test structure; (b) SCMP modeling details.
(1) The test model would remain linear elastic and the inter-story
drift response would meet the peak inter-story drift limit in
Chinese seismic code (CSDB, 2010), when the test model is sub­
jected to frequent ground motions corresponding to 63% proba­
bility of exceedance in 50 years.
(2) The structural damage would be mostly concentrated to the SW
elements, while other structural components would remain
elastic, including structural beams, columns and PTFs.
(3) The test model is able to re-center itself after tests with design
level ground motions, meaning the residual inter-story drift
should be less than 0.2%.
2.3. Test setup
The two-story test structure was tested on a shake table at the Multi-
functional Shaking Tables Lab of Tongji University in Shanghai, China.
The shake table has an overall table dimension of 6 m × 4 m, maximum
payload capacity of 700 kN, operational frequency range 0.1–50 Hz,
PGA of 1.5 g and displacement range of +/-500 mm. The overall view of
the test structure on the shaking table is shown in Fig. 9.
5
Engineering Structures 247 (2021) 113232
2.4. Finite element analysis and response prediction
A FE model has been developed using a nonlinear structural analysis
software - OpenSees [37] for specimen design, nonlinear response pre­
diction and test planning. In this FE model, only the seismic-force-
resisting system including the SCMPs and associated steel frames are
modeled while other gravity load carrying frames are not explicitly
modeled. Fig. 10 shows schematics of the FE model for the test structure.
A hysteretic uniaxial material model, Steel01, available in OpenSees has
been used for most steel members including through beams, columns,
HBEs and VBEs. For Grade Q345 steel used for these components, its
material properties are set as follows: modulus of elasticity, E0, of 2.06
× 105 MPa, yield stress, Fy, of 345 MPa and a strain-hardening ratio, b, of
1%. Here b is the ratio of post-yield modulus to initial elastic modulus.
Steel02 material has been adopted for steel PT cables, since one of its
parameters, sigInit, can be used to model the initial PT stress. Other
material properties of Steel02 for steel PT cables are set as follows: E0 =
2.06 × 105 MPa, Fy = 1720 MPa, b = 2%. Nonlinear fiber element is
used to model through beams, columns, HBEs and VBEs, while truss
element is used for PT cables. The beam-to-column connections and
column bases in the gravity load carrying steel frame are modelled as
hinged connections, because their rotational stiffness is fairly small due
to their simple connection configuration. A lean-on column is also
G. Chu et al. Engineering Structures 247 (2021) 113232

Displacement transducer

Accelerometer

(a) (c)

(d)
(b)

Fig. 11. Arrangement of instrumentation: (a) accelerometers and displacement transducers in BTSF; (b) strain gauges in BTSF; (c) location of strain gauges for
calculating column internal forces; (d) strain gauges and displacement transducers in the 1st-story SCMP.

Table 2
Summary of earthquake ground motions information.
Name M Year Recording station Records name Vs_30 PGAmax PGVmax Duration
(m/s) (g) (cm/s) (s)

Northridge 6.7 1994 Canyon Country-WLC NORTHR/LOS270 309 0.48 45 20.0

Superstition Hills 6.5 1987 Poe Road (temp) SUPERST/B-POE360 208 0.30 36 22.3

the approach by Ricles et al [41] has also been created to simulate the
gap opening and closing mechanisms of rocking members. For the SWs,
horizontal zero-length springs were used to simulate its nonlinear macro
hysteretic behavior. The Pinching4 material in the OpenSees software
[42] is uasually used to simulate wall with pinching characteristic [43].
In this study, the nonlinear springs simulating SWs are specified with the
Pinching4 material, whose parameters are calibrated with test result
from both numerical and experimental studies involving cyclic loading
of SWs.
Considering the reasonable arrangement of damping ratio of a bare
steel frame without non-structural components, a 0.50% damping ratio
is assigned to the FE model in the form of Rayleigh damping. Based on
modal analysis results, the natural period of calibrated FE model is
0.343 s.

2.5. Instrumentation
Fig. 12. Response spectrum of the selected ground motions.

2.5.1. Structural frame

included in the FE model to consider P-Δ effects.
Two uniaxial accelerometers are installed near the beam-to-column
SCMP model elements are connected to BTSF through rigid links
joints of interior frame on each floor to measure the dynamic response
between HBE and corresponding locations in the through beam. Fig. 10
and derive the story shear in the X direction based on inertial force.
(b) shows a SCMP model, in which the PT connection model similar to
Displacement transducers are installed to measure the drift at the ends of

6
G. Chu et al. Engineering Structures 247 (2021) 113232

Table 3
Shaking table test cases.
Case Input excitation PGA (g) Test duration (s)

1–1 White Noise 0.05 62.40

1–2 Northridge 0.11 29.49
1–3 White Noise 0.05 62.52
1–4 Northridge 0.21 29.04
1–5 White Noise 0.05 61.65
1–6 Northridge 0.31 29.57
1–7 White Noise 0.05 61.52
1–8 Northridge 0.54 29.30
1–9 White Noise 0.05 61.89
1–10 Northridge 0.78 29.40
1–11 White Noise 0.05 62.30
1–11 s Northridge 1.16 29.12
1-11ss White Noise 0.05 61.28
At completion of 1-11ss case, both seismic fuse members – SWs were replaced
1–12 White Noise 0.05 61.79
1–13 Northridge 0.11 28.89
1–14 White Noise 0.05 62.33
1–14 s Northridge 0.11 29.05
2–1 Superstition Hills 0.10 54.17
2–2 White Noise 0.05 61.25
2–3 Superstition Hills 0.22 54.03
2–4 White Noise 0.05 61.85
2–5 Superstition Hills 0.41 52.93
2–6 White Noise 0.05 61.11
2–7 Superstition Hills 0.48 53.74
2–7 s Superstition Hills 0.59 54.14
2–8 White Noise 0.05 61.92

Table 4
Summary of structure response from shake table tests.
Test PGA Peak acceleration Peak inter-story Residual inter-
Case (g) (amplification drift (%) story drift (%)
factor, βi)

1st- 2nd- 1st- 2nd- 1st- 2nd-

floor floor floor floor floor floor

Ground motion: Northridge – Canyon Country-WLC Fig. 13. Modal properties in white noise cases: (a) natural period; (b) mode
1–2 0.11 1.96 2.67 0.11 0.16 0.006 0.007 shapes of the initial test structure.
1–4 0.21 1.81 2.52 0.23 0.31 0.001 0.019
1–6 0.31 1.55 2.24 0.36 0.42 0.010 0.005
1–8 0.54 2.00 1.78 0.59 0.68 0.012 0.011
1–10 0.78 1.86 1.51 1.05 1.17 0.006 0.012
1–11 s 1.16 1.79 1.10 1.50 1.61 0.002 0.013
1–13 0.11 1.91 2.56 0.09 0.17 0.001 0.004
1–14 s 0.11 1.89 2.59 0.09 0.16 0.001 0.001
Ground motion: Superstition Hills – Poe Road (temp)
2–1 0.10 2.59 3.02 0.12 0.19 0.004 0.002
2–3 0.22 2.23 2.35 0.26 0.35 0.005 0.008
2–5 0.41 2.01 2.00 0.45 0.59 0.005 0.013
2–7 0.48 2.19 1.93 0.59 0.72 0.003 0.006
2–7 s 0.59 3.58 2.02 1.20 1.40 0.010 0.017

the beams in both X and Y directions. Fig. 11 (a) shows the locations of
these accelerometers and displacement transducers.
Fig. 11 (b) shows the locations of strain gauges in the main frame.
Strain gauges have been used to monitor the strain and yielding occur­ Fig. 14. Acceleration response amplification factor vs. base excitation PGA.
rence of structural members. Also, they can be used to calculate the
shear force in the columns, which can be calculated using Eq. (1). Fig. 11
2.5.2. SCMP
(c) shows the column internal forces and strain gauge location.
Fig. 11 (d) shows the locations of displacement transducers and
M2 + M1 strain gauges in the 1st-story SCMP while 2nd-story SCMP has a similar
V= (1)
l arrangement. Eight displacement transducers were used for each PTF to
measure the gap opening values at the PT connections. The one installed
M1 =
EWc (ε1 − ε2 )
(2) at the center of SW was used to monitor the out-of-plane displacement of
2 SW. The others are used to measure the relative slip between the PTF
and through-beams, as well as the relative slip between SW and PTF.
EWc (ε3 − ε4 )
M2 = (3) The strain gauges on the VBEs were used to calculate the horizontal
2
shear force in the VBEs while the strain gauges on SW were used to
monitor the strain of steel slats between slits and corners of SW.

7
G. Chu et al. Engineering Structures 247 (2021) 113232

Fig. 15. Acceleration time history of test structure in Case 1–8: (a) 1st-floor;
(b) roof.

Anchor cable dynamometers were used to monitor the tension force

of PT cables, which can only be measured when the structure is static
due to the special design of this type dynamometer device. However, the
measurement done between tests has been very helpful to determine
whether pre-tension force loss has happened after each test.

3. Test program

3.1. Ground motions

Two far-field earthquake ground motion records from Northridge

1994 and Superstition Hills 1987 respectively, are selected from FEMA
P695 [36] and their details are listed in Table 2. The ground motions are
scaled to the target PGA of 0.2 g for elastic time history analysis. Fig. 12
shows the response spectra of scaled ground motions and the elastic
design spectrum with PGA of 0.2 g and a 5% damping ratio. The vertical
dashed black line denotes the initial natural period of the test structure,
0.33 s.

3.2. Loading protocol
To investigate the seismic performance and resilience characteristics
of BTSF with SCMP under varying seismic hazard levels, the loading
protocol consisted of the selected ground motions with different scaling
factors, starting with low intensity (0.1 g) and increasing towards
greater amplitudes. Additionally, white noise excitation with peak ac­
celeration value of 0.05 g was also conducted to measure the natural
period and damping ratio of the test structure before and after each case
(with a few exceptions). The change of properties would offer insight
into the damage extents suffered by the test structure during the pre­
vious test and effect of replacing the damaged SWs. Table 3 lists the
details of the adopted loading protocol.
4. Experimental results
4.1. General result
Table 4 summarizes the recorded structural response from each test
case, including acceleration (amplification factor defined as maximum
floor acceleration value normalized by excitation ground motion’s PGA
value) and displacement (peak inter-story drift and residual inter-story
drift). Hysteretic response and damage pattern will be discussed later.
The input excitation to the shaking table was only applied in the X
Fig. 16. Measured inter-story drift time history of test structure in two tests: (a)
Case 1–11 s; (b) Case 2–7 s.
direction. It indicates that the installed tension-only braces along the Y
direction are strong enough to limit the undesired displacement and
rotation response. It should be also noted that the structure was sub­
jected to series of tests and the SWs were only replaced once (see
Table 3) in the tests, while in real structures the SWs would be replaced
after a strong earthquake.
4.2. Modal properties
Modal properties include the test structure’s natural period, T1,
viscous damping ratio, ξ and mode shapes. The change in modal prop­
erties might reflect the damage suffered by the test structure in previous
tests. Natural period and mode shapes are derived using the corre­
sponding transfer functions that relate the floor acceleration response
and base excitation during the white noise tests, while the viscous
damping ratio is derived from the free vibration attenuation duration
after the white noise excitation inputting. Fig. 13 (a) shows the change
of natural periods over the test cases. The vertical dotted line in this
figure denotes the timing when the SWs were replaced in the test
structure. Since the change of mode shape is very small even though the

8
G. Chu et al.
Fig. 17. Internal force response of stories, SCMPs, PTFs and gravity load carrying frame: (a) Case 1–2; (b) Case 1–11 s.
structure underwent severe earthquake loading in the tests, Fig. 13 (b)
only shows the mode shapes of the initial test structure. The viscous
damping ratio of the initial test structure is measured as 0.45%, and thus
it is rational to adopt the damping ratio as 0.50% in the FE model for the
first vibration mode.
The natural period of test structure is measured as 0.33 s. And it has
barely changed after completing the tests of Case 1–2 to Case 1–6,
suggesting that the test structure have no detectable damage in these
cases (these are also corroborated by visual inspection of test structures
after these tests and subsequent story shear hysteresis curve analysis).
After Case 1–8 to Case 1–11 s, it could be found that the structure has
suffered some damage, reflected by the increase in natural period,
changed to 0.38 s (measured in Case 1-11ss). After Case 1-11ss, the test
structure has been repaired by replacing the SWs, followed by a white
noise excitation test, Case 1–12. The results show that the natural period
has almost recovered. This is consistent with the observation that
damage in the test structure mainly occurred in the SWs and replacing
with new SWs makes the test structure return to its initial condition.
9
Engineering Structures 247 (2021) 113232
Case 1–13 and Case 1–14 s are also conducted to investigate the seismic
performance of the structure after SW replacement with repeated
experiments.
In the test series of Case 2 (Superstition Hills earthquake), the natural
period has remained stable before Case 2–5 (with PGA of 0.41 g). The
test structure has been damaged again in Case 2–5 and Case 2–7,
determined from the change in the measured dynamic parameters. The
natural period, measured in Case 2–8, has changed to 0.38 s.
4.3. Acceleration response
The acceleration response in Table 4 is presented by the amplifica­
tion factor, βi (where i is the floor number), which is defined as the ratio
of the maximum floor acceleration response value to base excitation
PGA. The amplification factors vary from 1.10 to 2.67 under the
Northridge cases, and the maximum amplification factor of 2.67
occurred at the second story in Case 1–2 with PGA of 0.11 g. Superstition
Hills earthquake has a far greater impact on the acceleration response
G. Chu et al. Engineering Structures 247 (2021) 113232

Fig. 17. (continued).

than Northridge, observed by the maximum acceleration response. The

Table 5
amplification factors range from 1.93 to 3.58, and the maximum one,
Summary of initial stiffness (kN/mm).
3.58, is in Case 2–7 s with PGA of 0.59 g. A scatter diagram of accel­
Story Story total PTF Frame SW eration amplification factor versus base excitation PGA is plotted in
1 24.35 10.64 2.15 11.56 Fig. 14, which includes results of the Northridge and Superstition Hills
2 15.07 7.96 1.89 5.22 ground motion cases. Fig. 15 shows the absolute acceleration time his­
tory of the test structure in Case 1–8, in which the excitation record’s

Fig. 18. Photos of PT connection states during test: (a) closed gap; (b) open upper gap; (c) open lower gap.

10
G. Chu et al.
Fig. 19. Gap opening width vs. inter-story drift of 1st-story under Case 1–11 s: (a) bottom in south side; (b) upper in south side.
Fig. 20. PT force vs. inter-story drift of 1st-story under Case 1–11 s: (a) bottom PT cables; (b) upper PT cables.
PGA value was near the moderate hazard level (0.56 g).
4.4. Displacement response
The peak inter-story drift and residual story drift in each case are
listed in Table 4. The peak inter-story drifts are 0.31% and 0.35% in Case
1–4 and Case 2–3 (with PGA of 0.2 g), respectively, which meet the drift
limit of 0.4% specified by CSDB (2010) [38].
In Case 1–11 s (Northridge, 1.16 g) and Case 2–7 s (Superstition,
0.59 g), the peak inter-story drifts are 1.61% and 1.40%, respectively.
However, the residual inter-story drifts are only 0.013% and 0.017%,
which are much less than 0.2% for structure re-centering limit, based on
the out-of-plumb tolerances for construction [44]. The test results show
that the test structure has such a strong re-centering force that can make
the structure return to the pre-earthquake position with the resistance
caused by plastic deformation of SWs after a severe earthquake.
By comparing the peak inter-story drifts in Case 1–2, Case 1–13 and
Case 1–14 s, it is found that the peak inter-story drifts, as well as the
acceleration amplification factors, are very close to each other. Since
SWs were replaced before Case 1–13, this verifies that the structural
condition of the retrofitted test structure with only damaged SWs being
replaced has been restored to pre-earthquake state.
Fig. 16 shows typical inter-story drift ratio (θ) time history curve in
Case 1–11 s and Case 2–7 s. As shown in Table 4, the distribution of
inter-story drift is quite uniform in all cases (including the cases with
relatively high PGA). It can be found that there are fairly small residual
11
Engineering Structures 247 (2021) 113232
inter-story drifts under any ground motions with different PGA and the
test structure can re-center itself.
4.5. Internal force response
The hysteretic behaviors of the test structure in the loading direction
are investigated, since the structure was only loaded in the X direction.
Two test cases, Case 1–2 and Case 1–11 s, are selected to demonstrate
the hysteretic behaviors of the test structure when subjected to North­
ridge ground motion under two different intensity levels. Fig. 17 shows
the hysteresis loops of different stories, SCMPs, PTFs and structural
frame in these cases. The story shear force of Story i is obtained as the
summation of the total inertia force of upper floors above Story i level.
The inertia force of each floor is calculated by multiplying the acceler­
ation with the corresponding floor mass. The story shear force in Fig. 17
has been normalized with the total weight, W, of the test structure (W =
324.7 kN). From the hysteretic response presented, the following ob­
servations are made:
(1) As shown in Fig. 17 (a), when the structure is subjected to
Northridge with PGA of 0.11 g (Case 1–2), the hysteretic curves
show linear behavior, which indicates that the structure
remained essentially elastic and did not dissipate much energy
through steel yielding. Furthermore, the contribution to lateral
resistance from the structural gravity load carrying beam-through
frame is relatively small in comparison with the corresponding
G. Chu et al. Engineering Structures 247 (2021) 113232

Fig. 21. Photos of test structure after test Case 1-11ss: (a) HBE-to-beam joint; (b) beam-to-column joint near SCMP; (c) middle beam-to-column joint in exterior
frame; (d) corner beam-to-column joint.

Fig. 18 shows the different states of PT connection during testing,

Table 6 being recorded by a camera attached to the specimen. Addi­
Measured pre-tensioning stress in PT cable groups (unit: MPa). tionally, gap opening was measured by corresponding displace­
Case (finishing) 1st-story PTF 2nd-story PTF ment transducers installed between the HBE and VBE at the
monitored PT connections. Fig. 19 shows the gap versus inter-
Lower Upper Lower Upper
story drift curves of the 1st-story PTF under Case 1–11 s.
Initial state 484.8 516.9 603.2 421.6
Clearly, the gap opened and closed as expected during the shake
1–4 486.5 516.4 610.2 415.5
1–6 486.0 516.7 611.1 415.0 table test. Fig. 20 shows the PT force versus inter-story drift curve
1–8 486.2 518.7 609.5 413.8 of the 1st-floor under Case 1–11 s. Here it should be noted that the
1–11 s 489.5 518.3 611.4 411.8 PT force was calculated as the initial PT force plus the increased
2–3 490.3 519.1 611.4 413.0 force due to gap opening and corresponding PT cable elongation.
2–5 489.9 519.1 608.5 414.4
2–7 s 489.5 519.0 608.8 413.0
Once gap opened, the inter-story drift response would increase,
which makes the SWs dissipate more energy. It is worth to note that the
story shear. Table 5 lists the initial elastic stiffness of each frame presents an elastic hysteresis behavior with no energy dissipation,
structural components derived from measured data. The results even when the peak inter-story drifts are 1.50% (the 1st-story) and
show that the gravity load carrying frame’s initial stiffness of the 1.61% (the 2nd-story) under Northridge with PGA of 1.16 g. It could
two stores are only 8.8% and 12.5% of the story initial stiffness, indicate that the members and connections of BTSF remain elastic with
respectively, and that PTFs and SWs provide the most seismic no visible damage, which is appealing for damage control purpose. One
lateral stiffness of the structure. inference is that the energy dissipation is predominantly provided by the
(2) When the test structure is subjected to Northridge earthquake SWs for the test structure, since BTSF and PTFs remained elastic. After
with PGA of 1.16 g (Case 1–11 s), hysteretic behavior of the the relatively large drifts in hysteretic cycles, the structure rocked
structure, as shown in Fig. 17 (b), is clearly different from that of around a balanced position in small cycles and eventually the structure
Case 2–1. The story hysteretic curves have flag-shaped behavior, returned to the pre-earthquake position. This suggests that the system
while the structural gravity load carrying frame remain elastic has effective self-centering capacity.
and its hysteresis still presents an elastic linear behavior the same
as Case 1–2. The hysteretic curves of PTF show an elastic bilinear
behavior. After gap opening in PT connection, the PTF exhibited a 4.6. Damage observation
reduced stiffness (post-gap-opening stiffness or recentering stiff­
ness) in its elastic bilinear hysteresis loop. The gap opening oc­ The BTSF, PTFs and floor slabs, including the components and con­
curs at inter-story drift of about 0.3% to 0.5%, which is designed nections, remain elastic (no damage) when the test structure was sub­
to be around 0.4% (the inter-story drift limit of elastic design in jected to base excitation ground motions with different intensity levels.
CSDB [38]) under frequent ground motions. The gaps would open No visible damage such as yielding and cracking were observed in the
when the seismic hazard level is greater than frequent ground BTSF, PTFs and floor slabs except for SWs, even when the specimen has
motions, and after that stiffness of the system decreases to post- been subjected to multiple tests and the peak inter-story drifts of 1.50%
gap-opening stiffness. PT connections gap would be closed after and 1.61% happened in Case 1–11 s for the 1st-story and the 2nd-story,
unloading because of the large re-centering force from PT cables. respectively. Sample photos taken from the test structure are shown in
Fig. 21.

12
G. Chu et al.
Fig. 22. Photo of SW: (a) 1st-story in Case 1–8; (b) 2nd-story in Case 1–8; (c) 1st-story in Case 1–11 s (removed); (d) 2nd-story in Case 1–11 s (removed).
The PT cables force was measured immediately after each test case.
The data in Table 6 shows that little prestress has been lost in the tests.
4.7. SW damage evaluation
The SW plays the role of seismic fuse and energy dissipation device in
the test structure. In the tests, the SWs remain elastic only in Case 1–2
(Northridge with PGA of 0.11 g) and Case 2–1 (Superstition Hills with
PGA of 0.10 g), as shown from measured strain data from local strain
gauges around steel slits. For illustration, Northridge earthquake exci­
tation cases (Case 1 series) are discussed here. In Case 1–2, Case 1–4 and
Case 1–6, no visible damage was found such as flaking of latex paint due
to plastic deformation of SWs. In general, the yielding of SWs begins at
the corner of steel slats between slits, and expands to the surrounding
corner region of SWs with increasing PGA levels. As illustrated in
13
Engineering Structures 247 (2021) 113232
Fig. 22, the photos of SWs taken after Case 1–8 and Case 1–11 s
(Northridge with PGA of 0.54 g and 1.16 g) are typical in showing the
different damage extents suffered by SWs.
The SWs in Case 1–8 were damaged slightly, as show in Fig. 23 (a)
and (b). The corners of slats between slits yielded (initiating at the end of
slits) and the severest damage happened in the outermost slats (judged
through the flaking off of white latex paint on the wall). Additionally,
the slats are slightly bulgy (out of plane deformation) upon close in­
spection. As described in Section 4.5, the 1st-story and the 2nd-story
sustained an inter-story drift of 1.50% and 1.61% respectively and the
SWs dissipated the seismic energy in Case 1–11 s as other components
remained elastic without energy dissipation. As shown in Fig. 23 (c) and
(d), the plastic zone of the SWs enlarged significantly, and the yielding of
the slat corners propagated along nearly 45 degrees line. In the center
area of SWs, yield path connected the upper and the bottom slats along
G. Chu et al. Engineering Structures 247 (2021) 113232

Fig. 22. (continued).

Fig. 23. inter-story drift time history response in FE analysis and test results in Case 1–11 s: (a) 1st-story; (b) 2nd-story.

an arc-like contour. Out-of-plane deformation of the slats also occurred,
and slight cracking occurred at the end of slits, as illustrated in Fig. 22
(c) and (d). It can be concluded that the SWs experienced large plastic
deformation and dissipated seismic energy.
During the shaking table tests, the damaged SWs were replaced once
by two crew members after Case 1–11 s. It took a total of 105 min to
remove the damaged SWs and install the new SWs (58 min for the 1st-
story SW and 47 min for the 2nd-story SW). It is noted that these
workers are not highly skilled in doing this as they were not trained
before. Replacing the SWs involves removing bolts with wrenches and
simple torque wrenches are needed to install the SWs, but otherwise no
special tools are required.
4.8. FE modeling validation
Simulation results for Case 1–11 s is presented here to showcase the
validity of the finite model. Fig. 23 shows the comparison of inter-story
drift time history responses from both FE analysis and experimental test
results. Fig. 24 shows the story shear vs. inter-story drift ratio curve and
the SW hysteresis curve. In general, a relatively good agreement is
observed between results of FE modeling and tests. And it can be
concluded that the FE models is validated to simulate the BTSF-SCMP-
SW structures with acceptable accuracy.

14
G. Chu et al.
Fig. 24. Hysteresis curves from FE analysis and test results in Case 1–11 s: (a) 1st-story shear; (b) 2nd-story shear; (c) 1st story SW shear; (d)2nd story SW shear.
5. Conclusion
This paper presents the experimental test results of shaking table
tests conducted on a two-story BTSF-SCMP model structure. The seismic
performance of the model structure subjected to strong ground motions
is investigated, including its self-centering and damage-control
behavior. In this system, the BTSF functions as a gravity load carrying
frame with negligibly-small lateral stiffness compared with the main
seismic-force-resisting system of SCMP comprised of PTF and SW. The
PTF provides mainly the re-centering capability through rocking actions
at PT connections, while the SW is adopted here as seismic fuse device to
dissipate seismic energy. The following conclusions can be drawn from
the present study.
(1) For the selected ground motion records with the PGA value lower
than frequent ground motion amplitude, the test structure
behaved with linear elastic behavior. After examining measured
strain values, it is found that all structural members had remained
elastic, including gravity load carrying frame members and SCMP
members. The test results also show that the inter-story drift
response under frequent ground motions met the limit of 0.4% in
the 2010 China Code, CSDB (2010) [38].
(2) In the cases with ground motion PGA values larger than that of
frequent ground motions, the system exhibited a flag-shaped
hysteresis curves in measured story shear vs. inter-story drift
ratio response, while the PTFs and gravity load carrying frame
showed nonlinear elastic behavior and linear elastic response,
respectively. The PT connections gap opening was enabled
through VBEs’ rocking about the HBE flanges. The replaceable
seismic fuse devices, SWs, yielded and the slats of SWs twisted out
of plane due to buckling.
(3) The test structure shows satisfactory self-centering and ductile
behavior. In Case 1–11 s (Northridge record scaled to PGA of
1.16 g), the peak inter-story drift ratios of the 1st-story and 2nd-
story were measured as 1.50% and 1.61%, respectively. After
completing the base excitation loading, the test structure was
found to return to the pre-loading position with negligibly small
residual inter-story drift ratio of 0.019%.
15
Engineering Structures 247 (2021) 113232
(4) Damage control, easy repair and resilience behavior of the test
structure are also inspected and experimentally validated
through shaking table test results. The seismic fuse device - SWs
dissipated energy by steel plastic deformation, while remaining
structural members and PTF elements in the test structure
behaved elastically during the tests. The initial pre-tensioning
force of PT cables remain largely unchanged after the tests. The
out-of-plane buckling and fracture in the steel slats between the
slits in the SWs reduced the lateral force capacity of SWs, and thus
due to this reduced force the plastic deformation of SWs did not
impede the structure’s ability of re-centering. The SW replace­
ment process was recorded, in which only the SWs need to be
replaced and each SW replacement took less than one hour for
two workers to execute. After repair, the dynamic properties and
inter-story drift response were found to be nearly identical to the
original test structure.
CRediT authorship contribution statement
Gongling Chu: Conceptualization, Methodology, Software, Formal
analysis, Investigation, Data curation, Writing – original draft, Visuali­
zation. Wei Wang: Conceptualization, Methodology, Supervision,
Project administration, Funding acquisition. Yunfeng Zhang: Concep­
tualization, Methodology, Validation, Investigation, Writing – review &
editing, Supervision.
Declaration of Competing Interest
The authors declare that they have no known competing financial
interests or personal relationships that could have appeared to influence
the work reported in this paper.
Acknowledgements
The financial supports from the National Natural Science Foundation
of China (NSFC) with Grant Nos. 51778459, 51820105013 and
52078366 are gratefully acknowledged. Supports for this study were
also provided by the State Key Laboratory of Disaster Reduction in Civil
G. Chu et al.
Engineering with Grant No. SLDRCE19-B-05 and Sustainable Structural
Engineering Research Funds from Tongji Architectural Design (Group)
Co. Ltd. .
References
[1] American Society of Civil Engineers (ASCE). Minimum Design Loads and
Associated Criteria for Buildings and Other Structures: ASCE/SEI 7-16; ASCE:
Reston, Virginia, USA, 2017.
[2] Tremblay R, Filiatrault A, Timler P, Bruneau M. Performance of steel structures
during the 1994 Northridge earthquake. Can J Civ Eng 1995;22(2):338–60.
[3] Tremblay R, Filiatrault A, Bruneau M, Nakashima M, Prion HGL, DeVall R. Seismic
design of steel buildings: Lessons from the 1995 Hyogo-ken Nanbu earthquake. Can
J Civ Eng 1996;23(3):727–56.
[4] Garlock MM, Ricles JM, Sause R. Experimental studies of full-scale posttensioned
steel connections. J Struct Eng 2005;131(3):438–48.
[5] Connor JJ, Wada A, Iwata M, Huang YH. Damage-controlled structures. I:
Preliminary design methodology for seismically active regions. J Struct Eng 1997;
123:423–31.
[6] Garlock MM, Ricles JM, Sause R. Cyclic load tests and analysis of bolted top-and-
seat angle connections. J Struct Eng 2003;129(12):1615–25.
[7] He X, Chen Y, Eatherton MR, Shao T. Experimental evaluation of replaceable
energy dissipation connection for moment-resisting composite steel frames.
J Struct Eng 2018;144(6):04018042. https://doi.org/10.1061/(ASCE)ST.1943-
541X.0002028.
[8] Qu B, Dai C, Qiu J, Hou H, Qiu C. Testing of seismic dampers with replaceable U-
shaped steel plates. Eng Struct 2019;179:625–39.
[9] Yao Z, Wang W, Fang C, Zhang Z. An experimental study on eccentrically braced
beam-through steel frames with replaceable shear links. Eng Struct 2020;206:
110185. https://doi.org/10.1016/j.engstruct.2020.110185.
[10] Ramirez CM, Miranda E. Significance of residual drifts in building earthquake loss
estimation. Earthquake Eng Struct Dyn 2012;41(11):1477–93.
[11] Miller DJ, Fahnestock LA, Eatherton MR. Development and experimental
validation of a nickel–titanium shape memory alloy self-centering buckling-
restrained brace. Eng Struct 2012;40:288–98.
[12] Chen J, Fang C, Wang W, Liu Y. Variable-friction self-centering energy-dissipation
braces (VF-SCEDBs) with NiTi SMA cables for seismic resilience. J Constr Steel Res
2020;175:106318. https://doi.org/10.1016/j.jcsr.2020.106318.
[13] Christopoulos C, Tremblay R, Kim H-J, Lacerte M. Self-centering energy dissipative
bracing system for the seismic resistance of structures: Development and
validation. J Struct Eng 2008;134(1):96–107.
[14] Zhu SY, Zhang YF. Seismic Behavior of Self-Centring Braced Frame Buildings with
Reusable Hysteretic Damping Brace. Earthquake Eng Struct Dyn 2007;36:1329–46.
[15] Clayton PM, Winkley TB, Berman JW, Lowes LN. Experimental investigation of
self-centering steel plate shear walls. J Struct Eng 2012;138(7):952–60.
[16] Chou C-C, Chen Y-C, Pham D-H, Truong V-M. Steel braced frames with dual-core
SCBs and sandwiched BRBs: Mechanics, modeling and seismic demands. Eng Struct
2014;72:26–40.
[17] Chou C-C, Tsai W-J, Chung P-T. Development and validation tests of a dual-core
self-centering sandwiched buckling-restrained brace (SC-SBRB) for seismic
resistance. Eng Struct 2016;121:30–41.
[18] Chou C-C, Hsiao C-H, Chen Z-B, Chung P-T, Pham D-H. Seismic loading tests of full-
scale two-story steel building frames with self-centering braces and buckling-
restrained braces. Thin-Walled Structures 2019;140:168–81.
[19] Wang W, Du XL, Zhang YF, Chu GL, Chen YY. Full-scale cyclic testing of self-
centering modular panels for seismic resilient structures. Key Eng Mater 2018;763:
339–46.
[20] Li J, Wang W, Qu B. Seismic design of low-rise steel building frames with self-
centering panels and steel strip braces. Eng Struct 2020;216:110730. https://doi.
org/10.1016/j.engstruct.2020.110730.
[21] Wang W, Kong J, Zhang Y, Chu G, Chen Y. Seismic behavior of self-centering
modular panel with slit steel plate shear walls: experimental testing. J Struct Eng
2018;144(1):04017179. https://doi.org/10.1061/(ASCE)ST.1943-541X.0001932.
16
Engineering Structures 247 (2021) 113232
[22] Wang W, Du X, Zhang Y, Chen Y. Experimental investigation of beam-through steel
frames with self-centering modular panels. J Struct Eng 2017;143(5):04017006.
https://doi.org/10.1061/(ASCE)ST.1943-541X.0001743.
[23] Hitaka T, Matsui C. Experimental study on steel shear wall with slits. J Struct Eng
2003;129(5):586–95.
[24] Jing L, Chen YY, Wang WH. Experimental study on full scale steel plate shear wall
with slits under low-frequency cyclic loads: Part I. Journal of Building Structures
2009;30(5):57–64.
[25] Wang W, Zhou Q, Chen Y, Tong L, Chan T-M. Experimental and numerical
investigation on full-scale tension-only concentrically braced steel beam-through
frames. J Constr Steel Res 2013;80:369–85.
[26] Chen Y, Wang W, Chen Y. Full-scale shake table tests of the tension-only
concentrically braced steel beam-through frame. J Constr Steel Res 2018;148:
611–26.
[27] Zhang R, Wang W, Ke Ke. Quantification of seismic demands of damage-control
tension-only concentrically braced steel beam-through frames (TCBSBFs) subjected
to near-fault ground motions based on the energy factor. Soil Dyn Earthquake Eng
2020;129:105910. https://doi.org/10.1016/j.soildyn.2019.105910.
[28] Chou C-C, Chen J-H. Seismic design and shake table tests of a steel post-tensioned
self-centering moment frame with a slab accommodating frame expansion.
Earthquake Eng Struct Dyn 2011;40(11):1241–61.
[29] Midorikawa M, Azuhata T, Ishihara T, Wada A. Shaking table tests on seismic
response of steel braced frames with column uplift. Earthquake Eng Struct Dyn
2006;35(14):1767–85.
[30] Ma X, Deierlein G, Eatherton M, et al. Large-scale shaking table test of steel braced
frame with controlled rocking and energy dissipating fuses. Proceedings of the 9th
U.S. National and 10th Canadian Conference on Earthquake Engineering. 2010.
Paper No 1248.
[31] Erochko J, Christopoulos C, Tremblay R, Kim H-J. Shake table testing and
numerical simulation of a self-centering energy dissipative braced frame.
Earthquake Eng Struct Dyn 2013;42(11):1617–35.
[32] Wiebe L, Christopoulos C, Tremblay R, Leclerc M. Mechanisms to limit higher
mode effects in a controlled rocking steel frame. 2: Large-amplitude shake table
testing. Earthquake Eng Struct Dyn 2013;42(7):1069–86.
[33] Lu X, Cui Ye, Liu J, Gao W. Shaking table test and numerical simulation of a 1/2-
scale self-centering reinforced concrete frame. Earthquake Eng Struct Dyn 2015;44
(12):1899–917.
[34] Dowden DM, Bruneau M. Dynamic shake-table testing and analytical investigation
of self-centering steel plate shear walls. J Struct Eng 2016;142(10):04016082.
https://doi.org/10.1061/(ASCE)ST.1943-541X.0001547.
[35] Qiu C, Zhu S. Shake table test and numerical study of self-centering steel frame
with SMA braces. B. Earthquake Eng Struct Dyn 2017;46(1):117–37.
[36] FEMA. 2009.
[37] Mazzoni S, McKenna F, Scott MH, Fenves GL. Open system for earthquake
engineering simulation user command-language manual. OpenSees version 2,
2006.
[38] Ministry of Housing and Urban-Rural Development of China. General
Administration of Quality Supervision, Inspection and Quarantine of the People’s
Republic of China, Code for Seismic Design of Buildings, GB 50011–2010. China:
Beijing; 2016.
[39] General Administration of Quality Supervision. Inspection and Quarantine of the
People’s Republic of China, Standardization Administration of China, Steel Strand
for Prestressed Concrete, GB/T 5224–2014. China: Beijing; 2014.
[40] Ministry of Housing and Urban-Rural Development of China. General
Administration of Quality Supervision, Inspection and Quarantine of the People’s
Republic of China, Code for Design of Concrete Structures, GB 50010–2010. China:
Beijing; 2015.
[41] Ricles JM, Sause R, Garlock MM, Zhao C. Posttensioned seismic-resistant
connections for steel frames. J Struct Eng 2001;127(2):113–21.
[42] Lowes LN, Mitra N, Altoontash A. A beam-column joint model for simulating the
earthquake response of reinforced concrete frames. Berkeley 2003.
[43] Macillo V, Shakeel S, Fiorino L, Landolfo R. Development and calibration of a
hysteretic model for cfs strap braced stud walls. Advanced Steel Construction 2018;
14(3):337–60.
[44] Applied Technology Council (ATC). Guidel ines for seismic performance
assessment of buildings: ATC-58 50% draft. Rep. No. 58, Washington DC, 2009.