J Fail. Anal. and Preven.
https://doi.org/10.1007/s11668-018-0418-4
TECHNICAL ARTICLE—PEER-REVIEWED
Failure Analysis of Composite Mono Leaf Spring Using Modal
Flexibility and Curvature Method
N. I. Jamadar . S. B. Kivade . Rakesh Raushan
Submitted: 25 October 2017 / in revised form: 2 January 2018
Ó ASM International 2018
Abstract Post-failure analysis of composite structures by
damage identification at pre-stage has gained considerable
interest in recent years. In the paper, modal flexibility and
modal curvature methods have been used for premature
failure analysis of composite mono leaf spring through
analytical and finite element approaches. Initially, finite
element model of healthy and cracked spring is used to
evaluate the eigenvalue and eigenvectors. Subsequently,
change in flexibility and absolute curvature difference
among both springs are evaluated after mass normalization.
The presence, location and severity of crack in a spring are
identified through differences in local flexibilities and
change in absolute modal curvatures. The modal curvature
method predicts location and severity of crack more precisely than the modal flexibility method.
M*e
Ke
K*e
le
K
K*
M
M*
U
[F]hx
[F]dx
Uxi
Keywords Failure analysis Modal flexibility
Modal curvature Composite mono leaf spring
Analytical Healthy and cracked spring
List of symbols
Me
Elemental mass of healthy beam
N. I. Jamadar (&)
Research Resource Center, VTU, Kalaburagi Region, Karnataka,
India
e-mail: jamadar94@gmail.com
S. B. Kivade
Basavakalyan Engineering College, Basavakalyan, Karnataka,
India
e-mail: sbkivade@gmail.com
R. Raushan
Dr. D. Y. Patil Institute of Technology, Pimpri,
Pune, Maharashtra, India
e-mail: prabhatpankaj1989@gmail.com
D[F]
M(x)
M*(x)
K(x)
K*(x)
D[K]
Uh1, Uh2, Uh3
and Uh4
Ud1, Ud2, Ud3
and Ud4
Kh1, Kh2, Kh3
and Kh4
Elemental mass of cracked beam
Elemental stiffness of healthy beam
Elemental stiffness of cracked beam
Elemental length of beam
Assembled stiffness matrix of healthy
beam
Assembled stiffness matrix of cracked
beam
Assembled mass matrix of healthy beam
Assembled mass matrix of cracked beam
Eigenvector
Flexibility matrix of healthy beam at
distance x
Flexibility matrix of cracked beam at
distance x
Mass normalized modal vector at
distance x
Change in flexibility
Bending moment of healthy beam at
distance x
Bending moment of cracked beam at
distance x
Curvature of healthy beam at distance x
Curvature of cracked beam at distance x
Absolute modal curvature difference
Are the normalized eigenvectors of
healthy beam for 1st, 2nd, 3rd and 4th
modes, respectively
Are the normalized eigenvectors of
cracked beam for 1st, 2nd, 3rd and 4th
modes, respectively
Are the modal curvatures of healthy
beam for 1st, 2nd, 3rd and 4th modes,
respectively
123
J Fail. Anal. and Preven.
Kd1, Kd2, Kd3
and Kd4
Are the modal curvatures of cracked
beam for 1st, 2nd, 3rd and 4th modes,
respectively
Introduction
Study of composite structural fatigue failures in service
condition has gained significance in scientific research
community. The failure of structures occurs due to reduction in local stiffness and leads in lowering the load
carrying capacity. The damage mechanism in these structures is complex in nature than the conventional materials.
The defects developed during manufacturing or in service
condition leads to low fatigue cycle failures in composite
structures. These defects include micro-cracks, delamination such as inter-laminar or de-bonding, micro-voids and
inclusions. These defects grow in progression during fatigue loading. The progressive assessment of such damages
at the intermediate status is one of the challenging tasks.
He et al. [1] explored damage identification in composite
beams through curvature mode difference via central difference operation with single and multiple damages on it.
The experimental and simulation results showed the
changes in modal difference between intact and damaged
beams. The presence, location, size of damage identified
accurately. Zhou et al. [2] proposed numerical method to
locate the damage in ocean drilling riser by modal curvature difference which showed accuracy in locating the
damages of intact and damaged element. Yang et al. [3]
presented Fourier spectral-based method, an alternate to
modal curvature method used for locating the damage of
any structures. The comparison of both the methods has
been demonstrated by experimental and numerical analysis
for accuracy. However, found weak in detecting small
damages. Yang et al. [4] proposed discrete Fourier transform-based modal curvature and scale wavenumber domain
filtering method to inspect the damage in carbon-reinforced
polymer beam with and without crack. Researcher concludes the traditional modal curvature method weaker in
dealing with the noisy data which deviates the accuracy in
detection of damage. Analytical method carried out and
validated with the experimental work. Dessi and Camerlengo [5] overviewed various damage identification
methods based on the natural frequency, curvature mode
shape and modal strain energy. Sensitivity and performance
analysis has been carried out to adjacent points of the
damage. Dawari and Vesmawala [6] applied both modal
curvature and modal flexibility methods to identify damage
in reinforced concrete beam and showed that modal curvature changes are high at region near the damage. Russo
[7] also investigated the homogeneity of the glass fiber-
123
reinforced polymer structural beam elements with different
cross sections. The damage location and its extent in these
structures were found based on their dynamic characteristics by curvature mode shapes. Ekinovic et al. [8]
determined the damage present in the cantilever beam
through modal parameters. The numerical model of beam
has been formulated to assess its dynamic modal characteristics. The modal curvatures and displacement mode
shapes have been calculated by central difference approximation. The results showed that the changes in mode shape
characteristics are the good indicators for location and
severity of damage. Lestari and Qiao [9] investigated
experimentally the damage location in fiber-reinforced
polymer (E-glass) sandwich beam based on curvature mode
shape with piezoelectric sensor and showed the lower
modal curvature decides accurately the damage location
than higher modal curvatures. Zhua and Xub [10] presented
numerical and experimental approach on 10-mono coupled
periodic structure by sensitivity-based method for damage
localization and its severity. To locate and quantify the
damage, curvature mode shape/slopes, natural frequencies
are used, respectively. It has been noticed that the modal
curvatures are more sensitive and accurate in locating the
damage than modal slopes. Lee and Eun [11] derived
analytical methods to update the flexibility matrix by the
expanded full degree of freedom. The detection of damage
carried out based on updated flexibility matrix. It is concluded that the second derivative of flexibility is accurate in
yielding results as compared to traditional method. Sung
et al. [12] presented new approach of damage detection
over cantilever beam using damage induced inter-storey
deflection estimated by modal flexibility. Gaith [13]
determined the effect of crack location, depth, fiber orientation and volume fraction on flexibility of graphite/epoxy
composite beam and determined the effect of these
parameters on natural frequency and mode shapes. Reynders and Roeck [14] determined the local stiffness directly
through flexibility to identify the presence and location of
damage in a reinforced concrete beam. Yu and Yin [15]
proposed a method for damage detection over a portal
frame by finite element model updating. This methodology
aimed to minimize the discrepancies among the experimental and analytical modal parameters. Abdo and Hori
[16] studied numerically the cantilever steel plate for
damage detection using changes in the rotation of mode
shapes instead of modal flexibility and curvature mode
shapes. It has been proved that the change in rotation mode
shape is the sensitive indicator for single and multiple
damages. Pandey and Biswas [17] used modal flexibility
method for locating damage in wide flanged beams and
concluded that only lower-frequency modes are suitable in
locating the damage in structures. Yuen [18] presented a
study of damage size and its location by changes in
J Fail. Anal. and Preven.
eigenvalue and eigenvector in damaged cantilever beam.
The changes in the eigenvalue and eigenvectors have
showed the indication of location and extent of damage.
Most of the research articles investigated the damage
presence, location and severity in a structure based on
modal curvatures, flexibility, slopes and rotations. The
main focus of these researchers was on steel plates, frames,
concrete beams and fiber-reinforced structures. The
extensive study on composites at early stage of damage
identification on automotive structural parts such as suspension leaf springs, chassis and body frames is essential.
Details of Composite Mono Leaf Spring
In the present work, composite mono leaf spring of utility
vehicle is analyzed. It is single leaf spring with T-3 tapered
trapezoidal cantilever [19] which has maximum width
125 mm, minimum width 72 mm, maximum thickness
18 mm and minimum thickness 12 mm. It is made of Eglass/epoxy material with unidirectional laminated fibers of
high strength/weight ratio than the conventional leaf
spring. Table 1 shows mechanical properties of unidirectional E-glass/epoxy composite material, and Fig. 1 shows
the actual mono composite leaf spring.
Artificial Crack Location
In practice, the failures of leaf springs are due to low
fatigue cycles resulted in weakening the portions of it
leading to premature failure. To understand the presence,
location and severity of damage, an artificial crack is created on the healthy spring.
Table 1 Mechanical properties of unidirectional E-glass/epoxy
material
Sr. no.
Properties
Value
1.
Density (Kg/m3)
2000
2.
Young’s modulus X direction
(MPa)
50,000
3.
Young’s modulus Y direction
(MPa)
10,000
4.
Young’s modulus Z direction
(MPa)
10,000
5.
Poisson’s ratio XY
0.3
6.
Poisson’s ratio YZ
0.4
7.
Poisson’s ratio XZ
0.3
8.
Shear modulus XY (MPa)
5000
9.
Shear modulus YZ (MPa)
3846.2
10.
Shear modulus XZ (MPa)
5000
11.
Ultimate tensile strength (MPa)
1390
Initially, finite element analysis of healthy composite
leaf spring is conducted to identify the damage sensitive
region to create artificial crack on it. This region is located
130 mm from the center and 380 mm from the eye end.
The crack dimensions are selected using the relation [20].
It is a = a/w where a = crack ratio (0.2–0.8), a = crack
length, w = width in direction of crack propagation. As the
composite material is orthotropic behavior in nature, it is
planned to check the crack behavior by creating small
crack of size 2 mm length and 2 mm depth on the surface.
The crack length is further increased to observe the effect
of crack size on propagation.
The cracked spring is subjected to set number of fatigue
cycles of 500, 1000 and so on in succession until the crack
evolution. After every 500 fatigue cycles, static testing has
been carried out to record the strains and corresponding
stiffness for maximum load of 5000 N. Finally, the full
length crack along the width has been created. Figure 2
illustrates the same.
Theoretical Background
The modal parameters such as natural frequencies, mode
shapes, flexibility, curvatures and modal rotations are used
for damage detection. Among these parameters, modal
flexibility and curvature mode shape methods are powerful
techniques for inspection of damage. The differences in
numerical values in the parameters are compared with
healthy spring.
Cantilever Beam Finite Element Model
The trapezoidal mono composite leaf springs are modeled
as healthy and cracked cantilever beams by finite element
method as shown in Figs. 1 and 2. It has been divided into
four equal parts by considering equivalent width and
thickness (t) across length of beam. The thickness has been
reduced to account for reduction in the cracked area over
the cracked beam. The numbers mentioned for each portions are the modal displacements, i.e., 1, 3, 5, 7, 9 and 2, 4,
6, 8, 10 are modal rotations, respectively. The numerical
analysis is carried out by considering modal displacements
only (Figs. 3 and 4).
The equation of motion for ‘n’ degrees of freedom of a
structure can be written as
:: :
M x þ C x þ K ¼ 0:
The healthy and cracked springs are modeled by
considering consistent mass. The elemental masses for
healthy and cracked beams can be expressed as
123
J Fail. Anal. and Preven.
Fig. 1 Composite mono leaf
spring
Fig. 2 Artificial crack on the
spring
Artificial Crack
118.75
Fig. 4 Finite element model of
cracked spring
118.75
78.63
118.75
105.125
t= 14.25
9,10
7,8
t=15.625
5,6
3,4
t=12.75
123
t= 14.25
118.75
118.38
1,2
91.88
105.125
118.38
t=12.75
118.75
5,6
3,4
1,2
118.75
t=17.25
118.75
78.63
118.75
91.88
Fig. 3 Finite element model of
healthy spring
9,10
7,8
t=13.625
t=17.25
J Fail. Anal. and Preven.
2
3
156
22le
54
13le
qAle 6
4l2e
13le
3l2e 7
6 22le
7
Me ¼
4
54
13le
156 22le 5
420
2
13l
4l2
2 e 3le 22le
3
156
22le
54
13le
qAle 6
4l2e
13le
3l2e 7
6 22le
7:
Me ¼
4
54
13le
156 22le 5
420
13le 3l2e 22le
4l2
The elemental stiffness for both the beam can be expressed
as
2
3
12
6le
12 6le
EI 6 6le
4l2e 6le 2l2e 7
7
Ke ¼ 3 6
4
12 6le 5
le 12 6le
6le
2l2e 6le 4l2
2
3
12
6le
12 6le
EI 6 6le
4l2e 6le 2l2e 7
7:
Ke ¼ 3 6
4
12 6le 5
le 12 6le
6le
2l2e 6le 4l2
The generalized characteristic equation or eigenvalue
problem for ‘n’ degrees of freedom of healthy and
cracked beam can be written as
½ðK Þ ki ðM Þ ¼ 0 and ½ðK Þ ki ðM Þ ¼ 0
Normalization of mass matrix of healthy and cracked beam
is given as UTMU and UTM*U.
Modal Curvature Method
It is also one of the damage evaluation methods to detect
the presence, location and severity in a damaged structure
or component by comparison with the healthy structures.
The modal curvatures for healthy and damaged structures
are determined using second-order central difference
method in modal displacement field. There is direct concurrency of damage location with the modal curvature
fields found, and the magnitude is directly proportional to
severity of damage. This method is effective for the lower
modal frequencies.
For instance, consider a cantilever beam with EIas
flexural rigidity, subjected to bending moment as M(x) and
curvature as K(x). Here, ‘x’ indicates the distance of any
point from fixed end of the beam. The curvature for healthy
cantilever beam can be given as
KðxÞ ¼
And the curvature at a distance x for cracked cantilever
beam can be given as
K ðxÞ ¼
KðxÞ ¼
½F hx ¼
1
X
1
U UT
2 xi xi
x
i
i¼1
½F dx ¼
1
X
1
U UT :
2 xi xi
x
i
i¼1
The change in flexibility matrix among the beams will be
D½F ¼ "
½F hx ½F dx # "
#
1
1
X
X
1
1
¼
Uxi UxiT
Uxi UxiT :
x2i
x2i
i¼1
i¼1
hx
M ðxÞ
EI
The mode shape curvatures at local level for healthy and
cracked beam are computed as follows based on the central
difference approximation.
Modal Flexibility Method
This method is one of the modal techniques to identify the
presence, location and severity of damage in structures. It
includes the determination of natural frequencies, mass
normalized mode shapes and flexibility matrix of structures. It yields better results for lower modes and converges
with the higher modes.
The flexibility matrix of healthy and cracked beam at
distance x along its length can be calculated as
MðxÞ
EI
/ðj þ 1Þi 2/ji þ /ðj 1Þi
l2
K ðxÞ ¼
/ðj þ 1Þi 2/ji þ /ðj 1Þi
l2
where i mode shape number, j node number, uji modal
displacement of node j at mode i and l distance between
nodes. The absolute modal curvature difference between
the healthy and cracked beam approximates the location of
damage.
uðj þ 1Þi 2uji þ uðj 1Þi
D½K ¼ ½K ð xÞ K ð xÞ ¼
l2
dx
uðj þ 1Þi 2uji þ uðj 1Þi
l2
hx
Based on the higher value of curvature difference, the
location of the crack can be identified in a beam.
Numerical Method
dx
Here, xi is the ith natural frequency. The increase in the
flexibility of corresponding element is used to detect the
presence and the location of damage.
To detect of presence, location and severity of crack in a
cracked spring, the modal flexibility and modal curvature
methods are applied.
123
J Fail. Anal. and Preven.
2
Modal Flexibility Method
The step-by-step procedure to calculate modal flexibility of
both the springs is given below.
Calculation of Assembled Mass and Stiffness Matrix
The elemental stiffness and mass matrices are assembled
for healthy springs as given below.
2
3
16408:68
9082:8
0
0
6 9082:8
7
19547:52 10464:72
0
7
K¼6
4
0
10464:72
22515
12050:28 5
0
12050:28
12050:28
2 0
3
0:265 0:046
0
0
6 0:046 0:259 0:044
0 7
7
M¼6
4 0
0:044 0:25 0:041 5
0
0
0:041 0:12
The elemental stiffness and mass matrices are assembled
for cracked spring as given below.
2
3
16408:68 9082:8
0
0
6 9082:8 16021:44 6938:64
7
0
7
K ¼ 6
4
0
6938:64 18988:92 12050:28 5
0
12050:28
12050:28
2 0
3
0:265 0:046
0
0
6 0:046 0:242 0:038
0 7
7
M ¼ 6
4 0
0:038 0:23 0:041 5
0
0
0:041 0:12
Determination of Eigenvalues and Eigenvectors by
MATLAB Programming
The eigenvalues and eigenvectors are computed by
MATLAB programming. Table 2 highlights the natural
frequencies and corresponding mode shapes for healthy
and cracked leaf spring.
Determination of Dynamic Flexibility of Healthy and
Cracked Spring
Flexibility matrices of healthy and cracked springs are
calculated after mass normalization.
Table 2 Modal frequencies of healthy and cracked spring
Mode shape
Healthy spring (Hz)
Cracked spring (Hz)
First
61.95
Second
200
198.74
Third
Fourth
346.70
502.72
333.07
467.1
123
58.01
3
0:021 0:026 0:026 0:026
6 0:026 0:043 0:048 0:048 7
4
7
½Fhx ¼ 6
4 0:02 0:048 0:06 0:065 5 10
0:026 0:048 0:065 0:072 3
2
0:019 0:023 0:021 0:021
6 0:023 0:037 0:04 0:039 7
6
7 104
½F dx ¼ 4
0:021 0:04 0:052 0:06 5
0:021 0:04 0:064 0:07
Dynamic Flexibility Change for Healthy and Cracked
Springs
The change in flexibility among both the spring is given
below.
D½F ¼ 2
½F hx ½F dx
0:019 0:023 0:021
6 0:023 0:037 0:04
¼6
4 0:021 0:04 0:052
0:021 0:04 0:064
3
0:021
0:039 7
7104
0:06 5
0:07
The increase in flexibility in corresponding element indicates the presence and location of crack.
Modal Curvature Method
The location of crack is found on the modeled healthy and
cracked cantilever beams after mass normalization. The
mass normalized eigenvectors of healthy beam are as
follows.
2
3
2
3
2
3
0:074
0:154
0:154
6 0:125 7
6 0:081 7
6 0:081 7
7
6
7
6
7
Uh1 ¼ 6
4 0:151 5 Uh2 ¼ 4 0:076 5 Uh2 ¼ 4 0:076 5
0:159
0:142
0:142
2
3
0:104
6 0:109 7
7
Uh3 ¼ 6
4 0:015 5
0:107
The mass normalized eigenvectors of cracked beam are as
follows.
2
3
2
3
2
3
0:065
0:150
0:120
6 0:110 7
6 0:110 7
6 0:12 7
6
6
7
7
7
Ud1 ¼ 6
4 0:148 5 Ud2 ¼ 4 0:08 5 Ud3 ¼ 4 0:011 5
0:156
0:14
0:087
2
3
0:014
6 0:033 7
7
Ud4 ¼ 6
4 0:068 5
0:09
The modal curvature of healthy and cracked beam for all
four nodal points can be calculated through central
difference approximation.
J Fail. Anal. and Preven.
2
3
2
3
1:042
15:76
Kh1 ¼ 4 01:73 5 Kh2 ¼ 4 05:83 5
01:25
2
3
26:32
3
22:01
5:486
Kh3 ¼ 4 21:32 5 Kh4 ¼ 4 12:36 5
01:94
19:58
2
3
2
3
2
3
1:38
13:19
25:0
Kd1 ¼ 4 0:486 5 Kd2 ¼ 4 10:42 5 Kd3 ¼ 4 24:24 5
2:08 3
09:03
0:76
2
4:23
Kd4 ¼ 4 10:28 5
17:99
The absolute modal curvature difference among the healthy
and cracked beam approximates the location of crack for
all four modes are given below
2
3
0:338
D½K 1 ¼ ½Kd1 Kh1 ¼ 4 1:244 5
0:83
2
3
2:57
D½K 2 ¼ ½Kd2 Kh2 ¼ 4 4:59 5
2:71
2
3
2:99
D½K 3 ¼ ½Kd3 Kh3 ¼ 4 02:92 5
02:7
2
3
1:256
D½K 4 ¼ ½Kd4 Kh4 ¼ 4 02:08 5
1:59
Fig. 5 Healthy leaf spring
where D[K]1, D[K]2, D[K]3 and D[K]4 are the absolute
modal curvature differences for the first four mode shapes.
Results and Discussion
The presence, location and severity of crack in composite
leaf spring are evaluated as follows.
Presence of Crack
In order to confirm the presence of crack at a global level,
the modal natural frequency is the suitable parameter. The
comparisons of natural frequencies of cracked spring with
the healthy for lower modes are made. It has been found
that natural frequency is decreased for cracked spring.
Figures 5 and 6 show frequency variation with mode
number for healthy and cracked leaf springs obtained by
analytical, ANSYS and experimental approaches, respectively. Figure 7 shows frequency change among both the
spring by analytical and ANSYS results.
Fig. 6 Cracked leaf spring
Location of Crack
Modal flexibility change and modal curvature difference
parameters are used to locate the position of the intentionally created crack on the mono composite leaf spring
using modal techniques. The modal flexibility change was
found to be maximum between the portions of the beam
0.3–0.4 m as shown in Fig. 8. Similarly, the modal curvature difference was observed to be maximum between
the portions of the beam 0.35–0.40 m as shown in Fig. 9.
123
J Fail. Anal. and Preven.
Fig. 7 Change in frequency among both springs
Fig. 9 Modal curvature method
Fig. 8 Modal flexibility method
Severity of Crack
The severity of crack is quantified in terms of modal
parameters such as modal flexibility and modal curvatures.
The severity of crack for modal flexibility is observed to be
15.38% as shown in Fig. 10. Similarly, severity of crack
for modal curvature found to be is 255.96 and 44.04% for
1st and 2nd modes, respectively, as shown in Fig. 11.
Conclusion
Failure analysis of composite mono leaf spring has been
carried out by modal flexibility and curvature mode shape
methods. The reduction in natural frequency showed the
presence of crack in the spring at global level. The modal
123
Fig. 10 Modal flexibility method
flexibility change was observed in the 3rd element at the
crack location. Similarly, the modal curvature change is
also observed in the same element at the crack region. The
curvature mode shape method is more precise in locating
the crack than the modal flexibility method. Thus, the crack
location can be detected using two independent methods.
The severity of crack was observed at 7th nodal point by
both the methods. The maximum severity was found at 1st
mode due to maximum amplitude of beam in the vicinity of
crack region. The amplitude of mode shape goes on
decreasing with the higher modes resulting in less severity
at the crack region.
J Fail. Anal. and Preven.
Fig. 11 Modal curvature method
References
1. M. He, Y. Tao Yang, Nondestructive identification of composite
beams damage based on the curvature mode difference. J.
Compos. Struct. (2017). https://doi.org/10.1016/j.comp.struct.
2017.05.040
2. X. Zhoua, D. Wangb, M. Duana, J. Gua, Y. Liuaa, Numerical
study on mode curvature for damage detection of a drilling riser
using transfer matrix technique. J. Appl. Ocean Res. 2017(63),
65–75 (2017)
3. Z.-B. Yang, M. Radzienski, P. Kudela, W. Ostachowicz, Fourier
spectral-based modal curvature analysis and its application to
damage detection in beams. J. Mech. Syst. Signal Process.
(2016). https://doi.org/10.1016/j.ymssp.2016.07.005
4. Z.-B. Yang, M. Radzienski, P. Kudela, W. Ostachowicz, Scalewave number domain filtering method for curvature modal
damage detection. J. Compos. Struct. (2016). https://doi.org/10.
1016/j.compstruct.2016.07.074
5. D. Dessi, G. Camerlengo, Damage identification techniques via
modal curvature analysis: overview and comparison. J. Mech.
Syst. Signal Process. 52–53, 181–205 (2015)
6. V.B. Dawari, G.R. Vesmawala, Modal curvature and flexibility
methods for honeycomb damage identification in reinforced
concrete beams. Proc. Eng. 51, 119–124 (2013)
7. S. Russo, Damage assessment of GFRP putruded structural elements. Compos. Struct. 96, 661–669 (2013)
8. E. Ekinovic, S. Ekinovic, R. Sunulahpasic, A Glance to a mode
shape based damage detection technique, in 17th International
Research/Expert Conference TMT 2013, Istanbul, Turkey
9. W. Lestari, P. Qiao, Damage detection of fiber-reinforced polymer honeycomb sandwich beams. Compos. Struct. 67, 365–373
(2005)
10. H.P. Zhu, Y.L. Xu, Damage detection of mono coupled periodic
structures based on sensitivity analysis of modal parameters. J.
Sound Vib. 285, 365–390 (2005)
11. E.-T. Lee, H.-C. Eun, Damage detection approach based on the
second derivative of flexibility estimated from incomplete mode
shape data. Appl. Math. Model. (2017). https://doi.org/10.1016/j.
apm.2017.02.014
12. S.H. Sung, K.Y. Koo, H.J. Jung, Modal flexibility-based damage
detection of cantilever beam-type structures using baseline
modification. J. Sound Vib. 333, 4123–4138 (2014)
13. M.S. Gaith, Nondestructive health monitoring of cracked simply
supported fiber-reinforced composite structures. J. Intell. Mater.
Syst. Struct. 22(18), 2207–2214 (2011)
14. E. Reynders, G. De Roeck, A local flexibility method for vibration- based damage localization and quantification. J. Sound Vib.
329, 2367–2383 (2010)
15. L. Yu, T. Yin, Damage identification in frame structures based on
FE model updating. J. Vib. Acoust. 132(5), 051007 (2010)
16. M.A.B. Abdo, M. Hori, A numerical study of structural damage
detection using changes in the rotation of mode shapes. J. Sound
Vib. 251(2), 227–239 (2002)
17. A.K. Pandey, M. Biswas, Experimental verification of flexibility
difference method for locating damage in structures. J. Sound
Vib. 184(2), 311–328 (1995)
18. M.M.F. Yuen, A numerical study of the Eigen parameters of a
damaged cantilever. J. Sound Vib. 103(3), 301–310 (1985)
19. Society of Automotive Engineers, Inc. 400 Commonwealth
Drive, Warrendale, PA, pp. 15096–0001.
20. Ansys Design nCode Manual, HBM, 2010
123