Accepted Manuscript

A unified nonlinear analytical solution of bending, buckling and vibration for

the temperature-dependent FG rectangular plates subjected to thermal load

Y.H. Dong, Y.H. Li

PII: S0263-8223(16)30877-7
DOI: http://dx.doi.org/10.1016/j.compstruct.2016.10.001
Reference: COST 7826

To appear in: Composite Structures

Received Date: 11 June 2016

Revised Date: 15 September 2016
Accepted Date: 3 October 2016

Please cite this article as: Dong, Y.H., Li, Y.H., A unified nonlinear analytical solution of bending, buckling and
vibration for the temperature-dependent FG rectangular plates subjected to thermal load, Composite Structures
(2016), doi: http://dx.doi.org/10.1016/j.compstruct.2016.10.001

This is a PDF file of an unedited manuscript that has been accepted for publication. As a service to our customers
we are providing this early version of the manuscript. The manuscript will undergo copyediting, typesetting, and
review of the resulting proof before it is published in its final form. Please note that during the production process
errors may be discovered which could affect the content, and all legal disclaimers that apply to the journal pertain.
 A unified nonlinear analytical solution of bending,
buckling and vibration for the
temperature-dependent FG rectangular plates
subjected to thermal load
Y.H.Dong, Y.H.Li∗
School of Mechanics and Engineering, Southwest Jiaotong University, Chengdu 610031, PR China

Abstract
This paper presents a unified nonlinear analytical solution of bending, buckling and vibra-
tion for the temperature-dependent functionally graded (FG) rectangular plates subjected
to thermal load. Geometric nonlinearity resulted from mid-plane stretching is considered.
Material properties of FG plates are assumed to vary with temperatures and the volume
fractions of the constituents. Newly proposed higher order shear deformation theories in
present available literature are sorted out and given a unified application. Three types
of mathematical models, P-FG, S-FG and E-FG models, describing effective material
properties of functionally graded materials (FGMs) are discussed. Finally, based on the
unified nonlinear analytical solution, influences of material heterogeneity, thermal load,
and plate geometry on bending, buckling and vibration of FG plates are studied. Out-
comes reveal that the nonlinear solution exhibits better accuracy in calculation of shear
stress in bending responses. The thermal load plays an important role in determining the
bending, buckling and vibration of the FG plate. In addition, the characteristics of these
three types of mathematical models to simulate effective material properties of FGMs are
numerically compared and discussed.
Keywords: Unified nonlinear analytical solution, FG rectangular plates, Thermal load,
Geometric nonlinearity

1. Introduction
The concept of FGMs is proposed by Japanese scientists [1] in the early 1980s. FGMs
are mixed materials, and the volume fractions of the constituents vary continuously in the
gradient direction. They have become increasingly popular in recent decades due to their
favorable mechanical properties. As a type of heat-shielding advanced structural materi-
als, FGMs find their applications in aerospace, nuclear fusions and biomaterial electronics,
etc. For example, thermal barrier plate structures for high temperature applications are

∗
Corresponding author. Tel.: +86 028 87600793;
E-mail addresses: yhli2007@sina.com(Y.H.Li).
Preprint submitted to Composite Structures October 4, 2016
mixtures of the ceramic and the metal. In this paper, it is found that the thermal load
acting on main boundaries of the FG plate has pronounced effect on its bending, buckling
and vibration behaviors.
Many studies for bending, buckling and free vibration of FG plates are available in
the literature. Zenkour [2] adopted the generalized shear deformation theory to present
displacements and stresses of a simply-supported FG rectangular plate subjected to a
transverse unified load. Wu and Li [3] developed a Reissner mixed variational theorem
(RMVT)-based third order shear deformation theory (TSDT). They discussed static anal-
ysis of simply-supported, multilayered FG plates under mechanical loads. Thai and Kim
[4] developed a new higher order shear deformation theory for bending and free vibration
analysis of FG plates. The theory has only four unknowns, but it accounts for a parabolic
variation of transverse shear strains through the thickness of plates. Nguyen [5] dis-
cussed the effects of the power law indexes and side to thickness ratios on the deflections,
stresses, critical buckling loads and natural frequencies of FG plates. Kulkarni et al. [6]
analyzed bending and buckling behaviors of FG plates using inverse trigonometric shear
deformation theory (ITSDT). However, the thermal effect and the temperature-dependent
material properties were not considered in Refs.[2–6].
The main applications of FGMs are in high temperature environment. Praveen and
Reddy [7] presented the nonlinear transient thermoelastic analysis of FG ceramic-metal
plates applying finite element method (FEM). Later Reddy [8] conducted theoretical
formulation and finite element models for FG plates based on the TSDT. However, the
variation of material properties caused by temperature distribution were not considered
in the analysis. Qian and Batra [9] studied the transient thermomechanical deformations
of thick FG plates loaded by either a thermal load or a combined thermal and mechanical
load on the top surface. They obtained the stresses and deformations induced by the
simultaneous application of the transient thermal and mechanical loads. Kim [10] studied
the temperature-dependent vibration characteristics of the FG rectangular plates. The
frequency equation was obtained using the Rayleigh-Ritz procedure based on the TSDT.
Soldatos [11] presented a general two-dimensional theory that is suitable for the static
and/or dynamic analysis of a transverse shear deformable plate made up of a homoge-
neous, monoclinic, linearly elastic material. Wu et al. [12, 13] derived, respectively, the
quasi-three-dimensional static and free vibration analysis of simply-supported, multilay-
ered composite and FG plates using the differential reproducing kernel interpolation. The
strong and the weak formulation of these problems are solved on the basis of the RMVT.
Mantar et al. [14] analyzed the bending response of FG plates adopting a new higher
order shear deformation theory. Asemi et al. [15] investigated buckling responses of het-
erogeneous FG plates under biaxial compression, shear, tension-compression, and shear-
compression load conditions. In their work, a full compatible three-dimensional Hermitian
graded finite element formulation was used, which was an extended form of their previ-
ous work (graded finite element and boundary element formulations) [16]. Lashkari and
Rahmani [17] studied the bending behavior of a sandwich beam with FG core employing
the higher order sandwich panel theory.
Some papers have been published recently about bending, buckling and vibration of
other FG structures. For instance, Aydogdu and Taskin [18] investigated the free vi-
bration of simply supported FG beams employing Navier-type analytical solution. Li et
2
al. [19] studied the bending of transversely isotropic FG circular plates subjected to a
transverse load without considering thermal effect. Pradhan and Chakraverty [20] dealt
with the free vibration of FG beams subjected to different sets of boundary conditions.
Eshraghi et al. [21] analyzed the static bending and free vibrations problems of function-
ally graded annular and circular micro-plates under a thermal loading. Results show that
the magnitude of thermal loading has a substantive effect on the mechanical response of
graded annular and circular micro-plates. Nasihatgozar et al. [22] presented the buck-
ling responses of piezoelectric cylindrical composite panels reinforced with carbon-tubes
employing classical laminated plate theory (CLPT).
Higher order shear deformation theories are always used to deal with the structural
behaviors of plate or beam-type structures combining Hamiltons principle and law of
variation of calculus. FGMs are composite materials intentionally designed so that the
theory investigating FGMs can be used to solve the structural behaviors of composite ma-
terial structures. Reddy [23] developed an improved shear deformation theory that gives
parabolic distribution of the transverse shear strains for laminated composite plates. Sa-
hoo and Singh [24] developed a new inverse trigonometric zigzag theory and implemented
it in the static analysis of laminated composite and sandwich plates. Grover et al. [25]
studied the structural behaviors of laminated composite and sandwich plates by present-
ing an analytical solution for various numerical examples. New nonpolynomial shear-
deformation theories (NPSDTs) were developed and tested. Besides, they [26] presented
the static and buckling analysis of laminated composite and sandwich plates using newly
proposed inverse hyperbolic shear deformation theory (IHSDT).
From an overall point of view, the thermal load is not considered in most of the pre-
vious works. In fact, the thermal load is important for FGMs. There are three types
of mathematical models, P-FG, S-FG and E-FG models, to characterize the variation of
effective material properties of FGMs, however, there is no paper published yet in litera-
ture giving a comparison and discussion about these three models, to the best of authors’
knowledge. According to the literature survey, it seems that geometric nonlinearity in
bending responses of FG plate has not been taken into consideration.
In the present work, we develop a unified nonlinear analytical solution of bending,
buckling and vibration for simply-supported FG rectangular plates subjected to thermal
load, from which equations governing bending, buckling and vibration of the FG plate can
be separated. Material properties are assumed to be temperature-dependent. Meanwhile,
it varies continuously along the thickness direction, which is described by three types
of mathematical models with respected to the volume fractions of the constituents. The
temperature distribution in the plate with the Dirichlet boundary is determined by solving
the heat conducting equation. Newly proposed higher order shear deformation theories in
present available literature are classified and given a unified application. The geometric
nonlinearity caused by mid-plane stretching is taken into account. Numerical examples are
performed to validate the present analyses. Numerical comparison and discussion about
three types of mathematical models simulating effective material properties of FGMs are
presented. The effects of material heterogeneity, thermal load, aspect ratio and span to
thickness ratio on bending, buckling and vibration are discussed, as well.

3
2. Displacement field and strains
Fig. 1 shows an FG rectangular plate constituted by a mixture of metal and ceramic
components. The length of the plate is l, width is b and total thickness is h. A local
Cartesian coordinate system x − y − z is fixed on the mid-plane of the plate where x−,
y− and z− axes are taken along the length, width and thickness direction of the plate,
respectively.

u2 Amplification

y u1 b Full metal

h/2
u3
x h/2

l Full ceramic
z

Fig. 1. A schematic figure of the FG plate

We start the analysis directly from the generalized displacement field. This theory
has five generalized displacement variables (u0 , v0 , w0 , θx and θy ), and it is given below
[11, 25–27]

u1 (x, y, z; t) = u0 (x, y; t) − zw0,x + [f (z) + zΓ] θx (x, y; t)

u2 (x, y, z; t) = v0 (x, y; t) − zw0,y + [f (z) + zΓ] θy (x, y; t) (1)
u3 (x, y, z; t) = w0 (x, y; t)

where ui (i = 1, 2, 3) stand for the displacement components at any point in x−, y−

and z− directions, respectively; u0 , v0 and w0 are displacements of the mid-plane; θx
and θy are shear deformations at the mid-plane; The commas in w0,x and w0,y denote
differentiation with respect to the indicated variables, respectively; f (z) and Γ represent
shape functions determining the distribution of the transverse shear strains and stresses
along the thickness.
Referring to corresponding literature, newly proposed higher order shear deformation
theories expressed in terms of shear-strain shape functions are classified into three types,
as shown in Table 1 where the common shear-strain shape functions for each type are
detailed. It is noted that, if we aim to develop new higher order shear deformation
theories, only the replacement of the shear-strain shape functions is needed.
According to the higher order shear deformation theory, the transverse normal strain
(i.e.εzz ) is negligible and thus the vector of the strain tensor εt = [εxx εyy εxy εxz εyz ]t

4
 Table 1 Shear-strain shape functions for shear deformation

Types Theories f (z) Γ

EBT 0 0
Polynomial FSDT z 0
TSDT[23] z − 4z 3 /3h2 0
HSDT[11] h sinh(z/h) − cosh(0.5)
Hyperbolic IHSDT[26] sinh−1 (rz/h) −2r/h(r2 + 4)1/2 , r = 3.0
cosh(z/h) cosh(0.5)
z cosh(0.5) sinh2 (0.5)
MHSDT[28] − sinh e 5h ( + )e 5h
h h 5h2
SFSDT[25] z sec(0.1z/h) − sec(0.05) [1 + 0.05 tan(0.05)]
Trigonometric ITSDT[25] cot−1 (rh/z) −4r/[h(4r2 + 1)], r = 0.46
SSDT[27] h sin(πz/h)/π 0

in terms of the generalized displacement variables can be expressed as

   0     
εxx  εxx   w0,xx   θx,x 
εyy = ε0 − z w0,yy + [f (z) + zΓ] θy,y
   0yy 
εxy εxy 2w0,xy θx,y + θy,x
   
(2)
   
εxz θ
= [f (z),z + Γ] x
εyz θy

where ε0ij (i, j = x, y) are nonlinear strain components induced by the mid-plane stretch-
ing.

z
x ds w0+w0,xdx
β

w0
u0 u0+u0,xdx
dx

Fig. 2. Deformation of a mid-plane differential element

A mid-plane differential element dx of a plate is depicted in Fig. 2, and it becomes ds

after the deformation. According to the definition of strain, mid-plane strain components
in x− direction can be expressed as

ε0xx = (ds/dx) − 1. (3)

5
 From Fig. 2

(dx − u0 ) + (u0 + u0,x dx) (w0 + w0,x dx) − w0

cos β = ; sin β = (4)
ds ds
can be established readily. Combining Eqs.(3) and (4) yields
q
ε0xx = (1 + u0,x )2 + w0,x
2
− 1. (5)

We assume the mid-plane strain component ε0xx is not zero. Because u0,x  1 and
2
w0,x 1, expressed as the form of taylor series, Eq.(5) can be rewritten as the following
form taking a approximate solution.

ε0xx = u0,x + (1/2)w0,x

2
(6)

Similarly, mid-plane strain components in y− and xy− directions have the following
expressions

ε0yy = v0,y + (1/2)w0,y

2
(7)
ε0xy = u0,y + v0,x + w0,x w0,y

3. Temperature field
Kiani and Eslami [29], Ebrahimi and Salari [30] proposed a linear temperature rise
solution (LTR). They assumed temperature field varies linearly along thickness from the
upper surface to the lower surface, which considers no thermal conductivity.
In the present study, the FG plate subjected to a thermal load. Considering ther-
mal conductivity, the temperature field T is a function of z only and in a steady-state,
therefore, the following thermal equilibrium equation should be satisfied [31]

qz,z = 0 (8)

According to the Fourier law of heat conduction, the thermal flux qz in z− direction
related to the temperature is given by

qz = −k(z)T,z (9)

where k(z) is the thermal conductivity coefficient which is assumed to be a constant for
different temperature fields in the present study.
Introducing Eq.(9) into Eq.(8), the one dimensional steady-state heat conduction equa-
tion can be obtained

− [k(z)T,z ],z = 0 (10)

Integrating Eq.(10) twice from the lower limit −h/2, the temperature field can be

6
derived
Z z
T = c1 k (ι)−1 dι + c2
−h/2

where c1 and c2 are integral constants and can be determined by the thermal boundary
conditions.
Considering the Dirichlet boundary (the temperature prescribed) [31] on the upper
and lower surfaces of the FG plate. We have T |z=−h/2 = Tm and T |z=h/2 = Tc , the
integral constants read
Z h/2
c1 = (Tc − Tm ) / k (ι)−1 dι, c2 = Tm
−h/2

and then temperature variation of the FG plate through the thickness can be derived as
following
 
Z z
(Tc − Tm )
∆Tz (z) =  R h/2 k (ι)−1 dι + Tm  − T0 (11)
−h/2
[dι/k (ι)] −h/2

where T0 is the value of temperature under the stress free state.

4. Equations governing motion of the plate

4.1. Constitutive relations
The stresses are given by
 t
σxx σyy σxy σxz σyz = Cεt (12)

where C = [Qij ]5×5 is the material stiffness matrix, as introduced in Appendix A for the
FG plate.
The plate can be stressed by the temperature variation. The stresses due to temper-
ature variation ∆Tz (z) are
 T   
σxx  Q11 Q12 0 1 0  
T α(z, T )
σ = − Q12 Q22 0  0 1 ∆Tz (z) (13)
 yyT  α(z, T )
σxy 0 0 Q66 0 0

where α(z, T ) is thermal expansion coefficient of the FG plate.

4.2. Derivation of plate equations

In order to obtain a unified nonlinear analytical solution for the temperature-dependent
FG plate under thermal load, Hamilton’s principle is employed with the following form
Z t2
(δU − δV − δK) dt = 0 (14)
t1

7
where δU = δUM + δUT is the variation of total strain energy; δUM is the variation of
strain energy due to vibration, and δUT is the variation of strain energy due to temperature
variation [10]; δV is the variation of work done by external loads and δK is the variation
of kinetic energy. Their expressions can be formulated as
Z Z h/2
δUM = [σxx δεxx + σyy δεyy + σxy δεxy + σxz δεxz + σyz δεyz ] dzdxdy (15a)
Ω0 −h/2

Z Z h/2  T T T

δUT = σxx δdxx + 2σxy δdxy + σyy δdyy dzdxdy (15b)
Ω0 −h/2
Z
!

δV = N + q δw0 dxdy (15c)
Ω0
Z Z h/2
δK = ρ (z) (u̇1 δ u̇1 + u̇2 δ u̇2 + u̇3 δ u̇3 ) dzdxdy (15d)
Ω0 −h/2

where Ω0 is the total cross section area of the FG plate; N is the external normal load
in-plane; q is the external transverse load. The dot-superscript convention indicates the
differentiation with respect to the time variable t. dij (i, j = x, y) in Eq.(15b) have the
following expression [10]

dij = u1,i u1,j + u2,i u2,j + u3,i u3,j (16)

In order to derive the FG plate’s governing differential equations expressed by gen-

eralized displacement variables, we present the governing equations with respected to
resultant forces and moments in Appendix B. To this end, the result is

A11 (u0,xx + w0,x w0,xx ) − B11 w0,xxx + (E11 + B11 Γ) θx,xx + A12 (v0,xy + w0,y w0,xy )
− B12 w0,xyy + (E12 + B12 Γ) θy,xy + A44 (u0,yy + v0,xy + w0,x w0,yy + w0,xy w0,y )
− 2B44 w0,xyy + (E44 + B44 Γ) (θx,yy + θy,xy ) + AT11 + AT12 u0,xx
!

! T T
! T T (17a)
θx,xx + AT22 + AT12 u0,yy


!

+ B11 + B12 (−w0,xxx + Γθx,xx ) + E11 + E12
! T T

! T T


+ B22 + B12 (−w0,xyy + Γθx,yy ) + E22 + E12 θx,yy
= L0 ü0 − L1 ẅ0,x + (L1 Γ + F0 ) θ̈x

A22 (v0,yy + w0,y w0,yy ) − B22 w0,yyy + (E22 + B22 Γ) θy,yy + A12 (u0,xy + w0,x w0,xy )
− B12 w0,xxy + (E12 + B12 Γ) θx,xy + A44 (u0,xy + v0,xx + w0,x w0,xy + w0,xx w0,y )
− 2B44 w0,xxy + (E44 + B44 Γ) (θx,xy + θy,xx ) + AT11 + AT12 v0,xx
!

! T T
! T T (17b)
θy,xx + AT22 + AT12 v0,yy


!

+ B11 + B12 (−w0,xxy + Γθy,xx ) + E11 + E12
! T T

! T T


+ B22 + B12 (−w0,yyy + Γθy,yy ) + E22 + E12 θy,yy
= L0 v̈0 − L1 ẅ0,y + (L1 Γ + F0 ) θ̈y

8
   
1 2
w0,xx A11 u0,x + w0,x − B11 w0,xx + (E11 + B11 Γ) θx,x − T11
2
  
1 2
+A12 v0,y + w0,y − B12 w0,yy + (E12 + B12 Γ) θy,y − T12
2
  
1 2
+ w0,yy A22 v0,y + w0,y − B22 w0,yy + (E22 + B22 Γ) θy,y − T22
2
  
1 2
+A12 u0,x + w0,x − B12 w0,xx + (E12 + B12 Γ) θx,x − T12
2
+ 2w0,xy [A44 (u0,y + v0,x + w0,x w0,y ) − 2B44 w0,xy + (E44 + B44 Γ) (θx,y + θy,x )]
2
!

+ B11 u0,xxx + w0,xx + w0,x w0,xxx − D11 w0,xxxx + (F11 + D11 Γ) θx,xxx
2
!

+ B12 v0,xxy + w0,xy + w0,y w0,xxy − D12 w0,xxyy + (F12 + D12 Γ) θy,xxy
2
!

+ B22 v0,yyy + w0,yy + w0,y w0,yyy − D22 w0,yyyy + (F22 + D22 Γ) θy,yyy
2
!

+ B12 u0,xyy + w0,xy + w0,x w0,xyy − D12 w0,xxyy + (F12 + D12 Γ) θx,xyy
2
!

+ 2B44 u0,xyy + v0,xxy + w0,xy + w0,x w0,xyy + w0,xxy w0,y + w0,xx w0,yy − 4D44 w0,xxyy
+ 2 (F44 + D44 Γ) (θx,xyy + θy,xxy ) + N + q + AT11 + AT12 w0,xx
!

! T T

! T T


+ B11 + B12 (u0,xxx + v0,xxy ) + D11 + D12 (−w0,xxxx + Γθx,xxx − w0,xxyy + Γθy,xxy )
! T T

! T T

! T T


+ F11 + F12 (θx,xxx + θy,xxy ) + A22 + A12 w0,yy + B22 + B12 (v0,yyy + u0,xyy )
! T T

! T T


+ D22 + D12 (−w0,yyyy + Γθy,yyy − w0,xxyy + Γθx,xyy ) + F22 + F12 (θy,yyy + θx,xyy )
 
= L0 ẅ0 + L1 ü0,x + L1 v̈0,y − L2 ẅ0,xx − L2 ẅ0,yy + (L2 Γ + F1 ) θ̈x,x + θ̈y,y
(17c)

(E11 + B11 Γ) (u0,xx + w0,x w0,xx ) − (F11 + D11 Γ) w0,xxx + I11 + D11 Γ2 + 2F11 Γ θx,xx
!

+ (E12 + B12 Γ) (v0,xy + w0,y w0,xy ) − (F12 + D12 Γ) w0,xyy + I12 + D12 Γ2 + 2F12 Γ θy,xy
!

+ (E44 + B44 Γ) (u0,yy + v0,xy + w0,xy w0,y + w0,x w0,yy ) − 2 (F44 + D44 Γ) w0,xyy
+ I44 + D44 Γ2 + 2F44 Γ (θx,yy + θy,xy ) − S44 + 2K44 Γ + A44 Γ2 θx
!
!

! T T

! T T

! 2
! T T


+ B11 + B12 Γu0,xx + D11 + D12 Γ θx,xx − Γw0,xxx + E11 + E12 u0,xx
! T T

! T T

! T T


+ F11 + F12 (2Γθx,xx − w0,xxx ) + I11 + I12 θx,xx + B22 + B12 Γu0,yy
! T T

! 2

! T T


+ D22 + D12 Γ θx,yy − Γw0,xyy + E22 + E12 u0,yy
! T T

! T T


+ F22 + F12 (2Γθx,yy − w0,xyy ) + I22 + I12 θx,yy
= (L1 Γ + F0 ) ü0 − (L2 Γ + F1 ) ẅ0,x + L2 Γ2 + 2F1 Γ + R θ̈x
!

(17d)

9
(E22 + B22 Γ) (v0,yy + w0,y w0,yy ) − (F22 + D22 Γ) w0,yyy + I22 + D22 Γ2 + 2F22 Γ θy,yy
!

+ (E12 + B12 Γ) (u0,xy + w0,x w0,xy ) − (F12 + D12 Γ) w0,xxy + I12 + D12 Γ2 + 2F12 Γ θx,xy
!

+ (E44 + B44 Γ) (u0,xy + v0,xx + w0,xx w0,y + w0,x w0,xy ) − 2 (F44 + D44 Γ) w0,xxy
+ I44 + D44 Γ2 + 2F44 Γ (θx,xy + θy,xx ) − S44 + 2K44 Γ + A44 Γ2 θy
!
!

! T T

! T T

! 2
! T T


+ B11 + B12 Γv0,xx + D11 + D12 Γ θy,xx − Γw0,xxy + E11 + E12 v0,xx
! T T

! T T

! T T


+ F11 + F12 (2Γθy,xx − w0,xxy ) + I11 + I12 θy,xx + B22 + B12 Γv0,yy
! T T

! 2
! T T


+ D22 + D12 Γ θy,yy − Γw0,yyy + E22 + E12 v0,yy
! T T

! T T


+ F22 + F12 (2Γθy,yy − w0,yyy ) + I22 + I12 θy,yy
= (L1 Γ + F0 ) v̈0 − (L2 Γ + F1 ) ẅ0,y + L2 Γ2 + 2F1 Γ + R θ̈y
!

(17e)

where
Z h/2
[Aij , Bij , Dij , Eij , Fij , Iij , Kij , Sij ] = Qij
−h/2

z2, f 2 (z), (f (z),z )2 dz

 
1, z, f (z), zf (z), f (z),z , (i, j = 1, 2, 4)
Z h/2
ATij , BijT , T
EijT , FijT , IijT
 
Dij , =− Qij α (z, T ) ∆Tz (z)
−h/2

z2, f 2 (z) dz (i, j = 1, 2)

 
1, z, f (z), zf (z),

5. Effective material properties of FG plates

Material properties of the FG plate are temperature-dependent and a typical property
’G’ of them (Gm for the metal component and Gc for the ceramic component) can be
expressed as a function of temperature field [10, 32]

G (T ) = G0 G−1 /T + 1 + G1 T + G2 T 2 + G3 T 3
!

(18)

where the coefficients Gi (i = 0, −1, 1, 2, 3) are unique to the constituent materials.

FGMs have constituent elements varying in thickness direction from the upper to
lower surface. The effective material properties change smoothly along the predetermined
direction and thus they can be described by continuous mathematical models. Based on
the rule of mixture, Young’s modulus E, mass density ρ, thermal expansion coefficient α
and the thermal conductivity coefficient k can be expressed as a function of z employing
the power law function [33] (P-FG), the sigmoid function [34] (S-FG), or the exponential
function [35, 36] (E-FG). The Poisson’s ratio of the plate is assumed to be a constant
since the effect of Poisson’s ratio on the deformation is much less, comparing with Young’s
modulus[37].
 s
z 1
G(z, T ) = [Gc (T ) − Gm (T )] + + Gm (T ) , −h/2 ≤ z ≤ h/2 (19a)
h 2

10
  s
1 2z

 [Gm (T ) − Gc (T )] 2 1 − h + Gc (T ) , 0 ≤ z ≤ h/2


G(z, T ) =  s (19b)
 1 2z
 [Gc (T ) − Gm (T )]
 1+ + Gm (T ) , −h/2 ≤ z ≤ 0
2 h
h  z i
G(z, T ) = G0 (T ) exp s 0.5 + , −h/2 ≤ z ≤ h/2 (19c)
h
The effective material properties expressed by the P-FG model are given in Eq.(19a),
where s is the power law index that is positive. The variation of material properties
through the thickness is displayed in Fig. 3. We can see that FG plates exhibit more
properties of ceramic for s ≤ 1, however lean to properties of metal for s ≥ 1. This
phenomenon has pronounced influence on bending, buckling and vibration of FG plates
as shown in the following discussions.

0.5
s=0.1

0.25 s=0.3
s=0.5
z/h

0 s=1.0
s=3.0
-0.25 s=5.0
s=10
-0.5
Gm Gc
Material properties

Fig. 3. Material properties for P-FG model

The sigmoid function is given in Eq.(19b) which defines the volume fraction using
two power-law functions that ensure smooth distribution of stresses among all the inter-
faces. Fig. 4(a) presents variation of material properties with different z coordinates. As
shown, when index s < 1, the change rate of material properties for intermediate section
perpendicular to z− axes is less than the change rate when index s > 1.
Many researchers have used the exponential function to describe the material prop-
erties of FGMs, and the exponential function about coordinate z is given in Eq.(19c)
where G0 represents material properties of the lower surface; s is the exponential index.
The mathematic model describing FGMs properties with the exponential function is only
associated with properties of the bottom surface. Variation of material properties for the
E-FG model is shown in Fig. 4(b). With the increase of index s, the E-FG model exhibits
more properties of ceramic, which results in the increase of stiffness of the FG plate. It
is noted that variation of the effective material property is linear as index s = 1, for all
those three FG models.

11
 0.5 0.5
s=0.1 s=0.1
s=0.3 s=0.3
s=0.5 s=0.5
0.25 s=1.0 0.25 s=1.0
s=3.0 s=3.0
s=5.0 s=5.0
z/h

z/h
0 s=10 0 s=10

-0.25 -0.25

-0.5 -0.5
Gm Gc Gm
(a) Material properties (b) Material properties

Fig. 4. Material properties for S-FG model (a) and E-FG model (b)

6. Analytical solution
In this section, Navier-type analytical solution is adopted to obtain the unified nonlin-
ear analytical solution of bending, buckling and vibration for a FG plate. The FG plate
considered is simply-supported at all four edges. The generalized displacement variables
u0 , v0 , w0 , θx and θy are expressed as products of undetermined coefficients and known
trigonometric functions, which satisfy the boundary conditions. The results are
   

 u0  
 rmn cos λx x sin λy y  
 v0   smn sin λx x cos λy y 
   
∞ X ∞ 
√
  X 
w0 = amn sin λx x sin λy y eiωt , i = −1 (20)
θ x cos λ x sin λ y
   

 x

 n=1 m=1 
 mn x y 



  
θy ymn sin λx x cos λy y
  

where rmn , smn , amn , xmn and ymn are generalized coefficients to be determined, and
mπ nπ
λx = , λy = .
l b
As mentioned in Eq.(15c), external in-plane load and transverse load are assumed in
the following forms[2, 6]
0 0
N = Nxx w0,xx + Nyy w0,yy (21)

N X
X M
q= qmn sin λx x sin λy y (22)
n=1 m=1
0 0
where Nxx and Nyy are the external normal loads in the mid-plane along x− and y−
directions. For convenience, we assume Ncr is the critical buckling load that makes the
0 0
relations Nxx = ηx Ncr and Nyy = ηy Ncr existing. When (ηx , ηy ) = (−1, 0), it is a case of
uniaxial compression. (ηx , ηy ) = (−1, −1) represents a case of equal biaxial compression.
12
 Any loads can be expanded as the double-Fourier series. For sinusoidally distributed
load (SDL), transverse load q can be rewritten as[2, 23]

q = q0 sin λx x sin λy y (23)

we have m = n = 1, and q11 = q0 , where q0 represents the intensity of the load at the
plate center.
Substituting Eqs.(20), (21) and (22) into Eq.(17), the nonlinear governing equation
about generalized displacement variables can be derived in a matrix form by using char-
acteristics of orthogonal family.
(3)
Υ + λN0 Υ(1) − ω 2 m Λ − ζΛ3 = q
 
(24)

where
(3) t
ymn ]t ; Λ3 = 0 0 a3mn

Λ = [rmn smn amn xmn 0 0 ;
q = [0 0 qmn 0 0]t ;

Stiffness matrix Υ and mass matrix m are symmetric, and Υ(1) is a 5 × 5 matrix. Their
nonzero upper triangular elements are given in Appendix C. λ is the buckling load factor,
and N0 is the unit in-plane load satisfying Ncr = λN0 . The coefficient ζ of the nonlinear
item is
ζ = A11 × sscc(λx ) × ssss(λy ) × λ4x + A12 × ssss(λx ) × sscc(λy ) × λ2x λ2y +


A11 × ssss(λx ) × sscc(λy ) × λ4y + A12 × sscc(λx ) × ssss(λy ) × λ2x λ2y −

4 × A44 × sscc(λx ) × sscc(λy ) × λ2x λ2y / [2 × ss(λx ) × ss(λy )]


where the expressions of functions sscc, ssss and ss are given in Appendix C. It is noted
that the temperature-dependent items in Υ are also symmetric.
Eq.(24) is a unified nonlinear equation governing bending, buckling and vibration
analyses of the FG plate subjected to external transverse, in-plane and thermal loads.

6.1. Bending
Setting the buckling load factor λ = 0 and omitting the mass matrix m, we can
separate the nonlinear equation governing bending response from Eq.(24). The result is
(3)
ΥΛ − ζΛ3 = q (25)

Solving this equation for the generalized coefficients rmn , smn , amn , xmn and ymn , the
displacement and stress fields of the FG plate can be obtained. Next, if we set the
coefficient ζ = 0, the linear solution of bending response can be derived, which is consistent
with the equation given in Refs. [6, 14] when thermal effect is not presented. The nonlinear
solution and the linear solution are numerically compared in the next section.

13
6.2. Buckling
Moreover, the linear stability problem can be carried out from Eq.(24) by setting the
coefficient ζ = 0 and omitting the mass matrix m and transverse load q. we obtain

Υ + λN0 Υ(1) Λ = 0
 
(26)

Eq.(26) is an eigenvalue problem about the buckling load factor. Uniaxial and biaxial
buckling problems can be studied on the base of this equation.
6.3. Vibration
If the coefficient ζ and the external transverse load q are neglected in Eq.(24), we can
study linear free vibration characteristics of the FG plate subject to thermal load and
external in-plane load. The governing equation is

Υ + Nex Υ(1) − ω 2 m Λ = 0
 
(27)
0 0
where Nex represents the external in-plane load which satisfies Nxx = ηx Nex and Nyy =
ηy Nex . The external in-plane load Nex can be regarded as a prestress. Solving this
characteristic problem, the natural frequencies of the FG plate can be derived readily.

7. Numerical results
In this section, numerical examples for bending, buckling and vibration of a FG rectan-
gular plate subjected to thermal load are performed applying the present unified nonlinear
analytical solution. The bending, buckling and vibration responses with different material
heterogeneities (s), thermal loads (Tm , Tc ), aspect ratios (b/l) and span to thickness ratios
(l/h) are investigated, respectively.
To unify the discussion, we take T0 as a reference value for temperature in the present
study.
7.1. Validation
In this subsection, numerical examples are given to show the accuracy of present
theory by omitting the temperature variation. To compare with the results of other liter-
ature, we take the P-FG model to describe variation of effective material properties and
take Aluminum(Al) as the metal component, Alumina(Al2 O3 ) as the ceramic component,
and their material properties are given as: Em =70GPa, Ec =380GPa, ρm =2702kg/m3 ,
ρc =3800kg/m3 , Poisson’s ratios ν = 0.3 for both materials. Without losing any gener-
ality, three representative shear-strain shape functions in each type of higher order shear
deformation theory are used. The following dimensionless variables of displacements,
stresses and frequencies are introduced

100Ec h3 10Ec h3
u (z) = u (0, 0.5l, z) ; w = w (0.5l, 0.5b) ;
q0 l 4 q0 l 4
h h
σ xx (z) = σxx (0.5l, 0.5b, z) ; σ xy (z) = σxy (0, 0, z) ;
q0 l q0 l
r
h ρc
σ xz (z) = σxy (0, 0.5b, z) ; ω = ωh ;
q0 l Ec
14
 With the help of Eq.(25), dimensionless displacements and stress components obtained
by linear solution and nonlinear solution for the square FG plate are shown in Table 2.
The FG plate considered is thick (l/h = 10) and subjected to SDL. The power law index
considered is 1. Present results calculated by three types of higher order shear deformation
theories are compared with results of the Quasi-3D theory [38], RMVT-based meshless
collocation solution [13] and discrete layer method based on a VK model [13]. It is
observed that the present results are in good agreement with the existing results. The
shear stresses σ xz calculated by nonlinear numerical solution more close to the existing
results than the ones computed by linear numerical solution.

Table 2 Dimensionless displacements and stress components for the square FG plate (s = 1,l/h = 10)

b/l Theories σ xz (h/6) σ xx (h/3) σ xy (−h/3) u (−h/4) w

Quasi-3D[38] 0.2510 1.5062 0.6081 0.6436 0.5875
11-nodes[13] 0.2509 1.5062 0.6112 0.6436 0.5876
VK model[13] 0.2510 1.5062 0.6081 0.6435 0.5875
TSDT 0.26076 1.48980 0.61111 0.64137 0.58895
1 linear HSDT 0.26059 1.48984 0.61112 0.64140 0.58895
SFSDT 0.26075 1.48981 0.61111 0.64137 0.58895
TSDT 0.25834 1.47596 0.60543 0.63541 0.58348
nonlinear HSDT 0.25817 1.47599 0.60544 0.63544 0.58348
SFSDT 0.25832 1.47596 0.60543 0.63541 0.58348
Quasi-3D[38] 0.2362 1.1985 0.5666 1.0541 0.8823
11-nodes[13] 0.2360 1.1935 0.5674 1.0548 0.8826
VK model[13] 0.2362 1.1985 0.5666 1.0541 0.8823
TSDT 0.25370 1.17936 0.56694 1.05021 0.88148
4 linear HSDT 0.25326 1.17946 0.56695 1.05029 0.88143
SFSDT 0.25366 1.17937 0.56694 1.05021 0.88147
TSDT 0.24918 1.15835 0.55684 1.03150 0.86578
nonlinear HSDT 0.24875 1.15845 0.55686 1.03159 0.86573
SFSDT 0.24914 1.15836 0.55684 1.03151 0.86577

The buckling analysis is carried out on the base of Eq.(26). For a simply-supported
FG rectangular plate with length l = 0.25m and width b = 0.5m in the x− and y−
directions, respectively. Uniaxial and biaxial critical buckling loads Ncr for different span
to thickness ratios are given in Table 3. It can be observed from Table 3 that present
results produce good results comparing with finite element (FE) results.
Assuming the external in-plane load Nex is nought in Eq.(27). Table 4 shows first two
dimensionless frequencies for the square FG plate with different power law indexes s and
span to thickness ratios l/h. As shown, the present results approximate those predicted
by Quasi 3D [39].

7.2. Numerical examples and discussions

Numerical examples based on the present unified nonlinear analytical solution are
now demonstrated. Without losing any generality, TSDT is applied in numerical results

15
 Table 3 Critical buckling load (×106 N) for the FG plate (s = 1)

l/h
(ηx , ηy ) Theories
5 10 20 30 40 50
FEM[15] 237.9880 32.3325 4.1289 1.2343 0.5215 0.2672
TSDT 239.14502 32.47209 4.14863 1.23427 0.52146 0.26716
(−1, 0)
HSDT 239.14448 32.47209 4.14863 1.23427 0.52146 0.26716
SFSDT 239.14495 32.47209 4.14863 1.23427 0.52146 0.26716
FEM[15] 190.3860 25.8680 3.3071 0.9839 0.4162 0.2131
TSDT 191.31602 25.97767 3.31890 0.98742 0.41717 0.21373
(−1, −1)
HSDT 191.31559 25.97767 3.31890 0.98742 0.41717 0.21373
SFSDT 191.31596 25.97767 3.31890 0.98742 0.41717 0.21373

Table 4 The first two dimensionless frequencies ω for square FG plate

Power law indexes (s)

Mode No. l/h Theories
0 0.5 1.0 4.0 10.0
Quasi 3D[39] 0.9400 0.8233 0.7477 0.5997 0.5460
TSDT 0.92972 0.81100 0.73560 0.59236 0.54117
2
HSDT 0.92966 0.81096 0.73556 0.59242 0.54117
SFSDT 0.92972 0.81101 0.73560 0.59237 0.54117
1
Quasi 3D[39] 0.05777 0.04917 0.04427 0.03811 0.03642
TSDT 0.05769 0.04901 0.04419 0.03807 0.03637
10
HSDT 0.05769 0.04901 0.04419 0.03807 0.03637
SFSDT 0.05769 0.04901 0.04419 0.03807 0.03637
Quasi 3D[39] 1.3777 1.2259 1.1166 0.8731 0.6995
TSDT 1.37768 1.22641 1.12016 0.87461 0.78948
2
HSDT 1.37768 1.22641 1.12016 0.87466 0.78948
SFSDT 1.37768 1.22641 1.12016 0.87461 0.78948
2
Quasi 3D[39] 0.2755 0.2473 0.2290 0.1848 0.1629
TSDT 0.27554 0.24724 0.22902 0.18475 0.16287
10
HSDT 0.27554 0.24724 0.22902 0.18476 0.16287
SFSDT 0.27554 0.24724 0.22902 0.18475 0.16287

16
discussion. An FG plate with temperature-dependent material properties is considered,
which is made of stainless steel (SUS304) on its lower surface and silicon nitride (Si3 N4 )
on its upper surface. The coefficients Gi (i = 0, −1, 1, 2, 3) in Eq.(18) for the temperature
range of 300K ≤ T ≤ 1100K are given in Table 5. Poisson’s ratios ν = 0.3 for both
materials, which are not affected by temperature variation. It is noted that the Young’s
moduli E decrease with the temperature.

Table 5 Temperature-dependent coefficients Gi of material properties for Si3 N4 and SUS304[32]

Mterial properties G0 G−1 G1 G2 G3

E(Pa) 348.43×109 0 -3.070×10−4 2.160×10−7 -8.946×10−11
ρ(kg/m3 ) 2370 0 0 0 0
Si3 N4
α(1/K) 5.8723×10−6 0 9.095×10−4 0 0
k(W/mK) 13.723 0 -1.032×10−3 5.466×10−7 -7.876×10−11
E(Pa) 201.04×109 0 3.079×10−4 -6.534×10−7 0
ρ(kg/m3 ) 8166 0 0 0 0
SUS304
α(1/K) 12.330×10−6 0 8.086×10−4 0 0
k(W/mK) 15.379 0 -1.264×10−3 2.092×10−6 -7.223×10−10

In this subsection, numerical results are presented using dimensionless values. The
following dimensionless variables are introduced

100Em (300) h3 10Em (300) h3 Ncr

û = u; ŵ = w; N̂cr = ;
q0 l 4 q0 l 4 2
10(m )Em (300)
s
h ρc (800)
[σ̂xx , σ̂xy , σ̂xz ] = [σxx , σxy , σxy ] ; ω̂ = ωh ;
q0 l Ec (800)

7.2.1. Bending
The bending responses of the FG plate are studied under the action of transverse SDL.
Fig. 5 shows the variations of the in-plane dimensionless displacement û and transverse
shear stress σ̂xz with the dimensionless coordinate z/h for the P-FG, the S-FG and the
E-FG square plates. Span to thickness ratios of the FG plate considered for above cases
are all l/h = 2. The indexes s considered are 0.1, 0.5, 1.0, 1.5 and 2.0. As shown, the
plane where in-plane displacement û = 0 locates near the mid-plane, and this zero plane
moves towards the metal-rich surface as the index s increases for all these three model
plates. It can be seen that the maximum in-plane dimensionless displacement increases
with s for the P-FG model and S-FG model, while decreases with s for the E-FG model.
That makes sense physically, since the stiffness of plate decreases with the index s for the
P-FG model and S-FG model, while increases with the indexes s for the E-FG model, as
observed in Figs. 3 and 4.
Fig. 6 plots variation of the normal stress σ̂xx with the dimensionless coordinate z/h
for the E-FG square plate. In Fig. 6, the maximum value of normal stresses increases
with index s. The normal stresses are compressive throughout the plate up to a plane
(perpendicular to z− axes) and then they become tensile. Coincidentally, the position of
the plane calculated by Zenkour [2] is z ≈ 0.153, which is consistent with the observation
17
 2.6 0.25
s=0.1
s=0.5
s=1.0 0.2 s=0.1
1.3 s=1.5 s=0.5
s=2.0 s=1.0
0.15 s=1.5
s=2.0

σ̂xz
û

0
0.1

-1.2
0.05

-2.4 0
-0.5 -0.25 0 0.25 0.5 -0.5 -0.25 0 0.25 0.5
z/h z/h
(a) P-FG

2.5 0.3
s=0.1
s=0.5
s=1.0 0.25
1.2 s=1.5
s=2.0 0.2
s=0.1
s=0.5
σ̂xz

0.15
û

0 s=1.0
s=1.5
0.1 s=2.0
-1.1
0.05

-2.3 0
-0.5 -0.25 0 0.25 0.5 -0.5 -0.25 0 0.25 0.5
z/h z/h
(b) S-FG

2.8 0.35
s=0.1
s=0.5 0.3
s=1.0
1.4 s=1.5 0.25
s=2.0
0.2
s=0.1
σ̂xz
û

0 s=0.5
0.15 s=1.0
s=1.5
0.1 s=2.0
-1.4
0.05

-2.7 0
-0.5 -0.25 0 0.25 0.5 -0.5 -0.25 0 0.25 0.5
z/h z/h
(c) E-FG

Fig. 5. Dimensionless in-plane displacements and transverse shear stresses versus dimensionless coordi-
nate z/h for P-FG (a), S-FG (b) and E-FG (c) square plates. (Tm =300K, Tc =800K)
18
in Fig. 6. Fig. 7 plots the dimensionless in-plane displacements and normal stresses as
functions of the dimensionless coordinate z/h. The indexes s considered are 0.1, 0.5, 1.0,
1.5 and 2.5. The nonlinear curves in Figs. 5, 6 and 7 are results of nonlinear properties
varying across the thickness as well as the geometric nonlinearity.

1.4
s=0.1
s=0.5
1
s=1.0
s=1.5
s=2.0
σ̂xx

0.5

-0.5
-0.5 -0.25 0 0.153 0.5
z/h
Fig. 6. Dimensionless normal stresses versus dimensionless coordinate z/h for E-FG plate. (Tm =300K,
Tc =800K)

Up to date, Refs. [2, 3, 8, 10, 14, 15, 20, 21, 29, 30] only adopted the P-FG model.
Ref. [31] only adopted the E-FG model. Refs. [5, 6] only adopted the P-FG and E-FG
models. Ref. [28] gave three type of mathematical models describing effective material
properties of FGMs, but only P-FG model was discussed in the numerical discussion
and a comparison of these three models was failed to give. The variation of volume
fractions of the constituents in each of these three models is characterized by a index s.
Comparing these three models, we find that bending responses in the E-FG model are
more sensitive to the index s than that in the P-FG and S-FG models. This phenomenon
is a mathematic fulfilment of material properties described by Eq.(19a), (19b) and (19c).
Although, material properties of the upper and the lower surfaces of these three models
are given and exactly alike, all the maximum values of bending responses derived by these
three models are different each other. Combining Figs. 5(c), 6 and 7 revels that the E-FG
model fails to simulate bending responses of FG plates, gradually, when the index s > 2.0.
With the help of Figs. 3 and 4, all the observation from Figs. 5, 6 and 7 can help the
optimization of FG plate structures such as maximizing or minimizing the displacements
of the plate.
Fig .8 displays variations of the central dimensionless transverse displacement with
indexes and span to thickness ratios for the P-FG square plate. It is observed that
the central transverse displacement changes significantly with the index s and span to
19
 2.8 2
s=0.1
s=0.5
1.5 s=1.0
1.4
s=1.5
1 s=2.5
s=0.1

σ̂xx
û

0 s=0.5
s=1.0 0.5
s=1.5
-1.4 s=2.5 0

-2.7 -0.5
-0.5 -0.25 0 0.25 0.5 -0.5 -0.25 0 0.25 0.5
z/h z/h
(a) (b)

Fig. 7. Dimensionless in-plane displacement (a) and dimensionless normal stress (b) versus dimensionless
coordinate z/h for E-FG plate. (Tm =300K, Tc =800K)

thickness ratio. Central transverse displacement increases with indexes s, which is due to
the fact that the metal-rich FG plate exhibits larger central transverse displacement than
the ceramic-rich plate does. Central transverse displacements of thicker plates are larger
than the thinner plates.
Fig. 9 shows the variations of dimensionless transverse displacement ŵ(l/2, b/2, 0) and
stresses σ̂xx (l/2, b/2, h/3), σ̂xy (0, 0, −h/3) and σ̂xz (0, b/2, h/6) with temperature field for
the E-FG model. The indexes s considered are 0.1, 0.5, 1.0 and 1.5. It is seen that the
transverse displacement and stress increases with the temperature at the same location.
The increase of index s results in the decrease of transverse displacement for the E-FG
model under a given temperature field, which is different from the observation for the
P-FG model. Consistent with the result in Figs. 5 and 6, the stress increases with the
index s.

7.2.2. Buckling
In this subsection, in-plane loads are applied to investigate buckling responses of the
FG plate. The variation of dimensionless critical buckling loads with the power law index
s for the square P-FG plate under specific temperature field is presented in Fig. 10. It is
seen that critical buckling loads decrease with the power law index s, for the reason that
the metal-rich plate subject smaller compressive in-plane loads than the ceramic-rich plate
does. As expected, the plate under the action of the uniaxial load exhibits larger critical
buckling load than the plate under the biaxial load or other loads(ηx = −1, ηy < −1)
do. The influences of aspect ratio and span to thickness ratio on dimensionless critical
buckling load are shown in Fig. 11. The biaxial load is considered for power law index
s = 1. As shown, the critical buckling loads increase with aspect ratios. Besides, the
aspect ratio in a thin plate has more significant effect on critical buckling loads than it
does in a thick plate. Thick FG plates exhibit larger critical buckling loads than thin FG
20
 1.8

1.4 l/h=1
l/h=2
l/h=4
l/h=8
ŵ
1
l/h=10

0.6

0.2
0 2 4 6 8 10
s
Fig. 8. Dimensionless transverse displacement versus span to thickness ratio for P-FG square plate
(Tm =300K, Tc =800K)

plates do, which can be observed in Table 3.

Figs. 12 and 13 plot the dimensionless critical buckling load as a function of temper-
ature for the P-FG and the E-FG models, respectively. The FG plate considered is under
the action of biaxial loads (ηx , ηy ) = (−1, −1), and its span to thickness ratio l/h=2.
Dimensionless critical buckling loads decrease with temperatures of both surfaces because
material stiffness decrease with temperature. In Fig. 12, critical buckling load of the
plate decrease with index s at the same temperature field, which is consistent with the
observation from Fig. 10. Whereas, critical buckling loads increase with index at the
same temperature field for the E-FG model as shown in Fig. 13. It is evident that the
ceramic-rich plate has larger stiffness value than metal-rich plate does(see Fig. 4(b)).

7.2.3. Free vibration

In this subsection, free vibration analysis is carried out by setting the external in-
plane load Nex = 0. The influences of power law index, aspect ratio, span to thickness
ratio and temperature field on dimensionless frequencies are shown in Figs. 14, 15 and
16, respectively. The first three dimensionless fundamental frequencies of the FG plate
decrease with the power law index, and when the index s exceed 2, the fundamental
frequencies have small change along with the increase of index, as shown in Fig. 14.
Fig. 15 shows that the increase of span to thickness ratio decreases the fundamental
frequencies. In addition, the lower span to thickness ratios can affect the fundamental
frequencies significantly. Whereas with further increase of span to thickness ratios, the
frequencies become less and less sensitive. It can be seen that the change of fundamental
frequencies are less sensitive to the change of aspect ratio. As can be seen from Fig.
21
 1.5 0.47
s=0.1
1.25 s=0.5 0.4

σ̂xx
ŵ 0.95 s=1.0
s=1.5 0.35
0.65 0.3
0.35 0.25
300 600 900 1100 300 600 900 1100
Tm Tm
0.2 0.42

0.35
0.17
σ̂xy

σ̂xz
0.15 0.29

0.13 0.22
300 600 900 1100 300 600 900 1100
Tm Tm

Fig. 9. Dimensionless displacements and stresses versus temperature variation for E-FG square plate
(Tc =800K, l/h=2)

16, there is an obvious decrease about the dimensionless fundamental frequency with the
increase of temperature at the bottom surface.

8. Conclusion
In this paper, a unified nonlinear analytical solution for simply-supported FG rectan-
gular plates subjected to thermal load are derived. Geometric nonlinearity resulted from
mid-plane stretching is considered. Material properties are assumed to be temperature-
dependent, and vary continuously through the thickness according to the P-FG, S-FG
or E-FG models. Newly proposed higher order shear deformation theories in present
available literature are sorted out and given a unified application. The characteristics of
the three types of mathematic models to simulate effective material properties of FGMs
are numerically compared and discussed, which can help the optimization of FG plate
structures such as maximizing or minimizing the displacements of the plate.
The results of this study can be summarized as follows:

(1) The thermal load plays an important role in determining the bending, buckling and
vibration of the FG plate.
(2) Simulating the shear stress of bending response for FG plates, the nonlinear solution
resulted from geometric nonlinearity is more accurate than the linear solution.
(3) Bending responses in the E-FG model are more sensitive to the index s than that in
the P-FG and S-FG models
(4) The maximum in-plane dimensionless displacement increases with index s for the
P-FG and S-FG models, while decreases with index s for the E-FG model.

22
 0.04 (η ,η )=(-1,0)
x y
(η ,η )=(-1,-1)
x y
(ηx,ηy)=(-1,-2)
0.03 (ηx,ηy)=(-1,-3)
ˆcr
N

0.02

0.01

0 2 4 6 8 10
s
Fig. 10. Dimensionless critical buckling load versus index s for P-FG square plate (Tm =300K, Tc =800K,
l/h=10)

(5) The aspect ratio in a thin plate has more significant effect on critical buckling loads
than it does in a thick plate. Thick FG plates exhibit larger critical buckling loads
than thin FG plates do.
(6) The dimensionless fundamental frequencies decrease with the temperature.

Acknowledgments
This work was supported by the National Natural Science Foundation of China (Grant
Nos.11372257, 11302183, 11102030 and 11602208).

23
 0.7
l/h=4
l/h=5
0.5 l/h=8
l/h=10
ˆcr
N

0.3

0.1
0
0.5 1 2 3 4 5
b/l
Fig. 11. Dimensionless critical buckling loads versus aspect and span to thickness ratios for P-FG plate
(Tm =300K, Tc =800K, s=1)

0.057
s=0.1
s=0.5
0.053 s=1.0
s=1.5
ˆcr

0.049
N

0.045

0.041
300 600 900 1100
Tc
Fig. 12. Dimensionless critical buckling loads versus temperature variation for P-FG square plate
(Tm =300K, l/h=2)

24
 0.065
s=0.1
0.055 s=0.5
s=1.0
0.045 s=1.5
ˆcr
N

0.035

0.025

0.017
300 600 900 1100
Tm
Fig. 13. Dimensionless critical buckling loads versus temperature variation for E-FG square plate
(Tc =1100K, l/h=2)

0.5
1st
0.4 2nd
3rd
0.3
ω̂

0.2

0.1

0
0 2 4 6 8 10
s
Fig. 14. Dimensionless frequencies versus power law indexes for P-FG square plate (Tm =300K, Tc =800K,
l/h=10)

25
 0.75
b/l=1
b/l=2
0.55 b/l=3
b/l=4
ω̂

0.35

0.15

0
2 10 20 30 40 50
l/h
Fig. 15. Dimensionless fundamental frequencies versus aspect and span to thickness ratios for P-FG
plate (Tm =300K, Tc =800K, s=2)

0.05 s=0.1
s=0.5
0.045 s=1.0
s=1.5
0.04
ω̂

0.035

0.03

0.024
300 600 900 1100
Tc
Fig. 16. Dimensionless fundamental frequency versus temperature variation for P-FG square plate
(l/h=10)

26
Appendix A.
The matrix of material stiffness for a FG plate is defined as

E(z) νE(z)
 
 1 − ν2 1 − ν2 0 0 0 
νE(z) E(z)
 
0 0 0
 
 1 − ν2 1 − ν2
 

 E(z) 
C= 0 0 0 0 (A.1)
 
 2(1 + ν) 


 0 E(z) 
0 0 0 

 2(1 + ν) 

 E(z) 
0 0 0 0
2(1 + ν)

where the nonzero elements are defined as Qij (i, j = 1, 2, 4, 5, 6).

Appendix B.
This section is devoted to present the governing equations expressed by resultant force
and moment. Substituting Eqs.(1), (2), (6), (7), (12), and (13) into Eqs.(15) and (16),
the variation of energy is calculated by
Z
δUM = [Nxx (δu0,x + w0,x δw0,x ) − Mxx δw0,xx + (Hxx + ΓMxx ) δθx,x
Ω0
+Nyy (δv0,y + w0,y δw0,y ) − Myy δw0,yy + (Hyy + ΓMyy ) δθy,y
(B.1)
+Nxy (δu0,y + δv0,x + w0,y δw0,x + w0,x δw0,y ) −
2Mxy δw0,xy + (Hxy + ΓMxy ) (δθx,y + δθy,x )
+ (Jxz + ΓNxz ) δθx + (Jyz + ΓNyz ) δθy ] dxdy

27
 Z
 T
δUT = Nxx (u0,x δu0,x + v0,x δv0,x + w0,x δw0,x )
Ω0
T
+Mxx (−u0,x δw0,xx − w0,xx δu0,x + Γu0,x δθx,x + Γθx,x u0,x
−v0,x δw0,xy − w0,xy δv0,x + Γv0,x δθy,x + Γθy,x v0,x )
T
w0,xx δw0,xx + Γ2 θx,x δθx,x − Γw0,xx δθx,x − Γθx,x w0,xx
!
+Hxx
+w0,xy δw0,xy + Γ2 θy,x δθy,x − Γw0,xy δθy,x − Γθy,x w0,xy


T
+Jxx (u0,x δθx,x + θx,x δu0,x + v0,x δθy,x + θy,x δv0,x )
T
+Kxx (2Γθx,x δθx,x − w0,xx δθx,x − θx,x δw0,xx
+2Γθy,x δθy,x − w0,xy δθy,x − θy,x δw0,xy )
T
+Pxx (θx,x δθx,x + θy,x δθy,x ) (B.2)
T
+Nyy (u0,y δu0,y + v0,y δv0,y + w0,y δw0,y )
T
+Myy (−v0,y δw0,yy − w0,yy δv0,y + Γv0,y δθy,y + Γθy,y v0,y
−u0,y δw0,xy − w0,xy δu0,y + Γu0,y δθx,y + Γθx,y u0,y )
T
w0,yy δw0,yy + Γ2 θy,y δθy,y − Γw0,yy δθy,y − Γθy,y w0,yy
!
+Hyy
+w0,xy δw0,xy + Γ2 θx,y δθx,y − Γw0,xy δθx,y − Γθx,y w0,xy


T
+Jyy (v0,y δθy,y + θy,y δv0,y + u0,y δθx,y + θx,y δu0,y )
T
+Kyy (2Γθy,y δθy,y − w0,yy δθy,y − θy,y δw0,yy
+2Γθx,y δθx,y − w0,xy δθx,y − θx,y δw0,xy )
T

+ Pyy (θy,y δθy,y + θx,y δθx,y ) dxdy
where N , M , H and J are stress results, moment results and higher order terms defined
as
Z h/2
[Nij , Mij , Hij , Jij ] = σij [1, z, f (z), f (z),z ] dz (i, j = x, y, z)
−h/2

Z h/2
NijT , MijT , HijT , JijT , KijT , PijT σijT
 
=
−h/2
2 2
 
1, z, z , f (z),
zf (z), f (z) dz (i, j = x, y)
Z Z
δV = N δw0 dxdy + qδw0 dxdy (B.3)
Ω0 Ω0

28
 Z
δK = [L0 (u̇0 δ u̇0 + v̇0 δ v̇0 + ẇ0 δ ẇ0 )
Ω0

+L1 −u̇0 δ ẇ0,x − ẇ0,x δ u̇0 + Γu̇0 δ θ̇x + Γθ̇x δ u̇0 −

v̇0 δ ẇ0,y − ẇ0,y δ v̇0 + Γv̇0 δ θ̇y + Γθ̇y δ v̇0

+L2 ẇ0,x δ ẇ0,x − Γẇ0,x δ θ̇x − Γθ̇x δ ẇ0,x + Γ2 θ̇x δ θ̇x +
 (B.4)
ẇ0,y δ ẇ0,y − Γẇ0,y δ θ̇y − Γθ̇y δ ẇ0,y + Γ2 θ̇y δ θ̇y
 
+F0 u̇0 δ θ̇x + θ̇x δ u̇0 + v̇0 δ θ̇y + θ̇y δ v̇0
 
+F1 −ẇ0,x δ θ̇x − θ̇x δ ẇ0,x + 2Γθ̇x δ θ̇x − ẇ0,y δ θ̇y − θ̇y δ ẇ0,y + 2Γθ̇y δ θ̇y
 i
+R θ̇x δ θ̇x + θ̇y δ θ̇y dxdy
where Li , Fj , R are inertia terms defined as
Z h/2
z2, f 2 (z) dz
 
[L0 , L1 , L2 , F0 , F1 , R] = ρ (z) 1, z, f (z), zf (z),
−h/2

Substituting Eqs(B.1), (B.3) and (B.4) into Eq.(14), integrating by parts and collecting
the coefficients of δu0 , δv0 , δw0 , δθx and δθy , the following governing equations of motion
for a FG plate can be obtained:
T T T
Nxx,x + Nxy,y + Nxx u0,xx + Mxx (−w0,xxx + Γθx,xx ) + Jxx θx,xx
T T T
(B.5)
+ Nyy u0,yy + Myy (−w0,xyy + Γθx,yy ) + Jyy θx,yy = L0 ü0 − L1 ẅ0,x + (L1 Γ + F0 ) θ̈x
T T T
Nyy,y + Nxy,x + Nxx v0,xx + Mxx (−w0,xxy + Γθy,xx ) + Jxx θy,xx
T T T
(B.6)
+ Nyy v0,yy + Myy (−w0,yyy + Γθy,yy ) + Jyy θy,yy = L0 v̈0 − L1 ẅ0,y + (L1 Γ + F0 ) θ̈y
Nxx,x w0,x + Nxx w0,xx + Nyy,y w0,y + Nyy w0,yy + Nxy,x w0,y + Nxy,y w0,x
T
+ 2Nxy w0,xy + Mxx,xx + Myy,yy + 2Mxy,xy + N + q + Nxx w0,xx
T T
+ Mxx (u0,xxx + v0,xxy ) + Hxx (−w0,xxxx + Γθx,xxx − w0,xxyy + Γθy,xxy )
T
+ Kxx T
(θx,xxx + θy,xxy ) + Nyy T
w0,yy + Myy (v0,yyy + u0,xyy ) (B.7)
T T
+ Hyy (−w0,yyyy + Γθy,yyy − w0,xxyy + Γθx,xyy ) + Kyy (θy,yyy + θx,xyy )
 
= L0 ẅ0 + L1 ü0,x + L1 v̈0,y − L2 ẅ0,xx − L2 ẅ0,yy + (L2 Γ + F1 ) θ̈x,x + θ̈y,y
T
Hxx,x + ΓMxx,x + Hxy,y + ΓMxy,y − Jxz − ΓNxz + Mxx Γu0,xx
T
! 2
T T T
+ Hxx Γ θx,xx − Γw0,xxx + Jxx u0,xx + Kxx (2Γθx,xx − w0,xxx ) + Pxx θx,xx
T T (B.8)
Γ2 θx,yy − Γw0,xyy + Jyy T T
!

+ Myy Γu0,yy + Hxx u0,yy + Kyy (2Γθx,yy − w0,xyy )
T
θx,yy = (L1 Γ + F0 ) ü0 − (L2 Γ + F1 ) ẅ0,x + L2 Γ2 + 2F1 Γ + R θ̈x
!

+ Pyy

29
 T
Hyy,y + ΓMyy,y + Hxy,x + ΓMxy,x − Jyz − ΓNyz + Mxx Γv0,xx
T
Γ2 θy,xx − Γw0,xxy + Jxx T T T
!

+ Hxx v0,xx + Kxx (2Γθy,xx − w0,xxy ) + Pxx θy,xx
T T (B.9)
Γ2 θy,yy − Γw0,yyy + JyyT T
!

+ Myy Γv0,yy + Hxx v0,yy + Kyy (2Γθy,yy − w0,yyy )
T 2
!

+ Pyy θy,yy = (L1 Γ + F0 ) v̈0 − (L2 Γ + F1 ) ẅ0,y + L2 Γ + 2F1 Γ + R θ̈y

Appendix C.

Υ11 = A11 λ2x + A44 λ2y + AT11 + AT12 λ2x + AT22 + AT12 λ2y ,
!
!

Υ12 = (A12 + A44 ) λx λy ,

Υ13 = −B11 λ3x − (B12 + 2B44 ) λx λ2y − B11
! T T

3 ! T T
λx λ2y ,


+ B12 λx − B22 + B12
Υ14 = (E11 + B11 Γ) λ2x + (E44 + B44 Γ) λ2y + B11
! T T T T

2
Γ + B12 Γ + E11 + E12 λx
! T T T T
λ2y ,


+ B22 Γ + B12 Γ + E22 + E12
Υ15 = (E12 + B12 Γ + E44 + B44 Γ) λx λy ,
Υ22 = A22 λ2y + A44 λ2x + AT11 + AT12 λ2x + AT22 + AT12 λ2y ,
!
!

Υ23 = −B22 λ3y − (B12 + 2B44 ) λ2x λy − B11

! T T

2 ! T T

3
+ B12 λx λy − B22 + B12 λy ,
Υ24 = Υ15 ,
Υ25 = (E22 + B22 Γ) λ2y + (E44 + B44 Γ) λ2x + B11
! T T T T

2
Γ + B12 Γ + E11 + E12 λx
! T T T T
λ2y ,


+ B22 Γ + B12 Γ + E22 + E12
Υ33 = D11 λ4x + D22 λ4y + (2D12 + 4D44 ) λ2x λ2y + AT11 + AT12 λ2x
!

! T T

! 4 (C.1)
λx + λ2x λ2y + AT22 + AT12 λ2y + D22
! T T

! 4
λy + λ2x λ2y ,

!


+ D11 + D12 + D12
Υ34 = − (F11 + D11 Γ) λ3x − (F12 + D12 Γ + 2F44 + 2D44 Γ) λx λ2y
! T T T T

3 ! T T T T
λx λ2y ,


− D11 Γ + D12 Γ + F11 + F12 λx − D22 Γ + D12 Γ + F22 + F12
Υ35 = − (F22 + D22 Γ) λ3y − (F12 + D12 Γ + 2F44 + 2D44 Γ) λ2x λy
! T T T T

2 ! T T T T

3
− D11 Γ + D12 Γ + F11 + F12 λx λy − D22 Γ + D12 Γ + F22 + F12 λy ,
Υ44 = I11 + D11 Γ2 + 2F11 Γ λ2x + I44 + D44 Γ2 + 2F44 Γ λ2y
!
!

+ S44 + 2K44 Γ + A44 Γ2 + D11

!
! T 2 T 2 T T T T

2
Γ + D12 Γ + 2F11 Γ + 2F12 Γ + I11 + I12 λx
! T 2 T 2 T T T T

2
+ D22 Γ + D12 Γ + 2F22 Γ + 2F12 Γ + I22 + I12 λy ,
Υ45 = I12 + D12 Γ2 + 2F12 Γ + I44 + D44 Γ2 + 2F44 Γ λx λy
!

Υ55 = I22 + D22 Γ2 + 2F22 Γ λ2y + I44 + D44 Γ2 + 2F44 Γ λ2x

!
!

+ S44 + 2K44 Γ + A44 Γ2 + D11

!
! T 2 T 2 T T T T

2
Γ + D12 Γ + 2F11 Γ + 2F12 Γ + I11 + I12 λx
! T 2 T 2 T T T T
λ2y .


+ D22 Γ + D12 Γ + 2F22 Γ + 2F12 Γ + I22 + I12
(1)
Υ33 = ηx λ2x + ηy λ2y (C.2)

30
m11 = L0 , m13 = −L1 λx , m14 = L1 Γ + F0 , m22 = m11 , m23 = −L1 λy ,
= L1 Γ + F0 , m33 = L0 + L2 λ2x + λ2y , m34 = − (L2 Γ + F1 ) λx ,
!

m25 (C.3)
= − (L2 Γ + F1 ) λy , m44 = L2 Γ2 + 2F1 Γ + R , m55 = m44 .
!

m35

2 sin(χπ) cos3 (χπ) − cos(χπ) sin(χπ) − χπ

sscc(χ) = −
4χ
3
2 sin (χπ) cos(χπ) + 3 cos(χπ) sin(χπ) − 3χπ
ssss(χ) = − (C.4)
4χ
sin(χπ) cos(χπ) − χπ
ss(χ) = − .
χ

References
[1] M Koizumu. The concept of fgm, ceramic transactions. Functionally Graded Materials, 34:3–10,
1993.
[2] Ashraf M Zenkour. Generalized shear deformation theory for bending analysis of functionally graded
plates. Applied Mathematical Modelling, 30(1):67–84, 2006.
[3] Chih Ping Wu and Hao Yuan Li. An rmvt-based third-order shear deformation theory of multilayered
functionally graded material plates. Composite Structures, 92(10):2591–2605, 2010.
[4] Huu-Tai Thai and Seung-Eock Kim. A simple higher-order shear deformation theory for bending
and free vibration analysis of functionally graded plates. Composite Structures, 96:165–173, 2013.
[5] Trung Kien Nguyen. A higher-order hyperbolic shear deformation plate model for analysis of func-
tionally graded materials. International Journal of Mechanics and Materials in Design, 11(2):203–
219, 2014.
[6] Kamlesh Kulkarni, B. N. Singh, and D. K. Maiti. Analytical solution for bending and buckling anal-
ysis of functionally graded plates using inverse trigonometric shear deformation theory. Composite
Structures, 134:147–157, 2015.
[7] G. N. Praveen and J. N. Reddy. Nonlinear transient thermoelastic analysis of functionally graded
ceramic-metal plates. International Journal of Solids and Structures, 35(33):4457–4476, 1998.
[8] J. N. Reddy. Analysis of functionally graded plates. International Journal for Numerical Methods
in Engineering, 47(47):663–684, 2000.
[9] L. F. Qian and R. C. Batra. Transient thermoelastic deformations of a thick functionally graded
plate. Journal of Thermal Stresses, 27(8):705–740, 2004.
[10] Young-Wann Kim. Temperature dependent vibration analysis of functionally graded rectangular
plates. Journal of Sound and Vibration, 284(3):531–549, 2005.
[11] KP Soldatos. A transverse shear deformation theory for homogeneous monoclinic plates. Acta
Mechanica, 94(3-4):195–220, 1992.
[12] Chih-Ping Wu and Kuan-Hao Chiu. Rmvt-based meshless collocation and element-free galerkin
methods for the quasi-3d free vibration analysis of multilayered composite and fgm plates. Composite
Structures, 93(5):1433–1448, 2011.
[13] Chih-Ping Wu, Kuan-Hao Chiu, and Yung-Ming Wang. Rmvt-based meshless collocation and
element-free galerkin methods for the quasi-3d analysis of multilayered composite and fgm plates.
Composite Structures, 93(2):923–943, 2011.
[14] JL Mantari, AS Oktem, and C Guedes Soares. Bending response of functionally graded plates by
using a new higher order shear deformation theory. Composite Structures, 94(2):714–723, 2012.
[15] K Asemi, M Shariyat, M Salehi, and H Ashrafi. A full compatible three-dimensional elasticity
element for buckling analysis of fgm rectangular plates subjected to various combinations of biaxial
normal and shear loads. Finite Elements in Analysis and Design, 74:9–21, 2013.

31
[16] H Ashrafi, K Asemi, M Shariyat, and M Salehi. Two-dimensional modeling of heterogeneous struc-
tures using graded finite element and boundary element methods. Meccanica, 48(3):663–680, 2013.
[17] MJ Lashkari and O Rahmani. Bending behavior of sandwich structures with flexible functionally
graded core based on high-order sandwich panel theory. Meccanica, pages 1–20, 2015.
[18] Metin Aydogdu and Vedat Taskin. Free vibration analysis of functionally graded beams with simply
supported edges. Materials and Design, 28(28):1651–1656, 2007.
[19] XY Li, HJ Ding, and WQ Chen. Elasticity solutions for a transversely isotropic functionally graded
circular plate subject to an axisymmetric transverse load qr k. International Journal of Solids and
Structures, 45(1):191–210, 2008.
[20] Karan K. Pradhan and S. Chakraverty. Generalized power-law exponent based shear deformation
theory for free vibration of functionally graded beams. Applied Mathematics and Computation, 268
(1):1240–1258, 2015.
[21] Iman Eshraghi, Serkan Dag, and Nasser Soltani. Bending and free vibrations of functionally graded
annular and circular micro-plates under thermal loading. Composite Structures, 137:196–207, 2016.
[22] Mohsen Nasihatgozar, Vahid Daghigh, Milad Eskandari, Kamran Nikbin, and Andy Simoneau.
Buckling analysis of piezoelectric cylindrical composite panels reinforced with carbon nanotubes.
International Journal of Mechanical Sciences, 107:69–79, 2016.
[23] J. N. Reddy. A simple higher-order theory for laminated composite plates. Journal of Applied
Mechanics, 51(4):745–752, 1984.
[24] R. Sahoo and B. N. Singh. A new shear deformation theory for the static analysis of laminated
composite and sandwich plates. International Journal of Mechanical Sciences, 75(4):324–336, 2013.
[25] Neeraj Grover, B. N. Singh, and D. K. Maiti. New nonpolynomial shear-deformation theories for
structural behavior of laminated-composite and sandwich plates. AIAA Journal, 51(8):1861–1871,
2013.
[26] Neeraj Grover, D. K. Maiti, and B. N. Singh. A new inverse hyperbolic shear deformation theory
for static and buckling analysis of laminated composite and sandwich plates. Composite Structures,
95(1):667–675, 2013.
[27] M Touratier. An efficient standard plate theory. International Journal of Engineering Science, 29
(8):901–916, 1991.
[28] Zakaria Belabed, Mohammed Sid Ahmed Houari, Abdelouahed Tounsi, SR Mahmoud, and O Anwar
Bég. An efficient and simple higher order shear and normal deformation theory for functionally
graded material (fgm) plates. Composites Part B: Engineering, 60:274–283, 2014.
[29] Y Kiani and MR Eslami. An exact solution for thermal buckling of annular fgm plates on an elastic
medium. Composites Part B: Engineering, 45(1):101–110, 2013.
[30] Farzad Ebrahimi and Erfan Salari. Nonlocal thermo-mechanical vibration analysis of functionally
graded nanobeams in thermal environment. Acta Astronautica, 113:29–50, 2015.
[31] XY Li, YH Dong, C Liu, Y Liu, CJ Wang, and TF Shi. Axisymmetric thermo-magneto-electro-
elastic field in a heterogeneous circular plate subjected to a uniform thermal load. International
Journal of Mechanical Sciences, 88:71–81, 2014.
[32] JN Reddy and CD Chin. Thermomechanical analysis of functionally graded cylinders and plates.
Journal of Thermal Stresses, 21(6):593–626, 1998.
[33] Nuttawit Wattanasakulpong, B Gangadhara Prusty, Donald W Kelly, and Mark Hoffman. Free
vibration analysis of layered functionally graded beams with experimental validation. Materials and
Design, 36:182–190, 2012.
[34] YL Chung and SH Chi. The residual stress of functionally graded materials. J Chin Inst Civil
Hydraulic Eng, 13:1–9, 2001.
[35] AM Zenkour. Benchmark trigonometric and 3-d elasticity solutions for an exponentially graded
thick rectangular plate. Archive of Applied Mechanics, 77(4):197–214, 2007.
[36] X-Y Li, H-J Ding, W-Q Chen, and P-D Li. Three-dimensional piezoelectricity solutions for uniformly
loaded circular plates of functionally graded piezoelectric materials with transverse isotropy. Journal
of Applied Mechanics, 80(4):041007, 2013.
[37] F Delale and F Erdogan. The crack problem for a nonhomogeneous plane. Journal of Applied
Mechanics, 50(3):609–614, 1983.
[38] Erasmo Carrera, Salvatore Brischetto, and Alessandro Robaldo. Variable kinematic model for the

32
 analysis of functionally graded material plates. AIAA journal, 46(1):194–203, 2008.
[39] Hiroyuki Matsunaga. Free vibration and stability of functionally graded plates according to a 2-d
higher-order deformation theory. Composite structures, 82(4):499–512, 2008.

33