Composite Structures 99 (2013) 76–87

Contents lists available at SciVerse ScienceDirect

Composite Structures
journal homepage: www.elsevier.com/locate/compstruct

Buckling and free vibration of exponentially graded sandwich plates resting

on elastic foundations under various boundary conditions
Mohammed Sobhy ⇑
Department of Mathematics, Faculty of Science, Kafrelsheik University, Kafr El-Sheikh 33516, Egypt

a r t i c l e i n f o a b s t r a c t

Article history: This paper deals with the vibration and buckling behavior of exponentially graded material (EGM) sand-
Available online 3 December 2012 wich plate resting on elastic foundations under various boundary conditions. New functions for midplane
displacements are suggested to satisfy the different boundary conditions. The elastic foundation is mod-
Keywords: eled as Pasternak’s type which can be either isotropic or orthotropic and as a special case it converges to
Exponentially graded material Winkler’s foundation if the shear layer is neglected. The present EGM sandwich plate is assumed to be
Various boundary conditions made of a fully ceramic core layer sandwiched by metal/ceramic EGM coat. The governing equations
Shear deformation plate theories
of the dynamic response of non-homogeneous composite plates are deduced by using various shear
Free vibration
Buckling loads
deformation plate theories. Numerical results for the natural frequencies and critical buckling loads of
several types of symmetric EGM sandwich plates are presented. The validity of the present solution is
demonstrated by comparison with solutions available in the literature. The influences of the inhomoge-
neity parameter, aspect ratio, thickness ratio and the foundation parameters on the natural frequencies
and critical buckling loads are investigated.
Ó 2012 Elsevier Ltd. All rights reserved.

1. Introduction sinusoidal shear deformation plate theory (SPT). An investigation

The conventional sandwich structures are generally fabricated
form three homogeneous layers, two face sheets adhesively
bonded to the core. However, the sudden change in material prop-
erties across the interface between different materials can result in
large interlaminar stresses. To overcome these adverse effects, a
new class of advanced inhomogeneous composite materials, that
compose of two or more phases with different material properties
and continuously varying composition distribution (using a simple
functional law or an exponential law), has been developed which is
referred to as functionally graded materials (FGMs). Such materials
were introduced as to take advantage of the desired material prop-
erties of each constituent material without interface problems. The
sandwich plate faces are typically made from a mixture of two
materials. While the core of this sandwich plate is fully homoge-
neous material.
Studies related to FGM sandwich structures are few in numbers.
Zenkour [1] was the first to introduce the sandwich structures with
functionally graded faces. He studied the mechanical bending re-
sponse, buckling and free vibration of simply supported FGM sand-
wich plate in that paper. Zenkour and Sobhy [2] investigated the
thermal buckling of various types of FGM sandwich plate using
⇑ Tel.: +20 1091904073.
E-mail addresses: msobhy@sci.kfs.edu.eg, msobhy2011@gmail.com
of bending response of a simply supported FGM viscoelastic sand-
wich beam with elastic core resting on Pasternak’s elastic founda-
tions was presented by Zenkour et al. [3]. Three-dimensional finite
element simulations for analyzing low velocity impact behavior of
sandwich panels with a functionally graded core were conducted
by Etemadi et al. [4]. Anderson [5] presented an analytical three-
dimensional elasticity solution method for a sandwich composite
with a functionally graded core subjected to transverse loading
by a rigid spherical indentor. An exact thermoelasticity solution
for a two-dimensional sandwich structures with functionally
graded coating was presented by Shodja et al. [6]. In Bhangale
and Ganesan [7], the buckling and vibration of a FGM sandwich
beam having viscoelastic layer was studied in thermal environ-
ment by using a finite element formulation.
Composite structures on elastic foundations have wide applica-
tions in modern engineering and pose great technical problems in
structural design. Winkler’s elastic foundation model, which con-
sists of infinitely many closed-spaced linear springs, is a one-
parameter model that is extensively used in practice. A number
of papers have dealt with vibration and buckling of plates on
Winkler’s foundation [8–10]. The limitation of this model is that
it assumes no interaction between the springs. To overcome this
problem, several two-parameter models have been suggested, such
as Filonenko-Borodich, Pasternak and Vlasov and Leontev
foundations, can be used [11]. Buckling loads, free vibrations and

0263-8223/$ - see front matter Ó 2012 Elsevier Ltd. All rights reserved.
http://dx.doi.org/10.1016/j.compstruct.2012.11.018
 M. Sobhy / Composite Structures 99 (2013) 76–87
vibrations with initial in-plane stresses for moderately thick, sim-
ply supported rectangular laminates resting on elastic foundations
(Pasternak’s type) were examined by Aiello and Ombres [12]. Lal
et al. [13] investigated the free vibration analysis of laminated
composite plates resting on elastic foundation undergoing large
amplitude oscillation with random system properties. Malekzadeh
and Karami [14] employed the first-order shear deformation the-
ory to study the free vibration of thick plates of continuously vary-
ing thickness on two-parameter elastic foundations using a
differential quadrature solution. Free vibration analysis of vertical
rectangular Mindlin plates resting on Pasternak’s elastic founda-
tions and fully or partially in contact with fluid on their one side
was investigated, for different combinations of boundary condi-
tions, by Hashemi et al. [15]. Chen et al. [16] studied the bending
and free vibration of arbitrarily thick beams resting on Pasternak’s
elastic foundations.
The most commonly used plate theory is the classical plate the-
ory. However, it neglects transverse shear strains and underpre-
dicts deflections and overpredicts natural frequencies and
buckling loads. In order to obtain accurate predictions of the global
response characteristics and adequately describe the motion of
plate-type structures, the first-order shear deformation plate the-
ory (FPT) is established by Reissner [17]. This theory does not sat-
isfy the stress-free boundary conditions on the surfaces of the plate
and requires an arbitrary shear correction factor. To overcome
these drawbacks, Reddy [18] proposed a third-order shear defor-
mation plate theory (TPT). Touratier [19] and Zenkour [1,20] chose
transverse strain distribution as a sine function. This theory may be
called trigonometric or sinusoidal shear deformation plate theory
(SPT). Hyperbolic shear deformation plate theory (HPT) was pro-
posed by Soldatos [21]. Finally, Karama et al. [22] suggested an
exponential variation (EPT) to investigate the effect of the trans-
verse shear deformation on the bending of composite beams.
Several studies have been performed to analyze the behavior of
composite structures using the above various shear deformation
theories. The mechanical and thermal buckling analysis of func-
tionally graded ceramic metal plates was presented by Zhao
et al. [23] using the FPT. Reddy [24] analyzed the static behavior
of FG rectangular plates based on his third-order shear deforma-
tion plate theory. Zenkour et al. [25–27] employed the SPT to ex-
plain the bending and thermal buckling behavior for various
structures resting on two-parameter elastic foundations. Free
vibrations of cross-ply laminated shells subjected to different sets
of edge boundary conditions have been investigated by Timarci
and Soldatos [28] on the basis of the HPT. In Akavci and Tanrikulu
[29], two hyperbolic displacement models have been used for the
buckling and free vibration analyzes of simply supported orthotro-
pic laminated composite plates. Aydogdu and Taskin [30] em-
ployed the EPT to discuss the free vibration analysis of simply
supported FG beam.
In the present paper, free vibrations and critical buckling
loads for various types of EGM sandwich plates are investigated.
The sandwich plate is assumed to be resting on isotropic or
orthotropic two-parameter elastic foundations. Material proper-
ties of the sandwich plate faces are assumed to vary in the
thickness direction only according to a new exponential law dis-
tribution in terms of the volume fractions of the constituents.
The core layer is still homogeneous and made of an isotropic
material. The governing equations of an EGM sandwich plate
are given based on the sinusoidal shear deformation plate the-
ory. The results obtained as per SPT are compared with those
obtained as per the FPT, TPT, EPT and HPT. Several kinds of sym-
metric sandwich plates are presented. Equilibrium equations of
EGM sandwich plates include the interaction between the plate
and the foundations. The influences of several parameters are
discussed.
77
2. Sandwich structures
An isotropic EGM sandwich plate of constant thickness of h with
cross-sectional dimensions a and b is considered and shown in
Fig. 1. The EGM sandwich plate is defined in the (x, y, z) coordinate
system with x- and y-axes located in the middle plane (z = 0) and
its origin is placed at the corner of the plate. The external bounding
planes of the sandwich plate are defined by z = ±h/2. The vertical
positions of the two interfaces between the core and faces layers
are denoted, respectively, by h1 and h2. The plate is assumed to at-
tach to the foundation so that no separation takes place in the pro-
cess of deformation. The load–displacement relation between the
plate and the supporting foundation is as follows [26,27]:
@ 2 w0 @ 2 w0
P ¼ Kw0  Gx 2
 Gy ; ð1Þ
@x @y2
where P is the density of the reaction force of elastic foundation, K is
the modulus of subgrade reaction (springs stiffness) and Gx, Gy are
the shear moduli of the subgrade (shear layer foundation stiffness).
If the foundation is homogeneous and isotropic, we will get Gx =
Gy = G. If the shear layer foundation stiffness is neglected, the
Pasternak’s foundation becomes the Winkler’s foundation.
The sandwich plate is made of three layers. Its faces are made of
an EGM with material properties varying smoothly in the z direc-
tion only. The EGM are composed from a mixture of metal and
ceramic while the core is fully ceramic. We assume that the com-
position is varied from the interfaces to the bottom and top sur-
faces, i.e. the bottom (z = h/2) and top (z = +h/2) surfaces of the
plate are metal-rich whereas the interfaces (h1, h2) are ceramic-
rich. The volume fraction of the sandwich plate faces varies accord-
ing to a simple power law function of z while that of the core
equals unity, and they are given as:
 k
2z þ 1
V ð1Þ ¼ ; 1=2 6 z 6 h1 ;
2h1 þ 1
V ð2Þ ¼ 1; h1 6 z 6 h2 ; ð2Þ
 k
2z  1
V ð3Þ ¼ ; h 2 6 z 6 1=2;
2h2  1
where z ¼ z=h; hi ¼ hi =h ði ¼ 1; 2Þ and k is the inhomogeneity
parameter which takes values greater than or equal to zero. It is
noted that the core is independent of the value of k which is fully
ceramic. The value of k equaling to zero represents a homogeneous
isotropic ceramic plate and the value of it equaling to infinity rep-
resents a metal–ceramic–metal (m–c–m) sandwich plate. The
above power law assumption reflects a simple rule of mixtures used
to obtain the effective properties of the metal–ceramic sandwich
plate (see Fig. 1).
The mechanical properties of FGMs are often being represented
in the exponentially graded form [31] and power law variations
one [1–3]. Based on a new exponential law distribution, the
Fig. 1. Geometry of the EGM sandwich plate resting on elastic foundations.
78 M. Sobhy / Composite Structures 99 (2013) 76–87

material properties P of the EGM, such as Young’s modulus E and h h

h1 ¼  ; h2 ¼ : ð5Þ
the material density q, are determined as: 6 6
!
  Pc
PðnÞ ðzÞ ¼ P m exp bV ðnÞ ; b ¼ ln ðn ¼ 1; 2; 3Þ; ð3Þ 2.1.3. The (1–2–1) EGM sandwich plate
Pm
Here the core thickness is twice the face thickness (see Fig. 2c).
where subscripts m and c refer to metal and ceramic, and the vol-
Thus, we have
ume fraction V(n) of each layer is given in Eq. (2). Poisson’s ratio m
is assumed to be a constant value through the plate thickness.
h h
h1 ¼  ; h2 ¼ : ð6Þ
2.1. Different types of EGM sandwich plates 4 4

Fig. 2 shows the through-the-thickness variation of the volume 2.1.4. The (1–3–1) EGM sandwich plate
fraction function of the material for k = 0.5, 1.0, 1.5, 3.5, 5.5. Note In this state, as shown in Fig. 2d, the face layers are one third the
that the core of the symmetric plates are fully ceramic while the core thickness layer. So, one gets
bottom and top surfaces of the plate are metal-rich.

3h 3h
2.1.1. The (1–0–1) EGM sandwich plate h1 ¼  ; h2 ¼ : ð7Þ
As shown in Fig. 2a the plate is made of two equal-thickness 10 10
layers, i.e. the core is not found. So, one gets
h1 ¼ h2 ¼ 0: ð4Þ 3. Formulation

In this study, the displacement field, taking into account the

2.1.2. The (1–1–1) EGM sandwich plate
shear deformation effect, is presented for EGM sandwich structures
Here the core thickness is equal to the face thickness (see
as:
Fig. 2b). Then

(1-0-1) (1-1-1)
(a) 0.5 (b) 0.5
0.4 0.4
0.3 0.3
0.2 0.2
0.5
k = 0.5
0.1 k = 5.5 0.1
k = 1.0
z/h

z/h

0 0 k = 1.5
k = 5.5 k = 3.5
-0.1 -0.1
k = 5.5
0.5
-0.2 -0.2
-0.3 -0.3
k = 1.0
-0.4 k = 1.5 -0.4
k = 3.5
-0.5 -0.5
0 0.1 0.2 0.3 0.4 0.5 0.6 0.7 0.8 0.9 1 0 0.1 0.2 0.3 0.4 0.5 0.6 0.7 0.8 0.9 1
Volume Fraction Function Volume Fraction Function

(1-2-1) (1-3-1)
(c) 0.5 (d) 0.5
0.4 0.4
0.3 0.3
0.2 0.2
k = 0.5 k = 0.5
0.1 0.1
k = 1.0 k = 1.0
z/h

z/h

0 k = 1.5 0 k = 1.5
k = 3.5 k = 3.5
-0.1 -0.1
k = 5.5 k = 5.5
-0.2 -0.2
-0.3 -0.3
-0.4 -0.4
-0.5 -0.5
0 0.1 0.2 0.3 0.4 0.5 0.6 0.7 0.8 0.9 1 0 0.1 0.2 0.3 0.4 0.5 0.6 0.7 0.8 0.9 1
Volume Fraction Function Volume Fraction Function

Fig. 2. Through-thickness distributions of volume fraction function for various values of the inhomogeneity parameter k and various types of EGM sandwich plates: (a) the
(1–0–1) EGM sandwich plate, (b) the (1–1–1) EGM sandwich plate, (c) the (1–2–1) EGM sandwich plate, and (d) The (1–3–1) EGM sandwich plate.
 M. Sobhy / Composite Structures 99 (2013) 76–87 79

@w0 The stress resultants Ni, Mi, Ri and Qj are expressed as:
uðx; y; zÞ ¼ u0  z þ f1 ðzÞu1 ;
@x 3 Z
X hn
@w
v ðx; y; zÞ ¼ v 0  z 0 þ f2 ðzÞv 1 ; ð8Þ f Ni ; M i ; Ri g ¼ rðnÞ
i f 1; z; f g dz;
n¼1 hn1
@y
3 Z
X hn
wðx; y; zÞ ¼ w0 ; b df
Qj ¼ K rjðnÞ dz; ði ¼ xx; yy; xy; j ¼ yz; xzÞ; ð14Þ
n¼1 hn1 dz
where u0, v0, w0 and u1, v1 have the same physical meaning as in the
first-order shear deformation plat theory (FPT); they denote the dis- where Kb is the shear correction factor of FPT, h0 = h/2 and h3 = h/2.
placements and rotations of transverse normals on the plane z = 0, Substituting Eqs. (9) and (12) into Eq. (14) gives the constitutive
respectively. Then the displacement field of FPT is obtained by relations as:
setting f1 = f2 = z. The functions f1(z) and f2(z) represent shape func- 8 9 2 38 ð0Þ 9
tions determining the distribution of the transverse shear strains < fNg >
> = ½A ½B ½C > < fe g >
=
6 7
and stresses along the thickness. In this study, the shear strain fMg ¼ 4 ½B ½D ½F 5 feð1Þ g ;
>
: >
; >
: ð2Þ >;
shape functions were chosen as the same in x and y directions fRg ½C ½F ½H fe g ð15Þ
(i.e. f1(z) = f2(z) = f(z)). For the higher-order shear deformation plate    ( ð0Þ
)
theories, they are given in Table 1.
Q yz I44 0 cyz
¼ ;
Moreover, the displacement of the classical plate theory (CPT) Q xz 0 I55 cð0Þ
xz
can be easily obtained by setting f(z) = 0. All of the generalized dis-
where
placements u0, v0, w0, u1, v1 are independent of z.
The displacement model (8) yields the following kinematic
relations: fNg ¼ f Nxx Nxy gT ;
fMg ¼ f M xx M yy M xy g ;
Nyy T

n oT
T
8 9 8 ð0Þ 9 8 ð2Þ 9 8 ð1Þ 9 fRg ¼ f Rxx Ryy Rxy g ; feð0Þ g ¼ eð0Þ xx
ð0Þ
eyy cð0Þ
xy ; ð16Þ
>
> exx >
> >
> e > > e > > e >
< = < xx >= >
< xx > =>
< xx >= n oT n oT
eyy ¼ eð0Þ
yy þ z eð1Þ
yy þ f ðzÞ eð2Þ
yy ; feð1Þ g ¼ eð1Þ
xx eð1Þ
yy
ð1Þ
cxy ; feð2Þ g ¼ eð2Þ
xx eð2Þ
yy
ð2Þ
cxy :
>
> > > > > > > >
:c > ; >
: ð0Þ >
; >
: ð1Þ >; >
: ð2Þ >;
xy cxy cxy cxy ð9Þ
( ) ( ) Here Aij denote the extensional stiffnesses, Dij the bending stiff-
ð0Þ
cyz df ðzÞ cyz nesses, Bij the bending–extensional coupling stiffnesses and Cij, Fij,
ezz ¼ 0; ¼ ; Hij are the stiffnesses associated with the transverse shear effects
cxz dz cð0Þ
xz

3 Z
X hn
where ðnÞ
fAij ; Bij ; Dij ; C ij ; F ij ; Hij g ¼ cij ðzÞf1; z; z2 ; f ; zf ; f 2 g dz;
8 9 9 8 n¼1 hn1
8 ð0Þ 9 @u0 8 ð1Þ 9 @ 2 w0 > >
> e > >
> >
> > e > 2 >
> >
> ði; j ¼ 1; 2; 6Þ; ð17Þ
>
< xx >= >
< >
=
@x >
< xx >
=
@x >
= >
<
ð0Þ @v 0 ð1Þ 2
@ w0
eyy ¼ @y ; eyy ¼  @y2 ; ð10Þ
 2
>
> > > > > > > > 3 Z
: ð0Þ >; >
>
:
>
@u0 >
;
>
: ð1Þ > ; >
>
>
>
> X hn
df
cxy @v 0
@x
þ @y cxy : 2 @ 2 w0 >
; b
Iii ¼ K
ðnÞ
cii dz; ði ¼ 4; 5Þ: ð18Þ
@x@y
n¼1 hn1 dz
8 ð2Þ 9 8 9 @u1
> > ( )
< exx >
> = >
< >
= @x ð0Þ
cyz

v1 
ð2Þ @v 1 4. Equations of motion
eyy ¼ ; @y ¼ : ð11Þ
> >
: ð2Þ ; >> > ð0Þ
cxz u1
: @v 1 @u1 >
;
c xy þ @x @y The virtual work principle of the present EGM sandwich plate
may be written as:
Hooke’s law for a plate is defined as: Z (Z 
t2
1  ðnÞ 2 ðnÞ ðnÞ ðnÞ ðnÞ ðnÞ

8 d c11 exx þ 2c12 exx eyy þ c22 e2yy þ c44 e2yz þ c55 e2xz þ c66 e2xy
>
>
rxx 9
>
>
ðnÞ 2
c11 c12 0 0 0
3ðnÞ 8
>
>
exx 9>
>
t1 V 2
>
> > > >
>
>
> r >
< yy >
>
>
6
6 c12 c22 0 0 0
7 >
7 > > e >
>
> € du þ v€ dv þ wdw
ðnÞ
þ q ðu € Þ dV
7 > >
yy
= 6 < = Z ! )
6 7 cyz ; 2
@ 2 w0 @ 2 w0
ryz ¼6 0 0 c44 0 0 7 ð12Þ @ w0
>
> >
> 6 7 > > þ Rxx þ R þ R þ P dw dX dt ¼ 0; ð19Þ
>
> rxz >
> 6 0 7 > >c > > @x2
yy
@y2
xy
@x@y
0
>
> >
> 4 0 0 c55 0 5 > >
>
>
xz >
>
X
>
: >
; >
: >
;
rxy 0 0 0 0 c66 cxy where Rxx ; Ryy and Rxy are the constant in-plane edge loads.
Substituting Eqs. (8) and (9) into Eq. (19) gives the general
where equations of motion for the shear deformation plate theories as:

ðnÞ ðnÞ EðnÞ ðzÞ ðnÞ ðnÞ ðnÞ ðnÞ ðnÞ EðnÞ ðzÞ @N xx @N xy @w€0
c11 ¼ c22 ¼ ; c12 ¼ mc11 ; c44 ¼ c55 ¼ c66 ¼ : ð13Þ þ ¼ .00 u
€ 0  .01 þ .10 u
€1 ;
1m 2 2ð1 þ mÞ @x @y @x
@N xy @N yy @w€0
þ ¼ .00 v€ 0  .01 þ .10 v€ 1 ;
@x @y @y
Table 1 @ 2 M xx @ 2 M xy @ 2 M yy @ 2 w0 @ 2 w0 @ 2 w0
Different transverse shear deformation functions. 2
þ2 þ 2
 Rxx 2
 Rxy  Ryy P
@x @x@y @y @x @x@y @y2
  !   ð20Þ
Theory Function € 0 @ v€ 0
@u @2 w
€ 0 @2 w €0 1 @ u1
€ @ v€ 1
¼ .00 w
€ 0 þ .01 þ  .02 þ þ . 1 þ ;
SPT f ðzÞ ¼ ph sin phz @x @y @x2 @y2 @x @y
h i
TPT f ðzÞ ¼ z 1  43 hz
2 @Rxx @Rxy @w€0
h i þ  Q xz ¼ .10 u
€ 0  .11 þ .20 u
€1 ;
EPT 2 @x @y @x
f ðzÞ ¼ z exp 2 hz
@Rxy @Ryy @w€0
HPT f ðzÞ ¼ h sinh z
 z sinh 1 þ  Q yz ¼ .10 v€ 0  .11 þ .20 v€ 1 ;
h 2 @x @y @y
80 M. Sobhy / Composite Structures 99 (2013) 76–87

where the inertias .ji are defined by: Table 3  2 pffiffiffiffiffiffiffiffiffiffiffiffi

Comparison of free vibration xpb2 qh=D of a simply supported homogeneous
3 Z hn
X square plate (a/b = 1, k = 0) resting on Pasternak’s elastic foundations (J2 = 10).
.ji ¼ qðnÞ zi f j dz; ði; j ¼ 0; 1; 2; i þ j 6 2Þ: ð21Þ
n¼1 hn1 r s h/b J1 Mindlin [32] Present
1 1 0.01 102 2.6551 2.6551
Using Eqs. (10), (11) and (15), we can get the governing
5  102 3.3400 3.3400
Eqs. (20) in terms of the displacements as:
0.1 2  102 2.7842 2.7842
@w0 103 3.9805 3.9806
L1 u0 þ AL2 v 0  L3 þ L4 u1 þ CL2 v 1 ¼ 0;
@x
@w0 2 1 0.01 102 5.5718 5.5718
AL2 u0 þ L5 v 0  L3 þ CL2 u1 þ L6 v 1 ¼ 0; 5  102 5.9287 5.9287
@y
    0.1 2  102 5.3043 5.3051
@u0 @ v 0 @u1 @ v 1 103 6.0078 6.0085
L3 þ þ L7 w0  L8 þ ¼ 0; ð22Þ
@x @y @x @y
@w0 2 2 0.01 102 8.5405 8.5405
L4 u0 þ CL2 v 0  L8 þ L9 u1 þ HL2 v 1 ¼ 0; 5  102 8.7775 8.7775
@x
@w0 0.1 2  102 7.7287 7.7311
CL2 u0 þ L6 v 0  L8 þ HL2 u1 þ L10 v 1 ¼ 0; 103 8.2214 8.2237
@y
where the operator Li are given by:
2
 2  .0 @ ; @2 @2
L 1 ¼ Ar x 0 L2 ¼ ðm þ m
Þ ; L3 ¼ Br2  .01 2 ; Table 4
@t2 @x@y @t  2 pffiffiffiffiffiffiffiffiffiffiffiffi
2 2 2 Comparison of free vibration xpb2 qh=D of a clamped homogeneous square plate
 2  .1 @ ;
L4 ¼ C r L5 ¼ Ar   . @ ; L6 ¼ C r
2 0  2  .1 @ ; (a/b = 1, k = 0) resting on Winkler’s elastic foundation (r = s = 1, h/b = 0.015, m = 0.15,
x 0 y 0 y 0
@t2 @t 2
@t 2 J2 = 0).
2 2 2
4 2 @ @ @ @2
L7 ¼ Dr þ ð.00  .02 r Þ 2 þ Rxx 2 þ Rxy þ Ryy 2 þ K  Gr2 ; J1 Classical [33] Present
@t @x @x@y @y
@2 2 2 1390.2 5.2510 5.3330
L8 ¼ F r2  .11 2 ;  2  .0 @  I; L10 ¼ Hr
L9 ¼ H r  2  .0 @  I; 2780.4 6.4686 6.5349
x 2 y 2
@t @t2 @t2
ð23Þ

in which
Table 5
2 ¼ @2 @2 2 ¼ m @2 @2 @2 @2
r x þm
 ; r y
 þ ; r2 ¼ 2 þ 2 ; Comparison of critical buckling load ½Ra2 =ðDp2 Þ of a simply supported thin
@x2 @y2 @x2 @y2 @x @y homogeneous square plate (a/h = 1000, k = 0) resting on Pasternak’s elastic founda-
tions (s = 1, f = 0).
r4 ¼ r2 ðr2 Þ;
ð24Þ
;  ; X66 ¼ m  ; I44 ¼ I55 ¼ I; r J1 J2 Green’s functions [34] Mindlin [35] Present
X11 ¼ X22 ¼ X X12 ¼ mX X
1 0 0 4.000 3.99998 3.99998
1m
m ¼ ðX ¼ A; B; C; D; F; HÞ: 2 100 18.92 18.9151 18.91506
2 1 100 0 5.027 5.02658 5.02658
2 100 19.17 19.1717 19.17171

5. Exact solutions for EGMs sandwich plates

The exact solution of Eqs. (22) for the EGMs sandwich plate un- Simply supported (S):
der various boundary conditions can be constructed. The plate is
assumed to have simply-supported (S), clamped (C) or free (F) v 0 ¼ w0 ¼ v 1 ¼ Nxx ¼ Mxx ¼ Rxx ¼ 0 at x ¼ 0; a;
ð25Þ
edges or have combinations of them, and they are given as: u0 ¼ w0 ¼ u1 ¼ Nyy ¼ Myy ¼ Ryy ¼ 0 at y ¼ 0; b:

Table 2
The admissible functions Xr(x) and Ys(y).

Boundary conditions The functions Xr and Ys

At x = 0, a At y = 0, b Xr(x) Ys(y)
SSSS X r ð0Þ ¼ X 00r ð0Þ ¼ 0 Y s ð0Þ ¼ Y 00s ð0Þ ¼ 0 sin(kx) sin(ly)
X r ðaÞ ¼ X 00r ðaÞ ¼ 0 Y s ðbÞ ¼ Y 00s ðbÞ ¼ 0
CSSS X r ð0Þ ¼ X 0r ð0Þ ¼ 0 Y s ð0Þ ¼ Y 00s ð0Þ ¼ 0 sin(kx)[cos(kx)  1] sin(ly)
X r ðaÞ ¼ X 00r ðaÞ ¼ 0 Y s ðbÞ ¼ Y 00s ðbÞ ¼ 0
CSCS X r ð0Þ ¼ X 0r ð0Þ ¼ 0 Y s ð0Þ ¼ Y 0s ð0Þ ¼ 0 sin(kx)[cos(kx)  1] sin(ly)[cos(ly)  1]
X r ðaÞ ¼ X 00r ðaÞ ¼ 0 Y s ðbÞ ¼ Y 00s ðbÞ ¼ 0
CCSS X r ð0Þ ¼ X 0r ð0Þ ¼ 0 Y s ð0Þ ¼ Y 00s ð0Þ ¼ 0 sin2(kx) sin(ly)
X r ðaÞ ¼ X 0r ðaÞ ¼ 0 Y s ðbÞ ¼ Y 00s ðbÞ ¼ 0
CCCC X r ð0Þ ¼ X 00r ð0Þ ¼ 0 Y s ð0Þ ¼ Y 00s ð0Þ ¼ 0 sin2(kx) sin2(ly)
X r ðaÞ ¼ X 0r ðaÞ ¼ 0 Y s ðbÞ ¼ Y 0s ðbÞ ¼ 0
FFCC X 00r ð0Þ ¼ X 000
r ð0Þ ¼ 0 Y s ð0Þ ¼ Y 0s ð0Þ ¼ 0 cos2(kx)[sin2(kx) + 1] sin2(ly)
X 00r ðaÞ ¼ X 000
r ðaÞ ¼ 0 Y s ðbÞ ¼ Y 0s ðbÞ ¼ 0

()0 Denotes the derivative with respect to the corresponding coordinates.

 M. Sobhy / Composite Structures 99 (2013) 76–87 81

Table 6
2
Comparison of critical buckling load ðRb =DÞ of a simply supported homogeneous plate (k = 0) resting on Pasternak’s elastic foundations (s = 1, f = 0).

a/b a/h (J1, J2)

Mindlin [35] Present
(0, 0) (100, 10) (1000, 100) (0, 0) (100, 10) (1000, 100)
0.5 1000 61.68481 152.2131 704.5892 61.68481 152.2131 704.5892
100 61.66411 152.1931 704.3862 61.66331 152.1921 704.3782
10 59.66291 150.1911 686.1712 59.58871 150.1171 685.5672
5 54.32071 – 643.5003 54.08591 144.6141 641.3803
1 1000 39.47821 69.61031 212.0142 39.47821 69.61031 212.0142
100 39.45701 69.58911 210.1612 39.45621 69.58831 211.9282
10 37.44771 67.57981 204.6512 37.37531 67.50741 204.4162
5 32.44141 55.02892 174.9763 32.23981 54.61162 174.3913
2 1000 39.47762 45.11082 85.25633 39.47752 45.11082 85.25623
100 39.39302 45.02622 85.09523 39.38972 45.02292 85.08893
10 32.44142 37.51823 72.82904 32.23982 37.85813 72.41174
5 19.22553 22.74764 – 19.04003 22.67784 52.22764

Mode shape (r) is denoted by the superscript numbers.

Table 7
Effects of elastic foundation stiffnesses J1 and J2 and side-to-thickness ratio a/h on the free vibration x⁄ of various types of simply supported sandwich square plates (k = 1.5).

Scheme Theory J1 = J2 = 0 J1 = 100, J2 = 0 J1 = 100, J2 = 100

a/h = 5 10 20 a/h = 5 10 20 a/h = 5 10 20
1–0–1 FPT 0.9547 1.0167 1.0347 1.4061 1.4610 1.4775 4.7803 4.8851 4.9134
TPT 0.9647 1.0198 1.0356 1.4121 1.4631 1.4781 4.7807 4.8854 4.9135
SPT 0.9655 1.0200 1.0356 1.4125 1.4633 1.4781 4.7808 4.8854 4.9135
EPT 0.9663 1.0203 1.0357 1.4131 1.4635 1.4782 4.7808 4.8854 4.9135
HPT 0.9643 1.0196 1.0355 1.4119 1.4630 1.4781 4.7805 4.8854 4.9135
1–1–1 FPT 1.0717 1.1367 1.1555 1.1563 1.5227 1.5401 4.6538 4.7513 4.7788
TPT 1.0807 1.1395 1.1563 1.4695 1.5247 1.5407 4.6537 4.7517 4.7789
SPT 1.0817 1.1396 1.1563 1.4697 1.5248 1.5407 4.6537 4.7517 4.7789
EPT 1.0815 1.1398 1.1563 1.4700 1.5249 1.5407 4.6537 4.7517 4.7790
HPT 1.0816 1.1398 1.1563 1.4703 1.5249 1.5407 4.6538 4.7518 4.7790
1–2–1 FPT 1.1797 1.2555 1.2775 1.5340 1.6027 1.6231 4.6102 4.7073 4.7355
TPT 1.1872 1.2578 1.2781 1.5392 1.6045 1.6236 4.6100 4.7076 4.7357
SPT 1.1872 1.2578 1.2781 1.5392 1.6045 1.6236 4.6099 4.7076 4.7357
EPT 1.1873 1.2579 1.2781 1.5392 1.6045 1.6236 4.6099 4.7076 4.7357
HPT 1.1922 1.2593 1.2785 1.5430 1.6057 1.6239 4.6105 4.7080 4.7358
1–3–1 FPT 1.2605 1.3460 1.3710 1.5912 1.6688 1.6920 4.5914 4.6898 4.7190
TPT 1.2666 1.3480 1.3716 1.5956 1.6704 1.6924 4.5911 4.6901 4.7192
SPT 1.2663 1.3479 1.3716 1.5954 1.6703 1.6924 4.5910 4.6901 4.7192
EPT 1.2662 1.3478 1.3715 1.5953 1.6703 1.6924 4.5909 4.6900 4.7192
HPT 1.2753 1.3506 1.3723 1.6024 1.6724 1.6930 4.5921 4.6907 4.7194

Table 8
Effects of elastic foundation stiffnesses J1 and J2 and side-to-thickness ratio a/h on the critical buckling R⁄ of various types of simply supported sandwich square plates (k = 0.5).

Scheme Theory J1 = J2 = 0 J1 =100 , J2 = 0 J1 = 100, J2 = 100

a/h = 5 10 20 a/h = 5 10 20 a/h = 5 10 20
1–0–1 FPT 2.5154 2.7987 2.8797 4.2783 4.5616 4.6427 39.0768 39.3601 39.4412
TPT 2.5592 2.8119 2.8832 4.3220 4.5748 4.6461 39.1206 39.3734 39.4447
SPT 2.5618 2.8127 2.8834 4.3247 4.5756 4.6463 39.1232 39.3741 39.4449
EPT 2.5652 2.8137 2.8837 4.3281 4.5766 4.6466 39.1266 39.3752 39.4451
HPT 2.5834 2.8193 2.8852 4.3463 4.5822 4.6481 39.1448 39.3808 39.4466
1–1–1 FPT 3.0560 3.4014 3.5003 4.8189 5.1643 5.2632 39.6175 39.9628 40.0617
TPT 3.1014 3.4151 3.5039 4.8643 5.1781 5.2668 39.6629 39.9766 40.0653
SPT 3.1030 3.4156 3.5040 4.8659 5.1785 5.2669 39.6644 39.9770 40.0655
EPT 3.1054 3.4163 3.5042 4.8683 5.1792 5.2671 39.6668 39.9777 40.0656
HPT 3.1399 3.4269 3.5070 4.9029 5.1898 5.2699 39.7014 39.9883 40.0684
1–2–1 FPT 3.4772 3.8906 4.0097 5.2401 5.6535 5.7726 40.0386 40.4520 40.5712
TPT 3.5165 3.9026 4.0129 5.2795 5.6655 5.7758 40.0780 40.4641 40.5744
SPT 3.5165 3.9026 4.0129 5.2795 5.6655 5.7758 40.0780 40.4640 40.5744
EPT 3.5176 3.9028 4.0130 5.2805 5.6657 5.7759 40.0790 40.4643 40.5744
HPT 3.5799 3.9220 4.0180 5.3429 5.6850 5.7810 40.1414 40.4835 40.5795
1–3–1 FPT 3.7922 4.2636 4.4004 5.5551 6.0265 6.1633 40.3537 40.8251 40.9618
TPT 3.8253 4.2738 4.4031 5.5882 6.0367 6.1660 40.3867 40.8353 40.9645
SPT 3.8243 4.2734 4.4030 5.5872 6.0364 6.1659 40.3857 40.8349 40.9644
EPT 3.8245 4.2735 4.4030 5.5875 6.0364 6.1659 40.3860 40.8349 40.9644
HPT 3.9114 4.3004 4.4101 5.6743 6.0634 6.1730 40.4728 40.8619 40.9716
82 M. Sobhy / Composite Structures 99 (2013) 76–87

Table 9
Effects of inhomogeneity parameter k and side-to-thickness ratio a/h on the free vibration x⁄ of (1–1–1) EGM sandwich plate (b/a = 2, J1 = J2 = 10).

B.C. Theory k=0 k = 0.5 k = 3.5

a/h = 5 10 20 a/h = 5 10 20 a/h = 5 10 20
FFCC FPT 2.4366 2.8347 2.9891 2.0924 2.3194 2.4003 1.8535 1.9790 2.0210
TPT 2.4366 2.8351 2.9891 2.1155 2.3289 2.4031 1.8780 1.9883 2.0237
SPT 2.4408 2.8354 2.9892 2.1164 2.3292 2.4032 1.8795 1.9889 2.0238
EPT 2.4432 2.8363 2.9895 2.1178 2.3297 2.4033 1.8811 1.9895 2.0240
HPT 2.5239 2.8742 3.0011 2.1359 2.3371 2.4055 1.8731 1.9865 2.0232
CCCC FPT 2.3359 2.6010 2.6940 2.0111 2.1586 2.2072 1.8022 1.8847 1.9107
TPT 2.3372 2.6012 2.6940 2.0250 2.1638 2.2087 1.8157 1.8895 1.9121
SPT 2.3379 2.6013 2.6941 2.0255 2.1640 2.2088 1.8165 1.8898 1.9121
EPT 2.3393 2.6018 2.6942 2.0262 2.1642 2.2089 1.8174 1.8901 1.9122
HPT 2.3937 2.6239 2.7006 2.0377 2.1684 2.2100 1.8133 1.8886 1.9118
CSCS FPT 2.3323 2.5262 2.5904 2.0265 2.1364 2.1712 1.8456 1.9117 1.9319
TPT 2.3328 2.5262 2.5904 2.0349 2.1395 2.1720 1.8531 1.9145 1.9327
SPT 2.3332 2.5263 2.5905 2.0352 2.1396 2.1721 1.8535 1.9146 1.9327
EPT 2.3341 2.5266 2.5905 2.0356 2.1398 2.1721 1.8540 1.9148 1.9328
HPT 2.3708 2.5404 2.5944 2.0431 2.1423 2.1728 1.8519 1.9140 1.9325
CCSS FPT 2.2578 2.5112 2.5997 1.9476 2.0880 2.1341 1.7472 1.8273 1.8519
TPT 2.2590 2.5114 2.5997 1.9610 2.0930 2.1355 1.7622 1.8320 1.8532
SPT 2.2596 2.5115 2.5998 1.9614 2.0932 2.1356 1.7630 1.8323 1.8533
EPT 2.2610 2.5120 2.5999 1.9621 2.0934 2.1357 1.7638 1.8326 1.8534
HPT 2.3133 2.5331 2.6060 1.9731 2.0974 2.1367 1.7599 1.8311 1.8530
CSSS FPT 2.2176 2.3921 2.4491 1.9319 2.0302 2.0610 1.7653 1.8242 1.8421
TPT 2.2181 2.3922 2.4493 1.9395 2.0330 2.0618 1.7720 1.8267 1.8428
SPT 2.2185 2.3923 2.4493 1.9397 2.0331 2.0618 1.7724 1.8268 1.8428
EPT 2.2193 2.3925 2.4494 1.9401 2.0332 2.0619 1.7728 1.8270 1.8429
HPT 2.2526 2.4048 2.4528 1.9468 2.0354 2.0625 1.7709 1.8262 1.8426
SSSS FPT 1.5386 1.5942 1.6113 1.3971 1.4304 1.4399 1.3361 1.3589 1.3653
TPT 1.5387 1.5947 1.6113 1.3989 1.4310 1.4401 1.3375 1.3594 1.3655
SPT 1.5388 1.5947 1.6113 1.3990 1.4310 1.4401 1.3376 1.3594 1.3655
EPT 1.5390 1.5948 1.6113 1.3991 1.4311 1.4401 1.3377 1.3595 1.3655
HPT 1.5482 1.5978 1.6121 1.4007 1.4315 1.4403 1.3373 1.3593 1.3654

Table 10
Effects of inhomogeneity parameter k and side-to-thickness ratio a/h on the critical buckling R⁄ of (1–1–1) EGM sandwich plate (b/a = 2, J1 = J2 = 10).

B.C. Theory k=0 k = 0.5 k = 3.5

a/h = 5 10 20 a/h = 5 10 20 a/h = 5 10 20
FFCC FPT 15.2693 20.4469 22.5889 10.5119 12.7398 13.5537 7.5375 8.4498 8.7524
TPT 15.3100 20.4524 22.5893 10.7685 12.8483 13.5856 7.5375 8.5326 8.7755
SPT 15.3262 20.4573 22.5906 10.7791 12.8519 13.5866 7.7767 8.5375 8.7769
EPT 15.3581 20.4701 22.5943 10.7947 12.8576 13.5882 7.7922 8.5427 8.7783
HPT 16.4396 21.0333 22.7718 10.9820 12.9404 13.6126 7.7130 8.5156 8.7709
CCCC FPT 12.9480 15.8035 16.8213 8.9602 10.1273 10.5061 6.5737 7.0322 7.1710
TPT 12.9640 15.8053 16.8214 9.1032 10.1789 10.5205 6.6922 7.0708 7.1813
SPT 12.9719 15.8075 16.8220 9.1086 10.1806 10.5209 6.6994 7.0730 7.1819
EPT 12.9892 15.8136 16.8237 9.1167 10.1833 10.5216 6.7073 7.0754 7.1826
HPT 13.6482 16.0906 16.9044 9.2238 10.2228 10.5326 6.6669 7.0629 7.1792
CSCS FPT 10.8802 12.4597 12.9692 7.6668 8.2896 8.4768 5.8071 6.0450 6.1130
TPT 10.8866 12.4603 12.9692 7.7464 8.3154 8.4837 5.8704 6.0641 6.1180
SPT 10.8906 12.4613 12.9695 7.7493 8.3163 8.4840 5.8742 6.0652 6.1183
EPT 10.9001 12.4644 12.9703 7.7536 8.3176 8.4843 5.8783 6.0663 6.1186
HPT 11.2895 12.6060 13.0096 7.8139 8.3374 8.4896 5.8572 6.0602 6.1170
CCSS FPT 12.7243 15.4988 16.4857 8.8365 9.9698 10.3370 6.5105 6.9555 7.0900
TPT 12.7398 15.5004 16.4858 8.9755 10.0199 10.3509 6.6256 6.9929 7.1000
SPT 12.7474 15.5026 16.4864 8.9808 10.0215 10.3514 6.6326 6.9950 7.1006
EPT 12.7642 15.5085 16.4880 8.9886 10.0241 10.3521 6.6402 6.9974 7.1012
HPT 13.4052 15.7773 16.5662 9.0927 10.0624 10.3627 6.6010 6.9853 7.0980
CSSS FPT 10.6131 12.0668 12.5311 7.5143 8.0853 8.2557 5.7275 5.9451 6.0069
TPT 10.6189 12.0673 12.5311 7.5876 8.1089 8.2620 5.7856 5.9624 6.0114
SPT 10.6225 12.0683 12.5314 7.5902 8.1096 8.2622 5.7891 5.9634 6.0117
EPT 10.6312 12.0711 12.5321 7.5942 8.1108 8.2625 5.7928 5.9645 6.0120
HPT 10.9920 12.2004 12.5678 7.6497 8.1289 8.2674 5.7735 5.9589 6.0105
SSSS FPT 7.5245 7.9088 8.0175 5.7741 5.9182 5.9575 4.8112 4.8643 4.8784
TPT 7.5252 7.9089 8.0175 5.7935 5.9237 5.9590 4.8259 4.8683 4.8795
SPT 7.5261 7.9091 8.0175 5.7942 5.9239 5.9590 4.8267 4.8685 4.8795
EPT 7.5284 7.9097 8.0177 5.7952 5.9242 5.9591 4.8277 4.8688 4.8796
HPT 7.6317 7.9407 8.0258 5.8101 5.9284 5.9602 4.8229 4.8675 4.8792
 M. Sobhy / Composite Structures 99 (2013) 76–87 83

Clamped (C): Substituting expressions (28) into the governing Eqs. (22) and
multiplying each equation by the corresponding eigenfunction
u0 ¼ v 0 ¼ w0 ¼ u1 ¼ v 1 ¼ 0 at x ¼ 0; a; y ¼ 0; b: ð26Þ then integrating over the domain of solution, we can obtain, after
Free (F): some mathematical manipulations, the following equations:
2 38 9
Mxx ¼ Mxy ¼ Q xz ¼ 0 at x ¼ 0; a; U 1rs V 1rs W 1rs W1rs U1rs >> U rs >
>
6 >
> >
>
Myy ¼ Mxy ¼ Q yz ¼ 0 at y ¼ 0; b:
ð27Þ
6 U 2rs V 2rs W 2rs W2rs U2rs 7
7>
>
< V rs >
>
=
6 7
6 U 3rs V 3rs W 3rs U3rs 7
The following representation for the displacement quantities, 6 W3rs W
7> rs > ¼ 0; ð29Þ
6 7> >
that satisfy the above boundary conditions, is appropriate in the 4 U 4rs V 4rs W 4rs W4rs U4rs 5>
>
> Wrs >>
>
>
: >
;
case of our problem: U 5rs V 5rs W 5rs W5rs U5rs Urs

@X r ðxÞ where the coefficients Uirs, Virs, Wirs, Wirs and Uirs (i = 1, 2, 3, 4, 5) are
fu0 ; u1 g ¼ fU rs ; Wrs g Y s ðyÞeðixtÞ ;
@x given in Appendix A. The non-trivial solution is obtained when the
@Y s ðyÞ ðixtÞ ð28Þ determinant jUirs Virs Wirs Wirs Uirsj equals zero. For the free vibration
fv 0 ; v 1 g ¼ fV rs ; Urs gX r ðxÞ e ;
@y problem, we have Rxx ¼ Ryy ¼ Rxy ¼ 0. While for the buckling anal-
fw0 g ¼ fW rs gX r ðxÞY s ðyÞeðixtÞ ; ysis, we put x ¼ Rxy ¼ 0; Rxx ¼ R and Ryy ¼ fR, i.e. f ¼ Ryy =Rxx .

where Urs, Vrs, Wrs, Urs and Wrs are arbitrary parameters and x = xrs
denotes the eigenfrequency associated with (rth, sth) eigenmode.
The functions Xr(x) and Ys(y) are suggested here to satisfy at least
the geometric boundary conditions given in Eqs. (25)–(27), and rep-
resent approximate shapes of the deflected surface of the plate.
These functions, for the different cases of boundary conditions,
are listed in Table 2 noting that k = rp/a, l = sp/b.
6. Numerical results and discussions
Numerical results for free vibration and buckling load are pre-
sented for symmetric rectangular EGM sandwich plates resting
on two-parameter elastic foundations with various cases of the
boundary conditions. The combination of materials consists of alu-
minum and alumina with the following material properties:

(1-0-1) (1-1-1)
(a) 5.1
(b) 4.87
5.05 4.84

5 4.81

4.95 4.78

4.9 4.75
ω*

ω*

4.85 4.72

4.8 4.69

4.66
4.75 k = 0.5 k = 0.5
k = 1.5 4.63
4.7 k = 1.5
k = 3.5 4.6 k = 3.5
4.65 k = 5.5 k = 5.5
4.57
4.6
5 7 9 11 13 15 17 19 21 5 7 9 11 13 15 17 19 21
a/h a/h

(1-2-1) (1-3-1)
(c) 4.795 (d) 4.76
4.77 4.74

4.745 4.72
4.7
4.72
4.68
4.695
4.66
ω*

ω*

4.67
4.64
4.645
4.62
4.62
k = 0.5 4.6 k = 0.5
4.595 k = 1.5 4.58 k = 1.5
k = 3.5 k = 3.5
4.57 4.56
k = 5.5 k = 5.5
4.545 4.54
5 7 9 11 13 15 17 19 21 5 7 9 11 13 15 17 19 21
a/h a/h
⁄
Fig. 3. Free vibration x versus the ratio a/h for various values of the inhomogeneity parameter k and various types of simply-supported EGM sandwich square plates resting
on elastic foundations (J1 = J2 = 100).
84 M. Sobhy / Composite Structures 99 (2013) 76–87

Em ¼ 70 GPa; qm ¼ 2707 kg=m3 ; for aluminum; Tables 3 and 4, respectively. The agreeability between the present
Ec ¼ 380 GPa; qc ¼ 3800 kg=m3 ; for alumina; results and the published ones can be clearly noted, especially for
the thin plate. It is also noted that the natural frequencies of sim-
mm ¼ mc ¼ m ¼ 0:3: ply-supported and clamped homogeneous square plates increase
The used non-dimensional parameters are as Winkler’s foundation parameter J1 increases. Table 5 shows
the uniaxial buckling load ½Ra2 =ðDp2 Þ for a simply supported thin
xa2 Ra2 Ka4 Gx a2 Gy a2 homogeneous square plate without or resting on elastic founda-
x ¼ ; R ¼ 3
; J1 ¼ ; J2 ¼ ¼ ; tions using the present theory, Green’s functions presented in
h 100h D D D
3 Ref. [34] and Mindlin plate theory introduced in Ref. [35]. Good
h Ec
D¼ : agreement is achieved between the present solution and the pub-
12ð1  m2 Þ
lished ones. Also, comparison of uniaxial critical buckling load
2
Many examples have been solved numerically using the follow- ðRb =DÞ is achieved in Table 6 for various values of the ratios a/b
ing fixed data (unless otherwise stated) a/h = 10, r = s = 1, and a/h. It is obvious that the results of the present theory closely
J1 = J2 = 100, f = 1. agree with Mindlin theory [35] results.
The general approach outlined in the previous sections for the Tables 7 and 8 exhibit the effects of the elastic foundation
free vibration and buckling analyses of the homogeneous and parameters J1 and J2 and side-to-thickness ratio a/h on the natural
EGM sandwich plates resting on elastic foundations is illustrated frequencies x⁄ and buckling load R⁄, respectively, of various types
in this section using the sinusoidal shear deformation plate theory of simply supported sandwich square plates on the basis of the dif-
and compared with those obtained by different FPT, TPT, EPT and ferent present shear deformation plate theories. It is clear that the
HPT. frequencies x⁄ and buckling load R⁄ increase with the increase of
As a check on the numerical accuracy ofthe p theory the core thickness of the EGM sandwich plates without or resting
2 ffiffiffiffiffiffiffiffiffiffiffiffiand formu-
lation, results of the natural frequencies xpb2 qh=D were ob- on elastic foundation excluding x⁄ for the plates resting on Paster-
tained for homogeneous isotropic plates (k = 0) resting on elastic nak’s foundations, the variation of them is reversed. In addition,
foundations by the present SPT and compared with those obtained the vibration frequencies and buckling load are increasing with
by Mindlin plate theory [32] and classical one [33] as shown in the existence of the elastic foundations. The inclusion of the

(1-0-1) (1-1-1)
(a) 10.5 (b) 10.5
9.9 9.9
9.3 9.3
k = 0.0
8.7 k = 0.0
k = 0.5 8.7 k = 0.5
8.1 k = 1.5
k = 1.5
k = 3.5 8.1
k = 3.5
R*

R*

7.5
7.5
6.9
6.9
6.3
6.3
5.7
5.7
5.1

4.5 5.1
5 7 9 11 13 15 17 19 21 5 7 9 11 13 15 17 19 21
a/h a/h

(1-2-1) (1-3-1)
(c) 10.5 (d) 10.5
9.9 9.9
k = 0.0
k = 0.0
9.3 k = 0.5
9.3 k = 0.5
k = 1.5
k = 1.5
8.7 k = 3.5
8.7 k = 3.5
R*

R*

8.1
8.1
7.5
7.5
6.9
6.9
6.3
6.3
5.7
5 7 9 11 13 15 17 19 21 5 7 9 11 13 15 17 19 21
a/h a/h
⁄
Fig. 4. Critical buckling R versus the ratio a/h for various values of the inhomogeneity parameter k and various types of simply-supported EGM sandwich square plates
resting on elastic foundations (J1 = J2 = 10, f = 1).
 M. Sobhy / Composite Structures 99 (2013) 76–87 85

(1-2-1) (1-1-1)
11
3.1 J1 = J2 = 100
FFCC
2.95 CCCC 10 J1 = 200, J2 = 100
J1 = 100, J2 = 120
2.8 9
CSCS
2.65
8
2.5
CCSS

ω*
7
ω*

2.35
CSSS 6 CCCC
2.2

2.05 5
SSSS
1.9 4
SSSS
1.75
3
0.4 0.7 1 1.3 1.6 1.9 2.2 2.5 2.8 3.1
1.6
5 7 9 11 13 15 17 19 21 b/a
a/h
Fig. 7. Free vibration x⁄ versus the aspect ratio b/a of simply-supported and
⁄
Fig. 5. Free vibration x versus the side-to-thickness ratio a/h of the (1–2–1) EGM clamped sandwich plate for different values of foundation stiffnesses J1 and J2
sandwich square plate resting on Winkler’s elastic foundation with various (a/h = 10, k = 0.5).
boundary conditions (J1 = 100, k = 0.5).

(1-2-1) (1-1-1)
65
14 J1 = J2 = 100
FFCC
62
J1 = 200, J2 = 100
13
J1 = 100, J2 = 120
59
12 CCCC
56
11 CCCC
53
10
R*
R*

CSCS
50
9
CCSS
8 47
CSSS
SSSS
7 44

6 SSSS 41

5 38
5 7 9 11 13 15 17 19 21 0.4 0.7 1 1.3 1.6 1.9 2.2 2.5 2.8 3.1
a/h b/a
⁄ ⁄
Fig. 6. Critical buckling R versus the side-to-thickness ratio a/h of the (1–2–1) EGM Fig. 8. Critical buckling R versus the aspect ratio b/a of simply-supported and
sandwich square plate resting on Winkler’s elastic foundation with various clamped sandwich plate for different values of foundation stiffnesses J1 and J2
boundary conditions (J1 = 100, k = 0.5, f = 1). (a/h = 10, k = 0.5, f = 1).

Pasternak’s foundation parameters gives results more than those
with the inclusion of Winkler’s foundation parameter. Regardless
of the types of the sandwich plates and elastic foundations, the re-
sults increase as the side-to-thickness ratio increases. Also, the re-
sults predicted by various shear deformation theories approach
each other as the ratio a/h increases as shown in Tables 7 and 8.
The vibration frequencies and critical buckling load of the
(1–1–1) EGM sandwich plate resting on two-parameter elastic
foundations under various boundary conditions using the different
shear deformation theories are depicted for different values of the
side-to-thickness ratio a/h and inhomogeneity parameter k as
shown in Tables 9 and 10. With the increase of the parameter k, a
decrement for the frequencies and critical buckling load can be
clearly observed. The results are the maximum for the free–
clamped plates and the minimum for the simply-supported plates.
Figs. 3 and 4 display the variations of the eigenfrequencies x⁄
and the critical buckling loads R⁄, respectively, versus the side-
to-thickness ratio a/h for different values of the inhomogeneity
parameter k. It can be seen that the frequencies x⁄ increase mono-
tonically as k increases while the change of the buckling loads R⁄ is
reversed. It is also observed that the differences between curves
are reduced as the core thickness increases. The effect of the
side-to-thickness ratio a/h on the buckling load of EGM sandwich
plate is reduced or may be canceled out for large values of it as
shown in Fig. 4.
The eigenfrequencies x⁄ and the buckling loads R⁄ of the
(1–2–1) EGM sandwich square plate resting on Winkler’s elastic
foundation with various boundary conditions are illustrated in
Figs. 5 and 6, respectively. It is noted that x⁄ and R⁄ increase
gradually as the side-to-thickness ratio a/h increases. The results
86 M. Sobhy / Composite Structures 99 (2013) 76–87

(1-1-1) 3. The results show that the presence of elastic foundations

15 leads to a significant increment in the variation of the
frequencies and buckling loads.
14
4. The vibration frequencies and buckling loads increase as the
13
side-to-thickness ratio increases and the aspect ratio decreases.
ζ = 0.0
5. The vibration frequencies and buckling loads for simply-
ζ = 0.5
12 supported EGM sandwich plates are lower than those for
ζ = 1.0
free and clamped EGM sandwich plates.
11 6. The critical buckling load for the plate under uniaxial com-
R*

pression is greater than the plate under biaxial compression.

10

9 Appendix A

8 e8 ÞA þ x2 e6 .00 ;
U 1rs ¼ ðe12 þ m
V 1rs ¼ ðm þ m
Þe8 A;
7
W 1rs ¼ ðe8 þ e12 ÞB  x2 e6 .01 ;
6 W1rs ¼ ðe12 þ me8 ÞC þ x2 e6 .10 ;
5 7 9 11 13 15 17 19 21
a/h U1rs ¼ ðm þ mÞe8 C;
⁄
Fig. 9. Critical buckling R versus the side-to-thickness ratio a/h of simply-
supported sandwich square plate for different values of f (J1 = J2 = 10, k = 0.5). U 2rs ¼ ðm þ m
Þe10 A;
e10 ÞA þ x2 e2 .00 ;
V 2rs ¼ ðe4 þ m
of the simply-supported sandwich plate is less than that of
the clamped–clamped and free–clamped sandwich plate. For the W 2rs ¼ ðe4 þ e10 ÞB  x2 e2 .01 ;
EGM sandwich plate with intermediate boundary conditions, the
W2rs ¼ ðm þ mÞe10 C;
results take the corresponding intermediate values.
Figs. 7 and 8 illustrate the variations of the natural frequencies U2rs ¼ ðe4 þ me10 ÞC þ x2 e2 .10 ;
and the buckling loads as functions of the aspect ratio b/a of the
SSSS and CCCC plate for various values of the elastic foundation U 3rs ¼ ðe13 þ e11 ÞB þ x2 e9 .01 ;
parameters. As it is well known, the clamped boundary condition V 3rs ¼ ðe11 þ e5 ÞB þ x2 e3 .01 ;
always overpredicts the vibration frequencies and buckling loads
magnitude. It can be noticed that the effect of the ratio b/a on W 3rs ¼ ðe5 þ 2e11 þ e13 ÞD  ðRxx  G þ x2 .02 Þe9
the frequencies and the buckling loads is not the same. The fre-  ðRyy  G þ x2 .02 Þe3  ðK  x2 .00 Þe1  Rxy e7 ;
quencies decrease directly as b/a increases, as shown in Fig. 7, W3rs ¼ ðe11 þ e13 ÞF  x2 e9 .11 ;
while the buckling loads decrease rabidly and then increase very
slowly as b/a increases (see Fig. 8). Obviously, both the frequencies U3rs ¼ ðe5 þ e11 ÞF  x2 e3 .11 ;
and the buckling loads are increasing with the increasing of the
foundation stiffnesses. e8 ÞC þ x2 e6 .10 ;
U 4rs ¼ ðe12 þ m
Finally, the influence of the parameter f on the critical buckling V 4rs ¼ ðm þ m
Þe8 C;
loads R⁄ is demonstrated in Fig. 9. As expected, the uniaxial buck-
W 4rs ¼ ðe8 þ e12 ÞF  x2 e6 .11 ;
ling load (f = 0) is greater than the biaxial one (f > 0) and that de-
creases as the parameter f increases. W4rs ¼ ðe12 þ me8 ÞH  ðI  x2 .20 Þe6 ;
U4rs ¼ ðm þ mÞe8 H;
7. Conclusions
U 5rs ¼ ðm þ m
Þe10 C;
In this work, investigations on the vibration frequencies and
e10 ÞC þ x2 e2 .10 ;
V 5rs ¼ ðe4 þ m
critical buckling loads of various types of EGM sandwich plates
with different cases of boundary conditions are presented using W 5rs ¼ ðe4 þ e10 ÞF  x2 e2 .11 ;
the sinusoidal shear deformation plate theory. The sandwich plates W5rs ¼ ðm þ mÞe10 H;
are assumed to be leaned on two-parameter elastic foundations.
The inhomogeneous plates are considered as EGM sandwich plates. U5rs ¼ ðe4 þ me10 ÞH  ðI  x2 .20 Þe2 ;
The upper layer and lower one are made of metal/ceramic EGMs, in which
which are graded according to a new exponential law distribution,
whereas the core is still ceramic. The results obtained by the SPT Z b Z a
are compared with present shear deformation theories and other ðe6 ; e8 ; e12 Þ ¼ X 0r Y s ; X 0r Y 00s ; X 000 0
r Y s X r Y s dx dy;
being in literature. The following conclusions may be drawn from 0 0
Z b Z a
the present analysis:
ðe2 ; e4 ; e10 Þ ¼ X r Y 0s ; X r Y 000 00 0 0
s ; X r Y s X r Y s dx dy;
0 0
1. The present results are very agreement with those being in Z b Z a  
0000

literature. ðe1 ; e3 ; e5 Þ ¼ X r Y s ; X r Y 00s ; X r Y s X r Y s dx dy;

0 0
2. The vibration frequencies and buckling loads for EGM sand- Z b Z a  
0000
wich plates are generally lower than the corresponding val- ðe7 ; e9 ; e11 ; e13 Þ ¼ X 0r Y 0s ; X 00r Y s ; X 00r Y 00s ; X r Y s X r Y s dx dy:
ues for homogeneous ceramic plates. 0 0
 M. Sobhy / Composite Structures 99 (2013) 76–87 87

References [17] Reissner E. On the theory of bending of elastic plates. J Math Phys 1944;23:
184–91.
[18] Reddy JN. A simple higher-order theory for laminated composite plates. J Appl
[1] Zenkour AM. A comprehensive analysis of functionally graded sandwich
Mech 1984;51(4):745–52.
plates: Part 1 – Deflection and stresses and Part 2 – Buckling and free vibration.
[19] Touratier M. An efficient standard plate theory. Int J Eng Sci
Int J Solids Struct 2005;42:5224–58.
1991;29(8):901–16.
[2] Zenkour AM, Sobhy M. Thermal buckling of various types of FGM sandwich
[20] Zenkour AM. Analytical solution for bending of cross-ply laminated plates
plates. Compos Struct 2010;93:93–102.
under thermo-mechanical loading. Compos Struct 2004;65:367–79.
[3] Zenkour AM, Allam MNM, Sobhy M. Bending analysis of FG visco-elastic
[21] Soldatos KP. A transverse shear deformation theory for homogeneous
sandwich beams with elastic cores resting on Pasternak’s elastic foundations.
monoclinic plates. Acta Mech 1992;94:195–200.
Acta Mech 2010;212:233–52.
[22] Karama M, Afaq KS, Mistou S. Mechanical behaviour of laminated composite
[4] Etemadi E, Khatibi AA, Takaffoli M. 3D finite element simulation of sandwich
beam by new multi-layered laminated composite structures model with
panels with a functionally graded core subjected to low velocity impact.
transverse shear stress continuity. Int J Solids Struct 2003;40:1525–46.
Compos Struct 2009;89:28–34.
[23] Zhao X, Lee YY, Liew KM. Mechanical and thermal buckling analysis of
[5] Anderson TA. A 3-D elasticity solution for a sandwich composite with
functionally graded plates. Compos Struct 2009;90:161–71.
functionally graded core subjected to transverse loading by a rigid sphere.
[24] Reddy JN. Analysis of functionally graded plates. Int J Numer Meth Eng 2000;
Compos Struct 2003;60:265–74.
47:663–84.
[6] Shodja HM, Haftbaradaran H, Asghari M. A thermoelasticity solution of
[25] Zenkour AM, Allam MNM, Sobhy M. Effect of transverse normal and shear
sandwich structures with functionally graded coating. Compos Sci Technol
deformations on a fiber-reinforced viscoelastic beam resting on two-
2007;67:1073–80.
parameter elastic foundations. Int J Appl Mech 2010;2:87–115.
[7] Bhangale RK, Ganesan N. Thermoelastic buckling and vibration behavior of
[26] Zenkour AM, Allam MNM, Sobhy M. Bending of a fiber-reinforced viscoelastic
functionally graded sandwich beam with constrained viscoelastic core. J Sound
composite plate resting on elastic foundations. Arch Appl Mech 2011;81:
Vib 2006;295:294–316.
77–96.
[8] Gupta US, Ansari AH, Sharma S. Buckling and vibration of polar orthotropic
[27] Zenkour AM, Sobhy M. Thermal buckling of functionally graded plates resting
circular plate resting on Winkler foundation. J Sound Vib 2006;297:457–76.
on elastic foundations using the trigonometric theory. J Therm Stresses
[9] Saha KN, Kart RC, Dattal PK. Dynamic stability of a rectangular plate on non-
2011;34:1119–38.
homogeneous Winkler foundation. Comput Struct 1997;63(6):1213–22.
[28] Timarci T, Soldatos KP. Comparative dynamic studies for symmetric cross-ply
[10] Lee HP. Dynamic response of a Timoshenko beam on a Winkler foundation
circular cylindrical shells on the basis of a unified shear deformable shell
subjected to a moving mass. Appl Acoust 1998;55(3):203–15.
theory. J Sound Vib 1995;187(4):609–24.
[11] Malekzadeh P, Karami G. A mixed differential quadrature and finite element
[29] Akavci SS, Tanrikulu AH. Buckling and free vibration analysis of laminated
free vibration and buckling analysis of thick beams on two-parameter elastic
composite plates by using two new hyperbolic shear-deformation theories.
foundations. Appl Math Model 2008;32:1381–94.
Mech Compos Mater 2008;44(2):145–54.
[12] Aiello MA, Ombres L. Buckling and vibrations of unsymmetric laminates
[30] Aydogdu M, Taskin V. Free vibration analysis of functionally graded beams
resting on elastic foundations under in-plane and shear forces. Compos Struct
with simply supported edges. Mater Des 2007;28:1651–6.
1999;44:31–41.
[31] Sankar BV. An elasticity solution for functionally graded beams. Compos Sci
[13] Lal A, Singh BN, Kumar R. Nonlinear free vibration of laminated composite
Technol 2001;61:689–96.
plates on elastic foundation with random system properties. Int J Mech Sci
[32] Xiang Y, Wang CM, Kitipornchai S. Exact vibration solution for initially stressed
2008;50:1203–12.
Mindlin plates on Pasternak foundation. Int J Mech Sci 1994;36:311–6.
[14] Malekzadeh P, Karami G. Vibration of non-uniform thick plates on elastic
[33] Leissa AW. The free vibration of plates. J Sound Vib 1973;31:257–93.
foundation by differential quadrature method. Eng Struct 2004;26:1473–82.
[34] Lam KY, Wang CM, He XQ. Canonical exact solutions for Levy-plates on two-
[15] Hashemi ShH, Karimi M, Taher HRD. Vibration analysis of rectangular Mindlin
parameter foundation using Green’s functions. Eng Struct 2000;22:364–78.
plates on elastic foundations and vertically in contact with stationary fluid by
[35] Akhavan H, Hashemi ShH, Taher HRD, Alibeigloo A, Vahabi Sh. Exact solutions
the Ritz method. Ocean Eng 2010;37:174–85.
for rectangular Mindlin plates under in-plane loads resting on Pasternak
[16] Chen WQ, Lü CF, Bian ZG. A mixed method for bending and free vibration of
elastic foundation. Part I: Buckling analysis. Comput Mater Sci 2009;
beams resting on a Pasternak elastic foundation. Appl Math Model 2004;28:
44:968–78.
877–90.