You are on page 1of 25

This article was downloaded by: [National Institute of Technology - Rourkela

]
On: 21 June 2013, At: 04:46
Publisher: Taylor & Francis
Informa Ltd Registered in England and Wales Registered Number: 1072954 Registered office: Mortimer House,
37-41 Mortimer Street, London W1T 3JH, UK
Mechanics of Advanced Materials and Structures
Publication details, including instructions for authors and subscription information:
http://www.tandfonline.com/loi/umcm20
Nonlinear Vibration Analysis of Simply Supported
Stiffened Plate by a Variational Method
Anirban Mitra
a
, Prasanta Sahoo
a
& Kashinath Saha
a
a
Department of Mechanical Engineering, Jadavpur University, Kolkata, India
Accepted author version posted online: 24 May 2012.Published online: 05 Mar 2013.
To cite this article: Anirban Mitra , Prasanta Sahoo & Kashinath Saha (2013): Nonlinear Vibration Analysis of Simply Supported
Stiffened Plate by a Variational Method, Mechanics of Advanced Materials and Structures, 20:5, 373-396
To link to this article: http://dx.doi.org/10.1080/15376494.2011.627640
PLEASE SCROLL DOWN FOR ARTICLE
Full terms and conditions of use: http://www.tandfonline.com/page/terms-and-conditions
This article may be used for research, teaching, and private study purposes. Any substantial or systematic
reproduction, redistribution, reselling, loan, sub-licensing, systematic supply, or distribution in any form to
anyone is expressly forbidden.
The publisher does not give any warranty express or implied or make any representation that the contents
will be complete or accurate or up to date. The accuracy of any instructions, formulae, and drug doses should
be independently verified with primary sources. The publisher shall not be liable for any loss, actions, claims,
proceedings, demand, or costs or damages whatsoever or howsoever caused arising directly or indirectly in
connection with or arising out of the use of this material.
Mechanics of Advanced Materials and Structures (2013) 20, 373–396
Copyright C Taylor & Francis Group, LLC
ISSN: 1537-6494 print / 1537-6532 online
DOI: 10.1080/15376494.2011.627640
Nonlinear Vibration Analysis of Simply Supported Stiffened
Plate by a Variational Method
ANIRBAN MITRA, PRASANTA SAHOO, and KASHINATH SAHA
Department of Mechanical Engineering, Jadavpur University, Kolkata, India
Received 28 July 2010; accepted 4 April 2011.
Large-amplitude free vibration analysis of uni-axially single stiffened rectangular plates subjected to transverse loading with simply
supported boundary conditions has been presented. The mathematical formulation is based on energy principle. Geometric nonlin-
earity is accounted for by consideration of nonlinear strain-displacement relations. The static problem is solved through an iterative
scheme and the dynamic problem is solved with the static displacement field as an initial guess. Influence of stiffener position, plate
aspect ratio, and stiffener to plate thickness ratio on the large amplitude dynamic behavior has been studied. The dynamic behavior
has been furnished in the form of backbone curves in a dimensionless frequency-amplitude plane.
Keywords: stiffened plate, nonlinear vibration, geometric nonlinearity, energy methods, backbone curves
Nomenclature
a length of the plate
b width of the plate
b
s
width of the stiffener
{d} vector of unknown coefficients
D flexural rigidity of the plate
e distance between the centroidal axis of the
plate and stiffener
E
p
elastic modulus of the plate material
E
s
elastic modulus of the stiffener material
{f } load vector
[K] stiffness matrix
[K
b
] stiffness matrix due to bending
[K
m
] stiffness matrix due to stretching
[K
sx
]
p
, [K
sy
]
q
stiffness matrices of the p-th stiffener
along x-direction and q-th stiffener along y-
direction, respectively
[M] mass matrix
[M
p
] mass matrix for plate
[M
sx
], [M
sy
] mass matrix for x- and y-direction stiffeners
nw, nu, nv number of constituent functions for w, u, and
v, respectively
p uniformly distributed transverse load
P concentrated transverse load
t
p
thickness of the plate
t
s
thickness of the stiffener
T total kinetic energy of the system
T
p
kinetic energy of the plate
Address correspondence to Prasanta Sahoo, Department of Me-
chanical Engineering, Jadavpur University, Raja S C Mullik
Road, Kolkata 700032, India. E-mail: psjume@gmail.com
T
s
total kinetic energy of the stiffeners
T
p
sx
, T
q
sy
kinetic energy of the p-th stiffener along x-
direction and q-th stiffener along y-direction,
respectively
u displacement along x-direction
U total strain energy of the system
U
b
strain energy due to bending of the plate
U
m
strain energy due to stretching of the plate
U
p
strain energy of the plate
U
s
strain energy of the stiffeners
U
p
sx
, U
q
sy
strain energy of the p-th stiffener along x-
direction and q-th stiffener along y-direction,
respectively
v displacement along y-direction
V work potential
w transverse displacement

i
set of functions defining approximate displace-
ment field u

i
set of functions defining approximate displace-
ment field v
␦ variational operator
ε predefined value of error limit for the iteration
scheme

i
set of functions defining approximate displace-
ment field w

i
set of temporal functions
␩ normalized coordinates in y-direction
␭ relaxation parameter
␲ total energy of the system

p
density of the plate material

s
density of the stiffener material
␶ time coordinate
␷ Poisson’s ratio
␻ natural frequency of the system
D
o
w
n
l
o
a
d
e
d

b
y

[
N
a
t
i
o
n
a
l

I
n
s
t
i
t
u
t
e

o
f

T
e
c
h
n
o
l
o
g
y

-

R
o
u
r
k
e
l
a
]

a
t

0
4
:
4
6

2
1

J
u
n
e

2
0
1
3

374 A. Mitra et al.

1
first linear natural frequency

nl
nonlinear frequency
␰ normalized coordinates in x-direction
1. Introduction
Large amplitude dynamic behavior of various structural ele-
ments differs significantly from its linear behavior. This is due
to the fact that the stiffness of a deformed structure changes
appreciably due to the effect of geometric nonlinearity. So, in
the case of the loaded natural frequencies, the behavior is also
significantly different from the linear counterparts. This leads
to the inference that linear free vibration analysis, carried
out to predict the dynamic response of a system, is not suf-
ficient to estimate the large amplitude vibration frequencies.
In the present article, the study of the large deformation free
vibration analysis of stiffened plates is presented. Stiffened
plates have wide applications in many branches of modern
civil, mechanical, and aerospace engineering. They are fre-
quently used to construct marine structures, such as floors of
bridges, bridge decks, ship hulls, etc., and aircraft structures.
Research work on dynamic behavior of stiffened plates
has gone through different phases and the evolution pro-
cess can be traced back with the help of review works of
Mukherjee and Mukhopadhyay [1], Mukhopadhyay and
Mukherjee [2], and Bedair [3]. Nonlinear analysis of stiffened
plates has been carried out by different researchers through
different techniques and methodologies. One of the earlier
studies on vibration of such plates was taken up by Wah [4]. He
used the orthotropic plate approach to evaluate the vibration
characteristics of beam-plate systems. The natural frequencies
of the first symmetric and first antisymmetric modes of a sim-
ply supported rectangular plate were determined by Kirk [5].
Pratap and Varadan [6] dealt with the large amplitude, free
flexural vibrations of clamped and simply supported stiffened
plates with two different in-plane edge conditions—movable
and immovable. The solutions were obtained on the basis of a
single termvibration mode shape by making use of Galerkin’s
method. Varadan and Pandalai [7] investigated large ampli-
tude free flexural vibrations of eccentrically stiffened elastic
rectangular plates with clamped boundary conditions and
movable in-plane edge conditions. They observed hardening
type nonlinearity for the system. A free vibration character-
istic of rectangular stiffened plates having a single stiffener
was examined by using the finite difference method by Aksu
and Ali [8]. Using a variational technique, they minimized the
total energy of the stiffened plate with respect to discretized
displacement components and obtained natural frequencies
and mode shapes as the solutions of a linear algebraic eigen-
value problem. Aksu [9] later included the effect of in-plane
inertia into free vibration analysis of stiffened plates. Koko
and Olson [10] developed a new numerical technique for large
deflection elastoplastic analysis of stiffened plates using super
finite elements. They also carried out free vibration analy-
sis of rectangular plates [11] with discrete stiffener utilizing
the super elements and including the effects of bending and
in-plane effects as well as beam torsion and lateral bending
motion. Harik and Guo [12] put forward a compound finite
element model to investigate free vibration of eccentrically
stiffened plates. Chen et al. [13] presented a spline compound
strip method for analyzing the free vibration problem of stiff-
ened plates. Vibration and stability analysis of stiffened plates
through a semi-analytical finite difference method was per-
formed by Mukhopadhyay [14, 15] with and without con-
sidering the effects of axial displacements. Mukherjee and
Mukhopadhyay [16] introduced an isoparametric quadratic
stiffened plate bending element for the free vibration analysis
of eccentrically stiffened plates. They included the effect of
shear deformation in the formulation and studied the effects
of eccentricity, shape, and torsional stiffness on the natural
frequencies of the system. Bedair and Troitsky [17] studied
the fundamental frequency characteristics of eccentrically and
concentrically stiffened simply supported plates on the basis
of energy formulation and mathematical programming tech-
nique. Bedair [18] also presented a methodology for deter-
mination of the lowest natural frequency of stiffened plates,
where the nonlinear strain energy function was transformed
into an unconstrained optimization problem and sequential
quadratic programming was used to achieve the solution. Ge-
ometric nonlinear analysis of stiffened plates was done by
Sheikh and Mukhopadhyay [19] utilizing the spline finite strip
method and von Karman nonlinear plate theory. The same au-
thors, in an earlier work, had undertaken free vibration anal-
ysis of eccentrically and concentrically stiffened plates having
arbitrary shapes [20]. More recently, Peng et al. [21] analyzed
stability and free vibration problem of stiffened plates via the
first-order shear deformation theory and mesh-free Galerkin
method. Sapountzakis andMokos [22] developedanimproved
model presenting a general solution for the dynamic analysis
of plates stiffened by arbitrarily placed parallel beams. Asemi-
analytical solution method based on the state-vector equation
theory for free vibration analysis of stiffened laminated plates
was introduced by Qing et al. [23]. Dozio and Ricciardi [24]
proposed a combined analytical-numerical method for pre-
diction of modal characteristics of rectangular plates rein-
forced by a small number of light stiffeners. They used the
assumed modes method to formulate the equations of motion
of the plate and ribs separately and obtained the equations
for the built-up structure by enforcing appropriate continuity
conditions at the interface. Xu et al. [25] introduced an an-
alytical method for the vibration analysis of plates stiffened
by any number of beams of arbitrary lengths and placement
angles considering coupling at an interface of the plate and
beam.
A review of the literature indicates that studies are pri-
marily focused on the free vibration problem, and quite a few
different techniques have been used to determine the natural
frequencies and their corresponding mode shapes. However,
the emphasis is mainly on developing different methodologies
to obtain the natural frequencies compared to studying the
variations in the dynamic behavior corresponding to changes
in system geometry and boundary conditions. It is also noted
that, works on large amplitude dynamic behavior, specifically
variation of loaded natural frequency with amplitude are
quite limited. The present study deals with large amplitude
dynamic behavior of stiffened plates and looks into amplitude
D
o
w
n
l
o
a
d
e
d

b
y

[
N
a
t
i
o
n
a
l

I
n
s
t
i
t
u
t
e

o
f

T
e
c
h
n
o
l
o
g
y

-

R
o
u
r
k
e
l
a
]

a
t

0
4
:
4
6

2
1

J
u
n
e

2
0
1
3

Nonlinear Vibration Analysis of Stiffened Plates 375
dependency of the loaded natural frequency. The nature of
change in the mode shape with increase in vibration amplitude
is also investigated. The effect of variation of systemgeometry
(namely, stiffener position, plate aspect ratio, and stiffener
to plate thickness ratio) on the said dynamic behavior is
studied. The present article documents the large amplitude
frequency parameters for the first six vibration modes of
a thin, isotropic stiffened plate with a single stiffener for
simply supported bending boundary conditions and immov-
able membrane boundary conditions. The present analysis
employs an energy approach to arrive at the appropriate
governing equations. The formulation is displacement based
and the unknown displacement fields are approximated by
finite linear combination of admissible orthogonal functions.
It should be mentioned here that the present semi-analytical
formulation is associated with the whole domain, which
implies that the assumed displacement fields are a function of
the whole physical domain. However, a domain decomposi-
tion technique, which divides the computational domain into
sub-domains according to the location of the stiffener, is used
to ensure an adequate number of computation points around
the location of the stiffener and boundaries of the plate so
that the displacement field can be captured accurately. The
results are furnished in the form of a graphical representation
of the large amplitude behavior, shown through backbone
curves in the dimensionless amplitude-frequency plane. The
vibration mode shapes are also provided corresponding to
the minimum and maximum amplitudes of vibration.
2. Analysis
Astiffened plate with a single stiffener parallel to y-axis along
with the notations for significant dimensions and coordinate
system used for the present analysis is shown in Figure 1. It
is assumed that the stiffeners are always parallel to the edges
of the plate and they are rigidly connected to the plate. The
mathematical formulation is further based on the following
assumptions:
1. Plate and stiffener materials are isotropic, homogeneous,
and linearly elastic.
2. Thicknesses of the plate and stiffener are uniform.
3. The thickness of the plate is sufficiently small compared to
the lateral dimensions, such that the effect of shear defor-
mation and rotary inertia may be neglected.
It is known that for nonlinear vibration, an increase in
amplitude of vibration changes the dynamic behavior of the
system. However, the maximum deflection is not the only in-
fluencing parameter as the dynamic behavior is also dependent
on the nature of the deflected profile, which again is a func-
tion of the boundary conditions of the plate and the type of
loading to which the plate is subjected. It can be said that
a change in amplitude or the deflection profile correspond
to a change in strain energy stored in the system and conse-
quently a change in the stiffness of the systemand its dynamic
behavior. The present large amplitude vibration problem is
analyzed in two steps. The static problem corresponding to
Fig. 1. Schematic drawing of a stiffened plate with notations for
significant dimensions and coordinate system.
a transverse loading is solved first and subsequently the dy-
namic problemis taken up with the known displacement field.
The mathematical formulation is based on variational form
of energy principle. Geometric nonlinearity is accounted for
by consideration of nonlinear strain-displacement relations.
Appropriate start functions are assumed and the necessary
higher-order constitutive functions are formed by following
the two-dimensional Gram-Schmidt orthogonalization proce-
dure, thus making the solution space complete. To predict the
large-amplitude vibration frequency of the stiffened plate, first
a static analysis is carried out and then knowing the deformed
shape, an eigenvalue problem is formulated corresponding
to the system stiffness to determine the loaded natural fre-
quency. The static problemis solved by employing an iterative
method with an appropriate relaxation technique. As the solu-
tion of the dynamic problem is obtained through the solution
of the static displacement field, the effect of statically imposed
large amplitude of vibration is incorporated into the dynamic
system.
2.1. Static Analysis
From the principle of minimum total potential energy of the
system,
␦(␲) = 0, (1)
where
␲ = U + V = total potential energy of the system;
U = U
p
+U
s
= total strain energy stored in the system, i.e.,
of plate and stiffener;
D
o
w
n
l
o
a
d
e
d

b
y

[
N
a
t
i
o
n
a
l

I
n
s
t
i
t
u
t
e

o
f

T
e
c
h
n
o
l
o
g
y

-

R
o
u
r
k
e
l
a
]

a
t

0
4
:
4
6

2
1

J
u
n
e

2
0
1
3

376 A. Mitra et al.
V = work function or potential of the external forces;
␦ = variational operator.
In the case of large displacement analysis of plates, both
bending and stretching effects are taken into consideration.
Thus, total strain energy stored in plate (U
p
) is given by U
p
=
U
b
+U
m
, where U
b
is the strain energy due to pure bending
and U
m
is the strain energy due to stretching of its middle
surface. The expressions of U
b
and U
m
have been mentioned
in an earlier study [26] for rectangular plates and are indicated
here once again.
U
b
=
D
2
(ab)
_
1
0
_
1
0
_
_
1
a
2
_

2
w
∂␰
2
_
+
1
b
2
_

2
w
∂␩
2
__
2
+2
× (1−␷ )
1
a
2
b
2
_
_

2
w
∂␰∂␩
_
2

_

2
w
∂␰
2
__

2
w
∂␩
2
_
__
d␰d␩,
(2)
U
m
=
E
p
t
p
2(1 −␷
2
)
(ab)
_
1
0
_
1
0
_
1
a
2
_
∂u
∂␰
_
2
+
1
a
3
_
∂u
∂␰
__
∂w
∂␰
_
2
+
1
b
2
_
∂v
∂␩
_
2
+
1
b
3
_
∂v
∂␩
__
∂w
∂␩
_
2
+
1
4
_
1
a
2
_
∂w
∂␰
_
2
+
1
b
2
_
∂w
∂␩
_
2
_
2
+2␷
_
1
ab
_
∂u
∂␰
__
∂v
∂␩
_
+
1
2a
2
b
_
∂v
∂␩
__
∂w
∂␰
_
2
+
1
2ab
2
_
∂u
∂␰
__
∂w
∂␩
_
2
_
+
1 −␷
2
_
1
b
2
_
∂u
∂␩
_
2
+
2
ab
_
∂u
∂␩
__
∂v
∂␰
_
+
1
a
2
_
∂v
∂␰
_
2
+
2
ab
2
_
∂u
∂␩
__
∂w
∂␰
__
∂w
∂␩
_
+
2
a
2
b
_
∂v
∂␰
__
∂w
∂␰
__
∂w
∂␩
_
__
d␰d␩, (3)
where E
p
, ␷ , and D(= E
p
t
3
p
/12(1 −␷
2
)) are the elastic mod-
ulus, Poisson’s ratio, and the flexural rigidity of the plate, re-
spectively. u, v, and w denote the displacements along x-, y-,
and z-directions, respectively.
U
s
is the total strain energy stored in all the stiffeners and
can be expressed as U
s
=

ns
x
p=1
U
p
sx
+

ns
y
q=1
U
q
sy
, where U
p
sx
,
U
q
sy
are strain energies stored in the p-th stiffener along x-
direction and q-th stiffener along y-direction.
U
p
sx
=
E
s
2
___
Vol
_
ε
s
xx
_
2
dV, (4a)
U
q
sy
=
E
s
2
___
Vol
_
ε
s
yy
_
2
dV, (4b)
where ε
s
xx
is the total axial strainof a stiffener along x-direction
andit includes axial straindue tobendingabout the major axis,
axial strain due to stretching of mid-plane, and axial strain due
to bending about the minor axis and can be expressed as:
ε
s
xx
=
du
dx
+
1
2
_
dw
dx
_
2
− z
s
d
2
w
dx
2
− y
s
d
2
v
dx
2
. (5)
Substituting Eq. (5) in (4a) and utilizing the following rela-
tions:
__
Area
z
2
dA = I
y
,
__
Area
y
2
dA = I
z
,
__
Area
yz dA = 0,
__
Area
z dA = Q,
__
Area
y dA = 0,
__
Area
dA = A,
the strain energy stored in a single stiffener is obtained as:
U
p
sx
=
E
s
a
2
_
1
0
_
I
p
y
1
a
4
_
d
2
w
d␰
2
_
2
+ I
p
yz
1
a
4
_
d
2
v
d␰
2
_
2
− Q
p
y
_
2
a
3
_
du
d␰
__
d
2
w
d␰
2
_
+
1
a
4
_
d
2
w
d␰
2
__
dw
d␰
_
2
_
+ A
p
y
_
1
a
2
_
du
d␰
_
2
+
1
4a
4
_
dw
d␰
_
4
+
1
a
3
_
du
d␰
__
dw
d␰
_
2
__
d␰. (6)
Similarly,
U
q
sy
=
E
s
b
2
_
1
0
_
I
q
x
1
b
4
_
d
2
w
d␩
2
_
2
+ I
q
xz
1
b
4
_
d
2
u
d␩
2
_
2
− Q
q
x
_
2
b
3
_
dv
d␩
__
d
2
w
d␩
2
_
+
1
b
4
_
d
2
w
d␩
2
__
dw
d␩
_
2
_
+ A
q
x
_
1
b
2
_
dv
d␩
_
2
+
1
4b
4
_
dw
d␩
_
4
+
1
b
3
_
dv
d␩
__
dw
d␩
_
2
__
d␩, (7)
where E
s
is the elastic modulus of the stiffener material. I
p
y
, I
q
x
and I
p
yz
, I
q
xz
are moment of inertia of the p-th (x-direction)
and q-th (y-direction) stiffener cross-section about the co-
ordinate axes of the plate. They can be determined by car-
rying out suitable transformation on the moment of inertia
about the centroidal axes of the stiffener cross-section. For
example, in case of a single y-direction stiffener as shown
in Figure 1, I
1
x
and I
1
xz
are determined by using the follow-
ing relations: I
1
x
= I
1
x
+ A
x
.e
2
, I
1
xz
= I
1
xz
+ A
x
.(a␰
1
stf
)
2
, where
I
1
x
(= b
s
t
3
s
/12) and I

xz
(= b
3
s
t
s
/12) are moment of inertia about
the major and minor axis of the stiffener cross-section and
A
x
= b
s
t
s
, ␰
stf
= x
stf
/a, ande = (t
p
+t
s
)/2. Q
p
y
, Q
q
x
are the first
moment of area about the plate coordinate axes and A
p
y
, A
q
x
are the cross-sectional areas of the p-th x-direction and q-th y-
direction stiffeners, respectively. Q
x
= A
x
.e for the case of the
uni-axially stiffened plate shown in Figure 1. Similar nomen-
clatures have been used for x-directional stiffeners also.
D
o
w
n
l
o
a
d
e
d

b
y

[
N
a
t
i
o
n
a
l

I
n
s
t
i
t
u
t
e

o
f

T
e
c
h
n
o
l
o
g
y

-

R
o
u
r
k
e
l
a
]

a
t

0
4
:
4
6

2
1

J
u
n
e

2
0
1
3

Nonlinear Vibration Analysis of Stiffened Plates 377
In the case of a stiffened plate under uniform transverse
pressure ( p) and a concentrated load (P), total potential en-
ergy is given by:
V = −Pw|
␰,␩
−(ab)
_
1
0
_
1
0
( pw)d␰d␩, (8)
where w|
␰,␩
is the deflection of the point of application of
the concentrated load. In Eqs. (2), (3), (6), (7), and (8), the
mid-plane coordinates are expressed in dimensionless form
as ␰ =
x
a
, ␩ =
y
b
, while the dimensions of all other physical
quantities, such as load, deflection, elastic modulus, etc., are
retained as such.
2.1.1. Approximate Displacement Fields
The displacement fields w, u, and v are expressed by linear
combinations of unknown parameters d
i
as follows:
w(␰, ␩) =
nw

i =1
d
i

i
(␰, ␩),
u(␰, ␩) =
nw+nu

i =nw+1
d
i

i −nw
(␰, ␩),
v(␰, ␩) =
nw+nu+nv

i =nw+nu+1
d
i

i −nw−nu
(␰, ␩), (9)
where ␾(␰, ␩), ␣(␰, ␩), and ␤(␰, ␩) are sets of nw, nu, and
nv numbers of orthogonal functions for w, u, and v, respec-
tively. Appropriate start functions for these orthogonal sets
are selected in such a way that they satisfy the flexural and
membrane boundary conditions of the plate. The functions
␾(␰, ␩) describe the displacements due to plate bending and
the base function for this comes from the flexural boundary
condition. The start functions for stretching of the plate (u
and v) come from the in-plane boundary conditions, and zero
displacement is assumed at the boundary edges, i.e., u = 0 at
␰ = 0, 1 and v = 0 at ␩ = 0, 1. The higher-order functions are
generated from the selected start functions following a two-
dimensional implementation of the Gram-Schmidt orthogo-
nalization scheme.
The displacement fields depicted in Eq. (9) are two dimen-
sional and compatible with the plate. But the stiffeners are a
one-dimensional element and the displacements at the junc-
tion of the plate and the stiffener are always equal. Hence,
to obtain the displacement fields of the stiffeners, an interpo-
lation function is used on the two-dimensional displacement
function of the plate at the location of the stiffener. For exam-
ple, in the case of the stiffener shown in Figure 1, the transverse
displacement function is given by w(␰, ␩)|

stf
= w(␰
stf
, ␩).
2.1.2. Governing System of Equations
Substituting Eqs. (2), (3), (6), (7), (8), and (9) in (1) gives the
governing set of equations in matrix form:
[K]{d} = { f }, (10)
where [K] = [K
b
] +[K
m
] +

ns
x
p=1
[K
sx
]
p
+

ns
y
q=1
[K
sy
]
q
is the
total stiffness matrix, [K
b
] and [K
m
] being the contributions
frombending andstretching actionof the plate, whereas [K
sx
]
p
and [K
sy
]
q
are stiffness matrices of the p-th stiffener along x-
direction and q-th stiffener along y-direction, respectively. { f }
represents the load vector and {d} is the vector of unknown
coefficients. Both these vectors and the total stiffness matrix
have an order of (nw + nu + nv). The details of the stiffness
matrices are as follows.
The form of [K
b
] is given by:
[K
b
] =




k
b
11
k
b
12
k
b
13
k
b
21
k
b
22
k
b
23
k
b
31
k
b
32
k
b
33




,
where
_
k
b
11
_
= D(ab)
nw

j =1
nw

i =1
_
1
0
_
1
0
__
1
a
4
_

2

i
∂␰
2
__

2

j
∂␰
2
_
+
1
b
4
_

2

i
∂␩
2
__

2

j
∂␩
2
_
+
1
a
2
b
2
_

2

i
∂␰
2
__

2

j
∂␩
2
_
+
1
a
2
b
2
_

2

i
∂␩
2
__

2

j
∂␰
2
__

(1 −␷ )
a
2
b
2
__

2

i
∂␰
2
__

2

j
∂␩
2
_
+
_

2

i
∂␩
2
__

2

j
∂␰
2
_
− 2
_

2

i
∂␰∂␩
__

2

j
∂␰∂␩
___
d␰d␩
_
k
b
12
_
=
_
k
b
13
_
=
_
k
b
21
_
=
_
k
b
22
_
=
_
k
b
23
_
=
_
k
b
31
_
=
_
k
b
32
_
=
_
k
b
33
_
= 0.
The form of [K
m
] is given by:
[K
m
] =



k
m
11
k
m
12
k
m
13
k
m
21
k
m
22
k
m
23
k
m
31
k
m
32
k
m
33


⎦,
where
_
k
m
11
_
=
E
p
t
p
2(1−␷
2
)
(ab)
nw

j =1
nw

i =1
_
1
0
_
1
0


1
a
4
_
nw

i =1
d
i
∂␾
i
∂␰
_
2
∂␾
i
∂␰
∂␾
j
∂␰
+
1
b
4
_
nw

i =1
d
i
∂␾
i
∂␩
_
2
∂␾
i
∂␩
∂␾
j
∂␩
+
1
a
2
b
2
_
nw

i =1
d
i
∂␾
i
∂␰
_
2
∂␾
i
∂␩
∂␾
j
∂␩
+
1
a
2
b
2
_
nw

i =1
d
i
∂␾
i
∂␩
_
2
∂␾
i
∂␰
∂␾
j
∂␰
+
2
a
3
_
nw+nu

i =nw+1
d
i
∂␣
i −nw
∂␰
_
∂␾
i
∂␰
∂␾
j
∂␰
D
o
w
n
l
o
a
d
e
d

b
y

[
N
a
t
i
o
n
a
l

I
n
s
t
i
t
u
t
e

o
f

T
e
c
h
n
o
l
o
g
y

-

R
o
u
r
k
e
l
a
]

a
t

0
4
:
4
6

2
1

J
u
n
e

2
0
1
3

378 A. Mitra et al.
+
2
b
3
_
nw+nu+nv

i =nw+nu+1
d
i
∂␤
i −nw−nu
∂␩
_
∂␾
i
∂␩
∂␾
j
∂␩
+
2␷
a
2
b
_
nw+nu+nv

i =nw+nu+1
d
i
∂␤
i −nw−nu
∂␩
_
∂␾
i
∂␰
∂␾
j
∂␰
+
2␷
ab
2
_
nw+nu

i =nw+1
d
i
∂␣
i −nw
∂␰
_
∂␾
i
∂␩
∂␾
j
∂␩
+
(1 −␷ )
ab
2
__
nw+nu

i =nw+1
d
i
∂␣
i −nw
∂␩
_
∂␾
i
∂␰
∂␾
j
∂␩
+
_
nw+nu

i =nw+1
d
i
∂␣
i −nw
∂␩
_
∂␾
i
∂␩
∂␾
j
∂␰
_
+
(1 −␷ )
a
2
b
__
nw+nu+nv

i =nw+nu+1
d
i
∂␤
i −nw−nu
∂␰
_
∂␾
i
∂␰
∂␾
j
∂␩
+
_
nw+nu+nv

i =nw+nu+1
d
i
∂␤
i −nw−nu
∂␰
_
∂␾
i
∂␩
∂␾
j
∂␰
_


d␰d␩
_
k
m
12
_
= [k
m
13
] = 0,
_
k
m
21
_
=
E
p
t
p
2(1 −␷
2
)
(ab)
nw+nu

j =nw+1
nw

i =1
_
1
0
_
1
0
_
1
a
3
_
nw

i =1
d
i
∂␾
i
∂␰
_
×
∂␾
i
∂␰
∂␣
j −nw
∂␰
+

ab
2
_
nw

i =1
d
i
∂␾
i
∂␩
_
∂␾
i
∂␩
∂␣
j −nw
∂␰
+
(1 −␷ )
ab
2
_
nw

i =1
d
i
∂␾
i
∂␰
_
∂␾
i
∂␩
∂␣
j −nw
∂␩
_
d␰d␩,
_
k
m
22
_
=
E
p
t
p
2(1 −␷
2
)
(ab)
nw+nu

j =nw+1
nw+nu

i =nw+1
_
1
0
_
1
0
_
2
a
2
∂␣
i −nw
∂␰
×
∂␣
j −nw
∂␰
+
(1 −␷ )
b
2
∂␣
i −nw
∂␩
∂␣
j −nw
∂␩
_
d␰d␩,
_
k
m
23
_
=
E
p
t
p
2(1 −␷
2
)
nw+nu

j =nw+1
nw+nu+nv

i =nw+nu+1
_
1
0
_
1
0
_
2␷
∂␤
i −nw−nu
∂␩
×
∂␣
j −nw
∂␰
+(1 −␷ )
∂␤
i −nw−nu
∂␰
∂␣
j −nw
∂␩
_
d␰d␩,
_
k
m
31
_
=
E
p
t
p
2(1 −␷
2
)
(ab)
nw+nu+nv

j =nw+nu+1
nw

i =1
_
1
0
_
1
0
_
1
b
3
_
nw

i =1
d
i
∂␾
i
∂␩
_
×
∂␾
i
∂␩
∂␤
j −nw−nu
∂␩
+

a
2
b
_
nw

i =1
d
i
∂␾
i
∂␰
_
∂␾
i
∂␰
∂␤
j −nw−nu
∂␩
+
(1 −␷ )
a
2
b
_
nw

i =1
d
i
∂␾
i
∂␰
_
∂␾
i
∂␩
∂␤
j −nw−nu
∂␰
_
d␰d␩,
_
k
m
32
_
=
E
p
t
p
2(1 −␷
2
)
nw+nu+nv

j =nw+nu+1
nw+nu

i =nw+1
_
1
0
_
1
0
_
2␷
∂␣
i −nw
∂␰
×
∂␤
j −nw−nu
∂␩
+(1 −␷ )
∂␣
i −nw
∂␩
∂␤
j −nw−nu
∂␰
_
d␰d␩,
_
k
m
33
_
=
E
p
t
p
2(1 −␷
2
)
(ab)
nw+nu+nv

j =nw+nu+1
nw+nu+nv

i =nw+nu+1
×
_
1
0
_
1
0
_
2
b
2
∂␤
i −nw−nu
∂␩
∂␤
i −nw−nu
∂␩
+
(1 −␷ )
a
2
∂␤
i −nw−nu
∂␰
∂␤
i −nw−nu
∂␰
_
d␰d␩.
The form of [K
sx
] is given by:
[K
sx
] =



k
sx
11
k
sx
12
k
sx
13
k
sx
21
k
sx
22
k
sx
23
k
sx
31
k
sx
32
k
sx
33


⎦,
where
_
k
sx
11
_
=
ns
x

p=1


E
s
a
2
nw

j =1
nw

i =1
_
1
0



I
p
y
2
a
4

2

i
∂␰
2

2

j
∂␰
2
− Q
p
y
2
a
4
_
nw

i =1
d
i
∂␾
i
∂␰
_
2

2

i
∂␰
2
∂␾
j
∂␰
− Q
p
y
1
a
4
_
nw

i =1
d
i
∂␾
i
∂␰
_
∂␾
i
∂␰

2

j
∂␰
2
+ A
p
y
1
a
4
_
nw

i =1
d
i
∂␾
i
∂␰
_
2
∂␾
i
∂␰
∂␾
j
∂␰
+ A
p
y
2
a
3
_
nw+nu

i =nw+1
d
i
∂␣
i −nw
∂␰
_
2
∂␾
i
∂␰
∂␾
j
∂␰



d␰


,
_
k
sx
12
_
=
ns
x

p=1


E
s
a
2
nw

j =1
nw+nu

i =nw+1
_
1
0
_
−Q
p
y
2
a
3
∂␣
i −nw
∂␰

2

j
∂␰
2
_
d␰


,
_
k
sx
21
_
=
ns
x

p=1


E
s
a
2
nw+nu

j =nw+1
nw

i =1
_
1
0
_
− Q
p
y
2
a
3

2

i
∂␰
2
∂␣
j −nw
∂␰
+ A
p
y
1
a
3
_
nw

i =1
d
i
∂␾
i
∂␰
_
∂␾
i
∂␰
∂␣
j −nw
∂␰
_
d␰


,
_
k
sx
22
_
=
ns
x

p=1


E
s
a
2
nw+nu

j =nw+1
nw+nu

i =nw+1
_
1
0
_
A
p
y
2
a
2
∂␣
i −nw
∂␰
∂␣
j −nw
∂␰
_
d␰


,
_
k
sx
33
_
=
ns
x

p=1


E
s
a
2
nw+nu+nv

j =nw+nu+1
nw+nu+nv

i =nw+nu+1
×
_
1
0
_
I
p
yz
2
a
4

2

i −nw−nu
∂␰
2

2

i −nw−nu
∂␰
2
_
d␰


,
_
k
sx
13
_
=
_
k
sx
23
_
=
_
k
sx
31
_
=
_
k
sx
32
_
= 0.
D
o
w
n
l
o
a
d
e
d

b
y

[
N
a
t
i
o
n
a
l

I
n
s
t
i
t
u
t
e

o
f

T
e
c
h
n
o
l
o
g
y

-

R
o
u
r
k
e
l
a
]

a
t

0
4
:
4
6

2
1

J
u
n
e

2
0
1
3

Nonlinear Vibration Analysis of Stiffened Plates 379
The form of [K
sy
] is given by:
[K
sy
] =



k
sy
11
k
sy
12
k
sy
13
k
sy
21
k
sy
22
k
sy
23
k
sy
31
k
sy
32
k
sy
33


⎦,
where
_
k
sy
11
_
=
ns
y

q=1


E
s
b
2
nw

j =1
nw

i =1
_
1
0



I
q
x
2
b
4

2

i
∂␩
2

2

j
∂␩
2
−Q
q
x
2
b
4
_
nw

i =1
d
i
∂␾
i
∂␩
_
2

2

i
∂␩
2
∂␾
j
∂␩
− Q
q
x
1
b
4
_
nw

i =1
d
i
∂␾
i
∂␩
_
∂␾
i
∂␩

2

j
∂␩
2
+ A
q
x
1
b
4
_
nw

i =1
d
i
∂␾
i
∂␩
_
2
∂␾
i
∂␩
∂␾
j
∂␩
+ A
q
x
2
b
3
_
nw+nu+nv

i =nw+nu+1
d
i
∂␤
i −nw−nu
∂␩
_
2
∂␾
i
∂␩
∂␾
j
∂␩



d␩


,
_
k
sy
22
_
=
ns
y

q=1


E
s
b
2
nw+nu

j =nw+1
nw+nu

i =nw+1
_
1
0
_
I
q
xz
2
b
4

2

i −nw
∂␩
2
∂␣
2
j −nw
∂␩
2
_
d␩


,
_
k
sy
33
_
=
ns
y

q=1


E
s
b
2
nw+nu+nv

j =nw+nu+1
nw+nu+nv

i =nw+nu+1
×
_
1
0
_
A
q
x
2
b
2
∂␤
i −nw−nu
∂␩
∂␤
i −nw−nu
∂␩
_
d␩


,
_
k
sy
31
_
=
ns
y

q=1


E
s
b
2
nw+nu+nv

j =nw+nu+1
nw

i =1
_
1
0
_
−Q
q
x
2
b
3

2

i
∂␩
2
∂␤
j −nw−nu
∂␩
+ A
q
x
1
b
3
_
nw

i =1
d
i
∂␾
i
∂␩
_
∂␾
i
∂␩
∂␤
j −nw−nu
∂␩
_
d␩


,
Fig. 2. Flowchart of solution methodology.
D
o
w
n
l
o
a
d
e
d

b
y

[
N
a
t
i
o
n
a
l

I
n
s
t
i
t
u
t
e

o
f

T
e
c
h
n
o
l
o
g
y

-

R
o
u
r
k
e
l
a
]

a
t

0
4
:
4
6

2
1

J
u
n
e

2
0
1
3

380 A. Mitra et al.
_
k
sy
13
_
=
ns
y

q=1


E
s
b
2
nw

j =1
nw+nu+nv

i =nw+nu+1
×
_
1
0
_
−Q
q
x
2
b
3
∂␤
i −nw−nu
∂␩

2

j
∂␩
2
_
d␩


,
_
k
sy
12
_
= [k
sy
21
] = [k
sy
23
] = [k
sy
32
] = 0.
The load vector { f } is of the form:
{ f } = { f
11
f
12
f
13
}
T
,
where
{ f
11
} =
nw

j =1
P␾
j
|
␰,␩
+ p(ab)
×
nw

j =1
_
1
0
_
1
0

j
d␰d␩ and { f
12
} = { f
13
} = 0,
as in-plane loading is absent. In the present analysis, re-
sults are individually generated for a single concentrated
load ({ f
11
} =

nw
j =1
P␾
j
|
␰,␩
) and uniformly distributed load
({ f
11
} = p(ab)

nw
j =1
_
1
0
_
1
0

j
d␰d␩).
2.1.3. Solution Methodology for Static Displacement Field
The set of governing equations (Eq. (10)) is nonlinear as the
stiffness matrix ([K]) is a function of unknown coefficients and
solved by a direct substitution technique using a successive
relaxation scheme. The solution technique for a particular
load step is elaborated through a flowchart shown in Figure 2.
At the beginning of each load-step, the values of the un-
known coefficients are assumed in order to evaluate the initial
stiffness matrix [K] and a set of new values of unknown co-
efficients are calculated from the expression {d} = [K]
−1
{ f }.
Calculated values are compared with the values obtained in
the previous iteration and if the difference is below a pre-
defined value of error limit (ε), it is assumed that the dis-
placement field has converged. Otherwise, the values of {d}
Fig. 3. Comparison of load deflection behavior of clamped stiffened plate at selective points.
D
o
w
n
l
o
a
d
e
d

b
y

[
N
a
t
i
o
n
a
l

I
n
s
t
i
t
u
t
e

o
f

T
e
c
h
n
o
l
o
g
y

-

R
o
u
r
k
e
l
a
]

a
t

0
4
:
4
6

2
1

J
u
n
e

2
0
1
3

Nonlinear Vibration Analysis of Stiffened Plates 381
are modified with a relaxation parameter following a relation
given by {d} = {d}
old
+␭({d} −{d}
old
) and this is used as the
next approximation for the values of {d}. Once convergence is
achieved for a particular load step, an increment is given to the
load and the present solution of {d} serves as the initial guess.
2.2. Dynamic Analysis
The governing set of equations for the dynamic problem is
derived following Hamilton’s principle, which can be mathe-
matically expressed as:

__

2

1
(T −U − V) d␶
_
= ␦
__

2

1
(T −␲) d␶
_
= ␦
__

2

1
Ld␶
_
= 0, (11)
where L represents the Lagrangian and T, U, and V are the
total kinetic energy of the system, total strain energy stored
in the system, and work function or potential of the external
forces, respectively. For free vibration analysis, the potential
of the external forces (V) is reduced to zero. ␦ and ␶ denote
the variational operator and time coordinate, respectively. The
total kinetic energy is summation of plate kinetic energy (T
p
)
and stiffener kinetic energy (T
s
) and is expressed as T = T
p
+
T
s
, where T
s
=

ns
x
p=1
T
p
sx
+

ns
y
q=1
T
q
sy
.
T
p
=
1
2

p
t
p
(ab)
_
1
0
_
1
0
_
_
∂w
∂␶
_
2
+
_
∂u
∂␶
_
2
+
_
∂v
∂␶
_
2
_
d␰d␩,
(12a)
T
p
sx
=
1
2

s
t
s
(ab
s
)
_
1
0
_
_
∂w
∂␶
_
2
+
_
∂u
∂␶
_
2
+
_
∂v
∂␶
_
2
_
d␰, (12b)
T
q
sy
=
1
2

s
t
s
(bb
s
)
_
1
0
_
_
∂w
∂␶
_
2
+
_
∂u
∂␶
_
2
+
_
∂v
∂␶
_
2
_
d␩. (12c)
The dynamic displacements w(␰, ␩, ␶), u(␰, ␩, ␶), and
v(␰, ␩, ␶) are assumed to be separable in space and time. They
are expressed as shown below:
w(␰, ␩, ␶) =
nw

i =1
d
i

i
(␰, ␩)␥
i
(␶) ,
u (␰, ␩, ␶) =
nw+nu

i =nw+1
d
i

i −nw
(␰, ␩)␥
i −nw
(␶) ,
v (␰, ␩, ␶) =
nw+nu+nv

i =nw+nu+1
d
i

i −nw−nu
(␰, ␩)␥
i −nw−nu
(␶). (13)
Here, {d} is a new set of unknown parameters to be evaluated,
which forms the eigenvectors of the dynamic problem in
matrix form. Hence, it can be said that this vector indicates
the contribution of the individual space functions in a
particular vibration frequency mode. The space functions
are completely known from the static analysis and the set of
harmonic temporal functions is expressed by ␥
i
(␶) = e
i ␻␶
,
where ␻ is the natural frequency of the system. Substituting
the above dynamic displacements along with Eqs. (2), (3), (6),
(7), and (8) in Eq. (11), the governing differential equation of
the dynamic system can be written in matrix form:
−␻
2
[M] {d} +[K]{d} = 0, (14)
where [M] is the mass matrix made up of contributions from
the plate ([M
p
]) and the x- and y-direction stiffeners ([M
sx
]
and [M
sy
]).
[M] = [M
p
] +[M
sx
] +[M
sy
].
The form of [M
p
] is given by:
[M
p
] =



M
p
11
0 0
0 M
p
22
0
0 0 M
p
33


⎦,
where
_
M
p
11
_
= ␳
p
t
p
(ab)
nw

j =1
nw

i =1
_
1
0
_
1
0

i

j
d␰d␩,
_
M
p
22
_
= ␳
p
t
p
(ab)
nw+nu

j =nw+1
nw+nu

i =nw+1
_
1
0
_
1
0

i −nw

j −nw
d␰d␩,
_
M
p
33
_
= ␳
p
t
p
(ab)
nw+nu+nv

j =nw++nu1
nw+nu+nv

i =nw+nu+1
×
_
1
0
_
1
0

i −nw−nu

j −nw−nu
d␰d␩.
Fig. 4. Comparison of deflected profile of simply supported
square stiffened plate under UDL.
D
o
w
n
l
o
a
d
e
d

b
y

[
N
a
t
i
o
n
a
l

I
n
s
t
i
t
u
t
e

o
f

T
e
c
h
n
o
l
o
g
y

-

R
o
u
r
k
e
l
a
]

a
t

0
4
:
4
6

2
1

J
u
n
e

2
0
1
3

382 A. Mitra et al.
Fig. 5. Backbone curves for the SSSS stiffened plate under UDL for different stiffener position: (a) ␰
stf
= 0.5, (b) ␰
stf
= 0.4,
(c) ␰
stf
= 0.3, and (d) ␰
stf
= 0.25.
The form of [M
sx
] is given by:
[M
sx
] =



M
sx
11
0 0
0 M
sx
22
0
0 0 M
sx
33


⎦,
where
_
M
sx
11
_
=
ns
x

p=1



s
t
s
(ab
s
)
nw

j =1
nw

i =1
_
1
0

i

j
d␰


,
_
M
sx
22
_
=
ns
x

p=1



s
t
s
(ab
s
)
nw+nu

j =nw+1
nw+nu

i =nw+1
_
1
0

i −nw

j −nw
d␰


,
_
M
sx
33
_
=
ns
x

p=1



s
t
s
(ab
s
)
nw+nu+nv

j =nw+nu+1
nw+nu+nv

i =nw+nu+1
×
_
1
0

i −nw−nu

j −nw−nu
d␰


.
D
o
w
n
l
o
a
d
e
d

b
y

[
N
a
t
i
o
n
a
l

I
n
s
t
i
t
u
t
e

o
f

T
e
c
h
n
o
l
o
g
y

-

R
o
u
r
k
e
l
a
]

a
t

0
4
:
4
6

2
1

J
u
n
e

2
0
1
3

Nonlinear Vibration Analysis of Stiffened Plates 383
Fig. 6. Backbone curves for the SSSS stiffened plate under concentrated load for different stiffener position: (a) ␰
stf
= 0.5, (b) ␰
stf
= 0.4,
(c) ␰
stf
= 0.3, and (d) ␰
stf
= 0.25.
The form of [M
sy
] is given by:
[M
sy
] =



M
sy
11
0 0
0 M
sy
22
0
0 0 M
sy
33


⎦,
where
_
M
sy
11
_
=
ns
y

q=1



s
t
s
(bb
s
)
nw

j =1
nw

i =1
_
1
0

i

j
d␩


,
_
M
sy
22
_
=
ns
y

q=1



s
t
s
(bb
s
)
nw+nu

j =nw+1
nw+nu

i =nw+1
_
1
0

i −nw

j −nw
d␩


,
D
o
w
n
l
o
a
d
e
d

b
y

[
N
a
t
i
o
n
a
l

I
n
s
t
i
t
u
t
e

o
f

T
e
c
h
n
o
l
o
g
y

-

R
o
u
r
k
e
l
a
]

a
t

0
4
:
4
6

2
1

J
u
n
e

2
0
1
3

384 A. Mitra et al.
_
M
sy
33
_
=
ns
y

q=1



s
t
s
(bb
s
)
nw+nu+nv

j =nw+nu+1
nw+nu+nv

i =nw+nu+1
×
_
1
0

i −nw−nu

j −nw−nu
d␩


.
The standard eigenvalue problem of Eq. (14) is solved numer-
ically to calculate the natural frequencies (␻
i
) by using Inter-
national Mathematics and Statistics Library (IMSL) routines.
3. Results and Discussion
The influence of stiffener position, plate aspect ratio (a/b),
and stiffener to plate thickness ratio (t
s
/t
p
) on the large deflec-
tion dynamic behavior of stiffened plates is investigated in the
present study. The analysis is based on solving the static dis-
placement of the stiffened plate and subsequently evaluating
the eigenvalues of the corresponding dynamic problem on the
basis of the known static displacement field. The square roots
of these eigenvalues represent the free vibration frequencies
of the stiffened plate at that statically deflected configuration.
The amplitude of free vibration is provided by the displace-
ment fields associated with each of the eigenvalues. The eigen-
vectors corresponding to the eigenvalues are also calculated
and through themthe mode shapes of the vibrating systemare
generated.
In order to keep the volume of the article within reasonable
limit, the results are generated for a plate stiffened with a single
stiffener and simply supported at all the boundaries (SSSS).
A centrally stiffened rectangular plate (Figure 3a) clamped
on all the edges is also analyzed for validation purpose. It
should be mentioned here that the formulation and solution
methodology are general in nature and can be applied for
any classical boundary condition. The start functions for the
definition of transverse deflection (w) come from the simply
supported ends and are given by ␾
1
(␰, ␩) = sin(␲␰). sin(␲␩).
The start functions for the stretching (u and v) come from
the membrane boundary conditions and they are modeled as
immovable by imposing zero in-plane displacements. These
selected start functions are used to generate the higher order
functions with the help of two-dimensional Gram-Schmidt
orthogonalization principle. The number of functions for
each of the plate displacements (u, v, and w) is taken as 25
(5 × 5), where numbers in the parentheses provide a break up
for the order of the functions corresponding totwoorthogonal
directions.
The solution methodology, especially adopted in the
present study for the nonlinear system, employs an iterative
numerical scheme using the technique of successive relaxation.
The tolerance value of the error limit (ε) for the numerical it-
eration scheme is taken as 0.50%and the relaxation parameter
(␭) is fixed at 0.50. Two types of transverse loading condition,
namely, uniformly distributed load (UDL) and concentrated
load at the plate center (␰ = 0.5, ␩ = 0.5) have been consid-
ered. The changes in dynamic behavior of the system due
to these two different types of loading are also noted. The
results are generated for the following material properties:
E
p
= E
s
=211 GPa, ␳
p
=␳
s
=7830 kg/m
3
, and ␷ =0.30 (for
both plate and stiffener).
3.1. Validation Study
The results of the present analysis are validated through
comparison with those of various researchers covering dif-
ferent aspects of the study. The results of the static analysis
(load–deflection curve) incorporating geometric nonlinearity
are compared with those of Sheikh and Mukhopadhyay [19],
Rao (taken from [19]), and Koko and Olson [10]. The de-
tails of the geometry and dimensions (in mm) of the clamped
stiffened plate under transverse pressure loading are shown in
Figure 3a. The comparison plots for two different locations,
Aand B, on the plate (as shown by Figure 3a) are presented in
Figures 3b and 3c and a fairly good agreement is observed for
both of the cases. The deflected shape of a simply supported
square plate with one stiffener under uniformly distributed
load, given by Bedair [27], is compared with the same gener-
ated by the present method and presented in Figure 4. The
nature of the deflected profiles is found to be identical and the
matching between the two sets of results is excellent.
The validation of the dynamic problemis provided by com-
paring the free vibration frequencies of different modes at
no loading condition, i.e., linear case. First, the dimension-
less free vibration frequency parameters (␻a
2
_
␳t
p
/D) of an
unstiffened square plate corresponding to linear analysis are
compared with Leissa [28]. For this particular case, the stiff-
ener dimensions are taken as zero. The results for the first
six modes are tabulated in Table 1 and fairly good agreement
is observed for all the cases. Second, in order to validate the
present formulation, the first three linear frequencies (Hz) for
Fig. 7. Variation of dimensionless free vibration frequency pa-
rameter for the first mode with change in stiffener position.
D
o
w
n
l
o
a
d
e
d

b
y

[
N
a
t
i
o
n
a
l

I
n
s
t
i
t
u
t
e

o
f

T
e
c
h
n
o
l
o
g
y

-

R
o
u
r
k
e
l
a
]

a
t

0
4
:
4
6

2
1

J
u
n
e

2
0
1
3

Nonlinear Vibration Analysis of Stiffened Plates 385
Fig. 8. Mode shape plots for SSSS stiffened rectangular plate with stiffener position ␰
stf
= 0.30 (corresponding to backbone curves in
Figure 5c) under UDL: (a) Mode 1, (b) Mode 2, (c) Mode 3, (d) Mode 4, (e) Mode 5, and (f) Mode 6.
D
o
w
n
l
o
a
d
e
d

b
y

[
N
a
t
i
o
n
a
l

I
n
s
t
i
t
u
t
e

o
f

T
e
c
h
n
o
l
o
g
y

-

R
o
u
r
k
e
l
a
]

a
t

0
4
:
4
6

2
1

J
u
n
e

2
0
1
3

386 A. Mitra et al.
Fig. 8. (Continued).
D
o
w
n
l
o
a
d
e
d

b
y

[
N
a
t
i
o
n
a
l

I
n
s
t
i
t
u
t
e

o
f

T
e
c
h
n
o
l
o
g
y

-

R
o
u
r
k
e
l
a
]

a
t

0
4
:
4
6

2
1

J
u
n
e

2
0
1
3

Nonlinear Vibration Analysis of Stiffened Plates 387
Fig. 9. Backbone curves for the SSSS stiffened plate under UDL for different plate aspect ratio: (a) a/b = 0.75, (b) a/b = 1.0,
(c) a/b = 1.5, and (d) a/b = 2.0.
Table 1. Validation of the results for the first six linear dimen-
sionless frequency parameters for an unstiffened square plate
with all edges simply supported
Boundary Present Leissa Saha et al.
condition Mode study [28] [26]
SSSS 1 19.7387 19.743 19.743
2 49.3568 49.344 49.369
3 49.3570 49.344 49.369
4 78.9664 78.944 78.996
5 99.1518 98.687 99.282
6 99.1518 98.687 99.282
a simply supported rectangular plate with a centrally placed
eccentric stiffener have been compared with the corresponding
results presented by various authors in Table 2. It is observed
that the results of different researchers are at variance to some
extent. However, it is found that the present results are very
close to some of themand the slight variations may be ascribed
to the difference in the solution methodologies adopted.
3.2. Large-Deflection Dynamic Behavior: Stiffener Position
The large amplitude dynamic behavior of stiffened plates is
shown graphically as the backbone curves in the dimensionless
D
o
w
n
l
o
a
d
e
d

b
y

[
N
a
t
i
o
n
a
l

I
n
s
t
i
t
u
t
e

o
f

T
e
c
h
n
o
l
o
g
y

-

R
o
u
r
k
e
l
a
]

a
t

0
4
:
4
6

2
1

J
u
n
e

2
0
1
3

388 A. Mitra et al.
Fig. 10. Backbone curves for the SSSS stiffened plate under concentrated load for different plate aspect ratio: (a) a/b = 0.75,
(b) a/b = 1.0, (c) a/b = 1.5, and (d) a/b = 2.0.
Table 2. Validation of the results for the first three linear dimensional frequencies (in Hz) for a centrally stiffened rectangular plate
with all edges simply supported (plate dimensions: a = 0.60 m, b = 0.41 m, t
p
= 0.00633 m; stiffener dimensions: b
s
= 0.0127 m,
t
s
= 0.02222 m; stiffener position: x
stf
= 0.3 m, parallel to y-axis)
Mode Present study Aksu [9] Harik and Guo [12] Mukherjee and Mukhopadhyay [16] Bedair [18]
1 256.9694 254.94 253.59 257.05 256.20
2 266.3567 269.46 282.02 272.10 —
3 529.5827 511.64 513.50 524.70 —
D
o
w
n
l
o
a
d
e
d

b
y

[
N
a
t
i
o
n
a
l

I
n
s
t
i
t
u
t
e

o
f

T
e
c
h
n
o
l
o
g
y

-

R
o
u
r
k
e
l
a
]

a
t

0
4
:
4
6

2
1

J
u
n
e

2
0
1
3

Nonlinear Vibration Analysis of Stiffened Plates 389
Fig. 11. Large-deflection dynamic behavior of SSSS stiffened plate (a/b = 1.0, t
s
/t
p
= 3.5, and ␰
stf
= 0.50) for the two types of loading:
(a) Mode 1 and (b) Mode 2.
amplitude-frequency plane. The ratio of the maximum plate
deflection to plate thickness is taken as the dimensionless am-
plitude (w
max
/t
p
), while the dimensionless frequency (␻
nl
/␻
1
)
is obtained by normalizing the nonlinear frequency (␻
nl
) with
the corresponding fundamental linear frequency (␻
1
). The
backbone curves are furnished for the first six modes and the
maximumvalue of dimensionless amplitude (w
max
/t
p
) is taken
as 2.0 for all of the cases. The backbone curves for different
stiffener positions are shown in Figures 5a–5d and 6a–6d for
stiffener positions ␰
stf
= 0.50, 0.40, 0.30, and 0.25 with a fixed
value of aspect ratio (a/b) and thickness ratio (t
s
/t
p
). The
plate and stiffener dimensions are taken identical to the case
as mentioned in the caption of Table 2. Figures 5 and 6 cor-
respond to uniformly distributed and concentrated loading,
respectively.
The general trend that can be obtained from any of the
backbone curves is that the free vibration frequency increases
as the deflection of the stiffened plate increases. The specific
reason for this trend can be attributed to the fact that
the plate stiffens with an increase in deflection due to the
effect of geometric nonlinearity, resulting in the increase in
nonlinear frequency. The phenomenon of mode switching
has been observed for stiffener position ␰
stf
= 0.30 and 0.25
for uniformly distributed loading. It can be further seen from
Figures 5c and 5d that the mode switching occurs between ␻
3
and ␻
4
. In Figure 5d, it is also apparent that switching takes
place between ␻
6
and ␻
7
giving the sixth backbone curve a
broken appearance in absence of the seventh mode. Similar
trends are observed for backbone curves corresponding to
stiffener position ␰
stf
= 0.30 in the case of concentrated
loading (Figure 6c). But the switching between ␻
3
and ␻
4
for ␰
stf
= 0.25 is absent in Figure 6d. It is understood that
the occurrence of mode switching and the modes between
which switching occurs are entirely dependent on the system
parameters.
As the stiffener position shifts, the system stiffness changes
causing the first linear natural frequency to change as well.
Thus, the normalizing parameter (␻
1
) for the cases of various
stiffener positions shown in Figures 5 and 6 are different. The
values of ␻
1
for the stiffener positions 0.25, 0.30, 0.40, and 0.50
are 175.204, 189.119, 224.163, and 256.969 Hz, respectively.
The influence of stiffener position on the dynamic behavior of
the systemis elaborated through Figure 7, where the variation
of dimensionless free vibration frequency parameter for the
first mode with change in stiffener position is plotted. It is ev-
ident from the figure that the shift in position of the stiffener
towards the boundary of the plate causes the frequency
parameter to decrease. The specific reason for this is that
the system stiffness decreases due to the offset position
of the stiffener and when a system becomes less stiff its
natural frequency decreases. Thus, it can be concluded that
maximum stiffness corresponds to a central stiffener position
(␰
stf
= 0.50).
Mode shape plots for the first six vibration modes of SSSS
stiffened plate under uniform transverse pressure have been
presented for the case ␰
stf
= 0.30 (Figure 5c) in Figures 8a–8f.
This particular geometry is chosen in order to elaborate the
mode switching phenomenon through the mode shapes. For
D
o
w
n
l
o
a
d
e
d

b
y

[
N
a
t
i
o
n
a
l

I
n
s
t
i
t
u
t
e

o
f

T
e
c
h
n
o
l
o
g
y

-

R
o
u
r
k
e
l
a
]

a
t

0
4
:
4
6

2
1

J
u
n
e

2
0
1
3

390 A. Mitra et al.
Fig. 12. Backbone curves for the SSSS stiffened plate under UDL for different thickness ratio: (a) t
s
/t
p
= 0.0, (b) t
s
/t
p
= 1.0,
(c) t
s
/t
p
= 2.0, and (d) t
s
/t
p
= 3.5.
each mode of vibration, two mode shape plots corresponding
to linear (w
max
/t
p
= 0) and nonlinear (w
max
/t
p
= 2.00)
frequencies are given. In each plot, the surface plot and its
corresponding contour plot for the vibrating plate have been
presented. It is also noted that the amplitude of vibration
for all the surface plots is normalized by the corresponding
maximum transverse displacement amplitude. Interchange of
linear and nonlinear mode shapes depicted in Figures 8c and
8d appropriately support the case of mode switching between
the third and fourth mode.
3.3. Large-Deflection Dynamic Behavior: Plate Aspect
Ratio (a/b)
Figures 9a–9d and 10a–10d show the backbone curves for
variation in plate aspect ratio (a/b) for uniformly distributed
D
o
w
n
l
o
a
d
e
d

b
y

[
N
a
t
i
o
n
a
l

I
n
s
t
i
t
u
t
e

o
f

T
e
c
h
n
o
l
o
g
y

-

R
o
u
r
k
e
l
a
]

a
t

0
4
:
4
6

2
1

J
u
n
e

2
0
1
3

Nonlinear Vibration Analysis of Stiffened Plates 391
Fig. 13. Backbone curves for the SSSS stiffened plate under concentrated load for different thickness ratio: (a) t
s
/t
p
= 0.0,
(b) t
s
/t
p
= 1.0, (c) t
s
/t
p
= 2.0, and (d) t
s
/t
p
= 3.5.
and concentrated loading, respectively. Corresponding to
each loading condition, the variations in the dynamic be-
havior is shown for aspect ratios a/b = 0.75, 1.0, 1.5, and 2.0
with a fixed value of thickness ratio (t
s
/t
p
= 3.5) and position
of the stiffener (␰
stf
= 0.50). The details of the geometry of the
stiffened plate are as follows: b = 0.40 m, t
p
= 0.006 m, b
s
=
0.012 m, t
s
=0.021 m. It should be mentioned here that varia-
tion of the aspect ratio is achieved by varying the length (a) of
the plate while keeping the width (b) constant. This is done to
keep the length of the stiffener spanning the width constant.
Normalization of the nonlinear frequency is obtained by
dividing with the fundamental frequency of the system and
the corresponding values for the aspect ratios 0.75, 1.00,
1.50, and 2.00 are as follows: 407.814, 338.934, 252.357, and
184.678 Hz, respectively. Figures 9 and 10 exhibit that, with
the increase in deflection, the natural frequencies increase
for all the cases. As in the previous case, for some specific
geometry the phenomenon of mode switching is observed. In
D
o
w
n
l
o
a
d
e
d

b
y

[
N
a
t
i
o
n
a
l

I
n
s
t
i
t
u
t
e

o
f

T
e
c
h
n
o
l
o
g
y

-

R
o
u
r
k
e
l
a
]

a
t

0
4
:
4
6

2
1

J
u
n
e

2
0
1
3

392 A. Mitra et al.
the case of a/b = 0.75 and 1.0 for pressure loading (Figures 9a
and 9b) and a/b = 0.75 for concentrated loading (Figure 10a)
switching occurs between ␻
3
and ␻
4
, whereas for a/b = 2.0
(Figure 10d) it occurs between ␻
4
and ␻
5
.
In order to see the effect of nature of loading on the
dynamic behavior, the backbone curves corresponding to
UDL and point loading are reproduced in a single plot.
The large-deflection dynamic behavior of SSSS stiffened plate
(a/b = 1.0, t
s
/t
p
= 3.5, and ␰
stf
= 0.50) for the two types of
loading are presented as backbone curves in Figures 11a and
11b for the first two vibration modes, respectively. The figures
show that a particular value of nondimensional amplitude
corresponds to a higher normalized nonlinear frequency for
pressure loading than point load. This can be attributed to
the difference in deflected profile and, hence, the stored strain
energy between the two types of loading.
3.4. Large-Deflection Dynamic Behavior: Thickness
Ratio (t
s
/t
p
)
The backbone curves for variation in thickness ratio (t
s
/t
p
) for
uniformly distributed and concentrated loading are presented
in Figures 12a–12d and 13a–13d, respectively. Corresponding
to each loading condition, the variations in the dynamic be-
havior is shown for thickness ratios t
s
/t
p
= 0.0, 1.0, 2.0, and
3.5 with a fixed value of thickness ratio (a/b = 1.5) and posi-
tionof the stiffener (␰
stf
= 0.50). The geometry witht
s
/t
p
= 0.0
actually represents an unstiffened plate. The details of the ge-
ometry of the stiffened plate are as follows: a = 0.60 m, b =
0.40 m, t
p
=0.006 m. The variation of the thickness ratio is ef-
fected by changing the stiffener thickness (t
s
) keeping the plate
thickness (t
p
) fixed. From Figures 12a–12d and 13a–13d it is
noted that the natural frequencies increase with an increase
in deflection for all the cases, which indicates hard spring be-
havior. Mode switching between ␻
4
and ␻
5
takes place for
the particular case of t
s
/t
p
= 2.0 as shown by Figures 12c
and 13c.
It is also observed that increase in thickness ratio causes the
backbone curves for different modes to come closer together
and this trend is observed for both types of loading. How-
ever, the plots in Figures 12 and 13 are in a nondimensional
plane and the normalizing parameters (first linear natural fre-
quency, ␻
1
) for the thickness ratios 0.0, 1.0, 2.0, and 3.5 are
133.657, 145.125, 181.282, and 252.357 Hz, respectively. The
trend indicates that the fundamental frequency increases with
increase in thickness ratio and this is supported by Figure 14,
which presents the variation of dimensionless free vibration
frequency parameter for the first mode with change in stiffener
to plate thickness ratio. With an increase in stiffener thickness,
the overall stiffness of the systemincreases and, hence, the fre-
quency parameter also moves towards higher values. It is seen
that as the thickness ratio tends to zero, the frequency pa-
rameter value converges towards that of an unstiffened plate.
Figure 14 also indicates that the dimensionless frequency pa-
rameter attains a constant value beyond a particular thickness
ratio. At this critical thickness ratio, the stiffener divides the
plate into two panels of equal width that vibrate with the same
natural frequency as the plate. In an earlier study by Bedair
Fig. 14. Variation of dimensionless free vibration frequency pa-
rameter for the first mode with change in plate to stiffener thick-
ness ratio.
[18], a similar trend is observed for simply supported stiffened
plates.
To highlight the effect of vibration amplitude on the dy-
namic behavior of the plate in greater detail, the mode shape
plots for all six vibration modes are presented corresponding
to thickness ratio, t
s
/t
p
= 1.00, and concentrated load
(Figure 13b) in Figures 15a–15f. As in Figure 8, two mode
shape plots corresponding to linear and nonlinear frequencies
are provided along with the corresponding contour plots
for the vibrating plate. It is expected that the degree of
nonlinearity is affected by load and is manifested through
a change in a particular mode shape. But it appears from
Figure 15 that the surface plots for a particular mode shape
for linear and nonlinear frequencies are identical in nature. In
order to establish the effect of nonlinearity on mode shapes,
the same results are presented in a different fashion in Figures
16a–16f. This figure is essentially a sectional viewof the earlier
mode shape plots at suitable locations. The centrally stiffened
plate being symmetric, the mode shapes for half of the plate
are shown in each case. For the first three modes of vibration,
the sectional view is taken along the x-axis (orthogonal to
stiffener direction) at y = 0.5. In order to avoid the section
line passing through the nodal locations, the sectional view
for the fourth, fifth, and sixth mode is obtained along x-axis
at y = 0.25. In all cases, mode shapes corresponding to two
different amplitudes of vibration (w
max
/t
p
= 0 (linear) and
2.0 (nonlinear)) are presented. For this particular case, such
a geometry is chosen for which no mode switching occurs
so as to avoid the corrective measure necessary to make the
comparisons. This study clearly reveals that nonlinear mode
shapes differ from the linear shape and there is a pronounced
effect of amplitude of vibration on mode shapes.
D
o
w
n
l
o
a
d
e
d

b
y

[
N
a
t
i
o
n
a
l

I
n
s
t
i
t
u
t
e

o
f

T
e
c
h
n
o
l
o
g
y

-

R
o
u
r
k
e
l
a
]

a
t

0
4
:
4
6

2
1

J
u
n
e

2
0
1
3

Nonlinear Vibration Analysis of Stiffened Plates 393
Fig. 15. Mode shape plots for SSSS stiffened rectangular plate with thickness ratio t
s
/t
p
= 1.00 (corresponding to backbone curves
in Figure 13b) under concentrated load: (a) Mode 1, (b) Mode 2, (c) Mode 3, (d) Mode 4, (e) Mode 5, and (f) Mode 6.
D
o
w
n
l
o
a
d
e
d

b
y

[
N
a
t
i
o
n
a
l

I
n
s
t
i
t
u
t
e

o
f

T
e
c
h
n
o
l
o
g
y

-

R
o
u
r
k
e
l
a
]

a
t

0
4
:
4
6

2
1

J
u
n
e

2
0
1
3

394 A. Mitra et al.
Fig. 15. (Continued).
D
o
w
n
l
o
a
d
e
d

b
y

[
N
a
t
i
o
n
a
l

I
n
s
t
i
t
u
t
e

o
f

T
e
c
h
n
o
l
o
g
y

-

R
o
u
r
k
e
l
a
]

a
t

0
4
:
4
6

2
1

J
u
n
e

2
0
1
3

Nonlinear Vibration Analysis of Stiffened Plates 395
Fig. 16. A detailed study on changes in mode shapes along x-axis with deflection amplitude (w
max
/t
p
) taken at suitable y-location.
D
o
w
n
l
o
a
d
e
d

b
y

[
N
a
t
i
o
n
a
l

I
n
s
t
i
t
u
t
e

o
f

T
e
c
h
n
o
l
o
g
y

-

R
o
u
r
k
e
l
a
]

a
t

0
4
:
4
6

2
1

J
u
n
e

2
0
1
3

396 A. Mitra et al.
5. Conclusion
In this article, a large-amplitude free vibration analysis of
stiffened plates subjected to transverse loading with all along
simply supported flexural boundary conditions and zero dis-
placement in-plane boundary conditions has been presented.
The mathematical formulation is based on energy method, the
underlying principle being the extremization of total energy of
the system in its equilibrium state. The analysis is carried out
in two steps. The first step involves solving the static displace-
ment of the stiffened plate and then the subsequent dynamic
study is taken up on the basis of known static displacement
field. It can be said that static analysis yields the initial de-
flection profile used in the subsequent free vibration analysis.
The vibration frequencies are obtained from the solution of
a standard eigenvalue problem. The results of both static and
dynamic analyses are validated with the published results of
other researchers, and fairly good agreement is observed in
almost all the cases. The dynamic behavior of stiffened plates
has been presented in the form of backbone curves in a di-
mensionless frequency-amplitude plane. The backbone curves
documented may be used by the practicing engineers as design
curves. Influence of stiffener position, plate aspect ratio, and
stiffener to plate thickness ratio on the large amplitude dy-
namic behavior has been studied. The phenomenon of mode
switching is observed for some specific geometries of the stiff-
ened plate. Three-dimensional mode shape plots along with
contour plots are provided in a few cases to get a better un-
derstanding of the nature of influence of vibration amplitude
on the dynamic behavior of the system. Astudy is also carried
out to establish the effect of amplitude of vibration on mode
shapes.
Acknowledgment
The first author acknowledges the research support re-
ceived from AICTE, India, vide File No.: 1-10/RID/NDF/
PG/(17)2008-09; Dated: 13.03.2009.
References
[1] A. Mukherjee and M. Mukhopadhyay, A review of dynamic be-
havior of stiffened plates, Shock Vib. Dig., vol. 18, no. 6, pp. 3–8,
1986.
[2] M. Mukhopadhyay and A. Mukherjee, Recent advances on the
dynamic behavior of stiffened plates, Shock Vib. Dig., vol. 21, no.
4, pp. 6–9, 1989.
[3] O.K. Bedair, Acontribution to the stability of stiffened plates under
uniform compression, Comput. Struc., vol. 66, no. 5, pp. 535–570,
1998.
[4] T. Wah, Vibration of stiffened plates, Aeronaut. Quart., vol. 15, pp.
285–298, 1964.
[5] C.L. Kirk, Natural frequencies of stiffened rectangular plates, J.
Sound Vib., vol. 13, no. 4, pp. 375–388, 1970.
[6] G. Prathap and T.K. Varadan, Large amplitude flexural vibra-
tion stiffened of plates, J. Sound Vib., vol. 57, no. 4, pp. 583–593,
1978.
[7] T.K. Varadan and K.A.V. Pandalai, Large amplitude flexural vibra-
tion of eccentrically stiffened plates, J. Sound Vib., vol. 67, no. 3,
pp. 329–340, 1979.
[8] G. Aksu and R. Ali, Free vibration analysis of stiffened plates using
finite difference method, J. Sound Vib., vol. 48, no. 1, pp. 15–25,
1976.
[9] G. Aksu, Free vibration analysis of stiffened plates by including the
effect of inplane inertia, J. Appl. Mech., vol. 49, no. 1, pp. 206–212,
1982.
[10] T.S. Koko and M.D. Olson, Non-linear analysis of stiffened plates
using super elements, Int. J. Numer. Meth. Eng., vol. 31, no. 2, pp.
319–343, 1991.
[11] T.S. Koko and M.D. Olson, Vibration analysis of stiffened plates by
super elements, J. Sound Vib., vol. 158, no. 1, pp. 149–167, 1992.
[12] I.E. Harik and M. Guo, Finite element analysis of eccentrically
stiffened plates in free vibration, Comput. Struc., vol. 49, no. 6,
1007–1015, 1993.
[13] C.J. Chen, W. Liu, and S.M. Chern, Vibration analysis of stiffened
plates, Comput. Struc., vol. 50, no. 4, pp. 471–480, 1994.
[14] M. Mukhopadhyay, Vibration and stability of stiffened plates by
semianalytic finite difference method, Part I: Considerationof bend-
ing displacement only, J. Sound Vib., vol. 130, no. 1, pp. 27–39, 1989.
[15] M. Mukhopadhyay, Vibration and stability of stiffened plates by
semianalytic finite difference method, Part II: Consideration of
bending and axial displacements, J. Sound Vib., vol. 130, no. 1,
pp. 41–53, 1989.
[16] A. Mukherjee and M. Mukhopadhyay, Finite element free vibration
of eccentrically stiffened plates, Comput. Struc., vol. 30, no. 6, pp.
1303–1317, 1988.
[17] O.K. Bedair and M. Troitsky, A study of fundamental frequency
characteristics of eccentrically and concentrically simply supported
stiffened plates, Int. J. Mech. Sci., vol. 39, no. 11, pp. 1257–1272,
1997.
[18] O.K. Bedair, Fundamental frequency determination of stiffened
plates using sequential quadratic programming, J. Sound Vib., vol.
199, no. 1, pp. 87–106, 1997.
[19] A.H. Shiekh and M. Mukhopadhyay, Geometric non-linear analy-
sis of stiffened plates by spline finite strip method, Comput. Struc.,
vol. 76, no. 3, pp. 765–785, 2000.
[20] A.H. Shiekh and M. Mukhopadhyay, Free vibration analysis of
stiffened plates with arbitrary planform by the general spline finite
strip method, J. Sound Vib., vol. 162, no. 1, pp. 421–449, 1993.
[21] L.X. Peng, K.M. Liew, and S. Kitipornchai, Buckling and free
vibration analyses of stiffened plates using the FSDT mesh-free
method, J. Sound Vib., vol. 289, no. 3, pp. 147–164, 2006.
[22] E.J. Sapountzakis and V.G. Mokos, An improved model for the
dynamic analysis of plates stiffened by parallel beams, Eng. Struc.,
vol. 30, no. 6, pp. 1720–1733, 2008.
[23] G. Qing, J. Qiu, and Y. Liu, Free vibration analysis of stiffened
laminated plates, Int. J. Solid Struc., vol. 43, pp. 1357–1371, 2006.
[24] L. Dozio and M. Ricciardi, Free vibration analysis of ribbed plates
by a combined analytical–numerical method, J. Sound Vib., vol.
319, pp. 681–697, 2009.
[25] H. Xu, J. Du, and W.L. Li, Vibrations of rectangular plates rein-
forced by any number of beams of arbitrary lengths and placement
angles, J. Sound Vib., vol. 329, no. 18, pp. 3759–3779, 2010.
[26] K.N. Saha, D. Misra, G. Pohit, and S. Ghosal, Large amplitude free
vibration study of square plates under different boundary condi-
tions through a static analysis, J. Vib. Contr., vol. 10, pp. 1009–1028,
2004.
[27] O.K. Bedair, Analysis of stiffened plates under lateral loading using
sequential quadratic programming (SQP), Comput. Struc., vol. 62,
no. 1, pp. 63–80, 1997.
[28] A. Leissa, The free vibration of rectangular plates, J. Sound Vib.,
vol. 31, no. 3, pp. 257–293, 1973.
D
o
w
n
l
o
a
d
e
d

b
y

[
N
a
t
i
o
n
a
l

I
n
s
t
i
t
u
t
e

o
f

T
e
c
h
n
o
l
o
g
y

-

R
o
u
r
k
e
l
a
]

a
t

0
4
:
4
6

2
1

J
u
n
e

2
0
1
3