You are on page 1of 16

Applied Mathematical Modelling 38 (2014) 1737–1752

Contents lists available at ScienceDirect

Applied Mathematical Modelling


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

Large deflection of magneto-electro-elastic laminated plates


A. Milazzo ⇑
Dipartimento di Ingegneria Civile, Ambientale, Aerospaziale, dei Materiali, Universitá di Palermo, Viale delle Scienze, Edificio 8, I90128 Palermo, Italy

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

Article history: A model for the large deflection analysis of magneto-electro-elastic laminated plates is
Received 15 May 2012 derived. The first order shear deformation theory and the von Karman stress function
Received in revised form 6 December 2012 approach are employed. A set of resolving partial differential equations involving kinemat-
Accepted 27 August 2013
ical variables and the stress function is obtained as a consequence of the preliminary
Available online 8 October 2013
condensation of the electro-magnetic state to the plate kinematics. A closed form solution
for simply-supported plates is presented. Numerical results are carried out for plates con-
Keywords:
sisting of piezoelectric BaTiO3 and piezomagnetic CoFe2 O4 layers. These results show the
Magneto-electro-elastic laminates
Smart structures
influence of large deflections on the plate response and could be useful in the analysis
Plate large deflection and design of layered magneto-electro-elastic composite plates.
von Karman plate theory Ó 2013 Elsevier Inc. All rights reserved.

1. Introduction

Smart structures are of interest in many technological fields, including the next generation of transport vehicles, health
monitoring, vibration control, sensor and actuator applications, robotics, medical instruments and energy harvesting only to
name a few. Indeed, they allow to design and employ multi-functional components as a mean to accomplish high reliability
and performance requirements. In this context, smart laminates belong to a very important and promising class of new
generation smart structures. In particular, magneto-electro-elastic composites containing piezoelectric and piezomagnetic
phases have recently emerged [1] owing to their ability to convert energy among the electric, magnetic and elastic
form and also as they exhibit a new property known as magnetoelectric effect, which refers to the capability of passively
coupling the electric and magnetic fields [2]. The laminated form of these composites appears more efficient with respect
to the bulk form and their modeling is essential to properly understand their behavior and provide for adequate analysis
and design.
Due to the multi-layer configuration and the multi-field behavior of the employed materials, structural modeling of smart
laminates requires special attention. The most accurate characterization of the smart laminates response is obtained by ana-
lytical exact or numerical 3-D solutions, when no assumptions are made on the variation of the involved field variables.
However, analytical exact solutions can be obtained for specific geometries and boundary conditions only, whereas 3-D finite
element solutions are carried out with meaningful computational costs. So 2-D efficient laminate theories play an important
role as they allow to reduce the analysis effort preserving a suitable level of accuracy and providing the basis to develop more
efficient numerical solutions.
For the analysis of magneto-electro-elastic laminates Pan [3] presented an exact 3-D solution for simply supported plates,
which was implemented by Pan and Heyliger [4] for the case of cylindrical bending. The same authors also proposed
approximate solutions for two and three-dimensional problems [5,6]. Wang and Shen [7] studied the general solution for
three-dimensional problems of transversely isotropic magneto-electro-elastic media by using potential functions. The state

⇑ Tel.: +39 9123896748.


E-mail address: alberto.milazzo@unipa.it

0307-904X/$ - see front matter Ó 2013 Elsevier Inc. All rights reserved.
http://dx.doi.org/10.1016/j.apm.2013.08.034
1738 A. Milazzo / Applied Mathematical Modelling 38 (2014) 1737–1752

space formulation was developed by Wang et al. [8]. More recently, Liu [9] proposed the employment of a 2-D plate theory
for homogeneous single-layer thin plates. Finite elements solutions have been presented by Lage et al. [10], Phoenix et al.
[11] and Carrera and Nali [12]; these are generally based on layerwise modelization. Also boundary elements approaches,
applicable to plate cylindrical bending, were proposed by Milazzo et al. [13] and Aimin et al. [14].
The literature survey evidences that the magneto-electro-elastic laminate solutions are prevalently obtained by 3-D ana-
lytical or finite elements layerwise-like theories. These assume the electro-magnetic state quantities as independent un-
knowns leading to increasing computational effort as the number of layer increases. This prompts for the development of
alternative multiphysics 2-D plate models, able to reduce the solution effort and providing efficient tools for the design
and optimization procedures. In this sense, from the mechanical point of view, equivalent single-layer theories for laminated
structures lead to plate models that have a solution complexity independent from the number of layers. Therefore, the exten-
sion of the equivalent single-layer theories to magneto-electro-elastic laminates, provided for a possible preliminary analytic
condensation of the electro-magnetic state to the mechanical variables, represents a sound underlying idea to develop an
efficient smart plate model. Additionally, such a modeling strategy could take advantage of the solution tools available
for the mechanics of multi-layered plates. Basing on these observations, the author and co-workers recently proposed the
use of equivalent single layer theories for magneto-electro-elastic laminated beams [15,16] and plates [17,18].
All of the above-cited approaches refer to the case of plate small displacements, whereas little attention was devoted to
the nonlinear large displacements case, and only very recently Xue et al. [19] presented a large deflection solution for homo-
geneous single-layer rectangular magneto-electro-elastic thin plates based on the von Karman’s plate assumptions. To the
best of the author’s knowledge, no closed form solutions for large deflection of multilayered magneto-electro-elastic lami-
nates have been proposed notwithstanding the large displacements influence can play important role in practical
applications.
Based on this consideration, in the present paper, an equivalent single-layer approach for the large deflection analysis of
multilayered magneto-electro-elastic laminates under static loads is proposed for the first time. It is based on the approach
previously presented by the author and co-workers for the small displacement case [17,18], which is here novelly formulated
for the large deflection case by using the von Karman’s plate assumptions and providing a closed form solution for simply-
supported magneto-electro-elastic laminated plates. In particular, the first order shear deformation theory is assumed to
model the plate deformation considering the non linear strain–displacement relations for large deflection of plates, whereas
the Maxwell equations for electrostatics and magnetostatics are used to model the electric and magnetic behavior. The plate
model is obtained by providing for a preliminary solution of the electro-magnetic state as a function of the mechanical vari-
ables. In turn, taking this result into account, the plate mechanical governing equations, namely equilibrium and compati-
bility equations, are written leading to a system of coupled differential equations whose solution determines the plate
mechanical response in terms of kinematical variables and stress function. Once the mechanical problem is solved the quan-
tities associated with the electro-magnetic behavior can be recovered by simple postprocessing. Some results for typical
magneto-electro-elastic simply-supported laminates are obtained by the Galerkin method and presented to show the effect
of the large deflection on the plate behavior.

2. Basic equations

Let us consider an N-layered, rectangular plate with side lengths Lx and Ly and thickness h. The plate is referred to a Carte-
sian coordinate system with the x and y coordinates spanning the midplane and the z-axis directed along the plate thickness.
The layers have constant thickness and are made of homogeneous and orthotropic magneto-electro-elastic materials with
electric and magnetic poling direction parallel to the z-axis and material principal directions parallel to the x- and y-axes.
Obviously, the case of piezoelectric, magnetoelastic or pure elastic layers is included. The faces of the kth layer are located

z
z

k-th layer H/2


Ly hk
hk-1
x
y y
Lx
H/2

Fig. 1. Laminate geometrical scheme.


A. Milazzo / Applied Mathematical Modelling 38 (2014) 1737–1752 1739

at z ¼ hk1 and z ¼ hk . The bottom and top surfaces of the plate are located at z ¼ h=2 and z ¼ h=2, respectively (see the
geometrical scheme of Fig. 1).

2.1. Basic assumptions

The following assumptions hold. The plate kinematics allows for rigid translations and rotations of the normals to the
midplane, which are considered inextensible along the thickness direction. These are the assumptions of the mechanical first
order shear deformation theory of plates (FSDT) [20]. The stress normal to the midplane is small when compared with the
other stress components and may be neglected [21]. From the electro-magnetic point of view, the effect of the in-plane com-
ponents of the electric field and magnetic induction is assumed negligible with respect to that of the corresponding compo-
nents along the thickness [9,19,17].

2.2. Kinematical model

Based on the stated assumptions the plate displacement components in the x-, y- and z-directions are given by
uðx; y; zÞ ¼ u0 ðx; yÞ þ z#x ðx; yÞ; ð1aÞ

v ðx; y; zÞ ¼ v 0 ðx; yÞ þ z#y ðx; yÞ; ð1bÞ

wðx; y; zÞ ¼ w0 ðx; yÞ; ð1cÞ


where u0 ; v 0 are the in-plane displacements at the mid-surface point, w0 is the transverse deflection and #x and #y are rota-
tions of the transverse normal around the y- and x-axis, respectively; all of them are functions of the mid-surface point
position.
Using the von Karman’s theory for large deflection of plates, the strain–displacement relations are written as [22]
 2
@u0 1 @w0 @#x
exx ¼ þ þz ¼ exx0 þ zjxx ; ð2aÞ
@x 2 @x @x
 2
@ v 0 1 @w0 @#y
eyy ¼ þ þz ¼ eyy0 þ zjyy ; ð2bÞ
@y 2 @y @y
 2
@w0 1 @w0
ezz ¼ þ ¼ 0; ð2cÞ
@z 2 @z

@w0 @w0 @w0 @w0


exz ¼ þ #x þ ¼ þ #x ; ð2dÞ
@x @x @z @x

@w0 @w0 @w0 @w0


eyz ¼ þ #y þ ¼ þ #y ; ð2eÞ
@y @y @z @y
 
@u0 @ v 0 @w0 @w0 @#x @#y
exy ¼ þ þ þz þ ¼ exy0 þ zjxy : ð2fÞ
@y @x @x @y @y @x
These strain components have to fulfill the compatibility equations [22], which for the present problem reduce to
!2
@ exx0 @ eyy0 @ exy0 @ 2 w0 @ 2 w0 @ 2 w0
þ  ¼  : ð3Þ
@y2 @x2 @x@y @x@y @x2 @y2

2.3. Magneto-electric state

For each layer the electro-magnetic state is described by introducing the electric potential Uhki ¼ Uhki ðx; y; zÞ and the scalar
magnetic potential Whki ¼ Whki ðx; y; zÞ. From here and in the following the superscript hki is employed to denote quantities
relevant to the kth layer. Extending to the magneto-electro-elastic case the widely employed assumption of zero in-plane
components of the electric field just formulated for the piezoelectric case, the in-plane electric field components, namely
Ex and Ey , and magnetic field components, namely Hx and Hy , are neglected and only the transverse electric field Ez and mag-
netic field Hz are considered in the present study. Thus, according to Maxwell theory, the gradient relations for the electric
and magnetic fields provide

@ Uhki
Ex ¼  ¼ 0; ð4aÞ
@x
1740 A. Milazzo / Applied Mathematical Modelling 38 (2014) 1737–1752

@ Uhki
Ey ¼  ¼ 0; ð4bÞ
@y

@ Uhki
Ez ¼  ; ð4cÞ
@z

@ Whki
Hx ¼  ¼ 0; ð5aÞ
@x

@ Whki
Hy ¼  ¼ 0; ð5bÞ
@y

@ Whki
Hz ¼  : ð5cÞ
@z
It is worth noting that the assumption of zero in-plane electric and magnetic field components is adequate for thin to mod-
erately thick plates, whereas for thick plates the contribution of these components becomes essential for accurate modeling
of the involved phenomena [3].

2.4. Magneto-electro-elastic constitutive law

By using Eqs. (2), (4) and (5) the constitutive law for each magneto-electro-elastic layer is written as [23]:
"  2 # "  2 #
hki
hki @u0 1 @w0 @#x hki @ v 0 1 @w0 @#y @ Uhki hki @ W
rxx ¼ C 11 þ þz þ C 12 þ þz þ ehki
31 þ q31 ; ð6aÞ
@x 2 @x @x @y 2 @y @y @z @z

"  2 # "  2 #
hki hki
hki @u0 1 @w0 @#x hki @ v 0 1 @w0 @#y hki @ U hki @ W
ryy ¼ C 12 þ þz þ C 22 þ þz þ e32 þ q32 ; ð6bÞ
@x 2 @x @x @y 2 @y @y @z @z

"  2 # "  2 #
hki
hki @u0 1 @w0 @#x hki @ v 0 1 @w0 @#y @ Uhki hki @ W
rzz ¼ C 13 þ þz þ C 23 þ þz þ ehki
33 þ q33 ; ð6cÞ
@x 2 @x @x @y 2 @y @y @z @z

 
hki @w0
rxz ¼ C 44 þ #x ; ð6dÞ
@x
 
hki @w0
ryz ¼ C 55 þ #y ; ð6eÞ
@y
   
hki @u0 @ v 0 @w0 @w0 hki @#x @#y
rxy ¼ C 66 þ þ þ C 66 z þ ; ð6fÞ
@y @x @x @y @y @x
 
hki @w0
Dx ¼ e15 þ #x ; ð6gÞ
@x
 
hki @w0
Dy ¼ e24 þ #y ; ð6hÞ
@y
"  2 # "  2 #
hki hki
hki @u0 1 @w0 @#x hki @ v 0 1 @w0 @#y hki @ U hki @ W
Dz ¼ e31 þ þz þ e32 þ þz  33  d33 ; ð6iÞ
@x 2 @x @x @y 2 @y @y @z @z

 
hki @w0
Bx ¼ q15 þ #x ; ð6jÞ
@x
 
hki @w0
By ¼ q15 þ #y ; ð6kÞ
@y
"  2 # "  2 #
hki hki
hki @u0 1 @w0 @#x hki @ v 0 1 @w0 @#y hki @ U hki @ W
Bz ¼ q31 þ þz þ q32 þ þz  d33  l33 ; ð6lÞ
@x 2 @x @x @y 2 @y @y @z @z
A. Milazzo / Applied Mathematical Modelling 38 (2014) 1737–1752 1741

where C ij are the elastic stiffness coefficients, ij and lij are the dielectric constants and magnetic permeability, respectively,
eij are the piezoelectric constants, qij are the piezomagnetic coupling coefficients and dij denotes the magnetoelectric
coupling coefficients.

2.5. Governing equations

Introducing the following definitions of the plate force and moment resultants per unit length:
Z h
2
Nab ¼ rab dz for a; b ¼ x; y; ð7aÞ
h2

Z h
2
M ab ¼ z rab dz for a; b ¼ x; y; ð7bÞ
2h

Z h
2
Q az ¼ raz dz for a ¼ x; y ð7cÞ
2h

the mechanical state of the plate is governed by the equilibrium equations, which can be deduced from the principle of
virtual displacements and read as [20]:
@Nxx @Nxy
þ ¼ 0; ð8aÞ
@x @y

@Nxy @Nyy
þ ¼ 0; ð8bÞ
@x @y
   
@Q xz @Q yz @ @w0 @w0 @ @w0 @w0
þ þ N xx þ Nxy þ Nxy þ Nyy þ q ¼ 0; ð8cÞ
@x @y @x @x @y @y @x @y

@M xx @Mxy
þ  Q xz ¼ 0; ð8dÞ
@x @y

@M xy @Myy
þ  Q yz ¼ 0; ð8eÞ
@x @y
where q ¼ qðx; yÞ denote the applied transverse load per unit area.
The electro-magnetic state of the plate layers is governed by the Gauss’ laws for electrostatics and magnetostatics, which
write as:
hki
@Dxhki @Dy @Dhki
þ þ z ¼ 0; ð9aÞ
@x @y @z

hki
@Bxhki @By @Bhki
þ þ z ¼ 0: ð9bÞ
@x @y @z
As no a priori distribution is assumed along the plate thickness for the electric and magnetic potentials, for multilayered
plates the Eq. (9), that hold for each single layer, have to be supplemented by the electro-magnetic interface continuity
conditions. At the interface located at z ¼ hk , these specify as:

Uhki ðx; y; hk Þ ¼ Uhkþ1i ðx; y; hk Þ; ð10aÞ

Dzhki ðx; y; hk Þ ¼ Dzhkþ1i ðx; y; hk Þ; ð10bÞ

Whki ðx; y; hk Þ ¼ Whkþ1i ðx; y; hk Þ; ð10cÞ

Bzhki ðx; y; hk Þ ¼ Bzhkþ1i ðx; y; hk Þ: ð10dÞ


Finally, the essential and natural boundary conditions associated with the mechanical equilibrium equations are those of the
first order shear plate theory [20], which are not reported here for the sake of brevity. The electro-magnetic boundary con-
ditions are applied at the bottom and top surface of the plate. They are represented by (i) prescribed values of either the
electric potential or normal electric displacements and (ii) prescribed values of either the magnetic potential or normal mag-
netic induction. They can be thought as the plate electro-magnetic loading and read as:
1742 A. Milazzo / Applied Mathematical Modelling 38 (2014) 1737–1752

Uðx; y; h=2Þ ¼ n1 ðx; yÞ or Dz ðx; y; h=2Þ ¼ n1 ðx; yÞ; ð11aÞ

Wðx; y; h=2Þ ¼ n2 ðx; yÞ or Bz ðx; y; h=2Þ ¼ n2 ðx; yÞ; ð11bÞ

Uðx; y; h=2Þ ¼ n3 ðx; yÞ or Dz ðx; y; h=2Þ ¼ n3 ðx; yÞ; ð11cÞ

Wðx; y; h=2Þ ¼ n4 ðx; yÞ or Bz ðx; y; h=2Þ ¼ n4 ðx; yÞ; ð11dÞ


where ni ðx; yÞ denotes prescribed quantities.

3. Formulation

The formulation of the model is based on the following scheme. At first the electro-magnetic state is determined as a
function of the mechanical state. In turn, this result is used to express the plate forces and moments to be substituted
into the governing equilibrium equations, which result in a system of coupled partial differential equations involving
mechanical quantities only. Once the solution for the plate mechanical state variables is accomplished, the electro-mag-
netic state can be determined by simple recovery procedures. The outlined approach has been presented by the author in
[17] and it is here extended and fit to the solution of plate large deflection. Some of the derivation presented in the
following are similar to those reported in [17] and they are also given here, in a concise form, to make the paper
self-contained.

3.1. Electro-magnetic state

Substituting Eq. (6) into the Gauss’ equations for electrostatics and magnetostatics, namely Eqs. (9a) and (9b), leads to the
following system of equations holding for each kth layer:

@ 2 Uhki 2 hki     2 2
hki hki @ W hki @#x hki @#y hki @ w0 hki @ w0
33 2
þ d33 2
¼ ehki
31 þ e15
hki
þ e32 þ e24 þ e15 2
þ e24 ; ð12aÞ
@z @z @x @y @x @y2

hki @ 2 Uhki @ 2 Whki  hki  @#


x
 
hki @#y
2
hki @ w0 @ 2 w0
d33 þ lhki
33 ¼ q31 þ qhki
15 þ qhki
32 þ q24 þ q15 þ qhki
24 : ð12bÞ
@z2 @z2 @x @y @x2 @y2
Integration of the Eq. (12) provides the expressions of the electric and magnetic potentials inside the kth layer
!
@#x @#y @ 2 w0 2
hki @ w0 z2 hki
Uhki ¼ A1hkiU þ A2hkiU þ Bhki
1U 2
þ B2U 2
þ ahki
U z þ bU ; ð13aÞ
@x @y @x @y 2

!
2 2
@#x hki @#y hki @ w0 hki @ w0 z2 hki
W hki
¼ A1hkiW þ A2W þ B1W þ B2W þ ahki
W z þ bW ; ð13bÞ
@x @y @x2 @y2 2
hki hki
where the coefficients Ahki hki hki hki
ab and Bab are defined in Appendix A, whereas aU ; bU ; aW and bW are integration constants. These
integration constants are determined by enforcing the electro-magnetic interface continuity conditions and the electro-mag-
netic boundary conditions at the bottom and top surfaces of the laminate. Actually, taking the constitutive law and the elec-
tric and magnetic potential expressions into account, the Eq. (10) written for the k ¼ 1; 2; . . . ; N  1 interfaces and the four
Eq. (11) provide for a linear system whose right-hand-side consists of a linear combinations of kinematical quantities and
electro-magnetic boundary conditions. The lengthy explicit expression of the involved equations are not given for the sake
of brevity and their form can be found in [17]. According with the form of the right-hand-side of the above mentioned sys-
tem, for the integration constants one can writes:

@ 2 w0 @ 2 w0 X 4
fhki hki hki hki hki hki
k ¼ fkm1 exx0 þ fkm2 eyy0 þ fkb1 jxx þ fkb2 jyy þ fkw1 2
þ fkhki 2
þ fkhki n;
ni i
ð14Þ
@x w2
@y i¼1

where the symbol f stands for a or b and the symbol k stands for U or W. It is worth noting that not all of the possible com-
bination of electro-magnetic boundary conditions are admissible in the present formulation. In particular, the open circuit
boundary conditions, that is zero electric displacement and magnetic induction on both the bottom and top laminate
hki hki
surfaces, lead to an overdetermined system for the integration constants ahki hki
U ; bU ; aW and bW , which admits only the trivial
solution of flexural displacements approaching to zero [24,17].
A. Milazzo / Applied Mathematical Modelling 38 (2014) 1737–1752 1743

3.2. Smart laminate constitutive equations

To take advantage from the compact matrix notation, let us introduce the in-plane resultant force per unit length vector
T  T
Q xz Q yz , the resultant moments per unit length
N ¼ f N xx N yy N xy g , the shear resultant force per unit length vector Q ¼
T
vector M ¼ f M xx M yy M xy g , the plate mid-plane strain vector e0 ¼ f exx0 eyy0 exy0 gT , the shear strain vector c ¼ f exz eyz gT ,
T
the plate curvatures vector j ¼ f jxx jyy jxy g , the plate transverse displacements second derivatives vector
n 2 2 2
oT
v ¼ @@xw20 @@yw20 @@x@y
w0
and the electro-magnetic loading vector n ¼ f n1 n2 n3 n4 gT . After the electro-magnetic state has been
expressed as a function of the plate kinematics and electro-magnetic loading (see Eq. 14), the smart laminate forces and mo-
ment resultants per unit length are obtained from Eq. (7) where Eq. (13) is substituted. One obtains:

N ¼ A e0 þ B j þ C v þ V n; ð15aÞ

M ¼ E e0 þ D j þ G v þ W n; ð15bÞ

Q ¼ Hc; ð15cÞ
where the expressions for the non-zero matrix coefficients are given in Appendix B. The Eq. (15) represent the smart lam-
inate equivalent single-layer mechanical constitutive law. They shows the same formal dependencies between the stress
resultant and the kinematical primary variables as the pure elastic case [20]. However, they involve magneto-electro-elastic
effective stiffness coefficients, namely A; B; E; D, which evidence two contributions: the first one depends on the mechanical
properties only and it coincides with the corresponding stiffness coefficient of the pure elastic theory, the second one de-
pends on the layers electro-magnetic properties and represents the effect of the electro-magnetic behavior affecting the
plate stiffness. Additionally, new terms exclusively associated with the piezoelectric and piezomagnetic behavior come
out, namely C and G. The electro-magnetic loading applied on the bottom and top surfaces of the plate contributes to the
stress resultants by the coefficients V and W.
To infer the smart plate resolving equations, according with the Von Karman large displacement plate theory, a stress
function F ¼ Fðx; yÞ is defined such as:

8 @2 F
9
>
> @y2 >
>
< =
N¼ @2 F ¼ ›2 F: ð16Þ
> @x2 >
>
: >
@2 F ;
 @x@y

Thus, the Eqs. (15a) and (15b) are rearranged as follows:

e0 ¼ f ejC
A ›2 F  B evV
e n; ð17aÞ

e ›2 F þ D
M¼E evþW
e jþG f n; ð17bÞ
where

f
A ¼ A 1 ; ð18aÞ

e ¼f
B A B; ð18bÞ

e¼f
C A C; ð18cÞ

e ¼f
V A V; ð18dÞ

e ¼ Ef
E A; ð18eÞ

D e
e ¼ D  E B; ð18fÞ

e ¼ G  E C;
G e ð18gÞ

f ¼ W  EV
W e: ð18hÞ
1744 A. Milazzo / Applied Mathematical Modelling 38 (2014) 1737–1752

3.3. Resolving equations

Due to the definition of the stress function F the plate in-plane equilibrium equations, namely Eqs. (8a) and (8b), are
trivially satisfied and their dual counterpart is the congruence equation, Eq. (3), which upon substitution of Eq. (17a) leads
to

4   @4F 4 3   3 3
e 22 @ F þ A
A e 33 þ A
e 12 þ A
e 21 e 11 @ F ¼ B
þA e 21 @ #x þ B e 11  Be 33 @ #x þ B e 12 @ #y
@x4 @x2 @y2 @y4 @x3 @x@y2 @y3
  3 @ 4 w0  e  @4w
þ B e 33 @ #y þ e
e 22  B C 21 þ C 11 þ e C 22
0
þeC 12
@x2 @y @x4 @x2 @y2
!2 !
@ 4 w0 @ 2 w0 @ 2 w0 @ 2 w0 X 4
e @ 2 ni e @ 2 ni
 þ  þ V 1i 2 þ V 2i 2 ; ð19Þ
@y4 @x@y @x2 @y2 i¼1
@y @x

where the notation Z ij indicates the elements of the matrix Z. Substitution of Eqs. (17b) and (15c), into the last three plate
equilibrium equations, namely Eqs. (8c), (8d) and (8e) provides:

@#x @#y @ 2 w0 @ 2 w0 @ 2 F @ 2 w0 @ 2 F @ 2 w0 @ 2 F @ 2 w0
H11 þ H22 þ H11 2
þ H22 2
þ 2 2
2 þ þ q ¼ 0; ð20aÞ
@x @y @x @y @x @y @x@y @x@y @y2 @x2

3   3 2   2 2 3 3
e 12 @ F þ E
E e 33 @ F þ D
e 11  E e 11 @ #x þ D e 33 @ #y þ D
e 12 þ D e 11 @ w0 þ G
e 33 @ #x þ G e 12 @ w0
@x 3 @x@y 2 @x 2 @x@y @y 2 @x 3 @x@y2
@w0 X f @ni
4
 H11 #x  H11 þ W 1i ¼ 0; ð20bÞ
@x i¼1
@x

3   3 2   2 2 3 3
e 21 @ F þ E
E e 33 @ F þ D
e 22  E e 22 @ #y þ D e 33 @ #x þ D
e 21 þ D e 22 @ w0 þ G
e 33 @ #y þ G e 21 @ w0
@y3 @x2 @y @y2 @x@y @x2 @y3 @x2 @y
@w0 X 4
 H22 #y  H22 þ f 2i @ni ¼ 0:
W ð20cÞ
@y i¼1
@y

In conclusion, the single-layer equivalent model for the smart laminate is given by Eqs. (19), (20a), (20b) and (20c), which
coupled with the plate mechanical boundary conditions allows to solve for the plate primary mechanical unknowns F; w0 ; #x
and #y . Once these are determined the other mechanical variables can be directly derived, whereas the electro-magnetic
state is recovered through Eq. (13).

4. Simply-supported plates

Consider a simply-supported magneto-electro-elastic multi-layered plate subjected to transverse mechanical load and
electro-magnetic loadings expressed by:

qðx; yÞ ¼ qmn sinðaxÞ sinðbyÞ; ð21Þ

ni ðx; yÞ ¼ Nimn sinðaxÞ sinðbyÞ; ð22Þ


where a ¼ mp=Lx and b ¼ np=Ly being m and n integers.

4.1. Kinematical variables

The plate kinematical variables of the solution are expressed as:

w0 ðx; yÞ ¼ W mn sinðaxÞ sinðbyÞ; ð23aÞ

#x ðx; yÞ ¼ Hxmn cosðaxÞ sinðbyÞ; ð23bÞ

#y ðx; yÞ ¼ Hy mn sinðaxÞ cosðbyÞ; ð23cÞ

4.2. Stress function and resolving equation

Upon substitution of Eq. (23) into Eq. (19) one deduces the corresponding expression of the stress function F, which reads
as:
A. Milazzo / Applied Mathematical Modelling 38 (2014) 1737–1752 1745


F ¼ K 1 cosð2axÞ þ K 2 cosð2byÞ þ K 3 x2 þ K 4 y2 W 2mn þ ½K 5 sinðaxÞ sinðbyÞW mn þ ½K 6 sinðaxÞ sinðbyÞHxmn
X
4
þ ½K 7 sinðaxÞ sinðbyÞHy mn þ ½K 7þi sinðaxÞ sinðbyÞNimn ; ð24Þ
i¼1

where the coefficients K i are given in Appendix C for the two conditions of immovable and movable edges. Substituting this
result and the kinematical variables expressions into Eq. (20) the following equations are obtained:

4a2 b2 W 3mn ½K 1 cosð2axÞ þ K 2 cosð2byÞ sinðaxÞ sinðbyÞ


  
 H11 aHxmn þ H22 bHy mn  ðH11 þ K 4 Þa2  ðH22 þ K 3 Þb2 W mn sinðaxÞ sinðbyÞþ
!
X 4  
2 2
þ 2a2 b2 K 5 W mn þ K 6 Hxmn þ K 7 Hy mn þ K 7þi Ni W mn sin ðaxÞ sin ðbyÞ  cos2 ðaxÞ cos2 ðbyÞ
i¼1

þ qmn sinðaxÞ sinðbyÞ ¼ 0; ð25aÞ

h   
e 12 a3 W 2 sinð2axÞ þ
8K 1 E e  G
K5 E e 11 a3  G
e 12 ab2  H11 a W mn þ K 6 E e  D
e 11 a2  D
e 33 b2  H11 Hxmn
mn 1 1
#
 h i  X4  
e  e e
þ K 7 E 1  D 12 þ D 33 ab Hy mn þ f e 
W 1i a þ K 7þi E 1 Ni cosðaxÞ sinðbyÞ ¼ 0; ð25bÞ
i¼1

h   
e 21 b3 W 2 sinð2byÞ þ
8K 2 E e  G
K5 E e 21 a2 b  G
e 22 b3  H22 b W mn Þ þ K 6 E e  De 21 ab  D
e 33 ab Hxmn
mn 2 2
#
  X 4  
e  D
þ K7 E e 22 b2  D
e 33 a2  H22 Hy þ W e  Ni sinðaxÞ cosðbyÞ ¼ 0;
f 2i b þ K 7þi E ð25cÞ
2 mn 2
i¼1

where
 
e ¼ E
E e 33  E
e 11 ab2  E
e 12 a3 ; ð26aÞ
1

 
e ¼ E
E e 33  E
e 22 a2 b  E
e 21 b3 : ð26bÞ
2

4.3. Solution

The problem is solved by determining the coefficients W mn ; Hxmn and Hy mn through the Bubnov–Galerkin method. Both
sides of Eqs. (25a), (25b) and (25c) are multiplied by the corresponding weighting functions sinðaxÞ sinðbyÞ, cosðaxÞ sinðbyÞ
and sinðaxÞ cosðbyÞ, respectively, and integrated over the whole plate area leading to the following non linear resolving alge-
braic system
8
3 2
>
< c11 W mn þ c12 W mn þ c13 W mn þ c14 W mn Hxmn þ c15 W mn Hy mn þ c16 Hxmn þ c17 Hy mn ¼ r 1
>
c21 W 2mn þ c22 W mn þ c23 Hxmn þ c24 Hy mn ¼ r 2 ð27Þ
>
>
:
c31 W 2mn þ c32 W mn þ c33 Hxmn þ c34 Hy mn ¼ r 3

whose coefficients expressions are given in Appendix D. Simple manipulations of Eq. (27) lead to a third order algebraic
equation whose real solution provides the value of W mn that can be used to calculate Hxmn and Hxmn from the second and
third equation of the system (27). Thus the closed form solution for the large deflection of simply-supported smart laminates
was determined.

5. Numerical results and discussion

In this section, some examples for magneto-electro-elastic laminated, simply-supported with immovable edges rectan-
gular plates subjected to an external transverse load q ¼ q sinðpx=Lx Þ sinðpy=Ly Þ and different electro-magnetic conditions
are presented. In particular, according to the most investigated configurations found in the literature (e.g. [3,10]), results
for three-layers plates consisting of equal thickness laminas of piezoelectric barium titanate BaTiO3 and piezomagnetic co-
balt ferrite CoFe2 O4 are presented. The employed material constants are those given in [3]. Taking B labels for BaTiO3 and F
labels for CoFe2 O4 , the investigated lay-ups are the two configurations B=F=B and F=B=F. Two distinct electro-magnetic
boundary conditions are considered: (i) the closed circuit one, which is referred to as EM-I, when the electric and magnetic
potentials are set to zero on both the top and bottom plate surfaces, and (ii) a mixed surface conditions, that is called EM-II,
when the electric and magnetic potentials are set to zero on the bottom surface, whereas zero electric normal displacement
1746 A. Milazzo / Applied Mathematical Modelling 38 (2014) 1737–1752

Table 1
Deflection w0 ðLx =2; Ly =2Þ=h at different loads q for an Lx =Ly ¼ 1:0 plate with different thickness ratio Ly =h. Comparison between linear and large deflection
theories.

=C 11;max
q B/F/B Laminate F/B/F Laminate

Ly =h ¼ 20 Ly =h ¼ 15 Ly =h ¼ 10 Ly =h ¼ 20 Ly =h ¼ 15 Ly =h ¼ 10

Nonlin. Lin. Nonlin. Lin. Nonlin. Lin. Nonlin. Lin. Nonlin. Lin. Nonlin. Lin.
0.1 3.455 3.520 1.953 1.974 0.865 0.869 3.198 3.249 1.804 1.821 0.798 0.801
0.2 6.589 7.041 3.794 3.947 1.707 1.738 6.135 6.498 3.520 3.641 1.577 1.602
0.3 9.294 10.561 5.464 5.921 2.509 2.607 8.707 9.748 5.093 5.462 2.324 2.402
0.4 11.611 14.081 6.954 7.894 3.260 3.476 10.935 12.997 6.512 7.283 3.030 3.203
0.5 13.618 17.602 8.280 9.868 3.959 4.345 12.879 16.246 7.786 9.103 3.691 4.004
0.6 15.381 21.122 9.468 11.841 4.606 5.214 14.595 19.495 8.934 10.924 4.308 4.805
0.7 16.954 24.643 10.541 13.815 5.205 6.083 16.129 22.745 9.974 12.745 4.882 5.606
0.8 18.375 28.163 11.517 15.788 5.761 6.952 17.518 25.994 10.925 14.565 5.418 6.406
0.9 19.672 31.683 12.414 17.762 6.279 7.821 18.788 29.243 11.800 16.386 5.918 7.207
1.0 20.867 35.204 13.243 19.735 6.763 8.691 19.959 32.492 12.609 18.207 6.388 8.008

Table 2
Deflection w0 ðLx =2; Ly =2Þ=h at different loads q for an Ly =h ¼ 15 plate with different side length ratio Lx =Ly . Comparison between linear and large deflection
theories.

=C 11;max
q B/F/B Laminate F/B/F Laminate
Lx =Ly ¼ 0:5 Lx =Ly ¼ 1:0 Lx =Ly ¼ 2:0 Lx =Ly ¼ 0:5 Lx =Ly ¼ 1:0 Lx =Ly ¼ 2:0

Nonlin. Lin. Nonlin. Lin. Nonlin. Lin. Nonlin. Lin. Nonlin. Lin. Nonlin. Lin.
0.1 0.271 0.271 1.953 1.974 4.108 4.334 0.250 0.250 1.804 1.821 3.817 3.999
0.2 0.541 0.542 3.794 3.947 7.365 8.667 0.500 0.500 3.520 3.641 6.919 7.997
0.3 0.811 0.813 5.464 5.921 9.868 13.001 0.749 0.750 5.093 5.462 9.341 11.996
0.4 1.080 1.084 6.954 7.894 11.874 17.335 0.997 1.000 6.512 7.283 11.298 15.995
0.5 1.347 1.355 8.280 9.868 13.552 21.668 1.244 1.250 7.786 9.103 12.938 19.993
0.6 1.612 1.626 9.468 11.841 14.999 26.002 1.489 1.500 8.934 10.924 14.355 23.992
0.7 1.875 1.896 10.541 13.815 16.276 30.335 1.733 1.750 9.974 12.745 15.607 27.991
0.8 2.136 2.167 11.517 15.788 17.423 34.669 1.975 2.000 10.925 14.565 16.732 31.989
0.9 2.394 2.438 12.414 17.762 18.467 39.003 2.215 2.250 11.800 16.386 17.755 35.988
1.0 2.649 2.709 13.243 19.735 19.427 43.336 2.452 2.500 12.609 18.207 18.697 39.986

Table 3
Deflection w0 Lx =2; Ly =2 =h at different loads q for an Lx =Ly ¼ 1:0 plate with different thickness ratio Ly =h and electro-magnetic boundary conditions (EM).

=C 11;max
q B/F/B Laminate F/B/F Laminate

Ly =h ¼ 20 Ly =h ¼ 15 Ly =h ¼ 10 Ly =h ¼ 20 Ly =h ¼ 15 Ly =h ¼ 10

EM I EM II EM I EM II EM I EM II EM I EM II EM I EM II EM I EM II
0.1 3.455 3.452 1.953 1.951 0.865 0.864 3.198 3.197 1.804 1.804 0.798 0.798
0.2 6.589 6.584 3.794 3.791 1.707 1.706 6.135 6.134 3.520 3.519 1.577 1.577
0.3 9.294 9.288 5.464 5.460 2.509 2.507 8.707 8.706 5.093 5.092 2.324 2.324
0.4 11.611 11.604 6.954 6.949 3.260 3.258 10.935 10.934 6.512 6.511 3.030 3.030
0.5 13.618 13.610 8.280 8.275 3.959 3.956 12.879 12.877 7.786 7.785 3.691 3.691
0.6 15.381 15.373 9.468 9.463 4.606 4.603 14.595 14.593 8.934 8.933 4.308 4.307
0.7 16.954 16.946 10.541 10.535 5.205 5.202 16.129 16.128 9.974 9.973 4.882 4.882
0.8 18.375 18.367 11.517 11.512 5.761 5.758 17.518 17.517 10.925 10.924 5.418 5.417
0.9 19.672 19.664 12.414 12.408 6.279 6.276 18.788 18.786 11.800 11.798 5.918 5.918
1.0 20.867 20.860 13.243 13.237 6.763 6.760 19.959 19.957 12.609 12.608 6.388 6.387

and magnetic normal induction components are set to zero on the plate top surface. Different side length ratios Lx =Ly and
thickness ratio Ly =h have been considered. In all of the applications the reference side length was chosen as Ly ¼ 1m.
Table 1 lists the central point deflection for the B=F=B and F=B=F plates with fixed side length ratio Lx =Ly ¼ 1:0 and dif-
ferent thickness ratio Ly =h. The results are given in terms of non-dimensional deflection w0 =h for different non-dimensional
transverse load amplitude obtained by dividing the load amplitude q  by the maximum value of C 11 in the plate. The
comparison with the corresponding linear solution shows that, as expected, the non linear large deflection effect is mean-
ingful for thin plates and becomes less important as the thickness plate increase. The effect of the in-plane plate geometry,
described by the side length ratio Lx =Ly , is investigated. In Table 2, the central point deflection for the B=F=B and F=B=F plates
A. Milazzo / Applied Mathematical Modelling 38 (2014) 1737–1752 1747

Fig. 2. Through-the-thickness distribution of the electric potential for the B=F=B laminate. Continuous line and open circles represent the linear and
nonlinear solution, respectively.

Fig. 3. Through-the-thickness distribution of the magnetic potential for the B=F=B laminate. Continuous line and open circles represent the linear and
nonlinear solution, respectively.

with fixed thickness ratio Ly =h ¼ 15 and different side length ratio is listed for different non-dimensional transverse load
amplitude. It is observed that for fixed load the nonlinear effect influence increases as Lx =Ly increases. The effect of the elec-
tro-magnetic boundary conditions is shown in Table 3, which reports the central point deflection for the B=F=B and F=B=F
plates with fixed lateral geometry Lx =Ly ¼ 1:0 and varying thickness ratio Ly =h. It is evidenced that the electro-magnetic
1748 A. Milazzo / Applied Mathematical Modelling 38 (2014) 1737–1752

Fig. 4. Through-the-thickness distribution of the electric potential for the F=B=F laminate. Continuous line and open circles represent the linear and
nonlinear solution, respectively.

Fig. 5. Through-the-thickness distribution of the magnetic potential for the F=B=F laminate. Continuous line and open circles represent the linear and
nonlinear solution, respectively.

boundary conditions slightly influences the amplitude of the plate response and do not modify the behavior trends with
respect to the large deflection effects, which is actually a mechanical driven effect.
Finally the through-the-thickness distributions of the electric potential and magnetic potential at the plate center for dif-
ferent thickness ratios Ly =h are plotted in Figs. 2 and 3 for the B=F=B laminate and in Figs. 4 and 5 for the F=B=F laminate,
A. Milazzo / Applied Mathematical Modelling 38 (2014) 1737–1752 1749

respectively. The plotted quantities are normalized with respect to their maximum obtained in the linear (small deflection)
case, namely U0 and W0 . Once again it is observed that the effect of the large deflection is important for thin plates whereas
as the plate thickness increases very high load amplitude are needed to appreciate this nonlinear effect.

6. Conclusions

A non linear large deflection model for magneto-electro-elastic laminated plates has been proposed. The model identifies
magneto-electro-elastic laminate effective stiffness coefficients, which take the electro-magnetic influence into account and
allow to obtain a resolving system in terms of mechanical variables only. Once these are determined the electro-magnetic
state can be recovered. A closed form solution for simply-supported plates has been derived and used to present some
numerical results. As expected, it is found that the effect of non linear large deflection depends on the plate thickness
and can meaningfully affect the prediction of the plate deflection and electro-magnetic response. Similar trends of the plate
behavior have been obtained for different electro-magnetic boundary conditions, which however slightly influence the vari-
ables amplitude. In the present paper, the proposed model has been solved for the case of bi-sinusoidal loadings however the
Bubnov–Galerkin solution procedure can be adapted also to other loading configurations. Finally, it is noticed that the pro-
posed model could be used to formulate finite elements, which present the same features of those employed for the elastic
analysis provided that the effective magneto-electro-elastic stiffnesses are considered.

Appendix A. Electric and magnetic potential coefficients

   
hki hki
e31 þ ehki hki hki hki
15 l33  q31 þ q15 d33
Ahki
1U ¼ 2
; ðA:1aÞ
33 l33  dhki
hki hki
33

   
hki hki
e32 þ ehki hki hki hki
24 l33  q32 þ q24 d33
Ahki
2U ¼ 2
; ðA:1bÞ
33 l33  dhki
hki hki
33

hki hki hki


e15 l33  qhki
15 d33
B1hkiU ¼ 2
; ðA:1cÞ
33 l33  dhki
hki hki
33

hki hki hki


e24 l33  qhki
24 d33
B2hkiU ¼ ; ðA:1dÞ
hki hki hki 2
33 l33  d33
   
hki hki hki hki hki hki
q31 þ q15 33  e31 þ e15 d33
hki
A1W ¼ 2
; ðA:2aÞ
33 l33  dhki
hki hki
33

   
hki hki hki
q32 þ q24 hki hki hki
33  e32 þ e24 d33
Ahki
2W ¼ 2
; ðA:2bÞ
33 l33  dhki
hki hki
33

hki
qhki hki hki
15 33  e15 d33
B1hkiW ¼ 2
; ðA:2cÞ
33 l33  dhki
hki hki
33

hki
qhki hki hki
24 33  e24 d33
B2hkiW ¼ 2
: ðA:2dÞ
33 l33  dhki
hki hki
33

Appendix B. Smart laminate stiffness coefficients

N h
X   i
hki hki hki hki hki hki hki
Aij ¼ C ij I 0 þ e3i aUmj þ q3i aWmj I 0 for i; j ¼ 1; 2; ðB:1aÞ
k¼1
1750 A. Milazzo / Applied Mathematical Modelling 38 (2014) 1737–1752

X
N
hki hki
A33 ¼ C 66 I0 ; ðB:1bÞ
k¼1

N n
X h    io
Bij ¼ C ijhki I 1hki þ hki hki
e3i hki hki
AUj þ q3i AWj I hki hki hki hki hki hki
1 þ e3i aUbj þ q3i aWbj I 0 for i; j ¼ 1; 2; ðB:2aÞ
k¼1

X
N
hki hki
B33 ¼ C 66 I 1 ; ðB:2bÞ
k¼1

N h
X  i h  i
hki hki hki hki
Cij ¼ e3i BUj þ q3i BWj I 1hki þ e3i
hki hki hki hki
aUwj þ q3i aWwj I 0hki for i; j ¼ 1; 2; ðB:3Þ
k¼1

N 
X 
hki hki hki hki
V ij ¼ e3i aUn þ q3i aWn I 0hki for i ¼ 1; 2 and for j ¼ 1; 2; 3; 4; ðB:4aÞ
j j
k¼1

N h
X   i
E ij ¼ C ijhki I 1hki þ ehki hki hki hki hki
3i aUmj þ q3i aWmj I 1 for i; j ¼ 1; 2; ðB:5aÞ
k¼1

X
N
hki hki
E 33 ¼ C 66 I 1 ; ðB:5bÞ
k¼1

N n
X h    io
Dij ¼ ðC ijhki I hki
2 þ
hki hki
e3i hki hki
AUj þ q3i AWj I hki hki hki hki hki hki
2 þ e3i aUbj þ q3i aWbj I 1 for i; j ¼ 1; 2; ðB:6aÞ
k¼1

X
N
hki hki
D33 ¼ C 66 I2 ; ðB:6bÞ
k¼1

N h
X    i
hki hki hki hki
Gij ¼ e3i BUj þ q3i BWj I hki hki hki hki hki hki
2 þ e3i aUwj þ q3i aUwj I 1 for i; j ¼ 1; 2; ðB:7Þ
k¼1

N 
X 
hki hki hki hki hki
W ij ¼ e3i aUn þ q3i aWn I 1 for i ¼ 1; 2 and for j ¼ 1; 2; 3; 4; ðB:8aÞ
j j
k¼1

X
N
hki hki
Hii ¼ C jj I 0 for i ¼ 1; 2 and j ¼ i þ 3; ðB:9Þ
k¼1

where
Z hk
hki
Im ¼ zm dz: ðB:10Þ
hk1

Appendix C. Stress function coefficients for simply-supported plate

1 1 b2
K1 ¼ ; ðC:1Þ
e 22 a2
32 A

1 1 a2
K2 ¼ ; ðC:2Þ
e 11 b2
32 A
8
<2eA 11 e
A 22 a2 K 1  e
A 21 e
A 11 b2 K 2
for immovable edges;
K3 ¼ eA 11 e
A 22  eA 12 e
A 21 ðC:3Þ
:0 for movable edges;
A. Milazzo / Applied Mathematical Modelling 38 (2014) 1737–1752 1751

8
< 2 e
A 12 e
A 22 a2 K 1  e
A 11 e
A 22 b2 K 2
for immovable edges;
K4 ¼ eA 11 e
A 22  eA 12 e
A 21 ðC:4Þ
:0 for movable edges;
 
e
C 21 a4 þ e C 11 þ eC 22 a2 b2 þ e
C 12 b4
K5 ¼   ; ðC:5Þ
e 22 a4 þ A
A e 33 þ A e 12 þ Ae 21 a2 b2 þ A e 11 b4

 
e 21 a3 þ B
B e 11  B
e 33 ab2
K6 ¼   ; ðC:6Þ
e 22 a4 þ A
A e 33 þ A
e 12 þ Ae 21 a2 b2 þ A
e 11 b4

 
e 12 b3 þ B
B e 33 a2 b
e 22  A
K7 ¼   ; ðC:7Þ
e 22 a4 þ A
A e 33 þ A
e 12 þ Ae 21 a2 b2 þ A
e 11 b4

e 1i b2 þ V
V e 2i a2
K 7þi ¼    for i ¼ 1; 2; 3; 4: ðC:8Þ
e 22 a4 þ A
A e 33 þ A e 12 þ Ae 21 a2 b2 þ A
e 11 b4

Appendix D. Non-linear resolving system coefficients for simply-supported plate

Lx Ly
c11 ¼  ðK 1 þ K 2 Þ; ðD:1Þ
8
(
8
abK 5 if m; n odd;
c12 ¼ 3 ðD:2Þ
0 otherwise;
8
>
>  X
4
< ðH11 þ K 3 Þa2 þ ðH22 þ K 3 Þb2 Lx Ly
þ 83 ab K 7þi Ni if m; n odd;
4
c13 ¼ i¼1 ðD:3Þ
>
:
> L L
2
ðH11 þ K 3 Þa þ ðH22 þ K 3 Þb2 x4 y otherwise;
(
8
3
abK 6 if m; n odd;
c14 ¼ ðD:4Þ
0 otherwise;
(
8
3
abK 7 if m; n odd;
c15 ¼ ðD:5Þ
0 otherwise;

Lx Ly
c16 ¼ H11 a ; ðD:6Þ
4

Lx Ly
c17 ¼ H22 b ; ðD:7Þ
4

Lx Ly
r1 ¼ qmn ; ðD:8Þ
4
(
8
3
ab if m; n odd;
c21 ¼ ðD:9Þ
0 otherwise;

Lx Ly n e  e 3 e o
c22 ¼ K 5 E 1  G 11 a  G 12 ab2  H11 a ; ðD:10Þ
4

Lx Ly n e  e o
e 33 b2  H11 ;
c23 ¼ K 6 E 1  D 11 a2  D ðD:11Þ
4
1752 A. Milazzo / Applied Mathematical Modelling 38 (2014) 1737–1752

Lx Ly n e  h e i o
e 33 ab ;
c24 ¼ K 7 E 1  D 12 þ D ðD:12Þ
4

X 4
Lx Ly n f o
e  Ni ;
r2 ¼  W 1i a þ K 7þi E 1 ðD:13Þ
i¼1
4
(
8
3
ab if m; n odd;
c31 ¼ ðD:14Þ
0 otherwise;

Lx Ly n e  e 2 o
e 22 b3  H22 b ;
c32 ¼ K 5 E 2  G 21 a b  G ðD:15Þ
4

Lx Ly n e  h e i o
e 33 ab ;
c33 ¼ K 6 E 2  D 21 þ D ðD:16Þ
4

Lx Ly n e  e o
e 33 a2  H22 ;
c34 ¼ K 7 E 2  D 22 b2  D ðD:17Þ
4

X 4
Lx Ly n f o
e  Ni :
r3 ¼  W 2i b þ K 7þi E 2 ðD:18Þ
i¼1
4

References

[1] S. Priya, R. Islam, S. Dong, D. Viehland, Recent advancements in magnetoelectric particulate and laminate composites, J. Electroceram. 19 (2007) 149–
166.
[2] C.-W. Nan, Magnetoelectric effect in composites of piezoelectric and piezomagnetic phases, Phys. Rev. B 50 (9) (1994) 6082–6088.
[3] E. Pan, Exact solution for simply supported and multilayered magneto-electro-elastic plates, J. Appl. Mech. Trans. ASME 68 (4) (2001) 608–618.
[4] E. Pan, P. Heyliger, Exact solutions for magneto-electro-elastic laminates in cylindrical bending, Int. J. Solids Struct. 40 (24) (2003) 6859–6876.
[5] P. Heyliger, E. Pan, Static fields in magnetoelectroelastic laminates, AIAA J. 42 (7) (2004) 1435–1443.
[6] P. Heyliger, F. Ramirez, E. Pan, Two-dimensional static fields in magnetoelectroelastic laminates, J. Intell. Mater. Syst. Struct. 15 (9–10) (2004) 689–709.
[7] X. Wang, Y.-P. Shen, The general solution of three-dimensional problems in magnetoelectroelastic media, Int. J. Eng. Sci. 40 (10) (2002) 1069–1080.
[8] J. Wang, L. Chen, S. Fang, State vector approach to analysis of multilayered magneto-electro-elastic plates, Int. J. Solids Struct. 40 (7) (2003) 1669–1680.
[9] M.-F. Liu, An exact deformation analysis for the magneto-electro-elastic fiber-reinforced thin plate, Appl. Math. Modell. 35 (5) (2011) 2443–2461.
[10] R. Garcia Lage, C.M. Mota Soares, C.A. Mota Soares, J.N. Reddy, Layerwise partial mixed finite element analysis of magneto-electro-elastic plates,
Comput. Struct. 82 (17–19) (2004) 1293–1301.
[11] S. Phoenix, S. Satsangi, B. Singh, Layer-wise modelling of magneto-electro-elastic plates, J. Sound Vib. 324 (3–5) (2009) 798–815.
[12] E. Carrera, P. Nali, Multilayered plate elements for the analysis of multifield problems, Finite Elem. Anal. Des. 46 (9) (2010) 732–742.
[13] A. Milazzo, I. Benedetti, C. Orlando, Boundary element method for magneto electro elastic laminates, CMES Comput. Model. Eng. Sci. 15 (1) (2006) 17–
30.
[14] J. Aimin, W. Guoquan, Q. Honglin, The boundary contour method for magneto-electro-elastic media with quadratic boundary elements, Int. J. Solids
Struct. 44 (18–19) (2007) 6220–6231.
[15] A. Milazzo, A one-dimensional model for dynamic analysis of generally layered magneto-electro-elastic beams, J. Sound Vib. 332 (2) (2013) 465–483.
[16] A. Milazzo, C. Orlando, A beam finite element for magneto-electro-elastic multilayered composite structures, Compos. Struct. 94 (12) (2012) 3710–
3721.
[17] A. Milazzo, An equivalent single-layer model for magnetoelectroelastic multilayered plate dynamics, Compos. Struct. 94 (6) (2012) 2078–2086.
[18] A. Milazzo, C. Orlando, An equivalent single-layer approach for free vibration analysis of smart laminated thick composite plates, Smart Mater. Struct.
21 (7) (2012) 075031.
[19] C. Xue, E. Pan, S. Zhang, H. Chu, Large deflection of a rectangular magnetoelectroelastic thin plate, Mech. Res. Commun. 38 (7) (2011) 518–523.
[20] J. Reddy, Mechanics of Laminated Composite Plates and Shells. Theory and analysis, CRC Press, 2004.
[21] J. Mitchell, J. Reddy, A refined hybrid plate theory for composite laminates with piezoelectric laminae, Int. J. Solids Struct. 32 (16) (1995) 2345–2367.
[22] S. Timoshenko, S. Woinowsky-Krieger, Theory of Plates and Shells, McGraw–Hill, New York, 1959.
[23] X.-M. Wang, Y.-P. Shen, The conservation laws and path-independent integrals with an application for linear electro-magneto-elastic media, Int. J.
Solids Struct. 33 (6) (1996) 865–878.
[24] M.-F. Liu, T.-P. Chang, Closed form expression for the vibration problem of a transversely isotropic magneto-electro-elastic plate, J. Appl. Mech. Trans.
ASME 77 (2) (2010) 024502-1–024502-8.

You might also like