Location via proxy:   [ UP ]  
[Report a bug]   [Manage cookies]                

Reddy

Download as pdf or txt
Download as pdf or txt
You are on page 1of 20

Composites Science and Technology 34 (1989) 205-224

Exact Solutions for the Transient Response of Symmetric


Cross-Ply Laminates Using a Higher-Order Plate Theory

A. A. Khdeir & J. N. R e d d y

Department of Engineering Science and Mechanics, Virginia Polytechnic Institute


and State University, Blacksburg, Virginia 24061, USA

(Received 11 February 1988; revised version received 7 June 1988;


accepted 29 June 1988)

ABSTRACT

The methods of the orthogonality of principal modes and the state variables
technique are used to investigate the dynamic response of symmetric cross-ply
laminated composite plates by solving the equation of a higher-order shear
deformation plate theory. Exact solutions of simply supported rectangular
plates are presented for various Ioadings. The results obtained using the
higher-order theory ( H S D T ) are compared with those obtained by the
Reissner-Mindlin shear deformation plate theory ( F S D T ) as well as the
classical laminate theory ( CPT).

1 INTRODUCTION

Composite laminates are increasingly used in a variety of engineering


structures, including automobiles, space vehicles and submarines. Com-
pared to metallic structures, composite laminates offer some unique
engineering properties while presenting interesting but challenging
problems for analysts and designers. The theory and analysis of laminated
composite plates subjected to static and dynamic loads have received
widespread attention in the last two decades. Most theories developed to
date are refinements of the classical laminate plate theory in which the
transverse stresses are neglected. A review of various higher-order theories
can be found in the works of Reddy. 1-4 The exact analysis of laminated
205
Composites Science and Technology 0266-3538/89/$03'50 © 1989 Elsevier Science Publishers
Ltd, England. Printed in Great Britain
206 A. A. Khdeir, J. N. Reddy

composite plates are often limited to either the classical laminate plate
theory or the first-order (i.e. Reissner-Mindlin) laminate plate theory
applied to simply supported rectangular plates (see Whitney and Pagano, 5
Pagano, 6 Whitney] Bert and Chen, a Reddy and Chao, 9 Reddy, 2-4 among
many others).
In the present paper exact solutions for the transient response of
symmetric cross-ply rectangular composite laminates are developed using
the third-order laminate plate theory proposed by Reddy. 2-4 Two
alternative approaches, namely, the method of orthogonality of principal
modes and the state variable approach, are used. The method of
orthogonality of principal modes was used by Sun and his co-workers 1°- 13
and Dobyns ~4 for the first-order theory. Spatially exact but temporally
approximate (e.g. the Newmark direct integration) method was used by
Reddy~ 5 - 17 and Bhimaraddi ~8, ~9 in the dynamic analysis of the first-order
and third-order theories. This is the first paper that deals with the
application of the method of orthogonality of principal modes as well as the
state variable technique to the dynamic analysis of the third-order theory.
The laminate is of rectangular planeform, symmetric cross-ply lamination
scheme, with simply supported boundary conditions and subjected to either
a uniform or sinusoidal dynamic load at the upper surface of the laminate.
Rectangular, triangular, exponential, and sinusoidal pulse loadings are
considered. A comparison of the dynamic response predicted by the
classical, first-order and third-order theories is presented. The closed-form
solutions can also be used by numerical analysts to validate their numerical
approaches.

2 BASIC E Q U A T I O N S

2.1 Kinematic relations

The higher-order shear deformation theory (HSDT) used in the present


study is based on the following representation of the displacement field
across the plate thickness: 2-4

[4,x- 5 Ox+E -/j


[
~3 ~ W

Here (u~,u~,u~) denote the displacement components in the (x,y,z)


Transient response of symmetric cross-ply laminates 207

directions respectively; (u, v, w) are the displacements of a point on the


midplane (x,y,0); @x and ~k~ are the rotations of the cross-sections
perpendicular to the x and y-axis, respectively. All of the generalized
displacements are functions of position (x,y) and time t.
The strains associated with the displacement field in eqn (1) are:

< = <o + ~(,~ + z~,~)


82 = 8~° + z(,~° + z~,~)
83=0
84 = 8~ + z2x~
85 = 8~° + z 2 ~
86 = 8~ + z(~6° + z 2 ~ }

where 8~° denote the m e m b r a n e strains, x~o the curvatures and x~ are the
second-order curvatures:

<o= ~au &O~


~o, _ ax
3h 2 + 8x2,]

av ~o _ a~,~ 4 (d~r ~2w'~


4 = a~ ay t¢22= 3h2 k 8y + Oy2J

~ = O, + ay (2)

~=0~+~ ~=-~
4( 0 ~ + ~
au av ~0~ O0~
~=~+~ ~ - a~ + ax

2.2 Constitutive equations


For a plate of constant thickness h and m a d e of an orthotropic material
where the plate possesses a plane of elastic symmetry parallel to the x - y
plane, the constitutive equations for the k-layer can be written as:

tY2 = 0~2 022 0 ~2


Q~J(k~Lg~J
~6 ~k) 0 0 066 (/0 ~6
Ol
208 A. A. Khdeir, J. N. Reddy

where (6i, ~ are the components of stress and strain and Qu are the plane
stress-reduced coefficients in the material axes of the layer,

E~
0 . ~ - 1 -v~ev2~ Q~ - 1 - v~:v~ Vl2E2 E2
022 -- 1 -- V~
2v:~ (4)

Q44 = G23 Q55 = G~3 Q66= G12


The lamina constitutive equations (3) can be expressed in terms of stresses
and strains referred to the laminate coordinates as
[ffl] [~~ Q12 QI6]~l] ~ ~
~2 = Q22 I
~ ~ ~6J[~;~j
where Qu are the transformed material constants.

2.3 E~uations of motion

Using Hamilton's principle, the equations of motion appropriate for the


displacement field (1) and the constitutive equations (5) are: 2-~

6.: ~~N~
+ ~ y. +¢~N6
g~=~,a+&G - ~ I4, ~ ~,~,

&: ~6 ~N~ 4 ~fi~


cx + ~?)~+g"=I'O+ i z b ~ - ~ I ' ~
OQI + ~Q2 + NI ~2W (?2w
6w: Ox ~3, ~ + R2 @2
_S[OR, ~Rz) 4 fO2P, ~2P6 ~"2)
+q /flk# x + @ / + ~ k ~ + 2 ~ + ~),~)
(4)~ /?~,i ' '~ff'~ 4 /'~ '"
=I,,i~- ~ /vi~+8),zj+~I4(~+~) (6)

+~ &[~

6¢x: ?M1
a~ + ~M6
~y e , + ~ e4, - ~ ~ + 4
~ ) fOP
+m1. ~P6~
_ .. 4 _ ~ff'
= [zff + I 3 ~ - ~ Is gx

b0,,:
-
8M 6
¢~Xq ,?M~
~, Q2+~ R2- ~(~e6+~P2)
3~: ~ e~ ~ /
+ m,, = ~ e + ~b',,- 4 _ ~c'
. 3h z Is ~y
Transient response of symmetric cross-ply laminates 209

where a superposed dot denotes differentiation with respect to time, (gx, gr)
and (mx, mr) are the body forces and moments, respectively, q is the
distributed transverse load, and (Ni, Mi, P3 are the stress resultants 2
N

(Ni, Mi, Pi) = a!'~(1, z, z3) dz (i = 1, 2, 6)


m
m=l

(Q~, R~) = ~ l ~ t , z~) dz (7)


~=1

( ~ , R9 = ~ ( 1 , z~I dz
~=l

The inertias I~ (i = 1, 2, 3, 4, 5, 7) are defined by

(I~, I~, Ia, I~, I5, I~)= 2;?


~=1
p~(1, z, z ~, z a, z~, za)dz

where p ~ is the material density of the ruth layer,

4 4
i2 = I2 -- ~ I, is = Is -- ~ I7
(8)
8 16
~ = I~-~Is +~I v

and ~ and ~z are the constant in-plane edge loads.

2.4 Expressions of stress resultants

The resultants are related to the total strains in eqn (2) by:

N, = A,s~~ + B,j~ + E,j~


M, = B,~? + o,~? + F,~,~ (i,j = 1, 2, 6)
Pi = EusY + Fij~? + HijK ~
~2 = A,j~Y + D~j~ (9)
~, = ~j~y + D~j~3 (j = 4, 5)
~ = D~ + ~3
R1 = Ds~y + Fsy~
210 A. A. Khdeir, J. N. Reddy

where Ais, Bu, etc., are the plate stiffnesses, defined by


N
(Air, Bit, Dij, Eit, Fij, Hit) = QI~I( 1, z, z ~, z 3, z '~, z 6) dz (i,j = 1, 2, 6)
n,i

i'll = 1

N
(Aij ,Dij,Fij) = ~ u) ~ 1~ , ~-~
, z 4,~dz (i,j=4,5) (10)
m=l
where ~u
O ~ are the stiffnesses of the ruth lamina in the plate coordinates.

2.5 Governing equations of symmetric cross-ply plates

Owing to the symmetry of the composite structure, the bending and


stretching states of stress are decoupled. For this case, the following plate
stiffnesses are identically zero:

Bit = E u -- 0 for i,j = 1, 2, 4, 5, 6


A16 = A26 = D16 -- 026 = F16 -- F26 = H16 = H26 = 0 (11)
A45 = D,~5 = F~5 = 0

and the equations governing the bending theory of symmetrically cross-ply


laminated plates are given by: 2

4 ~ ~x ~ 4 ~(~ ~.,~ ~,
3h~ ~F" ~ + H,,K-~jK~+ ex~j + F,2 8x~@
H { 4 ~ { ~3~y ~, ~ ~3~x
+ 12k--~Jk~ + ~X2 ~y2; + F12 By: 8x

H (--~{ ~3ex ~., ~ (~3¢y


+ ~ 3h:J~Sy~Sx +Sxz@2j+F~: @3
fl 4 ~f183@y 84w~ ~ f 83@), 8~x ~
+ "~t-~)t~ + ~) + z " 6 t ~ + ~y~o.~;
2H i 4 ~/~3Ox <730, + 2 8~w ] ]
+ 66~-~)~+ex~e~: e~a~/~
4r /e~w eO.] ( ~)(eO. e~w)
h~[D'5~ + 8x ) + F'5_-h2jt 8x + 8x~)
~., ~0,3 (_~(~0>, ~w~]
+ D.~+ ~), ) + F.k ~)k ~J' + ~")J
÷ + ~ ÷
÷ ~
I l ÷ + ~
+ + ÷
~I~ ~
~-I~
m,-~ ~
- +
ml ~ ~ ~ I--I ~ ~ ~ + II
I i ~
+ ~ ~ ~ ~ ~ ~ I I ~
-
~
~
N ~'J ~
C
~ ~ ~" I I ~ + ~ ~ ~ ~ -I- m~ m,
~1% ~ ~
~1~ ~ + ~ ~ ~
~ ~I T~M ~ ~~ ~
÷ ~ .~ ~1~
+ ~ ~ • ~ ~
. ~_ _~ ~ ~
+ ~ -~ ,~ ~ ~ • ~ ~
~ ~ ~I~ ~ ~- ~~ + + ~~
+ + + ~ ~ ~ ~ ~
~~- ~ I
~ ~
~~ ~~ ~-~ ~i ~
~+ ~ ~
I ÷ + ~
+ ~ +
ml~ ~ ÷
÷
~ ~I~_
~~I ~. + ~
~~-~ ~ ~
+ L ~ I
~ l
~ ~ ~~ + ~j~ + ÷ !.__l
~ ~ ÷ ÷
~ ~1 ~ ~ ~l l--J + .m
~ ~ ~ +
~ ,~ ~
~1 ~
+

+ ~l +
t ~
~
I I
212 ,4. A. Khdeir,J. N. Reddy

4 I F //020, ~72#~'1 H {' 4 ){Ozqt~, 020, ~?3w I


3h2[ 66k 0X2 AVOyOx) "~- 6 6 k - - 3 ~ ) k 0 ~ y -b 0----~-nt-2OxZOy/
~¢~+n { 4 5/~. e3w ~

+F22 020' O~wll


~ + H~( _ ~3h2)k
){~20,
ay ~ + ay3j j (12~contd.
4~
+ #[D44~
/Ow + ~)+F~(
_
4~{Ow
_ --;j~ + ff,)] + m,

~ a2ffr 4 _ O3w
-~3~ 3h~I~a~ ~
For the sake of comparison the counterpart of eqns (12), derived as per the
first-order shear deformation theory (FSDT) are also recorded here. They
read:

_ ~2w
K~a~, ~ + (KgA~ + N,) ~
~ _ a=w
+ ~#~,, + (~#A,, + N ~ ) ~ + q = ~,~

a=¢~. a=¢,. (~,~ + ~ . a~¢~


~,,~+~6~+ ~6~-~A,, ( ~.+~)+m~=136~
aw~

~ ' a2~x a2~Y a2~Y 2 / aW~


(DI2 + ~ 6 6 ) ~ + D66~+ D22~
¢~3)
where K~, K~ are the shear correction factors. Note that the third-order
theory described earlier does not require the use of any shear correction
factors.

3 SOLUTION PROCEDURES

3.1 State space approach

The state space representations of the dynamic systems in eqns (12) and (13)
will be used to analyze the dynamic response of symmetric cross-ply
rectangular panels simply supported (SS) along all four edges. We will
Transient response o f s y m m e t r i c cross-ply laminates 213

represent the u n k n o w n quantities so as to satisfy identically the SS


b o u n d a r y conditions defined for H S D T and F S D T as:
at x = O , a w = ~b~= M~ = 0 FSDT
w = ~ky= M 1 = P~ = 0 HSDT (14)
at y = 0 , b w = ~/x = M2 = 0 FSDT
W = ~x = M2 = P2 = 0 HSDT
The following form of the solution satisfies the b o u n d a r y conditions in eqn
(14),

w = y, W,..(t) sin ~x sin fly


?~l,tl ~ 1

X.,.(t) cos ~x sin fly (15)


ra,n = 1

Or = ~ Ym.(t)sin ~x cos fly


ra,n = !

where e = mrs/a, fl = n~/b. Where the transverse load q(x, y, t) is defined as:

q= ~ Q~(Osinexsin~y (16a)
~,~ = 1

and

Q~(t) = ( ~ N t ) for uniformly distributed pressure (16b)

qo~t) for sinusoidally distributed pressure (m = n = 1)


the time-dependent dynamic loading Nt) appearing in (16b) is defined as:

{ sn(r~t/tl) 0<t<t~sine
l>l 1 3
loading

0 < t < tl~


F(t) =
f: t>t~ j
0 < t < ta~
/- step loading
(16c)

triangular loading
t>l 1 ;
e - ~1
explosive blast loading
Here mx = m r = O.
214 A. A. Khdeir, J. N. Reddy

Substitution of (15) and (16a) in eqns (12) and (13) results in a system of
ordinary differential equations in the time domain, which may be presented,
for both H S D T and F S D T in a similar way as:
~)~ran= C1Wmn "q- C2Xmn "-~ C3Ymn q- blQ.,.(t)
Jd,.. = C4W.,. + CsXm. + C6Ym. + b2Q~.(t) (17)
~',.. = C7 Win. + CaXm. + C9 Y,.. + b3Q,..(t)
the coefficients (C1 C9, bl-b3) appearing in eqns (17) are displayed in
Appendix A for both theories.
In order to reduce the system of eqns (17) to a state-space form, we define
the components of the state vector Z(t) as
g l = Wren, Z2 = ~mn, Z3 = ~e~mn, Z4. = ~mn, Z5 = rmn, Z6 = ~rran (18)
upon employing (18), the system of equations (17) may be converted to the
form
Z=AZ+b (19)
where the matrix A and the load vector b for both F S D T and H S D T are
defined as:
0 1 0 0 0 0-
C1 0 C2 0 C3 0
0 0 0 1 0 0
A= (20)
C4 0 C5 0 C6 0
0 0 0 0 0 1
C7 0 C8 0 C9 0
b; = {0, b, Qm.(t), O, bzQ,..(t ), 0, b3Qm.(t)} (21)
The solution to eqn (19) is given by Refs 20 and 21:

Z(t) = eAU-t°)Z(to) + ~t e Art °b(z)dr (22)


o
where t o is the starting time and Z(to) the input response while e Ate-r) is
expressible in the form
o
e ~" -~ = [R] - ' - - e X ~ , - o [R] - ~ (23)

where 2~ denote the eigenvalues while JR] denotes the matrix of eigenvectors
of A. In the present case all the eigenvalues are distinct. However, in the case
of repeated eigenvalues, the Jordan canonical form associated with the
matrix in eqn (23) is to be used.
Transientresponseof symmetriccross-plyla,minates 215

3.20rthogonality of the principal modes


This method was used frequently in the literature for FSDT (e.g. Refs 10-14).
In this paper only HSDT will be treated. We shall express the displacement
quantities in terms of the free vibration modes by:
~t3

w(x,y, t) = y, W,..(x,y)T,.,(t)
~lJI = 1

~o

~b,:(x,y,t)= y, W,,,..(x,y)T,..(t) (24)


~¢1,~1 = 1

~k~(x,y,t)= L q2,,..(x,y)T,~.(t)
Iql,tl = 1

where W.., Wx,. and Wr.. denote the natural mode shapes while T~.{t), is a
time dependent generalized coordinate. In the case of the real distinct
natural frequencies ~ . and ~,~ the orthogonality condition of the principal
modes may be obtained by following the lines used in Ref. 10. For this case
the orthogonality condition of the principal modes reads:

( ~ . -- ~L) f o ; { [ I, W.. -- (4~)~I{O2W..v,, + ~dzW'"'


) + ~4 I~
-

{OW~.. OW""I] [ 4
+ Oy ]A W,, + &Wx.. 3h 2 & Ox J ~"

[
+ /3~y~n 3h 2 [5 Wr," dx dy = 0 (25)

The natural frequencies and the corresponding principal modes of the


associated free vibration problem are to be obtained. For the case of free
vibration, the generalized coordinates T..(t) and the principal modes
( W , . , ~ .... Wry.) associated to the homogeneous system of the governing
equations (12) and fulfilling the SS edge condition defined in eqn (14) may be
expressed as
T..(t) = e
Wx.. = A~. cos ~x sin fly
(26)
~r~. = B~. sin ~x cos fly
W~. = C~, sin ~x sin fly
where A~., B~., and C~. are arbitrary constants.
216 A. A. Khdeir, J. N. Reddy

Substitution of(24) considered in conjunction with (26) in the equations of


motion (12) (q = O)yields a set of homogeneous equations that may be solved
for the natural frequencies of vibration.
[L]{A} = 0 (27)
where
[L] = [C] - tn#.[M]
and
{A}r = {A,.., B,.., C.,.} (28)
The elements of the matrices [C] and [M] are defined in Appendix B.
Three eigenfrequencies and their respective eigenvectors result from eqn
(27) for each m,n pair. The distributed loads are now expanded in a series of
the principal modes

~ ( 4 "~2 ['6~2Wmn ~2Wran"~


q: 1..,,,[,,,...- +
lii,#I = 1

4 -/Oq~x,.. Oq~y...\-]
(29a)

mx= •
rtl,ll=|
[ 4 ~.~.q
f.~,(t) -[3Wxm. 3h2 -1-~~-x j (29b)

mr=
2
~,~= I
f~,(t) hWr~. 3h 2 s Oy J (29c)

The generalized forces f~(t) are determined by making use of the


orthogonality condition in eqn (25). Multiplying eqn (29a) by Win.,eqn (29b)
by Wx,.., and eqn (29c) by Wr.,., adding the results, integrating over the plate
area, and taking into account eqn (25), leads to the result
~ob ~oa (rnxV~,,..+ mrq?y... + qWm,)dxdy
f.~.(t) = (30)
N~..
where
~ f (I 4l \zl ( 02 mran ~2 Wmn~
uo.=jojol.,,W.~.-t,~),~t, ~ + ~ )~v.,.
4
~-~-i'5 ~ + ~ % Wmn+ r 3 ( e ~ . + e f ~ . )

4 L{OW.. W OW.. )}
3h 2 ~k~ ~.+~Vr~. dxdy (31)
Transient response of symmetric cross-ply laminates 217

Substituting eqns (24) into equations of m o t i o n (12), taking into account


eqns (29), gives for any m, n

~',..(t) + w~.T,..(t) = f,..(t) (32)

for zero initial conditions, the solution to eqn (32) reads:

T~..(t) = 1 I t f,..(r) sin ~o,..(t - ~) dr (33)


(~Omn do

The solution for w, ~,, ~by may be expressed formally (m~, = my = 0) as:

for uniform load q(x, y, t) = qoF(t)

1
Jmn(k)OOmn(k)
~I~,~,,/~ = ~.~° ~os ~ ~ ~ ~ ~mn(k)
~rnn(k)(3-)mn(k)
~y(x,y,t) J ..=1 I.sin~x c o s / / y J ,=1 Bmn(k)
Jran(k)O)mn(k)

x Jo F(r) sin W.,.tk,(t -- ~) dr (34)

for sinusoidal load q(x, y, t) = qo sin ~x sin flyF(t) (m = n = 1)

~x~,.,,~ =
~r(x,y,t)J
2
~=~
~
Jmn(k)Omnlk)A mn(k) ~ cos ~
B..~, ksin ~x
s~n# ~
cos ~yJ

x j~ ~ ) sin ~ . ~ ) ( t -- v) dr (3~)

when n o ~ a l i z e d to C~.~) where

J~ = I~ + ~ I~(e ~ + B ~) - ~ I ~ ( e A ~ + BB~.,~)

+ I- -3 ( A2~ ~+ B~.~)
2
(36)
We have to mention that the solutions in eqns (34) and (35) will be applicable
also to F S D T but J ~ in this case is defined as:
J~ = I~ + I a ( A2~ ~+ B 2~ ) (37)
and the eigenfrequencies with their respective eigenvectors should be
derived from the free vibration problem associated with eqns (13).
218 A. A. Khdeir, J. N. Reddy

4 N U M E R I C A L RESULTS

The two procedures described in the previous sections are used to investigate
the dynamic response o f three-layer cross-ply (0°/90°/0 °) square (a = b = 5h)
laminates under various loads. Both procedures give the same result. The
lamina properties used in the study are
E 1 = 172.369GPa (25 x 106 psi)
G12 -- G13 = 3"448 GPa (0"5 x 106 psi)
v~2 =0"25

Ez = 6"895 GPa (106 psi)


G23 -- 1"379 GPa (0.2 x 106 psi)
p = 1603"03 kg/m 3 (0.0579 1 lb/in 3)

1.2 1.2

::' ::///::/i/:/'~
~',, :,/ ,/' :i... /"
0.6 0.6

,,: ",! ,~ /, ,~:'


--?.
z~O.O ~t, ~i
,~
~
,~ !
.~ 0.0 , ~" i~ ,".' i ",:/'i~

~ : ~
-0.6 -0.6

)
-i .2 -1.2
C.O 0.002 0.004 01006 0.008 0.0 0.002 0.004 0.006 0.008
Time (sec) Time (see)

(a) (b)
0.6 1.0

0.3
0.5

~0.0
v •:~~ 0.0

-0.3 -0.5

-0.6 -I.0
0.0 0.002 0.004 0.006 0.00~ 0.0 0.002 0.004 0.006 O,OOB
Time (sec) Time (sec)
(c) (d)
Fig. 1. Variation of the center deflection as a function o f time, for various plate theories a n d
pulses. (a) step; (b) triangular; (c) sine; (d) exponential ( - H S D T ; . . . . F S D T ; C P T). O n e
in = 2'54 cm.
Transient response of symmetric cross-ply laminates 219

All layers are assumed to be o f the same thickness and material properties.
The total thickness o f the laminate is taken to be h = 0.1524m (6in).
The load is assumed to be sinusoidal in spatial domain and is assumed to
vary with time according to one of the expressions in eqn (16c) with
t~ =0.006s. The initial conditions are taken to be homogeneous. The
intensity of the transverse load is taken to be qo = 68.9476 MPa (104 psi). For
the blast loading, 7 is set to 330 s - t. The shear correction factors for the first-
order theory are taken to be 5/6.
Figures 1-3 contain plots of center transverse deflection, and non-
dimensionalized normal stress # 1 and transverse shear stress #4, respectively,
as functions of time for various loads and the three theories:
~ , = ~ ,(a/2, b/2, h/2)/qo
64 = a 4(a/2, O, O)/qo
The origin o f the coordinate system is taken to be at the lower left corner of
36 3?

~ ,~ ~ ~ : ~

~, ' :I,' ' ~l~i,


~ I I ~ , • . ,, ~,,, ,,,
Ip, ~ ~ :,, ,,,
t~ !,,i'..'., 16
,! i, ~i i/ ],, :,,
,

"
~
~
,
,
~ ~
,,
~
~,,
,,
~
"
,~
~,
,~
~ ,~

I~O ~0 1 i
~
',
.~
~,
.,
t
~s ~
~ ~ ,~
~~ '
i ~
[I
,

-18 -16

-36 -32
0.0 0.002 0.004 0.006 0.008 0.0 0.002 0.004 0.006 0.008
T~m~ ( s e c )
(a) T~me ( s e c )

(b)
20
30

I0 IS , , ,~ ~ , ~
~ ~ ~ ~
, , , ~ ~, :~ ~ ~,

~0 ~o i/
"~' ~l,i !~,/~ ~,~!/i~ ~ ~)I
.
i '~ ' ,

~10 -15

-20 -30
0.0 0.002 0.004. 0.006 0.00~ 0.0 0.002 0.004 0.006 0.008
Ttme (sec) Ttme ( s e c )

(c) (d)
Fig. 2. Variation of the normal stress 61 as a function of time, for various plate theories and
pulses. (a) step; (b) triangular: (c) sine; (d) exponential [ - H S D T : .... F S D T ~ - CPT).
220 A. A. Khdeir, J. N. Reddy

0.9 Z.6

0.45 1.3

~"o .i o ~,
~ ~

-0.45 -I,3
Fig. 3(a)

-0.9
-2 6
0.0 0.002 0.004 0.006 0.008 0.0 0.002 0.004 0.006 0.008
Time (sec) Time (sec)

(a) (a)

1.4

10
0.7

i~o v

-o.1 -I0
Fig. 4(b)
Fig. 3(b)

-20
-1.4
0.0 0.002 0.004 0.006 0.008 0.0 0.002 0.004 0.006 0.008
Time (sec) Time (secl

(b) (b)
Fig. 3. V a r i a t i o n o f the transverse shear Fig. 4. The effect of thickness on the center
stress ~4 as a function of time, for various deflection of square plate (a = 3 0 i n ) under
plate theories and pulses. (a) sine; (bt sinusoidally distributed load. (a) h = a/10; (bt
exponential ( - - HSDT; .... FSDT). h = a/20( HSDT; .... FSDT; CPT).
1 in = 2.54 cm.

the laminate with the x-axis horizontal. Figure 4 contains plots of center
deflection versus time for plates with different thicknesses subjected to sine
pulse.
From Figs 1-3 it is observed that the transverse deflection predicted by the
classical plate theory is significantly different from that predicted by the first-
order and higher-order theories. The difference between the stresses
obtained by the classical plate theory and the shear deformation theories is
not as large as the difference in the respective deflections. The stresses
predicted by the classical plate theory are closer to the first-order theory
than to those of the higher-order theory, which is the closest to the elasticity
solution. 2
Transient response of symmetric cross-ply laminates 221

The effect of thickness on the transient response can be seen from Fig. 4.
As the thickness increases, the difference between the center deflections
predicted by various theories also increases, although the first-order and
higher-order theories give solutions that are fairly close.

5 CONCLUSION

The exact solutions for the dynamic response of symmetrically laminated


cross-ply plates are presented using a higher-order theory. Comparison of
the numerical solutions for various loadings indicate that the classical plate
theory gives results that are significantly different from the first-order and
higher-order theories.

REFERENCES

1. Reddy, J. N., A Review of the literature on finite-element modeling of laminated


composite plates. The Shock and Vibration Digest, 17(4) (1985) 3-8.
2. Reddy, J. N., Energy and Variational Methods in Applied Mechanics. John
Wiley, NY, 1984.
3. Reddy, J. N., A simple higher-order theory for laminated composite plates. J.
Appl. Mech., 51 (1984) 745-52.
4. Reddy, J. N., A refined nonlinear theory of plates with transverse shear
deformation. Int. J. Solids Structures, 20(9/10) (1984) 881-96.
5. Whitney, J. M. & Pagano, N. J., Shear deformation in heterogeneous
anisotropic plates. A S M E J. Appl. Mech., 37 (1970) 1031 6.
6. Pagano, N. J. Exact solutions for composite laminates in cylindrical bending. J.
Composite Mater., 3(3) (1969) 398-411.
7. Whitney, J. M., The effect of transverse shear deformation on the bending of
laminated plates. J. Composite Mater., 3(3) (1969) 534-47.
8. Bert, C. W. &Chen, T. L. C., Effect of shear deformation on vibration of
antisymmetric angle-ply laminated rectangular plates. Int. J. Solids Structures,
14 (1978) 465 73.
9. Reddy, J. N. & Chao, C. W., A comparison of closed form and finite element
solutions of thick laminated anisotropic rectangular plates. Nucl. Engng. Des.,
64 (1981) 153-67.
10. Sun, C. T. & Whitney, J. M., Forced vibrations of laminated composite plates in
cylindrical bending. Journal of the Acoustical Society of America, 55 (1974)
1003-8.
11. Sun, C. T. & Chattopadhyay, S., Dynamic response of anisotropic plates under
initial stress due to impact of a mass. Transactions of the ASME, J. Appl. Mech.,
42 (1975) 693-8.
12. Sun, C. T. & Whitney, J. M., Dynamic response of laminated composite plates
under initial stress. AIAA Journal, 14 (1976) 268-70.
13. Whitney, J. M. & Sun, C. T., Transient response of laminated composite plates
222 A.A. Khde~,ZN. Reddy

subjected to transverse dynamic loading. Journal of the Acoustical Society o f


America, 61 (1977) 1014.
14. Dobyns, A. L., Analysis of simply-supported orthotropic plates subjected to
static and dynamic loads. A I A A Journal, 19 (1981) 642-50.
15. Reddy, J. N., On the solutions to forced motions of rectangular composite
plates. J. Appl. Mech., 49 (1982) 403-8.
16. Reddy, J. N., Geometrically nonlinear transient analysis of laminated
composite plates. A I A A Journal, 21(4) (1983) 621-9.
17. Reddy, J. N., Dynamic (transient) analysis of layered anisotropic composite
material plates. Int. J. Numer. Methods in Engng., 19 (1983) 237-55.
18. Bhimaraddi, A., Static and transient response of rectangular plates. Thin-
Walled Structures, 5 (1987) 125~13.
19. Bhimaraddi, A., Static and transient response of cylindrical shells. Thin-Walled
Structures, 5 (1987) 157-79.
20. Franklin, J. N., Matrix Theory. Prentice-Hall, Englewood Cliffs, New Jersey,
1968.
21. Brogan, W. L., Modern Control Theory. Prentice Hall, Englewood Cliffs, New
Jersey, 1985.

APPENDIX A

T h e coefficients Ci (i = 1-9), bi (i = 1-3) a p p e a r i n g in eqns (17) a n d associated


with H S D T

C a=(e~e~e~6-eseve~6 --e6e,,ex2)/Co
C 2 = (e2e~e~6 -- esese~6 - - e h e , , e l ~ ) / e o
C 3 = (e3e~e~6 - esege~6 - - e 6 e , , e , 4 ) / C o
C 4 = (e4eve~6 -- e~e~oe~6 - - e 6 e v e ~ + e6e~oe~)/Co
C 5 = (e4ese~6 -- e2e~oe~6 -- ehese~5 + e6etoe~3)/Co
C 6 = (e4ege~6 - e3e~oe~6 -- e h e g e ~ + e6e~oex4)/Co
C v = (eseve~5 - e~e~e~5 + e ~ e ~ e ~ -- e~e~oe~2)/Co
C 8 = (esese~5 -- e2e~e~5 + e~e~e~3--e~e~oea3)/Co
C 9 = (esege~5 + e4e~e~4 -- e s e ~ o e ~ - - e 3 e ~ e ~ s ) / C o
C 0 ~ ese~oe16 + e 6 e ~ l e l s - - e 4 e x ~ e l 6

b~ = ellel6/co, b 2 =--e~oe16/Co, b 3 = - e l ~ e ~ s / C o

where

eI = {~2055 + fl2O4a) -- 0¢2A~ - f12A44 --


(4)2
~ (°¢2F55 + ~2&4 )

(4) 2
- ~ [~H~ + 2(H~ + 2H~)~B ~ + B~H~] - ~N~ -B~N~
Transient response of symmetric cross-ply laminates 223

e2= o~Ds~--eA~- ~ eFs~ + - ~ [ e F ~ +e~2(F~2 + 2Fa~)]

- ~ [ ~ + ~(~ + ~ ]

8 (4) e 4
e~ = ~ D ~ -- BA~ - ~ BF~ + ~ [ ~ B ( F ~ + 2F~) + ~ F ~ ]

-- ~ [ ~ f l ( g l z + 2H66) + f13H22 ]

e,~=- [
I~+ ~-~ i~(~2+~2) ]
4 4
e 5 ~ 3-~ ~5, e6 = 3 ~ fl~'5, e7 = e2

8 (4) 2 8 2
es=-~-~O~-A~s-ot2Da,--~2066-- ~ F55+~ (~ F l l +f12F66)

- ~ (~g~ + B~g~)

e~ = ~B ~ ( F ~ + F ~ ) - D ~ - D~a - ~ (H~ + H~a)

elo ~ ¢5, ~11 ~ ~ 3 ~ ~12 ~ ~3~ ~13 ~ ~9

~,~ = ~ - ~ - ~ _ ~ - ~ ~. + ~(~ ~ + B~e~I

- ~ (B~H~+~H~)

el5 = e6, el6 ~ e l l

The coefficients C i (i= 1-9), bi (i= 1-3) appearing in eqns (17) and
associated with FSDT

C, = ( - o~ZK~ A ~ - ~ZK,~ A,,,, - ~2N, _ flz ~2)/I~


Cz = - ~K~A~ ~/I,, C~ = - ~K~A,,,,/I~, C4 = - ~K~A~ s/I~
C~ = (--o~2D11 -- ~2D66 - K~A~)/I~
c~ = - ~ ( D ~ + D ~ ) / ~ , C~ = -I~K~A~,,/~, C~ -- C~
C9 = (--~2D66 -- fl~Dz~ - K~A,,,,)/I~
ba = 1/I~, b~ = O, b~ = 0
224 A. A. Khdeir, J. N. Reddy

APPENDIX B

The coefficients of the matrices [C] and [M] in eqn (28)


C l l ~-~ C8~ C12 ~ C9~ ¢13 ~ C7, C21 ~ C12, C22 ~ C14, C23 ~ e l 2

c31 ~ c 1 3 , C32 ~ c 2 3 , c 3 3 z c1

roll ~-~ C l l , m 1 2 ~ 0, m 1 3 ~ c l O ~ m 2 1 ~-- m 1 2 ~ m 2 2 ~ c 1 6 , m 2 3 ~ c 1 5

m31 ~ m13 ~ m32 ~ m23 ~ m33 ~ c 4

where ei (i = 1-16) are defined in Appendix A.

You might also like