Download as pdf or txt
Download as pdf or txt
You are on page 1of 12

Original Research

Advances in Structural Engineering


2019, Vol. 22(7) 1579–1590
Structural health–monitoring system Ó The Author(s) 2018
Article reuse guidelines:
for roof structure of the Dalian sagepub.com/journals-permissions
DOI: 10.1177/1369433218819572

gymnasium journals.sagepub.com/home/ase

Hongnan Li1,2, Chaolin Yuan1 , Liang Ren1 and Tao Jiang1

Abstract
The roof of Dalian gymnasium was designed in the form of suspen-dome structure. A structural health–monitoring system has been
developed for the roof structure to guarantee the safety condition during construction process as well as in future service. In this arti-
cle, a monitoring scheme was proposed in detail according to the mechanical characteristics of the roof structure. Fiber Bragg grating
sensors, inclinometers, and accelerometers were applied to measure necessary structural information. In order to interrogate differ-
ent types of sensors, a novel data acquisition device of the structural health–monitoring system was also introduced and has achieved
multitudinous physical variable synchronization acquisition. By analyzing the data obtained during the construction and normal opera-
tion of the gymnasium, the structural health condition was evaluated and the structural damage could subsequently be located.

Keywords
data acquisition device, fiber Bragg grating sensors, finite element method, structural health–monitoring system, suspen-dome
structure

Introduction life-cycle (Cigada et al., 2010; Giurgiutiu and Cuc,


2005; Kang et al., 2003; Montalvao et al., 2006; Wu
Recent years in China, a large amount of large-span and Fujino, 2005).
spatial structures have been built such as gymnasiums In the early eighties of the last century, domestic
and convention centers due to their advantages of and overseas studies about SHM system were mainly
long-span, lightweight, and innovative design (Li focused on applications for bridges (Brownjohn et al.,
et al., 2015). These structures, are supposed to be sub- 1994; Ko and Ni, 2005; Li et al., 2003b, 2008; Pines
jected to human-induced vibrations apart from the and Aktan, 2002; Sumitoro et al., 2001). A total of 40
impacts of natural conditions like earthquakes and fiber Bragg grating (FBG) sensors divided into three
storms (Ren et al., 2016). If the structure has a local arrays were used on the Tsing Ma Bridge in Hong
damage, it is of possibility to cause structural failure Kong as the sensor sub-system to monitor the safety
leading to casualties and property loss. The situation condition of the bridge (Chan et al., 2006). The main
may be even worse when the structure is still under progress and problems that existed in the establishment
construction. Due to the disparity between a finite ele- of the SHM system designed for the Yangtse River
ment (FE) model and a real engineering structure, such Bridge in China were systematically discussed (Li et al.,
as simplification of the boundary conditions, in the 2003a). Meanwhile, as the building technology devel-
majority situations, the simulation results are unable ops, super-tall structures have been constructed all over
to accurately explain the real mechanical behavior of the world which relieves the pressure caused by the
the large-span structure. Thereby data acquired by the shortage of land resources. A complete SHM system
structural health–monitoring (SHM) system consist-
ing of sensor sub-system, data acquisition sub-sys- 1
Faculty of Infrastructure Engineering, Dalian University of Technology,
tem, and data processing sub-system are to some Dalian, China
2
extent of great value as a reference to conduct a fur- Shenyang Jianzhu University, Shenyang, China
ther research and modify the theoretical model (Li
Corresponding author:
and Li, 2002). Therefore, it has now become more Chaolin Yuan, Faculty of Infrastructure Engineering, Dalian University of
and more significant to utilize SHM system on Technology, No.2 Linggong Road, Dalian 116024, China.
the large-span spatial structures during their whole Email: chaolinyuan@mail.dlut.edu.cn
1580 Advances in Structural Engineering 22(7)

has been applied on the 600-m tall Canton tower (Su


et al., 2014; Xia et al., 2014). Data were collected to
analyze and evaluate the structural safety condition.
Accelerometers, displacement gauges, and strain sen-
sors were applied on Shanghai Tower to study struc-
tural behavior during construction (Zhang et al., 2013).
But compared to bridges and super-tall structures,
studies of health monitoring of suspen-dome structure
are relatively less due to the requirements of monitor-
ing multiple parameters such as acceleration, cable
force, and inclination of supports (Li et al., 2011).
During the last decade, large-span spatial structures
were built and corresponding researches have been con-
ducted. Initial practical progress has been made for
Figure 1. Components of the gymnasium roof structure.
grid structures by employing SHM system on Civil
Center in Shenzhen China (Zheng et al., 2007). The
anchorage dynamometers, vibration wire transducers,
other cultural activities held in the gymnasium. The
and vertical displacement gauges were installed on the
overall floor area of Dalian gymnasium is about
roof steel structures of the National Stadium to moni-
81,000 m2 allowing for 18,000 spectator capacity with
tor cable force, member strain, and displacement,
a total height of 41.26 m. The stand of the gymnasium
respectively (Qin et al., 2009). Gauges based on the
is a five-story reinforced concrete frame structure, on
FBG sensing technique and electromagnetic induction
top of which is the roof made of steel structure. The
were applied to monitor the strain of rigid rods and
roof with 145.4 m in span is suspen-dome structure
cable force separately during the steel structure installa-
and shaped like a spheroid. The upper part of the roof
tion of the Jinan Olympic Sports Center (Zhou, 2010).
is truss work, and the bottom part is cable system con-
Generally speaking, the most important part of an
sisting of circumferential cable, radial cable, and struts
SHM system is the data acquisition device which is in
(Figure 1).
charge of not only signal collection but also signal
The structure of the suspen-dome roof was compli-
interrogation. But as mentioned above, there are many
cated. Therefore, the construction was difficult to pro-
types of sensors, and they can be divided into two cate-
ceed with many risks. For safety reasons, SHM system
gories: one category is optic fiber sensor (Glisic and
was designed for the roof structure mainly monitoring
Inaudi, 2003) and the other one is electronic sensor
the stress changes of key components, vibration of the
including accelerometers, inclinometers, and so on. The
whole structure, and deformation of the supports.
problem of the existing SHM system was that the data
Accordingly, the structural response under any circum-
acquisition device was deficient in its signal collection
stance can be analyzed to provide technical supports
ability of both optic fiber sensors and electronic sensors
estimating the safety condition of the gymnasium.
synchronously (Dong, 2009; Zhou et al., 2016).
Therefore, in this work, a new data acquisition device
was developed to solve this problem. Monitoring scheme
This article first briefly introduced the structure
information of the Dalian Gymnasium. Then a moni- Monitoring parameters
toring scheme was proposed in detail for the gymna- The first step of establishing an SHM system is to
sium preventing from structural failure. A new way of define the objects of monitoring. According to the
measuring cable force using FBG sensors was also pre- characteristic of the suspen-dome roof structure of the
sented. After that the hardware and function of the Dalian gymnasium, the measurands of the SHM sys-
data acquisition device were discussed. The construc- tem are listed as a result of the following discussion.
tion progress of pre-tensioning and a load application Strain describes the local state of deformation in
were monitored, the measured results of which have continuous bodies. Moreover, as one of the most
been analyzed and concluded. basic mechanical parameter of a structure, it can make
great contribution to damage localization and load
identification.
General information of the gymnasium
Due to the long span of the roof structure, from the-
Dalian gymnasium and its ancillary facility are the oretical point of view, the deflection in the mid span is
most important projects of the Dalian sports center. supposed to be large. This would result in the inclina-
Until now, there have been more than 10 concerts and tion changes of the supports.
Li et al. 1581

Figure 2. Sensor locations.

Ambient vibration is of crucial importance to deter-


mine the dynamic property of a structure. In addition,
acceleration measurements are quite common in the
assessment and control of structural dynamics.
To determine the locations of sensors, a FE model
is established. Therefore, specific monitoring scheme is
presented as follows:

 Strain change of the cable;


 Strain change of the struts;
 Strain change of critical members;
 Inclination of the supports; Figure 3. Installation diagram of fiber Bragg grating sensors.
 Local vibration of the structure.

lB = 2nL ð1Þ
Monitoring scheme and sensor disposition
More than 200 sensors were installed on the roof struc- The environmental change induces the wavelength
ture of the gymnasium to monitor various parameters change of reflected light. Hence, the strain can be pre-
including FBG sensors, inclinometers and acceler- sented by
ometers (Figure 2).
e = KDlB ð2Þ

Stress measurement of chord members. Based on the FBG where e is the strain, K is the sensitivity coefficient of
sensing technology, fiber optic sensors were designed sensors, and DlB is the wavelength change.
and developed to monitor various physical quantities The cross-section of the truss work was in the shape
such as strain and temperature with the advantages of of an inverted triangle. The strain variation of chord
light weight, high accuracy, and high resistance to elec- members was monitored by FBG sensors (Figure 3).
tromagnetic interference. Due to concerns of a cross sensitivity between strain
Light propagates through the gratings with negligible and temperature, an FBG temperature sensor was
signal variation. Only those wavelengths that satisfy the mounted by the side of each FBG strain sensor for
Bragg condition are affected and strongly back-reflected. temperature compensation. The Young’s modulus of
The center wavelength of the reflected light lB is decided the members is 210 GPa, on the basis of which the
by the grating period L and the effective index n stress of the members can be then determined.
1582 Advances in Structural Engineering 22(7)

Measurement of support inclination. Under certain condi-


tions, small displacement and inclination are allowed
for the supports of the gymnasium to release the vibra-
tion energy. Displacement and inclination of the sup-
ports change along with the deformation or vibration
of the structure. Uniaxial inclinometer using current
output module was adopted to monitor the inclination
of the supports
 
Iout  Io
a = arcsin ð3Þ
K

where a is the inclination, Iout is the output current, Io Figure 4. Cable system of the roof structure.
is the output current of the initial status, and K is the
sensitivity of the sensor.
The relation between cable force DF and wavelength
of the sensor can be obtained according to equation (1)
Vibration measurement. The modal parameters were
and FBG sensor working principle (Li and Ren, 2008;
acquired by the accelerometers installed on the roof
Yi et al., 2011)
structure of the gymnasium to assist analyzing the
structural dynamic characteristics which could predict
DF = GDlB ð4Þ
the dynamic response.
where G is the sensibility coefficient of the cable force,
and DlB is the center wavelength change of FBG
Monitoring method of the cable system and measure point
sensor.
disposition. Due to the large span of the roof structure,
After the FBG sensor has been installed on the
cable system consisting of circumferential cables, radial
anchor, the cable was tensioned in the manufacturing
cables, and struts were built beneath the truss to reduce
factory. The sensibility coefficient of the cable force
the internal force and deflection of the central area
was 5.34 kN/nm calibrated through the tension test.
(Figure 4). The cable system was important to the
The sensitivity coefficient was also closely related
whole structure. In order to enhance the stiffness of
with the install location of FBG sensors because the
the cable system, proper pre-stress was applied so that
inappropriate location of the sensor would lead to
the structure would have the ability to undertake exter-
measurement error. The von Mises stress nephogram
nal load and support the truss structure. From the
of anchor is given in Figure 7 simulated by FE analysis
above, monitoring the pre-tension process of the cable
software ANSYS. In order to get large sensitivity coef-
system was necessary for safety construction.
ficient, sensors were installed on the pivot area where
The whole pre-tension process was divided into four
local stress concentration exists.
stages as follows. Stage 1 applied 10% of the designed
value on all of the radial cable which just straightened
the radial cable from relaxing status. As a result, there Data acquisition device of the SHM
was no strain changes monitored during this stage. system
Stages 2, 3, and 4 applied 50%, 70%, and 105% of the
designed value on the radial cable, respectively. An FE In the field of SHM, various kinds of new sensors such
model was built to simulate the last three stages of the as FBG sensors and electronic sensors have been devel-
pre-tension process. Figure 5 shows the cable force dis- oped to satisfy the requirement of sensing a quantity of
tribution in each stage. Based on the finite element physical parameters (Park et al., 2006). A special
method (FEM) results, the monitoring scheme for designed data acquisition device was in great demand
cable system is presented in Figure 6. to acquire signals from polytypic sensors to implement
In order to monitor the cable force, FBG sensors the function of data processing and structural safety
were installed on the anchor instead of the cable con- evaluation. In order to take the ability of SHM for
sidering the fact that the cable consisted of steel strand large-span structures, it was required that the data
and an outside steel tube. There existed a linear rela- acquisition device should be specified with the follow-
tionship between the strain of the anchor and that of ing goals according to the architectural characteristics:
the cable within the elastic range. Using epoxy, the
FBG sensor was glued on the anchor to monitor the 1. Due to the complexity of the large-span struc-
change of cable force. tures, parameters such as stress, acceleration,
Li et al. 1583

Figure 5. Cable force distribution during the pre-stress process.

Figure 6. Monitoring scheme for cable system.


1584 Advances in Structural Engineering 22(7)

3. For the harsh engineering environment, the


SHM system needed to keep regular work effi-
ciency, stability, and durability to fulfill the
long-term task from the construction stage.

As mentioned above, there needed to be three dif-


ferent types of sensors to get the safety status of the
roof structure. However, the fiber optic sensors signal
and the electric signal were usually incompatible in
one data acquisition device. Not to mention, it was
required to apply the SHM system on the structure
from the beginning of the construction with all kinds
of difficulties to overcome such as power supply. The
general solution of this problem now is the simulta-
neous use of more than one data acquisition unit in
order to collect all the desired measurements.
Figure 7. Finite element analysis for the anchor and FBG
sensors installation. According to these requirements in actual engineer-
ing projects, a data acquisition device (Figure 8) has
inclination, and deformation needed to be been researched and designed, which realizes the high
obtained with the aid of different types of level of synchronous integration of the signals of FBG
sensors. sensors and electronic sensors which include current,
2. Usually, the large-span structures were gigantic voltage, and so on for the roof structure of the Dalian
which required numerous sensors to cover the gymnasium. Figure 9 illustrates the working mechan-
key members. In this project, there were in total ism of this device. Several modules collected the signal
200 FBG sensors and 90 electronic sensors of optic sensors and electronic sensors. The sampling
including inclinometers and accelerometers. frequency of acceleration was usually much higher
More importantly, the data acquisition device than those of strain and inclination. The sampling fre-
had to guarantee the synchronization of data quency of acceleration was 512 Hz to capture the
acquisition and data sampling of so many structural dynamic response. But only meaningful
sensors. vibration information was recorded in order to save
storage space. Therefore, redundant information was

Figure 8. The picture of the sensing channels of the data acquisition device.

Figure 9. The principle of the data acquisition device.


Li et al. 1585

Table 1. The specific parameter of the data acquisition device.

Resolution Accuracy Protocol Bandwidth FBG channel Electronic sensor channel

1 pm 62 me TCP/IP 80 nm 15 48

FBG: fiber Bragg grating.

Figure 10. The picture of the demodulation module for electronic sensors.

filtered such as ambient noise. The strain and inclina- is shown in Figure 10. The integrated circuit
tion mainly depict the quasi-static process, and resam- board supported the demodulation for bridge
pling algorithm was adopted to decrease the sampling circuit, IEPE signal, voltage signal, and current
rate to 20 Hz for strain and inclination measurements. signal. The adjacent channel isolation could
Within each module, there was a real-time clock creat- reach up to 50 dB, and this module had the
ing a time-stamp. When data from multiple acquisition hardware-level filtering and noise reduction.
sources were to be combined, a universal timer with 3. In this device, the built-in modules acquiring
built-in 40 MHz clock installed in the field program- signals of FBG sensors and electronic sensors,
mable gate array (FPGA) accurately synchronized all respectively, were controlled by FPGA. Unlike
the individual clocks in the measurement modules. processors, FPGAs were truly parallel in
Time synchronization was 25 ns. Afterwards, the nature, so different processing operations did
obtained data were transferred to a buffer waiting for not have to compete for the same resources.
instructions from a real-time controller. The transmis- Each independent processing task, such as data
sion control protocol/Internet protocol enables the acquisition of acceleration and strain, was
data acquisition device to communicate with the calcu- assigned to a dedicated section of the chip,
lation terminal. At last, the monitoring results were and can function autonomously without any
processed to evaluate the structural safety condition. influence from other logic blocks. As a result,
In Table 1, specific parameter of the data acquisi- the performance of one part of the application
tion device is given. The characteristics of this device was not affected when more processing was
are summarized in detail as follows: added.
4. The real-time controller was embedded in this
1. In this project, a swept laser source with higher device with industrial-scale stability, in which
power and lower noise instead of the broad- equipped with an industrial processor in charge
band source enabled the data acquisition device of multi-rate management for data acquisition
to lengthen the sensing distance up to 20 km, and data transmission from FPGA to computing
interrogating the FBG signal. Meanwhile, the terminal through Transmission Control (TCP)/
measurable wavelength range was 80 nm from Internet Protocol (IP) protocol for further data
1510 to 1590 nm. Every channel of the 15 FBG processing and structural safety assessment.
channels can measure up to 100 sensors con-
nected in series in one single fiber. The swept In further data processing on the calculation termi-
laser source possessed the advantage of better nal, the criterion for structural safety assessment was
repeatability and accuracy without the need for compiled in the program. For strain measurement of
periodic recalibration. critical members, the safety evaluation was done by
2. As for the electronic sensors (current signal, comparison of the monitoring results with the steel
voltage signal, etc.), the demodulation module material strength to ensure the loading capacity. With
1586 Advances in Structural Engineering 22(7)

Figure 12. Stress changes of struts at each stage.

Figure 11. Percentages of cable force monitored at each stage contribute to the designed value, as it can be seen from
that contribute to the designed value. which that all the cable forces fail to reach 50% of the
designed value at stage 2. Considering the friction loss
regard to cable force measurement, the system com- and interaction in the cable system, it was normal and
pared the measured loads with the designed value. acceptable, and that was the reason for applying 105%
of the designed value at stage 4. However, when the pre-
tension construction was completed, sensor installed on
Analysis of the monitoring results
cable 20 showed the cable force was about 100 kN less
Analysis of the monitoring results of cable system than the designed value. Thorough inspections have
during the pre-tension construction process been carried out for the cable. This situation could be
probably caused by manual operation error due to the
Proper pre-stress for the cable system could enlarge
fact that the tension equipment was manually operated.
the stiffness of the cable system such that the structure
Hence compensation was made for cable 20 to reach the
possesses enough carrying capacity. Radial cable was
designed value. The error between monitored cable force
tensioned after the pre-tension construction of circum-
and target value at each stage was less than 10%.
ferential cable finished. Considering the fact that there
was a limited number of tension equipment, each stage
was carried out ring by ring to ensure the uniformity Monitoring results of struts. Stress changes of struts at
of the tension distribution. The tension construction each stage during the pre-tension construction process
lasted for 9 days in the process of which the SHM sys- are illustrated in Figure 12. The absolute maximum
tem was used to monitor the cable force. Parts of the stress change was 29.748 MPa less than the yield
monitoring results are listed as follows. strength 310 MPa.
Each end of the strut was equipped with one FBG
Monitoring results of radial cable. Figure 11 shows the strain sensor. The wavelength time-history of Z1-1
percentages of cable force monitored at each stage that strut in the second stage of pre-tension process is given

Figure 13. Stress time-history of Z1-1 strut in second stage of cable tensioning.
Li et al. 1587

Monitoring results of vibration. During the construction


process, the SHM system sent out an alert caused by
an unexpected blasting near the construction site
which was captured by the accelerometers. In Figure
15, the signal is analyzed in both time domain and
frequency domain. The peak acceleration of the
vibration was 0.17g, and the maximum response fre-
quency was around 15 Hz. Due to insufficient infor-
mation of the excitation source, it could only be
assumed that either the frequency of the local struc-
ture or the excitation source was 15 Hz. It indicated
that the SHM system could reflect the vibration
Figure 14. Inclination time-history of No. 34 support in pre- information of the structure.
tension process. The fundamental frequency calculated by the FE
model was 1.92 Hz. By analyzing the structural
responses to ambient vibration, the identified funda-
in Figure 13. The struts were designed to be a two- mental frequency was 2.5 Hz. The reason was that
force rod, meaning they should only be subjected to after the construction, the second ring of the truss sys-
axial load. However, the wavelength changes of these tem was reinforced once. The nodal stiffness of the sec-
two sensors indicated that one end of strut Z1-1 was in ond ring was revised in the FE model, and the
tension, while the other one was in compression. fundamental frequency was 2.547 Hz agreeing well
Hence, it could be deduced that the strut suffered from with the monitored result.
both axial load and moment at the same time, which
was contrary to the assumption. Generally speaking, if
a two-force rod was subjected to a bending moment, it
Analysis of the monitoring results after construction
reduced the loading capacity of the member and buck- has been completed
ling could probably occur. According to this situation, The monitoring system has been switched on from the
the theoretical model has been modified to take beginning of pre-tension construction. Part of the mon-
moment into consideration for stress redistribution in itoring results acquired through analyzing the monitor-
strut analysis. ing data is as follows after the construction process has
been completed.
Monitoring results of inclination. When the radial cable
was tensioned causing that the truss structure raised Long-term monitoring results. The FBG sensors were
up. The maximum inclination change during these mounted on the key chord members in order to get
9 days of No. 34 support was 0.914°, which satisfied structural health information in real time. Figure 16
the inclination limit 1.5° required by theoretical model shows the strain variation of parts of the chord mem-
analysis (Figure 14). bers for 77 days from October to early January. There

Figure 15. Vibration monitoring results of accelerometer AB1.


1588 Advances in Structural Engineering 22(7)

model display method showed the distribution of


strain on the whole roof structure, indicating that the
load application only affected the adjacent structures
and the entire roof was still under safe condition.
In Figure 18, the strain time-history is given out of
all the three stages, respectively. At the first stage, it
could be seen that in the end, the strain was restored to
initial status due to the fact that after the first 20 tons
of weight has been applied, the truss frame descended
to the ground again. During this period, the maximum
stress change was 6.5 MPa. The added truss structure
made small influence on the roof.

Conclusion
Figure 16. Strain variation from October to January. The measurement system has been successfully operat-
ing for nearly 2 years, during which the system has
was nearly no strain changes for the first 25 days. been updated once. The obtained data was used to
Afterwards, the local temperature began to decline gra- evaluate the safe condition of the roof structure.
dually leading to the contraction of the steel structure. Conclusions are as follows:
This explained the increase in compressive strain in this
period. From day 53, the gymnasium suffered from a 1. The data acquisition device of the SHM sys-
strong wind lasting for 5 days, as a result of which the tem applied on roof structure of the Dalian
strain of most members suddenly decreased. During gymnasium could measure signals from vari-
this period, the maximum stress change was ous types of sensors such as optical fiber sen-
14.54 MPa, indicating that the structure was still under sors and electronic sensors synchronously in
safe condition. real time.
2. Data have been stored for the stress time-
history of key members. During the pre-tension
Monitoring results of load application on roof construction process, the cable system was ana-
structure. During the operation period, an extra truss lyzed. The difference between monitored cable
structure was added (Figure 17). Roof structure force and designed value at each stage was less
designed primarily for distributed loads must be than 10%.
checked locally for point loads. So load applications 3. During the operation period, the truss work
were carried out on the roof structure divided into was monitored for load application and long-
three stages: 20, 35, and 40 tons of weight. These loads term monitoring results were also analyzed.
were applied separately on the truss structure hoisted The results indicated that the structure was
on the roof structure. The three-dimensional (3D) under safe condition.

Figure 17. Truss frame and strain distribution displayed in 3D model.


Li et al. 1589

detection, and failure prevention. Shock and Vibration


Digest 37(2): 505–511.
Glisic B and Inaudi D (2003) Integration of long-gage fiber-
optic sensor into a fiber-reinforced composite sensing
tape. In: Proceedings of SPIE—The international society
for optical engineering, San Diego, CA, 22 July.
Kang W, Chen Z, Lam HF, et al. (2003) Analysis and design
of the general and outmost-ring stiffened suspen-dome
structures. Engineering Structures 25(13): 1685–1695.
Ko JM and Ni YQ (2005) Technology developments in struc-
tural health monitoring of large-scale bridges. Engineering
Structures 27(12): 1715–1725.
Li A, Miao C, Li Z, et al. (2003a) Health monitoring system
for the Runyang Yangtse River Bridge. Journal of South-
Figure 18. Monitoring results of load application of all three east University (Natural Science Edition) 33(5): 544–548.
stages. Li DS, Zhang Y and Ren L (2011) Sensor deployment for
structural health monitoring and evaluation. Advances in
Mechanics 41(1): 39–50.
Acknowledgements Li HN and Li DS (2002) Safety assessment, health monitor-
The acquisition modules in the data acquisition device devel- ing and damage diagnosis for structures in civil engineer-
oped in this work was manufactured by National ing. Earthquake Engineering and Engineering Vibration
Instruments. 22(3): 82–90.
Li HN and Ren L (2008) Structural Health Monitoring Fiber
Optic Sensing Technology. Beijing, China: China Architec-
Declaration of Conflicting Interests ture & Building Press.
The author(s) declared no potential conflicts of interest with Li HN, Gao DW and Yi TH (2008) Advances in structural
respect to the research, authorship, and/or publication of this health monitoring system in civil engineering. Advances in
article. Mechanics 38(2): 151–166.
Li HN, Ren L, Jia ZG, et al. (2015) State-of-the-art in struc-
tural health monitoring of large and complex civil infra-
Funding
structures. Journal of Civil Structural Health Monitoring
The author(s) received no financial support for the research, 6(1): 3–16.
authorship, and/or publication of this article. Li Z, Li A, Tommy CH, et al. (2003b) Finite element model-
ing for health monitoring and condition assessment of
long-span bridges. Journal of Southeast University (Natu-
ORCID iD
ral Science Edition) 33(5): 562–572.
Chaolin Yuan https://orcid.org/0000-0001-5879-5211 Montalvao D, Maia NMM and Ribeiro AMR (2006) A
review of vibration-based structural health monitoring
with special emphasis on composite materials. Shock and
References
Vibration Digest 38(4): 295–326.
Brownjohn JMW, Bocciolone M, Curami A, et al. (1994) Park G, Farrar CR, Scalea FL, et al. (2006) Performance
Humber Bridge full-scale measurement campaigns 1990– assessment and validation of piezoelectric active-sensors
1991. Journal of Wind Engineering and Industrial Aerody- in structural health monitoring. Smart Materials and
namics 52(94): 185–218. Structures 15(6): 1673–1683.
Chan THT, Yu L, Tam HY, et al. (2006) Fiber Bragg grat- Pines D and Aktan AE (2002) Status of structural health
ing sensors for structural health monitoring of Tsing Ma monitoring of long-span bridges in the United States.
bridge: background and experimental observation. Engi- Progress in Structural Engineering and Materials 4(4):
neering Structures 28(5): 648–659. 372–380.
Cigada A, Moschioni G, Vanali M, et al. (2010) The mea- Qin J, Xu R and Xu Y (2009) Research on health monitoring
surement network of the San Siro Meazza stadium in system in national indoor gymnasium. Construction Tech-
Milan: origin and implementation of a new data acquisi- nology 38(3): 40–43.
tion strategy for structural health monitoring. Experimen- Ren L, Yuan CL, Li HN, et al. (2016) Structural health mon-
tal Techniques 34(1): 70–81. itoring system developed for Dalian stadium. International
Dong X (2009) Study on structural dynamic model updating and Journal of Structural Stability and Dynamics 16(4): 70–81.
damage detecting based on sensitivity analysis. PhD Thesis, Su JZ, Xia Y, Xu YL, et al. (2014) Settlement monitoring of
Lanzhou University of Technology, Lanzhou, China. a supertall building using the Kalman filtering technique
Giurgiutiu V and Cuc A (2005) Embedded non-destructive and forward construction stage analysis. Advances in
evaluation for structural health monitoring, damage Structural Engineering 17(6): 881–893.
1590 Advances in Structural Engineering 22(7)

Sumitoro S, Matsui Y, Kono M, et al. (2001) Long span Zhang QL, Lv J, Li H, et al. (2013) Numerical analysis and
bridge health monitoring system in Japan. Health Moni- monitoring research for construction progress of Shang-
toring and Management of Civil Infrastructure Systems hai Tower. Building Technique 42(15): 9–12.
4337: 517–524. Zheng J, Qu WL and Zhang QL (2007) Structural health
Wu Z and Fujino Y (2005) Structural health monitoring and monitoring for the steel bracket of Shenzhen Civil Center.
intelligent infrastructure. Smart Materials and Structures Building Structure 37(3): 57–60.
14(3): 448–464. Zhou GD, Yi TH and Chen B (2016) Innovative design of a
Xia Y, Zhang P, Ni YQ, et al. (2014) Deformation monitor- health monitoring system and its implementation in a
ing of a super-tall structure using real-time strain data. complicated long-span arch bridge. Journal of Aerospace
Engineering Structures 67: 29–38. Engineering 30(2): B4016006.
Yi TH, Li HN and Gu M (2011) Multiple optimization stra- Zhou X (2010) Structural design and health monitoring of
tegies based sensor placement method for structural Jinan Olympic Sports Center. Engineering Mechanics
health monitoring. Journal of Building Structures 32(12): 27(2): 105–113.
217–223.

You might also like