sensors

Communication
Real-Time Structural Health Monitoring and Damage
Identification Using Frequency Response Functions along with
Finite Element Model Updating Technique
Tarunpreet Singh 1 , Shankar Sehgal 1, * , Chander Prakash 2 and Saurav Dixit 3,4, *

1 University Institute of Engineering and Technology, Panjab University, Chandigarh 160014, India;
[email protected]
2 School of Mechanical Engineering, Lovely Professional University, Phagwara 144411, India;
[email protected]
3 Peter the Great St. Petersburg Polytechnic University, 195251 Saint Petersburg, Russia
4 Division of Research & Innovation, Uttaranchal University, Dehradun 248007, India
* Correspondence: [email protected] (S.S.); [email protected] or [email protected] (S.D.)

Abstract: Throughout service, damage can arise in the structure of buildings; hence, their dynamic
testing becomes essential to verify that such buildings possess sufficient strength to withstand
disturbances, particularly in the event of an earthquake. Dynamic testing, being uneconomical,
requires proof of concept; for this, a model of a structure can be dynamically tested, and the results are
used to update its finite element model. This can be used for damage detection in the prototype and
aids in predicting its behavior during an earthquake. In this instance, a wireless MEMS accelerometer
was used, which can measure the vibration signals emanating from the building and transfer these
signals to a remote workstation. The base of the structure is excited using a shaking table to induce
an earthquake-like situation. Four natural frequencies have been considered and six different types of
Citation: Singh, T.; Sehgal, S.;
damage conditions have been identified in this work. For each damage condition, the experimental
Prakash, C.; Dixit, S. Real-Time
responses are measured and the finite element model is updated using the Berman and Nagy method.
Structural Health Monitoring and
It is seen that the updated models can predict the dynamic responses of the building accurately. Thus,
Damage Identification Using
depending on these responses, the damage condition can be identified by using the updated finite
Frequency Response Functions along
with Finite Element Model Updating
element models.
Technique. Sensors 2022, 22, 4546.
https://doi.org/10.3390/s22124546 Keywords: finite element; model updating; structural dynamics; structural health monitoring

Academic Editor: Giuseppe

Lacidogna

Received: 13 May 2022 1. Introduction

Accepted: 14 June 2022 Over the course of time, several instances of damage can occur in an engineering
Published: 16 June 2022 structure. Condition monitoring and damage identification of engineered structures are
Publisher’s Note: MDPI stays neutral vital for the longevity of the structures, as well as in reducing the overall maintenance
with regard to jurisdictional claims in cost [1]. Technological advancements in the discipline of structural health monitoring
published maps and institutional affil- (SHM) have been growing rapidly in the last two decades. SHM has been employed in the
iations. monitoring of almost every engineered structure, from rotating machinery [2,3] to bridges
and buildings [4,5]. This technique includes various damage detection methods, namely
vibration-based sensors, embedded sensors, acoustic emissions, lamb wave method, and
comparative vacuum monitoring [6]. Humans and the environment largely influence the
Copyright: © 2022 by the authors. health of these structures and, with time, damage can arise in these structures which can be
Licensee MDPI, Basel, Switzerland. attributed to aging, inadequate service conditions, and erroneous manufacturing.
This article is an open access article
Cawley et al. [7] proposed studying the natural frequencies of a structure to identify
distributed under the terms and
and locate damage. They also studied the growth of damage using the proposed tech-
conditions of the Creative Commons
nique [8]. Kessler et al. [9] compared the dynamic responses of the control and damaged
Attribution (CC BY) license (https://
specimens, and postulated that the reduction in the frequency response was attributed to
creativecommons.org/licenses/by/
the presence of damage.
4.0/).

Sensors 2022, 22, 4546. https://doi.org/10.3390/s22124546 https://www.mdpi.com/journal/sensors

Sensors 2022, 22, 4546 2 of 14

Many techniques of SHM have been comprehensively researched and used for dam-
age detection, and it has been established that vibration-based damage detection is a
well-recognized and uncomplicated technique to study the dynamic characteristics of an
engineered structure. Various researchers have utilized a wide variety of methods to ana-
lyze them. The dynamic characteristics include the damping ratio, natural frequencies, and
mode shapes, which are a function of mass and stiffness distribution of the structure. Any
variation in the mass and stiffness distribution, due to any external influence, will result in a
deviation of the dynamic characteristics from the control structure. A direct correlation has
been acknowledged between damage and a decrease in the natural frequency of the system.
Intelligent monitoring techniques have been used for damage localization and quantifi-
cation in numerous operational bridges [10], buildings [11], and aerospace structures [12],
by identifying the deviation from the optimal working conditions and determining remain-
ing life. Mishra [13] systematically reviewed the advantages of combining test data with
machine learning for structural health monitoring and damage prognosis, to ensure the
longevity of heritage buildings. Gopinath et al. [14] also reviewed the various long-term
and short-term techniques for damage identification and localization in heritage struc-
tures. Numerous researchers have also focused their study on the condition monitoring of
heritage structures [15,16].
Kessler et al. [9] compared the dynamic responses of control and damaged spec-
imens, and postulated that the reduction in the frequency response was attributed to
damage. The loss of stiffness assisted in the detection and localization of the damage [17].
Fan et al. [18] compared various vibration-based methods for damage detection in compos-
ite materials. It was proposed that the natural frequency-based methods could localize
the damages, whereas curvature and mode shape-based methods called for optimization
algorithms for localization. Yan et al. [19] also proposed the effectiveness of combining
conventional vibrational theory with other methods, such as artificial intelligence, control
theory, and signal processing, etc., to increase the accuracy of vibration-based structural
health monitoring techniques.
Alavi et al. [20] employed finite element modelling and a probabilistic neural network
approach. They interpreted data from a wireless sensor and identified damage in a simply
supported beam in the complex case of bridge gusset plate. Tran-Ngoc et al. [21] proposed a
hybrid metaheuristic algorithm approach to overcome the limitations of genetic algorithms,
and used an improved cuckoo search technique for solving optimization issues and damage
detection in a bridge structure. Hsu et al. [22] studied the vibration response of a building
structure using a wireless sensor for damage localization and quantification in a building
structure. Feng et al. [23] compared transmissibility function and cross-correlation analysis
for damage detection in a tunnel structure, and proposed using both approaches for
validation of the results.
The major novelty of this work is applying the combined use of the modified Berman
and Nagy direct method of finite element model updating with wireless sensors in real-
time health monitoring along with subsequent damage identification under the event of an
earthquake or seismic disturbance by studying the change in frequency response functions
(FRFs), which has not been explored previously under the aforementioned combinations.
This paper will focus on designing and developing a real-time monitoring and subse-
quent damage identification technique, using the structural frequency response and finite
element model updating by exciting a scaled-down two-story building model on a vibrating
table. This proposed technique will not only apply to buildings, but to every structure for
which a finite element model can be prepared and a vibrational analysis performed.

2. Analytical Methodology
2.1. Finite Element Method (FEM)
FEM is the most functional and acknowledged method used to analyze the perfor-
mance of the building model and for finding the approximate solutions for field value
problems in engineering [24,25]. The FEM model of the two-story building was generated
Sensors 2022, 22, 4546 3 of 14

using 34 2D frame elements with two translational degrees of freedom in the x and y
directions, and one rotational degree of freedom, to model in-plane dynamics.
The finite element modelling of the structure can be subcategorized into three cate-
gories: pre-processing, processing, and post-processing.

2.1.1. Pre-Processing
In this section, the input file created by the user defines the material properties, such
as E, I, ρ, A, and L. The model structure, i.e., the coordinates of each element in the frame
and its connectivity with other elements, is established. The degree of freedom of the nodes
and initial matrices for an element are described by the user for processing in MATLAB.

2.1.2. Processing
In the processing phase, the input given by the user is processed by MATLAB to
fabricate the global mass and stiffness matrices, and boundary conditions are applied
and solved for producing eigenvalues and eigenvectors. Symmetric global stiffness and
mass matrices of size n produce n eigenvalues λ and corresponding eigenvectors φ, which
satisfies equation [26].    
K λi = λi M φi

2.1.3. Post-Processing
In this phase, the eigenvalues and the eigenvectors are utilized for producing the
natural frequencies and mode shapes of the building frame under free-vibration conditions.
The following operation is performed on the eigenvalue matrix for obtaining the natural
frequency in rad/sec or Hz. q
[ω ] = [λ]
[ω ]
[f] =
2π
The omega matrix [ω] will provide the natural frequency along its diagonal in rad/sec
and [ f ] will give the natural frequency of the model in Hz.

2.2. Modal Analysis

Asymmetric global stiffness and mass matrices were assembled, which produced
eigenvalues and corresponding eigenvectors for yielding the natural frequencies and mode
shapes of the building frame under free-vibration conditions. The first 4 natural frequencies
of the analytical building model are listed in Table 1. Figure 1 illustrates the undeformed
structure and the first 4 eigenmodes of the building model plotted in MATLAB.

Table 1. Analytical FEM based first 4 natural frequencies in Hertz.

Mode Number 1 2 3 4
Analytical
natural 90.357 325.91 381.84 455.27
frequency (Hz)

2.3. Finite Element Model Updating

Inaccuracies in FE models can be attributed to errors such as: false material properties
of constant values, inferior quality mesh, improper modeling of complex shapes and joints,
and simplification and rounding off in computation [27]. The technique used to correct
or update the analytical model (so that it can predict the dynamic responses accurately
consistent with the experimental result) is known as finite model updating (FEMU) [28].
This technique can be broadly categorized into two categories: direct methods, such as the
Baruch and Bar-ltzhack [29] and Berman and Nagy [30] methods, and indirect methods,
such as the sensitivity method [28,31]. FEMU techniques have been reviewed [27,32,33]
Sensors 2022, 22, 4546 4 of 14

and
Sensors 2022, 22, x FOR PEER REVIEWthe results show that FEMU should be combined with structural health monitoring
4 of 14for
updating vibrating-based FE models for damage localization and quantification, for the
maintenance of the structure and the prognosis of remaining life [34,35].

Figure 1. (A) Undeformed frame structure (all dimensions are in mm), (B–E) first four eigenmodes.
Figure 1. (A) Undeformed frame structure (all dimensions are in mm), (B–E) first four
eigenmodes.
In the Berman and Nagy method, both the mass and stiffness matrices were updated
using the eigenvector matrix φ that was updated as shown
Table 1. Analytical FEM based first 4 natural frequencies in Hertz.
h i −1
Mode Number T 1
φ = φm φm Ma φm
2 2 3 4
Analytical natural frequency (Hz) 90.357 325.91 381.84 455.27
T
Ma = φm Ma φm
2.3. Finite Element Model Updating
−1

−1 T
Mu = M
Inaccuracies in FE models a+
can beM a φm Ma to
attributed I− Ma such
errors Ma φas: mMfalse
a material properties
of
and constant values, inferior quality mesh, improper modeling of complex shapes and
joints, and K simplification and
T rounding off
T in computation
T [27].
T The technique
T used to
u = K a − K a φφ Ma + Ma φφ K a + Ma φφ K a φφ Ma + Ma φλφ Ma
correct or update the analytical model (so that it can predict the dynamic responses accu-
ratelyInconsistent
this case,with
φ = corrected eigenvector
the experimental result)matrix, φm =asexperimental
is known finite model eigenvector matrix.
updating (FEMU)
M
[28]. This = analytical mass matrix, M
a technique can be broadly categorized
u = updated mass matrix.
into two categories: direct methods, such
as theKBaruch
a = analytical
and Bar-ltzhack [29] andKBerman
stiffness matrix, u = updated mass[30]
and Nagy matrix.
methods, and indirect meth-
ods, In thisas
such study,
the the Berman and
sensitivity Nagy[28,31].
method methodFEMU was modified by updating
techniques the eigenvector
have been reviewed
matrix φ using
[27,32,33] the results
and the structure’s
showfrequency
that FEMU response.
should be combined with structural health
monitoring for updating vibrating-based FE models for damage localization and quanti-
3. Experimental
fication, Methodology
for the maintenance of the structure and the prognosis of remaining life [34,35].
Mild
In the steel
Bermanwasand
selected
Nagy as the workpiece
method, for preparing
both the mass the building
and stiffness model
matrices were and was
updated
securely mounted on the vibrating table. The
using the eigenvector matrix ϕ that was updated as shown vibrating table provided continuous and
constant vibration stimuli which acted as an excitation−1mechanism for the modal testing of
𝑇
the structure. The vibrational analysis 𝜙 =of𝜙the𝑚 [𝜙building
𝑚 𝑀𝑎 𝜙𝑚 ] model
2 was performed by employing

̅̅̅̅ 𝑇
𝑀𝑎 = 𝜙𝑚 𝑀𝑎 𝜙𝑚
 3. Experimental Methodology
Mild steel was selected as the workpiece for preparing the building model and was
securely mounted on the vibrating table. The vibrating table provided continuous and
Sensors 2022, 22, 4546 5 of 14
constant vibration stimuli which acted as an excitation mechanism for the modal testing
of the structure. The vibrational analysis of the building model was performed by em-
ploying a wireless MEMS accelerometer sensor. The acceleration measured by the sensor
a wireless MEMS accelerometer sensor. The acceleration measured by the sensor in the 3
in the 3 axes is in the axes
formis of voltage
in the and
form of converted
voltage to a value
and converted of ‘g’.
to a value TheThe
of ‘g’. entire
entireexperi-
experimental setup
mental setup is shownisinshown
Figure 2.
in Figure 2.

Figure 2. Experimental Setup.

Figure 2. Experimental Setup.
A BeanDevice Wilow AX-3DS was employed for measuring the vibrations of the
A BeanDevice Wilow AX-3DS
two-story was employed
building for measuring
model. A BeanDevice Wilow theAX-3DS
vibrations of the two-
is a micro-electromechanical
story building model. system
A BeanDevice Wilow AX-3DS is a micro-electromechanical
IoT based sensor, shown in Figure 3. It is an ultra-low-power,system
Wi-Fi enabled, tri-
axial accelerometer dedicated to condition monitoring, vibrational analysis, and structural
health monitoring. It has a measuring range of ±16 g with a maximum sampling of 1.6 k
samples per second per axis. It integrates a 2.4 GHz antenna diversity, which provides
a maximum wireless range of 200 m and a sensor frequency response (−3 dB) of DC to
800 Hz.
 IoT based sensor, shown in Figure 3. It is an ultra-low-power, Wi-Fi e
celerometer dedicated to condition monitoring, vibrational analysis, an
Sensors 2022, 22, x FOR PEER REVIEW 6 of 14
monitoring. It has a measuring range of ±16 g with a maximum sampli
per second per axis. It integrates a 2.4 GHz antenna diversity, which pr
Sensors 2022, 22, 4546 6 of 14
IoT based sensor,
wireless range shown in Figure
of 200 m and3. It is
a an ultra-low-power,
sensor frequencyWi-Fi enabled, tri-axial
response (−3 dB) ac- of DC
celerometer dedicated to condition monitoring, vibrational analysis, and structural health
monitoring. It has a measuring range of ±16 g with a maximum sampling of 1.6 k samples
per second per axis. It integrates a 2.4 GHz antenna diversity, which provides a maximum
wireless range of 200 m and a sensor frequency response (−3 dB) of DC to 800 Hz.

Figure 3. BeanDevice Wilow AX-3DS, MEMS Accelerometer.

Figure 3. BeanDevice Wilow AX-3DS, MEMS Accelerometer.

4. Experimental Results
Figure 3. BeanDevice Wilow AX-3DS, MEMS Accelerometer.
4. Experimental
Once the true Results
experimental setup was established and the accelerometer was wire-
lesslyOnce
connected
the trueto the personalsetup
experimental computer, the vibrating
was established table
and the was operated
accelerometer wastowirelessly
supply a
4. Experimental
constant
connected and continuous
to the personal Results
excitation
computer, frequency
the vibrating to the building
table model to
was operated at supply
100 to 400 rpm in
a constant
50 rpm
and incremental
continuous steps. The
excitation acceleration
frequency to theresponse
buildingwas measured
model and
at 100 to 400then
rpmtransformed
in 50 rpm
Oncesteps.
from time-domain
incremental
thesignal
truetoexperimental
frequency-domain
The acceleration response was
setup
signal
was and established
for obtaining
measured the natural
then
and from
the accele
frequencies
transformed
lessly
time-domainconnected
of the model. to the personal
signal to frequency-domain signalcomputer,
for obtaining thethe vibrating
natural frequenciestable
of thewas op
model. The experimental results from the accelerometer sensor for the undamaged setup
constant and continuous excitation frequency to the building model a
wereThe experimental
exercised to updateresults
the from the accelerometer
FE model of the two-story sensor for the structure
building undamaged setup
using thewere
Ber-
50
manrpm
exercised
and toincremental
update
Nagy the FEwhich
method, steps.
model The
of the
improved acceleration
two-story
the FE building response
model tostructure
predict using was
dynamic measured
theresponses
Berman and
ac- an
Nagy method, which improved the FE model to predict
from time-domain signal to frequency-domain signal for obtaining the
curately. dynamic responses accurately.
AA sample
sample of of damage
damage introduced
introduced to to the
the building
building model
model hashas been
been shown
shown in in Figure
Figure 44
of
and
and
the
the model.
seven conditions of the building are summarized in Table 2. The damage hashas
conditions of the building are summarized in Table 2. The damage di-
2 and was introduced using a cutting tool. The seven conditions
Theofofexperimental
dimensions
mensions 1010××6 6mm mm 2 and wasresults
introduced fromusingthe accelerometer
a cutting tool. The sevensensorconditionsfor the
included
included oneone undamaged
undamaged condition
condition and and sixsix damaged
damaged conditions
conditions ofof the
the building
building model,
model,
were
and
exercised to update the FE model of thethetwo-story building struc
and the
the dynamic
dynamic response
response of of the
the building
buildingwas wasacquired
acquiredusing
using theaccelerometer.
accelerometer.
man and Nagy method, which improved the FE model to predict dyn
curately.
A sample of damage introduced to the building model has been
and the seven conditions of the building are summarized in Table 2. T
mensions of 10 × 6 mm2 and was introduced using a cutting tool. Th
included one undamaged condition and six damaged conditions of t
and the dynamic response of the building was acquired using the acce

Figure 4. A sample of damage introduced to the building model.

 Sensors 2022, 22, x FOR PEER REVIEW 7
Sensors 2022, 22, x FOR PEER REVIEW 7
Sensors 2022, 22, x FOR PEER REVIEW 7
Sensors 2022, 22, x FOR PEER REVIEW 7
Figure 4. A sample of damage introduced to the building model.
Figure 4. A sample of damage introduced to the building model.
Sensors 2022, 22, 4546 Figure 4. A sample of damage introduced to the building model. 7 of 14
Table 2.
Figure 4. Damage
A conditions
sample of damageofintroduced
the building
to model
the (All dimensions
building model. are in mm).
Table 2. Damage conditions of the building model (All dimensions are in mm).
Table 2. Damage conditions of the building model (All dimensions are in mm).
Sr No. Condition Image Description
Sr No. Condition Image
Table 2. Damage Description
conditions of the building model (All dimensions are in mm).
Sr No. Condition Image
Table 2. Damage conditions of Description
the building model (All dimensions are in mm).
Sr No. Condition Image Description
Sr No. Condition Image Description

The two-story building model was made from M

The two-story building model was made from M
flats two-story
The building
of 50 mm width andmodel
12 mm was made from
thickness. TheM
1. Undamaged flats of 50 mm width and 12 mm thickness. The
1. Undamaged flats of 50 mm width
The two-story
The two-story
mensions and
of building
(mm)building 12 mm
model
model
the scaled-down thickness.
wasmade
was The
from
building moM
1. Undamaged mensions (mm) of flats
the scaled-down building mo
flats made
mensions fromare
of 50 (mm)
mm MS
ofshown
theand
width ofin
50 mm
scaled-down
12the
mm width
buildingThe
thickness.
image. modd
1 1. Undamaged
Undamaged and 12 mmare shownThe
thickness. in the image.
dimensions
are shown in the image.
mensions (mm) of the scaled-down building mod
(mm) of the scaled-down building
are shown
model are shownininthe
the image.
image.

The damage condition (Type−1) was introduced

The The
damage
damage condition
condition (Type−1)
(Type −1) was introduced
wasintroduced
The damage
the walls condition
of the first-floor(Type−1)
of the was
building model,
the introduced
walls of the first-floor
to first-floor
the walls ofof of
thethe building
first-floor model,
Damaged the walls
The
shown, to of
damage the
condition
study its dynamic(Type−1)the building model,
was introduced
responses. The dam
2. Damaged shown, to study its dynamic responses. The dam
of the building model, as shown, to
Damaged
2. Damaged
(Type−1) shown,
the walls to
was established study
of the its dynamic
first-floor of
using a responses.
study its dynamic
responses.
the
cutting tool building
The
The dam
model,
to cut slots of
2 2. (Type−1) was established using a cutting tool to cut slots of
(Type−1)
(Type−Damaged
1) wasmmestablished
6shown, to
2 along
damage study
the
was using
its
width aof
establishedcutting
dynamic the
using tool
responses.
columns
a to cut
cutting slots
ofThe
the of
dama
mod
2. 6 mm22 along the width of the columns 2 of the mod
(Type−1) 6 mm
was along
toola to
established theslots
cut
height width
using
of 104.78of
of a10 ×the
6 mm
cutting
mm columns
fromalong
tool to ofbase.
cut
the the mod
slots of
the a height
width of of 104.78
the columns mm of from
the modelthe base.
a height of 104.78 mm
6 mm along the width of the columns of the mod
2 from the base.
at a height of 104.78 mm from the base.
a height of 104.78 mm from the base.

The Thedamage
damage
condition
condition
(Type−2)
(Type−2)was
was introduced
wasintroduced
The damage condition (Type−2)
The damage
the introduced
walls of the condition
to first-floor
the (Type−2)
walls ofof thethe was introduced
building model,
first-floor
the walls of the first-floor of the building model,
Damaged the walls
The
shown, theof
damage
of to the
study
buildingfirst-floor
condition
its model,
dynamic ofshown,
(Type−2)
as the building
was
responses. model,
to introduced
The dam
Damaged
3 Damaged shown,study to study
its its dynamic responses. The dam
3 3 Damaged
(Type−2) shown,
was walls to of
theestablished thedynamic
study a responses.
its dynamic
first-floor
using cutting thetoolThe
of responses. to cutThe
building dam
model,
slots of
3 (Type−2)
(Type−2) was established
damage using
was using a cutting
established usingtooltool to cut slots of
a cutting
(Type−2)
Damaged was established
6shown,
mm along
2 to study a cutting
its dynamic
the width of the responses. to cut
2 along ofThe
columns slots of
the dama
mod
3 6 mm22tool along theslots
to cut widthof 10of×the
6 mmcolumns of the mod
(Type−2) 6 mm
was along
established
a
the width
the
height width
using
of theof a of
58.74 the
cutting
mm columns
tool
from to
the of
cut the mod
slots
base. of
a height of columns
58.74 mm of the
from model
the base.
6 mmat a
2 along height
the of
width 58.74
of mm
the from
columns
a height of 58.74 mm from the base. the ofbase.
the mod
a height of 58.74 mm from the base.

The The damage

damage condition
condition (Type−3)was
(Type−3) wasintroduced
The damage condition (Type−3) was introduced
Theintroduced
the damage
walls to first-floor
the walls(Type−3)
condition
of the ofofthethefirst-floor
was introduced
building model,
the walls ofbuilding
of theof the first-floor
model, asofshown,
the building
to model,
Damaged the walls
The damage
shown, to studythe first-floor
condition of
(Type−3)
its dynamic the building model,
was introduced
responses. The dam
4 4 Damaged
Damaged shown,study its dynamic
to study responses.
its dynamic The
responses. The dam
4 (Type−Damaged
(Type−3)
3) shown,
was damage to of
theestablished
walls study
was its dynamic
theusing
first-floor of responses.
a cutting
established usingthe to cutThe
a building
tool
cutting dam
model,
slots of
4 (Type−3) was established using a cutting tool to cut slots of
(Type−3)
Damaged was established
shown, to
tool study
to cut using
its
slots a cutting
dynamic
of 10 6 tool
responses.
6 mm 2 along the width of the columns of the mod
2 × mm to
2 along cut slots
The of
dama
4 6 mmthe along the width of theofcolumns of the mod
(Type−3) 6 mm
was 2 along
width
established the
a height width
ofusing
the a of
columns
of 150.81 the
mmcolumns
cutting the
frommodel
tool to ofbase.
cut
the the mod
slots of
a height
at2 a height of of 150.81
150.81 mm mm
from from
the the
base. base.
6 mm along a height of 150.81
the width mmcolumns
of the from theofbase.
the mod
a height of 150.81 mm from the base.
Sensors 2022, 22, 4546 8 of 14

Sensors 2022, 22, x FOR PEER REVIEW 8

Sensors 2022, 22, x FOR PEER REVIEW 8
Sensors 2022, 22, x FOR PEER REVIEW 8
Table 2. Cont.

Sr No. Condition Image Description

The damage condition (Type−4)

(Type−4)waswas introduced
The The damage
damage condition
condition (Type−4) wasintroduced
The
the damage
walls of the condition
second-floor (Type−4)
of the was introduced
building mode
the walls introduced to the wallsofofthe
of the second-floor thebuilding mode
Damaged the
shown, to study its dynamic responses. Themode
walls of the
second-floor second-floor
of the buildingof the
model,building
as dam
5 Damaged shown, to study
shown, itsits
to study dynamic responses. The dam
Damaged
5 Damaged
5 (Type−4) shown,
was to study
established using adynamic
its dynamic
cutting responses.
responses.
tool to cutThe dam
slots of
5 (Type−4) was established
The damage using
was a cutting tool
established usingtoacut slots of
(Type−4)
(Type−4) was
6 mmestablished
2 along theusing
width aof
cutting
the tool to 2cut
columns of slots
the modof
6 mmcutting
2 alongtool
2 along the width
to of the
cut slots × 6 mm of the mod
of 10columns
6 mmalong the
a height width
of 288.93
the width
of the columns
mm from
of the columns
of the
the base.
of the
mod
a height of 288.93 mm from the base.
a height
model of 288.93
at a height mmmm
of 288.93 fromfromthe base.
the base.

The damage condition (Type−5)

(Type−5)was was introduced
The The damage
damage condition
condition (Type−5) wasintroduced
The
the damage
walls of the
introduced condition
second (Type−5)
floor
to the walls of
ofof the
thethe was introduced
building
second mode
the walls of the second floor building mode
Damaged the
shown, to study its dynamic responses. Themode
walls
floor of
of the
the second
building floor
model, of the
as building
shown, dam
6 Damaged shown, to study its dynamic responses. The dam
6 6 Damaged
Damaged
(Type−5) shown,
was to study
established itsusing
to study dynamic
its dynamicresponses.
a cutting toolThe
responses.
to cutThe dam
slots of
6 (Type−5)
(Type−5) was established
damage was using a cutting
established using tool
a to cut slots of
cutting
(Type−5) was
6 mmestablished
2 along theusing
width aof
cutting
the tool
columns to
2 along
cut
of slots
the modof
6 mm22tool
along theslots
to cut width of×the
of 10 6 mmcolumns of the mod
6 mmthealong the
a height
width
width
of 249.24
of the
of
columns
the
mm columns
from
of the
of the
the base.
model
mod
a height of 249.24 mm from the base.
a height
at a height of 249.24
of 249.24 mmmm fromfrom the base.
the base.

The damage condition (Type−6) was

The damage condition
introduced (Type−6)
to the(Type−6) was introduced
walls of the
The damage condition was introduced
The
the damage
walls of thecondition
second-floor second-floor
of the (Type−6)
building thewas
ofmodel, asintroduced
building mode
the walls of the second-floor of the building mode
Damaged the walls
shown, toofstudy
shown, the
to second-floor
itsits
study dynamic
dynamic of the building
responses.
responses. Themode
dam
Damaged
7 Damaged shown, todamage
study its dynamic responses. The dam
7 7 (Type−Damaged
(Type−6)
6)
shown,
The to study its
was dynamic
establishedresponses.
using
was established using a cutting tool to 2cut slotsa The damof
7 (Type−6) was established
cutting using
tool to a cutting
cut slots of 10 ×tool
6 mmto cut slots of
(Type−6) was established using a cutting tool
6 mm2 along the width of the columns of the mod
2 to cut slots of
6 mmalong
alongthethe width
width of columns
of the the columnsof theof the mod
6 mm along
2 the width
a height of 328.62 of the
mmcolumns
from theofbase.
the mod
model at a height
a height of 328.62
of 328.62 mmmm fromfromthe base.
a height ofthe 328.62
base. mm from the base.

Utilizing the experimental results, a universal analytical model for the two-s
Utilizing theUtilizing the experimental
experimental
Utilizing
results, analytical
results, a universal a universalmodel
analytical
for themodel for the two-s
two-story
building modelthe
canexperimental
be generated.results, a universal
The natural analytical
frequencies model
of the first formodes
four the two-s
of vi
building modelbuilding
can model
be can be generated.
generated. The Thefrequencies
natural natural frequencies
of the of the
first first
four four modes
modes of of vi
building model can be generated. The natural frequencies of the
tion are given in Table 3, which is true for all the excitation frequencies.first four modes of v
vibration are tion
tion
areingiven
given Tablein
are given 3, Table
in which3,
Table
which
3,iswhich
is true
true for
is all
true
forexcitation
the all the excitation frequencies.
frequencies.
for all the excitation frequencies.
Table 3. Updated natural frequencies (Hz) of undamaged and damaged analytical models.
Table
Table 3. Updated 3. Updated
natural natural
frequencies frequencies
(Hz) (Hz) of
of undamaged andundamaged and damaged
damaged analytical analytical models.
models.
Table 3. Updated natural frequencies (Hz) of undamaged and damaged analytical models.
Natural Frequency (Hz) of the Building Model
Natural Natural Frequency
Model (Hz) of the Building Model
Mode Un- (Hz) ofNatural
No.Frequency Frequency
the Building (Hz) Damaged
of the Building Model
Mode No. Un- Damaged
Mode No. Un- Mode No. DamagedUn- Damaged Type−2
Type−1 Damaged
Type−3 Type−4 Type−5 Type
Damaged Type−1 Type−2 Type−3 Type−4 Type−5 Type
Damaged Type−1 Damaged Type−1
Type−2 54.15Type−353.84
1. Type−2
Type −4
53.75 Type−3
Type−5 Type−4
53.71 53.66 TypeType−5
−53.66
6 Type
53.6
1. 54.15 53.84 53.75 53.71 53.66 53.66 53.6
1. 54.15 53.84 1.53.75
2. 54.15 53.71264.69
266.24 53.84 53.75
264.24
53.66 53.71
263.84 53.66
262.53
53.66 262.53 53.66 53.66
262.16 53.6
262.
2. 266.24 264.69 264.24 263.84 262.16 262.
2. 266.24 264.69 2.264.24 307.67
3. 266.24263.84304.80
264.69 264.24
262.53
304.09 263.84 262.53 262.04
262.16 297.81
302.40 262.16
296.51 262.
296.
3.304.09 307.67 302.40304.80 304.09 302.40 297.81 296.51 296.
3. 307.67 304.80 3.
4. 307.67
368.04 304.80
361.36 304.09
297.81
361.01 302.40
296.51
354.89 297.81
341.64 296.51
296.33
338.89 296.
337.
4. 368.04 361.36 4.361.01 368.04 354.89361.36 361.01
341.64 354.89
338.89 341.64 338.89 337.
4. 368.04 361.36 361.01 354.89 341.64 337.99
338.89 337.
Sensors 2022, 22, 4546 9 of 14

Sensors 2022, 22, x FOR PEER REVIEW 9 of 14

5. Discussion
5. Discussion
The results from Table 3 were employed to update the undamaged FE model to
the various damagedThe results from Table
conditions of the3 were employed
two-story to update
building the undamaged
structure using theFEBerman
model to the
various The
and Nagy method. damaged
pseudoconditions
code for of plotting
the two-story
and building
comparing structure
FRFs using the Berman
is discussed in and
Figure 5. Figures 6–11 predict the structural dynamic behavior of the building modelFigure
Nagy method. The pseudo code for plotting and comparing FRFs is discussed in
5. Figures
under undamaged and 6–11 predictconditions
damaged the structural dynamic
through FRFsbehavior of the building
for a better modelstudy.
comparative under un-
damaged and damaged conditions through FRFs for a better comparative study. The fre-
The frequency response function, generally known as FRF, is a complex mathematical
quency response function, generally known as FRF, is a complex mathematical relation-
relationship between the input signal and the output signal for the system [36]. If the
ship between the input signal and the output signal for the system [36]. If the measured
measured response is in the form displacement, then the corresponding FRF is called
response is in the form displacement, then the corresponding FRF is called a receptance
a receptance (admittance, dynamic
(admittance, dynamic compliance,
compliance, or dynamic
or dynamic flexibility)
flexibility) FRF. Otherwise,
FRF. Otherwise, velocity or ac-
velocity or acceleration response signals can be used to produce mobility
celeration response signals can be used to produce mobility or accelerance or accelerance
FRF, respec-
FRF, respectively.
tively.Accelerance
Accelerance FRFFRFisisalso
alsosometimes
sometimes written
written as inertance
as inertance FRF.
FRF. The FRFThe
canFRFsimilarly
can similarly be
be defined
defined as asthe
thevibration
vibration response
response of aoflocation
a location
owingowing to excitation
to excitation at thatatlocation
that or
location or any other
any otherlocation
locationininthe
the system.
system.

Figure 5. Pseudo code5.for

Figure extracting
Pseudo and
code for plottingand
extracting FRFs.
plotting FRFs.

As discussed previously, dynamic responses are the function of stiffness and mass dis-
tribution of the structure. Variation in these, due to any external influence, will result in devi-
ation of the dynamic response from the control structure; this was established after studying
the dynamic response of the building following the introduction of damage conditions.
After analyzing the dynamic responses and FRFs of the system, it can be established
that with the introduction of damage along the column, the natural frequencies of the
system decrease. Furthermore, the magnitude of reduction in the natural frequency is
directly proportional to the mode of vibration, i.e., the higher the mode of vibration, the
larger the reduction in the frequency will be. It can also be postulated that the magnitude
of reduction in the frequency of mode of vibration, in the case of a building structure
due to damage, depends on three main factors: elevation of the damage from the base
of the structure, extent of damage, and the presence of damage in the vicinity of that
damage condition.
Sensors 2022,22,
Sensors2022, 22,4546
x FOR PEER REVIEW 10 of
10 of 14
14
Sensors 2022, 22, x FOR PEER REVIEW 10 of 14
Sensors 2022, 22, x FOR PEER REVIEW 10 of 14

Figure 6.
Figure 6. Prediction
Predictionofofstructural
structuraldynamic
dynamicbehavior of building
behavior model
of building under
model undamaged
under and dam-
undamaged and
Figure 6. Prediction
aged (Type−1) of structural
conditions throughdynamic
FRFs. behavior of building model under undamaged and dam-
damaged (Type − 1) conditions through FRFs.
Figure 6. Prediction of structural dynamic behavior of building model under undamaged and dam-
aged (Type−1) conditions through FRFs.
aged (Type−1) conditions through FRFs.

Figure 7. Prediction of structural dynamic behavior of building model under undamaged and dam-
Figure
Figure 7. Prediction
Prediction
aged (Type−2) ofofstructural
conditions throughdynamic
structural dynamic
FRFs. behavior of building
behavior model
of building under
model undamaged
under and dam-
undamaged and
Figure 7. Prediction of structural dynamic behavior of building model under undamaged and dam-
aged (Type−2)
damaged (Typeconditions
− through
2) conditions FRFs.FRFs.
through
aged (Type−2) conditions through FRFs.

Figure 8. Prediction of structural dynamic behavior of building model under undamaged and dam-
Figure 8. Prediction
aged (Type−3) of structural
conditions throughdynamic
FRFs. behavior of building model under undamaged and dam-
Figure 8. Prediction
Predictionofofstructural dynamic
structural behavior
dynamic of building
behavior model
of building under
model undamaged
under and dam-
undamaged and
aged (Type−3) conditions through FRFs.
aged (Type−3)
damaged (Typeconditions through
−3) conditions FRFs.FRFs.
through
Sensors 2022, 22, 4546 11 of 14
Sensors 2022, 22, x FOR PEER REVIEW 11 of 14
Sensors 2022, 22, x FOR PEER REVIEW 11 of 14

Figure 9. Prediction
Figure Predictionofofstructural dynamic
structural dynamicbehavior of building
behavior model
of building under
model undamaged
under and dam-
undamaged and
Figure
aged 9. Prediction
(Type−4) of structural dynamic behavior of building model under undamaged and dam-
damaged (Typeconditions
−4) conditionsthrough FRFs.FRFs.
through
aged (Type−4) conditions through FRFs.

Sensors 2022, 22, x FOR PEER REVIEW 12 of 14

Figure 10. Prediction of structural dynamic behavior of building model under undamaged and dam-
Figure
Figure
aged 10. Prediction
10. Prediction
(Type−5) ofofstructural dynamic
structural dynamicbehavior
behaviorof building model
of building under
model undamaged
under and dam-
undamaged and
reduction
aged (Type−5)in conditions
the naturalthrough
conditions through
FRFs.
frequency of the first four modes of vibration for respective dam-
FRFs.
damaged (Type−5) conditions through FRFs.
age conditions.
As discussed previously, dynamic responses are the function of stiffness and mass
As discussed previously, dynamic responses are the function of stiffness and mass
distribution of the structure. Variation in these, due to any external influence, will result
distribution of the structure. Variation in these, due to any external influence, will result
in deviation of the dynamic response from the control structure; this was established after
in deviation of the dynamic response from the control structure; this was established after
studying the dynamic response of the building following the introduction of damage con-
studying the dynamic response of the building following the introduction of damage con-
ditions.
ditions.
After analyzing the dynamic responses and FRFs of the system, it can be established
After analyzing the dynamic responses and FRFs of the system, it can be established
that with the introduction of damage along the column, the natural frequencies of the
that with the introduction of damage along the column, the natural frequencies of the
system decrease. Furthermore, the magnitude of reduction in the natural frequency is di-
system decrease. Furthermore, the magnitude of reduction in the natural frequency is di-
rectly proportional to the mode of vibration, i.e., the higher the mode of vibration, the
rectly proportional to the mode of vibration, i.e., the higher the mode of vibration, the
larger the reduction in the frequency will be. It can also be postulated that the magnitude
larger the reduction in the frequency will be. It can also be postulated that the magnitude
of reduction in the frequency of mode of vibration, in the case of a building structure due
of reduction in the frequency of mode of vibration, in the case of a building structure due
to damage, depends on three main factors: elevation of the damage from the base of the
to damage, depends on three main factors: elevation of the damage from the base of the
structure, extent of damage, and the presence of damage in the vicinity of that damage
structure, extent of damage, and the presence of damage in the vicinity of that damage
condition.
condition.
The magnitude of reduction is directly proportional to the elevation of the damage
The magnitude of reduction is directly proportional to the elevation of the damage
from the base of the structure and the extent of damage. It is inversely proportional to the
from the
Figure 11.base of the structure
Prediction and the extent of damage. It model
is inversely proportional todam-
the
presence
Figure 11. of damageof instructural
of structuraldynamic
the vicinity
dynamic behavior
of that
behaviorof of
damage. building
Table under
4 displays
building model undamaged
the
under elevationand
undamagedof each
and
presence
aged of of damage
(Type−6) in the
conditions vicinity
through of that damage. Table 4 displays the elevation of each
FRFs.
type
damaged damage
(Type − condition
6) conditions introduced
through FRFs. to the building model. Figure 12 exhibits the
type of damage condition introduced to the building model. Figure 12 exhibits the
Table 4. Elevation (mm) of all damage conditions.

Damaged
Type−1 Type−2 Type−3 Type−4 Type−5 Type−6
Condition
Sensors 2022, 22, 4546 12 of 14

The magnitude of reduction is directly proportional to the elevation of the damage

from the base of the structure and the extent of damage. It is inversely proportional to
the presence of damage in the vicinity of that damage. Table 4 displays the elevation of
Figure 11. Prediction of structural dynamic behavior of building model under undamaged and da
each type of damage condition introduced to the building model. Figure 12 exhibits the
aged (Type−6)
reduction conditions through
in the natural FRFs.
frequency of the first four modes of vibration for respective damage
conditions.
Table 4. Elevation (mm) of all damage conditions.
Table 4. Elevation (mm) of all damage conditions.
Damaged
Damaged Type−1
Type−1
Type−2
Type−2
Type−3 Type−Type−4
Type−3 4 Type−5
Type−5
Type−6
Type−6
Condition
Condition
Elevation
Elevation
(mm)
104.78
104.78 58.74
58.74 150.81 288.93288.93249.24
150.81 249.24
328.62 328.62
(mm)

FigureFigure
12. Reduction in in
12. Reduction frequencies offirst
frequencies of firstfour
four modes
modes of vibration.
of vibration.

As analyzed in the case of Type−1 and Type−2 damage conditions, the reduction in
As analyzed in the case of Type−1 and Type−2 damage conditions, the reduction
the frequency of a mode is lesser than that for Type−3 damage because of its elevation. The
the frequency of a mode
highest reduction is lesser
in frequency can bethan
seenthat
in thefor
caseType−3
of Type−damage because of
4 damage conditions, its elevatio
owing
The highest reduction
to its altitude and zero indamages
frequency in itscan be seen
vicinity. Type−in5 and
the Type
case−of Type−4conditions
6 damage damageare condition
also present at a higher level of elevation, but the Type − 4 damage
owing to its altitude and zero damages in its vicinity. Type−5 and Type−6 damage concondition already exists,
tions due
are to which
also reduction
present at ainhigher
the frequency
level ofdecreases.
elevation,Initially,
but thewhen smallerdamage
Type−4 damage wascondition
created on the building model, the reduction in magnitude was found to be insignificant.
ready exists,
It is also evident from the experimental results that the natural frequency of the smal
due to which reduction in the frequency decreases. Initially, when
modes of vibration from the lower end tries to gradually attain a constant value, following
which they are unaffected by the introduction of new damage conditions in the case of the
building structure.

6. Conclusions
In this research, a scaled-down version of a two-story building model was developed
experimentally and a finite element model was generated analytically. The experimental
setup of the building model was subjected to vibratory stimuli from a shaking table at
different frequencies. Modal testing was performed using wireless sensors in combination
Sensors 2022, 22, 4546 13 of 14

with the finite element model updating method of Berman and Nagy, so that the updated
analytical model can predict the experimental dynamic response values accurately. Several
damage conditions were introduced to the columns of the building model and the dynamic
responses were scrutinized to analyze the magnitude of reduction in natural frequency
and the factors influencing it. It was realized that the magnitude of reduction is directly
proportional to the elevation of the damage from the base of the structure and extent of
damage condition, and inversely proportional to the presence of other damages in the
vicinity of that damage. It is also apparent that the natural frequency of the first mode of
vibration gradually tries to attain a stagnant value, following which it is unaffected by the
introduction of new damage conditions in the case of the building structure.

Author Contributions: All authors contributed to the study conception and design. Material prepa-
ration, data collection, and analysis were performed by T.S. The first draft of the manuscript was
written by T.S. and all authors commented on previous versions of the manuscript. All authors have
read and agreed to the published version of the manuscript.
Funding: The research is partially funded by the Ministry of Science and Higher Education of
the Russian Federation as part of the World-class Research Center program: Advanced Digital
Technologies (contract No. 075-15-2022-311 dated 20 April 2022). Authors are also grateful to the
Department of Science and Technology & Renewable Energy, India, for financial support of this
research work vide research project grant as per S&T&RE/RP/147(19-20)/Sanc/10/2019/1703-1710
dated 31 October 2019.
Conflicts of Interest: The authors declare that they have no known competing financial interests or
personal relationships that could have appeared to influence the work reported in this paper.

References
1. Mishra, M.; Lourenço, P.B.; Ramana, G.V. Structural health monitoring of civil engineering structures by using the internet of
things: A review. J. Build. Eng. 2022, 48, 103954. [CrossRef]
2. Wu, Z.; Zhang, Q.; Cheng, L.; Hou, S.; Tan, S. The VMTES: Application to the structural health monitoring and diagnosis of
rotating machines. Renew. Energy 2020, 162, 2380–2396. [CrossRef]
3. Zhou, H.; Huang, X.; Wen, G.; Lei, Z.; Dong, S.; Zhang, P.; Chen, X. Construction of health indicators for condition monitoring of
rotating machinery: A review of the research. Expert Syst. Appl. 2022, 203, 117297. [CrossRef]
4. Al-Nasar, M.K.R.; Al-Zwainy, F.M.S. A systematic review of structural materials health monitoring system for girder-type bridges.
Mater. Today Proc. 2022, 49, A19–A28. [CrossRef]
5. He, Z.; Li, W.; Salehi, H.; Zhang, H.; Zhou, H.; Jiao, P. Integrated structural health monitoring in bridge engineering. Autom.
Constr. 2022, 136, 104168. [CrossRef]
6. Singh, T.; Sehgal, S. Structural Health Monitoring of Composite Materials. Arch. Comput. Methods Eng. 2021, 29, 1997–2017.
[CrossRef]
7. Adams, R.D.; Cawley, P.; Pye, C.J.; Stone, B.J. A Vibration Technique for Non-Destructively Assessing the Integrity of Structures. J.
Mech. Eng. Sci. 1978, 20, 93–100. [CrossRef]
8. Cawley, P.; Adams, R.D. The location of defects in structures from measurements of natural frequencies. J. Strain Anal. Eng. Des.
1979, 14, 49–57. [CrossRef]
9. Kessler, S.S.; Spearing, S.M.; Atalla, M.J.; Cesnik, C.E.S.; Soutis, C. Structural health monitoring in composite materials using
frequency response methods. Nondestruct. Eval. Mater. Compos. V 2001, 4336, 1–11. [CrossRef]
10. Goi, Y.; Kim, C.-W. Damage detection of a truss bridge utilizing a damage indicator from multivariate autoregressive model. J.
Civ. Struct. Health Monit. 2017, 7, 153–162. [CrossRef]
11. González, M.P.; Zapico, J.L. Seismic damage identification in buildings using neural networks and modal data. Comput. Struct.
2008, 86, 416–426. [CrossRef]
12. Ochôa, P.A.; Groves, R.M.; Benedictus, R. Effects of high-amplitude low-frequency structural vibrations and machinery sound
waves on ultrasonic guided wave propagation for health monitoring of composite aircraft primary structures. J. Sound Vib. 2020,
475, 115289. [CrossRef]
13. Mishra, M. Machine learning techniques for structural health monitoring of heritage buildings: A state-of-the-art review and case
studies. J. Cult. Herit. 2020, 47, 227–245. [CrossRef]
14. Gopinath, V.; Ramadoss, R. Review on structural health monitoring for restoration of heritage buildings. Mater. Today Proc. 2020,
43, 1534–1538. [CrossRef]
15. Clemente, P. Extending the life-span of cultural heritage structures. J. Civ. Struct. Health Monit. 2018, 8, 171–179. [CrossRef]
16. Cigada, A.; Dell’Acqua, L.C.; Castiglione, B.M.V.; Scaccabarozzi, M.; Vanali, M.; Zappa, E. Structural Health Monitoring of an
Historical Building: The Main Spire of the Duomo Di Milano. Int. J. Archit. Herit. 2017, 11, 501–518. [CrossRef]
Sensors 2022, 22, 4546 14 of 14

17. Zou, Y.; Tong, L.; Steven, G.P. Vibration-based model-dependent damage (delamination) identification and health monitoring for
composite structures—A Review. J. Sound Vib. 2000, 230, 357–378. [CrossRef]
18. Fan, W.; Qiao, P. Vibration-based damage identification methods: A review and comparative study. Struct. Health Monit. 2011, 10,
83–111. [CrossRef]
19. Yan, Y.J.; Cheng, L.; Wu, Z.Y.; Yam, L.H. Development in vibration-based structural damage detection technique. Mech. Syst.
Signal Process. 2007, 21, 2198–2211. [CrossRef]
20. Alavi, A.H.; Hasni, H.; Lajnef, N.; Chatti, K.; Faridazar, F. An intelligent structural damage detection approach based on
self-powered wireless sensor data. Autom. Constr. 2016, 62, 24–44. [CrossRef]
21. Tran, H.; Khatir, S.; Thang, L.; De Roeck, G.; Bui-Tien, T.; Abdel Wahab, M. Finite element model updating of a multispan
bridge with a hybrid metaheuristic search algorithm using experimental data from wireless triaxial sensors. Eng. Comput. 2021.
[CrossRef]
22. Hsu, T.Y.; Huang, S.K.; Lu, K.C.; Loh, C.H. A damage detection algorithm integrated with a wireless sensing system. J. Phys. Conf.
Ser. 2011, 305, 12042. [CrossRef]
23. Feng, L.; Yi, X.; Zhu, D.; Xie, X.; Wang, Y. Damage detection of metro tunnel structure through transmissibility function and cross
correlation analysis using local excitation and measurement. Mech. Syst. Signal Process. 2015, 60–61, 59–74. [CrossRef]
24. Lucena, C.; Queiroz, P.; Debs, A.; Mendonca, A. Dynamic analysis of buildings using the Finite Element Method. In Proceedings
of the 10th World Congress on Computational Mechanics, São Paulo, Brazil, 8–13 July 2012; pp. 4712–4726. [CrossRef]
25. Pathak, M.V.; Bhaskar, G.B. Finite element analysis program of frames. Int. J. Technol. Res. Eng. 2016, 3, 2455–2459.
26. Jain, A.K. Dynamics of Structures with MATLAB® Applications; Pearson Education: Chennai, India, 2017; ISBN 9789332558557.
27. Sehgal, S.; Kumar, H. Structural Dynamic Model Updating Techniques: A State of the Art Review. Arch. Comput. Methods Eng.
2016, 23, 515–533. [CrossRef]
28. Bais, R.S.; Gupta, A.K.; Nakra, B.C.; Kundra, T.K. Studies in dynamic design of drilling machine using updated finite element
models. Mech. Mach. Theory 2004, 39, 1307–1320. [CrossRef]
29. Baruch, M.; Bar Itzhack, I.Y. Optimal weighted orttiogonalization of measured modes. AIAA J. 1978, 16, 346–351. [CrossRef]
30. Berman, A.; Nagy, E.J. Improvement of a large analytical model using test data. AIAA J. 1983, 21, 1168–1173. [CrossRef]
31. Arora, V.; Singh, S.P.; Kundra, T.K. Damped model updating using complex updating parameters. J. Sound Vib. 2009, 320, 438–451.
[CrossRef]
32. Simoen, E.; De Roeck, G.; Lombaert, G. Dealing with uncertainty in model updating for damage assessment: A review. Mech.
Syst. Signal Process. 2015, 56, 123–149. [CrossRef]
33. Teughels, A.; De Roeck, G. Damage detection and parameter identification by finite element model updating. Arch. Comput.
Methods Eng. 2005, 12, 123–164. [CrossRef]
34. Gesualdi, G.; Cardone, D.; Rosa, G. Finite element model updating of base-isolated buildings using experimental results of in-situ
tests. Soil Dyn. Earthq. Eng. 2019, 119, 422–432. [CrossRef]
35. Tian, W.; Weng, S.; Xia, Q.; Xia, Y. Dynamic condensation approach for response-based finite element model updating of
large-scale structures. J. Sound Vib. 2021, 506, 116176. [CrossRef]
36. Hosoya, N.; Kato, J.; Kajiwara, I. Spherical projectile impact using compressed air for frequency response function measurements
in vibration tests. Mech. Syst. Signal Process. 2019, 134, 106295. [CrossRef]