2 Laminated composite plate formulation based on Bézier extraction for NURBS

respectively. Therefore, the UHSDT can be called the unconstrained inverse

trigonometric shear deformation theory (UITSDT) and the unconstrained sinusoidal

shear deformation theory (USSDT), respectively. By the assumption that a linear

constitutive relationship is employed for the analysis of the laminated composite plate

embedded in piezoelectric layers, the formulation for each field is approximated

separately.

The in-plane strain vector ε p is thus expressed by the following equation

[ε xx ε yy γ xy ]T= ε 0 + zε1 + f ( z )ε 2

ε=

p



(4. 1)



and the transverse shear strain vector γ has the following form

=

γ γ xz



T



γ yz  =

ε 0s + f ' ( z )ε1s



(4. 2)



where f ' ( z ) is derivative of f(z) function and



 u2, x 

 u0, x 

 u1, x 

u1 + w, x 













ε 0 =  v0, y  , ε1 =  v1, y  , ε 2 =  v2, y  , ε 0s = 

,

+

v

w

y

1

,





v2, x + u2, y 

v0, x + u0, y 













v1, x + u1, y 



(4. 3)



u 

ε1s =  2 

 v2 

A weak form of the static model for the plates under transverse loading q0 can be

written as



∫



Ω



δεTp Dε p dΩ + ∫ δγ T =

D s γ dΩ

Ω



∫



Ω



δ wq0dΩ



(4. 4)



where q0 is the transverse loading per unit area.

From Hooke’s law and the linear strains given by Eqs.(4. 1) and (4. 2), the stress

is computed by

σ p 

=

σ =



τ



 D 0  ε p 



 = cε

Ds   γ 

0



ε



c



61



(4. 5)



where σ p and τ are the in-plane stress component and shear stress; D and Ds are

material constant matrices given in the form of

A B E 

A

D =  B D F  ; Ds =  s

 Bs

 E F H 



Bs 

Ds 



(4. 6)



in which



( Aij , Bij , Dij , Eij , Fij , H ij ) = ∫



h /2



− h /2



(1, z , z 2 , f ( z ), zf ( z ), f 2 ( z ))Qij dz;



i, j = 1, 2, 6



( Asij , Bsij , Dsij ) = ∫



h /2



− h /2



(4. 7)



1, f ′( z ),( f ' ( z )) 2  Qij dz; i, j = 4,5



where Q ij is the transformed material constant matrix (see section 3.3.2 for further

details).

For forced vibration analysis of the plates, a weak form can be derived from the

following undamped dynamic equilibrium equation as follows:



∫



Ω



T

δεTp Dε p dΩ + ∫ δγ T Ds γdΩ + =

∫ δ u mu dΩ

Ω



Ω



∫



Ω



δ wq( x, y, t )dΩ



(4. 8)



where the mass matrix m is calculated in a consistent form as follows

m



 I1 I 2 I 4 

=



 I 2 I 3 I 5  , ( I1 , I 2 , I 3 , I 4 , I 5 , I 6 )

 I 4 I 5 I 6 



h /2



∫ ρ (1, z, z



− h /2



2



, f ( z ), zf ( z ), f 2 ( z ) )dz



(4. 9)



in which ρ is the mass density,

 u1 

 

u 2  , u1

u = =

u 

 3



u0 

 

v0  ; u 2

=

w

 



u1 

 

v1  ; u 3

=

0

 



u2 

 

 v2 

 

0



(4. 10)



and q(x,y,t) is the transverse loading per unit area which is the function depending on

time and space.

It should be noted that no external forces are required in the free vibration

problems, and the terms on the right-hand side of Eq.(4. 8) is thus equivalent to zero.

4.2.2 Approximated formulation based on Bézier extraction for NURBS



62



By using the Bézier extraction for NURBS, the displacement field u of the plate

is approximated as follows

m×n



u h (ξ ,η ) = ∑ RAe (ξ ,η )d A



(4. 11)



A



where n×m is the number of basis functions, RAe (ξ ,η ) is a NURBS basis function

for two-dimensional problems which is written in form of the linear combination of

Bézier extraction operator and Bernstein polynomials, PA is the control point A and



d A = [u0 A v0 A u1 A v1 A u2 A v2 A



wA ] is the vector of nodal degrees of freedom

T



associated with control point A.

By substituting Eq. (4. 11) with Eq.(4. 1), the in-plane and shear strains can be

rewritten as

T



m×n



ε p γ  = ∑ B mA

A=1



T



B sA1 B sA2  q A



BbA1 BbA2



(4. 12)



in which



B mA



 RA, x

0 0 0 0 0 0





0

RA, y 0 0 0 0 0  , BbA1

=

 RA, y RA, x 0 0 0 0 0 







 0 0 RA, x



0 0 0

 0 0 RA, y





 0 0 0 0 RA, x



BbA2 = 0 0 0 0 0

 0 0 0 0 RA, y





B sA1



0

RA, y

RA, x



0

RA, y

RA, x



0 0 0



0 0 0

0 0 0 



(4. 13)



0



0

0 



 0 0 RA 0 0 0 RA, x  s 2  0 0 0 0 RA

=

 , BA 

0 0 0 0 0

 0 0 0 RA 0 0 RA, y 



0

RA



0

0 



By substituting Eq. (4. 12) with Eq.(4. 4), the formulation of static analysis is

obtained in the following form



63



Kd = F



(4. 14)



where the global stiffness matrix K is given by

 B m T

 

=

K ∫  Bb1 

Ω

 b2 

 B 



m

 A B E  B 

s1 T

 B D F  Bb1  +  B 



   B s 2 

 E F H  Bb 2   

 



 As

B

 s





Bs   B 

 dΩ

Ds  B s 2 



s1



(4. 15)



and the load vector F is calculated as



∫



=

F



Ω



q0 R 0dΩ



(4. 16)



in which

R 0 = [ 0 0 0 0 0 0 RA ]



(4. 17)



For free vibration analysis, one has

 + Kd = 0

Md



(4. 18)



where the global mass matrix M is described as



  N 0 T



=

M ∫   N1 

Ω

N 

 2 



 I1

I

 2

 I 4



I2

I3

I5



I 4  N0  



I 5   N1   dΩ

I 6   N 2  





(4. 19)



with

 RA

N 0 =  0

 0



N1



0

RA

0



0 0 0 0 0

0 0 0 0 0  ;

0 0 0 0 RA 



 0 0 RA 0 0 0 0 

=



 0 0 0 RA 0 0 0  ; N 2

0 0 0

0 0 0 0 



 0 0 0 0 RA

0 0 0 0 0



0 0 0 0 0



(4. 20)

0

RA

0



0

0 

0 



And for forced vibration analysis, undamped dynamic discrete equations can be

expressed from Eq.(4. 8)

 + Kd = F(t )

Md



64



(4. 21)



To solve this second order time-dependent problem, several methods have been

proposed such as Wilson, Newmark, Houbolt, Crank-Nicholson, etc. Here, Eq.(4. 21)

is solved by the Newmark direct integration scheme.

4.3 Theory and formulation of the piezoelectric laminated composite plates

4.3.1 Variational forms of piezoelectric composite plates

The summation of kinetic energy, strain energy, dielectric energy and external

work named the generalized energy function is written in the following form [28]



L

=



1



1



1



∫  2 ρu u − 2 σ ε + 2 D E + u f

T



T



T



T



s





− φq s dΩ + ∑ uT Fp − ∑ φQ p





(4. 22)



where ρ is the mass density, u and u are the mechanical displacement and velocity;



φ is the electric potential; f s and Fp are the mechanical surface loads and point loads;



q s and Q p are the surface charges and point charges.

The Galerkin weak form of piezoelectric structures derived by using

Hamilton’s variational principle can be possibly written as



δΠ =0



(4. 23)



The material behavior of actuators and sensors made of the piezoelectric

composite can be modeled as the following constitutive equations [125]

 σ  c −eT   ε 

 

 D = 

  e g   E 



(4. 24)



in which c, the elasticity matrix, is defined as



A B L 0

B G F 0



c = L F H 0



 0 0 0 As

 0 0 0 B s



0

0 

0



Bs 

Ds 



(4. 25)



in which



1, z , z 2 , f ( z ), zf ( z ), f 2 ( z ) ) Qij dz i, j

( A, B, G, L, F, H ) ∫=

(

− h /2

h /2

2

=

1, f ′( z ), ( f ′( z ) ) ) Qij dz i, j 4,5

( A s , B s , Ds ) ∫=

(

− h /2

h /2



65



1, 2,6



(4. 26)



where Qij is calculated as in Eq.(3. 18).

4.3.2 Approximated formulation of electric potential field

To approximate the electric potential field, each thin piezoelectric layer is

discretized into a lot of finite sublayers through the thickness dimension. Besides, the

electric potential variation is assumed to be linear in each sublayer and is

approximated throughout the piezoelectric layer thickness as follows [150]:



φ i ( z ) = Nφi φi



(4. 27)



where Nφi is the shape functions for the electric potential with p = 1, and φi is the

vector containing the electric potentials at the top and bottom surfaces of the i-th



=

φi =

φ i −1 φ i  (i 1, 2,...., nsub ) in which nsub is the number of

sublayer,

piezoelectric layers.

For each piezoelectric sublayer element, values of electric potentials are

assumed to be equal at the same height along the thickness [125]. The electric field

E can be rewritten as

E = −∇Nφi φi = −Bφ φi



(4. 28)





Bφ = 0 0





(4. 29)



in which



1 



hp 



in which hp is the thickness of piezoelectric layer. Note that, for the type of

piezoelectric materials considered in this work the piezoelectric constant matrix e and

the dielectric constant matrix g of the kth orthotropic layer in the local coordinate

system are written as follows [150]

(k )



e(



k)



 0 0 0 0 e15 

(k )

=



g

e

0

0

0

0

;

15





e31 e32 e33 0 0 



 p11

0



 0



0

p22

0



0 

0 

p33 



(k )



(4. 30)



However, the laminate is usually made of several orthotropic layers with

different directions of orthotropy and consequently different characteristic directions



66