See

discussions, stats, and author profiles for this publication at: https://www.researchgate.net/publication/289573622

Hermite-Lagrangian finite element

formulation to study functionally graded
sandwich beams

Article January 2016

DOI: 10.1016/j.compstruct.2016.01.015

CITATIONS READS

5 83

3 authors:

Jorge Yarasca JL Mantari

Universidad Nacional de Ingeniera (Peru) Universidad de Ingeniera y Tecnologa (Peru)
6 PUBLICATIONS 15 CITATIONS 62 PUBLICATIONS 1,148 CITATIONS

SEE PROFILE SEE PROFILE

Roman A. Arciniega
Universidad Peruana de Ciencias Aplicadas (
14 PUBLICATIONS 390 CITATIONS

SEE PROFILE

Some of the authors of this publication are also working on these related projects:

Comparative study of the potential of natural endemic fibers of Peru for its use as reinforcement in
laminated composite materials View project

Design of a devise to mitigate glaucoma through computational mechanics. Fondecyt-UK. View

project

All content following this page was uploaded by Jorge Yarasca on 30 April 2016.

The user has requested enhancement of the downloaded file.

 Composite Structures 140 (2016) 567581

Contents lists available at ScienceDirect

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

HermiteLagrangian finite element formulation to study functionally

graded sandwich beams
J. Yarasca a,, J.L. Mantari b,, R.A. Arciniega c
a
Department of Mechanical Engineering, National University of Engineering, Av. Tpac Amaru 210, Rimac, Lima, Peru
b
Department of Mechanical Engineering, University of Engineering and Technology, Jr. Medrano Silva 165, Barranco, Lima, Peru
c
Department of Civil Engineering, Universidad Peruana de Ciencias Aplicadas, Prolongacin Primavera 2390, Monterrico Santiago de Surco, Lima, Peru

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

Article history: This paper presents a static analysis of functionally graded single and sandwich beams by using an effi-
Available online 7 January 2016 cient 7DOFs quasi-3D hybrid type theory. The governing equations are derived by employing the princi-
ple of virtual works in a weak form and solved by means of the Finite Element Method (FEM). A C 1 cubic
Keywords: Hermite interpolation is used for the vertical deflection variables while C 0 linear interpolation is
Layered structures employed for the other kinematics variables. Convergence rates are studied in order to validate the finite
Elasticity element technique. Numerical results of the present formulation are compared with analytical and FEM
Finite element analysis (FEA)
solutions available in the literature.
2016 Elsevier Ltd. All rights reserved.

1. Introduction In order to reproduce the characteristic smoothness of the stres-

Functionally graded materials (FGMs) are advanced composite
materials, in which each constituent material varies gradually over
one or more directions. The material properties of FGMs vary con-
tinuously, which allows avoiding abrupt changes in the stress and
displacement distributions. This particular feature distinguishes
them from laminated composite materials, where abrupt change
in the material properties across the layer interfaces leads to large
interlaminar stresses. Thus, several researchers have analyzed the
mechanical behavior of FG structures [117].
Carrera et al. [18] investigated the influence of the stretching
effect on the static responses of functionally graded (FG) plates
and shells, which is especially significant for thick FG plates. Thai
et al. [1921] proposed new higher-order shear deformation theo-
ries for bending, buckling and vibration of functionally graded
plates. Apetre et al. [22] compared four different sandwich beam
theories: (a) an equivalent single-layer (ESL) first-order shear
deformation theory, (b) an ESL third-order shear deformation
theory developed by Reddy [23], (c) FourierGalerkin method
developed by Zhu and Sankar [24], and (d) the higher-order theory
of Frostig et al. [25]. Filippi et al. [26] based on Giunta et al. [1416]
beam formulation (1D Carreras unified formulation), evaluated
trigonometric, polynomial, exponential and miscellaneous expan-
sions for various structural problems.
Corresponding authors. Tel.: +51 1 4602092; mobile: +51 966748841.
E-mail address: jmantari@utec.edu.pe (J.L. Mantari).
ses via FEM, a proper model needs to be considered. For this rea-
son, Hermite-cubic polynomial is used to ensure C1 continuous
elements. The generation of Hermite (or Hermitian) polynomials
is described in many mathematical books [2729] and works have
been presented providing a systematic approach to allow simpler
coding (see Augarde [30]).
As far as the authors are aware, there is limited work available
for bending analysis of FG sandwich beams considering the thick-
ness stretching effect investigated via Finite Element Method
(FEM). Vo et al. [31] used a quasi-3D polynomial theory and orga-
nized it to build a finite element formulation with six degree-of-
freedom (DOF) to investigate the static behavior and the effect of
normal strain in FG sandwich beams for various power-law
indexes, skin-core-skin thickness ratios and boundary conditions.
Both Carrera et al. [18] and Vo et al. [31] established the impor-
tance of the transverse normal strain effects in higher-order theo-
ries for one-layered plates, multilayered plates, shells embedding
FGM layers (see Ref. [18]) and FG sandwich beams (see Ref. [31]).
The influence of non-polynomial or hybrid type shear strain shape
functions were not explored to study the bending analysis of FG
beams along with C1 quasi-3D hybrid type generalized shear defor-
mation theory solved via HermitianLagrangian finite element
technique. This paper aims to contribute in this sense.
In this paper, a hybrid type quasi-3D theory with both shear
deformation and thickness stretching effects for the bending anal-
ysis of FG beams is presented. Many quasi-3D hybrid type HSDTs
(polynomial, non-polynomial, and hybrid), including the thickness

http://dx.doi.org/10.1016/j.compstruct.2016.01.015
0263-8223/ 2016 Elsevier Ltd. All rights reserved.
568 J. Yarasca et al. / Composite Structures 140 (2016) 567581

expansion can be derived by using the present generalized theory. 2.3. Type C sandwich beams with FG skins and ceramic core
Convergence rates are studied in order to evaluate the performance
of the finite element formulation with various quasi-3D hybrid The bottom and top skin of sandwich beams is composed of a
type shear deformation models. The theory complies with the tan- FG material, while, the core is ceramic (Fig. 1c) with V c given by
gential stress-free boundary conditions on the beam boundary sur- [31]:
face, and thus a shear correction factor is not required. The beam
p
z  h0
governing equations of layered structures and its boundary condi- Vc z 2 h=2; h1  bottom skin
tions are derived by employing the principle of virtual works in a h1  h0
weak form and solved via FEM for sandwich beams subjected to V c 1 z 2 h1 ; h2  core 4
transverse load for simply supported, clampedclamped and
p
z  h3
clamped-free boundary conditions. The results are compared with Vc z 2 h2 ; h=2 top skin
Navier solutions and other FE solutions based on a quasi-3D HSDT. h2  h3
Further referential results for the displacement and stresses of FG
sandwich beams are obtained. 2.4. Theoretical displacement field

2. Analytical modeling of FG beams The displacement field satisfying the free surfaces boundary
conditions of transverse shear stresses (and hence strains) vanish-
A FG beam of length a, width b and a total thickness h made of a ing at a point (x; h=2) on the outer (top) and inner (bottom)
mixture of metal and ceramic materials is considered in the pre- surfaces of the beam, is given as follows:
sent analysis. Youngs modulus E is assumed to vary continuously  
@wb @h @ws @wb
through the thickness and obtained by the rule of mixtures [31]: ux; z u z y q  f z
@x @x @x @x 5
Ez Ec  Em V c z Em 1
wx; z wb ws gzh
where subscripts m and c represent the metallic and ceramic con-
where u, ws , wb and h are four unknown displacements at the mid-
stituents, V c is the volume fraction of the ceramic phase of the
plane of the beam. Table 1 presents different couples of shear strain
beam. For comparison reasons, three types of FG beams are consid-
shape functions to be evaluated in this paper. The first couple of
ered, see Fig. 1.
shear strain shape functions have polynomial nature, the second
one is trigonometric [7], and the last one is of a hybrid type (which
2.1. Type A FG beams
combines polynomial with non-polynomial shear strain shape func-
tions); i.e. three quasi-3D HSDTs. For simplicity, the theories are
The beam is composed of a FG material (Fig. 1a) with Vc given
called HSDT1, HSDT2 and HSDT3 (see Table 1). The constants y ,
by:

p y and q are obtained by considering the criteria to reduce the
2z h
V c z 2 number of unknowns in HSDTs as in Reddy and Liu [32]. These
2h
parameters are obtained from the shear strain shape functions,
0  
f z and gz, i.e. y y  1, y f 2h and q g 2h .
2.2. Type B sandwich beams with homogeneous skins and FG core For deriving the equations, small deformations are assumed, i.e.
displacements and rotations are small, and the material obeys
The bottom and top skin of sandwich beams is metal and cera- Hookes law. The starting point of the present generalized quasi-
mic, while, the core is composed of a FG material (Fig. 1b) with V c 3D HSDT is the 3D elasticity theory [33].
given by [31]:
V c 0 z 2 h=2; h1  bottom skin exx e0xx ze1xx f ze2xx

p
z  h1 ezz g 0 ze5zz 6
Vc z 2 h1 ; h2  core 3
h2  h1 cxz e0xz gze3xz f 0 ze4xz
V c 1 z 2 h2 ; h=2 top skin

(a) Type A (b) Type B

(c) Type C
Fig. 1. Geometry and coordinate of a FG sandwich beam.
 J. Yarasca et al. / Composite Structures 140 (2016) 567581 569

Table 1 where r and e are the stresses and the strain vectors defined in
Shear strain shape functions. Eq. (11) and (8) respectively, q is the distributed transverse load.
Model Function f z and gz A; B ; C; D; E; F; G; H ; I; J; K; L; M; N ; O; P; Q  ; R; S; T and
Polynomial HSDT1 4z3 2 U are the resultants of the following integrations:
f z 2 gz 1  4z2
N Z
3h h
Trigonometric HSDT2 f z p sinzp
1 gz coszp X zk
0
Hybrid HSDT3
z
f z tan 4h gz 1  4z2
2 A; B ; C; D; E; F Q k 1; z; f z; gz; f z; g 0 zdz;
h
k1 zk1

N Z
X zk
0
where G; H ; I; J; K Q k z2 ; zf z; zgz; zf z; zg 0 zdz;
2 2 2 2 zk1
@u @ w @ h @ ws @ wb k1
e0xx e1xx y 2b q  e2xx N Z
X zk
@x @x @x2 @x2 @x2 L; M; N ; O
0
Q k f z; f zgz; f zf z; f zg 0 zdz;
2

e5zz h 7 k1 zk1

@wb @h @h @w N Z
X zk
e0xz y q e3xz e4xz b P; Q  ; R
0
Q k g 2 z; gzf z; gzg 0 zdz;
@x @x @x @x
k1 zk1
In vector form this can be expressed as, N Z
X zk
0 0 0
S; T Q k f zf z; f zg 0 zdz;
e fexx ezz cxz g
T
8
k1 zk1

N Z
X zk
e e0 ze1 f ze2 gze3 f 0 ze4 g 0 ze5 9 U Q k g 0 zg 0 zdz;
k1 zk1
where
8 0 9 8 1 9 8 2 9 17
< exx = < exx = < exx =
e0 0 e1 0 e 0
2 where k represent a single-layer in the case of functionally graded
: 0 ; : ; : ; sandwich beam.
e
8 xz 9
0
8 9
0
8 9 10
<0 = <0 = <0 =
e3
0 e 4
0 e5 e5zz 3. Solution procedure
: 3 ; : 4 ; : ;
exz exz 0
For the finite element analysis, a two-noded C 1 beam element
The stresses can be written using the same approach with 7 DOFs per node is developed. Linear polynomial N is used
rT frxx rzz sxz g 11 for the unknown u and Hermite-cubic polynomial H is used for
the unknowns wb ; ws and h. The generalized displacements within
The linear constitutive relations become: an element are expressed as:
rQe 12 u N ue
where wb H web
2 3 18
Q 11 Q 13 0 ws H wes
6 7 h Hhe
Q 4 Q 13 Q 33 0 5 13
0 0 Q 55 where the shape functions are

The Q ij expressions are given below: N N1 N 2  H H1 H2 H3 H4  19

Ez and the nodal displacements of an element are
Q 11 z Q 33 z
1  m2 T T
Ezm ue u1 u 2 web wb1 wb2 wb3 wb4 
Q 13 z 14 T T
20
1  m2 web ws1 ws2 ws3 ws4  he h1 h2 h3 h4 
Ez
Q 55 z
21 m Table 2
Comparison between Navier, FI and SI solutions (Type A, HSDT1, a=h 100).
The elastic coefficients Q ij vary through the thickness according
to Eq. (1). p Theory w rxx rzz rxz
Considering the static version of the principle of virtual work, 1 Navier 5.6935 115.9213 3.2135 0.7482
FI 5.6931 115.9883 3.1934 0.7498
the following expressions can be obtained:
SI 5.6931 115.9883 3.1934 0.7498
Z h=2 Z Z
2 Navier 7.2432 135.3580 3.4532 0.6855
0 deT r dx dz  qdwdx; 15 FI 7.2424 135.4352 3.4295 0.6754
h=2 x x
SI 7.2424 135.4352 3.4295 0.6754
Z
T
deo Ae0 deo B e1 deo C e2 deo De3 de0 TEe4 deo F e5
T T T T T
0 Table 3
x
Comparison between Navier, FI and SI solutions (Type A, HSDT1, a=h 200).
de1 B e0 de1 Ge1 de1 He2 de1 Ie3 de1 J e4 de1 K e5
T T T T T T

p Theory w rxx rzz rxz

de2 C e0 de2 He1 de2 Le2 de2 M e3 de2 N e4 de2 Oe5
T T T T T T

T T T T T  4 T 1 Navier 5.6927 231.8363 6.4441 0.7484

de De de Ie de M e de Pe de Q
3 0 3 1 3 2 3 3 3
e de Re 3 5
FI 5.6922 231.9702 6.4039 0.7500
de4 Ee0 de4 J e1 de4 N e2 de4 Q  e3 de4 Se4 de4 T e5
T T T T T T
SI 5.6922 231.9702 6.4039 0.7500
5T 5T 5T 5T 5T 5T
deZ F e de K e de Oe de Re de T e de U e5 dx
0 1 2 3 4
2 Navier 7.2421 270.7074 6.9263 0.6857
FI 7.2413 270.8614 6.8798 0.6751

q dwb ws  hq dx; 16 SI 7.2413 270.8614 6.8798 0.6751
x
570 J. Yarasca et al. / Composite Structures 140 (2016) 567581

Z
In order to solve Eq. (16) the matrices Bib and Bis are defined. T T T T T T
0 duT Bo A B0 Bo B B1 Bo C B2 Bo D B3 B0 E B4 Bo F B5
These matrices are obtained by substituting Eq. (18) in Eq. (10), x

obtaining the following equation: 1T T T T

B B B0 B1 G B1 B1 H B2 B1 I B3 B1 J B4 B1 K B5
T T

ev Bv u; where m 0; 1 . . . ; 5: 21 T T T T T T
B2 C B0 B2 H B1 B2 L B2 B2 M B3 B2 N B4 B2 O B5
where u is the vector with all the nodal displacements of an T T T T T T

element: B3 D B0 B3 I B1 B3 M B2 B3 P B3 B3 Q  B4 B3 R B5
T T T T T T

u ue web wes he 
T
22 B4 E B0 B4 J B1 B4 N B2 B4 Q  B3 B4 S B4 B4 T B5
T T T T T T
All matrices of type, Bv , associated with the expressions in B5 F B0 B5 K B1 B5 O B2 B5 R B3 B5 T B4 B5 U B5 dxdu
Z
Eq. (21), for example B0 or B1 , are given in Appendix A.
 qduT Bp dx; 23
Substituting Eq. (21) into Eq. (16) the following expression is x
obtained:

9.58

9.57

9.56

w
9.55

9.54 FEM
Navier

9.53
5 10 15 20 25 30 35 40 45 50 55 60

Ne
Fig. 2. Convergence of vertical displacement to the analytical solution (Type A, HSDT1, p 10, a=h 20).

0.66

0.655

0.65
xz

0.645
FEM
Navier
0.64

5 10 15 20 25 30 35 40 45 50 55 60

Ne
Fig. 3. Convergence of shear stress to the analytical solution (Type A, p 10, HSDT1, a=h 5).

0.685

0.68

0.675

0.67

xz 0.665

0.66

0.655
FEM
0.65 Navier

0.645
5 10 15 20 25 30 35 40 45 50 55 60

Ne
Fig. 4. Convergence of shear stress to the analytical solution (Type A, p 10, HSDT1, a=h 20).
 J. Yarasca et al. / Composite Structures 140 (2016) 567581 571

where To implement this technique, the vector e in Eq. (8) is divided in

T their bending and shear components:
Bp 0 H H 
q H  24
eTb fexx ezz 0g
and du are the virtual variations of the nodal displacements. 25
In order to attenuate the effects of shear locking due to low eTs f0 0 cxz g
thickness values, many techniques are used. Among these, the
selective integration (SI) proved to be one of the simplest remedies eb e0b ze1b f ze2b g 0 ze5b
for avoiding shear locking. This method is based on a reduced 26
Gauss integration of the terms of the stiffness matrix that are
es e0s gze3s f 0 ze4s
related to shear.

9.6

9.58

9.56
w
9.54

FEM
9.52
Navier

9.5
5 10 15 20 25 30 35 40 45 50 55 60

Ne
Fig. 5. Convergence of vertical displacement to the analytical solution (Type A, HSDT2, p 10, a=h 20).

41

40.5

40

xx 39.5

39
FEM
38.5 Navier

38
5 10 15 20 25 30 35 40 45 50 55 60

Ne
Fig. 6. Convergence of axial stress to the analytical solution (Type A, p 10, HSDT2, a=h 20).

-0.8

-0.9

-1

-1.1

zz -1.2

-1.3

-1.4 FEM
Navier
-1.5

-1.6
5 10 15 20 25 30 35 40 45 50 55 60

Ne
Fig. 7. Convergence of normal stress to the analytical solution (Type A, p 10, HSDT2, a=h 20).
572 J. Yarasca et al. / Composite Structures 140 (2016) 567581

where Analogous to Eq. (15), the following expressions can be

8 9 8 9 8 9 8 9 obtained:
>
< exx >
0
= >
< exx >
1
= >
< exx >
2
= >
< 0 >
= Z h=2 Z Z h=2 Z Z
e0b 0 e1b 0 eb 0
2
e5b e5zz
>
: > ; >
: > ; >
: > ; >
: >
; 0 deTb rb dxdz deTs rs dxdz  qdw dx; 28
0 0 0 0 h=2 x h=2 x x
8 9 8 9 8 9 27
>
<0 > = >
<0 > = >
<0 > = The solution of Eq. (28) is obtained in similar manner as
e0s 0 e3s 0 es 0
4
Eqs. (16)(24). The results obtained from full integration (FI) and
>
: 0 > ; >
: 3 > ; >
: 4 > ;
exz exz exz SI are compared for thin beams (a=h 100 and a=h 200) in
Tables 2 and 3. The kinematic model solved via FEM using

0.69

0.68

0.67
xz
0.66

0.65 FEM
Navier

0.64
5 10 15 20 25 30 35 40 45 50 55 60

Ne
Fig. 8. Convergence of shear stress to the analytical solution (Type A, p 10, HSDT2, a=h 20).

9.59

9.58

9.57

9.56
w
9.55

9.54
FEM
9.53 Navier

9.52
5 10 15 20 25 30 35 40 45 50 55 60

Ne
Fig. 9. Convergence of vertical displacement to the analytical solution (Type A, HSDT3, p 10, a=h 20).

41

40.5

40

xx 39.5

39
FEM
38.5 Navier

38
5 10 15 20 25 30 35 40 45 50 55 60

Ne
Fig. 10. Convergence of axial stress to the analytical solution (Type A, p 10, HSDT3, a=h 20).
 J. Yarasca et al. / Composite Structures 140 (2016) 567581 573

Table 4 Table 7
Comparison of the maximum vertical displacement of FG SS beams (Type A). Comparison of the shear stress rxz 0; 0 of FG SS beams (Type A).
Method Theory p0 p1 p2 p5 p 10 Method Theory p0 p1 p2 p5 p 10
a/h = 5 a/h = 5
Navier Vo et al. [31] 3.1397 6.1338 7.8606 9.6037 10.7578 Navier Vo et al. [31] 0.7233 0.7233 0.6622 0.5840 0.6396
FEM Vo et al. [31] 3.1397 6.1334 7.8598 9.6030 10.7572 FEM Vo et al. [31] 0.7291 0.7291 0.6661 0.5873 0.6439
HSDT1 3.1397 6.1334 7.8598 9.6029 10.7572 HSDT1 0.7233 0.7233 0.6623 0.5841 0.6397
HSDT2 3.1397 6.1334 7.8598 9.6030 10.7573 HSDT2 0.7235 0.7235 0.6625 0.5843 0.6399
HSDT3 3.1397 6.1334 7.8597 9.6025 10.7569 HSDT3 0.7224 0.7224 0.6613 0.5829 0.6386
a/h = 20 a/h = 20
Navier Vo et al. [31] 2.8947 5.7201 7.2805 8.6479 9.5749 Navier Vo et al. [31] 0.7432 0.7432 0.6809 0.6010 0.6583
FEM Vo et al. [31] 2.8947 5.7197 7.2797 8.6471 9.5743 FEM Vo et al. [31] 0.7466 0.7466 0.6776 0.6036 0.6675
HSDT1 2.8947 5.7197 7.2797 8.6471 9.5742 HSDT1 0.7443 0.7443 0.6800 0.6019 0.6614
HSDT2 2.8947 5.7197 7.2797 8.6471 9.5742 HSDT2 0.7443 0.7443 0.6800 0.6019 0.6614
HSDT3 2.8947 5.7197 7.2797 8.6471 9.5743 HSDT3 0.7433 0.7433 0.6790 0.6014 0.6529

HermiteLagrangian shape functions appear to be not affected by
shear locking as shown in both tables. For this reason, FI is used
in the subsequent part of this paper.
4. Numerical results and discussions
The target of this paper is to present: (a) the mathematical for-
mulation for the HermitianLagrangian finite element formulation
of a generalized quasi-3D hybrid type HSDT with 7 DOF per node;
and (b) numerical results of the theoretical formulation using poly-
nomial, non-polynomial and quasi-3D hybrid type shear strain
shape functions for various FG sandwich beams. It was not the
intent of this paper to present the best FE formulation having
7DOF. However, so far, very accurate simple FE formulation of
quasi-3D HSDT with 7DOF can be obtained by simply using refer-
ential shear strain shape functions developed previously by the
authors [7,23].
In case of the present quasi-3D hybrid type HSDT (HSDT3) it is
important to properly select its shear strain shape function in order
to get accurate results. For some non-polynomial shear strain
shape functions the bending results of the present finite element
formulation may present inadequate convergence rates and in
some cases inaccurate results. Hence, careful analysis is needed
to select the appropriate shear strain shape functions. This threat
is alleviated when one of the hybrid shears strain function is poly-
nomial as in HSDT3.
It is important to remark that FE formulation of quasi-3D HSDT
with 6 DOFs can be obtained as special case of the present gener-
alized finite element formulation. This can be achieved by selecting
a g(z) so that q 0. By doing that the number of DOFs is reduced
from 7 to 6 as in the interesting formulation presented by Vo et al.
[31]. Readers can easily verify this point in Eq. (5).
As mentioned before, in this paper, the FEM is used to investi-
gate the bending behavior of FG sandwich beams with various

Table 5
Comparison of the axial stress rxx a=2; h=2 of FG SS beams (Type A).
Method Theory p0 p1 p2 p5 p 10
a/h = 5
Navier Vo et al. [31] 3.8005 5.8812 6.8818 8.1140 9.7164
FEM Vo et al. [31] 3.8020 5.8840 6.8860 8.1190 9.7220
HSDT1 3.8023 5.8846 6.8860 8.1190 9.7221
HSDT2 3.8028 5.8855 6.8871 8.1202 9.7234
HSDT3 3.8021 5.8843 6.8857 8.1184 9.7215
a/h = 20
Navier Vo et al. [31] 15.0125 23.2046 27.0988 31.8137 38.1395
FEM Vo et al. [31] 15.0200 23.2200 27.1100 31.8300 38.1600
HSDT1 15.0195 23.2182 27.1155 31.8338 38.1621
HSDT2 15.0197 23.2184 27.1158 31.8341 38.1624
HSDT3 15.0195 23.2181 27.1154 31.8337 38.1636

Table 6
Comparison of the normal stress rzz a=2; h=2 of FG SS beams (Type A).
Method Theory p0 p1 p2 p5 p 10
a/h = 5
Navier Vo et al. [31] 0.1352 0.0670 0.0925 0.0180 0.0181
FEM Vo et al. [31] 0.1352 0.0672 0.0927 0.0183 0.0179
HSDT1 0.1357 0.0680 0.0937 0.0196 0.0164
HSDT2 0.1357 0.0712 0.0972 0.0234 0.0122
HSDT3 0.1356 0.0680 0.0936 0.0193 0.0167
a/h = 20
Navier Vo et al. [31] 0.0337 0.5880 0.6269 1.1698 1.5572
FEM Vo et al. [31] 0.0338 0.5874 0.6261 1.1690 1.5560
HSDT1 0.0359 0.5839 0.6215 1.1637 1.5511
HSDT2 0.0364 0.5830 0.6205 1.1627 1.5499
HSDT3 0.0358 0.5839 0.6215 1.1638 1.5500
574 J. Yarasca et al. / Composite Structures 140 (2016) 567581

0.5

0.3

0.1
z/h
-0.1 Navier
HSDT1
HSDT2
-0.3
HSDT3

-0.5
-20 -10 0 10 20 30 40

xx

(a) Axial stress

0.5

0.3

0.1 Navier
HSDT1
z/h HSDT2
-0.1 HSDT3

-0.3

-0.5
0 0.1 0.2 0.3 0.4 0.5 0.6 0.7

xz

(b) Shear stress

Fig. 11. Variation of axial and transverse shear stress through the thickness of FG sandwich SS beams under uniform load (p 10, Type A, a=h 20).

Table 8
Comparison of the maximum vertical displacement of FG CC and C-F beams (Type A).

Method Theory p0 p1 p2 p5 p 10
a/h = 5
CC Vo et al. [31] 0.8327 1.5722 2.0489 2.6929 3.1058
HSDT1 0.8339 1.5742 2.0510 2.6951 3.1084
HSDT2 0.8339 1.5741 2.0509 2.6951 3.1084
HSDT3 0.8340 1.5743 2.0511 2.6950 3.1086
CF Vo et al. [31] 28.5524 56.2002 71.7295 86.1201 95.7582
HSDT1 28.5666 56.2243 71.7523 86.1412 95.7830
HSDT2 28.5664 56.2240 71.7518 86.1412 95.7833
HSDT3 28.5674 56.2257 71.7537 86.1414 95.7843
a/h = 20
CC Vo et al. [31] 0.5894 1.1613 1.4811 1.7731 1.9694
HSDT1 0.5896 1.1616 1.4813 1.7732 1.9694
HSDT2 0.5896 1.1615 1.4813 1.7732 1.9694
HSDT3 0.5896 1.1616 1.4813 1.7736 1.9653
CF Vo et al. [31] 27.6217 54.6285 69.5266 82.4836 91.2606
HSDT1 27.6200 54.6255 69.5214 82.4750 91.2490
HSDT2 27.6199 54.6254 69.5214 82.4750 91.2490
HSDT3 27.6201 54.6258 69.5217 82.4811 91.1994

theories. Displacements and stresses of symmetric and non-
symmetric sandwich beams with FG material in the core or skins
are calculated. Various power-law indexes, skin-core-skin thick-
ness ratios and boundary conditions are considered. FG sandwich
beams made of Aluminum as metal (Al: Em 70 GPa, mm 0:3)
and Alumina as ceramic (Al2 O3 : Ec 380 GPa, mm 0:3) with two
slenderness ratios, a=h equal to 5 and 20, are considered. For con-
venience, the following non-dimensional terms are used to calcu-
late the vertical displacement of beams under uniformly
distributed load q:
 J. Yarasca et al. / Composite Structures 140 (2016) 567581 575

Table 9 distributed load are considered; p 10 is considered since conver-

The maximum vertical displacement of (1-8-1) FG SS sandwich beams (Type B). gence rates are higher for this case due to the fact that such a high
Method Theory p0 p1 p2 p5 p 10 exponent results practically in homogeneous ceramic beam.
a/h = 5 Figs. 24 present the convergence rates for HSDT1. From Fig. 2,
Navier Vo et al. [31] 3.9374 6.5505 7.7721 8.8089 9.2426 it is clear that the vertical displacement obtained by FEM
FEM Vo et al. [31] 3.9374 6.5499 7.7711 8.8078 9.2417 converges to the Navier solution. Figs. 3 and 4 show that more ele-
HSDT1 3.9373 6.5498 7.7711 8.8078 9.2417 ments are needed in order to get accurate results for transverse
HSDT2 3.9374 6.5498 7.7711 8.8079 9.2419
HSDT3 3.9373 6.5499 7.7710 8.8071 9.2408
shear stresses of thin beams.
Figs. 58 present the convergence rates for HSDT2. It is impor-
a/h = 20
Navier Vo et al. [31] 3.6841 6.1383 7.2143 7.9435 8.1710
tant to remark that the shear strain shape function has to be care-
FEM Vo et al. [31] 3.6840 6.1377 7.2133 7.9425 8.1700 fully chosen in order to obtain good results. Figs. 9 and 10 present
HSDT1 3.6840 6.1377 7.2133 7.9424 8.1700 the convergence rates for HSDT3. It is clear from this figures that
HSDT2 3.6840 6.1377 7.2133 7.9424 8.1700 the FEM converge the Navier solution very well.
HSDT3 3.6840 6.1377 7.2133 7.9424 8.1699

4.2. Type A: FG beams

100Em h a
3
 FG beams (Type A) under uniformly distributed load are consid-
w w ;z ; 29a
qa4 2 ered. The vertical displacement and stresses obtained from the
and the axial, normal and shear stresses:
h a Table 12
rxx rxx ; z 29b Comparison of the shear stress rxz 0; 0 of (1-8-1) FG sandwich SS beams (Type B).
qa 2
Method Theory p0 p1 p2 p5 p 10
h a
rzz rzz ; z 29c a/h = 5
qa 2 Navier Vo et al. [31] 0.7486 0.7219 0.6365 0.5262 0.5733
FEM Vo et al. [31] 0.7568 0.7272 0.6395 0.5286 0.5766
h HSDT1 0.7487 0.7220 0.6366 0.5262 0.5734
rxz rxz 0; z 29d HSDT2 0.7489 0.7221 0.6368 0.5265 0.5737
qa HSDT3 0.7477 0.7210 0.6355 0.5249 0.5721
a/h = 20
4.1. Convergence studies Navier Vo et al. [31] 0.7683 0.7418 0.6543 0.5414 0.5900
FEM Vo et al. [31] 0.7777 0.7412 0.6454 0.5399 0.5963
HSDT1 0.7709 0.7418 0.6518 0.5409 0.5924
A large number of convergence studies and comparison to ana- HSDT2 0.7709 0.7418 0.6518 0.5410 0.5924
lytical solutions were developed in order to assess the imple- HSDT3 0.7699 0.7408 0.6507 0.5396 0.5871
mented finite elements. FG beams (Type A) under uniformly

Table 10
Comparison of the axial stress rxx of (1-8-1) FG sandwich SS beams (Type B).
Method Theory p0 p1 p2 p5 p 10
a/h = 5
Navier Vo et al. [31] 4.4603 6.0069 6.5253 6.8927 7.2292
FEM Vo et al. [31] 4.4620 6.0100 6.5290 6.8970 7.2330
HSDT1 4.4626 6.0103 6.5294 6.8971 7.2335
HSDT2 4.4634 6.0113 6.5304 6.8981 7.2345
HSDT3 4.4624 6.0101 6.5290 6.8965 7.2329
a/h = 20
Navier Vo et al. [31] 17.6318 23.7073 25.6848 26.9703 28.2298
FEM Vo et al. [31] 17.6400 23.7200 25.7000 26.9900 28.2500
HSDT1 17.6409 23.7210 25.7008 26.9875 28.2469
HSDT2 17.6411 23.7212 25.7011 26.9878 28.2472
HSDT3 17.6408 23.7209 25.7007 26.9873 28.2475

Table 11
Comparison of the normal stress rzz a=2; h=2 of (1-8-1) FG sandwich SS beams (Type B).
Method Theory p0 p1 p2 p5 p 10
a/h = 5
Navier Vo et al. [31] 0.0872 0.1043 0.1277 0.0619 0.0001
FEM Vo et al. [31] 0.0873 0.1045 0.1279 0.0622 0.0001
HSDT1 0.0879 0.1054 0.1289 0.0633 0.0012
HSDT2 0.0905 0.1085 0.1322 0.0666 0.0044
HSDT3 0.0878 0.1053 0.1288 0.0630 0.0009
a/h = 20
Navier Vo et al. [31] 0.2904 0.4373 0.4179 0.8042 1.1450
FEM Vo et al. [31] 0.2901 0.4367 0.4170 0.8032 1.1440
HSDT1 0.2879 0.4330 0.4122 0.7987 -1.1402
HSDT2 0.2871 0.4321 0.4113 0.7978 1.1394
HSDT3 0.2879 0.4330 0.4123 0.7988 1.1398
576 J. Yarasca et al. / Composite Structures 140 (2016) 567581

0.5

0.3

0.1 Navier
HSDT1
z/h HSDT2
-0.1 HSDT3

-0.3

-0.5
-2 -1.8 -1.6 -1.4 -1.2 -1 -0.8 -0.6

zz
(a) Normal stress

0.5

0.3

0.1 Navier
HSDT1
z/h HSDT2
-0.1 HSDT3

-0.3

-0.5
0 0.2 0.4 0.6 0.8 1 1.2 1.4

xz
(b) Shear stress
Fig. 12. Variation of transverse normal and shear stress through the thickness of (1-8-1) FG sandwich SS beams under uniform load (p 10, Type B, a=h 20).

Table 13
The maximum vertical displacement of (1-8-1) FG CC and CF sandwich beams (Type B).

Method Theory p0 p1 p2 p5 p 10
a/h = 5
CC Vo et al. [31] 1.0046 1.6539 2.0122 2.4595 2.7089
HSDT1 1.0064 1.6559 2.0141 2.4613 2.7108
HSDT2 1.0064 1.6559 2.0141 2.4614 2.7110
HSDT3 1.0065 1.6561 2.0142 2.4604 2.7104
CF Vo et al. [31] 36.1509 60.2081 71.0316 79.0886 81.9813
HSDT1 36.1746 60.2321 71.0525 79.1058 81.9987
HSDT2 36.1747 60.2317 71.0518 79.1059 81.9994
HSDT3 36.1755 60.2335 71.0535 79.1020 81.9969
a/h = 20
CC Vo et al. [31] 0.7472 1.2447 1.4669 1.6283 1.6843
HSDT1 0.7474 1.2449 1.4671 1.6284 1.6843
HSDT2 0.7474 1.2449 1.4671 1.6284 1.6843
HSDT3 0.7475 1.2449 1.4671 1.6282 1.6824
CF Vo et al. [31] 35.1767 58.6432 68.9096 75.7851 77.8811
HSDT1 35.1765 58.6403 68.9049 75.7782 77.8731
HSDT2 35.1765 58.6403 68.9049 75.7782 77.8731
HSDT3 35.1767 58.6405 68.9051 75.7761 77.8516

different HSDTs are given in Tables 48 along with the results from
previous studies (Ref. [31]). Table 4 present results of non-
dimensionalized maximum beam deflection; the results using
FEM have good agreement with the Navier solutions and those of
previous paper [31]. Tables 57 show that the stresses obtained
from FEM are very close to Navier solution especially for HSDT1
and HSDT3. The comparisons of the distribution of axial and shear
stresses through the beam thickness are shown in Fig. 11. The ver-
tical displacement obtained from the different HSDTs for clamped
clamped (CC) and clamped-free (C-F) beams are given in Table 8.
The present results fully agree with those of the referential article
of this paper (Ref. [31]).
 J. Yarasca et al. / Composite Structures 140 (2016) 567581 577

Table 14
The maximum vertical displacement of FG sandwich SS beams (Type C).

p Theory a=h 5 a=h 20

1-2-1 2-2-1 1-2-1 2-2-1
Navier FEM Navier FEM Navier FEM Navier FEM
0 Vo et al. [31] 3.1397 3.1397 2.8947 2.8947
HSDT1 3.1397 3.1397 3.1397 3.1397 2.8947 2.8947 2.8947 2.8947
HSDT2 3.1397 3.1397 3.1397 3.1397 2.8947 2.8947 2.8947 2.8947
HSDT3 3.1397 3.1397 3.1397 3.1397 2.8947 2.8947 2.8947 2.8947
1 Vo et al. [31] 5.3612 5.7777 5.0975 5.5040
HSDT1 5.3611 5.3611 5.7777 5.7776 5.0975 5.0975 5.5040 5.5039
HSDT2 5.3612 5.3612 5.7777 5.7777 5.0975 5.0975 5.5040 5.5039
HSDT3 5.3611 5.3611 5.7777 5.7776 5.0975 5.0975 5.5040 5.5039
2 Vo et al. [31] 6.6913 7.4629 6.4235 7.1790
HSDT1 6.6913 6.6913 7.4630 7.4629 6.4235 6.4235 7.1791 7.1789
HSDT2 6.6913 6.6913 7.4630 7.4629 6.4235 6.4235 7.1791 7.1789
HSDT3 6.6913 6.6913 7.4630 7.4629 6.4235 6.4235 7.1791 7.1790
5 Vo et al. [31] 8.4276 9.6459 8.1589 9.3498
HSDT1 8.4276 8.4276 9.6462 9.6459 8.1589 8.1589 9.3501 9.3498
HSDT2 8.4277 8.4277 9.6461 9.6459 8.1589 8.1589 9.3501 9.3498
HSDT3 8.4276 8.4276 9.6463 9.6460 8.1589 8.1589 9.3501 9.3498
10 Vo et al. [31] 9.3099 10.6769 9.0413 10.3715
HSDT1 9.3099 9.3099 10.6772 10.6768 9.0413 9.0413 10.3719 10.3715
HSDT2 9.3099 9.3099 10.6771 10.6767 9.0413 9.0413 10.3719 10.3715
HSDT3 9.3099 9.3099 10.6773 10.6770 9.0413 9.0413 10.3719 10.3715

Table 15
Comparison of the axial stress rxx of FG sandwich SS beams (Type C).
p Theory a=h 5 a=h 20
1-2-1 2-2-1 1-2-1 2-2-1
Navier FEM Navier FEM Navier FEM Navier FEM
0 Vo et al. [31] 3.8005 3.8005 15.0125 15.0125
HSDT1 3.8005 3.8023 3.8005 3.8023 15.0126 15.0195 15.0126 15.0195
HSDT2 3.8011 3.8028 3.8011 3.8028 15.0128 15.0197 15.0128 15.0197
HSDT3 3.8004 3.8021 3.8004 3.8021 15.0126 15.0195 15.0126 15.0195
1 Vo et al. [31] 1.2315 1.2459 4.8797 4.9360
HSDT1 1.2315 1.2320 1.2459 1.2466 4.8797 4.8819 4.9360 4.9386
HSDT2 1.2317 1.2323 1.2462 1.2468 4.8798 4.8820 4.9361 4.9387
HSDT3 1.2314 1.2320 1.2459 1.2465 4.8797 4.8819 4.9360 4.9386
2 Vo et al. [31] 1.5505 1.5849 6.1526 6.2882
HSDT1 1.5505 1.5512 1.5850 1.5858 6.1526 6.1554 6.2883 6.2918
HSDT2 1.5508 1.5515 1.5853 1.5862 6.1527 6.1555 6.2884 6.2919
HSDT3 1.5504 1.5511 1.5849 1.5858 6.1526 6.1554 6.2883 6.2918
5 Vo et al. [31] 1.9672 2.0160 7.8185 8.0100
HSDT1 1.9672 1.9681 2.0160 2.0172 7.8186 7.8222 8.0100 8.0148
HSDT2 1.9676 1.9685 2.0165 2.0177 7.8187 7.8223 8.0101 8.0149
HSDT3 1.9672 1.9681 2.0160 2.0172 7.8186 7.8222 8.0100 8.0144
10 Vo et al. [31] 2.1788 2.2161 8.6655 8.8094
HSDT1 2.1788 2.1798 2.2162 2.2176 8.6656 8.6696 8.8095 8.8150
HSDT2 2.1793 2.1803 2.2166 2.2180 8.6657 8.6697 8.8097 8.8151
HSDT3 2.1788 2.1798 2.2161 2.2175 8.6656 8.6696 8.8418 8.8338

4.3. Type B: Sandwich beams with homogeneous skins FG core
In this example, bending analysis of (1-8-1) sandwich beams of
Type B is performed. The results are given in Tables 913. It is clear
that the vertical displacement and stresses obtained via FEM is
similar with the Navier solution. In Table 11 can be seen that the
normal stress obtained from the different HSDTs are somewhat dif-
ferent. However, Fig. 12 shows similar distribution of the axial and
shear stress along the thickness for the different HSDTs. Therefore,
it can be said that the present theories, in general, can be accept-
able for this kind of problem. Table 13 presents some results for
CC and C-F beams, it is clear that the results are in excellent
agreement with referential results published in the literature
(see Ref. [31]).
4.4. Type C: Sandwich beams with FG skins ceramic core
Finally, two types of symmetric (1-2-1) and non-symmetric
(2-2-1) sandwich beams of Type C are considered. The vertical
displacement and stresses for various HSDTs are given in Tables
1419. As expected, the deflection and different stresses obtained
via FEM are similar with the Navier solution. Figs. 13 and 14 show
578 J. Yarasca et al. / Composite Structures 140 (2016) 567581

Table 16
Comparison of the normal stress rzz of FG sandwich SS beams (Type C).
p Theory a=h 5 a=h 20
1-2-1 2-2-1 1-2-1 2-2-1
Navier FEM Navier FEM Navier FEM Navier FEM
0 Vo et al. [31] 0.1352 0.1352 0.0337 0.0337
HSDT1 0.1352 0.1357 0.1352 0.1357 0.0338 0.0359 0.0338 0.0359
HSDT2 0.1370 0.1375 0.1370 0.1375 0.0343 0.0364 0.0343 0.0364
HSDT3 0.1351 0.1356 0.1351 0.1356 0.0338 0.0358 0.0338 0.0358
1 Vo et al. [31] 0.0447 0.0286 0.0111 0.0625
HSDT1 0.0447 0.0449 0.0286 0.0288 0.0112 0.0118 0.0625 0.0617
HSDT2 0.0454 0.0456 0.0294 0.0296 0.0114 0.0121 0.0623 0.0615
HSDT3 0.0447 0.0448 0.0286 0.0288 0.0112 0.0118 0.0625 0.0617
2 Vo et al. [31] 0.0564 0.0341 0.0141 0.0895
HSDT1 0.0564 0.0566 0.0341 0.0344 0.0141 0.0149 0.0895 0.0884
HSDT2 0.0574 0.0576 0.0352 0.0355 0.0144 0.0152 0.0892 0.0881
HSDT3 0.0564 0.0566 0.0341 0.0344 0.0141 0.0149 0.0895 0.0884
5 Vo et al. [31] 0.0712 0.0454 0.0178 0.1010
HSDT1 0.0712 0.0714 0.0455 0.0458 0.0178 0.0189 0.1009 0.0994
HSDT2 0.0725 0.0728 0.0468 0.0472 0.0182 0.0192 0.1006 0.0990
HSDT3 0.0712 0.0714 0.0455 0.0458 0.0178 0.0189 0.1009 0.0995
10 Vo et al. [31] 0.0783 0.0518 0.0195 0.0998
HSDT1 0.0783 0.0786 0.0518 0.0522 0.0196 0.0208 0.0997 0.0978
HSDT2 0.0798 0.0801 0.0533 0.0537 0.0200 0.0212 0.0993 0.0974
HSDT3 0.0783 0.0786 0.0518 0.0522 0.0196 0.0208 0.0780 0.0846

Table 17
Comparison of the shear stress rxz of FG sandwich SS beams (Type C).
p Theory a=h 5 a=h 20
1-2-1 2-2-1 1-2-1 2-2-1
Navier FEM Navier FEM Navier FEM Navier FEM
0 Vo et al. [31] 0.7233 0.7233 0.7432 0.7432
HSDT1 0.7233 0.7233 0.7233 0.6944 0.7433 0.7443 0.7433 0.7145
HSDT2 0.7235 0.7235 0.7235 0.6945 0.7433 0.7443 0.7433 0.7145
HSDT3 0.7223 0.7224 0.7223 0.6938 0.7423 0.7433 0.7423 0.7138
1 Vo et al. [31] 0.7993 0.8342 0.8193 0.7432
HSDT1 0.7993 0.7994 0.8342 0.8010 0.8194 0.8206 0.8553 0.8221
HSDT2 0.7995 0.7996 0.8344 0.8011 0.8194 0.8206 0.8553 0.8221
HSDT3 0.7983 0.7984 0.8333 0.8004 0.8185 0.8197 0.8544 0.8215
2 Vo et al. [31] 0.8349 0.8920 0.8556 0.9142
HSDT1 0.8349 0.8350 0.8920 0.8564 0.8557 0.8570 0.9143 0.8781
HSDT2 0.8351 0.8352 0.8922 0.8565 0.8557 0.8570 0.9143 0.8781
HSDT3 0.8340 0.8341 0.8911 0.8560 0.8549 0.8561 0.9135 0.8776
5 Vo et al. [31] 0.8763 0.9683 0.8986 0.9927
HSDT1 0.8763 0.8764 0.9683 0.9297 0.8987 0.8999 0.9928 0.9512
HSDT2 0.8764 0.8765 0.9684 0.9297 0.8987 0.9000 0.9928 0.9512
HSDT3 0.8754 0.8755 0.9676 0.9293 0.8975 0.8986 0.9900 0.9501
10 Vo et al. [31] 0.8980 1.0148 0.9214 1.0405
HSDT1 0.8980 0.8981 1.0148 0.9743 0.9214 0.9227 1.0406 0.9953
HSDT2 0.8981 0.8982 1.0149 0.9743 0.9214 0.9227 1.0406 0.9953
HSDT3 0.8972 0.8973 1.0140 0.9740 0.9227 0.9246 1.0396 1.0202

similar distributions of axial, normal and shear stresses along the
thickness for symmetric and non-symmetric sandwich beams.
Tables 18 and 19 present the results for CC and C-F beams. As
expected, based on the previous test problems, the results are
accurate and inline with results published in the literature. So,
the present FE formulation of a generalized quasi-3D hybrid type
HSDT can be further utilized or extended to cover other mechanical
or physical problems.
5. Conclusions
The HermitianLagrangian finite element formulation of a gen-
eralized quasi-3D hybrid type HSDT with 7DOF is presented in this
paper. The governing equations and boundary conditions are
derived by employing the principle of virtual work in a weak form
and solved by means of the FEM for several functionally graded
sandwich beams subjected to distributed load for simply sup-
ported, clampedclamped and clamped-free boundary conditions.
The important conclusions that emerge from this paper can be
summarized as follows:
(a) Multiple shears strain shape function can be efficiently eval-
uated by using the present theory.
(b) The polynomial, non-polynomial and hybrid shear strain
functions proposed converge to the analytical value as the
number of elements is increased.
 J. Yarasca et al. / Composite Structures 140 (2016) 567581 579

0.5

0.3
Navier
HSDT1
0.1 HSDT2
z/h HSDT3
-0.1

-0.3

-0.5
-25 -20 -15 -10 -5 0 5 10 15 20 25

xx
(a) Axial stress

0.5

0.3 Navier
HSDT1
HSDT2
0.1 HSDT3
z/h
-0.1

-0.3

-0.5
-0.06 -0.04 -0.02 0 0.02 0.04 0.06

zz
(b) Transverse normal stress
0.5

0.3

0.1 Navier
HSDT1
z/h HSDT2
-0.1 HSDT3

-0.3

-0.5
0 0.1 0.2 0.3 0.4 0.5 0.6 0.7 0.8 0.9 1

xz

(c) Transverse shear stress

Fig. 13. Variation of stresses through the thickness of (1-2-1) FG sandwich SS beams under uniform load (p 10, Type C, a=h 20).

(c) The kinematic model proposed solved via FEM using Hermi-
tian shape functions appear to be not affected by shear lock-
ing. Full integration and selective integration results are
similar, hence no integration technique appear to be neces-
sary in this model. Further studies need to be carried out in
this sense.
(d) The hybrid type shear strain function presented in this paper
are more accurate than trigonometric ones by having one
polynomial function as shear strain shape function.
Further studies need to be performed for different types of FG
sandwich beams with the present generalized finite element for-
mulation, such as metallic beams or classical laminated compos-
ites, which is not available in the literature so far.
Appendix A
Definition of Matrices of type, Biv
The matrices associated with the Eq. (21) are the following:
580 J. Yarasca et al. / Composite Structures 140 (2016) 567581

0.5

0.3 Navier
HSDT1
HSDT2
0.1 HSDT3

z/h
-0.1

-0.3

-0.5
-20 -15 -10 -5 0 5 10 15 20 25 30

xx
(a) Axial stress
0.5

Navier
0.3 HSDT1
HSDT2
HSDT3
0.1
z/h
-0.1

-0.3

-0.5
-2.5 -2 -1.5 -1 -0.5 0

zz
(b) Transverse normal stress
0.5

0.3
Navier
HSDT1
0.1
HSDT2
z/h HSDT3
-0.1

-0.3

-0.5
0 0.2 0.4 0.6 0.8 1 1.2 1.4

xz
(c) Transverse shear stress
Fig. 14. Variation of stresses through the thickness of (2-2-1) FG sandwich SS beams under uniform load (p 10, Type C, a=h 20).

2 @N @N 2
3
@x
1
@x
0 0 0 0 0 0 0 0 0 0 0 0
6 7
B0 4 0 0 0 0 0 0 0 0 0 0 ... 0 0 0 0 5;
0 0 y @H
@x
2
y @H
@x
3
y @H
@x
4
y @H
@x
5
0 0 0 0 q @H
@x
1
q @H@x21 q @H
@x
3
q @H
@x
4

2 2 2 2 2 2 2 2 2 2 2 2 2 3
0 0 y @@xH21 y @@xH22 y @@xH23 y @@xH24  @@xH21  @@xH22  @@xH23  @@xH24 q @@xH21 q @@xH22 q @@xH23 q @@xH24
1 6 7
B 40 0 0 0 0 0 0 0 ... 0 0 0 0 0 0 5;
0 0 0 0 0 0 0 0 0 0 0 0 0 0
 J. Yarasca et al. / Composite Structures 140 (2016) 567581 581

Table 18 References
The maximum vertical displacement of FG sandwich CC beams (Type C).

p Theory a=h 5 a=h 20 [1] Mantari JL, Oktem AS, Guedes Soares C. Bending response of functionally
graded plates by using a new higher order shear deformation theory. Compos
1-2-1 2-2-1 1-2-1 2-2-1 Struct 2012;94:71423.
[2] Mantari JL, Guedes Soares C. A novel higher-order shear deformation theory
0 Vo et al. [31] 0.8327 0.8327 0.5894 0.5894
with stretching effect for functionally graded plates. Compos Struct
HSDT1 0.8339 0.8339 0.5896 0.5896 2012;94:19912000.
HSDT2 0.8339 0.8339 0.5896 0.5896 [3] Mantari JL, Guedes Soares C. Evaluation of layer-wise mixed theories for
HSDT3 0.8340 0.8340 0.5896 0.5896 laminated plates analysis. Composite B 2013;45:26881.
1 Vo et al. [31] 1.3077 1.4002 1.0293 1.1108 [4] Mantari JL, Guedes Soares C. Generalized hybrid quasi-3D shear deformation
theory for the static analysis of advanced composite plates. Compos Struct
HSDT1 1.3109 1.4036 1.0299 1.1115
2012;94:256175.
HSDT2 1.3109 1.4035 1.0299 1.1115
[5] Mantari JL, Guedes Soares C. Finite element formulation of a generalized higher
HSDT3 1.3110 1.4037 1.0299 1.1115
order shear deformation theory for advanced composite plates. Compos Struct
2 Vo et al. [31] 1.5853 1.7540 1.2936 1.4450 2013;96:54553.
HSDT1 1.5897 1.7587 1.2946 1.4460 [6] Mantari JL, Guedes Soares C. Five-unknowns generalized hybrid-type quasi-3D
HSDT2 1.5897 1.7587 1.2946 1.4460 HSDT for advanced composite plates. Appl Math Modell 2015.
HSDT3 1.5899 1.7589 1.2946 1.4460 [7] Swaminathan K, Naveenkumar DT, Zenkour AM, Carrera E. Stress, vibration
and buckling analysis of FGM plates- a state of the art review. Compos Struct
5 Vo et al. [31] 1.9416 2.2080 1.6390 1.8779 2015;120:1031.
HSDT1 1.9474 2.2142 1.6404 1.8794 [8] Neves AMA, Ferreira AJM, Carrera E, Cinefra M, Roque CMC, Jorge RMN, et al.
HSDT2 1.9474 2.2142 1.6404 1.8794 Static, free vibration and buckling analysis of isotropic and sandwich
HSDT3 1.9476 2.2145 1.6401 1.8775 functionally graded plates using a quasi-3D higher-order shear deformation
theory and a meshless technique. Compos Part B 2013;44:65774.
10 Vo et al. [31] 2.1211 2.4242 1.8145 2.0818 [9] Carrera E, Brischetto S. Modeling and analysis of functionally graded beams,
HSDT1 2.1275 2.4311 1.8162 2.0835 plates and shells: part II. Mech Adv Mater Struct 2011;18:12.
HSDT2 2.1274 2.4310 1.8162 2.0835 [10] Zenkour AMA. A comprehensive analysis of functionally graded sandwich
HSDT3 2.1275 2.4312 1.8177 2.2805 plates: Part 1 Deflection and stresses. Int J Solids Struct 2005;42(18
19):522442.
Table 19 [11] Zenkour AMA. A comprehensive analysis of functionally graded sandwich
The maximum vertical displacement of FG sandwich CF beams (Type C). plates: Part 2 Buckling and free vibration. Int J Solids Struct 2005;42(18
19):524358.
p Theory a=h 5 a=h 20 [12] Neves AMA, Ferreira AJM, Carrera E, Cinefra M, Roque CMC, Jorge RMN, et al.
Free vibration analysis of functionally graded shells by a higher-order shear
1-2-1 2-2-1 1-2-1 2-2-1 deformation theory and radial basis functions collocation, accounting for
0 Vo et al. [31] 28.5524 28.5524 27.6217 27.6217 through-the-thickness deformations. Eur J Mech A/Solids 2013;37:2434.
HSDT1 28.5666 28.5666 27.6200 27.6200 [13] Fazzolari FA, Carrera E. Refined hierarchical kinematics quasi-3D Ritz models
HSDT2 28.5664 28.5664 27.6200 27.6200 for free vibration analysis of doubly curved FGM shells and sandwich shells
with FGM core. J Sound Vib 2014;333:1485508.
HSDT3 28.5674 28.5674 27.6201 27.6201
[14] Giunta G, Belouettar S, Carrera E. Analysis of FGM beams by means of classical
1 Vo et al. [31] 49.7281 53.6540 48.6985 52.5878 and advanced theories. Mech Adv Mater Struct 2010;17:62235.
HSDT1 49.7746 53.7037 48.7040 52.5938 [15] Giunta G, Crisafulli D, Belouettar S, Carrera E. Hierarchical theories for the free
HSDT2 49.7744 53.7035 48.7040 52.5938 vibration analysis of functionally graded beams. Compos Struct
HSDT3 49.7760 53.7053 48.7042 52.5940 2011;94:6874.
[16] Giunta G, Crisafulli D, Belouettar S, Carrera E. A thermo-mechanical analysis of
2 Vo et al. [31] 62.4386 69.7271 61.3855 68.6176 functionally graded beams via hierarchical modelling. Compos Struct
HSDT1 62.5085 69.8018 61.3974 68.6303 2013;95:67690.
HSDT2 62.5082 69.8013 61.3974 68.6303 [17] Arciniega RA, Reddy JN. Large deformation analysis of functionally graded
HSDT3 62.5103 69.8039 61.3977 68.6306 shells. Int J Solid Struct 2007;44:203652.
[18] Carrera E, Brischetto S, Cinefra M, Soave M. Effects of thickness stretching in
5 Vo et al. [31] 79.0288 90.5335 77.9869 89.3942
functionally graded plates and shells. Compos Part B 2011;42:123.
HSDT1 79.1260 90.6361 78.0071 89.4143 [19] Thai Huu-Tai, Nguyen Trung-Kien, Vo Thuc P, Lee Jaehong. Analysis of
HSDT2 79.1257 90.6350 78.0071 89.4143 functionally graded sandwich plates using a new first-order shear
HSDT3 79.1282 90.6386 77.9993 89.3764 deformation theory. Eur J Mech A/Solids 2014;45:21125.
10 Vo et al. [31] 87.4501 100.3340 86.4264 99.1718 [20] Thai Huu-Tai, Kim Seung-Eock. A simple higher-order shear deformation
HSDT1 87.5594 100.4478 86.4505 99.1954 theory for bending and free vibration analysis of functionally graded plates.
Compos Struct 2013;96:16573.
HSDT2 87.5590 100.4461 86.4505 99.1953
[21] Thai Huu-Tai, Vo Thuc P. A new sinusoidal shear deformation theory for
HSDT3 87.5598 100.4479 86.4841 99.4892
bending, buckling, and vibration of functionally graded plates. Appl Math
Modell 2013;37:326981.
[22] Apetre NA, Sankar BV, Ambur DR. Analytical modeling of sandwich beams
2 3 with functionally graded core. J Sandwich Struct Mater 2008;10.
@ 2 H1 @ 2 H2 @ 2 H3 @ 2 H4 [23] Reddy NA. A simple higher-order theory for laminated composite plates. J Appl
0 0 0 0 0 0 0 0 0 0
6 @x2 @x2 @x2 @x2 7 Mech 1984;51:74552.
B2 4 0 0 0 0 0 0 0 0 0 0 0 0 0 0 5; [24] Zhu H, Zankar BV. A combined fourier series-Galerkin method for the analysis
of functionally graded beams. J Appl Mech 2004;71(3):4214.
0 0 0 0 0 0 0 0 0 0 0 0 0 0
[25] Frostig Y, Baruch M, Vinay O, Sheinman I. High-order theory for sandwich
beams behaviour with transversely flexible core. J Eng Mech 1992;118
2 3 (5):102643.
0 0 0 0 0 0 0 0 0 0 0 0 0 0 [26] Filippi M, Carrera E, Zenkour AM. Static analyses of FGM beams by various
3 6 0 7
B 40 0 0 0 0 0 0 0 0 0 0 0 0 5; theories and finite elements. Compos Part B 2015;72:19.
@H1 @H2 @H3 @H4 [27] Jacques I, Judd C. Numerical analysis. London: Chapman & Hall; 1987.
0 0 0 0 0 0 0 0 0 0 @x @x @x @x [28] Morris JL. Computational methods in elementary numerical
analysis. Chichester: John Wiley & Sons; 1983.
2 3 [29] Spanier J, Oldham KB. An atlas of functions. London: Hemisphere; 1987.
0 0 0 0 0 0 0 0 0 0 0 0 0 0
[30] Augarde CE. Generation of shape functions for straight beam elements.
6 0 0 0 0 0 0 07
B4 4 0 0 0 0 0 0 0 5; Comput Struct 1998;68(6):55560.
@H1 @H2 @H3 @H4 [31] Vo P Thuc, Thai Huu-Tai, Nguyen Trung-Kien, Inam Fawad, Lee Jaehong. Static
0 0 0 @x @x @x @x
0 0 0 0 0 0 0 behaviour of functionally graded sandwich beams using a quasi-3D theory.
Compos Part B 2015;68:5974.
2 3
0 0 0 0 0 0 0 0 0 0 0 0 0 0 [32] Reddy JN, Liu CF. A higher-order shear deformation theory of laminated elastic
6 7 shells. Int J Eng Sci 1985;23:31930.
B5 4 0 0 0 0 0 0 0 0 0 0 H1 H2 H3 H4 5; [33] Reddy JN. Mechanics of laminated composite plates: theory and analysis. 2nd
ed. Boca Raton (FL): CRC Press; 2004.
0 0 0 0 0 0 0 0 0 0 0 0 0 0

View publication stats