Structures 57 (2023) 105093

Contents lists available at ScienceDirect

Structures
journal homepage: www.elsevier.com/locate/structures

Seismic control of cross laminated timber (CLT) structure with shape

memory alloy-based semi-active tuned mass damper (SMA-STMD)
Luyue Yan a, Yi Li a, Wen-Shao Chang b, Haoyu Huang c, *
a
Faculty of Architecture, Civil and Transportation Engineering, Beijing University of Technology, Beijing 100124, China
b
Lincoln School of Architecture and the Built Environment, University of Lincoln, Lincoln LN6 7TS, UK
c
School of Engineering, Newcastle University, Newcastle upon Tyne NE1 7RU, UK

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

Keywords: Cross laminated timber (CLT) is becoming increasingly popular for constructing multiple-story buildings due to
Semi-active control its alignment with Sustainable Development Goals of United Nations. However, such structures can be suscep­
Cross laminated timber structure tible to excessive vibrations during earthquakes. While tuned mass dampers (TMDs) can be used to control these
Tuned mass damper
vibrations in CLT structures, their effectiveness can be reduced due to the lightweight nature of CLT and changes
Shape memory alloy
Seismic control
in its structural mass or deterioration of the wood. To address this issue, this paper proposes the use of shape
memory alloy-based semi-active tuned mass dampers (SMA-STMDs) for controlling vibrations in CLT structures.
This study includes the design and experimental testing of a full-scale, novel spring-pendulum combined SMA-
STMD system, which utilises the mechanical properties of SMAs that change with temperature to achieve
semi-active control. Finite element method simulations demonstrate that the proposed SMA-STMD system can
effectively reduce structural seismic vibrations in CLT structure. By adjusting the SMA temperature, the SMA-
STMD system can effectively address the issue of TMD system off-tuning caused by changes in the CLT struc­
tural properties. Overall, the proposed SMA-STMD system offers a promising solution for seismic control in CLT
structures and has the potential to promote the development of CLT structures.

1. Introduction timber structures are being designed and built, with increasing heights
and number of stories.
In recent years, with the emergence of new engineered wood prod­ Compared to traditional building materials like concrete and steel,
ucts, the height of timber structures has been increasing, making timber CLT structures have a good strength-to-weight ratio and high ductility.
structure a popular trend in the global construction industry. Cross However, due to lower stiffness of timber, multi-storey CLT structures
laminated timber (CLT) is a type of high-strength solid wood panel could be susceptible to have larger responses under earthquake loads
created by bonding wood boards in alternating directions. It can be used [3,4]. Therefore, further development of multi-storey CLT structures
as shear wall in the construction of modern timber structures, as well as requires a reduction in their seismic response. Researchers worldwide
in steel-timber and concrete-timber hybrid structures. In 2012, Australia have gradually completed CLT connection tests [5,6], CLT shear wall
built a 9-story CLT building called Forte, measuring 32.2 m. Compared tests [7,8], CLT structural pseudo-dynamic tests [9,10], and CLT struc­
to reinforced concrete structures, it reduced carbon dioxide emissions by tural shaking table tests [11]. Austria and Japan have conducted shaking
1,451 tons [1]. In 2017, the University of British Columbia in Canada table tests on full-scale 3-storey and 5-storey CLT structures [12,13]. In
built an 18-story CLT student apartment measuring 54 m. By using 2013, the CNR-IVALSA research institute in Italy conducted shaking
prefabricated components in construction, it reduced building waste by table tests on a full-scale 7-storey CLT structure at NIED Miki in Japan as
two-thirds and decreased carbon dioxide emissions by 2,432 tons part of the SOFIE project [14]. The research results have shown that CLT
compared to reinforced concrete structures [2]. Therefore, building materials have excellent ductility, and the seismic response of CLT
more modern timber structures can be a significant way to reduce the structures is better than expected. However, shaking table tests have
carbon emissions and achieve the net zero target and the Sustainable found that the peak acceleration of multi-storey CLT structures is high.
Development Goals of United Nations. In the future, more and more tall In a very short time period, the top storey acceleration of the full-scale 7-

* Corresponding author.
E-mail address: [email protected] (H. Huang).

https://doi.org/10.1016/j.istruc.2023.105093
Received 5 May 2023; Received in revised form 25 July 2023; Accepted 17 August 2023
Available online 23 August 2023
2352-0124/© 2023 The Author(s). Published by Elsevier Ltd on behalf of Institution of Structural Engineers. This is an open access article under the CC BY license
(http://creativecommons.org/licenses/by/4.0/).
L. Yan et al. Structures 57 (2023) 105093

layer CLT structure reached 3.8 g, which could cause serious damage to tested, and the results showed that the SMA-STMD can effectively
non-structural components and cause some injuries to occupants address the problem of offtuning caused by changes in the controlled
[14,15]. The SOFIE project [14] concluded that reducing top storey structure’s mass [35]. However, the research on application of SMA-
acceleration response is a key research direction for future seismic STMD to a full-scale multi-storey CLT structure to address the poten­
research on multi-storey CLT structures and suggested that dampers tial offtuning problems is still scarce, and previous research only focused
could be used for seismic control in future studies. on the feasibility research on scaled and laboratory-size models. To
The tuned mass damper (TMD) system is a convenient and overcome this limitation, this paper will adopt the full-scale SMA-STMD
economical vibration control measure that has gained attention and to improve the seismic performance of multi-storey CLT structures. This
research across various fields since its concept was introduced by Den paper aims to reduce the excessive acceleration response in CLT struc­
Hartog in 1956 [16]. Studies have found that TMD systems have sig­ tures under seismic action using SMA-STMD and enhance the robustness
nificant vibration reduction effects in structures and have rapidly of tuning by avoiding the offtuning during use.
developed [17]. Poh’sié et al. [15,18] have innovatively applied the This paper can be outlined as follows: 1) Controller, 2) Controlled
TMD system to a full-scale, multi-storey CLT structure to resist seismic Object, and 3) Application of the Controller to the Object. Specifically,
action and found that under optimal tuning, and the TMD system can 1) Controller: Section 2 focuses on the design and testing of the SMA-
effectively reduce the seismic response of multi-storey CLT structures. STMD. 2) Controlled Object: Section 3.1 is dedicated to the modelling
However, there are limitations to the application of TMD systems in of the CLT structure. 3) Application of the Controller to the Object is
multi-storey CLT structures: (1) During the structural design process, explored in Sections 3.2, 3.3, and 3.4, which delve into the application of
calculating the natural frequency of the structure accurately is chal­ the SMA-STMD to the CLT structure. Section 4 is the Discussion.
lenging because timber is a natural material [19], and the installation of
TMD can cause a certain degree of frequency drift [16], which can 2. Test of SMA-STMD
significantly affect the control effect of TMD; (2) CLT structures are
lightweight [3]. The mass of the structure can easily change due to 2.1. SMA material
human movement and facility relocation, which can affect the natural
frequency of the CLT structure (previous studies by the authors have Among various families of SMA, Ni-Ti SMA has matured in its
reported that the natural frequency of timber structural systems can development, and possesses excellent material properties with good
easily change due to changes in mass [20,21]); (3) The stiffness of CLT temperature sensitivity [28] and long fatigue life [36], making it suit­
structures can easily change due to factors such as material aging, able for semi-active control in STMD. The experimental results [36]
moisture content, and further reinforcement during use [22]. Changes in affirm that both SMAs with different phase transformation temperatures
stiffness and mass can have a significant impact on the seismic reduction exhibit stable mechanical behaviours and possess a long fatigue life.
effect of TMD. In general, the natural frequencies of the TMD system and Additionally, Ni-Ti SMA has smaller grain size, which reduces the risk of
the controlled structure must be the same or similar to achieve reso­ grain boundary fracture. Therefore, all SMAs used in this study is Ni-Ti
nance effects. Otherwise, the TMD system will face serious offtuning (Ni56% and Ti44%) SMA, and they are provided by Xi’an SaiTe Metal
problems, leading to greater structural vibration. Therefore, how to Materials Development Co., Ltd. Differential Scanning Calorimetry
design TMD systems to control the seismic performance of multi-storey (DSC) tests were conducted on a 13.6 mg sample of the Ni-Ti SMA, with
CLT structures is a research problem that needs to be addressed. temperature settings ranging from − 40℃ to 100℃. The phase trans­
The application of semi-active tuned mass dampers (STMD) can solve formation temperatures of SMA were found to be Mf = -11.1℃, Ms =
the problem of offtuning in TMD systems. STMDs can adjust their own -9.3℃, As = -2.2℃, and Af = 25.7℃. The temperatures Ms and Mf refer
dynamic characteristics, expand their frequency domain control range, to the start and finish temperatures of the Martensite transformation,
and effectively adapt to the conditions where the dynamic characteris­ which is a phase transformation that occurs in SMA when it undergoes a
tics of the controlled structure are changed. STMDs have been cooling process. On the other hand, As and Af denote the start and finish
researched and verified to have better vibration reduction performance temperatures of the Austenite transformation. Austenite is a phase at
than passive TMDs [23]. Recently, innovative materials have been elevated temperatures. Through a monotonic tensile test on a Ni-Ti SMA
investigated for structural engineering applications, such as functionally dog bone bar with a diameter of 10 mm and an effective length of 30 mm
graded porous, composite materials and shape memory alloy (SMA) at room temperature (20℃), the elastic modulus and strength were
[24–27]. Their performances have been investigated subjected to ther­ determined to be 35.4 GPa and 822 MPa, respectively. Fig. 1 depicts the
mal and mechanical loading fields. This study aims to explore the
application of a new smart material - SMA - in STMDs to form an SMA-
STMD system, as the dynamic characteristics of the STMD can be
adjusted by changing the temperature of SMA [23,28]. Compared with
traditional STMDs [29,30], the advantages of SMA-STMDs mainly focus
on three aspects: (1) SMAs have superelastic performance, which means
that even after yielding, the material has the ability to recover to the
stress–strain origin, which can greatly reduce the residual deformation
of the damper material [31–33] and enhance the robustness of STMDs;
(2) SMA characteristics can be controlled by temperature, which can
solve the problem of offtuning in the structure. Previous research has
also found that SMA temperature control has a fast response speed [34];
(3) SMA can provide semi-active control relying on its own material
properties, with simple mechanical structures that will not cause sig­
nificant loads or space burdens on the structure.
Our pilot studies have explored whether the SMA-STMD can be used
to control vibration by adjusting the temperature of the SMA in order to
mitigate the structural vibration. Through vibration control experiments
on a cantilever steel beam structure, the feasibility of the concept of
using SMA temperature control for STMD vibration control was verified
[28]. Subsequently, a scaled two-storey steel frame structure has been Fig. 1. Stress–strain relationship of Ni-Ti SMA bar.

2
L. Yan et al. Structures 57 (2023) 105093

stress–strain hysteresis of the Ni-Ti SMA bar under cyclic tensile loading [ ][ •• ] [ ][ • ] [ ][ ]
using displacement control at room temperature, exhibiting typical m1 + m2 m2 l x1 (t)
••
c
+ 1
0 x1 (t)
• + 1
k 0 x1 (t)
superelastic behaviour. m2 l m2 l2 θ (t) 0 c2 h2
θ(t) 0 m2 gl + k2 h2
θ(t)
[ ]
F (t)
= 1
0
2.2. Setting-up of the SMA-STMD
(7)
The SMA-STMD system designed in this study is a combination of The natural frequency of the SMA-STMD can be calculated in
pendulum-type TMD and spring-type TMD, and it is an improvement Equation (8):
over the design presented in [23]. In [23], the mass block was suspended √̅̅̅̅̅̅̅̅̅̅̅̅̅̅̅̅̅̅̅̅̅̅̅̅̅̅
( )2
by the SMA bar, which caused the mass block to provide a pre-stress to 1 g k2 h
f2 = + (8)
the SMA bar. However, this approach is not feasible in a full-scale TMD 2π l m2 l
where the block mass is heavy. For SMA-STMD in this study, steel cables
where g denotes the acceleration of gravity. The derived equations
have been used to support the weight of the steel block, while SMA bars
play a crucial role in demonstrating the design considerations and pa­
are placed under the steel block and utilised to adjust the dynamic
rameters of the SMA-STMD. These equations form the foundation for the
characteristics of the STMD. The mechanical model of application of the
subsequent modelling of the SMA-STMD.
SMA-STMD to a structure is presented in Fig. 2. The function of SMA
The controlled structure is chosen to be the 7-storey CLT structure in
bars can be simplified into springs. The controlled structure is simplified
the SOFIE project [14]. The controlled structure has the natural fre­
as a single-degree-of-freedom system with mass m1 , stiffness k1 , and
quency of 2.34 Hz in x direction and 3.34 Hz in y direction. To design a
damping c1 . The external load applied to the controlled structure is
TMD for that, the optimal frequency ratio under seismic loading is set to
denoted as F1 . The stiffness of the SMA-STMD system, k2 , is provided by
be 0.96 for a mass ratio of 0.01 according to [15]. Thus, the natural
four Ni-Ti SMA bars placed in parallel to provide bending stiffness. The
frequencies of the TMD system in the x and y directions aim to be around
mass of the steel block is denoted as m2 , with damping c2 , and l repre­
2.25 Hz and 3.21 Hz, respectively. Based on Equation (8), the di­
sents the distance between the mass centre of the steel mass block and
mensions of the SMA-STMD system can be determined and the con­
the top of the steel frame. h is distance between the universal joint on the
structed laboratory model is illustrated in Figs. 3 and 4. It consists of six
top of the SMA bars and the top of the steel frame.
main components: a steel frame, four steel cables, a steel mass block,
The motion equations can be derived from the mechanical model
four universal joints, four Ni-Ti SMA bars, and a steel bottom plate. The
shown in Fig. 2 using the principle of virtual work:
mass block is a steel cube with a weight of 0.9648 t, suspended by four
− δWs = k2 θ(t)hδ(θ(t)h )+k1 x1 (t)δx1 (t) = k2 h2 θ(t)δθ(t)+k1 x1 (t)δx1 (t) (1) steel cables to support its entire weight. As shown in Fig. 4 (a) and (b),
the four steel cables are hinged to the top of a steel frame with di­
• • • •
− δWD = c2 θ(t)hδ(θ(t)h)+c1 x1 (t)δx1 (t) = c2 h2 θ(t)δ(θ(t))+c1 x1 δx1 (t) (2) mensions of 0.9 m × 0.9 m × 0.9 m. Four SMA bars are placed under the
steel cube, as seen in Fig. 3 (a) and Fig. 4 (c). The parameters of the SMA-
( ) STMD were designed based on Equation (8). The key dimensions are:
•• •• ••
− δWI = m2 lθ (t) + x 1 (t) δ(lθ(t) + x1 (t) ) + m1 x1 (t)δx1 (t) steel cube side length a = 503 mm, steel cable length b = 150 mm,
•• •• •• •• length between steel cube centre and oscillation centre l = 477 mm,
= m2 l2 θ (t)δθ(t) + m2 lx1 (t)δθ(t) + m2 lθ (t)δx1 (t) + m2 x1 (t)δx1 (t) length between hinge centre and oscillation centre h = 789 mm and
steel frame height H = 900 mm. The top of the SMA bars is connected to
••
+ m1 x1 (t)δx1 (t)
(3) the steel cube using universal joints which are hinges, while the foot of
the SMA rods is fixed on the steel bottom plate. This boundary condition
δWp = − m2 gθ(t)δ(lθ(t) ) = − m2 glθ(t)δθ(t) + F1 (t)δx1 (t) (4) ensures that the SMA bars bend instead of being sheared when the steel
mass block sways. The cross-section of the upper and lower gripping
Combine Equations (1), (2), (3) and (4):
sections of the SMA bars is round and larger than that in the effective
•• zone, as presented in Fig. 3 (b), to facilitate the bending occurring in the
(5)
•• • ••
m1 x1 (t) + c1 x1 (t) + k1 x1 (t) + m2 lθ (t) + m2 x1 (t) = F1 (t)
effective zone. In terms of SMA bars, each gripping section plus the
•• • ( ) transition zone is 20 mm long, and the effective zone is 30 mm long. To
(6)
••
m2 l2 θ (t) + c2 h2 θ(t) + k2 h2 + m2 gl θ(t) + m2 lx 1 (t) = 0 enable the TMD system to function in both the x and y directions, the
By using Equations (5) and (6), the motion equation of the SMA- effective cross-section of the SMA rod is designed as a rectangle with
STMD can be derived as Equation (7): dimensions of 15 mm (x direction) × 25 mm (y direction), providing
different bending stiffness in the x and y directions for the TMD system.
After a free vibration testing, the natural frequencies of the SMA-STMD
are 2.09 Hz in x direction and 3.32 Hz in y direction, respectively, which
are similar with the target frequencies.

2.3. Methods

To control the SMA-STMD by changing the temperature of SMA, it is

important to investigate the effect of temperature changes on the dy­
namic characteristics of the SMA-STMD system. Also, the effective fre­
quency domain for semi-active control should be obtained.
The dynamic characteristics were determined by free vibration. This
study chose x direction as a case study for control. An initial displace­
ment of 50 mm was applied to the steel mass block in the x direction to
enable free vibration. An accelerometer was installed on the steel mass
block to obtain the acceleration time history, with a sampling frequency
Fig. 2. Mechanical model of SMA-STMD applied to a controlled structure. of 1000 Hz. The temperature of the SMA can be adjusted. The cooling

3
L. Yan et al. Structures 57 (2023) 105093

Fig. 3. (a) Schematic diagram of SMA-STMD; (b) SMA bar in the SMA-STMD.

was achieved by spraying 1,1,1,2-Tetrafluoroethane, and the heating SMA. According to the study conducted by Shaw and Kyriakides [39],
was achieved by using electric blowers. Four thermocouples were the SMA undergoes a phase transformation into Martensite at temper­
attached to the outer surface of the middle section of the four SMA bars. atures at below 0℃, resulting in a low and constant loading trans­
After the temperature was raised or lowered, the temperature of the formation stress. In the Martensite phase, it can be observed that the
SMA bars was kept constant for 2 min before the free vibration test transformation stress of the SMA becomes low and its stiffness remains
began. This step allowed the internal temperature of the material to relatively constant [39]. Consequently, the relationship between stiff­
equilibrate and accounted for the heat transfer effects. By stabilising the ness and temperature should follow a quadratic function. The paper [39]
temperature of the SMA, this process contributed to maintaining the also observed that there was a linearly declined unloading trans­
accuracy and reliability of the experimental results. formation stress with the lower temperature, and the hysteretic loop is
The first step of the test was to measure the acceleration time history becoming larger. As a result, a linear function is employed to represent
of the SMA-STMD system in the x-direction free vibration under room the damping ratio with the temperature.
temperature (20℃) for 20 s. The second step was to control the working Based on the analysis above, by curve fitting, the relationship be­
temperature of the SMA bars to vary from − 40℃ to 80℃, with a tem­ tween the natural frequency and the SMA temperature as well as the
perature interval of 10℃, and measure the acceleration time history of relationship between the damping ratio and the SMA temperature can be
the SMA-STMD system in the x-direction free vibration for 20 s in each depicted by equations. As a result, the dynamic characteristics of SMA-
temperature. The natural frequency and damping ratio of the SMA- STMDs, in terms of natural frequency and damping ratio, can be effec­
STMD can be analysed by the Matrix Pencil method [37,38] using the tively adjusted by varying the temperature. This temperature-sensitive
measured acceleration timber history. behaviour of SMA offers a promising way to improve the performance
of structural vibration control systems.

2.4. Results 2.5. Discussions

The variations of the natural frequency and the damping ratio of the In this study, four SMA bars were used. In practical applications,
SMA-STMD are shown in Fig. 5, indicating that the natural frequency of more SMA bars can be placed in parallel to widen the frequency and
the SMA-STMD system increased and the damping ratio decreased with damping ratio range. It is also possible to use SMAs with different phase
the temperature rise. When the temperature increases from − 40℃ to transformation temperatures to achieve various temperature sensitiv­
80℃, the natural frequency increased from 1.88 Hz to 2.30 Hz (Fig. 5 ities. Since the size of the SMA bars may vary in practical use, the
(a)). The reason is that the temperature can change the elastic modulus transfer of heat and the effect of temperature on stiffness may differ. In
of the SMA, which has been proved in the previous research [23]. the future, more advanced and effective heating or cooling measures can
From Fig. 5 (b), the damping ratio decreased from 12.86% to 2.69% be adopted, such as using direct current voltage control for Joule heating
from − 40℃ to 80℃. The Ni-Ti SMA used in the test had phase transition of the SMA bars.
temperatures of Mf = -11.1℃, Ms = -9.3℃, As = -2.2℃, and Af = 25.7℃.
From − 2.2℃ to 25.7℃, the SMA underwent a phase transformation 3. Application of SMA-STMD to the CLT structure for seismic
from martensite to austenite, and at this stage, it was in a mixed state of control
martensite and austenite. From 25.7℃, it was the disappearance of
martensite, and the SMA exhibited its superelastic properties. When the The Section 3 revolves around assessing the effectiveness of the SMA-
temperature was reduced below − 11.1℃, SMA behaved in shape STMD on the CLT structure. To achieve this, the paper relies on FEM
memory effect (SME), and it had a performance with larger energy analysis, Section 3.1 models the CLT structure and Section 3.2 models
dissipation. the installation of the SMA-STMD. The testing is detailed in Sections 3.3
In Fig. 5, the influence of temperature on natural frequency is and 3.4. Section 3.3 serves as the methodology, and Section 3.4 then
modelled as a quadratic function while the influence of temperature on presents the results of this analysis.
damping ratio is modelled as a linear function. The reason for this
behaviour can be attributed to the phase transformation property of the

4
L. Yan et al. Structures 57 (2023) 105093

Fig. 4. Construction diagram of (a) the SMA-STMD; (b) the steel cables; (c) the SMA bars.

Fig. 5. Natural frequency (a) and damping ratio (b) change of the SMA-STMD with temperature change.

5
L. Yan et al. Structures 57 (2023) 105093

3.1. The CLT structure FEM modelling methods can be found in [15] in details. The S4R elastic
shell elements with isotropic properties were used to simulate CLT floor
This study aims to use finite element method (FEM) to investigate the slabs and wall panels. The elastic modulus of elasticity was calculated
vibration control performance of the SMA-STMD on the CLT structure. using the Blass-Fellmoser method [41], which averages the stiffness of
The controlled structure chosen in this study is the seven-storey CLT the CLT panels in the vertical and horizontal directions, respectively, at
structure in the SOFIE project [14], as shown in Fig. 6, which was 370 MPa and 11000 MPa. The elastic modulus of the five-layer CLT
simulated using FEM software ABAQUS. Fig. 7 shows the plan and south panels is estimated to be 5685 MPa [15]. There are three types of metal
elevation view of the CLT structure, with overall dimensions of 13.44 m connectors between the CLT panels in the SOFIE experiment, namely,
in the x-direction and 7.68 m in the y-direction, and a structural height angle brackets, hold-downs, and screws. In the FEM model, each shell
of 23.20 m. The thickness of CLT panels decreasing along the height: element is connected to another through elastic spring elements, simu­
142 mm for the first and second floors, 125 mm for the third and fourth lating the connectors between CLT timber. The stiffness of each spring
floors, and 85 mm for the roof and other parts. The details in terms of element is based on the values provided by [15]. The damping ratio of
mass of the CLT structure in the SOFIE experiment is also shown in [14], the structure is set to 2% using Rayleigh damping.
with a total weight of 278 t for the CLT structure simulated in this study Fig. 8 shows the FEM model and mesh partition of the CLT structure.
using finite element analysis. Other details of the CLT structure are The natural frequencies of the FEM model of the CLT structure in the x
presented in [14]. and y directions are 2.18 Hz and 3.62 Hz, respectively. During the SOFIE
Currently, there are three main methods for FEM modelling of CLT full-scale test [14], the natural frequencies of the CLT structure in the x
structures, which are the combination of elastic shell elements and and y directions were 2.34 Hz and 3.32 Hz, respectively, with an error of
nonlinear spring elements, the combination of elastic shell elements and less than 10% compared to the FEM modelling results. This level of error
linear spring elements, and a simplified method using nonlinear spring is acceptable given the uncertainties in the finite element simulation.
elements to simulate walls. In this study, the methods of elastic shell However, a more complex nonlinear model could yield more precise
elements and elastic spring elements were used which referred to [15]. results. When subjected to the JMA Kobe earthquake at the bottom of the
Fragiacomo et al. [40] have conducted a case study of the elastic seismic CLT structure, the peak accelerations in the x and y directions are 3.9 g
analysis of a multi-storey crosslam building under seismic actions. The and 1.6 g, respectively. During the SOFIE test, the CLT structure expe­
results found that the elastic modelling led to an acceptable approxi­ rienced peak accelerations of 3.8 g in the x direction and 2.4 g in the y
mation of the wooden panel behaviour. Also, there was absence of non- direction under the JMA Kobe earthquake, with an error within an
linear damage in the structural elements after the shaking table tests acceptable range.
[14]. The absence of damage in the structural elements observed after
the shaking-table tests allows for the use of an elastic model. Relative 3.2. Application of the SMA-STMD

A common mass ratio of 1–5% between TMD and the main structure
is recommended because a ratio that is too low would have little effect,
while a ratio that is too high would place a heavy loading burden on the
floor slab. However, since the CLT floor has a lower stiffness compared
to concrete or steel floors, a mass ratio of 1% was chosen for this study.
To achieve this ratio, three SMA-STMDs were used and placed on the top
floor of the CLT structure. Since the mass of each SMA-STMD is 0.964 t,
applying three of them results in a mass ratio of 1.06%.
The FEM model of the SMA-STMD is presented in Fig. 10. To
construct the steel frame, the B31 beam element was utilised to simulate
the steel cable, universal joint sleeve, hinge block, cross hinge, SMA
bars, and mass block. The bottom plate was modelled using the C3D8
solid element. All dimensions of the model parts were consistent with
those of the full-scale SMA-STMD laboratory model in Section 2. The
stress state of both the universal joint and SMA bar in the SMA-TMD
system is of utmost importance. To accurately simulate the SMA-TMD
system using FEM, the mesh division for these components is inten­
tionally dense. Additionally, mesh refinement techniques were
employed in areas with stress concentration, such as the connections
between various parts. To achieve an optimal mesh, adaptive mesh
refinement was utilised, and manual adjustments were made through
face partitioning. Moreover, considering the irregular shapes of the
universal joint, SMA bar components, and mass blocks, the C3D10M
mesh generation mode was applied. The finer mesh size is 5 mm. The
bottom plate, on the other hand, was meshed using the C3D8R mesh
generation mode.
All materials used in FEM are identical to those in Section 2. To
simplify the modelling, all the materials were modelled to be elastic. The
energy dissipation property was modelled by setting the Rayleigh
damping for the entire SMA-STMD system. The steel frame, universal
joint, base plate, and steel cables were modelled using steel material,
with their elastic modulus set to 2.06 × 105 N/mm2 and Poisson’s ratio
set to 0.3. The SMA bar was modelled explicitly as shown in Fig. 11 (a),
and the elastic modulus and dimensions of SMA were the same as the
value in Section 2.
Fig. 6. The 7-storey CLT structure in SOFIE project [14]. Fig. 11 (b) shows the FEM simulation of the universal joint of the top

6
L. Yan et al. Structures 57 (2023) 105093

Fig. 7. (a) Plan views of the CLT structure; (b) South elevation view of the CLT structure [14].

Fig. 9. Installation of the SMA-STMDs on the CLT structure.

the x and y directions, with natural frequencies of 2.09 Hz and 3.33 Hz,
respectively. The natural frequencies in both directions were close to
those measured in laboratory tests (2.09 Hz and 3.32 Hz), indicating that
Fig. 8. FEM model of the CLT structure in Abaqus.
the FEM modelling was effective. This provides a foundation for the
subsequent FEM simulation of the SMA-STMD system applied to CLT
of the SMA bar, which consists of three parts: upper and lower sleeves, a structures.
central hinge block, and a cross slide block. Hard contact is used to Based on the free vibration tests of the SMA-STMD system in Section
simulate the contact surfaces between the cross hinge and the upper and 2, the natural frequency and damping ratio of the system are found to
lower sleeves, and between the hinge block and the sleeves. Smooth vary with changes in temperature of the SMA bar. In Abaqus, it is
contact is set to simulate the characteristic of the universal joint being possible to simulate this effect by adjusting the elastic modulus of the
able to rotate freely. SMA to reflect the change in natural frequency and modifying the
The modal analysis results of the SMA-STMD system are shown in Rayleigh damping to account for the change in damping ratio caused by
Fig. 12. The first and second mode shapes were found to be swinging in temperature.

7
L. Yan et al. Structures 57 (2023) 105093

(Imperial Valley) in NS direction with PGA of 0.35 g, and the magnitude

is 7.1.
The earthquake was performed in x direction (Fig. 9) as a case study,
which is in line with the deformation direction of the structural first
mode shape. Although it is important to acknowledge the need for
comprehensive coverage, analysing the y direction would be redundant
as it would essentially involve re-examining the same control theory.
Therefore, only x direction is tested in this study. The seismic perfor­
mance of the CLT structure was evaluated through various tests, as
depicted in Fig. 13 and summarised in Table 1. Test 1 examined the
structural behaviour without TMD, while Test 2 involved the use of a
tuned SMA-STMD. Tests 3 and 4 were conducted with offtuned SMA-
STMD. The offtuning scenarios were induced by modifying the CLT
structural properties. In Test 3, the natural frequency decrease was
achieved by reducing the elastic modulus of the CLT, which could occur
in real-life scenarios due to material deterioration caused by high
moisture content or aging. In Test 4, the natural frequency increase was
Fig. 10. FEM model of the SMA-STMD system. modelled by decreasing the structural mass, which could result from
human or equipment relocation in the building. Tests 2, 3 and 4 spe­
cifically investigate the performance of a passive TMD (traditional TMD)
under consideration. In these tests, the semi-active control function was
turned off and it became a passive TMD. Finally, Tests 5–8 explored the
retuning by adjusting SMA temperature on the SMA-STMD accordingly.
The details of the retuning are described in Sections 3.4.2 and 3.4.3.

3.4. Results

3.4.1. Effectiveness of SMA-STMD

Fig. 14 shows the application of SMA-STMD on the top of CLT
structure under El Centro earthquake. Fig. 14 (a) presents the acceler­
ation time history curves in the x-direction of the mass centre of the top
floor in Tests 1 and 2. As shown in Table 1, the reduction in top-storey
peak acceleration amplitudes is 28.00% from Test 1 to Test 2, while the
Fig. 11. (a) FEM model of SMA bar; (b) FEM model of the universal joint.
reduction in RMS acceleration amplitudes is 39.24%. Fig. 14 (b) shows
the peak acceleration per storey of the CLT structure in Tests 1 and 2,
and it proves that the tuned SMA-STMD is effective to reduce the vi­
3.3. Testing plan
bration in each storey and the effectiveness become more significant in
higher storey. The results are aligned with the expectations.
In this section, the dynamic performance of the CLT structure with
In summary, Fig. 14 provides strong evidence that SMA-STMD is an
and without three SMA-STMDs was investigated under seismic loading.
effective solution for reducing the vibration response of CLT structures
All the investigations were based on Abaqus modelling, and the CLT
under earthquake loading. The results suggest that SMA-STMD can be a
model and the SMA-STMD model built in Sections 3.1 and 3.2 were
viable alternative for seismic protection of CLT structures.
employed. The El Centro earthquake can cause strong vibrations lasting
between 1.5 and 5.5 s, with a dominant frequency range of 0.39–6.39
3.4.2. Offtuning caused by deteriorating and its retuning
Hz. The first and second natural frequencies of the CLT structure were
The results presented in Fig. 15 (a) and Table 1 reveal that the peak
within this frequency range. In structural engineering applications,
acceleration at the top of the CLT structure increased significantly from
seismic forces are random and have a broad frequency range. Therefore,
1.471 g in the tuned state of Test 2 to 2.025 g in the offtuning condition
this is the reason why the non-scaled El Centro earthquake was selected
of Test 3. This represents a 37.66% increase in peak acceleration and a
to validate the effectiveness of the SMA-STMD measures on the CLT
57.24% increase in RMS acceleration. 2.025 g is close to the peak ac­
structure. The earthquake wave for input was El Centro seismic wave
celeration of 2.043 g at the top of the CLT structure without the TMD

Fig. 12. Modal analysis of SMA-STMD system: (a) First mode shape in x direction; (b) Second mode shape in y direction.

8
L. Yan et al. Structures 57 (2023) 105093

Fig. 13. Testing plan.

Table 1
SMA-STMD system applied to CLT structure tuning, offtuning and retuning.
Without With SMA-STMD
TMD With tuned SMA- Natural frequency of CLT Cooling SMA to Natural frequency of CLT Heating SMA to
(Test 1) STMD reduced retune increased retune
(Test 2) (Test 3) Test 5 Test 7 (Test 4) Test 6 Test 8

Mass of the CLT structure (t) 278 278 278 278 278 259 259 259
Elastic Modulus of CLT (MPa) 5685 5685 4000 4000 4000 5685 5685 5685
f1 (Hz) 2.18 2.18 1.99 1.99 1.99 2.25 2.25 2.25
Temperature of SMA (℃) n/a 20 20 − 40 − 80 20 50 50
f2 (Hz) n/a 2.09 2.09 1.90 1.81 2.09 2.16 2.16
ξ2 (%) n/a 7.54 7.54 12.86 17.78 7.54 4.01 15.00
Peak acceleration (mm/s2 ) 2.043 1.471 2.025 1.954 1.828 1.833 1.846 1.600
Reduction of the peak Reference 28.00 0.88 4.36 10.52 10.28 9.64 21.68
acceleration (%)
RMS acceleration (gal) 60.809 36.950 58.099 50.921 45.881 45.579 48.662 39.981
Reduction of the RMS Reference 39.24 4.45 16.30 24.55 25.05 19.98 34.25
acceleration (%)

f1 : Natural frequency of the CLT structure.

f2 : Natural frequency of the SMA-STMD.
ξ2 : Damping ratio of the SMA-STMD.

system. The acceleration per storey also showed a similar trend, indi­ exercised to avoid excessive modification of the SMA stiffness.
cating that the SMA-STMD system became ineffective in the offtuning
condition. The results show a risk of using TMD system. 3.4.3. Offtuning caused by reducing structural weight and its retuning
Two plans (Tests 5 and 7) were conducted to retune the SMA-STMD The second offtuning was caused by reducing the CLT structural mass
system as shown in Table 1. Test 5 cooled down the SMA to − 40℃ and from 278 t to 259 t. According to Table 1 and Fig. 16, by adjusting the
Test 7 cooled down the SMA to − 80℃. Test 5 involved retuning the structural mass in Test 4, the peak acceleration at the top of the CLT
system to the optimal frequency ratio of 0.95, while Test 7 used a structure increased to 1.833 g and the RMS acceleration at top increased
slightly lower frequency ratio of 0.91, but with a higher damping ratio of to 45.579 Gal compared to the tuned condition in Test 2. It can be found
17.78% achieved through cooling to have a more significant SME. The that the seismic performance with the offtuned SMA-STMD system
parameters of the retuned SMA-STMD in Tests 5 and 7 are presented in became worse.
Table 1, based on the effect of temperature on the dynamic character­ To address the problem, the retuning was performed to raise the
istics of the system as tested in Section 2.4. frequency of the SMA-STMD system by increasing the temperature of
From Table 1, the results indicate that Test 5 was effective in SMA to 50℃ in Test 6. Based on the relationship concluded in Section
reducing the RMS acceleration by 12.35% and the peak acceleration by 2.4, at this temperature, the frequency of the SMA-STMD system was
3.51% compared to Test 3, confirming the success of the retuning pro­ 2.16 Hz, where the frequency ratio is 0.96. As shown in Table 1 and
cess. In Test 7, the peak acceleration was reduced by 9.73% and the RMS Fig. 16, when El Centro earthquake was applied to the CLT structure, the
acceleration by 21.03% compared to Test 3, demonstrating the effective difference of the acceleration response between Test 4 and 6 is little. The
retuning of the SMA-STMD system. In comparison, retuning in Test 7 is reason is that the damping ratio of the SMA-STMD system decreased
more effective. Importantly, the results suggest that taking advantage of from 7.54% to 4.01% after heating SMA, which is less effective for
the SME property of SMA in the retuning process can further enhance the seismic control.
energy dissipation. This was observed in Fig. 5, where cooling the SMA Test 8 demonstrated that the natural frequency of the SMA-STMD
allowed for greater utilisation of SME. However, caution should be can be adjusted by heating the SMA, while maintaining a sufficient

9
L. Yan et al. Structures 57 (2023) 105093

Fig. 14. (a) Acceleration time history of the CLT structure with SMA-STMD (Test 2) and without TMD (Test 1); (b) Peak acceleration per storey of the CLT structure
with SMA-STMD (Test 2) and without TMD (Test 1).

damping ratio (15%) for TMD. This assumes that the SMA can maintain shown in Fig. 17(a). The pre-bending process can use a hydraulic or
a high damping such as being prestressed or staying in SME without mechanical press to bend the SMA bars to a certain degree. The amount
undergoing phase transformation. Fig. 16 and Table 1 show that under of pre-bending required depends on the specific application and desired
the retuning condition, the peak acceleration of the CLT structural top level of prestress. As shown in Fig. 17(b), prestressing the SMA bars
storey can be reduced by 12.71% compared to Test 4. Additionally, the causes a shift in the coordinate point to the top right, indicating a ‘jump’
RMS acceleration was reduced by 12.28%, indicating a satisfactory over the elastic region. As a result, the SMA material can dissipate more
reduction effect. Thus, the SMA-STMD system with SMAs in higher energy and exhibit higher damping behaviour. To accurately achieve the
damping of 15% can be effectively retuned. It can be found that using targeted damping ratio, the implementation of temperature control can
SMAs in higher damping provides an added advantage over passive be utilised. The findings presented in Section 2.4 can serve as a valuable
damping systems that cannot be retuned. The potential for further im­ guide for adjusting the temperature accordingly. The temperature con­
provements in damping through the use of SMAs is evident, and this trol can be achieved by electric heating and cooling and is aimed to be
study offers valuable insights into the efficacy of SMA-STMD systems in experimentally tested in the future.
reducing structural vibrations. In order to further enhance the damping properties of SMA-based
The further direction for improving the SMA-STMD is to increase the systems, it is possible to select higher damping SMA materials. Addi­
damping capacity. Several solutions can be explored to address the tionally, the damping ratio of the SMA-STMD system can be increased by
limitation of lower damping in higher temperatures of SMA. One incorporating additional damping materials or components. The
effective way of achieving higher damping in SMA is to apply a prestress damping behaviour of Ni-Ti SMA is influenced by factors such as ma­
to the SMA bars. This technique involves pre-bending the SMA bars, as terial composition, excitation frequency, and heat treatment method.

10
L. Yan et al. Structures 57 (2023) 105093

Fig. 15. (a) Acceleration time history curve at top storey and (b) peak acceleration per storey of the CLT structure during tuning (Test 2), offtuning (Test 3) and
retuning (Tests 5 and 7).

Therefore, it would be beneficial to study and understand the factors bars in parallel to accumulate damping coefficients. This method helps
that affect the damping of Ni-Ti SMA, and subsequently select SMA to multiply damping coefficients and improve the overall performance
materials with higher damping properties, such as those with a broader of the system. Overall, the benefits of increasing damping in SMA-STMD
range of Martensite. Increasing the phase transformation temperatures are evident. The aforementioned solutions emphasise the potential for
of the SMA can also have significant benefits for maintaining a high level further improvement and development of SMA-STMD.
of damping because it will be in its SME state. This can be achieved by
altering the composition of the SMA material, which involves changing 4. Discussion
the ratios of various elements presenting in the alloy. For example,
changing the nickel content or changing elements such as cobalt can To enhance the practicality of the SMA-STMD and make it more
change the transformation temperature, and an increase in annealing applicable for a wider range of applications, several measures can be
temperature also helps increasing the transformation temperature [42]. taken to improve the dynamic characteristics of the system. One po­
In addition to altering the composition of the SMA, another way to in­ tential method is to enhance the heating and cooling equipment, which
crease the phase transformation temperatures is through mechanical can expand the range of achievable temperatures and improve the
processing. Furthermore, shape memory polymer (SMP) composites can versatility of the system. For example, incorporating more advanced
enhance the performance of SMA-STMDs. By embedding the SMA within heating and cooling systems, such as thermoelectric coolers or fluidic
an SMP matrix, the composite material can provide improved energy systems, can allow for more precise control of temperature gradients and
dissipation capabilities due to the damping properties of the polymer enable the SMA-STMD to operate more effectively across a broader
[43]. If there were accumulated residual strain in SME after cyclic range of temperatures. Currently, there is a limited pool of references
deformation, heating would be used to remain the resilience. To available for meaningful comparisons of the dynamic control technics
enhance the damping of SMA-STMD, another option is to add more SMA on tall CLT structures. Moving forward, our research aims to incorporate

11
L. Yan et al. Structures 57 (2023) 105093

Fig. 16. (a) Acceleration time history curve at top storey and (b) peak acceleration per storey of the CLT structure during tuning (Test 2), offtuning (Test 4) and
retuning (Tests 6 and 8).

a broader range of dynamic control methods in order to facilitate a To enhance the practicality and applicability of SMA-STMD, the
comprehensive comparative analysis. SMA-STMD can be designed to be adjusted to suit the practicality. For
Another way to broaden the frequency band of the SMA-STMD is to example, the sleeve of the bottom plate in the SMA-STMD system is
increase the number of SMA bars used in the system. This can help to designed to be detachable, allowing for easy replacement of SMA bars to
create various combinations to achieve different mechanical and ther­ adjust their size. Additionally, the length of the steel cable can be
momechanical properties. By doing these, the dynamic characteristics of adjusted using bolts. In the future, the SMA-TMD system can be effec­
the system can be more versatile. Additionally, by optimising the di­ tively combined with Structural Health Monitoring (SHM) technology to
mensions and configuration of the SMA bars and steel cable used in the address the issue of maladjustment in specific frequency domains, which
SMA-STMD, it may be possible to further improve its performance and can have serious consequences for TMD systems. By implementing ac­
effectiveness. Other potential areas for development of the SMA-STMD celeration detection on controlled structures using SMA-TMD systems, it
include exploring alternative materials and manufacturing techniques becomes possible to monitor structural behaviour. Signal processing
for the SMA bars and steel cable. For instance, the use of advanced techniques, coupled with artificial intelligence, machine learning, and
composites or alloys may allow for greater strength and stiffness while related technologies, can be employed to rapidly calculate seismic and
also reducing the weight and overall size of the system. structural information. By adjusting the SMA temperature, the TMD

12
L. Yan et al. Structures 57 (2023) 105093

Fig. 17. (a) Prestressing SMA in the SMA-STMD; (b) The effect of prestressing on the stress–strain hysteresis of SMA.

system can be quickly retuned, enabling semi-active control of the This paper holds innovations. It proposes the theory of using SMA-
controlled structure and effectively mitigating vibrations. STMD as a solution for controlling vibrations in CLT structures. The
In practice, earthquakes are considered unpredictable events. Re­ use of SMA-STMD is a novel approach that takes advantage of the me­
searchers have conducted several studies [44,45] that aim to establish chanical properties of SMA, which change with temperature, to achieve
optimal equations for TMD parameters under different types of seismic semi-active control of structural vibrations. Additionally, the study of­
loadings. These studies provide valuable insights into achieving desir­ fers a comprehensive design and examination of a full-scale SMA-STMD
able performance of the TMD system. Building upon this existing system. Both experimental and numerical methods are provided to
research, this study aims to implement the semi-active control tech­ assess the system’s behaviour. Also, the study includes the specific FEM
niques to maintain these parameters of TMD at their optimal values. In simulation techniques used to investigate the TMD system and its
other words, this study provides the control theory to maintain the structural application. The proposed SMA-STMD system addresses the
TMD’s parameters to be optimal for suiting different types of earthquake issue of excessive vibrations in CLT structures during earthquakes. It
waves. overcomes the limitations of traditional TMDs by utilising SMA and its
ability to adjust properties based on temperature changes.
5. Conclusion All in all, this technology holds immense potential for advancing the
development of multi-storey CLT structures, further promoting sus­
This study investigated the application of the SMA-STMD to the CLT tainable construction practices. In the future, several solutions such as
structure for seismic vibration control. An innovative pendulum-spring prestressing will be explored to increase the energy dissipation of SMA.
combined SMA-TMD system was designed and tested through experi­ More accurate and quicker temperature control approaches will be
mental tests, and the effect of temperature on its dynamic characteristics tested for the full-scale SMA-STMD system.
was obtained. Then, the SMA-STMD system was simulated using FEM.
By conducting numerical modelling for a CLT structure, the effective­
Declaration of Competing Interest
ness of the SMA-STMD control on the CLT structure was verified by
simulating the tuning, offtuning, and retuning process.
The authors declare that they have no known competing financial
The main conclusions are as follows:
interests or personal relationships that could have appeared to influence
the work reported in this paper.
The material properties of Ni-Ti SMA bar can be changed by varying
temperature. The natural frequency of SMA-STMD decreased from
Acknowledgement
2.30 Hz to 1.88 Hz, and the damping ratio increased from 2.69% to
12.86% when the SMA temperature changed from 80℃ to − 40℃.
The authors would like to thank the financial support from the Start-
The SMA-STMD system can effectively mitigate seismic vibration in
Up Funding at Newcastle University.
CLT structures compared to the non-TMD condition. When subjected
to the El Centro earthquake, the peak acceleration was reduced by
28.00%, and the RMS acceleration was reduced by 39.24%. References
Changing the main structural properties can cause offtuning of the
[1] Woodsolutions. Forte Living. 2023; Available from: https://www.woodsolutions.
TMD, which can increase structural vibration. Retuning the SMA- com.au/case-studies/forte-living.
STMD by cooling the SMA reduced the peak acceleration by 9.73% [2] Nations, U., Circularity concepts in wood construction ed. C.o.F.a.t.F. Industry. 2022,
and the RMS acceleration by 21.03% compared to the offtuning Geneva: United Nations.
[3] Reynolds, T., et al., Ambient vibration tests of a cross-laminated timber building.
condition. Proceedings of the Institution of Civil Engineers-Construction Materials, 2015. 168
By elevating the level of damping in the SMA, the effectiveness of the (3): p. 121-131.
SMA-STMD in both cooling and heating can be enhanced, resulting [4] Feldmann, A., et al., Dynamic properties of tall timber structures under ambient
vibration, in World Conference on Timber Engineering. 2016: World Conference on
in an improved retuning process. Timber Engineering.

13
L. Yan et al. Structures 57 (2023) 105093

[5] Gavric I, Fragiacomo M, Ceccotti A. Cyclic behaviour of typical metal connectors [26] Xiao X, et al. Analytical model for the nonlinear buckling responses of the confined
for cross-laminated (CLT) structures. Mater Struct 2015;48(6):1841–57. polyhedral FGP-GPLs lining subjected to crown point loading. Eng Struct 2023;
[6] Huang HY, Chang WS. Seismic resilience timber connectionadoption of shape 282:115780.
memory alloy tubes as dowels. Struct Control Health Monit 2017;24(10). [27] Xiao X, et al. Nonlinear in-plane instability of the confined FGP arches with
[7] Hristovski V, et al. Full-Scale Shaking-Table Tests of XLam Panel Systems and nanocomposites reinforcement under radially-directed uniform pressure. Eng
Numerical Verification: Specimen 1. J Struct Eng 2013;139(11):2010–8. Struct 2022;252:113670.
[8] Sun XF, et al. Performance evaluation of multi-storey cross-laminated timber [28] Huang HY, Chang WS, Mosalam KM. Feasibility of shape memory alloy in a
structures under different earthquake hazard levels. J Wood Sci 2018;64(1):23–39. tuneable mass damper to reduce excessive in-service vibration. Struct Control
[9] Lauriola, M.P. and C. Sandhaas. Quasi-static and pseudo-dynamic tests on XLAM walls Health Monit 2017;24(2):1–14.
and buildings. in COST E29 International workshop on earthquake engineering on [29] Nagarajaiah S, Varadarajan N. Novel semiactive variable stiffness tuned mass damper
timber structures. 2006. Coimbra. with real time tuning capability, t.E.M.C. Virginia.: ASCE; 2000.
[10] Popovski M, Gavric I. Performance of a 2-Story CLT House Subjected to Lateral [30] Sun C, et al. Hardening Duffing oscillator attenuation using a nonlinear TMD, a
Loads. J Struct Eng 2016;142(4). semi-active TMD and multiple TMD. J Sound Vib 2013;332(4):674–86.
[11] Ceccotti A. New Technologies for Construction of Medium-Rise Buildings in [31] Aguiar RAA, Savi MA, Pacheco PMCL. Experimental investigation of vibration
Seismic Regions: The XLAM Case. Struct Eng Int 2008;18(2). reduction using shape memory alloys. J Intel Mat Syst Str 2013;24(2):247–61.
[12] Flatscher, G. and G. Schickhofer, Shaking-table test of a cross-laminated timber [32] Mani Y, Senthilkumar M. Shape memory alloy-based adaptive-passive dynamic
structure. Proceedings of the Institution of Civil Engineers-Structures and Buildings, vibration absorber for vibration control in piping applications. J Vib Control 2015;
2015. 168(11): p. 878-888. 21(9):1838–47.
[13] Kawai N, et al. Full scale shake table tests of five story and three story CLT building [33] Huang, H., W. Xie, and W.S. Chang, Materials characterisation of NiTi and CuAIMn
structures, in World Conference on Timber Engineering. Vienna: Austria; 2016. shape memory alloy bars under dynamic bending, in ASM International - International
[14] Ceccotti A, et al. SOFIE project-3D shaking table test on a seven-storey full-scale Conference on Shape Memory and Superelastic Technologies. 2015, ASM International:
cross-laminated timber building. Earthq Eng Struct Dyn 2013;42(13):2003–21. Oxford, UK.
[15] Poh’sie GH, et al. Application of a Translational Tuned Mass Damper Designed by [34] Flemming L, Mascaro S. Analysis of hybrid electric/thermofluidic inputs for wet
Means of Genetic Algorithms on a Multistory Cross-Laminated Timber Building. shape memory alloy actuators. Smart Mater Struct 2013;22(1):014015.
J Struct Eng 2016;142(4). [35] Huang H. A Temperature Controlled Semi-active Tuned Mass Damper using Shape
[16] Den Hartog JP, Mechanical vibrations.. New York. London: McGraw-Hill; 1956. Memory Alloy for Vibration Reduction Applications. Department of Architecture
[17] Rahimi F, Aghayari R, Samali B. Application of Tuned Mass Dampers for Structural and Civil Engineering. Bath, UK: University of Bath:; 2017.
Vibration Control: A State-of-the-art Review. Civil Engineering Journal-Tehran [36] Huang HY, Zhu YZ, Chang WS. Comparison of Bending Fatigue of NiTi and CuAlMn
2020;6(8):1622–51. Shape Memory Alloy Bars. Adv Mater Sci Eng 2020;2020.
[18] Poh’sie GH, et al. Optimal design of tuned mass dampers for a multi-storey cross [37] Zielinski TP, Duda K. Frequency and Damping Estimation Methods - an Overview.
laminated timber building against seismic loads. Earthq Eng Struct Dyn 2016;45 Metrology and Measurement Systems 2011;18(4):505–28.
(12):1977–95. [38] Zivanovic S, Pavia A. Probabilistic Assessment of Human Response to Footbridge
[19] Qian C, et al. Development of a Vibroacoustic Stochastic Finite Element Prediction Vibration. Journal of low frequency noise, vibration and active control 2009;28(4):
Tool for a CLT Floor. Applied Sciences-Basel 2019;9(6). 255–68.
[20] Huang H, Chang W-S. Application of pre-stressed SMA-based tuned mass damper to [39] Shaw JA, Kyriakides S. Thermomechanical Aspects of Niti. J Mech Phys Solids
a timber floor system. Eng Struct 2018;167:143–50. 1995;43(8):1243–81.
[21] Huang, H. and W.-S. Chang, Reducing footfall-induced vibration in the timber floor [40] Fragiacomo M, Dujic B, Sustersic I. Elastic and ductile design of multi-storey
system using a pre-stressed shape memory alloy-based tuned mass damper, in World crosslam massive wooden buildings under seismic actions. Eng Struct 2011;33(11):
Conference on Timber Engineering 2018: Seoul. 3043–53.
[22] Porteous J, Kermani A. Structural timber design to Eurocode 5. 2nd ed. Chichester: [41] Blass, H.J. and P. Fellmoser, Design of solid wood panels with cross layers, in 8th
Wiley-Blackwell; 2013. World Conference on Timber Engineering 2004: Lahti, Finland.
[23] Huang HY, Mosalam KM, Chang WS. Adaptive tuned mass damper with shape [42] Thompson SA. An overview of nickel-titanium alloys used in dentistry. Int Endod J
memory alloy for seismic application. Eng Struct 2020;223. 2000;33(4):297–310.
[24] Zhang Q, et al. Stability of submarine bi-material pipeline-liner system with novel [43] Butaud P, Foltete E, Ouisse M. Sandwich structures with tunable damping
polyhedral composites subjected to thermal and mechanical loading fields. Mar properties: On the use of Shape Memory Polymer as viscoelastic core. Compos
Struct 2023;90:103424. Struct 2016;153:401–8.
[25] Li Z, et al. Buckling performance of the encased functionally graded porous [44] Tsai H-C, Lin G-C. Optimum tuned-mass dampers for minimizing steady-state
composite liner with polyhedral shapes reinforced by graphene platelets under response of support-excited and damped systems. Earthq Eng Struct Dyn 1993;22
external pressure. Thin-Walled Struct 2023;183:110370. (11):957–73.
[45] Sadek F, et al. A method of estimating the parameters of tuned mass dampers for
seismic applications. Earthq Eng Struct Dyn 1997;26(6):617–35.

14