Applied Mathematical Modelling 34 (2010) 39914011

Contents lists available at ScienceDirect

Applied Mathematical Modelling

journal homepage: www.elsevier.com/locate/apm

Static response and free vibration analysis of FGM plates using higher order shear deformation theory
Mohammad Talha, B.N. Singh *
Department of Aerospace Engineering, Indian Institute of Technology, Kharagpur 721 302, India

a r t i c l e

i n f o

a b s t r a c t
Free vibration and static analysis of functionally graded material (FGM) plates are studied using higher order shear deformation theory with a special modication in the transverse displacement in conjunction with nite element models. The mechanical properties of the plate are assumed to vary continuously in the thickness direction by a simple power-law distribution in terms of the volume fractions of the constituents. The fundamental equations for FGM plates are derived using variational approach by considering traction free boundary conditions on the top and bottom faces of the plate. Results have been obtained by employing a continuous isoparametric Lagrangian nite element with 13 degrees of freedom per node. Convergence tests and comparison studies have been carried out to demonstrate the efciency of the present model. Numerical results for different thickness ratios, aspect ratios and volume fraction index with different boundary conditions have been presented. It is observed that the natural frequency parameter increases for plate aspect ratio, lower volume fraction index n and smaller thickness ratios. It is also observed that the effect of thickness ratio on the frequency of a plate is independent of the volume fraction index. For a given thickness ratio non-dimensional deection increases as the volume fraction index increases. It is concluded that the gradient in the material properties plays a vital role in determining the response of the FGM plates. 2010 Elsevier Inc. All rights reserved.

Article history: Received 14 July 2009 Received in revised form 24 February 2010 Accepted 30 March 2010 Available online 18 April 2010 Keywords: Functionally graded material Higher order shear deformation theory Finite element method Independent eld variables

1. Introduction Composite materials have been successfully used in aircraft and other engineering applications for many years because of their excellent strength to weight and stiffness to weight ratios. Recently, advanced composite materials known as functionally graded material have attracted much attention in many engineering applications due to their advantages of being able to resist high temperature gradient while maintaining structural integrity [1]. The functionally graded materials (FGMs) are microscopically inhomogeneous, in which the mechanical properties vary smoothly and continuously from one surface to the other. They are usually made from a mixture of ceramics and metals to attain the signicant requirement of material properties. Due to the increased relevance of the FGMs structural components in the design of aerospace structures, their static and vibration characteristics have attracted the attention of many scientists in recent years. It is observed from the literature that the amount of such work carried out for isotropic plates are considerable, and limited literature is available on composite plates. However, the literature on the analysis of the FGMs plate is very few. Reddy [2] presented theoretical formulation and nite element models based on third order shear deformation theory for static and dynamic analysis of the FGM plates.

* Corresponding author. Tel.: +91 3222 283026; fax: +91 3222 255303. E-mail address: bnsingh@aero.iitkgp.ernet.in (B.N. Singh). 0307-904X/$ - see front matter 2010 Elsevier Inc. All rights reserved. doi:10.1016/j.apm.2010.03.034

3992

M. Talha, B.N. Singh / Applied Mathematical Modelling 34 (2010) 39914011

He obtained Navier solutions for a simply supported square plate under sinusoidally distributed load including the effect of shear deformation. Vel and Batra [3] presented a three-dimensional analytical solution for free and forced vibrations of simply supported FGM rectangular plates. Roque et al. [4] performed the free vibration analysis of the FGM plates by using the multiquadric radial basis function method, i.e., meshless method and a higher order shear deformation theory (HSDT). Efraim and Eisenberger [5] derived the equations of motion for annular plates using the rst-order shear deformation theory (FSDT) which includes the effect of shear deformations to solve vibration frequencies and modes for various combinations of boundary conditions. Abrate [6] analyzed the problems of free vibrations, buckling, and static deections of the FGM plates. The analysis of these problems is made using the classical laminated plate model, the FSDT model, and the HSDT model. Batra and Jin [7] used the rst-order shear deformation theory coupled with the nite element method to study the free vibrations of rectangular anisotropic FGM plate. Malekzadeh [8] presented the free vibration analysis of thick FGM plates supported on twoparameter elastic foundation. Fazelzadeh et al. [9] investigated the applicability of differential quadrature method for vibration analysis of aero-thermoelastic thin-walled blades made of FGM materials. First-order shear deformation theory is used to solve the governing equations of beams which include the effects of the rotary inertias and the blade presetting angle. Navazi and Haddadpour [10] analytically investigated the aero-thermoelastic stability margins of FGM panels in thermal environment. They employed piston theory of aerodynamics to model quasi-steady aerodynamic loading. Ng et al. [11] presented the parametric resonance of FGM rectangular plates under harmonic in-plane loading. The problem is formulated using Hamiltons principle and the material properties are graded according to a simple power-law distribution in the thickness direction only. It is concluded that the origin of the instability regions can be easily controlled by correctly varying the power-law exponent. Reddy and Cheng [12] studied the harmonic vibration problem of FGM plates by means of a three-dimensional asymptotic approach, and the formulation is based in terms of transfer matrix. Nguyen et al. [13] have proposed the FSDT model in which transverse shear factors is obtained through these plate models by using energy equivalence methods. Park [14] derived the frequency equation for the in-plane vibration of the clamped circular plate of uniform thickness with an isotropic material in the elastic range. Zenkour [15] studied the static response of a simply supported functionally graded rectangular plate subjected to a transverse uniform load using the generalized shear deformation theory. Uymaz and Aydogdu [16] presented three-dimensional vibration solutions for rectangular FGM plates with different boundary conditions based on the small strain linear elasticity theory. Chebyshev polynomials based on Ritz energy approach have been used in the analysis. Li et al. [17] studied free vibration of FGM sandwich rectangular plates with simply supported and clamped edges, and the formulation is based on the three-dimensional elasticity theory. Ferreira et al. [18] have used collocation multiquadric radial basis functions to analyse static deformations of simply supported functionally graded plate modelled by a higher order shear deformation theory. The determination of accurate behavior of the FGMs largely depends on the theory used to model the structure because in the FGMs, material properties vary continuously as a function of position in the preferred direction. Various concepts have been developed to inculcate the appropriate analysis of the FGM plates. The classical Kirchhoff plate theory neglects transverse shear deformation and gives acceptable results for relatively thin plates. In order to circumvent this problem an earlier attempts were made by Reissner [19] and Mindlin [20]. However, a shear-correction factor must be incorporated to overcome the problem of a constant transverse shear stress distribution and its value depends on various parameters, such as applied loads, boundary conditions and geometric parameters, etc. The inaccuracy occurs due to neglecting the effects of transverse shear and normal strains in the plate [21]. Due to continuous variation in material properties, the rst-order shear deformation theory and higher order shear deformation theory may be conveniently used in the analysis. It is noted that the rst-order shear deformation theory proposed by Mindlin [20] does not satisfy the parabolic variation of transverse shear strain in the thickness direction. Subsequently, many higher order theories were proposed, notable among them are [22 25]. The higher order theories assume the in-plane displacements as a cubic expression of the thickness coordinate and the out-of-plane displacement to be constant. Thus, the development of higher order shear deformation theory to assimilate the behaviour of FGM structures has been of high importance to the researchers. Therefore, keeping this view point in mind the present study aims to develop a higher order shear deformation theory with a special modication in the transverse displacement which contributes additional freedom to the displacements through the thickness and fundamentally eradicates the over-correction. To implement this theory a suitable C0 continuous isoparametric nite element with 13 degrees of freedom (DOFs) per node is proposed in order to reduce the computational efforts required in the formation of element matrices without affecting the solution accuracy. It is assumed that the plate is made of an isotropic material and its properties varying only through the thickness direction according to a simple powerlaw function of the position. Numerical results have been presented for ceramicmetal graded plates with different boundary and loading conditions. The convergence and comparison studies manifest the accuracy and precision of the present proposed method. The effects of aspect and thickness ratios, volume fraction index, various boundary and loading conditions on free vibration and static responses are investigated. Some results are presented in the form of tables and gures, which can suit as a benchmark for the future research. The paper is organised as follows. Section 2 gives the brief description of geometric conguration and material properties of the plate. Theoretical formulations of the problem are given in Section 3. In Section 4 solution methodology is presented

M. Talha, B.N. Singh / Applied Mathematical Modelling 34 (2010) 39914011

3993

followed by the convergence and validation studies in Section 5. Section 6 explains the results and discussion and Section 7 accomplishes the conclusions. 2. Geometric conguration and material properties Here we consider an FGM plate of length a, width b and total thickness h made from a mixture of metal and ceramics, in which the composition is varied from the top to the bottom surface as shown in Fig. 1. The top surface (Z = h/2) of the plate is ceramic-rich, whereas the bottom surface (Z = h/2) is metal rich. All formulation are conned here with the assumption of a linear elastic material behavior and small displacements and strains. The elastic material properties vary through the plate thickness according to the volume fractions of the constituents. Power-law distribution is commonly used to describe the variation of material properties (Fig. 2), which is expressed as

Pz Pc Pm V c Pm ;  n z 1 V c z 0 6 n 6 1; h 2

1 2

where P denotes the effective material property, Pm and Pc represents the properties of the metal and ceramic, respectively, Vc is the volume fraction of the ceramic and n is the volume fraction exponent. The effective material properties of the plate, including Youngs modulus E, density q vary according to Eq. (1) and m is assumed to be constant. 3. Theoretical formulation 3.1. Displacement eld and strains In the present study, system of governing equations for FGM plate is derived by using variational approach. The origin of the material coordinates is at the middle of the plate as shown in Fig. 1. For accurate analysis of transverse shear effects in the mathematical formulation the HSDT model has been used with a special modication in the transverse displacement.  for the plate are assumed as, ; v  and the transverse displacement w The in-plane displacements u

x; y; z; t u0 x; y; t zwx x; y; t z2 nx x; y; t z3 qx x; y; t; u  x; y; z; t v 0 x; y; t zwy x; y; t z2 ny x; y; t z3 qy x; y; t ; v  x; y; z; t w0 x; y; t zwz x; y; t z2 nz x; y; t ; w

; v  denote the displacements of a point along the (x, y, z) coordinates. u0, v0, and w0 are corresponding dis and w where u placements of a point on the mid-plane. wx and wy are the rotations of normal to the mid-plane about the y-axis and x-axis, respectively. The functions nx, ny, qx and qy are the higher order terms in the Taylor series expansion dened in the mid-plane of the plate. The higher order terms are determined by vanishing the transverse shear stresses sxz = s4 and syz = s5 on the top and bottom surfaces of the plate and by applying this boundary condition the displacement eld is modied as

 u0 f1 zwx f2 zax f3 zbx f4 zhx ; u  v 0 f1 zwy f2 zay f3 zby f4 zhy ; v  w0 f5 zwz f6 zaz ; w 4

a b

z
h 2 h 2
Fig. 1. Geometry and dimensions of the plate.

3994

M. Talha, B.N. Singh / Applied Mathematical Modelling 34 (2010) 39914011

Fig. 2. Variation of the volume fractions Vc through the thickness.

where f1(z) = C1z C2z3, f2(z) = C3z2, f3(z) = C4z3, f4(z) = C5z3, f5 = C1z, f6 = C1z2, C1 = 1, C2 = C4 = 4/3h2, C3 = 1/2, C5 = 1/3 and nz = az. To assure the eld variables are continuous within the element for C0 continuous nite element modelling, the out of plane derivatives are replaced by incorporating the following relations in Eq. (4) as:

ax

owz ; ox

bx

ow ; ox

hx

onz ; ox

ay

owz ; oy

by

ow ; oy

hy

onz : oy

However, the above substitution (Eq. (5)) imposes an articial constraints which are enforced variationally through a penalty approach [26] as:

owz ax 0; ox

owz ay 0; oy

ow bx 0; ox

ow by 0; oy

onz on hx 0 z hy 0: ox oy

From Eq. (4), the eld variables (basic unknowns) are interpreted as u, v, w, wx, wy, wz, ax, ay, az, bx, by, hx and hy for structural deformation. Mathematically, it can be represented as

T fKg u; v ; w; wx ; wy ; wz ; ax ; ay ; az ; bx ; by ; hx ; hy ;
where {K} is named as displacement vector. The linear strains corresponding to the displacement eld as given in Eq. (4) is expressed as

g fxx ; yy ; zz ; cyz ; cxz ; cxy gT ; f

or, it may be written as

8 19 8 29 8 39 k1 > > > k1 > > > > > > k1 > > > > > > > > > > > > > > > > > > > > > > > > > > > > > > 1> > > > > > 2> 3> > > > > > k k2 > > > > > 2> k > > > > > 2> > > > > > > > > > > > > > > > > > > > > > > > > > > > > > > 1> > > > > <0> < k3 = <0= = 2 3 z ; z z > > > > > > > > > > 2 > > > > > > > > > cyz > k e0 > > > > k1 > > > > > 0 4 4> > > > > > > > > > > 4> > > > > > > > > > > > > > > > > > > > > > > > > > > > > > > > > > 2> > > > > > > > > > > > > > > 0> > > > > > > > > > > 1 c e > > > > > > > > xz > k 0 5> > > > > > > > > > > k 5 > > > > > 5> > > > > > > > > > > > > > > > > > > > > > > > > > > > > > > > :c > : 0> : > > ; > ; > > 3; : : ; ; 2 e 1 xy k 6 k 6 k6 6 8 9 exx > > > > > > > > > > > > > > > > eyy > > > > > > > > > > > > > > > e < zz > = 8 9 > e0 1> > > > > > > > > > > > > > > > e0 > 2 > > > > > > > > > > > 0> > < e3 > =
where

M. Talha, B.N. Singh / Applied Mathematical Modelling 34 (2010) 39914011

3995

e0 1

ou0 ; ox ow 1 k1 x ; ox

ov 0 ow0 ow0 ou0 ov 0 e0 ; e0 ; e0 ; e0 ; 3 wz ; 4 wy 5 wx 6 oy oy ox oy ox owy owy ow ow ow 1 1 1 1 1 k2 ; k3 2az ; k4 z ay ; k5 z ax ; k6 x ; oy oy ox oy ox ow oay oaz oaz 2 2 2 2 k1 C 3 x ; k2 C 3 ; k4 hy 3C 2 wy by ; k5 hx 3C 2 wx bx ; ox oy oy ox       owy oby oax oay owx obx ohx ohy 2 3 3 k6 C 3 ; k1 C 2 C5 ; k2 C 2 C5 ; oy ox ox ox ox oy oy oy     owx owy obx oby ohx ohy 3 k6 C 2 C5 : oy ox oy ox oy ox

e0 2

3.2. Constitutive relations The linear constitutive relations are [27]

8 rxx 9 > > > > > > ryy > > > > > > > > > > < =

Q 11

Q 12 Q 22 Q 23 0 0 0

Q 13 Q 23 Q 33 0 0 0

0 0 0 Q 44 0 0

0 0 0 0 Q 55 0

0 0 0 0 0 Q 66

6Q 6 12 6 6 Q 13 rzz 6 6 0 > ryz > > > 6 > > > > 6 > > > > 4 0 r xz > > > > : ; rxy 0
where

9 38 exx > > > > > > 7> > eyy > > > > > 7> > > > 7> 7< ezz = 7 7> c >; 7> > yz > > > 7> > > 5> c > > > xz > > > :c > ;
xy

10

Q 11 Q 22 Q 33 Q 12 Q 13 Q 23 Q 44 Q 55 Q 66

Ez1 m2 ; 1 3m2 2m3 Ezm1 m ; 1 3m2 2m3 Ez : 2 1 m

11

The modulus E and the elastic coefcients Qij vary through the plate thickness according to Eqs. (1) and (2). 3.3. Strain energy The strain energy of the FGM plate is given by,

Z 1 fegT i frgi dV : 2 v

12

3.4. Kinetic energy The kinetic energy of the FGM plate can be expressed as

1 2

Z
V

 T  _ _ dV ; u q u

13

 g are the density and global displacement vector of the plate. The global displacement eld model as given by where q and fu Eq. (4) may be represented as,

g N fKg; fu
where {K} is as dened in Eq. (7) and the function of thickness coordinate N is dened as

14 3 15

2 1 0 0 6 N 40 1 0 0 Z Z
A Z

f1 z 0 0

0 f1 z 0

0 0 f5 z

f 2 z 0 0

0 f 2 z 0

0 0 f6 z

f3 z 0 0

0 f3 z 0

f 4 z 0 0

7 f4 z 5; 0

0 1

by substituting Eq. (14) into Eq. (13), the kinetic energy becomes

1 2

_ q K

Z n oT n o n o T T n o  _ dz dA 1 _ m K _ dA: N N K K 2 A

16

3996

M. Talha, B.N. Singh / Applied Mathematical Modelling 34 (2010) 39914011

7 4 1

8 5 2

9 6

Fig. 3. Node number for nine-noded rectangular element in natural coordinate system. Table 1 Properties of the FGM components. Material Properties E (N/m2) Aluminium (Al) Alumina (Al2O3) Zirconia (ZrO2) Silicon carbide (SiC) Ti6Al4V Stainless steel (SUS304) Silicon nitride (Si3N4) 70 10 380 109 151 109 427 109 105.7 109 207.78 109 322.27 109
9

m
0.30 0.30 0.30 0.17 0.298 0.3177 0.24

q (kg/m3)
2707 3800 3000 3210 4429 8166 2370

Table 2 q 2 2 2 4  x 121 v 2 qc x2 Comparison of linear frequency x l a b =p Ec h for various boundary conditions, n = 5 for (Al/ZrO2) square plate with different mess size. Mesh size Boundary conditions SSSS Present (2 2) Present (3 3) Present (4 4) Present (5 5) Uymaz and Aydogdu [16] Percentage difference 1.4564 1.4321 1.4222 1.4165 1.4106 0.416 CCCC 2.5371 2.2669 2.1944 2.1540 2.1447 0.431 SCSC 2.1457 2.0260 1.9287 1.8161 1.8055 0.583

Table 3  xa2 =h qm 1 m2 =Em 1=2 with volume fraction index n of (Ti6Al4V/ZrO2) square plate for simply supported (SSSS) Comparison of linear frequency x boundary condition, with different mess size. Mesh size Volume fraction index (n) ZrO2 Present (2 2) Present (3 3) Present (4 4) Present (5 5) Huang and Shen [30] Percentage difference 8.6061 8.3961 8.3245 8.2877 8.273 0.17 0.5 7.4217 7.2380 7.1758 7.1440 7.139 0.07 1 6.9653 6.7907 6.7319 6.7022 6.657 0.67 2 6.5789 6.4143 6.3589 6.3310 6.286 0.71 Ti6Al4V 5.6446 5.5066 5.4596 5.4102 5.400 0.18

Table 4  xa2 =hqc =Ec 1=2 with volume fraction index n of (SUS304/Si3N4) square plate for simply supported (SSSS) boundary Comparison of linear frequency x condition, with different mess size. Mesh size Volume fraction index (n) 0 Present (2 2) Present (3 3) Present (4 4) Present (5 5) Zhao et al. [31] Percentage difference 5.8914 5.7841 5.6769 5.6523 5.6148 0.663 0.5 4.1041 4.0223 3.9870 3.9201 3.8947 0.647 1 3.6913 3.4910 3.4894 3.4415 3.4242 0.502 2 3.2603 3.1696 3.1411 3.1062 3.0813 0.801 5 2.9539 2.8728 2.8472 2.8210 2.8058 0.538 8 2.8500 2.7721 2.7476 2.7258 2.7129 0.473 10 2.8099 2.7331 2.7089 2.6973 2.6768 0.760

M. Talha, B.N. Singh / Applied Mathematical Modelling 34 (2010) 39914011

3997

Table 5 Comparison of the centroidal deection of a simply supported (SSSS) (Al/ZrO2) square plate for various volume fraction index n and a/h = 5, with different mess size. Mesh size Volume fraction index (n) Ceramic Present (2 2) Present (3 3) Present (4 4) Present (5 5) Ferreira et al. [18] Percentage difference 0.0285 0.0352 0.0269 0.0250 0.0247 1.20 0.5 0.0357 0.0439 0.0336 0.0319 0.0313 1.88 1 0.0394 0.0487 0.0373 0.0358 0.0351 1.81 2 0.0430 0.0536 0.0409 0.0393 0.0388 1.27 Metal 0.0596 0.0738 0.0564 0.0541 0.0534 1.29

Fig. 4. Comparison of the computed non-dimensional center deection of an Al/SiC plate with [32].

Table 6 q 2 2 2 4  x 121 v 2 qc x2 Variation of the frequency parameter x l a b =p Ec h with the volume fraction index n for (CCCC) square (Al/ZrO2) FGM plates (a/h = 10). Boundary condition Mode n Ceramic CCCC 1 2 3 4 5 3.4173 6.7029 6.7029 9.4603 11.6199 0.5 3.0897 6.0788 6.0788 8.5930 10.5751 1 2.9436 5.7875 5.7875 8.1787 10.0627 5 2.7191 5.2956 5.2956 7.4474 9.1182 10 2.6461 5.1529 5.1529 7.2463 8.8667 Metal 2.4849 4.8682 4.8682 6.8670 8.4264

Table 7 q 2 2 2 4  x 121 v 2 qc x2 Variation of the frequency parameter x l a b =p Ec h with the volume fraction index n for (SSSS) square (Al/ZrO2) FGM plates (a/h = 10). Boundary condition Mode n Ceramic SSSS 1 2 3 4 5 1.9362 4.7231 4.7231 7.2780 9.1111 0.5 1.7566 4.2631 4.2631 6.5850 8.2711 1 1.6853 4.0609 4.0609 6.2752 7.8766 5 1.5692 3.7653 3.7654 5.7862 7.0493 10 1.5189 3.6664 3.6666 5.6308 6.8266 Metal 1.4099 3.4363 3.4367 5.2921 6.5701

3998

M. Talha, B.N. Singh / Applied Mathematical Modelling 34 (2010) 39914011

3.5. Work done due to transverse load The external work done on the plate by uniformly applied load p0 may be written as

W ext

1 2

Z
A

p0 x; yfwg dA:

17

Table 8 q 2 2 2 4  x 121 v 2 qc x2 Variation of the frequency parameter x l a b =p Ec h with the volume fraction index n for (CCCC) square (Al/ZrO2) FGM plates (a/h = 20). Boundary condition Mode n Ceramic CCCC 1 2 3 4 5 3.7343 7.8219 7.8219 11.4172 14.9035 0.5 3.3670 7.0664 7.0664 10.3233 13.5009 1 3.2102 6.7349 6.7349 9.8377 12.8641 5 2.9929 6.2424 6.2424 9.0950 11.8399 10 2.9124 6.0737 6.0737 8.8488 11.5115 Metal 2.7185 5.6899 5.6899 8.3029 10.8283

Table 9 q 2 2 2 4  x 121 v 2 qc x2 Variation of the frequency parameter x l a b =p Ec h with the volume fraction index n for (SSSS) square (Al/ZrO2) FGM plates (a/h = 20). Boundary condition Mode n Ceramic SSSS 1 2 3 4 5 1.9943 5.0388 5.0392 7.9752 10.3318 0.5 1.8074 4.5415 4.5417 7.1971 9.3352 1 1.7348 4.3322 4.3322 6.8699 8.9029 5 1.6215 4.0467 4.0467 6.4057 8.2691 10 1.5693 3.9372 3.9375 6.2286 8.0441 Metal 1.4530 3.6695 3.6698 5.8072 7.5198

Table 10 p  xa2 =h qc =Ec with the volume fraction index n for (CCCC) square (SUS304/Si3N4) FGM plates (a/h = 10). Variation of the frequency parameter x Boundary condition Mode n Ceramic CCCC 1 2 3 4 5 10.1599 19.9367 19.9367 28.1367 34.6017 0.5 7.0202 13.7978 13.7978 19.4845 23.9945 1 6.1489 12.0812 12.0812 17.0625 20.9992 5 4.9816 9.7440 9.7440 13.7350 16.8507 10 4.7457 9.2841 9.2841 13.0873 16.0556 Metal 4.4410 8.7107 8.7107 12.2919 15.1084

Table 11 p  xa2 =h qc =Ec with the volume fraction index n for (SSSS) square (SUS304/Si3N4) FGM plates (a/h = 10). Variation of the frequency parameter x Boundary condition Mode n Ceramic SSSS 1 2 3 4 5 5.7523 14.0336 14.0354 21.6188 27.1449 0.5 3.9701 9.6890 9.6906 14.9404 18.7691 1 3.4845 8.4903 8.4918 13.0959 16.4526 5 2.8351 6.8941 6.8952 10.6102 13.3057 10 2.6973 6.5669 6.5680 10.1053 12.6668 Metal 2.5154 6.1361 6.1370 9.4515 11.8664

Table 12 p  xa2 =h qc =Ec with the volume fraction index n for (CCCC) square (SUS304/Si3N4) FGM plates (a/h = 20). Variation of the frequency parameter x Boundary condition Mode n Ceramic CCCC 1 2 3 4 5 11.0932 23.2389 23.2389 33.8879 44.3223 0.5 7.6528 16.0474 16.0474 23.4088 30.6566 1 6.7074 14.0622 14.0622 20.5156 26.8554 5 5.4592 11.4138 11.4138 16.6363 21.7155 10 5.1991 10.8706 10.8706 15.8439 20.6790 Metal 4.8513 10.1598 10.1598 14.8146 19.3671

M. Talha, B.N. Singh / Applied Mathematical Modelling 34 (2010) 39914011

3999

4. Solution methodology 4.1. Finite element model 4.1.1. Introduction  and v  involve the derivatives of out-of-plane disFrom Eq. (4), it is seen that the expressions for in-plane displacement u placement w0. As a result of this, second order derivatives would be present in the strain vector, thus necessitating the employment of C1 continuity for nite element analysis. The complexity and difculty involved with making a choice of C1 continuity is well known, and C0 continuity permits easy isoparametric nite element formulation and consequently can be applied for non-rectangular geometries as well. Here in the present study, the derivatives are assumed as independent eld variables by treating ax, ay, bx, by, hx and hy as a separate DOFs. In this process, DOFs increase from 11 to 13 for the HSDT. However, the strain vector will be having only the rst-order derivatives, and hence a C0 continuous element would be sufcient for the present model. However, articial constraints are imposed which can be enforced variationally through a penalty approach in order to satisfy the constraints emphasized in Eq. (6).

Table 13 p  xa2 =h qc =Ec with the volume fraction index n for (SSSS) square (SUS304/Si3N4) FGM plates (a/h = 20). Variation of the frequency parameter x Boundary condition Mode n Ceramic SSSS 1 2 3 4 5 5.9240 14.9636 14.9651 23.6605 30.6973 0.5 4.0853 10.3158 10.3158 16.3181 21.1826 1 3.5865 9.0439 9.0439 14.3108 18.5757 5 2.9237 7.3701 7.3712 11.6551 15.1021 10 2.7815 7.0188 7.0198 11.0973 14.3809 Metal 2.5910 6.5449 6.5456 10.3487 13.4247

Table 14 Non-dimensional center deection with aspect ratio (b/a) for clamped-free (CFCF) Al/ZrO2 FGM plate. Aspect ratio (b/a) wc n=0 0.5 1.0 1.5 2.0 2.5 3.0 0.1153 0.5189 0.9742 1.3412 1.5923 1.7557 n = 0.5 0.1447 0.6556 1.2343 1.7016 2.0217 2.2301 n=1 0.1626 0.7352 1.3832 1.9062 2.2644 2.4976 n=2 0.1814 0.8135 1.5254 2.0984 2.4904 2.7452 n=5 0.2026 0.8966 1.6724 2.2948 2.7196 2.9953 n = 10 0.2158 0.9545 1.7804 2.4430 2.8951 3.1886

Fig. 5. Variation of frequency parameter with volume fraction index n and a/h ratio for clamped (CCCC) Al/Al2O3 square plate.

4000

M. Talha, B.N. Singh / Applied Mathematical Modelling 34 (2010) 39914011

A nine-noded isoparametric element is employed for nite element modeling. In the nite element, the domain is discretized into a set of nite elements. Over each of the elements, the displacement vector and element geometry of the model is expressed by

fKg

NN X i1

Ni fKgi ;

NN X i1

N i xi ;

NN X i1

Ni yi ;

18

where Ni is the interpolation function (shape function) for the ith node, {K}i is the vector of unknown displacements for the ith node, NN is the number of nodes per element and xi and yi are Cartesian coordinate of the ith node. The shape functions which are used in the nite element approximation are mentioned below [28] and the node numbering are shown in Fig. 3.

N1

1 2 1 n n g2 g ; N2 1 n2 g2 g ; 4 2 1 1 N3 n2 n g2 g ; N4 n2 n 1 g2 ; 4 2 1 N5 1 n2 1 g2 ; N6 n2 n 1 g2 ; 2 1 1 N7 n2 n g2 g ; N8 1 n2 g2 g ; 4 2 1 N9 n2 n g2 g : 4

4.1.2. Strain energy of the plate The strain energy of the FGM plate is given by,

NE X e1

U e :

19

Here NE is number of elements used for messing the plate U(e) is the elemental strain energy which can be obtained using Eqs. (10), (12) and (14) and expressed as

NE 1X T fKge K e fKge : 2 e1

20

Here [K](e) and {K}(e) are dened as linear stiffness matrix and displacement vector for the eth element, respectively. 4.1.3. Kinetic energy of the plate P e Summing over total number of element NE, kinetic energy of vibrating plate may be written as T NN e1 T , where

Fig. 6. Variation of frequency parameter with volume fraction index n and a/h ratio for clamped (CCCC) Al/Al2O3 rectangular plate.

M. Talha, B.N. Singh / Applied Mathematical Modelling 34 (2010) 39914011

4001

T e

1 2

Z
A
e

n oeT n oe _ _ K me K dA:

21

Here [m](e) is the inertia matrix of element. 4.1.4. Work done due to external transverse load The work done by external uniformly applied mechanical load p0(x, y) is given by

V W ext

1 2

Z
A

p0 x; yw dA:

22

Using the nite element notation model, above equation may be written as

NE X e 1

V e ;
R
Ae

23
fKgT fP g dA fKgeT fP ge with {P}(e) = (0 0 p0 0 0 0 0 0 0 0 0 0 0)T(e).

where V e

Fig. 7. Variation of frequency parameter with volume fraction index n and a/h ratio for simply supported (SSSS) Al/Al2O3 square plate.

Fig. 8. Variation of frequency parameter with volume fraction index n and a/h ratio for simply supported (SSSS) Al/Al2O3 rectangular plate.

4002

M. Talha, B.N. Singh / Applied Mathematical Modelling 34 (2010) 39914011

4.2. Governing equation The governing equation for free vibration and static problem of FGM plate can be derived using variational principle, which is generalisation of the principle of virtual displacement. Lagrange equation for a conservative system can be written as [29]

8 3 9 2 owz 2 wz 2 > > > > a bx oo x ox > > x > > > >   Z 6  2 7 < 7 6 2 d oT oU o c 6 h onz a owz 7 = oV d v 0; x y 7 6 ox oy _ ig 7 > ofqg 6 dt ofq ofqi g ofqi g > 2 >  2  2 5 > > > 4 > > > > w : ; by o hy oonyz oy

24

for i = 1, 2, . . .. _ i g are the generalized coordinates and generalized velocities, respectively, and c is the penalty parameter Here {qi} and fq that enforces constraints, as stated in Section 3.1 (Eqs. (5) and (6)). The equilibrium equation for free vibration and static

Fig. 9. Variation of frequency parameter with volume fraction index n and a/h ratio for simply supported and clamped (SSCC) Al/Al2O3 square plate.

Fig. 10. Variation of frequency parameter with volume fraction index n and a/h ratio for simply supported and clamped (SSCC) Al/Al2O3 rectangular plate.

M. Talha, B.N. Singh / Applied Mathematical Modelling 34 (2010) 39914011

4003

analysis with small deformation (i.e., small displacements and strain relation) can be obtained by substituting Eqs. (20), (21) and (23) in Eq. (24) as follows:

g K cK c fqg fF g; Mfq

25

g and {F} are global mass matrix, global linear stiffness matrix, global linear stiffness matrix arising where [M], [K], [Kc] {q}, fq due to constrains, global displacement, global acceleration and force vector, respectively. The generalized governing equation (24) can be employed to study the free vibration and static analysis by dropping the appropriate terms as: 1. Generalized eigenvalue problem for a system vibrating in principle mode with natural frequency x, can be expressed as

K cK c fqg kMfqg;
with k = x , where x is dened as frequency of natural vibration. 2. Linear static analysis
2

26

K cK c fqg fF g:

27

Fig. 11. Non-dimensional center deection vs. volume fraction index n for clamped (CCCC) Al/Al2O3 square plate.

Fig. 12. Non-dimensional center deection vs. volume fraction index n for simply supported plate (SSSS) Al/Al2O3 square plate.

4004

M. Talha, B.N. Singh / Applied Mathematical Modelling 34 (2010) 39914011

5. Numerical results The free vibration and static responses of the FGMs plate are computed using the proposed mathematical model in conjunction with FEM. A computer programme has been developed in MATLAB 7.5.0 (R2007b) environment. The validation and efcacy of the proposed algorithm is examined by comparing the results with those available in the literature. A nine-noded Lagrange isoparametric element, with 117 degrees of freedom per element for the present HSDT model has been used for discretizing the plate. For the computation of results full integration schemes (3 3) are used for thick plates and selective integration schemes (2 2) for thin plates. In the present analysis, various boundary conditions are used to check the efcacy of the model. However, the formulation and code do not put any limitations. Table 1 shows the properties of the FGMs constituents given at room temperature (300 K), which have been used for the computation of the results throughout the study, unless specied otherwise. Fig. 2 shows the volume fraction of the ceramic phase through the dimensionless thickness. It is assumed that the materials are perfectly elastic throughout the deformation. The boundary conditions used in the analysis are as follows: Simply supported (SSSS): u0 = w0 = wy = ax = az = by = hx = 0, at x = 0 and a. v0 = w0 = wx = ay = az = bx = hy = 0, at y = 0 and b.

Fig. 13. Non-dimensional center deection vs. volume fraction index n for simply supported and clamped (SSCC) Al/Al2O3 square plate.

Fig. 14. Non-dimensional center deection vs. volume fraction index n for clamped (CCCC) Al/Al2O3 rectangular plate.

M. Talha, B.N. Singh / Applied Mathematical Modelling 34 (2010) 39914011

4005

Clamped (CCCC): u0 = v0 = w0 = wx = wy = wz = ax = ay = az = bx = by = hx = hy = 0, at x = 0, a and y = 0, b. Simply supported and clamped (SSCC): u0 = w0 = wy = ax = az = by = hx = 0, at x = 0 and a. u0 = v0 = w0 = wx = wy = wz = ax = ay = az = bx = by = hx = hy = 0, at y = 0, b. Clamped-free (CFCF): u0 = v0 = w0 = wx = wy = wz = ax = ay = az = bx = by = hx = hy = 0, at x = 0, and y = 0. u0 v0 w0 wx wy wz ax ay az bx by hx hy 0, at x = a and y = b. Simply supported-clamped (SCSC): u0 = w0 = wy = ax = az = by = hx = 0, at x = 0 and y = 0. u0 = v0 = w0 = wx = wy = wz = ax = ay = az = bx = by = hx = hy = 0, at x = a, y = b. 5.1. Convergence and validation study To make certain the accuracy and prociency of the present nite element formulation, three test examples have been analyzed for free vibration of the FGM plates.

Fig. 15. Non-dimensional center deection vs. volume fraction index n for simply supported (SSSS) Al/Al2O3 rectangular plate.

Fig. 16. Non-dimensional center deection vs. volume fraction index n for simply supported and clamped (SSCC) Al/Al2O3 rectangular plate.

4006

M. Talha, B.N. Singh / Applied Mathematical Modelling 34 (2010) 39914011

5.1.1. Free vibration analysis

Example 1. We rst consider the accuracy of the present nite element formulation by comparing the results with those given by Uymaz and Aydogdu [16] which is based on the small strain linear elasticity theory. A convergence study is also presented. A FGM plate with different boundary conditions is analyzed with the proposed nite element formulation. In this example, the analysis is performed with n = 5, aspect ratio a/b = 1 and thickness ratio a/h = 5, where n is the volume fraction index and h is the thickness of the plate as dened earlier. The top face of the plate is ceramic-rich, whereas the bottom face is metal rich. The plate is comprised of metal (aluminium) and ceramic (zirconia). The material properties as given in [16], are Eb = 70 109 N/m2, qb = 2702 kg/m3 for aluminium, and Et = 151 109 N/m2, qt = 3000 kg/m3 for zirconia. The Poissons ratio m is assumed to be constant as 0.3. The non-dimensional frequency parameters are obtained by applying three 2 121 m2 qc x2 a2 b2 =p4 Ec h2 . Based on the convergence, it is concluded that (5 5) dimensional continuum method as: x mesh is acceptable for free vibration analysis of the FGM plate. The accomplished results are in good agreement with the published results, as shown in Table 2.

Fig. 17. Non-dimensional deection due to uniformly applied load vs. non-dimensional length for Al/ZrO2 rectangular FGM plate (a/h = 5) with simply supported-clamped (SCSC) boundary condition.

Fig. 18. Non-dimensional deection due to uniformly applied load vs. non-dimensional length for Al/ZrO2 rectangular FGM plate (a/h = 10) with simply supported-clamped (SCSC) boundary condition.

M. Talha, B.N. Singh / Applied Mathematical Modelling 34 (2010) 39914011

4007

Example 2. In this example, we consider the free vibration analysis of a square FGM plate made of zirconium oxide (ZrO2) and titanium alloy (Ti6Al4V), as given by Huang and Shen [30]. The top surface of the plate is assumed to be ceramic-rich, whereas the bottom surface is metal rich. The side and thickness of the square plate are a = 0.2 m and h = 0.025 m, respectively. Poissons ratio m is assumed to be constant and is equal to 0.3. Youngs modulus and mass density listed in Table 1. The  xa2 =h plate is simply supported at all edges and non-dimensional frequency parameter are dened as: x 1=2 2 qm 1 m =Em . The calculated frequency parameters are compared with Ref. [30], as shown in Table 3, again a good agreement between the results are accomplished. Example 3. In this example, we consider the FGM plate made of silicon nitride (Si3N4) and titanium alloy (SUS304), as given by Zhao et al. [31] for free vibration. They employed rst-order shear deformation plate theory to account for the transverse shear strain and rotary inertia. Mesh-free kernel particle functions are used to approximate the two-dimensional displacement elds. The side to thickness ratio (a/h) is equal to 10. The Poissons ratio for Si3N4 and SUS304 is assumed as 0.24 and 0.3117, respectively. Youngs modulus and mass density are listed in Table 1. The plate is simply supported at all four edges  xa2 =hqc =Ec 1=2 . The calculated frequency parameters are and non-dimensional frequency parameter are dened as: x compared with Ref. [31], as shown in Table 4, again a good agreement between the two results are obtained. These three comparison studies show that the present result matches very well with the established one.

Fig. 19. Non-dimensional deection due to uniformly applied load vs. non-dimensional length for Al/ZrO2 rectangular FGM plate (a/h = 15) with simply supported-clamped (SCSC) boundary condition.

Fig. 20. Non-dimensional deection due to uniformly applied load vs. non-dimensional length for Al/ZrO2 rectangular FGM plate (a/h = 20) with simply supported-clamped (SCSC) boundary condition.

4008

M. Talha, B.N. Singh / Applied Mathematical Modelling 34 (2010) 39914011

5.1.2. Static analysis Two test examples have been analyzed for static analysis of the FGM plates to check the accuracy and prociency of the present nite element formulation. Example 1. Here we consider the square FGM plate simply supported at all four edges with the proposed nite element formulation. In this example, the analysis is performed for thickness ratio (a/h) = 5, and aspect ratio a/b = 1 with different volume fraction indices n = 0, 0.5, 1, 2 and 1, where a and b are the sides of the plate and h is the thickness of the plate. For this problem, the convergence along with validation is also conducted. The various non-dimensionalized parameters used 4 are: central deection, w/h, load parameter p0 p=Em h and thickness coordinate,  z z=h. The analysis is carried out for FGM plate made of aluminium and zirconia. The obtained results with the present outlined approach are presented in Table 5 and compared with that of Ferreira et al. [18] that is based on the collocation multiquadric radial basis functions by a third order shear deformation theory. It is observed that both results agree well. Based on the convergence, it is observed that (5 5) mesh is sufcient for static analysis. Hence for all computation, a (5 5) mesh is adopted.

Fig. 21. Non-dimensional deection due to uniformly applied load vs. non-dimensional length for Al/ZrO2 rectangular FGM plate (a/h = 5) with clampedfree (CFCF) boundary condition.

Fig. 22. Non-dimensional deection due to uniformly applied load vs. non-dimensional length for Al/ZrO2 rectangular FGM plate (a/h = 10) with clampedfree (CFCF) boundary condition.

M. Talha, B.N. Singh / Applied Mathematical Modelling 34 (2010) 39914011

4009

Example 2. In this example, the numerical and analytical solutions for the centroid transverse deection of the plate is compared with Qian et al. [32]. The top face of the plate is loaded by a normal pressure given by q0sin px/asin py/a. The plate is comprised of Al/SiC with side to thickness ratio a/h = 5 and all edges are simply supported. The transverse displacement w
Em h z  100 w;  z2 . The non-dimensional center deecand thickness coordinate z have been non-dimensionalized as: w h 12a4 1m2 q
m 0 3

tions with volume fraction index n along with the gure layout are preferred as used in Ref. [32] for direct comparison. Again, a good agreement is observed between the two results as shown in Fig. 4.

6. Results and discussion Based on the analysis of foregoing sections it is observed that (5 5) mesh has been found to give good convergence for the FGM plates as mentioned earlier. These have been used for accomplishing the results, unless otherwise stated. Tables 6 and 7 show the variation of non-dimensional frequency parameter with the volume fraction index n for the Al/ZrO2 square FGM plate. Choosing n = 0 and 1 corresponds to the frequency of the ceramics and metallic components, respectively. The

Fig. 23. Non-dimensional deection due to uniformly applied load vs. non-dimensional length for Al/ZrO2 rectangular FGM plate (a/h = 15) with clampedfree (CFCF) boundary condition.

Fig. 24. Non-dimensional deection due to uniformly applied load vs. non-dimensional length for Al/ZrO2 rectangular FGM plate (a/h = 20) with clampedfree (CFCF) boundary condition.

4010

M. Talha, B.N. Singh / Applied Mathematical Modelling 34 (2010) 39914011

results are computed only for the rst ve modes with side to thickness ratios a/h = 10 for CCCC and SSSS boundary conditions. It is observed that as the volume fraction index increases the frequencies in all ve modes decrease. This is due to the fact that, the larger volume fraction index means the plate has a smaller ceramic component and hence the stiffness is reduced. It is also observed that increasing constraints on the boundaries increase the frequency parameter. Variation of the frequency parameter with the volume fraction index of Al/ZrO2 FGM plate for a/h = 20 with CCCC and SSSS boundary condition is presented in Tables 8 and 9, respectively. It is noticed that frequency parameter decreases with the increase of volume fraction index n as expected. It is anticipated that there is approximately 9% rise in frequency parameter for CCCC boundary condition as a/h ratio changes from 10 to 20 and nearly 3% rise for SSSS boundary condition. In Tables 1013, the frequency characteristics for square SUS304/Si3N4 FGM plates are investigated for ve modes only at a/h = 10 and 20 for CCCC and SSSS boundary conditions. It is observed that the frequency parameter decreases as volume fraction index n increases, as expected. This decrease is different for CCCC and SSSS boundary conditions considered. The considerable decrease is found for CCCC boundary conditions. Table 14 shows the non-dimensional center deection of CFCF, Al/ZrO2 FGM plates with volume fraction index n for aspect ratio (a/b) ranges from 0.5 to 3. The plates are subjected to a sinusoidal load and side to thickness ratio (a/h) is equal to 5. It is found that the deection increases as volume fraction index n increases. This is due to the fact that the bending stiffness is the maximum for fully ceramic plate, i.e., (n = 0) and degrades gradually as n increases. Fig. 5 illustrates the effect of volume fraction index and side to thickness ratio (a/h) on the fundamental natural frequency parameter of Clamped square Al/ZrO2 FGM plates. It is observed that with a particular volume fraction index, the frequency parameter increases as the side to thickness ratio increases. The rise in frequency parameter is observed up to around a/ h = 20, beyond this no changes in frequency parameter are distinguished. Fig. 6 shows the variation of fundamental natural frequency parameter with a/h ratio and volume fraction index n for CCCC boundary conditions of rectangular Al/ZrO2 FGM plates. Bestowing to this gure the frequency parameter increases with increasing side to thickness (a/h) ratio, for (a/h) less than 20 as observed in Fig. 5. Figs. 7 and 8 show the inuence of volume fraction index and side to thickness ratio on the fundamental frequency parameters for SSSS square and rectangular Al/Al2O3 plates, respectively, and inuence of volume fraction index and side to thickness ratio on the fundamental frequency parameters for SSCC square and rectangular Al/Al2O3 plates are shown in Figs. 9 and 10, respectively. It is found that the similar frequency parameter behavior is noticed to those obtained in Fig. 5. It is therefore accomplished that effects of the side to thickness ratio on the natural frequency parameter of plates is independent of the variation in the volume fraction index. Figs. 1113 represent the variation of the non-dimensional central deection with volume fraction index for the aluminiumalumina square plates having CCCC, SSSS and SSCC boundary conditions. The load parameter is taken as (+1). It can be ascertain that all of the curves that represent the various boundary conditions show a similar behaviour with the deection enhances as the volume fraction index increases. This is due to the fact that the larger volume fraction index means the plate has a smaller ceramic component and hence the stiffness is reduced. Figs. 1416 show the effects of volume fraction index and side to thickness ratio on the non-dimensional central deection for CCCC, SSSS and SSCC rectangular Al/Al2O3 plates. It is observed that the rectangular plates have slightly higher central deection than square plate for same boundary conditions and material properties. Figs. 1720 show the non-dimensional deection due to uniformly applied load vs. non-dimensional length for rectangular plates (a/b = 0.5) with varying volume fraction index n. The FGM plate is made up of Al/ZrO2 and the side to thickness ratio (a/h) of the plate is varying from 5 to 20. Simply supported-clamped boundary conditions have been used in the analysis. It is perceptible from the gure that the plate deection lies between those made of ceramic and metal. The variation of the non-dimensional deection due to uniformly applied load vs. non-dimensional length for rectangular plates (a/b = 0.5) with varying volume fraction index n is described in Figs. 2124. The plate is comprised of Al/ZrO2 and the side to thickness ratio (a/h) of the plate is varying from 5 to 20 with CFCF boundary condition. The deection characteristics shown in these four gures are similar to those in Figs. 1720. 7. Conclusions An extensive study of the free vibration and static analysis of square and rectangular functionally graded plates is presented, which is based on the higher order shear deformation theory with a special modication in the transverse displacement in conjunction with nite element models. The systems of algebraic equations are derived using variational approach for the free vibration and static problem. A C0 isoparametric Lagrangian element with 117 degrees of freedom per element is developed and implemented for both the problems. Convergence and validation studies have been carried out to inculcate the accuracy of the present formulation. The obtained result shows a good agreement with those available in the literature ranging from thin to thick plates. The following conclusions are noted from the typical results obtained for different volume fraction indices, the aspect ratios, the thickness ratios and different combinations of the boundary conditions.  The frequency parameter increases with the increase in plates aspect ratio (a/b) and smaller side to thickness (a/h) ratio.  The frequency parameter decreases with the increases of volume fraction index, n.

M. Talha, B.N. Singh / Applied Mathematical Modelling 34 (2010) 39914011

4011

 The frequency parameter increases with increasing (a/h) ratio for a/h 6 20 and becomes approximately insensitive to a/h ratio greater than 40.  There is an increases in frequency parameter as support condition changes from all edges simply supported (SSSS) to all edges clamped (CCCC).  The non-dimensional central deection for a given thickness ratio increases as the volume fraction index n increases.  The non-dimensional central deection decreases with the increase of plates aspect ratio (a/b).  The non-dimensional central deection with non-dimensional length shows that the plates with intermediate properties undergo corresponding intermediate values of center deection.

References
[1] [2] [3] [4] [5] [6] [7] [8] [9] [10] [11] [12] [13] [14] [15] [16] [17] [18] [19] [20] [21] [22] [23] [24] [25] [26] [27] [28] [29] [30] [31] [32] M. Koizumi, FGM activities in Japan, Compos. Part B 28 (1997) 14. J.N. Reddy, Analysis of functionally graded plates, Int. J. Numer. Method Eng. 47 (2000) 663684. S.S. Vel, R.C. Batra, Three-dimensional exact solution for the vibration of functionally graded rectangular plates, J. Sound Vib. 272 (2004) 703730. C.M.C. Roque, A.J.M. Ferreira, R.M.N. Jorge, A radial basis function approach for the free vibration analysis of functionally graded plates using a rened theory, J. Sound Vib. 300 (2007) 10481070. E. Efraim, M. Eisenberger, Exact vibration analysis of variable thickness annular isotropic and FGM plates, J. Sound Vib. 299 (2007) 720738. S. Abrate, Free vibration buckling and static deections of functionally graded plates, Compos. Sci. Technol. 66 (2006) 23832394. R.C. Batra, J. Jin, Natural frequencies of a functionally graded anisotropic rectangular plate, J. Sound Vib. 282 (2005) 509516. P. Malekzadeh, Three-dimensional free vibration analysis of thick functionally graded plates on elastic foundations, Compos. Struct. 89 (2009) 367 373. S.A. Fazelzadeh, P. Malekzadeh, P. Zahedinejad, M. Hosseini, Vibration analysis of functionally graded thin-walled rotating blades under high temperature supersonic ow using the differential quadrature method, J. Sound Vib. 306 (2007) 333348. H.M. Navazi, H. Haddadpour, Aero-thermoelastic stability of functionally graded plates, Compos. Struct. 80 (2007) 580587. T.Y. Ng, K.Y. Lam, K.M. Liew, Effects of FGM materials on the parametric resonance of plate structures, Comput. Methods Appl. Mech. Eng. 190 (2000) 953962. J.N. Reddy, Z.Q. Cheng, Frequency of functionally graded plates with three-dimensional asymptotic approach, J. Eng. Mech. 129 (8) (2003) 896900. T.M. Nguyen, K. Sab, G. Bonnet, First-order shear deformation plate models for functionally graded materials, Compos. Struct. 83 (2008) 2536. Chan Park II, Frequency equation for the in-plane vibration of a clamped circular plate, J. Sound Vib. 313 (2007) 325333. A.M. Zenkour, Generalised shear deformation theory for bending analysis of functionally graded plates, Appl. Math. Model. 30 (2006) 6784. B. Uymaz, M. Aydogdu, Three-dimensional vibration analysis of functionally graded plates under various boundary conditions, J. Reinf. Plastic Compos. 26 (18) (2007) 18471863. Q. Li, V.P. Iu, K.P. Kou, Three-dimensional vibration analysis of functionally graded material sandwich plates, J. Sound Vib. 311 (2008) 498515. A.J.M. Ferreira, R.C. Batra, C.M.C. Roque, L.F. Qian, P.A.L.S. Martins, Static analysis of functionally graded plates using third order shear deformation theory and a meshless method, Compos. Struct. 69 (2005) 449457. E. Reissner, The effect of transverse shear deformation on the bending of elastic plates, J. Appl. Mech. 12 Trans (1945) A69A77. R.D. Mindlin, Inuence of rotatory inertia and shear on exural vibrations of isotropic elastic plates, J. Appl. Mech. 73 (1951) 3138. H. Matasunaga, Analysis of functionally graded plates, Compos. Struct. 82 (2008) 499512. R.B. Nelson, D.R. Larch, A rened theory of laminated orthotropic plates, J. Appl. Mech. 41 (1974) 177183. E. Reissner, On transverse bending of plates including the effect of transverse shear deformation, Int. J. Solids Struct. 11 (1975) 569573. K.H. Lo, R.M. Christensen, E.M. Wu, A high-order theory of plate deformation part I: homogeneous plates, J. Appl. Mech. 44 (1977) 663668. J.N. Reddy, A simple higher-order theory for laminated composite plates, J. Appl. Mech. 51 (1984) 745752. R.D. Cook, D.S. Malkus, M.E. Plesha, R.J. Witt, Concept and Applications of Finite Element Analysis, fourth ed., Wiley, Singapore, 2007. R.M. Jones, Mechanics of Composite Materials, second ed., Taylor & Francis, Philadelphia, PA, 1999. J.N. Reddy, An Introduction to Finite Element Method, third ed., Tata McGraw-Hill Edition, New York, 2009. D. Thorby, Structural Dynamics and Vibration in Practice: An Engineering Handbook, rst ed., Elsevier, Amsterdam, 2008. X.L. Huang, H.S. Shen, Nonlinear vibration and dynamic response of functionally graded plates in thermal environments, Int. J. Solids Struct. 41 (2004) 24032427. X. Zhao, Y.Y. lee, K.M. Liew, Free vibration analysis of functionally graded plates using the element free kp Ritz method, J. Sound Vib. 19 (2009) 918 939. L.F. Qian, R.C. Batra, L.M. Chen, Static and dynamic deformations of thick functionally graded elastic plates by using higher-order shear and normal deformable plate theory and meshless local PetrovGalerkin method, Compos. Part B 35 (2004) 685697.