Author's Accepted Manuscript

Nonlinear free vibration analysis of point

supported laminated composite skew plates
A. Naghsh, M. Azhari

www.elsevier.com/locate/nlm

PII: S0020-7462(15)00109-2
DOI: http://dx.doi.org/10.1016/j.ijnonlinmec.2015.05.008
Reference: NLM2496

To appear in: International Journal of Non-Linear Mechanics

Received date: 20 December 2014

Revised date: 7 April 2015
Accepted date: 22 May 2015

Cite this article as: A. Naghsh, M. Azhari, Nonlinear free vibration analysis of
point supported laminated composite skew plates, International Journal of Non-
Linear Mechanics, http://dx.doi.org/10.1016/j.ijnonlinmec.2015.05.008

This is a PDF file of an unedited manuscript that has been accepted for
publication. As a service to our customers we are providing this early version of
the manuscript. The manuscript will undergo copyediting, typesetting, and
review of the resulting galley proof before it is published in its final citable form.
Please note that during the production process errors may be discovered which
could affect the content, and all legal disclaimers that apply to the journal
pertain.
 Nonlinear free vibration analysis of point supported laminated composite
skew plates

A. Naghsh, M. Azhari*

Department of Civil Engineering, Isfahan University of Technology, 84156-83111 Isfahan, Iran

Abstract

The element-free Galerkin (EFG) method is employed to analyze the large amplitude free vibration of

point supported laminated composite skew plates. The geometrical nonlinearity is considered based

on the von Karman’s assumptions and the point support boundary conditions are satisfied through the

use of Lagrange multiplier method and orthogonal transformation technique. Assuming a periodic

solution and applying the weighted residual method, the nonlinear governing equations are used, and

the problem is solved by direct iteration technique. The obtained results are first validated against the

available solutions of simply supported and clamped plates, and then the point supported plates are

dealt with. To this end, four types of point support arrangements are considered and nonlinear

frequency ratios for different numbers of point supports, amplitude ratios, skew angles, aspect ratios

and fiber orientations are calculated which may serve as benchmarks for future research.

Keywords: Nonlinear free vibration; Point supported skew plates; Laminated composite plates;

Element-free Galerkin method.

*
Corresponding author. Tel.: +98 3133913804; fax: +98 3133912700.
E-mail address: mojtaba@cc.iut.ac.ir (M. Azhari).

1
1. Introduction

Laminated composite skew plates are widely used as the main components in
aerospace, marine, and other modern industries. These materials allow high-strength
structures with a minimum weight, thus forming thin plate components that are prone
to large amplitude oscillations. Analysis of large amplitude free vibration of these
plates has proven practically important in structural designs, and has recently received
increasing attention of the researchers.
During vibration of a plate, by increasing the deflection, tensile membrane stresses
develop in the thickness of the plate, which makes the plate stiffer and results in
changing the frequencies and mode shapes. Consequently, in large amplitude
vibrations, we have to deal with a nonlinear problem, where the frequencies are
amplitude-dependent.
Most studies reported on the nonlinear vibration of laminated composite plates
have been concerned with rectangular shaped plates, and have employed different
methods for deriving and solving the equations of motion. However, a few works can
also be found in the current literature on the large amplitude vibration of laminated
composite skew plates, some of which are briefly touched on below. As a case in
point, Singha and Ganapathi [1] dealt with nonlinear free vibration of thin composite
skew plates using finite element method. They investigated the effects of boundary
conditions and various geometrical parameters of the plate on its frequency and mode
shape. Singha and Daripa [2] reported large amplitude free vibration analysis of
composite skew plates based on using 3 different approaches, namely 1) satisfying the
equation of motion at the point of maximum displacement, 2) taking the weighted
residual of the nonlinear governing equation over a period, and 3) performing time
history analysis. They showed that the frequencies obtained by second approach are in
good agreement with those obtained from time history analysis and other available
analytical solutions. Malekzadeh [3] presented nonlinear frequencies of laminated
composite skew thin plates having simply supported, clamped and mixed boundary

2
conditions by using a differential quadrature method. Later Malekzadeh extended the
differential quadrature method to free vibration analysis of moderately thick laminated
skew plates based on the first order shear deformation theory [4].
Based on the available literature, all the works addressing the large amplitude free
vibration of the composite rectangular or skew plates encompass the simply supported
and clamped boundary conditions, plates with edges elastically restrained, or plates
resting on elastic foundation. However, despite having practical significance,
composite plates having edge or interior point supports have not been deservedly
investigated through the large amplitude free vibration studies. It must be noted again
that point supports are used in many cases to connect structural components, or are
employed as an additional fulcrum in the plates having edge support to increase their
loading capacity.
The linear vibration of point supported composite plates has been studied by
several researchers. Narita and Iwato [5] presented free vibration analysis of
composite circular and elliptical plates resting on elastic or rigid point supports by
employing the Ritz method. Liew [6] employed the Rayleigh-Ritz method and the
Lagrange multiplier technique for free vibration analysis of angle-ply thin trapezoidal
plates over point supports. Abrate [7] analyzed the free vibration of rectangular
laminated composite plates with internal line supports and discrete point supports
using the Rayleigh-Ritz method and the Lagrange multiplier technique. On the basis
of three-dimensional elasticity considerations, Ye and Soldatos [8] studied free
vibration of simply supported laminated plates and cylindrical panels with lateral
surfaces point supports. They showed that for very thin plates on point supports, the
two-dimensional classical theory provides very good results in accordance with three-
dimensional analysis. Cheung and Zhou [9] investigated the free vibration of
rectangular laminated composite plates with interior and edge point supports by using
the Rayleigh-Ritz approach and employing a set of admissible functions. Setoodeh
and Karami [10] reported a layer-wise finite element method to study the static, free
vibration and buckling analysis of rectangular thick composite plates with elastic line
3
and point supports. Kong [11] carried out free vibration analysis of rectangular thin
isotropic and composite plates with various boundary conditions, including elastic
point supported conditions by using the Rayleigh-Ritz approach. Bahmyari and
Khedmati [12] studied free vibration of nonhomogeneous moderately thick
rectangular plates with point supports resting on elastic foundation by using Element-
Free Galerkin Method. They employed Penalty method for imposing essential
boundary conditions.
In recent years, mesh-free methods have been developed and successfully applied
to the problems of solid mechanics, and the problems dealing with large deflection of
plates. In mesh-free methods, the field approximation is based on a set of nodes
scattered on the problem domain without any need for mesh generation or
connectivity information among nodes. These methods have made it possible to
overcome some drawbacks of finite element method (FEM), for example mesh
distortion in large deformation problems and difficulties in remeshing of problems
with moving discontinuities [13, 14]. Mesh-free methods also suffer from their own
drawbacks, i. e., construction shape functions is more time-consuming in contrast to
FEM, and there are difficulties in handling essential boundary conditions. However,
they often yield more accurate results and provide higher rate of convergence [15].
One of the popular approaches in the class of mesh-free numerical methods is the
element-free Galerkin (EFG) method developed by Belytschko et al. [15] . In the EFG
method, the field approximation is based on the moving least squares (MLS)
approach, and essential boundary conditions are introduced into the analysis by means
of Lagrange multipliers technique. This method has been used to solve various
problems of shell and plate structures in the literature [16-18]. One of the advantages
of EFG method in solving thin plate problems is that the derivatives of the
displacement field do not appear as the unknown parameters in the analysis and thus
the size of the total problem is reduced. This issue will become more important in
problems that require iterative procedures.

4
 In this paper, large amplitude vibration characteristics of thin laminated composite
skew plates on point supports are investigated. The formulation is based on the
classical plate theory in conjunction with the von Karman strain–displacement
displacement
assumptions. The nonlinear equations of motion are solved using the EFG method
through employing direct
irect iteration technique. Some numerical examples of
rectangular and skew plates are presented for different
erent point support conditions and
the effects of number of point supports, skew angle, fiber orientation and aspect ratio
on the nonlinear frequency of laminated skew plates are reported.

2. Formulation

The plate considered here is a thin symmetrically laminated skew plate with length
a and width b, resting on a set of arbitrarily located point supports (Fig. 1). The plate
is composed of several orthotropic layers with the same thickness and mechanical
properties. The fiber orientation in each layer is indicated by angle θ with respect to
the x axis.

Fig. 1. The skew plate supported at discrete points.

Based on the thin classical plate theory and considering the von Karman large
deformation assumption, the strain of the plate may be defined in the following form

5
  u0, x   w, x 2   − w, xx 
  1 2   
ε = ε 0 L + ε 0 NL + zκ =  v0, y  +  w, y  + z  − w, yy  . (1)
u + v  2 2 w w  −2 w 
 0, y 0, x   ,x , y   , xy 

In the above relation, u0 , v0 are in-plane displacements of the plate mid-surface in

x and y directions and w is out of plane displacement. ( ), x and ( ), y represent the
partial differentiation with respect to x and y, respectively. ε 0L and ε0NL are linear and

nonlinear part of the in-plane axial strain, respectively, and κ is the curvature change.
For a symmetric laminated composite plate, the constitutive relations are
 N   A 0  ε 0 L + ε 0 NL 
 =  . (2)
M   0 D   κ 
T T
where N = N x N y N xy is the vector of in-plane forces, and M = M x M y M xy is the

vector of bending and twisting moments defined per unit length. A and D are the
axial and flexural rigidity matrices defined according to the properties and
arrangement of layers through the thickness, and may be found in [19].
The equations of motion for large deformation free vibration of the plate may be
derived by using principle of virtual work. The virtual work done by internal forces
and inertia forces are denoted by Ws and Wk , respectively, and are expressed as

Ws = ∫ (δ ε 0 L + δ ε 0 NL ) T A (ε 0 L + ε 0 NL ) dA + ∫ (δ κ ) T D(κ ) dA (3)
A A

Wk = ∫ ρ h(δ u0u

0 + δ v0v

0 + δ ww

) dA (4)
A

where, ρ is the mass per unit volume, h is the total thickness of the plate, A is the area
and dots denote differentiation with respect to time t. After approximation of
displacement field by EFG method and using strain equations (1), the virtual works
may be expressed in a discretized form in terms of nodal displacement parameters. By
setting the sum of the virtual works equal to zero and considering the virtual
displacement to be arbitrary, the following system of equations of motion will be
derived

6


= 0
(K L + K1 NL + K2 NL )∆ + (M )∆ (5)

where, M is the mass matrix, K L is the linear stiffness matrix which is constant
during vibration, and K1NL and K2 NL are time-dependent nonlinear stiffness matrices

which are linear and quadratic function of displacement w, respectively. These

stiffness matrices are presented in Appendix A. ∆ is the vector of nodal displacement
parameters which contains the unknown coefficients of ( u0i , v0i , wi ) per each node i.

Considering a periodic vibration with only one harmonic, the vector ∆ may be written
as ∆ = ∆ sin ωt , where ω is the frequency of vibration and the vector ∆ contains
value of the nodal displacement parameters at the instant maximum displacement
occurs. By this assumption, Eq. (5) may be rewritten as
 (K L + K1
  2 2

NL sin ω t + K2 NL sin ω t ) − ω M  ∆ sin ω t = 0 (6)

To eliminate the time parameter t from Eq. (6), the weighted residual method [2] is
employed. To this end, the left hand side of Eq. (6) is defined as the residual vector,
R , and the sin ωt term as the weight function. Then, the integral of the R sin ωt in
the timeframe [0, T / 4] is set to be zero (Eq. (7)).
T /4

∫ R sin ωt dt = 0 (7)
0

where T = 2π / ω is the period of vibration. This method gives a set of equations of

motion in the frequency domain as follows
8  3 
[(K L + K1 NL + K2 NL ) − ω 2M ]∆ = 0
 3π  4  (8)
K

Eq. (8) represents an eigenvalue problem including the effect of change of the

plate’s stiffness during vibration. K1 
and K2 are nonlinear stiffness matrices of
NL NL

order one and two, respectively, and the mass matrix, M , includes the in-plane inertia
effects.

7
3. EFG method

The element-free Galerkin method was introduced by Belytschko et al. [15]

through making some modifications to the diffuse element method (DEM) [20]. In
this method, a set of scattered nodes are used to represent the problem domain and
boundaries. The moving least squares (MLS) method is employed to construct shape
functions. The required integrations are performed by a Gauss quadrature scheme
through using a mesh of cells called background mesh, and essential boundary
conditions are imposed via Lagrange multipliers technique.

3.1. MLS approximations

MLS is a general method to approximate a function from a set of scattered data.

This method is employed by researchers to produce shape functions in different
numerical methods. In this method, the approximation displacement, u h ( x) , of the
displacement u ( x) at the arbitrary point x T = (x, y), is defined by
m
u h ( x ) = ∑ p j ( x )a j ( x ) = p T ( x)a ( x ) (9)
j =1

where p(x) is polynomial basis here taken as

p T(x)= [1, x, y , x 2 , xy , y 2 ] m=6 (10)

a( x) is the unknown coefficient vector whose entries are functions of x . In order to

obtain a( x), a functional of weighted square of residual is defined as follows
n 2
(
J = ∑ w( x − x I ) p T ( x I ) a ( x ) − u I
I =1
) (11)

where w( x − x I ) is weight function, u I is the nodal parameter at node I, and n is the

number of nodes in a local domain in the neighborhood of x called the domain of

influence of x (Fig. 2).

8
 Fig. 2. Mesh-free
free model of the skew plate and the influence domain of a quadrature point X
X.

The unknown vector a (x) is determined by minimizing the functional J, which

leads to the following linear equation system

A(x)(
a x)= B(x)u (12)

A(x) and B(x) are matrices with dimensions m × m and m × n , respectively

respectively, defined
as
n
A(x)= ∑ w( x − x I )p ( x I )p T ( x I ) (13)
I =1

B(x)= [ B1(x), B 2(x), ... , B n(x)] , B I(x)= w( x − x I )p ( x I ) (14)

u T = [u1 , u2 , ... , un ] (15)

Calculating a( x) from Eq. (12)

(1 and substituting into Eq. (9) results in

A −1(x)B(x)u = Φ(x)u
u h ( x) = pT ( x) 
(16)
a(x)

in which Φ(x) is vector of shape functions. The shape function correspond

corresponding to each
node I may be written as
ΦI(x)= p T ( x ) A −1(x)B I(x) (17)

The first and second order partial derivatives of the shape functions
function appear in the
formulation of stiffness matrix and are required to be calculated. To this end, ffrom Eq.
(16)) the vector of shape function is written in the form of

9
Φ(x)= p T ( x ) A −1(x)B(x)= γ T ( x)B(x)
 (18)
γT ( x)

The first and second order partial derivatives of shape function Φ(x) are obtained
as

Φ,i(x)= γ T ,i ( x )B(x)+ γ T ( x)B ,i(x) (19)

Φ,ij(x)= γ T ,ij ( x )B(x)+ γ T ,i ( x)B , j(x)+ γ T , j ( x )B ,i(x)+ γ T ( x )B ,ij(x) (20)

where the comma-subscript represents partial differentiation, and i and j indicate

coordinates x or y. The partial derivatives of γ ( x) are needed in Eqs. (19, 20). From
Eq. (18), we have
A(x)γ ( x) = p( x) (21)

and the partial derivatives of γ ( x ) may be calculated from Eqs. (22, 23).

A(x)γ ,i ( x) = p,i ( x) − A ,i(x)γ ( x) (22)

A(x)γ ,ij (x) = p ,ij ( x) − A ,i(x)γ , j ( x) − A , j(x)γ ,i (x) − A ,ij(x)γ ( x) (23)

3.2. Weight function and integration scheme

Weight functions directly affect the ability of shape functions in approximation of

displacement field. An improper choice of weight function, may lead to inaccurate
results and convergence problems. The weight function should be positive and
monotonically decreasing in the domain of influence of point x , which causes the
nodes closer to x in Eq. (11) to have higher weight [21]. Also, the order of continuity
of displacement derivatives depends on the characteristics of weight functions.
Various exponential and polynomial weight functions are proposed and used by
researchers. The cubic, quartic, and seventh order splines are the most common
weight functions in the literature, each of which has a different shape and supplies a
certain order of continuity.

10
 In this paper, a circular influence domain with the seventh-order spline weight
function [22] which provides C3 continuity is used.
1 − 4.7 r 2 + 12r 4 − 10r 5 + 0.5r 6 + 1.2r 7 0 ≤ r ≤ 1
w(x − x I ) = w( r ) =  (24)
 0 r >1

r is normalized distance defined as

x − xI
r= (25)
R
where x − x I is the distance of two points x and x I , and R is radius of influence

domain which controls the number of nodes used to approximate the displacement
field at point x. This parameter is usually defined as R = β d where β is a scaling
parameter and d is the average distance between nodes in x and y directions. β
should be chosen so that sufficient number of nodes fall into the influence domain of
point x , and the matrix A( x) is invertible.
The usual numerical integration in the EFG method is based on a Gauss quadrature
scheme which is implemented through the development of a mesh of cells around the
physical domain of the plate. In this method, only those Gauss points that fall within
the physical domain are taken into account. In the present study, to keep the total
number of used Gauss points under control, a mapping scheme is employed. The
physical coordinate system x-y is transformed into the natural coordinate system
ξ − η by employing the following transformation equations
4 4
1
x = ∑ xi N i (ξ ,η ), y = ∑ yi N i (ξ ,η ), N i (ξ ,η ) = (1 + ξiξ )(1 + ηiη ) (26)
i =1 i =1 4

where xi , yi are Cartesian coordinates and ξi ,ηi are natural coordinates of the point i

(Fig. 3). The integral of arbitrary function f(x,y) in the physical coordinate may be
expressed in the natural coordinate system by
1 1

∫A f ( x, y )dxdy = ∫ ∫ f (ξ ,η ) J dξ dη (27)
−1 −1

11
  ∂x ∂y 
 ∂ξ ∂ξ 
J (ξ ,η ) =  
(28)
 ∂x ∂y 
 ∂η ∂η 
 

where J is determinant of the Jacobian matrix J and is generally a function of ξ

and η ; but, for a skew plate,
plate it will be simplified to J = 0.25ab cos φ . This

transformation enables us to perform the numerical Gauss integration scheme over a

simple squared area. Therefore,
Therefore the background mesh is first constructed in the natural
coordinate system and the integral of Eq. (27) is stated as
1 1 NG NG

∫∫ f (ξ ,η ) J dξ dη = ∑ wi f (ξi ,ηi ) J i = ∑ wi f ( xi , yi ) J i (29)

−1 −1 i =1 i =1

In the above equation, NG is the total number of Gauss points and wi is weight

coefficient of Gauss point i. The natural coordinate system is used to determine the
location of Gauss points and their weight coefficients. Then, the coordinate of each
Gauss point is specified in the physical domain according to Eq. (26)
( and the MLS
shape functions are calculated in the physical domain based on the physical
coordinates off nodes and a circular influence domain (Fig. 4).

Fig. 3.. Mapping of a quadrilateral plate into natural coordinate system.

12
 Fig. 4. A schematic design of background mesh for quadrature points in physical and natural coordinate systems.

3.3. Boundary conditions

Using MLS method to approximate the displacement field makes the nodal
concept i.e. u I ≠ u h ( x I ) . Therefore, the essential
parameters u I have no physical concept,

boundary conditions cannot be satisfied in the same way as in the conventional finite
element method. In the method of Lagrange multipliers, the essential boundary
conditions are imposed by defining the following weak form equation

∫ (δ λ
T
d)ds = 0 (30)
Sd

where λ is vector of the Lagrange multipliers and a function of coordinates, and d is

the vector of displacements and rotation to be satisfied, which is given by
λ T = [λu , λv , λw , λφ ]
(31)
d T = [u , v , w , ϕ ]
In Eq. (31), ϕ = ∂w / ∂n is rotation about the tangent to the boundary and n
indicates normal to the essential boundary, S d . The Lagrange multipliers at any point

of essential boundary are interpolated using linear Lagrange polynomials as follows

λ = ∑ NI λ I
I ∈S
(32)
d

where I represents the nodes on the essential boundary and N I is the matrix of shape

functions. Here, linear polynomials

ynomials are used to interpolate Lagrange multipliers on the
boundary.

13
 Substituting the approximated form of displacements and Lagrange multipliers
from Eq.(16) and Eq.(32) into Eq.(30), and considering ∆ as the vector of nodal
parameters, the following system of equations is derived

H (4 NB×3 N ) ∆ (3 N ×1) = 0 (33)

where N is the number of total points and NB is the number of constraint points. H is
a singular matrix composed of shape functions of displacements and Lagrange
multipliers, which for the clamped boundaries may be expressed as
ΦJ 0 0 
0 ΦJ 0 
H IJ = ∫ NI  dS (34)
Sd
0 0 ΦJ 
 
0 0 ΦJ , n 

For simply supported boundaries, the parameter ΦJ ,n in Eq. (34) is replaced by 0.

For the case of point supports, the integral on essential boundaries in Eq. (30) is
converted to summation on the point supports, and Lagrange multipliers are defined
directly at each point support without any need for interpolation. Eq. (33) should be
considered as a constraint for Eq. (8). This is implemented by a transformation
technique. Matrix H is written in the following form by using the singular value
decomposition technique [21].
Σ 0
H (4 NB×3 N ) = U (4 NB×4 NB )  r×r  V(3T N ×3 N ) (35)
 0 0 (4 NB×3 N )

where r is rank of H , Σ is a diagonal matrix containing singular values of H , and U

and V are orthogonal matrices. Matrix V is written as
V T = {V3 N ×r , V3 N ×(3 N − r ) }T (36)

Using the transformation,

∆ = V3 N ×(3 N − r ) ∆ , (37)

constraint Eq. (33) is satisfied and the following eigenvalue equation is obtained

(K − ω 2 M )∆ = 0 (38)

14
where K and M are
K = V T (3 N − r )×3 N K V3 N ×(3 N − r ) (39)

Μ = V T (3 N − r )×3 N Μ V3 N ×(3 N − r ) (40)

Direct iteration technique [23] is employed here to obtain the fundamental

nonlinear frequency of the plate under a specified amplitude of vibration (
wmax = α × h ). At first, by ignoring the nonlinear stiffness matrices, the eigenvalue

problem of Eq. (38) is solved and the fundamental linear frequency and its associated
eigenvector ( ∆ and consequently ∆ ) are determined. Thereafter, the eigenvector is
scaled in such a way that the maximum out of plane displacement, wmax , reaches the

predefined amplitude, α × h . Then, this eigenvector and the EFG shape functions are
used to calculate the out of plane displacements at the Gauss points and subsequently
the nonlinear stiffness matrices are formed. These matrices are used to update the total
stiffness matrix, and calculate new frequency and eigenvector. New scaled
eigenvector is employed again to form new nonlinear stiffness matrices which in turn
result in another eigenvector. This iterative process continues until in the two
consecutive iterations, the relative difference of the two frequencies is less than a
certain value, taken as 10−6 in this paper.

4. Numerical examples and discussion

The EFG method is applied to obtain the fundamental nonlinear frequency of

various skew plates with point supports. The radius of influence domain is considered
equal to five times the average distance between the points in x any y directions. A
8 × 8 background mesh with a 4 × 4 Gauss integration scheme is used to obtain the
mass and stiffness matrices.
Four types of point support arrangements illustrated in Fig. 5 are investigated. It is
assumed that point supports impose zero displacement conditions ( u0 = v0 = w = 0 ).

15
 Fig. 5.. Four types of point support arrangements.

The mechanical properties of composite materials used in this paper are as follows
EL / ET = 40, GLT / ET = 0.6, GTT / ET = 0.5, ν LT = 0.25

where E, G and ν are the Young’s modulus, shear modulus and Poisson ratio
ratio,
respectively, and subscripts L and T denote the longitudinal and transvers directions of
fibers, respectively.
To the best of the authors’ knowledge, there are no available results in the open
literature for large amplitude vibration
vibr of point supported skew plates. So, to ensure
the validity and accuracy of the present method, examples of nonlinear free vibration
of simply supported and clamped isotropic and laminated plates, and an example of
linear vibration of point supported plate are analyzed at first. Thereafter
Thereafter, nonlinear
free vibration of point supported plates is performed and different point support
patterns, skew angles, and aspect ratios are considered in the analysis.

16
4.1. Simply supported and clamped plates

Example of large amplitude free vibration of isotropic square plate has been
widely studied in the literature to investigate the validity and accuracy of the results
obtained by different methods. Table 1 presents the fundamental non-dimensional
linear frequency Ω L , and the nonlinear frequency ratio ωNL / ωL for different
amplitude ratios ( wmax / h ) along with other results for the fully simply supported and

clamped boundary conditions. The convergence of the results is also shown for
various node densities. The results are in good agreement with those published
previously by finite element method [24], spline finite strip method [25], and
analytical solution [26]. Also, a 13 × 13 node density is adequate to achieve a desirable
convergence.

Table 1
*
The fundamental non-dimensional linear frequency ( Ω L = ωL a 2 ρ h / D ) and nonlinear frequency ratios ( ω NL / ωL ) for a
isotropic square plate ( ν = 0.3 )
Node wmax / h
ΩL
density 0.2 0.4 0.6 0.8 1
Simply supported plate
9×9 19.7299 1.0199 1.0777 1.1679 1.2843 1.4208
11 × 11 19.7345 1.0200 1.0780 1.1686 1.2855 1.4226
Present study
13 × 13 19.7347 1.0200 1.0779 1.1684 1.2851 1.4219
15 × 15 19.7359 1.0200 1.0779 1.1683 1.2850 1.4218
FEM [24] - - 1.0197 1.0767 1.1659 1.2800 1.4167
FSM [25] - - 1.0197 1.0768 1.1662 1.2813 1.4173
Analytical solution [26] - - 1.0196 1.0766 1.1666 1.2840 1.4242
Clamped plate
9×9 36.0525 1.0072 1.0283 1.0625 1.1085 1.1650
11 × 11 36.0234 1.0074 1.0293 1.0648 1.1124 1.1705
Present study
13 × 13 35.9926 1.0073 1.0290 1.0641 1.1113 1.1690
15 × 15 36.0015 1.0074 1.0290 1.0642 1.1114 1.1691
[24] - - 1.0072 1.0286 1.0632 1.1098 1.1669
[25] - - 1.0073 1.0287 1.0634 1.1100 1.1672
3 2
* D = Eh / 12(1 − ν )

17
Another example is dedicated to the nonlinear free vibration of five-layered angle-ply
[45° / − 45° /45° / −45° /45° ] thin rhombic plate with fully simply supported and

clamped boundary conditions. The present results are compared with the existing
results of References [1-3, 24] in Table 2. It should be noted that, in References [1, 3],
the equations of motion at the instant of maximum displacement are used to calculate
the natural frequency. This means that variations of nonlinear stiffness during
vibration is not included in the analysis and thus higher nonlinear frequency ratios are
expected. Therefore, to make a comparison with the results of References [1, 3], the
parameter sin ωt in Eq. (6) is kept to 1 and the coefficients 8 / 3π and 3 / 4 in Eq. (8)
are replaced with 1. It may be concluded from the obtained results that, for higher
nonlinear frequency ratios, the relative difference between the two mentioned
approaches will be increased. Also, the present nonlinear frequency ratios have good
conformity with other results, especially with those of References [3, 24].

Table 2
The nonlinear frequency ratios ( ωNL / ωL ) for angle-ply [45° / − 45° /45° / −45° /45° ] rhombic plates.
Skew Skew angle wmax / h
angle 0.2 0.4 0.6 0.8 1
Simply supported plate
Present study 1.0150 1.0589 1.1283 1.2191 1.3269
[24] 1.0147 1.0578 1.1260 1.2154 1.3219
0 [2] 1.0143 1.0562 1.1229 1.2102 1.3150
Present study* 1.0197 1.0767 1.1658 1.2806 1.4149
[1]* 1.0190 1.0740 1.1603 1.2720 1.4034
Present study 1.0226 1.0875 1.1877 1.3153 1.4631
[2] 1.0215 1.0833 1.1796 1.3029 1.4466
30 Present study* 1.0295 1.1132 1.2404 1.3994 1.5807
[3]* 1.0293 1.1128 1.2395 1.3981 -
[1]* 1.0284 1.1095 1.2332 1.3884 1.5661
Clamped plate
Present study 1.0094 1.0372 1.0819 1.1414 1.2135
[2] 1.0099 1.0387 1.0849 1.1458 1.2215
45 Present study* 1.0123 1.0484 1.1059 1.1816 1.2723
[3]* - 1.0487 - 1.1827 1.2740
[1]* 1.0124 1.0488 1.1067 1.1832 1.2749

18
*
The equations of motion at the instant of maximum displacement.
displacement
Results of References [2, 24] based on a shear deformable finite element approach for thickness ratio a/h=100.

4.2. Point supported skew plate

The free vibration analysis of point supported rhombic plates is carried out to
investigate the validity of the present method for the skew plates with this type of
boundary condition. The first five non-dimensional
non linear frequencies of the plates
supported at four and eight equally spaced points on the edges are reported in Table 3
for different skew angles. The frequencies match well with the results reported earlier
by Mizusawa [27] and Abrate [28].

Table 3
dimensional natural frequencies of the point supported rhombic plates ( Ω = ωL a 2 ρ h / D )
The first five non-dimensional

Point support
Skew angle Ω1 Ω2 Ω3 Ω4 Ω5
condition a* 7.1131 15.7729 15.7729 19.5966 38.4582
0 [27]] 7.111 15.77 15.77 19.60 38.43
[28]] 7.1109 15.7703 15.7703 19.5961 -
*
a 9.10481 11.4027 22.6508 25.1824 30.2967
30
[27]] 9.104 11.40 22.65 25.18 30.29
*
a 10.2086 11.8203 27.764 29.7824 33.3507
45
[27]] 10.21 11.82 27.76 29.78 33.34
*
a 9.43431 14.6641 36.2774 39.3177 39.9546
60
[27]] 9.433 14.66 36.29 39.32 39.94
*
a 17.8757 34.9124 34.9124 38.3974 60.2422
0 [27]] 17.91 34.99 34.99 38.43 60.27
[28]] 18.150 35.3580 35.3580 38.4323 60.9431-
a* 19.9089 42.0237 42.8600 44.9505 48.9842
30
[27]] 19.94 42.21 42.96 44.98 49.10
*
a 24.3401 42.8726 47.8954 51.4239 61.2479
45
[27]] 24.37 42.88 48.13 51.98 61.39
*
a 35.3364 39.5013 52.3898 57.5381 80.7762
60
[27]] 35.41 39.57 52.94 58.46 81.24
*
a = Present study ( 15 × 15 node density)

19
4.3. Point supported skew plate with point arrangement type (I)

The point support arrangement (I) is applied to obtain fundamental linear and
nonlinear frequencies of four layered composite cross-ply
cross [0° /90° /90° /0° ] and angle-
ply [45° / − 45° / − 45° /45° ] skew plates ( a / b = 1 ). The problem is solved for different
d
numbers of point supports, skew angles and amplitude ratios ( wmax / h ), and the

results are presented in Tables 4 and 5 together with the results of simply supported
plate. The mode shape contours of the cross-ply
cross and angle-ply
ply skew plates with 12
ply simply supported plate at the amplitude ratio wmax / h = 1
point supports and angle-ply

are shown in Figs. 6 and 7, respectively.

respectively

Table 4
dimensional linear frequency ( Ω L = ωL a 2 ρ / ET / h ) and nonlinear frequency ratios ( ωNL / ωL ) of
The fundamental non-dimensional
cross-ply [0° / 90° / 90° / 0° ] skew plates with point support type (I).
Number of point Skew wmax / h
ΩL
support angle 0.2 0.4 0.6 0.8 1
0 16.2982 1.0119 1.0468 1.1027 1.1768 1.2663
30 17.9006 1.0122 1.0478 1.1040 1.1777 1.2650
45 20.8939 1.0132 1.0517 1.1122 1.1907 1.2830
0 18.2009 1.0116 1.0458 1.1003 1.1723 1.2586
30 20.9319 1.0130 1.0508 1.1110 1.1898 1.2839
45 26.7904 1.0142 1.0556 1.1208 1.2056 1.3055
0 18.6290 1.0149 1.0584 1.1277 1.2182 1.3265
30 21.6493 1.0162 1.0635 1.1380 1.2351 1.3497
45 28.2397 1.0171 1.0666 1.1442 1.2445 1.3622
0 18.7651 1.0174 1.0679 1.1477 1.2511 1.3738
30 21.9118 1.0188 1.0734 1.1590 1.2691 1.3986
45 28.7489 1.0208 1.0808 1.1736 1.2920 1.4296
0 18.8177 1.0182 1.0710 1.1540 1.2617 1.3893
30 22.0232 1.0208 1.0810 1.1748 1.2948 1.4353
45 28.9819 1.0229 1.0885 1.1897 1.3183 1.4669
0 18.8919 1.0319 1.1227 1.2607 1.4335 1.3443
30 22.2347 1.0355 1.1356 1.2858 1.4721 1.3911
45 29.3300 1.0426 1.1607 1.3341 1.5431 1.4385

Table 5
dimensional linear frequency ( Ω L = ωL a 2 ρ / ET / h ) and nonlinear frequency ratios ( ωNL / ωL ) of
The fundamental non-dimensional
angle-ply [45° / −45° / −45° / 45° ] skew plates with point support type (I).

20
Number of point Skew wmax / h
ΩL
support angle 0.2 0.4 0.6 0.8 1
0 17.8368 1.0113 1.0438 1.0943 1.1587 1.2332
30 19.5328 1.0088 1.0345 1.0759 1.1312 1.1979
45 20.8939 1.0132 1.0517 1.1122 1.1907 1.2830
0 20.7066 1.0098 1.0386 1.0846 1.1451 1.2173
30 21.7889 1.0121 1.0475 1.1038 1.1792 1.2661
45 26.7904 1.0142 1.0556 1.1208 1.2056 1.3055
0 21.6606 1.0113 1.0444 1.0971 1.1668 1.2497
30 22.5666 1.0151 1.0589 1.1280 1.2180 1.3241
45 28.2397 1.0171 1.0666 1.1442 1.2445 1.3622
0 22.0245 1.0120 1.0470 1.1027 1.1763 1.2635
30 22.9255 1.0174 1.0679 1.1470 1.2493 1.3690
45 28.7489 1.0208 1.0808 1.1736 1.2920 1.4296
0 22.2006 1.0134 1.0527 1.1150 1.1962 1.2932
30 23.1069 1.0186 1.0725 1.1567 1.2651 1.3913
45 28.9819 1.0229 1.0885 1.1897 1.3183 1.4669
0 22.4473 1.0186 1.0724 1.1561 1.2636 1.3891
30 23.4902 1.0318 1.1214 1.2559 1.4216 1.6083
45 29.3300 1.0426 1.1607 1.3341 1.5431 1.4385

Fig. 6. Fundamental nonlinear mode shape contour of skew plates with 12 point supports type (I) ( wmax / h = 1 ).

21
 Fig. 7. Fundamental nonlinear mode shape contour of angle-ply simply supported skew plates ( wmax / h = 1 ).

The results show that increasing skew angle will cause an increase in the nonlinear
frequency ratios. Comparison of the results indicates that the angle-ply
ply plate provides
greater linear and nonlinear frequencies towards the cross-ply
cross ply plate but the ratio of
nonlinear to linear frequency is higher in the case of cross-ply
cross ply plate. Also
Also, it is worth
noting that for the rhombic plate with skew angle of 45° , the results for cross
cross-ply
[0° /90° /90° /0° ] and angle-ply
ply [45° / − 45° / − 45° /45° ] plates are identical, which arises

from the similar condition of plate for the two fiber arrangements.
Based on the results, by increasing the amplitude of vibration ( wmax / h ), the
nonlinear frequency ratio ωNL / ωL increases. But there are exceptions in the case of
the simply supported plate at the amplitude ratio wmax / h = 1 , in which the nonlinear
frequency ratio decreases compared to the amplitude ratio wmax / h = 0.8 . This

condition is associated with an overall change in the mode shape such that the
maximum displacement transfers from the center of the plate toward one side and the
mode shape becomes asymmetric
asymmetric.. This phenomenon is known as “mode switching”
and has already been addressed by researchers for the nonlinear vibration and post
post-
buckling analysis of plates [1,, 29].
It may be further concluded that increasing the number of point supports will lead
to an increase in fundamental linear frequency and nonlinear frequency ratios, which
is associated with approaching the values
valu of simply supported plate. In this regard,
regard 20
point supports are sufficient to reach 97% of linear frequency values of simply
supported plate. But, this
his is different for the case of nonlinear frequency ratios as more

22
point supports are required to approach the values of simply supported plate,
especially by increasing the skew angles and amplitude ratios. Accordingly, it can be
said that this type of point supported plates shows less nonlinear frequency ratios
compared with the simply supported plate except in the cases that mode switching
occurs.
It is of further interest to mention that the number of nodes required to achieve
convergence is much more in the case of the point supported plates than that of the
fully simply supported plates. Also, in the nonlinear analysis of point supported plates
more nodes should be used compared with linear analysis to obtain converged results.
This means that in the nonlinear analysis of point supported plates we are dealing with
a more difficult problem compared with the nonlinear analysis of plates having usual
boundary conditions. This matter may be deduced from Table 6, which presents the
convergence process of nonlinear frequency ratio for the cross-ply composite plates
with 8 point supports. Accordingly, the nonlinear frequency ratio shows a decreasing
trend when the node density increases. This increasing is continued until the
maximum relative difference between two consecutive nonlinear frequency ratios is
0.3%. The results also indicate that, for higher amplitude ratios, more nodes should be
employed.

Table 6
Convergence study of nonlinear frequency ratio ( ωNL / ωL ) for cross-ply [0° / 90° / 90° / 0° ] plates with 8 point supports
type (I).
Skew Node density
wmax / h
angle
13 × 13 15 × 15 17 × 17 19 × 19 21 × 21 23 × 23 25 × 25

0.2 1.0152 1.0142 1.0135 1.0129 1.0125 1.0120 1.0119

0.4 1.0597 1.0556 1.0532 1.0509 1.0492 1.0474 1.0468
0 0.6 1.1305 1.1217 1.1165 1.1115 1.1079 1.1041 1.1027
0.8 1.2234 1.2088 1.2000 1.1917 1.1857 1.1793 1.1768
1 1.3345 1.3131 1.3005 1.2882 1.2795 1.2701 1.2663
0.2 1.0153 1.0149 1.0143 1.0139 1.0137 1.0134 1.0132
0.4 1.0597 1.0579 1.0559 1.0543 1.0534 1.0525 1.0517
45
0.6 1.1291 1.1254 1.1211 1.1176 1.1158 1.1139 1.1122
0.8 1.2184 1.2123 1.2053 1.1996 1.1965 1.1934 1.1907

23
 1 1.3227 1.3140 1.3040 1.2958 1.2913 1.2869 1.2830

The effect of fiber orientation is also studied for the rhombic plate with four plies
and 16 point supports. The variation of nonlinear frequency ratio of the plate at the
amplitude ratio wmax / h = 1 and for two cases of fiber arrangement
arrangement, namely
[θ / − θ / − θ /θ ] and [θ /θ +90° / θ + 90° /θ ] , is illustrated in Fig. 8. The obtained results

are different for square and skew plates. The results of the square plate exhibit a
symmetric curve with a minimum ratio at θ = 45° while the minimum ratio of the
skew plates occurs at θ = 90° for both fiber arrangements.

Fig. 8. Effect of fiber orientation on the nonlinear frequency ratio of skew plates with 16 point supports ( wmax / h = 1 ) :

a) fiber arrangement [θ / − θ / − θ /θ ] , b) fiber arrangement [θ /θ +90° / θ + 90° /θ ]

4.4. Point supported skew plate with point arrangement type (II)

The second type of support conditions considered here includes equally spaced
point supports located on the diagonals of skew plate. Tables 7 and 8 show the
fundamental linear frequency and the nonlinear frequency ratios for cross
cross-ply and
angle-ply rhombic plates. The results are presented for plates with 8, 9, 12, 13 and 16
point supports.

Table 7
dimensional linear frequency ( Ω L = ωL a 2 ρ / ET / h ) and nonlinear frequency ratios ( ωNL / ωL ) of
The fundamental non-dimensional
cross-ply [0° / 90° / 90° / 0° ] skew plates with point support type (II).
Number of point Skew wmax / h
ΩL
support angle 0.2 0.4 0.6 0.8 1
0 21.3409 1.0057 1.0225 1.0498 1.0866 1.0532

24
 30 22.8540 1.0062 1.0244 1.0539 1.0935 1.0382
45 25.2348 1.0081 1.0317 1.0685 1.1164 1.0446
0 26.8706 1.0058 1.0230 1.0512 1.0897 1.0679
30 27.9014 1.0080 1.0317 1.0702 1.1223 1.0359
45 29.5016 1.0149 1.0578 1.1246 1.2103 1.0509
0 28.2973 1.0066 1.0263 1.0588 1.1035 1.0798
30 29.1377 1.0087 1.0343 1.0757 1.1314 1.1074
45 30.54900 1.0164 1.0628 1.1329 1.2194 1.1544
0 28.6833 1.0073 1.0290 1.0647 1.1140 1.0791
30 29.4549 1.0090 1.0355 1.0785 1.1363 1.1081
45 30.7879 1.0161 1.0615 1.1310 1.2185 1.1562
0 28.8078 1.0081 1.0322 1.0719 1.1266 1.0792
30 29.5529 1.0094 1.0372 1.0822 1.1433 1.1140
45 30.8426 1.0161 1.0618 1.1322 1.2204 1.1665

Table 8
dimensional linear frequency ( Ω L = ωL a 2 ρ / ET / h ) and nonlinear frequency ratios ( ωNL / ωL ) of
The fundamental non-dimensional
angle-ply [45° / −45° / −45° / 45° ] skew plates with point support type (II).
Number of point Skew wmax / h
ΩL
support angle 0.2 0.4 0.6 0.8 1
0 8.1227 1.0751 1.2710 1.5359 1.8333 2.1443
30 18.7946 1.0172 1.0658 1.1393 1.2301 1.3328
45 25.2348 1.0081 1.0317 1.0685 1.1164 1.0446
0 8.1227 1.0752 1.2714 1.5366 1.8343 2.1457
30 19.0164 1.0199 1.0764 1.1623 1.2694 1.3926
45 29.5016 1.0149 1.0578 1.1246 1.2103 1.0509
0 8.1227 1.0753 1.2717 1.5372 1.8352 2.1467
30 19.1494 1.0211 1.0806 1.1709 1.2831 1.4114
45 30.5490 1.0164 1.0628 1.1329 1.2194 1.1544
0 8.1227 1.0753 1.2718 1.5375 1.8356 2.1472
30 19.2175 1.0215 1.0824 1.1743 1.2883 1.4185
45 30.7879 1.0161 1.0615 1.1310 1.2185 1.1562
0 8.1227 1.0754 1.2720 1.5377 1.8359 2.1476
30 19.2577 1.0218 1.0835 1.1765 1.2918 1.4233
45 30.8426 1.0161 1.0618 1.1322 1.2204 1.1665

It is observed that by increasing the number of point supports, the linear frequency
and the nonlinear
inear frequency ratios increase with a lower rate for more numbers
number of
point supports. Increasing the amplitude of vibration will be followed by higher
nonlinear frequency ratios except in the cases that mode switching occurs. For the

25
skew angles less than 45° , angle-ply
ang ply plate provides lower linear frequency with higher
nonlinear frequency ratios compared with cross-ply
cross ply plate, and for the skew angles of
45° , the results are the same. Also,
Also for the cross-ply
ply plate the higher skew angle
results in higher nonlinear frequency ratios but this trend is reversed for the angle-ply
angle
plate.
It is worth mentioning that for the angle-ply
angle square plate, the diagonals
gonals of which
are in line with the fibers, the linear frequency is independent of the number of point
supports, and the nonlinear frequency ratio changes
change very slightly. This case is
associated with a significant reduction in linear frequency and a great increase in
nonlinear frequency ratio, i.e., at the amplitude ratio wmax / h = 1 the nonlinear

frequency is about 2.14 times the linear frequency. The mode shape contours
associated with plates having 12 point supports at the amplitude ratio wmax / h = 1 are

shown in Fig. 9.

Fig. 9. Fundamental nonlinear mode shape contour of cross-ply skew plates with
th 12 point supports type (II).
(II)

4.5. Point supported skew plate with point arrangement type (III)

In the third case of support conditions, the point supports are located along the four
sides of the plate as in the first case but with the difference that a set of point supports
are added parallel to the peripheral
eripheral sides. The distance between the two adjacent point

26
supports is specified by the parameter α as shown in Fig. 5. The fundamental linear
frequency and the nonlinear frequency ratios of cross-ply and angle-ply
ply rhombic
plates are presented in Tables 9 and 10 for α = 0.05 with the results of clamped plate
plate.

Table 9
dimensional linear frequency ( Ω L = ωL a 2 ρ / ET / h ) and nonlinear frequency ratios ( ωNL / ωL ) of
The fundamental non-dimensional
cross-ply [0° / 90° / 90° / 0° ] skew plates with point support type (III). ( α = 0.05 )
Number of point Skew wmax / h
ΩL
support angle 0.2 0.4 0.6 0.8 1
0 35.1151 1.0049 1.0195 1.0433 1.0757 1.1157
30 40.0434 1.0060 1.0236 1.0522 1.0910 1.1385
45 50.4028 1.0064 1.0251 1.0553 1.0956 1.1445
0 38.6216 1.0060 1.0237 1.0527 1.0918 1.1403
30 43.7949 1.0068 1.0271 1.0601 1.1045 1.1595
45 56.1589 1.0074 1.0294 1.0647 1.1120 1.1696
0 40.5557 1.0063 1.0251 1.0556 1.0967 1.1475
30 46.4261 1.0073 1.0289 1.0639 1.1108 1.1687
45 59.6007 1.0087 1.0342 1.0754 1.1302 1.1968
0 41.9356 1.0064 1.0252 1.0559 1.0972 1.1485
30 48.0816 1.0079 1.0313 1.0692 1.1197 1.1819
45 61.4939 1.0096 1.0379 1.0834 1.1437 1.2169
0 44.9116 1.0068 1.0269 1.0596 1.1034 1.1578
30 49.7804 1.0083 1.0328 1.0724 1.1251 1.1899
45 62.9478 1.0105 1.0413 1.0908 1.1560 1.2350
0 41.4909 1.0081 1.0319 1.0705 1.1220 1.1849
30 46.1355 1.0099 1.0392 1.0861 1.1481 1.2225
45 57.0510 1.0131 1.0514 1.1119 1.1909 1.2852

Table 10
dimensional linear frequency ( Ω L = ωL a 2 ρ / ET / h ) and nonlinear frequency ratios ( ωNL / ωL ) of
The fundamental non-dimensional
angle-ply [45° / −45° / −45° / 45° ] skew plates with point support type (III). ( α = 0.05 )

Number of point Skew wmax / h

ΩL
support angle 0.2 0.4 0.6 0.8 1
0 35.6274 1.0058 1.0231 1.0511 1.0890 1.1353
30 39.0516 1.0071 1.0280 1.0616 1.1064 1.1606
45 48.3188 1.0071 1.0281 1.0621 1.1078 1.1638
0 38.9124 1.0069 1.0270 1.0591 1.1016 1.1528
30 43.6425 1.0077 1.0305 1.0672 1.1163 1.1754
45 57.7283 1.0074 1.0294 1.0652 1.1131 1.1721

27
 0 40.2036 1.0079 1.0309 1.0676 1.1158 1.1733
30 45.4574 1.0088 1.0348 1.0764 1.1316 1.1985
45 62.2372 1.0087 1.0344 1.0760 1.1312 1.1988
0 40.8718 1.0082 1.0320 1.0700 1.1201 1.1797
30 46.3518 1.0098 1.0385 1.0844 1.1448 1.2178
45 64.5060 1.0096 1.0379 1.0833 1.1434 1.2166
0 41.9341 1.0089 1.0348 1.0759 1.1300 1.1940
30 47.1635 1.0106 1.0418 1.0914 1.1564 1.2346
45 65.9437 1.0105 1.0412 1.0905 1.1552 1.2338
0 36.7420 1.0089 1.0350 1.0763 1.1306 1.1947
30 41.1229 1.0122 1.0479 1.1042 1.1770 1.2640
45 57.0510 1.0131 1.0514 1.1119 1.1909 1.2852

By comparing the results of support conditions type (III) and (I),

(I) it can be
concluded that the support type (III) yields higher linear frequency and lower
nonlinear frequency ratios, which indicates the reduction of nonlinear effects. Similar
to the general trend observed in the previous support conditions,, increasing the
amplitude of vibration leads to an increase in the nonlinear frequency ratios. In both
cases of cross-ply and angle--ply plates, this increase is greater
ater for the skew plate
plates
compared to square plates.
To examine the effect of α , different values of this parameter are used to evaluate the
linear frequency and nonlinear frequency ratios of cross-ply
cross ply plate support
supported at 52
points, and the results are shown in Fig. 10. It is observed that higher values of
parameter α provide higher fundamental linear frequency even more than the
frequency of the clamped plate. Unlike the linear frequency, nonlinear
onlinear frequency ratio
is not much dependent on the parameter α , and the values are less than those of the
clamped plate. However,, when the number of point supports increases,, the nnonlinear
frequency ratios approach the corresponding values of clamped plate.

28
Fig. 10. Effect of parameter α on the fundamental
undamental linear frequency and nonlinear frequency ratios of cross-ply
cross ply skew plates with
52 point supports. a) fundamental non-dimensional linear frequency, b) nonlinear frequency ratios

4.6. Point supported skew plate with point arrangement type (IV)

Nonlinear free vibration analysis of cross-ply rhombic plates with point support
condition type (IV) shown in Fig. 5 is carried out and the results are presented in
Table 11.. This support condition is composed of four point supports at corners and
variable numbers of point supports located in a circular pattern at the center of the
plate. Based on the results, the nonlinear
n frequency ratios show an increasing trend
when the plate experiences larger out of plane displacement until the mode switching
at the amplitude ratio wmax / h = 1 happens. It is notable that in the case of support

condition type (IV), increasing the skew angle and number of point supports has a
negligible effect on the results.
results
Comparison of the two types of supports that have middle point supports, namely
type (II) and (IV), shows that for the same number of point supports, support type (II)
presents higher nonlinear frequency ratios and is more sensitive to the changing of
skew angle.

Table 11
dimensional linear frequency ( Ω L = ωL a 2 ρ / ET / h ) and nonlinear frequency ratios ( ωNL / ωL ) of
The fundamental non-dimensional
cross-ply [0° / 90° / 90° / 0° ] skew plates with point support type (IV).
Number of point Skew wmax / h
ΩL
support angle 0.2 0.4 0.6 0.8 1
0 25.2270 1.0058 1.0229 1.0509 1.0889 1.0266
30 24.6819 1.0061 1.0241 1.0527 1.0904 1.0396

29
 45 26.0810 1.0064 1.0248 1.0533 1.0897 1.0341
0 26.3910 1.0057 1.0225 1.0502 1.0880 1.0241
30 24.7509 1.0061 1.0240 1.0525 1.0901 1.0349
45 25.8774 1.0066 1.0255 1.0549 1.0928 1.0397
0 26.2404 1.0057 1.0258 1.0508 1.0892 1.0242
30 25.0290 1.0059 1.0233 1.0512 1.0880 1.0372
45 26.2557 1.0065 1.0252 1.0543 1.0915 1.0402
0 26.3892 1.0057 1.0226 1.0504 1.0886 1.0245
30 25.0596 1.0060 1.0234 1.0513 1.0883 1.0371
45 26.2471 1.0065 1.0253 1.0545 1.0920 1.0409‫د‬
0 26.3957 1.0057 1.0228 1.0509 1.0894 1.0245
30 25.1080 1.0059 1.0234 1.0512 1.0881 1.0372
45 26.3030 1.0065 1.0253 1.0545 1.0920 1.0411

4.7. Effect of aspect ratio

To investigate the effect of aspect ratio on the large amplitude free vibration of
point supported skew plates, one case from
f m each point support types (I, II, III) is
selected and the nonlinear frequency ratios at the amplitude ratio wmax / h = 1 and for

the aspect ratios equal to 1, 1.25,

1.25 1.5, 1.75 and 2 are obtained
btained and presented in Table
12.. The bold numbers indicate occurrence of mode switching phenomenon.

Table 12
The nonlinear frequency ratios ( ωNL / ωL ) of skew plates with different aspect ratios ( wmax / h = 1 )
Point support arrangement Skew a/b
angle 1 1.25 1.5 1.75 2
0 1.3265 1.4727 1.5885 1.7570 1.8506
Cross-
ply 30 1.3497 1.4837 1.6139 1.7096 1.8226
45 1.3622 1.4647 1.5862 1.6892 1.7557
0 1.2497 1.2687 1.2939 1.3232 1.3515
Angle-
ply 30 1.3241 1.3380 1.3479 1.3700 1.3880
45 1.3622 1.3703 1.3772 1.2238 1.1902
0 1.0792 1.1596 1.2341 1.2044 1.1537
Cross-
ply 30 1.1140 1.1846 1.1987 1.1528 1.1060
45 1.1665 1.1949 1.1504 1.0935 1.0506
0 2.1476 1.8004 1.5008 1.3517 1.2727
Angle-
ply 30 1.4233 1.3342 1.2606 1.0837 1.0524
45 1.1665 1.1031 1.0604 1.0346 1.0186
0 1.1403 1.1938 1.2330 1.2897 1.3222
Cross-
30 1.1595 1.2121 1.2566 1.2906 1.3302

30
 ply 45 1.1696 1.2171 1.2593 1.2915 1.3146
0 1.1528 1.1612 1.1667 1.1756 1.1811
Angle-
ply 30 1.1754 1.1747 1.1689 1.1704 1.1753
45 1.1721 1.1664 1.1635 1.0885 1.0713

The results show that the aspect ratio has a significant effect on the values of
nonlinear frequency ratio. In the case of cross-ply plates, higher aspect ratios exhibit
higher nonlinear frequency ratios apart from when the mode switching phenomenon
occurs, but for the angle-ply plates this is true only for the point support type (I). The
skew plates with greater aspect ratios are more prone to experience mode switching.
Also, in the plates with a greater aspect ratio, mode switching phenomenon will lead
to a greater reduction in the nonlinear frequency ratio.
It should be noted that some conclusions that were drawn for the rhombic plate are
no longer valid for the plates with aspect ratios greater than 1; for example in the case
of point support plate type (I), by increasing the aspect ratio the square plate yields
higher nonlinear frequency ratios than the skew plates. Also, it is worth mentioning
that in the case of angle-ply rectangular plate with point support type (II), increasing
the aspect ratio is associated with a rapid decrease in nonlinear frequency ratio, and
the plate with aspect ratio a / b = 1 gives the maximum nonlinear frequency ratio.

5. Conclusions

The element-free Galerkin method was utilized to study large amplitude free
vibration of thin laminated composite skew plates supported at discrete points. To this
end, four types of point support arrangements were considered and nonlinear
frequency ratios of four layered composite plates with different numbers of point
supports, amplitude ratios, skew angles, aspect ratios and fiber orientations were
calculated. The results show that, in general, increasing the amplitude of vibration will
be followed by increasing the nonlinear stiffness, and thus higher nonlinear frequency
ratios would be achieved. But, this is not valid when the mode switching phenomenon

31
happens. The obtained results reveal that plates with greater aspect ratios and skew
angles are more prone to experience mode switching. Also, it was concluded that
cross-ply and angle-ply plates offer different behaviors for various types of point
support arrangements. For the plates with the skew angle less than 45° , the cross-ply
plates with support conditions type (I) yield lower linear frequency and higher
nonlinear frequency ratio than angle-ply plates. This condition is reversed for the
plates with support conditions type (II) and for the rhombic plates with point support
type (III).
The general trend observed in the results implies an increase in linear frequency
and nonlinear frequency ratio with the increasing of the number of point supports as
was expected. Comparison of the results of four types of support conditions indicates
that the highest nonlinear frequency ratio occurs in the point support arrangement type
(I) for the cross-ply plates. For the angle-ply laminates, the square plate with point
support type (II) gives the maximum nonlinear frequency ratio. It is further inferred
that the support type (II) presents the highest sensitivity to the skew angle among
other types of point support arrangements.

Appendix A:

The stiffness matrix of the plate is formed by assembling the submatrices (K ij )3×3

associated with two nodes i and j which fall within the influence domain of a certain
Gauss point. The strain vector at this Gauss point may be written as
ε + ε  n B
ε =  0 L 0 NL  = ∑  0 p {}
0.5B1   u 
  v 

(A.1)
 κ   0 B 0b  I  w
   I
I =1

where u , v and w are nodal parameters of node I, and n is number of nodes in the
influence domain of the Gauss point. The matrices B 0 p , B1 and B0b are defined in the

following form

32
  ∂N i   ∂w   ∂ 2 Ni 
 0   0  − 2 
 ∂x   ∂x   ∂N i   ∂x 
 ∂N i   ∂w   ∂x   ∂2N 
. (B 0 p ) i =  0 , (B1 )i =  0  , (B 0 b ) i =  − 2 i  .
∂y  ∂y   ∂N i 
(A.2)
   ∂y 
 ∂N i ∂N i   ∂w ∂w   ∂y   ∂2 N 
     −2 i

 ∂y ∂x   ∂y ∂x   ∂x∂y 

After calculating δ ε and expanding Eq. (3), the stiffness submatrices are obtained
as

(B T0 p )i A(B 0 p ) j 0 
(K L )ij =  T  ,
 0 (B 0b )i D(B 0b ) j 
 1 T 
 0 (B 0 p )i A(B1 ) j 
(K1NL )ij = 2 ,
 T  (A.3)
(B1 )i A(B 0 p ) j 0 
0 0 
(K2 NL )ij =  
0 1 (B1T )i A(B1 ) j 
 2 

These submatrices are calculated at a certain Gauss point, which should be

assembled together to obtain the total stiffness matrix.

References

[1] Singha MK, Ganapathi M. Large amplitude free flexural vibrations of laminated
composite skew plates. International Journal of Non-Linear Mechanics.
2004;39:1709-20.
[2] Singha MK, Daripa R. Nonlinear vibration of symmetrically laminated composite
skew plates by finite element method. International Journal of Non-Linear Mechanics.
2007;42:1144-52.

33
[3] Malekzadeh P. A differential quadrature nonlinear free vibration analysis of
laminated composite skew thin plates. Thin-Walled Structures. 2007;45:237-50.
[4] Malekzadeh P. Differential quadrature large amplitude free vibration analysis of
laminated skew plates based on FSDT. Composite Structures. 2008;83:189-200.

[5] Narita Y, Iwato N. Vibration studies for symmetrically laminated circular and
elliptical plates resting on elastic point supports. Composites Science and Technology.
1990;39:75-88.

[6] Liew KM. A hybrid energy approach for vibrational modelling of laminated
trapezoidal plates with point supports. International Journal of Solids and Structures.
1992;29:3087-97.

[7] Abrate S. Vibration of composite plates with internal supports. International

Journal of Mechanical Sciences. 1994;36:1027-43.

[8] Ye J, Soldatos KP. Three-dimensional vibration of laminated composite plates and

cylindrical panels with arbitrarily located lateral surfaces point supports. International
Journal of Mechanical Sciences. 1996;38:271-81.

[9] Cheung YK, Zhou D. The free vibrations of rectangular composite plates with
point-supports using static beam functions. Composite Structures. 1999;44:145-54.
[10] Setoodeh AR, Karami G. Static, free vibration and buckling analysis of
anisotropic thick laminated composite plates on distributed and point elastic supports
using a 3-D layer-wise FEM. Engineering Structures. 2004;26:211-20.
[11] Kong J. Vibration of isotropic and composite plates using computed shape
function and its application to elastic support optimization. Journal of Sound and
Vibration. 2009;326:671-86.
[12] Bahmyari E, Khedmati MR. Vibration analysis of nonhomogeneous moderately
thick plates with point supports resting on Pasternak elastic foundation using element
free Galerkin method. Engineering Analysis with Boundary Elements. 2013;37:1212-
38.

[13] Nguyen VP, Rabczuk T, Bordas S, Duflot M. Meshless methods: A review and
computer implementation aspects. Mathematics and Computers in Simulation.
2008;79:763-813.

34
[14] Liew KM, Zhao X, Ferreira AJM. A review of meshless methods for laminated
and functionally graded plates and shells. Composite Structures. 2011;93:2031-41.
[15] Belytschko T, Lu YY, Gu L. Element-free Galerkin methods. International
Journal for Numerical Methods in Engineering. 1994;37:229-56.

[16] Belinha J, Dinis LMJS. Nonlinear analysis of plates and laminates using the
element free Galerkin method. Composite Structures. 2007;78:337-50.

[17] Jaberzadeh E, Azhari M, Boroomand B. Inelastic buckling of skew and rhombic

thin thickness-tapered plates with and without intermediate supports using the
element-free Galerkin method. Applied Mathematical Modelling. 2013;37:6838-54.

[18] Liu L, Chua LP, Ghista DN. Element-free Galerkin method for static and
dynamic analysis of spatial shell structures. Journal of Sound and Vibration.
2006;295:388-406.

[19] Reddy JN. Mechanics of Laminated Composite Plates and Shells: Theory and
Analysis, Second Edition: Taylor & Francis; 2004.
[20] Nayroles B, Touzot G, Villon P. Generalizing the finite element method: Diffuse
approximation and diffuse elements. Computational Mechanics. 1992;10:307-18.

[21] Liu GR. Mesh Free Methods: Moving Beyond the Finite Element Method: Taylor
& Francis; 2002.

[22] Belinha J, Dinis LMJS. Analysis of plates and laminates using the element-free
Galerkin method. Computers & Structures. 2006;84:1547-59.
[23] Han W, Petyt M. Geometrically nonlinear vibration analysis of thin, rectangular
plates using the hierarchical finite element method—I: The fundamental mode of
isotropic plates. Computers & Structures. 1997;63:295-308.
[24] Singha MK, Daripa R. Nonlinear vibration and dynamic stability analysis of
composite plates. Journal of Sound and Vibration. 2009;328:541-54.

[25] Sheikh AH, Mukhopadhyay M. Large amplitude free flexural vibration of

stiffened plates. AIAA Journal. 1996;34:2377-83.

[26] Shooshtari A, Razavi S. A closed form solution for linear and nonlinear free
vibrations of composite and fiber metal laminated rectangular plates. Composite
Structures. 2010;92:2663-75.

35
[27] Mizusawa T, Kajita T. Vibration of skew plates resting on point supports. Journal
of Sound and Vibration. 1987;115:243-51.
[28] Abrate S. Vibration of point-supported rectangular composite plates. Composites
Science and Technology. 1995;53:325-32.

[29] Chen H, Yu W. Postbuckling and mode jumping analysis of composite laminates

using an asymptotically correct, geometrically non-linear theory. International Journal
of Non-Linear Mechanics. 2006;41:1143-60.

Figure Captions

Fig. 1. The skew plate supported at discrete points.

Fig. 2. Mesh-free model of the skew plate and the influence domain of a quadrature
point X.

Fig. 3. Mapping of a quadrilateral plate into natural coordinate system.

Fig. 4. A schematic design of background mesh for quadrature points in physical and
natural coordinate systems.

Fig. 5. Four types of point support arrangements.

Fig. 6. Fundamental nonlinear mode shape contour of skew plates with 12 point
supports type (I) ( wmax / h = 1 ).
Fig. 7. Fundamental nonlinear mode shape contour of angle-ply simply supported
skew plates ( wmax / h = 1 ).

Fig. 8. Effect of fiber orientation on the nonlinear frequency ratio of skew plates with
16 point supports ( wmax / h = 1 ). a) fiber arrangement [θ / − θ / − θ /θ ] , b) fiber
arrangement [θ /θ +90° / θ + 90° /θ ] .

36
Fig. 9. Fundamental nonlinear mode shape contour of cross-ply skew plates with 12
point supports type (II).

Fig. 10. Effect of parameter α on the fundamental linear frequency and nonlinear
frequency ratios of cross-ply skew plates with 52 point supports. a) fundamental non-
dimensional linear frequency, b) nonlinear frequency ratios.

37
Table 1
*
The fundamental non-dimensional linear frequency ( Ω L = ωL a 2 ρ h / D ) and nonlinear frequency ratios ( ω NL / ωL ) for a
isotropic square plate ( ν = 0.3 )
Node wmax / h
ΩL
density 0.2 0.4 0.6 0.8 1
Simply supported plate
9×9 19.7299 1.0199 1.0777 1.1679 1.2843 1.4208
11 × 11 19.7345 1.0200 1.0780 1.1686 1.2855 1.4226
Present study
13 × 13 19.7347 1.0200 1.0779 1.1684 1.2851 1.4219
15 × 15 19.7359 1.0200 1.0779 1.1683 1.2850 1.4218
FEM [23] - - 1.0197 1.0767 1.1659 1.2800 1.4167
FSM [24] - - 1.0197 1.0768 1.1662 1.2813 1.4173
Analytical solution [25] - - 1.0196 1.0766 1.1666 1.2840 1.4242
Clamped plate
9×9 36.0525 1.0072 1.0283 1.0625 1.1085 1.1650
11 × 11 36.0234 1.0074 1.0293 1.0648 1.1124 1.1705
Present study
13 × 13 35.9926 1.0073 1.0290 1.0641 1.1113 1.1690
15 × 15 36.0015 1.0074 1.0290 1.0642 1.1114 1.1691
[23] - - 1.0072 1.0286 1.0632 1.1098 1.1669
[24] - - 1.0073 1.0287 1.0634 1.1100 1.1672
* D = Eh3 / 12(1 − ν 2 )

Table 2
The nonlinear frequency ratios ( ωNL / ωL ) for angle-ply [45° / − 45° /45° / −45° /45° ] rhombic plates.
Skew wmax / h
angle 0.2 0.4 0.6 0.8 1
Simply supported plate
Present study 1.0150 1.0589 1.1283 1.2191 1.3269
[24] 1.0147 1.0578 1.1260 1.2154 1.3219
0 [2] 1.0143 1.0562 1.1229 1.2102 1.3150
Present study* 1.0197 1.0767 1.1658 1.2806 1.4149
[1]* 1.0190 1.0740 1.1603 1.2720 1.4034
Present study 1.0226 1.0875 1.1877 1.3153 1.4631
[2] 1.0215 1.0833 1.1796 1.3029 1.4466
30 Present study* 1.0295 1.1132 1.2404 1.3994 1.5807
[3]* 1.0293 1.1128 1.2395 1.3981 -
[1]* 1.0284 1.1095 1.2332 1.3884 1.5661
Clamped plate
Present study 1.0094 1.0372 1.0819 1.1414 1.2135
[2] 1.0099 1.0387 1.0849 1.1458 1.2215
45 Present study* 1.0123 1.0484 1.1059 1.1816 1.2723
[3]* - 1.0487 - 1.1827 1.2740
[1]* 1.0124 1.0488 1.1067 1.1832 1.2749

38
*
The equations of motion at the instant of maximum displacement.
displacement
Results of References [2, 24] based on a shear deformable finite element approach for thickness ratio a/h=100.
=100.

Table 3
dimensional natural frequencies of the point supported rhombic plates ( Ω = ωL a 2 ρ h / D )
The first five non-dimensional

Point support Skew angle Ω1 Ω2 Ω3 Ω4 Ω5

condition a* 7.1131 15.7729 15.7729 19.5966 38.4582
0 [26]] 7.111 15.77 15.77 19.60 38.43
[27]] 7.1109 15.7703 15.7703 19.5961 -
*
a 9.10481 11.4027 22.6508 25.1824 30.2967
30
[26]] 9.104 11.40 22.65 25.18 30.29
*
45 a 10.2086 11.8203 27.764 29.7824 33.3507
[26]] 10.21 11.82 27.76 29.78 33.34
*
60 a 9.43431 14.6641 36.2774 39.3177 39.9546
[26]] 9.433 14.66 36.29 39.32 39.94
a* 17.8757 34.9124 34.9124 38.3974 60.2422
0 [26]] 17.91 34.99 34.99 38.43 60.27
[27]] 18.150 35.3580 35.3580 38.4323 60.9431-
*
a 19.9089 42.0237 42.8600 44.9505 48.9842
30
[26]] 19.94 42.21 42.96 44.98 49.10
*
a 24.3401 42.8726 47.8954 51.4239 61.2479
45
[26]] 24.37 42.88 48.13 51.98 61.39
a* 35.3364 39.5013 52.3898 57.5381 80.7762
60
[26]] 35.41 39.57 52.94 58.46 81.24
*
a = Present study ( 15 × 15 node density)

39
Table 4
dimensional linear frequency ( Ω L = ωL a 2 ρ / ET / h ) and nonlinear frequency ratios ( ωNL / ωL ) of
The fundamental non-dimensional
cross-ply [0° / 90° / 90° / 0° ] skew plates with point support type (I).
Number of point Skew wmax / h
ΩL
support angle 0.2 0.4 0.6 0.8 1
0 16.2982 1.0119 1.0468 1.1027 1.1768 1.2663
30 17.9006 1.0122 1.0478 1.1040 1.1777 1.2650
45 20.8939 1.0132 1.0517 1.1122 1.1907 1.2830
0 18.2009 1.0116 1.0458 1.1003 1.1723 1.2586
30 20.9319 1.0130 1.0508 1.1110 1.1898 1.2839
45 26.7904 1.0142 1.0556 1.1208 1.2056 1.3055
0 18.6290 1.0149 1.0584 1.1277 1.2182 1.3265
30 21.6493 1.0162 1.0635 1.1380 1.2351 1.3497
45 28.2397 1.0171 1.0666 1.1442 1.2445 1.3622
0 18.7651 1.0174 1.0679 1.1477 1.2511 1.3738
30 21.9118 1.0188 1.0734 1.1590 1.2691 1.3986
45 28.7489 1.0208 1.0808 1.1736 1.2920 1.4296
0 18.8177 1.0182 1.0710 1.1540 1.2617 1.3893
30 22.0232 1.0208 1.0810 1.1748 1.2948 1.4353
45 28.9819 1.0229 1.0885 1.1897 1.3183 1.4669
0 18.8919 1.0319 1.1227 1.2607 1.4335 1.3443
30 22.2347 1.0355 1.1356 1.2858 1.4721 1.3911
45 29.3300 1.0426 1.1607 1.3341 1.5431 1.4385

Table 5

40
The fundamental non-dimensional linear frequency ( Ω L = ωL a 2 ρ / ET / h ) and nonlinear frequency ratios ( ωNL / ωL ) of
angle-ply [45° / −45° / −45° / 45° ] skew plates with point support type (I).
Number of point Skew wmax / h
ΩL
support angle 0.2 0.4 0.6 0.8 1
0 17.8368 1.0113 1.0438 1.0943 1.1587 1.2332
30 19.5328 1.0088 1.0345 1.0759 1.1312 1.1979
45 20.8939 1.0132 1.0517 1.1122 1.1907 1.2830
0 20.7066 1.0098 1.0386 1.0846 1.1451 1.2173
30 21.7889 1.0121 1.0475 1.1038 1.1792 1.2661
45 26.7904 1.0142 1.0556 1.1208 1.2056 1.3055
0 21.6606 1.0113 1.0444 1.0971 1.1668 1.2497
30 22.5666 1.0151 1.0589 1.1280 1.2180 1.3241
45 28.2397 1.0171 1.0666 1.1442 1.2445 1.3622
0 22.0245 1.0120 1.0470 1.1027 1.1763 1.2635
30 22.9255 1.0174 1.0679 1.1470 1.2493 1.3690
45 28.7489 1.0208 1.0808 1.1736 1.2920 1.4296
0 22.2006 1.0134 1.0527 1.1150 1.1962 1.2932
30 23.1069 1.0186 1.0725 1.1567 1.2651 1.3913
45 28.9819 1.0229 1.0885 1.1897 1.3183 1.4669
0 22.4473 1.0186 1.0724 1.1561 1.2636 1.3891
30 23.4902 1.0318 1.1214 1.2559 1.4216 1.6083
45 29.3300 1.0426 1.1607 1.3341 1.5431 1.4385

Table 6
Convergence study of nonlinear frequency ratio ( ωNL / ωL ) for cross-ply [0° / 90° / 90° / 0° ] plates with 8 point supports
type (I).
Skew Node density
wmax / h
angle
13 × 13 15 × 15 17 × 17 19 × 19 21 × 21 23 × 23 25 × 25

0.2 1.0152 1.0142 1.0135 1.0129 1.0125 1.0120 1.0119

0.4 1.0597 1.0556 1.0532 1.0509 1.0492 1.0474 1.0468
0 0.6 1.1305 1.1217 1.1165 1.1115 1.1079 1.1041 1.1027
0.8 1.2234 1.2088 1.2000 1.1917 1.1857 1.1793 1.1768
1 1.3345 1.3131 1.3005 1.2882 1.2795 1.2701 1.2663
0.2 1.0153 1.0149 1.0143 1.0139 1.0137 1.0134 1.0132
0.4 1.0597 1.0579 1.0559 1.0543 1.0534 1.0525 1.0517
45 0.6 1.1291 1.1254 1.1211 1.1176 1.1158 1.1139 1.1122
0.8 1.2184 1.2123 1.2053 1.1996 1.1965 1.1934 1.1907
1 1.3227 1.3140 1.3040 1.2958 1.2913 1.2869 1.2830

41
Table 7
dimensional linear frequency ( Ω L = ωL a 2 ρ / ET / h ) and nonlinear frequency ratios ( ωNL / ωL ) of
The fundamental non-dimensional
cross-ply [0° / 90° / 90° / 0° ] skew plates with point support type (II).
Number of point Skew wmax / h
ΩL
support angle 0.2 0.4 0.6 0.8 1
0 21.3409 1.0057 1.0225 1.0498 1.0866 1.0532
30 22.8540 1.0062 1.0244 1.0539 1.0935 1.0382
45 25.2348 1.0081 1.0317 1.0685 1.1164 1.0446
0 26.8706 1.0058 1.0230 1.0512 1.0897 1.0679
30 27.9014 1.0080 1.0317 1.0702 1.1223 1.0359
45 29.5016 1.0149 1.0578 1.1246 1.2103 1.0509
0 28.2973 1.0066 1.0263 1.0588 1.1035 1.0798
30 29.1377 1.0087 1.0343 1.0757 1.1314 1.1074
45 30.54900 1.0164 1.0628 1.1329 1.2194 1.1544
0 28.6833 1.0073 1.0290 1.0647 1.1140 1.0791
30 29.4549 1.0090 1.0355 1.0785 1.1363 1.1081
45 30.7879 1.0161 1.0615 1.1310 1.2185 1.1562
0 28.8078 1.0081 1.0322 1.0719 1.1266 1.0792
30 29.5529 1.0094 1.0372 1.0822 1.1433 1.1140
45 30.8426 1.0161 1.0618 1.1322 1.2204 1.1665

Table 8
dimensional linear frequency ( Ω L = ωL a 2 ρ / ET / h ) and nonlinear frequency ratios ( ωNL / ωL ) of
The fundamental non-dimensional
angle-ply [45° / −45° / −45° / 45° ] skew plates with point support type (II).
Number of point Skew wmax / h
ΩL
support angle 0.2 0.4 0.6 0.8 1
0 8.1227 1.0751 1.2710 1.5359 1.8333 2.1443
30 18.7946 1.0172 1.0658 1.1393 1.2301 1.3328
45 25.2348 1.0081 1.0317 1.0685 1.1164 1.0446
0 8.1227 1.0752 1.2714 1.5366 1.8343 2.1457
30 19.0164 1.0199 1.0764 1.1623 1.2694 1.3926
45 29.5016 1.0149 1.0578 1.1246 1.2103 1.0509
0 8.1227 1.0753 1.2717 1.5372 1.8352 2.1467
30 19.1494 1.0211 1.0806 1.1709 1.2831 1.4114
45 30.5490 1.0164 1.0628 1.1329 1.2194 1.1544
0 8.1227 1.0753 1.2718 1.5375 1.8356 2.1472
30 19.2175 1.0215 1.0824 1.1743 1.2883 1.4185
45 30.7879 1.0161 1.0615 1.1310 1.2185 1.1562
0 8.1227 1.0754 1.2720 1.5377 1.8359 2.1476
30 19.2577 1.0218 1.0835 1.1765 1.2918 1.4233
45 30.8426 1.0161 1.0618 1.1322 1.2204 1.1665

42
Table 9
The fundamental non-dimensional linear frequency ( Ω L = ωL a 2 ρ / ET / h ) and nonlinear frequency ratios ( ωNL / ωL ) of
cross-ply [0° / 90° / 90° / 0° ] skew plates with point support type (III). ( α = 0.05 )
Number of point Skew wmax / h
ΩL
support angle 0.2 0.4 0.6 0.8 1
0 35.1151 1.0049 1.0195 1.0433 1.0757 1.1157
30 40.0434 1.0060 1.0236 1.0522 1.0910 1.1385
45 50.4028 1.0064 1.0251 1.0553 1.0956 1.1445
0 38.6216 1.0060 1.0237 1.0527 1.0918 1.1403
30 43.7949 1.0068 1.0271 1.0601 1.1045 1.1595
45 56.1589 1.0074 1.0294 1.0647 1.1120 1.1696
0 40.5557 1.0063 1.0251 1.0556 1.0967 1.1475
30 46.4261 1.0073 1.0289 1.0639 1.1108 1.1687
45 59.6007 1.0087 1.0342 1.0754 1.1302 1.1968
0 41.9356 1.0064 1.0252 1.0559 1.0972 1.1485
30 48.0816 1.0079 1.0313 1.0692 1.1197 1.1819
45 61.4939 1.0096 1.0379 1.0834 1.1437 1.2169
0 44.9116 1.0068 1.0269 1.0596 1.1034 1.1578
30 49.7804 1.0083 1.0328 1.0724 1.1251 1.1899
45 62.9478 1.0105 1.0413 1.0908 1.1560 1.2350
0 41.4909 1.0081 1.0319 1.0705 1.1220 1.1849
30 46.1355 1.0099 1.0392 1.0861 1.1481 1.2225
45 57.0510 1.0131 1.0514 1.1119 1.1909 1.2852

Table 10
dimensional linear frequency ( Ω L = ωL a 2 ρ / ET / h ) and nonlinear frequency ratios ( ωNL / ωL ) of
The fundamental non-dimensional
angle-ply [45° / −45° / −45° / 45° ] skew plates with point support type (III). ( α = 0.05 )
Number of point Skew wmax / h
ΩL
support angle 0.2 0.4 0.6 0.8 1
0 35.6274 1.0058 1.0231 1.0511 1.0890 1.1353
30 39.0516 1.0071 1.0280 1.0616 1.1064 1.1606
45 48.3188 1.0071 1.0281 1.0621 1.1078 1.1638
0 38.9124 1.0069 1.0270 1.0591 1.1016 1.1528
30 43.6425 1.0077 1.0305 1.0672 1.1163 1.1754
45 57.7283 1.0074 1.0294 1.0652 1.1131 1.1721
0 40.2036 1.0079 1.0309 1.0676 1.1158 1.1733
30 45.4574 1.0088 1.0348 1.0764 1.1316 1.1985
45 62.2372 1.0087 1.0344 1.0760 1.1312 1.1988
0 40.8718 1.0082 1.0320 1.0700 1.1201 1.1797
30 46.3518 1.0098 1.0385 1.0844 1.1448 1.2178
45 64.5060 1.0096 1.0379 1.0833 1.1434 1.2166
0 41.9341 1.0089 1.0348 1.0759 1.1300 1.1940

43
 30 47.1635 1.0106 1.0418 1.0914 1.1564 1.2346
45 65.9437 1.0105 1.0412 1.0905 1.1552 1.2338
0 36.7420 1.0089 1.0350 1.0763 1.1306 1.1947
30 41.1229 1.0122 1.0479 1.1042 1.1770 1.2640
45 57.0510 1.0131 1.0514 1.1119 1.1909 1.2852

Table 11
dimensional linear frequency ( Ω L = ωL a 2 ρ / ET / h ) and nonlinear frequency ratios ( ωNL / ωL ) of
The fundamental non-dimensional
cross-ply [0° / 90° / 90° / 0° ] skew plates with point support type (IV).
Number of point Skew wmax / h
ΩL
support angle 0.2 0.4 0.6 0.8 1
0 25.2270 1.0058 1.0229 1.0509 1.0889 1.0266
30 24.6819 1.0061 1.0241 1.0527 1.0904 1.0396
45 26.0810 1.0064 1.0248 1.0533 1.0897 1.0341
0 26.3910 1.0057 1.0225 1.0502 1.0880 1.0241
30 24.7509 1.0061 1.0240 1.0525 1.0901 1.0349
45 25.8774 1.0066 1.0255 1.0549 1.0928 1.0397
0 26.2404 1.0057 1.0258 1.0508 1.0892 1.0242
30 25.0290 1.0059 1.0233 1.0512 1.0880 1.0372
45 26.2557 1.0065 1.0252 1.0543 1.0915 1.0402
0 26.3892 1.0057 1.0226 1.0504 1.0886 1.0245
30 25.0596 1.0060 1.0234 1.0513 1.0883 1.0371
45 26.2471 1.0065 1.0253 1.0545 1.0920 1.0409‫د‬
0 26.3957 1.0057 1.0228 1.0509 1.0894 1.0245
30 25.1080 1.0059 1.0234 1.0512 1.0881 1.0372
45 26.3030 1.0065 1.0253 1.0545 1.0920 1.0411

44
Table 12
The nonlinear frequency ratios ( ωNL / ωL ) of skew plates with different aspect ratios ( wmax / h = 1 )
Point support arrangement Skew a/b
angle 1 1.25 1.5 1.75 2
0 1.3265 1.4727 1.5885 1.7570 1.8506
Cross-
ply 30 1.3497 1.4837 1.6139 1.7096 1.8226
45 1.3622 1.4647 1.5862 1.6892 1.7557
0 1.2497 1.2687 1.2939 1.3232 1.3515
Angle-
ply 30 1.3241 1.3380 1.3479 1.3700 1.3880
45 1.3622 1.3703 1.3772 1.2238 1.1902
0 1.0792 1.1596 1.2341 1.2044 1.1537
Cross-
ply 30 1.1140 1.1846 1.1987 1.1528 1.1060
45 1.1665 1.1949 1.1504 1.0935 1.0506
0 2.1476 1.8004 1.5008 1.3517 1.2727
Angle-
ply 30 1.4233 1.3342 1.2606 1.0837 1.0524
45 1.1665 1.1031 1.0604 1.0346 1.0186
0 1.1403 1.1938 1.2330 1.2897 1.3222
Cross-
ply 30 1.1595 1.2121 1.2566 1.2906 1.3302
45 1.1696 1.2171 1.2593 1.2915 1.3146
0 1.1528 1.1612 1.1667 1.1756 1.1811
Angle-
ply 30 1.1754 1.1747 1.1689 1.1704 1.1753
45 1.1721 1.1664 1.1635 1.0885 1.0713

45
 Highlights
1- Nonlinear free vibration of skew plates with four types of point support

arrangements is investigated.

2- The effects of number of point supports, skew angle, fiber orientation and

aspect ratio are reported.

3- For the higher amplitude ratios, more nodes should be employed to achieve

convergence.

4- Aspect ratio has a significant effect on the nonlinear frequency ratio.

46