Structural Engineering and Mechanics, Vol. 53, No.

4 (2015) 725-744
DOI: http://dx.doi.org/10.12989/sem.2015.53.4.725 725

A new damage detection indicator for beams based on mode

shape data
O. Yazdanpanah1a, S.M. Seyedpoor1 and H. Akbarzadeh Bengar2b
1
Department of Civil Engineering, Shomal University, Amol, Iran
2
Department of Civil Engineering, University of Mazandaran, Babolsar, Iran

(Received March 26, 2014, Revised November 7, 2014, Accepted November 25, 2014)

Abstract. In this paper, a new damage indicator based on mode shape data is introduced to identify damage
in beam structures. In order to construct the indicator proposed, the mode shape, mode shape slope and
mode shape curvature of a beam before and after damage are utilized. Mode shape data of the beam are first
obtained here using a finite element modeling and then the slope and curvature of mode shape are evaluated
via the central finite difference method. In order to assess the robustness of the proposed indicator, two test
examples including a simply supported beam and a two-span beam are considered. Numerical results
demonstrate that using the proposed indicator, the location of single and multiple damage cases having
different characteristics can be accurately determined. Moreover, the indicator shows a better performance
when compared with a well-known indicator found in the literature.
Keywords: damage detection; damage indicator; beam structure; mode shape data

1. Introduction

Structural damage detection is of a great importance, because early detection and repair of
damage in a structure can increase its life and prevent it from an overall failure. Structural damage
detection consists of four different levels (Rytter 1993). The first level determines the presence of
damage in the structure. The second level includes locating the damage, while the third level
quantifies the severity of the damage. The final level uses the information from the first three steps
to predict the remaining service life of the damaged structure. After discovering the damage
occurrence, damage localization is more important than damage quantification. In the last years,
numerous methods have been proposed to accurately locate structural damage (Cawley and Adams
1979, Rizos et al. 1990, Doebling et al. 1992, Doebling et al. 1998, Abdo and Hori 2002, Koh and
Dyke 2007, Sinou 2007). The damage location assurance criterion (DLAC) and multiple damage
location assurance criterion (MDLAC) based on the changes in natural frequencies have been
proposed by Messina et al. (1992, 1998). The DLAC and MDLAC criteria are used to locate single
and multiple damages, respectively. The modal assurance criterion (MAC) was used by West

Corresponding author, Ph.D., E-mail: [email protected]

a
M.Sc., E-mail: [email protected]
b
Ph.D., E-mail: [email protected]

Copyright © 2015 Techno-Press, Ltd.

http://www.techno-press.org/?journal=sem&subpage=8 ISSN: 1225-4568 (Print), 1598-6217 (Online)
726 O. Yazdanpanah, S.M. Seyedpoor and H. Akbarzadeh Bengar

(1984) to determine the level of correlation between modes from the test of an undamaged Space
Shuttle Orbiter body flap and the modes from the test of the flap after it has been exposed to an
acoustic loading. The MAC criterion can’t show explicitly where the source of the damage in the
structure lies. Therefore, the coordinate modal assurance criterion (COMAC) has been proposed
by Lieven and Ewins (1988). Damage detection from changes in mode shape curvature has been
proposed by Pandey et al. (1991). The curvature of mode shapes was investigated as a possible
candidate for identifying and locating damage in a beam structure. The results showed that the
absolute changes in the mode shape curvature are localized in the region of damage and hence can
be used to detect damage in the structure. Damage detection in structures using changes in
flexibility has been presented by Pandey and Biswas (1994). Appraisal of changes in the flexibility
matrix of a structure was presented as a candidate method for recognizing the presence and
locating the damage. Some other related criteria based on the mode shape curvature as damage
indicators including mode shape amplitude comparison (MSAC), flexibility index (FI), mode
shape slope (MSS), and mode shape curvature square (MSCS) have been proposed by Ho and
Ewins (2000). The changes of modal flexibility matrix and modal strain energy of flexural
members before and after damage have been used by Shih et al. (2009) as a basis for locating the
structural damage. Results showed that the proposed method is effective in multiple damage
assessment in beam and plate structures. Locating structural damage through an adaptive neuro-
fuzzy inference system (ANFIS) has been introduced by Fallahian and Seyedpoor (2010). Damage
detection using an efficient correlation based index and a modified genetic algorithm has been
proposed by Nobahari and Seyedpoor (2011). A two-stage method for determining structural
damage sites and extent using a modal strain energy based index (MSEBI) and particle swarm
optimization (PSO) has been proposed by Seyedpoor (2012). An efficient method for structural
damage localization based on the concepts of flexibility matrix and strain energy of a structure has
been suggested by Nobahari and Seyedpoor (2013). An efficient indicator for structural damage
localization using the change of strain energy based on static noisy data (SSEBI) has been
proposed by Seyedpoor and Yazdanpanah (2014). The acquired results clearly showed that the
proposed indicator can precisely locate the damaged elements.
The main purpose of this study is to introduce an efficient mode shape data based indicator for
estimating the damage location in beam structures. Two test examples with and without
considering measurement noise are considered to assess the efficiency of the proposed indicator
for accurately locating the damage. Numerical results demonstrate that the proposed indicator can
well determine the location of single and multiple damage cases having different characteristics.

2. Vibration based damage detection techniques

There are various vibration based damage detection techniques that can be utilized to identify
the damage in a beam structure. In this section, three more common methods including natural
frequency based methods, mode shape based methods and mode shape curvature based methods
are described briefly.

2.1 Natural frequency based methods

The change of natural frequencies of a structure before and after damage can be considered as
an index for identifying structural damage. One of the most advantages of this technique is that
 A new damage detection indicator for beams based on mode shape data 727

frequency measurements can be quickly and easily conducted (Gholizadeh and Barzegar 2013).
However, there are quite a few limitations for using natural frequencies. The practical frequencies
are a global parameter and do not provide any spatial information about a structure. Also, the
sensitivity of frequencies to damage is relatively low, especially for large structures. Numerous
researchers used the changes in natural frequencies for damage detection. For example, Messina et
al. (1992) proposed the damage location assurance criterion (DLAC) as follows

TA B (i)

DLAC (i) 
2

(TA A ) (B (i )T B (i))

(1)

where ΔωA is the experimental frequency shift vector and ΔωB is the theoretical frequency change
vector for damage that is situated at the ith position.

2.2 Mode shape based methods

The mode shapes are another basic dynamic characteristic of a structure. Mode shape vectors
are a spatial characteristic and can provide information regarding both the existence and location
of damage. Mode shape changes before and after damage can be considered as one of prevalent
damage detection indicator in structures. For example, the coordinate modal assurance criterion
(COMAC) is a mode shape based damage indicator proposed by Lieven and Ewins (1988). The
COMAC factor at a point i between two sets of mode shape is given by

  ( A )ij ( B )ij 
 nm 
2

 
COMACi  nm 
j 1

 ( A )ij2 .  ( B )ij2
(2)
nm

j i j i

where nm defines the number of correlated mode shapes. Also, (ΦA)ij and (ΦB)ij indicate the value
of the jth mode shape at a point i of the structure for the states A and B, respectively.

2.3 Mode shape curvature based methods

Mode shape derivatives such as curvature are widely used as an alternative to mode shape in
order to obtain spatial information about vibration characteristic changes for damage
identification. It is first noted that for beams, plates and shells there is a direct relationship between
curvature and bending strain. The value of curvature at a point in the structure is equal to M/EI, so
if the stiffness at a point reduces by damage, the curvature at that point will increase. This can be
used to detect and locate damage in the structure. Pandey et al. (1991) demonstrated that the
absolute change in mode shape curvature can be used as an efficient indicator of damage. They
estimated the mode shape curvature using the central finite difference approximation (the second
derivative of mode shape) as
 q 1,i  2 q ,i   q 1,i
q ,i  (3)
le 2
where le is the distance between the measurement co-ordinates or it can be the element length.
728 O. Yazdanpanah, S.M. Seyedpoor and H. Akbarzadeh Bengar

Fig. 1 A simply supported beam having a crack located at xcr from the left end

Fig. 2 Variation of EI due to the crack in an element with length le

Also, Φq,i represents the modal displacement for the ith mode shape at the measurement co-ordinate
q. Then, they proposed a damage indicator named mode shape curvature (MSC) as

MSCq   ( damaged )"  ( undamaged

nm
)" (4)
i 1
q ,i q ,i

where nm is the number of modes to be considered.

Moreover, for considering the effect of higher modes of a structure, Abdel Wahab and De
Roeck (1999) proposed the curvature damage factor (CDF) as

CDFq   ( damaged )"  ( qundamaged

1 nm
)" (5)
nm i 1
q ,i ,i

3. Damage (crack) modeling in beams

In this study, it is assumed that damage occurs by a transverse surface crack located at xcr from
the left end of a beam as shown in Fig. 1. For crack modeling, a fully open transverse surface
crack model, illuminated by Sinha et al. (2002), is adopted. The effect of the crack on the mass is
small and can be neglected. The crack only leads to local stiffness reduction in a specified length
adjacent to the crack. It is assumed that the reduction of stiffness due to the crack is inside one
 A new damage detection indicator for beams based on mode shape data 729

element. Considering one cracked element as shown in Fig. 2, the flexural rigidity EI of the
cracked element varies linearly from the cracked position towards both sides in an effective length
lc. The stiffness matrix of the damaged element can be represented as
K ce  K ue  K ce (6)

where K ue represents the element stiffness matrix of the intact element; K ce is the stiffness
reduction on the intact element stiffness matrix due to the crack. According to Euler–Bernoulli
beam element, the element stiffness matrix of the intact beam is expressed as
6 3le 
 le2 
3le -6

K ue  3 o 
2
2le -3le
le  symmetric 6 -3le 
2EI (7)
 
 2le2 

By using the linear variation of EI as proposed by Sinha et al. (2002), the reduction on the
element stiffness matrix can be obtained as
 K11 K12 K11 K14 
 K 22 K12 K 24 
K ce  
 symmetric K11 K14 
(8)
 
 K 44 

where the stiffness factors are given by

12 E ( I o  I c )  2lc3  2 xc  
K11   2  3lc   1  ,
2

I e4  le  le  

12 E ( I o  I c )  lc3  7 xc 6 xc2  
K12    
c   2  ,
 le  le  
l 2
I e3 2
le

12 E ( I o  I c )  lc3  5 xc 6 xc2  
K14    
c 1   2  ,
 le  le  
l
I e3 2
le

2 E ( I o  I c )  3lc3  3xc  
K 22      
 ,
2

c
 le  le  
2l 2
I e2 2

2 E ( I o  I c )  3lc3  9 xc 9 xc2 
K 24   2  
c   2  ,
 le  le  
2l 2
I e2 le

2 E ( I o  I c )  3lc3  3xc  
K 44   2  2lc   1  ,
2

I e2  le  le   (9)
where xc is the crack location in the local coordinate, le is the length of the element and lc is the
effective length of the stiffness reduction. The value of lc is assumed to be 1.5 times the beam
730 O. Yazdanpanah, S.M. Seyedpoor and H. Akbarzadeh Bengar

height. Also, E is the Young’s modulus, Io=wh3/12 and Ic=w(h−hc)3/12 are the moment of inertia
of the intact and cracked cross sections, respectively; w and h are the width and height of the intact
beam and hc is the crack depth. For cases of more than one cracked elements, the same procedure
can be followed. The global stiffness matrix of the beam Kc is obtained by assembling the element
stiffness matrices including those of cracked elements.

4. The proposed damage detection indicator

In this paper, damage detection of a prismatic beam with a specified length is studied. First, the
beam is divided into a number of finite elements. Then, mode shapes of the healthy beam in
measurement points are evaluated using the finite element method. A MATLAB (R2010b) code is
prepared here for this purpose. Henceforward, consider the nodal coordinates (xq, q=1,2,...,n+1)

x ,    ( x ,  
and ith mode shape (Фh(q,i), q=1,2,...,n+1) obtained for the healthy beam as follows

q h ( q ,i ) 1 h (1, i ) ), ( x2 ,  h ( 2,i ) ), . . ., ( xn 1 ,  h ( n 1,i ) )

Now by having the mode shapes, the slope of mode shapes (the first derivative of mode shapes,
Ф′=dФ/dx) of healthy beam can be achieved using the central finite difference approximation as
 h ( q 1,i )   h ( q 1,i )
h ( q ,i )  (10)
2le
where i, q and le are defined in Eq.(3).
Also, the mode shape curvature (the second derivative of mode shapes) of healthy beam can
now be determined using the central finite difference approximation as
 h ( q 1,i )  2 h ( q ,i )   h ( q 1,i )
h ( q ,i )  (11)
le 2
This process can also be repeated for damaged beam. It should be noted, it is assumed that the
damage decreases the stiffness and therefore can be simulated by a reduction in the moment of

an element. So, consider the nodal coordinates and ith mode shape (  d ( q ,i ) , q  1, 2, ..., n  1 )
inertia (I) at the location of damage. In this paper, it is supposed the damage occurs in the center of

x ,    ( x ,  
obtained for the damaged beam as follows:

q d ( q ,i ) 1 d (1, i ) ), ( x2 ,  d ( 2,i ) ), . . ., ( xn 1 ,  d ( n 1,i ) )

Now by having the damaged mode shapes, the mode shape slope of damaged beam can be
achieved using the central finite difference method as
 d ( q 1,i )   d ( q 1,i )
d ( q ,i )  (12)
2le
Also, the mode shape curvature of damaged beam can now be approximated as
 d ( q 1,i )  2 d ( q ,i )   d ( q 1,i )
d ( q ,i )  (13)
le 2
 A new damage detection indicator for beams based on mode shape data 731

Finally, using the dynamic responses (mode shapes, slope and curvature of mode shapes) obtained


  h ( q ,i )    d ( q ,i )      d ( q ,i )  h ( q ,i ) 
for two above states, a new mode shape data based indicator (MSDBI) is introduced here as

    
  h ( q ,i ) 

nm 2
2

MSDBI q  i 1
d ( q ,i )
(14)
nm
where nm is the number of mode shapes considered.
Assuming that the set of the MSDBI of all points (MSDBIq, q=1,2,...,n+1) represents a sample
population of a normally distributed variable, a normalized form of MSDBI can be defined as
follows
  MSDBI q  mean ( MSDBI ) 
nMSDBI q  max 0 ,   , q  1,2,..., n  1
  
(15)
std ( MSDBI )
where MSDBIq is defined by Eq. (14). Also, mean (MSDBI) and std (MSDBI) represent the mean
and standard deviation of (MSDBIq, q=1,2,...,n+1), respectively.
In this paper, the result of proposed indicator given by Eq. (15) is compared with that proposed
by Abdel Wahab and De Roeck (1999) given by Eq. (5). Therefore, it is required the CDF to be
normalized in a similar way as
  CDFq  mean (CDF ) 
nCDFq  max 0 ,   , q  1,2,..., n  1
  
(16)
std (CDF )

5. Numerical examples

In order to assess the efficiency of the proposed indicator for damage detection of beams, the
result of nMSDBI is compared with that of nCDF. For this, two test examples including a simply
supported beam and a two-span beam are considered. Various parameters that may affect the
performance of the method are studied.

5.1 Example 1: a simply supported beam

A simply supported beam with span L=6 m shown in Fig. 3 is selected as the first example.
Parameters of the beam consist of width w=0.2 m, height h=0.2 m, elasticity modulus E=200 GPa
and the mass density ρ=7850 kg/m3. The beam is discretized by twenty 2D-beam elements leading
to 40 DOFs. In order to assess the efficiency of the proposed indicator, four different damage
cases listed in Table 1 are considered. It should be noted that damage in the cracked beam is
simulated here by reducing the height and thereinafter the inertia moment (I) at the crack location.

5.1.1 Damage identification without considering noise

Damage identification charts of the simply supported beam for cases 1 to 4 are shown in Figs.
4-7, respectively. As shown in the figures, the value of nMSDBI is further in vicinity of some
elements that this indicates, damage occurs in these elements. As shown in the damage
identification charts of cases 1 to 4, for precisely locating the damage, 1, 2, 3 and 4 vibration
732 O. Yazdanpanah, S.M. Seyedpoor and H. Akbarzadeh Bengar

Fig. 3 (a) Geometry of the simply supported beam (b) Cross-section of the beam

Table 1 Four different damage cases induced in simply supported beam

Case 1 Case 2 Case 3 Case 4
Element Damage Element Damage Element Damage Element Damage
number ratio* number ratio number ratio number ratio
2 0.10 5 0.10 3 0.20 4 0.15
- - 17 0.15 10 0.15 9 0.10
- - - - 16 0.10 13 0.20
- - - - - - 18 0.10
*Damage ratio is hc where hc is the crack depth
h

Fig. 4 Damage identification chart of 20-element beam for damage case 1 considering one mode

modes of the beam are necessitated to be considered, respectively. It should be noted that with
considering higher vibration modes, the results of nMSDBI and nCDF do not change significantly.
Hence, the study of considering the higher vibration modes after exact identification has been
neglected.
In addition, for verifying the proposed indicator, the result of nMSDBI can be compared with
that of nCDF. As can be observed in the figures, the efficiency of the proposed indicator for
damage localization is high when comparing with the available damage indicator.

5.1.2 The effect of measurement noise

Measurement noise can’t be avoided practically. Hence, the effect of noise is considered here
by perturbing the responses of damaged structure. In this example, 3% noise is assumed in case 1.
 A new damage detection indicator for beams based on mode shape data 733

(a)

(b)
Fig. 5 Damage identification chart of 20-element beam for damage case 2 considering: (a) one mode,
(b) two modes

(a)

(b)
Fig. 6 Damage identification chart of 20-element beam for damage case 3 considering: (a) one mode,
(b) two modes and (c) three modes
734 O. Yazdanpanah, S.M. Seyedpoor and H. Akbarzadeh Bengar

(c)
Fig. 6 Continued

(a)

(b)

(c)
Fig. 7 Damage identification chart of 20-element beam for damage case 4 considering: (a) one mode,
(b) two modes, (c) three modes and (d) four modes
 A new damage detection indicator for beams based on mode shape data 735

(d)
Fig. 7 Continued

Fig. 8 Damage identification chart of 20-element beam for damage case 1 considering: one mode and
3% noise

Fig. 9 (a) Geometry of the two-span continuous beam (b) Cross-section of the beam

Fig. 8 shows damage identification chart for the damage scenario 1 considering 3% noise. When
comparing it with the one shown in Fig. 4 for scenario 1 (state without noise), it can be indicated
that there is a good compatibility between them. In other words, the noise has a negligible effect
on the performance of nMSDBI.

5.2 Example 2: a two-span beam

An indeterminate beam with two spans shown in Fig. 9 is considered as the second example.
Parameters of the beam are length L=12 m, width w=0.2 m, height h=0.2 m, elasticity modulus
736 O. Yazdanpanah, S.M. Seyedpoor and H. Akbarzadeh Bengar

Table 2 Four different damage cases induced in two-span continuous beam

Case 1 Case 2 Case 3 Case 4
Element Damage Element Damage Element Damage Element Damage
number ratio number ratio number ratio number ratio
21 0.20 8 0.10 2 0.10 3 0.15
- - 34 0.10 21 0.15 12 0.20
- - - - 37 0.20 20 0.15
- - - - - - 39 0.10

(a)

(b)

(c)
Fig. 10 Damage identification chart of 40-element beam for damage case 1 considering: (a) one mode,
(b) two modes and (c) three modes

E=200 GPa and the mass density ρ=7850 kg/m3. The beam is discretized by forty 2D-beam
elements leading to 79 DOFs. For assessing the robustness of the proposed indicator, four different
damage cases, including single and multiple damage cases, listed in Table 2 are considered.
 A new damage detection indicator for beams based on mode shape data 737

5.2.1 Damage identification without considering noise

Damage identification charts of the beam for cases 1 to 4 are shown in Figs. 10-13,
respectively. As shown in the figures, the value of nMSDBI is further in vicinity of some elements
that this indicates there is damage in these elements. As shown in the damage identification charts
of cases 1 to 4, for locating the damage precisely, 3, 5, 6 and 8 vibration modes of the beam are
required to be considered, respectively. It should be noted that with considering the higher
vibration modes, the results of nMSDBI and nCDF do not change. Hence, the effect of higher
vibration modes after exact identification has been neglected.
In addition, for verifying the proposed indicator, the performance of nMSDBI is compared with
that of nCDF. As can be observed in the figures, the proposed indicator is capable of identifying
all damage cases correctly, but when damage occurs in the elements that have placed in the
vicinity of the middle support, the nCDF indicates other elements which damage don’t happen in
them. It seems that the use of nMSDBI for these elements can be better than nCDF. In addition, as
can be observed in Fig. 10 (c), even with considering three modes of the beam, the nCDF can not
indicate the crack location correctly. In other word, three damaged elements are identified by this
figure. It reveals that the results obtained by the proposed indicator are more reliable.
As can be observed in Fig. 11 (e), even with considering five vibration modes for the structure,
the result of nCDF can not show the crack location properly. In other word, six elements are
introduced by this figure as damaged ones.

5.2.2 The effect of measurement noise

Measurement noise can’t be avoided. Hence, the effect of noise is considered here by

(a)

(b)
Fig. 11 Damage identification chart of 40-element beam for damage case 2 considering: (a) one mode,
(b) two modes, (c) three modes, (d) four modes and (e) five modes
738 O. Yazdanpanah, S.M. Seyedpoor and H. Akbarzadeh Bengar

(c)

(d)

(e)
Fig. 11 Continued

(a)
Fig. 12 Damage identification chart of 40-element beam for damage case 3 considering: (a) one mode,
(b) two modes, (c) three modes, (d) four modes, (e) five modes and (f) six modes
A new damage detection indicator for beams based on mode shape data 739

(b)

(c)

(d)

(e)
Fig. 12 Continued
740 O. Yazdanpanah, S.M. Seyedpoor and H. Akbarzadeh Bengar

(f)
Fig. 12 Continued

(a)

(b)

(c)
Fig. 13 Damage identification chart of 40-element beam for damage case 4 considering: (a) one mode,
(b) two modes, (c) three modes, (d) four modes, (e) five modes, (f) six modes, (g) seven modes and (h)
eight modes
A new damage detection indicator for beams based on mode shape data 741

(d)

(e)

(f)

(g)
Fig. 13 Continued
742 O. Yazdanpanah, S.M. Seyedpoor and H. Akbarzadeh Bengar

(h)
Fig. 13 Continued

Fig. 14 Damage identification chart of 40-element beam for damage case 4 considering 8 modes and
3% noise

perturbing the responses of damaged structure. In this example, 3% noise is assumed in case 4 of
Table 2. Fig. 14 shows damage identification chart for the damage scenario 4 considering 3%
noise. When comparing it with the one shown in Fig. 13 for scenario 4 (state without noise), it can
be indicated that there is a good compatibility between them. In other words, the measurement
noise has a negligible effect on nMSDBI.

6. Conclusions

In this paper, damage identification of beams using a mode shape data based indicator
(MSDBI) proposed here has been investigated. The effects of many parameters, may affect the
efficiency of the damage indicator, have been assessed with considering a simply supported and an
indeterminate beam as test examples. Based on the numerical studies, the following results can be
concluded:
• The MSDBI is sensitive to the stiffness reduction (moments of inertia), in other words, it has
characteristics from damaged area and can be used as a good indicator for damage detection. As
the identification charts revealed, the proposed indicator could identify all damage scenarios
properly.
 A new damage detection indicator for beams based on mode shape data 743

• In order to verify the proposed indictor, the results of MSDBI have been compared with those
of a well-known indicator, CDF. Results have shown a better performance for the proposed
indicator when compared to that of CDF.
• One of important characteristics of the proposed indicator is that it can precisely detect
damaged elements adjacent to middle supports in continues beams. Whereas the CDF identifies
other elements which damage don’t occur in them. It seems that the use of MSDBI for these cases
is better than CDF.
• As can be observed in identification charts, with increasing damaged elements, more vibration
modes were required to be considered to exactly identify the elements. Of course, after
identification of all damaged elements by a few modes, by considering the higher vibration modes,
the results of MSDBI do not change significantly. In other words, the indicator has a high
convergence in terms of number of modes.
•Measurement noise has a negligible effect on the efficiency of the proposed indicator for
damage detection.

References

Abdo, M.A.B. and Hori, M. (2002), “A numerical study of structural damage detection using changes in the
rotation of mode shapes”, J. Sound Vib., 251(2), 227-239.
Abdel Wahab, M.M. and De Roeck, G. (1999), “Damage detection in bridges using modal curvatures:
application to a real damage scenario”, J. Sound Vib., 226(2), 217-235.
Cawley, P. and Adams, R.D. (1979), “The location of defects in structures from measurements of natural
frequency”, J. Strain Anal. Eng. Des., 14(2), 49-57.
Doebling, S.W., Farrar, C.R., Prime, M.B. and Shevits, D.W. (1996), “Damage identification and health
monitoring of structural and mechanical systems from changes in their vibration characteristics: A
literature review”, Los Alamos Natl. Lab., USA, 1, 1-136.
Doebling, S.W., Farrar, C.R. and Prime, M.B. (1998), “A summary review of vibration-based damage
identification methods”, Shock Vib. Dig., 30(2), 91-105.
Fallahian, S. and Seyedpoor, S.M. (2010), “A two stage method for structural damage identification using an
adaptive neuro-fuzzy inference system and particle swarm optimization”, Asian J. Civil Eng., Build.
Hous., 11(6), 797-810.
Gholizadeh, S. and Barzagar, A. (2013), “Shape optimization of structures for frequency constraints by
sequential harmony search algorithm”, Eng. Optim., 45(6), 627-646.
Ho, Y.K. and Ewins, D.J. (2000), “On the structural damage identification with mode shapes”, Proceedings
of the European COST F3 Conference on System Identification and Structural Health Monitoring, 1,
Madrid, Spain.
Koh, B.H. and Dyke, S.J. (2007), “Structural health monitoring for flexible bridge structures using
correlation and sensitivity of modal data”, Comput. Struct., 85(3-4), 117-130.
Lieven, N.A.J. and Ewins, D.J. (1988), “Spatial correlation of mode shapes: the coordinate modal assurance
criterion (COMAC)”, Proceedings of the 6th International Modal Analysis Conference, 1, Kissimmee,
Florida, USA.
Messina, A., Jones, I.A. and Williams, E.J. (1992), “Damage detection and localization using natural
frequency changes”, Proceedings of the Conference on Identification in Engineering System, Cambridge,
UK, 1, 67-76.
Messina, A., Williams, E.J. and Contursi, T. (1998), “Structural damage detection by a sensitivity and
statistical-based method”, J. Sound Vib., 216(5), 791-808.
Nobahari, M. and Seyedpoor, S.M. (2011), “Structural damage detection using an efficient correlation based
index and a modified genetic algorithm”, Math. Comput. Model., 53(9-10), 1798-1809.
744 O. Yazdanpanah, S.M. Seyedpoor and H. Akbarzadeh Bengar

Nobahari, M. and Seyedpoor, S.M. (2013), “An efficient method for structural damage localization based on
the concepts of flexibility matrix and strain energy of a structure”, Struct. Eng. Mech., 46(2), 231-244.
Pandey, A.K., Biswas, M. and Samman, M.M. (1991), “Damage detection from changes in curvature mode
shapes”, J. Sound Vib., 145(2), 321-332.
Pandey, A.K. and Biswas, M. (1994), “Damage detection in structures using changes in flexibility”, J. Sound
Vib., 169(1), 3-17.
Rytter, A. (1993), “Vibration Based Inspection of Civil Engineering Structures”, PhD Thesis, Aalborg
University, Denmark.
Rizos, P.F., Aspragathos, N. and Dimarogonas , A.D. (1990), “Identification of crack location and
magnitude in a cantilever beam from the vibration modes”, J. Sound Vib., 138(3), 381-388.
Sinou, J.J. (2007), “a robust identification of single crack location and size only based on pulsations of the
cracked systems”, Struct. Eng. Mech., 25(6), 691-716.
Shih, H.W., Thambiratnam, D.P. and Chan, T.H.T. (2009), “Vibration based structural damage detection in
flexural members using multi-criteria approach”, J. Sound Vib., 323(3-5), 645-661.
Seyedpoor, S.M. (2012), “A two stage method for structural damage detection using a modal strain energy
based index and particle swarm optimization”, Int. J. Nonlin. Mech., 47(1), 1-8.
Seyedpoor, S.M. and Yazdanpanah, O. (2014), “An efficient indicator for structural damage localization
using the change of strain energy based on static noisy data”, Appl. Math. Modelling, 38(9-10), 2661-
2672.
Sinha, J.K., Friswell, M.I. and Edwards, S. (2002), “Simplified models for the location of cracks in beam
structures using measured vibration data”, J. Sound Vib., 251(1), 13-38.
West, W.M. (1984), “Illustration of the use of modal assurance criterion to detect structural changes in an
orbiter test specimen”, Proceedings of the Air Force Conference on Aircraft Structural Integrity, 1, Palm
Springs, USA.
MATLAB (R2010b), The language of technical computing (software), Math Works Inc.

CC