Professional Documents
Culture Documents
R=19700009156 2019-11-22T17:42:42+00:00Z
NASA SP-160
PL
shapes of free vibration of plates would be provided for the design or develop-
ments and advice during the course of the work were also greatly appreciated.
I particularly wish to thank Messrs. Milton Vagins and S. G . Sampath, who
did all the necessary work so that I could be free for the actual summarization
and writing. Without their efforts in supervising the procurement of papers,
in manuscript editing, and in providing technical criticism, this work would
not have been possible. I wish to recognize the contributions of the project
advisory panel, which consisted of Mr. Michel, Drs. Robert Fulton, W. H.
Hoppmann, T. C. Huang, Eric Reissner, and Howard Wolko, who generously
met with me twice during the course of the project and offered their comments.
P also thank my colleagues, Drs. C. T. West and F. W. Niedenfuhr, for their
technical advice. Finally, the enormous editorial assistance of Mr. Chester
Ball, Mrs. Ada Simon, and Miss Doris Byrd of The Ohio State University is
gratefully acknowledged.
ARTHURW. EEISSA
T h e Ohio State University
0 S
CHAPTER PAGE
un en so
The classical differential equation of motion I n the case of a plate supported by (or
for the transverse displacement w of a plate is embedded in) a massless elastic medium (or
given by (see app. A): foundation), equation (1.1) becomes
DV4w +p -=O
d2W
at2
where D is the flexural rigidity and is defined by where K is the stiffness of the foundation
Eh3 measured in units of force per unit length of
D= deflection per unit area of contact. If the
12(1-v2)
foundation has significant mass, then its differ-
E is Young's modulus, h is the plate thickness, ential equation must also be written and a
v is Poisson's ratio, p is mass density per unit coupled system of differential equations solved,
area of the plate, t is time, and V4=VzV2, where which is beyond the scope of the present work.
V2 is the Laplacisn operator. Assuming the deflection form (eq. (1.3)) and
When free vibrations are assumed, the mo- substituting into equation (1.8) again results in
tion is expressed as equation (1*4), where now
w= w cos wt 11.3)
where w is the circular frequency (expressed in
radianslunit time) and W is a function only of Thus, all results presented in this section as
the position coordinates. Substituting equa- pertaining to the classical plate equation (eq.
tion (1.33 into equation (1.1) yields (1.1)) can also apply to the case of elastic
foundations by the simple use of equation (1.9)
(V4- k3 W= O (1-4) in place of equation (1.5).
where k is a parameter of convenience defined as
The location of a point P in polar coordinates
(1.5) is shown in figure 1.1.
0.7) FIGURE
l.l.-Polar coordinate system.
1
2 VIBRATION OF PLATES
- t ~ n I n ( b ) + D n K ( ~cos
)l
5 [&Jn(kr) f EYn(bT)
4-Tb=l
+CIn(kr)+EKn(kr)]sin n8 (1.18)
he strain energy of bending and twisting of
a plate expressed in polar coordinates is .P EtLlPTfCAL COORDlNATES
Elliptical coordinates 4 , are
~ shown in figure
1.2 and are related to rectangular coordinates
x,y by the relation
x + i y = ~ eosfi ( t + i q > (i= J-1)(1.19)
where 2c is the interfocal distance. Separating
real and imaginary parts of equation (1.19)
yields
where dA=r dr de. x=c cosh cos q
(1.20)
y=c sinb sin 11
.f .P Solutions
When Fourier components in 6 are assu 3.1 Classical Equations
The Laplacian operator in elliptical co-
ordinates is (refs. 1.2 to 1.4)
-
2- 2
cZ(c0sh 25-cos 2q)
FUNDAMENTAL EQUATIONS OF CLASSICAL PLATE 'i%EORY 3
Bending and twisting moments are related to the displacements by
2D d2w b2w (l--v)sinh2f *e (l--v)sin2q bw
Mi= -2(Cosh 2 , $ - ~ 0 ~ 2 q )bpSYbqZ-(cosb 2t-cos2q) dt (cosh2f-cos 2q)&
20 d2w b2w (l--v)sinh2f _- bw (l--v)sin2q bw
M,= -c2(cosh2.5-cos 2q) " ~ ~ ~ ~ ( c o s h 2 f 29) o s (cosh2~-cos 2q) &
- c df
(1.22)
2D (1-Y)
~~~=--cz(cosh2f-cos2q)~
b2W
(cosh 2,$-cos 2q)
3
and the transverse shearing forces are given by (ref. 1.4)
242D
--(cosh 2f-cos 2q
QE=C3(cosh25-cos 2 ~ ) ~ "
(1.23)
23D
"=c3(cosh 2F-cos 2q)5/2
COS 21)
=Wl+ w2 (1.26)
Y
where dA=dx dy.
1.3.2 Solutions
General solutions to equation (1.4) in rec-
tangular coordinates may be obtained by
assuming Fourier series in one of the variables,
say x; that is,
( 1.42)
M,,=-D--- (I-Y)
--sin.
d2W
a
b2W
COSa (b[bQ at The strain energy of bending and twisting of
8 plate expressed in skew coordinates is
(1.40)
1.4.2 Sofutions
--9 P There are no known general solutions to
I
I equation (1.4) in skew coordinates which allow
I
I a separation of variables.
1.2. NASH,W. A.: Bending of an Elliptical Plate by 1.5. MCLACHLAN, N.: Theory and Application of
Edge Loading. J. Appl. Mech., vol. 17, no. 3, Mathieu Functions. Oxford Univ. Press (Lon-
Sept. 1950, pp. 269-274. don), 1947.
1.3. GALERKIN, 3. G.: Berechnung der frei gelagerten 1.6. MORLEY,L. S. D.: Skew Plates and Structures.
elliptischen Platte auf Biegung. ZAMM, Bd. 3, ' Macmillan Co., Inc., 1963.
1923, pp. 113-117. 1.7. ~ D M A N , S. T. A.: Studies of Boundary Value
1.4. CHENG,SHUN: Bending of an Elliptic Plate Under Problems. Part 11, Characteristic Functions of
a Moment a t the Center. Tech. Sum. Rept. Rectangular Plates. Proc. N R 24, Swedish
No. 444, Math. Res. Center, Univ. Wisconsin, Cement and Concrete Res. Inst., Roy. Inst.
Dec. 1963. Tech. (Stockholm), 1955, pp. 7-62.
Chapter 2
es
Wn=[AnJn(kr)
+CJ,(kr)] COS ne (2.1)
(2.3)
TABLE
2.1.-Values of XZ=wa2 Jn
for a Clamped CtiTcUlaT Plate
A3 for values of 1~ of-
0 1 2
~-
10.2168 21.26 34.88
39.771 60.82 84.68
89.104 120.08 163.81
168.183 199.06 242.71
247.005 287.77 361.38
365.568 416.20 479.65
483.872 654.37 627.75
631.914 712.30 795.62
799.702 889.95 883.07
987.216 1087.4 190.4
-
circle. I t is seen from equations (2.2) that the plate be a (see fig. 2.2). The boundary
frequency does not depend upon Poisson’s ratio conditions are
in the clamped case. An accurate transcen-
dental approximating equation for additional W(a)=O
roots of equation (2.5) is given in reference 2.5. M,(a)=Q
The mode shapes of equation (2.1) are
determined from either of equations (2.2). Substituting equation (2.1) and equation (1.11)
Using the first of equations (2.2) into equations (2.9) and noting that dZw/bfl2=0
on the boundary give the equations
A n J n ( X ) + Cnrn(X>=O
where the X values are taken from table 2.1.
The radii of nodal circles p=r/a are determined
from the equation
(2.10)
0 1 2 3
I
1 1
1 1
.550 .613
1 1
.692 .726
.378 .443
1 t 1
.736 .765 .787
.469 .528 .570
.204 .288 .348
FIGURE
2.3.-Free circular plate.
___-
I 2 3 0 1 2 3 4 5
- --
- - - - - - -. 5.513 12.75 1.680 L781 0.822 1.847 1.863 0.881
20.41 35.28 53.16 .841 -871 .8897 .925 -926 .993
59.74 84.38 112.36 .391 .4972 .562 .605 .635 .663
118.88 153.29 191.02 .893 .932 .936 .939 .943 .947
196.67 241.99 289.51 .591 .643 .678 .704 .726 .745
296.46 350.48 408.16 .257 .351 .414 .460 .498 .529
414.86 478.73 545.83 ~941 .946 .95Q .951 .955 .958
553.00 626.75 703.63 .691 .723 .746 .763 .779 ,793
710.92 794.51 881.20 .441 .498 -540 .572 .600 .623
888.58 982.01 1078.5 .I92 -272 .330 .374 .411 f443
.952 .956 .959 .960 .963 .966
.752 .773 .790 .803 .814 .825
. 52 .590 .620 .644 .644 .682
The radii p=r/a of the nodal circles may .352 .407 .449 .483 .512 .536
be found from reference 2.20: .154 .222 -274 .316 .351 .381
-
x>-n2Jn(A) I +X2J,(1)
(1-v) [ATn(
J,(b>=
LI’(W
- n 2 I (A>
L (2.19)
Jn(XP) Jn(b>
(2.18) where is the pth root of the equation
Table 2.7 gives values of p=r/a for v=O.33 Jn(X)
=O.
computed from equation (2.18). Experirnen tal results were tained for a free
For large values of n and s it has been shown circular brass plate (ref. 2.21). The ratios of
(ref. 2.20) that the radii of nodal circles can be frequencies of free vibration o to the funda-
computed from the .approximate formula mental frequency w o are presented in table 2.8
12 VIBRATION OF PLATES
CIRCULAR PLATES 13
and figure 2.4 taken from reference 2.21. Radii
of nodal circles p=r/a are given in table 2.9.
Other experimental data are presented in
references 2.20 and 2.22 to 2.28. Further dis-
cussion of this problem is given in references
2.10, 2.11, 2.12, 2.15, 2.17, 2.29, 2.30, and2.31.
2.1.4 Plates With Elastic Edge Supports
Consider a circular plate of radius a. sup-
ported elastically by springs uniformly dis-
tributed about its contour as shown in figure
2.5, Translation in the direction of w is op-
posed by springs having distributed stiffness
K , (force/(unit length)'). Edge rotation fi is
opposed by spira1 springs having distributed
stzness K+ (moment/unit length).
Number of Diameters, n
ABLE 2.9.-Experimentally Determined Radii of Nodal Circles p=rfa for a Completely Free
Circular Brass Plate
W ( a ,e>=o
2Kwa2
a 0 i
I 1
.t 2
2(3--v)n2--X2---
D J*(N 8= 1
~-
39.7 21.2 34.8
39.7 21.2 34.8
39.1 20.9 34.2
35.2 18.6 30.8
30.8 15.0 26.7
29.7 13.9 25.6
4
2(2
+-j -3~Inz+ X2- 9.1.5 Plater ClamDed
x
and Simply Supported Along Remainder
(2.223
ows 5 circular plate whi
~ o r m ~ a t i of
o nthe second-order characteristic de- s edge for the interval
terminant for the frequencies from equations on y<e<2*--~.
(2.29) and 42.22) is a trivial operation. In the artlett (ref. 2.33)
case n=O, the frequency equation simplifies to by an interesting variational approach to give
CIRCULAR PLATES 15
-
x2 for values of y of-
Bound -_
0 r/S 2r/8 3r/8 4~18 5u/S 6r18 7r/8 ?r
- ---_________-----
_ ___ __ ___ __ __ 5. 871 6. 350
upper _ - _ _ _ _ _ _ _ _ _ _ _ _ 6.880 7.508 8.231 9. 120 9. 885 10. 21
Lower _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ 4.862 5.842 6.335 6. 864 7.480 8. 162 8.880 9. 126 ________
16 VIBRATION OF PLATES
X2=0a2@=3.752
X’“14.8
=49.4
TABLE
2.12.-Vdues of X2=uWa2m for Axiqrnmtric Vibrations of a Free Circular Plate Fixed
at the Center; v=l/S
.- I_-
~~-
0
3. 752
1
20. 91
2
61. 2
3
120.6 1
1 4
199. 9
5
298. 2
6
416.6
7
555. 1
I 10
1088
Xp
S
______
1
3. 97
2
7. 08
3
10. 20
4
13. 33
5
16. 49
6
19. 61
7
22. 75
~- g iI g
25. 90 29. 04
10
32. 18
I
CIRCULAR PLATES 17
The frequency equation for the clamped plate is acteristic determinant equation. The roots of
the determinant are found by evaluating it
by computer €or many values of X for a given
b/a ratio.
The numerical solution of this problem is
reported in reference 2.19 for the fundamental
mode. The frequency parameter X2 is plotted
which has as its first two roots: in figure 2.9 and mode shapes for three repre-
sentative b/a ratios are shown in figure 2.10,
A'= 22.7 both for ~'113.
=61.9
9.1.8 Plates With Concentrated Mass a t Center
9.1.7 Plates Supported on Circle of Arbitrary
The problems of free and clamped circular
Radius
plates having a concentrated mass m a t the
A circular plate having a free outside edge of center were solved by Roberson (refs. 2.39 and
radius a is supported on a concentric ring 2.40) for the case of axisymmetric modes. The
having a radius b as shown in figure 2.8. The concentrated mass was treated as an impulse in
solution of this problem is very straightforward. the mass density function. The impulsive
One can recognize symmetry and take change in density makes it convenient to solve
the problem by Laplace transform methods.
Wn,=An,Jn(b)+ Bn,Yn(b)+ Q n J n ( b )
In the case of the plate having free edges, it
+D&n(h) (i=1,2) (2.30) is shown (ref. 2.39) that the frequency equation
takes the f o m
from equation (1.18), where the subscript 1
refers to the region O<r<b and the subscript 2 (2.32)
refers to b<r<a; Bnf and Dnt are discarded
to satisfy regularity conditions at r=O. where
The remaining six boundary and continuity
conditions
w(b)=wz(b)=O L
--x (1-v,>Y,(x>r,(x)
bw,(b)
-=- atop)
Br ar
M,(4=V,,(a)=o
are satisfied by substituting equation (2.30) into
equatious (2.31) and forming a sixth-order ehar-
(2.33)
(2.34)
WO
FIQURE
2.1l.-Values of V=ua2&JB for various mass
FIQURE2.9.-Vdues of h2=wa2dTD for a circular ratios for a free circular plate having a concentrated
plate of radius a supported on a concentric circle mass at the center; v=0.3. (After ref. 2.39)
of radius b (for fundamental mode); V= 1/3. (After
ref. 2.19)
IW (2.35)
s
I- 00
S for values of p
I
of-
0
-
Error of
eq. (2.32) estimate, eq. (2.32) estimate,
2. 47 -33.8 9.006 9. 87 9. 6
22.20 6. 2 38.44 39.48 2. 7
61.69 I. 9 87.76 88.83 1. 2
120.91 1. 0 156. 75 157.90
199.85 246.74
298.56 355.32
416.98 483.60
W
a
3
io
a 0 0.05 0.10
FIQURE 2.12.-First mode shape for a free plate having The clamped case having a general concen-
a concentrated mass a t its center; v=0.3. (After
ref. 2.39)
trated impedance at the center was discussed
in reference 2.41, though no numerical results
were presented therein.
4.9 A ~ PLATES
~ ~ ~ A ~
An annular plate consists of a circular outer
boundary and a concentric circular inner bound-
ary. Throughout this work the radii a and b
will define the outer and inner boundaries,
respectively.
There exist nine possible combinations of sim-
ple boundary conditions (i.e., clamped, simply
supported, or free) for the two boundaries. An
outstanding set of results was given by Raju
F r a n ~2.13.-Values
~ of P = w u *for ~ various mass (ref. 2.42) for all nine combinations of boundary
ratios for a clamped circuhr plate having a con- conditions for a Foisson's ratio of 1/3, and the
centrated mass at the center. (After ref. 2 . 9 results which follow draw h e a d y from his work.
20 VIBRATION OF PLATES
Joga-Rao and Pickett (ref. 2.43) also evaluated I n addition, Southwell (ref. 2.37) presented
the exact characteristic determinants in the results for the outside-free, inside-clamped case;
axisymmetric case when the outside boundary Hort and Koenig (ref. 2.47) and Kumai (ref.
is clamped, simply supported, or free and the 2.48) gave theoretical and experimental results
inside boundary is free. Their results closely for annular plates of given dimensions; reference
match those of Raju and will not be repeated 2.47 deals with the free-free case and reference
here. They also analyzed these cases for 2.48, with the case for both edges either clamped
aJb=0.5 by the Rayleigh-Ritz method and or simply supported.
obtained confirming results.
2.2.1 Annular Plates Clamped on Outside and
Two-term Rayleigh-Ritz solutions were used
inside
in reference 2.44 to obtain approximate axi-
symmetric frequency parameters for all but the Substituting the complete solution (eq.
free-free eases. These results are summarized (1.18)) for the cos ne terms into the boundary
in table 2.16 for v = 113 and are compared with conditions W=dWJdr=O a t r=a and r=
exact solutions. The b/a ratio is 0.5 throughout yields four homogeneous equations in A,, B,,
the table. C,, and D, for which a nontrivial solution can
Sakharov (ref. 2.45) solved the cases for exist only if the determinant of coefficients is
zero. Using recursion relationships of the types
plates with the outside clamped or simply sup-
in equations (2.4) and equations (2.6), deriva-
ported and the inside free, and Gontkevich tives of the Bessel functions can be expressed
(ref. 2.6) presented results for four additional in terms of functions of the zeroth and first
cases but omitted those for the simply supported orders. The frequency determinants for n=O
inside boundary. VogeI and Skinner (ref. 2.46) (axisymmetric), n = 1 (one diametral node), and
in a recent paper also obtained exact solutions n = 2 (two diametral nodes) are given below
for all nine cases. (ref. 2.6).
TABLE
2.16.-Azisynmetrie Frequeny Parametersfor Annular Plates; v = I / S ; b/a=0.6
~
Boundary
conditions a
Deflection function W ( r )
Exact Rayleigh-
r=u r=b solution Ritz
solution
c e 89.30 89.42
c SS 64.06 65.17
c F 17.51 17.56
SS c 59.91 61.81
SS ss 40.01 43.19
ss F 5.040 5.062
F (2 13.05 13.59
F ss 4.060 4.084
For n=op
where a= ala.
CIRCULAR PLATES 21
For n=1,
For n=2,
=O
Fundamental roots for these three frequency equations are given in table 2.17.
10 TABLEB.IS.-Values of A2=oa2mD f o r a
Clumped, Simply Supported Annulus; Y = 1JS
9
xz for values of b/a of-
8 -
0.2 0.3 0.4 0.5 0.6
_I_- ______
40 01 02 03 04 0.5 06 07 08
1 A more comprehensive set of results is given
FIGURE2.15.--Values of A= ( p 1 2 / D ) ~ ffor
~ aa clamped, in table 2.20 (see ref. 2.46).
simply supported annulus; Y= 1/3. (After ref. 2.42)
The frequency d e ~ e r ~ ~ for ~ ~ s1, and 2 taken from reference 2.45 for plates clamped
n an=O,
on the outside and free on the inside are as follows:
=O
where
orn=l
CIRCULAR PLATES 23
=O
ax 3 i - v ax 34-v
A*=~--Ac, A=---, F=B-AD, B=-+--,
4 203% 4 2aA
48 ( 1 - v ) A
D= 12 ( 1-v) f (74- v) + -(ox)*
C= 1 2 ( 1 - + - (cyh)4’ 12(1- v ) 2 - ( a ~ ) 4
Eigenvalues from reference 2.42 are given in table 2.21 and figure 2.16. Results for b/a=O
are also given in the section entitled “Completely Free Rates” (2.1.3).
TABLE
2.21.-Vdues of k 2 = u a z ~ ~ f ao rChmped, Free Annulus; v = 1/3
TABLE
2.23.-Va~ues of X2=wa2@ for a Simply Supported, Clamped Annulus; v=l/S
where
reference 2.42 are given in table 2.23 and figure 2.17. Eigenvalues for
mode of n=O, tak from reference 2.6, are also given in figure 2.17. Additional infor-
mation appears in table 2
CIRCULAR PLATES 25
9
9
8
8
7
7
A 6
x
5 6
4
5
3
4
*O 01 02 03 04 05 06 07 08 09 IO
!I
3O 01 0.2 0.3 0.4 05 06 07 0.8
FIGURE2.16.-Values of X= ( p ~ a / D ) ~ ’for
4 a a clamped,
free annulus; Y = 1/3. (After ref. 2.42)
FIGURE2.17.-VaIues of X=(p3/D)1/‘a for a simply
supported, clamped annulus; Y= 1/3. (After ref.
2.42)
A more comprehensive set of results is given in table 2.24 (see ref. 2.46).
TABLE2.24.-Frequency Parameters w a 2 J m for a Simply Supported, Clamped Annular Plate
A more comprehensive set of results is given in table 2.26 (see ref. 2.46).
ABLE 2.26.-Frequency Parameters w a 2 mf o r a Simply SupporteE, Simply Supported Annular
Plate
1
n S
,_
I
__
0.1 0.3 0.5 1 0.7 0.9
-I
14. 5 21. 1 40. 0 110 988
16. 7 23. 3 41. 8 112 988
25. 9 30. 2 47. 1 116 993
40. 0 42. 0 56. 0 122 998
51. 7 81. 8 159 439 3948
56. 5 84. 6 161 441 3948
71. 7 933 167 444 3952
94.7 108 171 453 3958
where
=O
CIRCULAR PLATES 27
where
I ~~~~~~~
0.1
4.933
0.2
4.726
0.3
4.654
0.4
4.752
0.5
5.040
0.6
5.664
0.7
6.864
. 0.8
9.431
0.9
17. 81
13.91 12. 60 ________ 11. 66 _ _ _ _ - _ - 12.
_ 27 ________ 17. 05 ________
25. 43 24. 97 ________ 23. 09 ________ 22. 20 ________ 29. 92 ________
TABLE
2.28.-Frequenc.y Parameters u a 2 - J sf o r a Simply Supported, Free Annular Plate
15
*0 01 02 03 04 05 06 07 OS 09 10
3 i
FIGURE 2.18.--Values of k= (po2iD)W~for a simply sup- PEGTIRE2.19.-Values of k= (pd/D)If4a for a simply
ported, simply supported annulus; V= 1/3. (After supported, free annulus; V= 113. (After ref. 2.42)
ref. 2.42)
28 VIBRATION OF PLATES
The frequency determinants for n=Q, 1, 2 taken from reference 2.6 are as follows:
=Q
For n=1,
where
For 1x=2~
igenvalues from reference 2.42 are given in table 2.29 and figure 2.20. Accurate values for
b/a=Q are given in the section entitled ((Plates Clamped a t Center With Various Conditions
on Contour” (2.1.6).
TABLE
2.29.-Values f o r a Free, Clamped Annulus; v=1/3
of X2=~~ZJplD
n
I
I
X2 for values of bla of-
Additional data for this case are available from the work of Southwelf (ref. 2.371, who saved
considerable effort in computation of the Bessel functions by assuming arguments of X and then
finding the b/a ratios to which these correspond. These additional data are presented in table 2.31
tor v = 0 . 3 . Results appear also in table 2.16. This problem was also discussed in reference 2.15.
TABLE
2.31.-Additional Values of X 2 = o a 2 mfor a Free, CZQrrqped Annulus; v=O.S
1
n=O n=l n=2 n=3
--
bla 1 ??
The ease of annular plates free on the outside and simply supported on the inside is not discussed
in reference 2.6. Eigenvalues from reference 2.42 are given in table 2.32 and figure 2.21. Additional
information appears in table 2.16.
TABLE
2.32.-Values of X 2 = w a 2 mf o r a Free, Simply Supported Annulus; v = I / S
30 VIBRATION --
FIGURE
2.21.-Values of A= ( p d / D ) l ” a for a free, simply
supported annulus; v= 113. (After ref. 2.42)
A more comprehensive set of results is given in table 2.33 (see ref. 2.46).
TABLE JSfor
2.33.-Frequency Parameters @a2
a Free, Simply Supported Annular Pkte
-
n 8 -
1 0 a 2 m for values of bla of-
where
CIRCULAR PLATES 31
=O
where
A=’-(&) x 3+u C Q=
48 (1 -v)X
12 ( 1 -2 ) --x
-ba
FIGURE
2.22.-VaIues of h= ( p ~ 2 / D ) ~ f hfor
a a free, free
annulus; Y= 113. (After ref. 2.6)
32 VIBRATION OF PLATES
TABLE
2.34.-Values of X 2 = w a 2 m for an Annular Plate Free on Both Edges; v = I / S
~ ~~~
TABLE 2.35.-Frepuency Parameters wa2-JplD where M is the total mass of the rigid insert. In
for a Free, Free Annular Plate the general case the condition of zero slope a t
the junction with the rigid mass would be
replaced by an equation of motion relating the
% $ -
integral of the components of torque along the
0.1 0.3 0.5 0.7 0.9 edge r=b about a diametral axis to the product
-_-__ - - ~
___- of the mass moment of inertia and the rotational
2 0 5.30 4.91 4.28 3.57 2.94 acceleration about the axis.
3 0 12.4 12.26 11.4 9.86 8. 14 Letting n=Q in equation (1.18) and substi-
0 1 8.77 8.36 9. 32 13.2 34.9 tuting into equation (2.37) result in a fourth-
1 1 20.5 18. 3 17. 2 22.0 55.7 order frequency determinant. Expanding this
2 1 34.9 33.0 31. 1 37.8 93.8
3 1 53.0 51. 0 47.4 55.7 135
by making use of the recursion formulas for
0 2 38.2 50.4 92. 3 251 2238 derivatives of Bessel functions yields a char-
1 2 59.0 58.8 96. 3 253 2240 acteristic equation which was given by
Handelman and Cohen (ref. 2.49):
where
X = (w2p/D) /*a (2.39)
and
s=b&
2.15. PFEIFFER, F.: ubergang zu Systemen mit unend- 2.33. BARTLETT, C. C.: The Vibration and Buckling of
lich vielen Freiheitsgraden. Ch. 4, Handbuch a Circular Plate Clamped on Part of I t s
der Phys., Julius Springer (Berlin), 1928, pp. Boundary and Simply Supported on the Re-
337-402. mainder. Quart. J. Mech. Appl. Math., vol.
2.16. COURANT, F.; A N D HILBERT, D.: Metho& of 16, pt. 4, 1963, pp. 431440.
Mathematical Physics. Vol. I. Julius Springer 2.34. NOBLE,BEN: The Vibration and Buckling of a
(Berlin), 1924. Interscience Publ., Inc. (New Circular Plate Clamped on Part of Its Boundary
York, N.Y.), 1953. and Simply Supported on the Remainder.
2.17. RAYLEIGH, LORD:Theory of Sound. Vol. I and Proc. 9th Midwest. Conf. Solid and Fluid
vol. II. Dover Pub., 1945. (Originally pub- Mech., Aug. 1965.
lished in 1877.) 2.35. NOWACKI,W.; A N D OLESIAK,Z.: Vibration,
2.18. POLYA,G.; AND SZEGO, G.: Isoperimetric In-
Buckling, and Bending of a Circular Plate
equalities in Mathematical Physics. Princeton
Clamped Along Part of Its Periphery and
Univ. Press (Princeton, N.J.), 1951.
Simply Supported on the Remaining Part.
2.19. BODINE,R. Y.: The Fundamental Frequencies
Bull. Acad. Pol. Sci., el. IV, vol. 4, no. 4, 1956,
of a Thin Flat Circular Plate Simply Supported
pp. 247-258.
Along a Circle of Arbitrary Radius. ASME
Paper no. APMW-10, J. Appl. Mech., vol. 26, 2.36. NOWACKI, W.; AND OLESIAK,Z.: The Problem of
Dec. 1959, pp. 666-668. a Circular Plate Partially Clamped and Par-
2.20. COLWELL, R. C.; A N D HARDY,H. C.: The Fre- tially Simply Supported Along the Periphery.
quencies and Nodal Systems of Circular Plates. Arch. Mech. Stos., vol. 8, 1956, pp. 233-255.
Phil. Mag., ser. 7, voi. 24, no. 165, Dec. 1937, (In Polish.)
pp. 1041-1055. 2.37. SOUTHWELL, R. V.: On the Free Transverse
2.21. WALLER,MARYD.: Vibrations of Free Circular Vibrations of a Uniform Circular Disc Clamped
Plates. Proc. Phys. SOC. (London), vol. 50, at Its Centre and on the Effect of Rotation.
1938, pp. 70-76. Proc. Roy. SOC. (London), ser. A, vol. 101,
2.22. GRINSTED,B. : Nodal Pattern Analysis. Proc. 1922, pp. 133-153.
Inst. Mech. Eng., ser. A, vol. 166, 1952, 2.38. SAKHAROV, I. E.: Dynamic Stiffness in the Theory
pp. 309-326. of Axisymmetric Vibrations of Circular and
2.23. KIRCHHOFF, G.: Uber das Gleichgewicht und die Annular Plates. Izv. An SSSR, OTN, Mekh.
Bewegung einer elastischen Scheibe. Math. i Mashin., no. 5, 1959, pp. 90-98. (In Russian.)
J. (Crelle), Bd. 40, no. 5, 1850, pp. 51-58. 2.39. ROBERSON, R. E.: Transverse Vibrations of a
2.24. WOOD,A. B.:An Experimental Determination of Free Circular Plate Carrying Concentrated
the Frequencies of Free Circular Plates. Mass. J. App1. Mech., vof. 18, no. 3, Sept.
Proc. Phys. SOC. (London), vol. 47, no. 5, 1951, pp. 280-282.
1935, pp. 794-799.
2.40. ROBERSON,R. E.: Vibrations of a Clamped
2.25. COLWELL, R. C.; STEWART, J. K.; AND ARNETT,
Circular Plate Carrying Concentrated Mass.
H. D.: Symmetrical Sand Figures on Circular
Plates. J. Acoust. SOC. Am., vol. 12? Oct.
J. Appl. Mech., vol. 18, no. 4, Dec. 1951, pp.
349-352.
1940, pp. 260-265.
2.26. COLWELC,R. C.: The Vibrations of a Circular 2.41. TYUTEKIN,V. V.: Flexural Oscillations of a
Plate. 3. Franklin Inst., vol. 213, no. 1276- Circular Elastic Plate Loaded a t the Center.
1277, 1932, pp. 373-380. Akusticheskii Zhurnal, vol. 6, no. 3, July 1960,
2.27. STEWART, 3. X.; A N D COLWELC, R. C.: The CaH- pp. 388-391. (In Russian.)
culation of Chladni Patterns. J. Acoust. Soc. 2.42. RAJU, P. N.: Vibrations of Annular Plates. J.
Am., V O ~ . 11, July 1939, pp. 147-151. Aeron. SOC. India, vol. 14, no. 2, May 1962,
2.28. CHLADNI,E. F. F.: Die Akustik. Leipzig, 1802. pp. 37-52.
2.29. COLWELL,R. C.; STEWART, J. K.; A N D FRIEND, 2.43. JOGA-RAO, 6. V.; A N D PICKETT, GERALD: Vibra-
A. W.: Symmetrical Figures on Circular tions of Plates of Irregular Shapes and Plates
Plates and Membranes. Phil. Mag., sei. 7, With Holes. J. Aeron. Soc. India, vol. 13,
V O ~ 27,
. 1939, pp. 123-128. no. 3, 1961, pp. 83-88.
2.30. TIMOSHENKO, S.; A N D WOINOWSKY-KRIEOER, S.:
Theory of Plates and Shells. Second ed., 2.44. JOGA-RAO,C. V.; A N D VIJAYAKUMAR, K.: On
McGraw-Hill Book Go., Inc., 1959. Admissible Functions for Flexural Vibration
2.31. KIRCHHOFF, G.: Ges. Abhandl. (Leipzig), 1882, p. and Buckling of Annular Plates. J. Aeron.
259. Soc. India, vol. 15, no. 1, Feb. 1963, pp. 1-5.
2.32. KANTEIAM, C. L.: Bending and Vibration of Elas- 2.45. SAKHAROV, I. E. : Natural Vibrations Frequencies
tically Restrained Circular Plates. J. Franklin of Annular Plates. Izv. An SSSR, OTN, no. 5,
Inst., vol. 265, no. 6, June 1958, pp. 483491. 1957, pp. 107-3.10. (In Russian.)
lCIRCULAR PLATES 35
2.46. VOGEL, S. M.; AND SKINNER,D. W.: Natural 2.48. KUMAI,T.: The Flexural Vibrations of a Square
Frequencies of Transversely Vibrating Uniform Plate With a Central Circular Hole. Proc. 2d
Annular Plates. J. Appl. Mech., vol. 32, Dee. Jap. Nat. Congr. Appl. Mech., 1952, pp. 339-
1965, pp. 926-931. 342.
2.49. HANDELMAN, 6.;A N D COHEN,H.: On the Effects
2.47. WORT, W.; A N D KOENIG, M.: Studien iiber of the Addition of Mass to Vibrating Systems.
Schwingungen von Kreisplatten und Ringen I. AFOSR T N 56-387 (ASTIA DOC.NO.96045),
2. Tech. Phys., ser. 373, Bd. 9, Heft 10, 1928, Sept. 1956. Also Proc. 9th Int. Congr. Appl.
pp. 373-382. Mech., vol. VIP, 1957, pp. 509-518.
Chapter 3
es
TABLE
3.3.-Approximate ~ r e ~ u e Parameters
nc~ X2= w a 2 m for a Clamped Elliptical Plate
bla 1.0 0.9 0.8 0.7 0.6 0.5 0.4 0.3 0.2 0. 1
- - ~ ~ _ _ _ ~ - ~ ~
S
Frequency for value of n of- I Frequency for value of n of-
3 4
2. 58 4 7
5. 68 8.29
2 _ _ _ _ 4 . 2 5 6.57
9 . 4 3 12. 6
3 _ _ _ _ 10.6 14 ____
-_ __
_._ _ _.
___.
4 _ _ _ _ 17 22
es
with y2 as defined in equation (4.6). aries as nodal lines, except when the boundary
cf, Free a t x = O and Simply Supported is free.
at x=@: The frequency w is given by reference 4.1 as
X
X(x)=l-- (m=1) (4.14)
a
sinhyz($--i) (m=2,3,4, . a> (4.15) where G,, H,, and J, are functions determined
from table 4.1 according to the conditions at
with y2 as defined in equation (4.6). z=O and x=a.
The functions Y(y> are similarly chosen by The qunntities Gu,H,, and Jv are obtained
the conditions at y=O and y = a by replacing from table 4.1 by replacing x by y and m by n.
x by y, a by b, and m by n in equations (4.2) to An0 ther comprehensive set of s o htions was
(4.15). The indicators n and m are seen to be later given by Janich (ref. 4.2). Fundamental
the number of nodal lines lying in the x- and frequencies were obtained for 18 combinations
y-directions, respectively, including the bound- of boundary conditions. He, too, used the
TABLE
4.1.-Frequency Coefiients in Equation (4.16)
m G, HZ JZ
0 0 0 0
1 0 0 121s
2 1.506 1. 248 5.017
1
3, 4,5, .. ~
m-5
3
2, 3, 4, . . I
m-z
1 0 0 32.J
3
2, 3, 4, . . ~
m-z
1 0.599 -0.0870 0.471
2 1.494 1.347 3. 284
I 1
3, 4, 5, * . m- -
~
-2 2
'2= 0.
bx=a.
RECTANGULAR PLATES 43
Rayleigh method, but used simple trigono- SS-3'-F-F, and SS-SS-F-F) yield such poor
metric functions which satisfied only the geo- results with mode shapes of the same type that
metric boundary conditions. The mode shapes they were not included in reference 4.2. The
used in reference 4.2 are given in table 4.2. force-type boundary conditions as well as the
The frequency o is given in reference 4.2 for geometric are satisfied in reference 4.1; this
v=0.25 by usually improves the accuracy of the solution,
K
u*=-7r4D- but occasionally makes it worse. The results
(4.17)
dP N determined from table 4.1 will decrease in
accuracy for higher mode shapes (increasing
with K and N given in table 4.2. values of m and n ) .
The results of references 4.1 and 4.2 are both A partial summary of vibration frequencies
obtained by the Rayleigh method and, hence, for rectangular plates was given in reference 4.3.
yield upper bounds on the frequency values.
However, it must be pointed out that both sets 4.1 ss-ss-ss-ss
of results have limitations in accuracy. The The problem of plates with all sides SS is the
three cases not included in table 4.2 (F-F-F-F, most simple to solve for the rectangular plate.
TABLE
4.2.-F'requency Coe&ients for Equation (4.17)and Diferent Mode Shapes; v=0.25
2
(cof3""-1);
a .50 2.67+0.304
2nz
cos -- 1 1-50 8
a
2a 3z-c0s 2
2b 1.00
3UX
cos --cos
2a
-
UZ
2a
1-cos "y
2b
~ 227
44 VIBRATION OF PLATES
+0.0071(0'
(cos ---cos
3rz
2a
'>
2a
sin
50 1.28+1.25 (:)a+0.50 (:y
3rx
(cos %-COB ") .333 0.853+0.190
2a b (b)a
3rx *X
cos --cos
2a
-
2a
1.130 2.56
+0.1134 (is
(1-cosz); .0756 0.0104+0.0190 2
(by
*z
I-coe - .2268 0.0313
2a
sin -
TZ
a
sin 9
b
.25
TX
sin - .50 .50
a
The boundary conditions are satisfies the boundary conditions, where A,, is
an amplitude coefficient determined from the
w=O, M,=O (forx = ~ ,a ) initid conditions of the problem and m and n
(4.18)
w=O, Mv=O (for y=O, b ) are integers. Substituting equation (4.19) into
equation (1.4) gives the frequency
When equations (1.29) are use
(4.19) w=
2
+ (4.28)
RECTANGULAR PLATES 45
A plot of four frequency parameters as a func-
tion of the b/u ratio was made by Vet (ref. 4.4)
and is shown in figure 4.1.
090 w a z m
0
+B, c o s - , l k ~ y + C m
sinhd-y
b
+D, cosh&%& sin cu;l: (4.21)
m=2 .n. 2
Applying the remaining four homogeneous
FIGURE
$.Z.--Nodal patterns for SS rectangular plate boundary conditions results in a set of fourth-
with a> b.
order characteristic determinants, one for each
value of a. Each determinant has an infinity
4.4 TWO OPPOSITE SIDES SS of solutions for the eigenvalues k. Any of the
There are six combinations of boundary con- four edges being free is a necessary and SUE-
ditions for which two opposite sides are SS. cient condition for the frequency parameter to
One of these (for the plate with all sides SS depend upon Poisson's ratio.
which has a simple, exact solution) has already The first straightforward, comprehensive
been discussed. The remaining five cases also solution of these five cases by the method out-
have exact (although more difficult) solutions. lined above was given by Fletcher, Woodfield,
When the edges x = O and x=a are SS, it is and Larsen in reference 4.7 and in reference 4.8.
seen that the conditions at these boundaries, In reference 4.7 an exc n t analysis is made of
as well as the differential equation of the prob- the conditions which lead to k2<ru2 requiring
46 VIBRATION OF PLATES
For a nontrivial solution the determinant of the coe eients of equations (4.26) must vanish; that is,
(4.28)
which, when expanded, yields the characteristic os,)which would separate some
equation table 4.4 if a complete, sequ
available. These can be obt
O S cash A&-
~ X ~ X ~ ( C Alb 1) work of Odman (ref. 4.13) who
+(X;-A;) sin Xlb sinh X&=O (4.29) (4.29) with less accuracy than did
extracted the first six roots for
Iguchi (ref. 4.9) solved this problem in essen- The corresponding frequency parameters are
tially the same manner and obtained the first listed in table 4.5.
six frequency parameters for the case of the Nishimura (ref. 4.14) achieved accurate results
square. They are presented in table 4.3. for the square using relatively coarse finite dif-
For the frequency a,%, the subscript m identifies ference grids. He obtained o a 2 4 m = 2 8 . 9 7 4
the number of half-sine waves in the 2-direction for the fundamental mode by solving only third-
and the subscript n identifies the nth lowest root order finite-difference determinants.
for a fixed value of m. The results of table 4.3 For nonsquare plates, fundamental frequen-
are also verified in references 4.7 and 4.13. cies are available for various aspect ratios.
These are listed in table 4.6 (see also ref. 4.9).
TABLE 4.4.-16 Higher Frequency Parameters Aamada (ref. 4.15) used a variational approach
A = a a 2 ( m > (not a Complete Set) f o r SS- and Kanazawa and Kawai (ref. 4.16) used an
6-8s-C Square Plate
TABLE
4.5.-Frequency Parameters ua2JplD for
SS-C-SS-C Square Plate '
TABLE
4.6.-Fundamental Frequency Parameters for a SS-C-SS-C Rectangular Plate
Boundary conditions
&(&m
for value of alb of-
> I
1 1.5 2 2.5
integral formulation to obtain confirming results Eliminating three of the constants (e.g.,
or several alb ratios. I n reference 4.16, results B,, C,,, and 0,) in equations (4.26) in favor of
are also obtained for the mode antisymmetric a fourth (e.g., A,) leaves one equation giving
about p=O, for alb 2 1 . Unfortunately, this is the eigenfunctions, or mode shapes, for this
the second antisymmetric ' mode shape of the case. From reference 4.7 it is known to be:
plate. These frequency parameters are given
in table 4.7. W ( X ,y) =[(Gosh Azb-cos Arb) (AI sinh XZY
The first six roots of equation (4.29) for -Az sin Xy)
m = l , 2, . . ., 6 and for a/b=0.5, 1.5, and 2.0
were found in reference 4.13. The correspond- -(Al sinh AI sin Alb) (cosh by-cos X,y)j sin az
ing frequency parameters are listed in table (4.30)
4.8.
By using equation (4.24)) one can apply Substitution of AI and Az determined from
stddity results to this problem. Fundamental equations (4.27) into equation (4.30),using the
frequencies are listed in table 4.9 for various frequencies from the tables of this section,
a/b ratios as given on page 367 of reference 4.17. completely determines the mode shapes. Mode
TABLE
4.8.-Frequency Parameters cob2JpID for SS-C-SS-C Rectangukr Plate
a
- m
I 5 I 6
b
1 2 3 4
I5
10
05
c
- 0
X Y 5 0 9.P
-,-
" b - O b
3
-05
-1 0
-I 5
I5 15
10 10
05 05
5s -Ea ,-bl
0
E.2
-g
0"
o " b =
0"
0
-0 5 -05
-I 0 -I 0
-I 5 -I 5
I5 I5
10 10
05 05
-
E 0
-ma .-lb
g-05
-t 0 -1 0
-1 5 -1.5
(4 nl=3
(f)
ne6
1
waz&i- -
alb
---------_ _ _ _ _ _ _
I 0*4
0.5
13.718
0.6
15. 692
0.7
18. 258
0.8
20. 824
0.9
24.080
50 VIBRATION OF PLATES
40
20
.
0
0 10
08
-1.5 06
04
FIGURE4.6.-Variation in Y,(p) with alb for the mode
m=6, n = 5 for a SS-C-SS-C rectangular plate.
(After ref. 4.13) 02
10 20 40 60 80 100 xx) 400 6 0 0 8 0 0 1 ~
090 w b 2 m
1
B
(m,nintegers) (4.31) 6b
I
I
Other approximate formulas are given pre- I
viously in equations (4.16) and (4.17). Fre-
quency parameters obtained from equation
-x
(4.16) are given in reference 4.4 and are re-
produced as figure 4.7. FIGURE
4.8.-SS-GSS-SS plate.
1. 06 1. 02 1. 01 1. 00 1. 00
Mode _ _ _ _ _ _ _ _ _ _ _ _ oa2 a23 03s 1. 09 1. 05 1. 02 1. 02 1. 01
1. 09 1. 06 1. 03 1. 02 1. 02
1. 09 1. 06 1. 04 1. 03 1. 02
Ungar (ref. 4.22) presented an interesting 1. 08 1. 06 1. 05 1. 03 1. 02
table which shows the ratio of the frequencies 1. 07 1. 06 1. 05 1. 04 1: 02
of the SS-CSS-SS plate to those of the
(d,/o) __ - - - - - - - - - - - -
X= @a2 23. 646 15.573 12.918 11. 754 11. 142 9.869
X*=obz(dD) _ _ _ _ _ _ _ _ - _ _ _ _ 23. 646 18.899 17.330 16. 629 16. 254 15. 425
_. I
The mode shapes are (ref. 4.7) eters obtained from equation (4.16) are given
in figure 4.9 (ref. 4.4). The problem was also
(z, y) =(sin X,b sinh X2y-sinh X,b sin XI y> sin cyz discussed in references 4.23 and 4.24.
(4.34)
4.9.3 SS-C-SS-F
When k2+Cu2>>1, equation (4.33) reduces The boundary conditions for SS-C-SS-F
to (ref. 4.7) rectangular plates (fig. 4.10) a t y=O, b are
dW
(2, o>=- d?J (2,o>= (2,b)=Vw(z,b)=O
FJGURE
4.10.-SS-GSS-F plate.
sinh k2bsin X y
} sin ax
(4.42)
FIQTJRE
4.11.-SS-SS-SS-F plate.
n integers) (4.43)
(m,
TABLE4.15.-Fundamentu.l Frequency Param- Other approximate formulas are given by
eters for SS-0-8s-F Rectangular Plate; Y =0.66 equations (4.16) and (4.17).
By using equation (4.24), one can apply
stability results to this problem. Fundamental
frequencies given in reference 4.17 (p. 362) and
1.0_ _ _ _ _ _ _ _ _ _ 12. 859 18. 258 reference 4.25 (p. 297) are listed in table 4.17
1.1_ _ _ _ _ _ _ _ _ _ 13.520 19. 343 for various alb ratios for v=0.25.
1.2_ _ _ _ _ _ _ _ _ _ 14. 310 20.527
1.3_ _ _ _ _ _ _ _ _ _ 15. 198 21.910
1.4_ _ _ _ _ _ _ _ _ _ 16. 086 23. 192 4.2.5 SS-F-SS-F
1.5_ _ _ _ _ _ _ _ _ _ 17. 172 26. 153 The boundary conditions for SS-F-SS-F
rectangular plates (fig.4.12) at y=O, b are
2
*
d
-
H
2
1"
c
$
56 VIBRATION OF PLATES
0.1
%
3
M
M
n!
d
00
0
r-
w
E9
0
v?
0
RECTANGULAR PLATES 57
TABLE
4.21.--c11-itiCal Gonstank Detemining When k 2 < 2 for SS-F-SS-F Plate
(b') tc')
The detailed mode shapes are (ref. 4.7): Other approximate formulas are given in equa-
tions (4.16) and (4.17).
Zeissig (ref. 4.28) reported many experimen-
tal results which essentially substantiated his
analytical calculations. The problem was also
formulated in references 4.10 and 4.24.
4.3 OTHER SIMPLE EDGE CONDITIONS
4.3.1 All Sides Clamped (C-C-C-C)
The problem of C-C-C-C rectangular plates
(fig. 4.18) has received a voluminous treatment
in the literature, especially for the case of the
square plate. The first reasonably accurate
results for the square plate were given in 1931 by
Sezawa (ref. 4.21), who used the series method.
He used functions which exactly satisfied the
differential equation (eq. (1.1)) and the bound-
(4.47) ary condition of zero deflection along all edges
and required the slope to be zero only a t the
Mode shapes were computed and plotted in midpoints of the edges. This initial work has
reference 4.13 for the six roots of equation been followed by a host of Japanese publica-
(4.45) for m=1, 2, . . ., 6 and v=1/6. Unfor- tions on the problem; for example, see references
tunately, it was assumed that for the lowest root 4.9, 4.15,4.16,4.20, and 4.29 to 4.33.
(symmetry about y=O) for each value of m, the Some variation of the series method was used
plate behaves exactly like a beam and, conse- in references 4.9,4.20, 4.21,4.29,4.30,4.32, and
quently, these cases were omitted in the results. 4.34 to 4.40. Particularly notable is Tomotika’s
Thus, the plotted mode shapes begin with those work (refs. 4.30 and 4.41); he determined
antisymmetrical about T=O. Plots are given the fundamental frequency for the square plate
in reference 4.13 for a/b=0.5, 1.0, 1.5, and 2.0 with extreme accuracy. Like Sezawa, he chose
and those for a/b= 1.0 reproduced in figure 4.16. functions which satisfied the deflection condi-
The mode shapes are represented as the products tions exactly and set up an infinite characteris-
Wrne(Z, =Xrn(Z)Yn(g, where Z and ij are tic determinant for the slope conditions.
measured with the point at the center of Convergence of results from a sequence of deter-
the plate taken as origin (see fig. 4.12). minants obtained by truncating the infinite case
Each of the six parts of figure 4.16 corre- was used to get extreme accuracy. Me also
sponds to one value of m. The f i s t six modes used the Rayleigh and Weinstein methods to ob-
g that value of m are then determined tain the frequency bounds 35.9855< ( u a 2 d a )
from the separate curves Y,(jj). The curves <36.09 for a square of dimension a x a.
for Y,(5) do not change markedly €or Finite difference techniques were used in
variations in alb in the range 0.5<a/b<2.0. references 4.14, 4.38, 4.42, and 4.43; the Galer-
The maximum variations for the 36 modes kin technique, in references 4.13, 4.33, 4.44,
shown are illustrated in figure 4.17, which cor- 4.45, and 4.46; the Rayleigh or Rayleigh-Ritz
responds to m=5 and n = 5 . method, in references 4.1, 4.2, 4.47, and 4.48;
>
When k2/a2> 1, equation (4.45) reduces to Weinstein’s method, in reference 4.49; integral
(ref. 4.7)
equations, in reference 4.16; and a variational
approach, in reference 4.15. Other publica-
tions include references 4.18, 4.31, and 4.58 t o
4.56. A notable lack of experimental results
n integers) (4.48)
(m, exists.
RECTANGULAR PLATES 59
15 I5
IO 10
0.5 05
0
g2 0
g o B,I n l
-0-
c a b a b
g n
-05 -0 5
-10 -I 0
-I 5 -I 5
(a) (d)
15
85
10 IO
-
0
05
-c 05
-
y o
0
n
n
73 '9E $ 0
s
-05 -05
-I 0 -I 0
-1 5 -1 5
15 1.5
10 10
05 55
-
c
-.- -
g o
0
E l i
-pg o a P
O b L
2
T.6
-05 -05
-LO -I 0
-I 5 -1 5
(4 (f)
P
!‘6
Mode 1 1 2 3 4
-
5 6
Ampli- All= 1.0000 Am= 1.OOOO Azz= 1.0000 Ai3= 1.0000 Ail= -0.0280 Alz= - 0.0406
tude eo- A13=0.0142 Al4=0.0101 Azr=0.0326 Ai5=0.0085 A13= 1.0000 Air= -0.0105
efficient b A16=0.0020 A16=0.0020 Aza=0.0073 A31= - 1.OOOO A is= 0.0055 Ala= -0.0017
A31=0.0142 A~z=0.0406 A42=0.0326 Aa5= -0.0141 Aa1= 1.0000 Aa= 1.0000
A33= -0.0031 Aa*= -0.0022 A44= -0.0019 &I= -0.0085 A.qs=0.1267 A34= 0.0560
A36= -0.0007 Ala= -0.0010 A63=0.0141 A35= 0.0118 A3~,=0.0141
A.q=0.0070 Aaz=0.0073 A51= 0.0055 A52= 0.0238
A54= -0.0011 A@= -0.0010 A53=0.0118 A#= -0.0011
Am= - 0.0005 Am= -0.0006 A55= -0.0018 Aae= -0.0009
(mS-3‘>l’d;
I
1062.5
omn=2 a2 - (4.50) 1147. 1
1315. 4
Bazley, Fox, and Stadter (ref. 4.58) used a 1393.4
method developed in reference 4.59 to compute
Iower bounds for the fist 15 frequencies of the
following symmetry class of a square: With an C)dman (ref. 4.13) used a variation of the
?@-coordinate system having its origin at the Galerkin method and mode shapes of the form
plate center and axes parallel t o the edges, the W(Z,J)=X(Z)YY(jji),where
modes are symmetric with respect to both Z
and 5 and are unaltered by interchange of Z X(Z>=A,cosh p1Z+A2sinh filf
and 7 (fourfold symmetry). (Thus, the first +A, cosh &+AI sinh p2Z
and fifth modes of table 4.22 would be the only (4.51)
modes shown which would fall into this sym- Y(jj)=BIcoshp31H-B2sinhp3jj
metry class.) They also obtained extremely +B3 coshpL4jj+B4sinh
p477
accurate upper bounds by the Rayleigh-Ritz
method by taking the first 50 admissible prod- and where pr,. . ., p4are determined by applying
ucts of C-C beam functions. Double-precision the Galerkin formula to the differential equation
arithmetic (16 signifcant figures) was used in of motion for the plate. The 36 frequencies
the computations where necessary. Results w,,(m, n= 1, , . ~,6) computed by this method
are listed in table 4.25. I n this table results in reference 4.13 are upper bounds and are
from the Rayleigh-Ritz method are given using given in table 4.27. It is interesting to note
both 25 and 50 admissible functions to show that, in spite of apparent numerical precision,
the rate of convergence. &man did not detect two separate frequencies
Another significant contribution was made for w13, as did Young (table 4.22).
by Aronszajn (ref. 4.49),who used Weinstein’s For computing fundamental frequencies of
method to obtain accurate lower bounds for clamped rectangular plates of arbitrary a/b ratio,
the first 10 frequencies of a square plate. The there exists, in addition, Warburton’s (ref. 4.1)
Rayleigh-Ritz method was used to obtain and Janich’s (ref. 4.2) formulas, equations (4.16)
upper bounds. These results are summarized and (4.17). Frequencies obtained from War-
in table 4.26. burton’s formula were plotted in reference 4.4.
62 VIBRATION OF PLATES
TABLE
4.26.-F'requency Parameters for a C-C-C-C SpwtTe Phte
~~
Mode symmetry
Lower bound Upper bound Mean value Maximum
error, percent
I 1
2
3
131. 902
210. 526
l 4 5
309. 038
1 6
(428)
165. 023
242. 66 340. 59 458.27
220. 06
296.35 393.36 509.9
._________371.38
____ 467. 29 583. 83
562. 18 (676)
------------. 792. 5
A simple formula derived by Galin (ref. 4.45) plained previously. More extensive results are
for this case is obtained in reference 4.13 and are also listed in
w=12 4; (4.52)
table 4.29.
Mode shapes in the form W,,(&, 1J>=
X,(Z)Y,(F) corresponding to w,, were found in
For a square this reduces to wa2&@=36, which reference 4.13. The components X,(Z)& and
compares favorably with the accurate value of Yn(jj)-,hare shown in figure 4.19 for a/b=1.0.
35.9866 from table 4.22. Variation in these curves with alb is very small
A summary of the literature for frequencies for the range 0.5 Salb 5 2.0. The magnitude of
of nonsquare C-C-C-C rectangular plates is this variation is shown by figure 4.20 for the
presented in table 4.28. Neither Iguchi (ref. components X4(Z)& and Y,(jj)G. Figure
4.9) nor Kanazawa and Kawai (ref. 4.16) recog- 4.21, taken from reference 4.60, shows the
nized the existence of the other mode having x
frequency parameter -= w a 2 ( w D ) / a 2plotted
one symmetry axis and one antisymmetry axis 2
which is not shown in the table. as a function of alb and bla. For a/b=Q, the
Sixteen frequency parameters for a/b=0.25 frequencies are given by reference 4.60:
and 0.50 are computed in reference 4.60. These
are given in table 4.29, with m and n as ex-
RECTANGULAR PUTES 63
TABLE
4.28.--wb2%1mfor C-6-6-6 Rectangular Plates
wb2dm for values of a/b of-
Source Mode (a>b) ~~
1.5 2 2.5 3 m
_ _ _ ~
a
Iguchi (ref. 4.9) 27.00 24. 56 23. 76 23. 19 22.37
b
I I I
0 0 5 I 05 0
FIGURE4.19.-Mode shape components X,(Z)& or 0- !!,
Yn(@)& for a C-GC-C rectangular plate. (After ref. b D
4.13)
4.2l.--X/nz= w a 2 / G ( d a ) for a C-C-C-C
FIGURE rec-
tangular plate.
m n 0.50
0.667
(ref.
Ref. Ref. 4.13)
4.60 4.13
___
o/b
1 24. 09 24. 58 27. 01
2 31.40 31. 83 41. 72 FIGURE 4.22.-Frequency parameters A/$= w a 2 / 9 ( . \ l D )
3 44.35 44. 78 66. 53 for modes symmetric about both 2- and g- axes for
4 63. 00 63. 34 100.81 R G C - C - C rectangular plate. (After ref. 4.35)
5 .- - - - - - . 87. 26 144 21
6 - ------. (117) ( 195)
1 63. 93 (64. 1) (65. 5)
2 70. 90 71.08 79.81
3 82. 90 (83. 2) (103)
Y
4 100. !F: 100.80 136. 10
5 - ------- ( 124. 2) (178)
6 .- - - - - - . 151. 91 230. 04 b x
1 123. 07 ( 124) (126)
2 130. 13 130. 35 138. 64 a
4.3.2 C-C-C-SS
Three sources of numerical data are available
olb for the problem of the C-C-C-SS plate (fig.
FIGURE4.25.-Frequency parameters x / ~ z =oa2/rz(dplo) 4.27). Results are listed in table 4.33 for the
for modes antisymmetric about both 3- and g-axes case of the square.
for a C-C-G-C rectangular plate. (After ref. 4.35) Some higher frequencies for the square were
obtained by Kaul and Cadambe (ref. 4.61) as a
third symmetric-antisymmetric frequencies special case of the parallelogram plate by using
cross in the vicinity of a/b=0.84. Such an the Rayleigh-Ritz method and beam functions
intersection point is termed a “transition (see sec. 5.1.1). Frequencies for four higher
point.” I t is the contention of Claassen and modes are presented in table 4.34.
Thorne that these curves do not actually cross For a general rectangle, a spectrum of funda-
at transition points but only approach each mental frequency parameters is given in
other closely before [‘v g away” or being table 4.35.
“repelled.” Very small ements of alb are Frequencies for the first antisymmetric mode
taken in reference 4.36 in the vicinity of th with respect to z=a/2 are given in table 4.36
transition points and corresponding values (ref. 4.16). However, it is obvious that this
frequency parameter X are computed which is at least the third mode of all mode shapes of
appear t o substantiate this. The details of a plate for a/b$1. No detailed mode shapes
this phenomenon can be seen in table 4.31. are available in the literature, but for a/bS 1
From the table it is seen that the two curves the second mode clearly must have a nodal line
approach each other most closely a t a/b=0.834. essentially parallel to the x-axis and located
I t is the opinion of the writer that, although above ‘y= b / 2 .
extremely precise work was performed in refer- Approximate formulas for frequencies are
ence 4.36, certain questions of convergence of given previously in equations (4.16) and (4.17).
the series approach used need to be answered Frequency parameters obtained from equation
before the transition-point phenomena de- (4.6) are plotted in figure 4.28 (from ref. 4.4).
scribed above can be accepted. For more information on this problem, see
In figure 4.26 are shown nodal lines for one the discussion of the antisymmetric modes of a
quadrant of the plate for various a/b ratios in 6-C-C-C rectangular plate in the preceding
the vicinity of transition points (ref. 4.36). section (see. 4.3.1). Straighb nodal lines of anti-
In these figures the center of the plate is at symmetry duplicate SS boundary conditions.
(0,O) and the C and 7 coordinates have been
nondimensionalized to Z/a and Gib, respectively. 4.3.3 C-C-C-F
The rapid change from one mode form to The only known results for the problem of the
another with smdl variation in alb is interest- 6-C-C-F plate (fig. 4.29) are the approximate
ing. Precise node-line coordinates used for formulas, equations (4.16) and (4.17).
figure 4.26 and other nodal patterns are gven
in reference 4.36. 4.3.4 C-C-ss-ss
Accurate upper and lower bounds for the Four sources of numerical data are available
doubly symmetric modes of a rectangle (see for fundamental frequencies of 6 - C - S S S S rec-
discussion earlier in this section) are reported tangular plates (fig. 4.30). The results are
66 VIBRATION OF PLATES
RECTANQULAR PLATES 67
,
68 VIBRATION OF PLATES
0.5
05
04
03
-
-bY
02
01
0
0 02 03 04 05
Id) i/G
RECTANGULAR PLATES 69
summarized in table 4.37. Kanazaiva and and retained C,,, C,,, Gal, and CZ3. Nishimura
Kawai (ref. 4.16) used an integral equation (ref. 4.14)used finite difference equations. Ap-
formulation. Hamada (ref. 4.15) used a varia- proximate formulas, equations (4.16) and (4.171,
tional approach. Hwato (ref. 4.62) used the may also be used. Frequency parameters ob-
itz method and mode shapes of the tained from equation (4.16) are plotted in figure
form 4.31 (from ref. 4.4) for four modes.
For more information on this problem, see
the discussion of the doubly antisymmetric
modes of a 6 - 6 6 6 rectangular plate (sec.
-cos __ (COS 3- COS
3nry
~
2b
(4.553
4.3.11. Straight node lines of antisymmetry
duplicate simply supported boundary conditions.
70 VIBRATION OF PLATES
3;
Eg
4P
RECTANGULAR PLATES 71
TABLE
4.33.-Frequency Parameters wa2JplD for TABLE4.34.-Frequency Parameters for Higher
a C-C-C-SS Square Plate Mode Shapes of a C-C-C-SS Square Plate
I I I I I
w a 2 m D for mode- Mode 6 7 8 9
Source ~-~
I
1 2 3 4 waz&p _______ 130.84 152. 75 160. 00 209.97
-----
Dill and Pister
(ref. 4.24) _ _ _ _ _ 31. 83 63. 33 71. 08 100. 8 116. 4
Kanasawa and 4.3.5 C-C-SS-F
Kawai (ref.
4.16)--__------31.88 - - - - - - 7 1 . 2 6 _ _ _ _ _ _ - _ _ _ _ _ The only known results €or the problem of
Hamada (ref.
4.15)---------- 31. 83 . . . . . . . . . . . . . . . . . . . . . . . . the G G S S F plate (fig*4.32) are the approxi-
mate formulas, equations (4.16) and (4.17).
TABLE
4.35.-Fun&mental Frequency Parameters w a 2 mf o r a C-C-C-SS Rectangular Plate
~ O I
0.333
_
0.4 i 0.5
_
1 0.667 1 1
_
1 1.5
~
2
-
~
Dill and Pister (ref. 4.24) _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ 24.49 _ _ _ _ _ _ _ _ 31. 83 ________ 73. 07
Kanasawa and Kawai (ref. 4.16) _ _ _ _ 22.39 23.40 23.76 2 4 48 26.23 31.87 ________ ________
Hsmada (ref. 4.15) _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ 25. 85 31. 83 48. I _ _ _ _ _ _ _ _
TABLE
4.36.--FundamentaE Frequency Parameters wa2J- f o r the First Antisymmetric Mot& of a
C-C-C-SS Rectangular Phte
61.781 63. 947 6 4 366 65. 161 66. 971 71. 259
Q
TABLE
4.37.-Fund~mental Frequency Parameters wa2&@ for a C-C-SS-SS Rectangular Plate
72 VIBRATION OF PLATES
V Y
X -x
0.90w b ' m
0.90 w b 2 m
Y v
TABLE
4.39.-Eigenfunetion Parameters for a
C-F Beam
1. 8751041
-X
4. 6940911
7. 8547574
FIGURE4.33.-GGF-F plate.
10.9955407
14. 1371684
(2r- l ) r / 2
A fundamental frequency of large error is
also computed in reference 4.48 by use of the
R ayleigh-Ri tz method.
Results from using the Galerkin method are
Y
given in reference 4.46; these results also appear
to have considerable error, particularly for the
fundamental mode. Approximate formulas,
equations (4.16) and (4.17), may also be used.
4.3.7 c-ss-c-F
The approximate formulas, equations (4.16)
and (4.171, may be used for the problem of a
6-SSC-F rectangular plate (fig. 4.34). Addi-
tional information can be obtained from an
antisymmetric mode of the case ob the C-F-
6-F plate (sec. 4.3.10). Straight node lines of FIGURE4.34.-GSS-GF plate.
TABLE4.38.--First Five Sets of Frequency Parameters, Nodal Lines, and Amplitude Coe@cients for
a 6-6-F-F Square Plate; v=O.S
2 3 4 5
308-337 0-7-
74 VIBRATION OF PLATES
4.3.10 C-F-C-F
Claassen and Thorne (ref. 4.36) used the
series method described in the section for
FIGURE
4.37.--C-F-C-F plate.
FIGURE4.35.--C-SS-SS-F plate.
(c ) (4
RECTANGULAR PLATES 75
from table 4.40 that the doubly symmetric TABLE 4.41.-Frequency Parameters A= ua2JpID
frequencies all vanish as b/a+O. However, as and X*= ub2JTDfor the Symmetric-Antisym-
b is held fixed and a becomes infinite, it is metric Modes of a C-F-C-F Rectangular Plate
obvious that the boundary conditions at x = O
and x= m are no longer significant, and the I Mode
fundamental frequency becomes that of an
a / b for A,
infinite strip having two node lines parallel to bla for X *
the x-axis. Additional frequency parameters in
the vicinity of “transition points” (see see.
4.3.1) and detailed coordinates of nodal lines
are given in reference 4.36. 26.40 79. 8 79. 8
Approximate values of frequency parameters 2 2 10 68. 4 76. 1
are given by equations (4.16) and (4.17). 18. 22 58. 2 72. 9
14.75 49. 3 65. 2
11.68 41. 8 49. 5
8. 99 35. 5 36. 2
TABLE 4.4Q.-Frequency Parameters X= 6. 65 30. 6 25. 1
W a 2 m and k*=ub2,1plD f o r the Doubly 4. 63 26. 8 16. 3
Symmetric Modes of a C-F-6-F Rectangular 2. 88 24 2 9. 4
Plate 1. 36 22. 7 4.1
.0 22. 4 .o
I Mode
TABLE
4.44.-Frequency Parameters for a C-F-
SS-F Square Plate; v = 0 . 3
Mode
TABLE4.45.-First Five Sets of Frequency Parameters, Nodal Lines, and Amplitude Coe$cients for
a Square Cantilever Plate; v=O.S
1 4 5
2 1 3
Nodal lines
__-_____-___-____ 1.8751041 _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _
__-____--__--____ 4. 6940911 _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _
0. 98250222 7. a547574 4.7300408
1. 00077731 10.9955407 7. 8532046
. 99996645 14. 1371684 10. 9956078
1. 00000145 (2m- 1)1r/2 14. 1371655
o a 2 q D for m o d e
Type of data
b
+ enY
e n Y cos -
cosh -
b
I I I
-cos~-a, sinhL--sin*)
e x
e ax ( a a
where the values of A,, are given in table
4.45 and those of a and e are given in table
4.46.
In references 4.65 t o 4.68, Reissner's varia-
tional principle (ref. 4.69) is modified and
applied to the square plate. As in the RayIeigh-
Ritz method, generalized force boundary COIF
ditions may or not be satisfied here. In reference
4.66, moment boundary conditions were satis- 0
112 2 5 2 .i I
Mode
shape shape
3.508 3.472 3.450 94.49 563.9
Fifth mode
Second mode
5
2 5 Mode
Mode shape
shape 105.9
14.93 34.73 -__
A Experimental
0 Theoretical
FIGURE
4.41.-Theoretical and experimental mode shapes for a square cantilever plate.
I
Mode Upper bounds Upper bounds
Lower
bounds bounds
Ref. 4.75 Ref. 4.76 Ref. 4.75 Ref. 4.76
FIGURE
4.43.-Frequency parameter X/a2=wa2dm ( ~ 2 ) for a rectangular cantilever plate; v= 0.3.
(After ref. 4.77)
Aspect
I
l X and X* for mode-
Aspect ratio
1 A and A* for mode-
ratio / 2 ! 3 1 4 1 5
l
1 1
1 1 1 2 1 3 ! 4
alb x x
_-
1. 00 _ _ _ _ _ _ _ 3.472 21. 29 27. 2 54. 3 61. 3 31. 1 64. 2 71. 1
0.95 - _ _ _ _ _3.474- 21. 13 25. 3 51. 8 61. 2 30. 3 58. 6 69. 8
0. 90 _ _ _ _ _ _ _ 3.476 20.74 23. 7 49. 2 61. 2 29. 5 53. 2 68. 7
0. 85_ _ _ _ _ _ _ 3.477 19. 85 22. 7 46. 9 61. 1 28. 7 48. 0 67. 7
0.80 _ _ _ _ _ _ - 3.479 18.49 22. 2 44. 5 61. 1 28. 0 43. 0 66. 6
0. 75_ _ _ _ _ _ _ 3. 481 16.98 22. 1 42. 1 61. 0 27. 2 38. 5 64. 1
0. 70 _ _ _ _ _ _ _ 3.484 15.48 22. 0 40. 0 60. 3 26. 5 34. 1 59. 8
0. 65_ _ _ _ _ _ _ 3.486 14. 06 21. 9 37. 9 54. 0 25. 7 30. 3 55. 4
0. 60 _ _ _ _ _ _ _ 3.488 12. 68 21. 9 35. 7 46. 7 24. 4 27. 2 51. 1
0. 55_ _ _ _ _ _ _ 3.491 11.41 21. 9 33. 7 40. 1 22. 0 25. 6 47. 1
0. 50 _ _ _ _ _ _ _ 3. 493 10.22 21. 9 31. 5 34. 1 19. 0 24. 8 43. 2
0.45 _ _ _ _ _ _ _ 3.496 9. 13 21. 8 27. 6 30. 7 16. 4 24 1 39. 6
0.40 _ _ _ _ _ _ _ 3.498 8. 11 21. 5 22. 0 28. 8 13. 8 23. 6 34. 6
0. 35 _ _ _ _ _ _ _ 3. 501 7. 18 18. 3 21. 9 27. 2 11. 5 23. 1 27. 7
0.30 _ _ _ _ _ _ _ 3. 503 6. 32 14. 52 21. 4 25. 8 9. 62 21. 0 23. 2
0.25 _ _ _ _ _ _ _ 3.506 5. 57 11.31 15. 3 -____ 7. 91 15. 8
0. 20 _ _ _ _ _ _ _ 3. 508 4. 85 8. 65 _____ 6. 42 11. 58
0.15_ _ _ _ _ _ _ 3.511 4. 28 6. 5 -____ 5. 20
______ --___ _ _ ._ _ _
0. 05_ _ _ _ _ _ _ _ _ _ _ _ _ _ _ ___-__ _ - _ ____ _._ _
0. 00 _ _ _ _ _ _3.- 5160 --___ _____. _____
bla x* x*
0 1,220 2,390
1, 120 2,233
8. 9 7. 0
1 1,743 2,970
I, 676 2,804
4. 0 5. 9
2 3,280 4, 660
3, 160 4,428
3. 8 5. 3
3 4,810 5,950 7,450
4,773 5,739 7,069
0. 8 3. 7 5. 4
4 8,870
8,685
2. 1
__
Frcc
VB
1 2 3
39.8 cps 104.5cps 249.5 cps 324 cps 367 cps 646 cps
MODES 1(0/0) 2( I/O) 3(0/1) 4(2/0) 5(1/1) 6(2/1)
__----
40.6 cps 95 cps 229.5 cps 254.5 cps 336 cps 5 \ 8 5 CPS
MODES 1(0/0) 2(1/0) 3(2/0) 4(0/1) 5(J/I) 6(3/0)
(b)
(C)
_e-----
FrGnRE 4.46.-Experimentally determined cyclic frequencies and nodal patterns for rectangular cantilever plates.
(a) a/b=l.O. (b) a/b=0.8. ( e ) ~ / b = 0 . 6 . (d) a/b=0.4.
86 VIBRATION O F PLATES
I 1
Mode I 1
1 I "xi. 1
2.00 2.50 3.33 5.00 ~ _ _ _ ~
Frequency,
- cps-_ 71. 9 175 552
wa24p/D- _ _ _ _ _ _ 3. 34 8. 23 56 25. 97
3. 50 3. 50 3. 50
3. 45
21. 7 21. 6 21. 1
21. 5
60. 5 60. 4 59. 8
59. 3
118.7 117. 5 116. 5
115. 2 TABLE4.58.-ExperimentaEly Determined Fre-
196.0 _ _ _ _ _ _ _ 195. 0
190.0 quency Parameters for a Square Cantilever
292. 0 _ _0_ _ _ . Plate
- - - _ _ _ _ _ _ 281.
14. 5
48. 1
92. 3
17. 3
54. 8
22. 5
69. 6
125.0
101. 5
32. 0
98. 0
169. 0
Mode
1
- I_ - 2_ 1 _ 3 _ 1 ~ 4 - ~
~
5
154. 0 _ _ _187.0
_ _ _248.
. 0
228. 0
319-324
for a mild steel plate having a/b= 1.86 are shown
92. 8
125. 1 in figure 4.48.
176. 0 Walton (ref. 4.88) used the method developed
244.0 by Houbolt (ref. 4.89) to determine the first
five frequencies for the cantilevered square.
246. 0
274. 0 This method is a numerical development of
319-324 the Rayleigh-Ritz method in which deriva-
tives are replaced by finite differences and
area integrals are replaced by double summa-
less, figure 4.47 showing the nodal lines is tions. In table 4.59 are given the first five
reproduced, partly as an estimate of accuracy cyclic frequencies for an aluminum-alloy plate
for further results on parallelogram and tri- as determined: (1) Experimentally, (2) by the
angular plates. method of reference 4.89, using 30 internal
Gustafson, Stokey, and Zorowski (ref. 4.87) grid points in the finite-difference mesh, and
experimentally determined the first five fre- ( 3 ) by using Warburton's formula (eq. (4.16)).
quencies of a square steel plate having dimen- No plate dimensions are given in reference 4.88.
sions BO inches by 10 inches by 0.0627 inch
and the foIlowing material properties: TABLE4.59.--Theoretical and E x p e r i m e n ~ ~ ~
Cyclk Frequencies for a Square Aluminurn-
odulus of elasticity in x-direction: 29.3 X
Alloy Cantilever Plate; v=Q.28
IO6 psi
Modulus of elasticity in y-direction: 31.5X
Frequency, cps
106psi -
Weight density: 0.282 16/i11.~ Mode
Poisson's ratio (assumed) : 0.29
Frequency parameters wa2JplD are listed in
table 4.58. The arithmetic
i 23 21
i 21
oduIi given above was u 48 50 56
I18 121 132
xural rigidity D. 162 163 171
Grinsted (ref. 4.79) obtained considerable ex- 173 177 190
perimental data. Frequencies and nodal patterns
RECTANGULAR PLATES 87
For a comparison of frequencies of a rectan- were used, where A?,(?) and Ym(ij)are the F-F
gular cantilever plate in air, water, or vacuum, beam functions expressed in terms of a nor-
see the chapter entitled “Other Considerations” malized (i.e., z=x/a, where a = l ) xy coordi-
(ch. 12). nate system having the origin at the plate
The problem was also discussed in references center (fig. 4.51) ; that is,
4.90 to 4.94.
cosh kmcos kmf+cos k, cosh k,E
4.3.1 3 SS-SS-F-F X,(E)=
dcosh2km+cosak,,,
The only specific result directly available for ( m even)
the problem of the SS-SS-F-F plate (fig. 4.49)
is the approximate formula, equation (4.16). sinh k, sin k,Z+sin k, sinh k,Z
X,(Z)=
For more information on this problem, see the Jsinh2 k,- sin2k,
discussion of the doubly antisymmetric modes ( modd)
of a completely free rectangular plate (sec. (4.58)
4.3.15). Straight node lines of antisymmetry
duplicate SS boundary conditions. The function Y,(ij>is obtained from equations
(4.58) by replacing 2 by ?/ and m by n. The
4.3.1 4 SS-F-F-F values k, are the roots of the equations
The only specific result directly available for
the problem of the SS-F-F-F plate (fig. 4.50)
is the approximate formula, equation (4.16).
tan k,+tanh k,=O
tan km-tanh k,=0
( meven)
( modd) ) (4.59)
4.3.1 5 F-F-F-F and are listed in table 4.60.
The problem of the completely free plate (fig. Results were obtained in mference 4.110 by
4.51) has a rich history. The first significant using six or more terms of equation (4.57)
work examining nodal patterns on rectangular and four M e r e n t values of Poisson’s ratio.
plates of any kind was produced by Chladni These data are given in table 4.61.
in 1787 (ref. 4.95) for completely free bound- Ritz (ref. 4.108) himself computed many
aries and extended in references 4.96 to 4.98. more frequencies for the square. Table 4.62
Other early experimental work on this problem lists frequency parameters, nodal patterns,
was performed by Strehlke (refs. 4.99 to and the approximate mode shapes used, again
4.103), Konig (ref. 4.1041, and Tanaka (ref. in terms of equations (4.57) and (4.5%). All the
4.105). Wheatstone (ref. 4.106) in 1833 made
nodal patterns in table 4.62 are either doubly
symmetric or doubly antisymmetric about the
an attempt to explain the Chladni patterns
bisectors of the square Z=O, ij=O.
in terms of the anodes of F-F beams, and these
Frequencies and mode shapes am computed
geometrid studies were extended by Rayleigh for v=0.225, using the number of terms listed
(ref. 4.107). (E, 5). Small variations in v from the
Ritz (ref. 4.108) in 1909 used the problem
to demonstrate his famous method for ex- TABLE
4.60.-Eigenvdues of a F-F Beam
tending the Rayleigh principle €or obtaining
upper bounds on vibration frequencies. This
innovation resulted in several following papers
(e.g., refs. 4.109 to 4.112) which used the
method to solve the problem in great
Lemke (ref. 4.1 10) computed frequencies
and mode shapes for the six modes of a square.
Functions of the type
88 VIBRATION OF PLATES
(0)
Wl
c-3
RECTANGULAR PLATES 89
-
? in terms of figure 4.51, where i=:/u, q = Y / b , and
I
i
+O8O1
c (4.61)
with
FIGURE 4.4?.-Exp
terns for the first
plate; material,
(a) ExperimentaJ
(6) Mode I ; fi=7 Mode 2 ; /,=I75 cps.
(d) Mode 3; f3=437 cps. (e) Mode 4; f4=552 eps. The boundary conditions are
308-3137 0-70----7
90 VIBRATION OF PLATES
om WI 012
64 405 1,120 2,233 3,736 3,573 7,750
2O
; .2/l 2/2 213 214 215
1,606 3,160 4,428 6,009 7,859
. . .-
310 3/i 312 313 3/4
4,235 4,773 3,739 7,069
fundamental 64 cps
Number of Nodal Lines$
4.49.-SS-SS-F-F
FIGURE plate. 4.51).-SS-F-F-F
FIGURE plate.
RECTANGULAR PLATES 91
The last of these is identically satisfied by
equation (4.60). Applying equations (4.63 (6))
and (4.63(d)) gives
b
1 cosh TA,,(
uan(F) =-(Zn-V d ? L Z >
xu, T
sinh - A,,
2
FIGURE
4.51.-F-F-F-F plate.
(4.64) c
@,31=n=1,3,. . .4- 1) * I%(U sin m-77
where a,, a:, b,, and b z are un -v,(q> sin ~ T S ] (4.68)
constants. where u, and v, are given in equation (4.64)
Applying equations (4.63(a)) and (4.63( and a,, A,, A,: and so forth are given in
results in an infinite characteristic determinant table 4.65.
92 VIBRATION O F PLATES
TABLE4.61.-6 Frequency Parameters and Mode Shapes for a Completely Free Square Plate
1
_____
13.10 13.086 12.94 12.927 12.64
0.360 0.390
11 4 11 11
TABLE
4 6 1 . 4 Frequency Parameters and Mode Shapes f o r a Completely Free Square Plate-Con.
0.390
15 6 15 15
I I
24.58 24.73 24.66 24.80
I I
0.360 0.390
I I -
I
No. terms _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _
II II I
I
l5
6 15 15
0.343
I 0.360 0.390
-- I
3 10 1 3 10 10
-I
67.804 66.820
- -
1.0000 1.0000
-. 01634 -. 02008
-. 00623 -. 008235
. 00709 00826
~
. 00470 . 005495
. 000315 . 000322
-. 00367 -. 00419
-. 002934 -. 00333
-. 000236 -. 000241
. 0000421 . 000043
RECTANGULAR PLATES 97
TABLE
4.61 .-6 Frequency Parameters and Mode Shapes for a Completely Free Square Plate-Con.
15 1 6 15 15
77.730 77.380
I 77.683 77.309 77.162
-
TABLE
4.62.-Doubly Symmebic and Doubly Antisymmetric Frequencies and Mode Shapes for a
Completely Free Square Plate; v=0.226
316. 1-270 8v
378- 57 6 v
1554 + +
o . o o 9 x l Y ~ -0.075(x1Y34-X J Y J X3Y3- 0 . 0 5 7 ( x I Y 6 XSYJ
+0.121(X3Y&+X~Y3) - 0.007X5Y6
2713
2945
5570
6303
13 674
RECTANGULAR PLATES 99
TABLE4.62.-Doubly Symmetric and Doubly' Antisymmetric Frequencies and Mode Shapes for a
Completely Free Square Plate; ~=0.225--Continued
26.40
35. 73$20. 8 6v
266.0-274 6v
886 xo Y4 -x4 Yo
1702
2020 XzY4*X*Y2
5480 4Y ,
x
100 VIBRATION O F PLATES
TABLE4.62.-Doubly Symmetric and Doubly Antisymmetric Frequencies and Mode Shapes for a
Completdy Free Square Plate; ~=0.225-Continued
5500
5640 XoYa+XaYo
7840
13 840
15 120
28 740
RECTANGULAR PLATES 101
TABLE4.63.-~Yymmetrie-AntisymmetricFrequencies and Mode Shapes f o r a Completely Free Square
Plate; v = 0.225
w2a4p
~
W(Z,y, - W(Y, 2,
16D nodal pattern
Mode shape Nodal pattern
O.O6?8X1Yz+ X3Yo-0.015OX3Yz
SO.O355X,Y4+ O.OOOOX~Y~+O.OlOOX~I'~
+
-0.0007XsYz 0.0016X5Y4
- 0 . 0 7 0 9 X ~ Y ~ + O . O 2 1 4 X ~ YXaY,
~+
-0 . 1 2 6 0 X J - 0.0038X5Y,+ 0. i234X3Y4
-0.0095 XsY2-0.0100 X5Yd
I02 VIBRATION O F PLATES
-
da4p
16D
Mode shape Nodal pattern
3 _____415 6
-
For modes symmetric with respect to 5=8 The first four of these frequencies and the
(fig. 4.51) and antisymmetric aboutT=O (asym- amplitude parameters are listed in table 4.66
metric with respect to the diagonals) : (ref. 4.113).
The four nodal patterns corresponding to
table 4.66 are shown in gure 4.52; also shown
are interesting patterns which arise by taking
the linear combinations :
W(S,5)-KW(7,Z)
RECTANGULAR PLATES 103
c
0
I1
k
Y
FIQIJRE
4.52.-Superposition of mode shapes for a completely free square plate; v=0.3. (After ref. 4.113)
104 VIBRATION OF PLATES
TABLE4.65.-Frequency Parameters and Mode Shapes for a Completely Free square plate; v = o . 3
__
Nodal pattern I w a z m 1 1
n an
I xn
-. 11966 _ _ _ _ _ _ _ _ __ __ -_-_ _ _ _ _ _ - _ - _ -
1. 0000 3. 23309 81. 56615;
63. 6870 . 03422 4.73844 3. 08985
. 01065 6.51558 5.43573
. 00473 8. 39362 7. 58598
-8. 81714 - - - -- -- -- - -. _ _ _ _ _ _ _ _ _ - - - _ -
1.00000 4. 05046 2. 89935;
122.4449 -1. 19356 5. 32975 1.89572
-. 08213 6.95746 4.85734
-. 02402 8.74107 7. 18288
-. 07482 _ _ _ _ _ _ - - - - - - -_- - - - - - - - - - -
1.00000 4.59037 3.61545;
168.4888 .44885 5. 75078 1. 03513.1:
.03590 7.28502 4.35069
. 01347 9. 00397 6. 85044
- 8. 90424 - _ _ - _ _ _ - - - -.---- - - - - - - - - - - -
1.00000 5. 86426 5. 13707;
299. 9325 -. 59521 6.81099 3.79335;
- 1. 39192 8. 14998 2.36864
-. 13703 9. 71543 5.79745
Detailed mode shapes showing contour lines Taking a coordinate system as in figure 4.51,
for 16 of the modes described in the foregoing the modes are antisymmetric with respect to
paragraphs are shown in figure 4.53 (ref. 4.113). both Z and 5 and are unaltered by interchange
Grauers (ref. 4.114) in an early work also of Z and i j (symmetric about the diagonals).
attempted to solve the problem using solutions Five nodal patterns of this type are shown
to the differential equation but obtained in- in the third part of table 4.65. They also
accurate results. obtained extremely accurate upper bounds by
Upper and lower bounds for the fundamental the Rayleigh-Ritz method, using the first 50
frequency were obtained in references 4.115 admissible products of free-free beam functions.
and 4.116 and were improved to extreme Double precision arithmetic was used in the
accuracy in reference 4.117. For v=0.225, computations where necessary. Results are
these bounds are listed in table 4.67 for v=0.225 and v = Q . 3 .
Herein results from the Rayleigh-Ritz pro-
14.1Q28<wa2&@< 14.1165
cedure are given; both 25 and 50 admissi-
Bazley, Fox, and Stadter (ref. 4.118) used ble functions are used to show the rate of
a method developed in reference 4.59 to com- convergence.
pute lower bounds for the first 18 frequencies Siguito (ref. 4.76) showed that more precise
of the following symmetry class of B square: upper bounds can be obtained with the Ray-
RECTANGULAR PLATES 105
TABLE
4.65.-Fraquency Parametersand Mode Shapes fora Completely Free Square Plate; v=0.3--Con.
-
Nodal pattern 1 w a z m 1 i an I
Modes symmetric about coordinate axes, antisymmetric about diagonals
0 - 19.46060 .- - - - - - - - - - - - _--____----___
2 1.00000 2.44653 1.41933
19.5961 4 .00264 4.24093 3.74359
6 - .00487 6.16324 5.83219
8 - .00290 8.12315 7.87493
0 3.93698 - -_- - - - - - _ - - - -
_ _ _ _ - _ _ _ - -- _
2 1.00000 3.25932 * 1.619263'
65.3680 4 -. 09935 4.75638 3.06216
6 -. 01507 6.52864 5.42004
8 -. 00451 8.40376 7.57475
3.84826 ___-_____-_-
2 1. 00000 3.98317 2.804582'
117.1093 4 -. 48091 5.27879 2.03331
6 - .02845 6.91850 4.91267
S - .00453 8.71009 7.22041
leigh-Ritz procedure by using Legendre fune- to plus or minus signs in equation (4.70).
tions rather than beam functions. Results Vahes given above the main diagonal of the
from this approach are also listed in table 4.67. array are for the minus sign, and values below
Waller (ref. 4.119) obtained experimental the diagonal are for the plus sign. Numbers OR
frequencies and mode shapes for square brass the diagonal of the table are then for m=n.
plates (v=%). Consider the mode shapes as I n reference 4.79 are plotted the experimental
being approximated by free membrane mode frequency ratios of reference 4.119. This plot
shapes ; for example, is reproduced as figure 4.54. Experimentally
observed mode shapes corresponding to many
of these frequencies are shown in figure 4.55
(ref. 4.119). Other experimental results for
(4.70) the square are given in references 4.110,
4.113, 4.120, and 4.121.
in Germs of figure 4.51. Theratio of fre- Waller (ref. 4.122) observed the transition
quencies relative to the fundamental are given points in sudden nodal. pattern change in the
in table 4.68 for various m/n ratios. The pllus fundamental mode as alb varies for the com-
or minus signs after m/n in the table correspond pletely free plate. This had been observed
308437 0-70-8
106 VIBRATION OF PLATES
TABLE
4.65.-Frequency Parameters and Mode Shapes .for a Completely Free Square Plate; v=0.3--
Concluded
294. 9247 I 8
3
5
7
1.00000
1275. 527
-346. 402
-20. 133
5.55717
6. 23555
7.40825
8.88156
5. 37421.1'
4. 569703
2. 20955i
4. 37240
-
RECTANGULAR PLATES 107
wa2@= 299.9320 ( K = O )
FIGURE
4.53.--Contour lines for 16 modes of a completely free square plate; v=0.3. (After ref. 4.113)
W a = m i n 1 A: m Bm
*
Am
---
1
_ _ _ _ ~ ~ - - - _ _ _ _ - -
' 7
I ,01182
.00430
,00183
3.53922
5.34098
7.24749 '
2.33964
4.63399 ,
6.74343 1
'1 2
4
6
.29218
.01218
2.74337 0.68841
4.41884 3.53184
. 00321 6.28698 5.69859
1 8 .00133 8.21743 7.77650
61.0932 _ _ _ _ _ _ _ _ _ _ _ _ _ _
I
1
1
3
5
1.00000
-.86028
-. 04348
2.68145
3.89746
5.58482
2.27819 1
1.67626
4.33703 ,
0
2
4
-7. 62932 _ _ _ _ - _ - - -- -_ _ _ - _ - - - -
-. 56588 3.19221 1.47992i
-. 01392 4.71065 3.13207
7 -. 00974 7.42901 6. 54292 6 -. 01076 6.49540 5.45984
8 -. 00595 8.37796 7.60328
105.4634-- - - - ----- 1 1.00000 3.41843 3.1121% 0 .14106 . . . . . . . . . . . . . . . . . . . .
3 -1.43311 4.43685 1.298342 2 -3. 03882 3.83219 2.58567i
5 -. 07788 5.97375 3.78343 4 -. 01973 5.16582 2.30528
7 -. 02391 7.72565 6.18986 6 -.02166 6.83269 5.03133
8 -.01136 8.64209 7.30167
131.4697_ _ _ _ _ _ _ _ _ _ _ _ _ 1 1.00000 3.78427 3.510083' 0 -.07630 _ _ _ _ _ _ _ _ _ _ - _ _ _ _ _ - _ _ _
3 .09333 3.72448 2.078633' 2 -.07815 4.16181 3.052983'
5 .00250 6.19037 3.41750 4 . 17972 5.41486 1.63687i
7 .00245 7.89434 5.97322 .01323 7.02287 4.76227
.00382 8.79322 7.11894
theoretically for other boundary conditions Xm@)and Yn@)are shown in figure 4.57 for
(see secs. 4.3.1 and 4.3.12). In figures 4.56(a) a/b= 1.O. The curves of figure 4.57 do change
and 4.56(b) are shown the nodal patterns of slightly between the different modes and with
two brass plates having the same width, but varying a/b ratio. Thirty-six precise sets of
the length in figure 4.56(a) is slightly greater. curves for Wmn(Zl Q) are plotted in reference
The a/b ratio is approximately 1.93. The cyclic 4.13, but is is not felt that the variations are
frequencies in figures 4.56(a) and 4.56(b) were sufEcient to justify their detailed repetition here.
548.8 and 558 cps, respectively. It was found An estimate of this variation can be obtained by
that by gradually filing down the longer side looking a t the edges where the variation is
the nodal patterns in figures 4.56(c), 4.56(d), usually the greatest. One of the mode compo-
and 4.56(e) could be produced. It is esti- nents having relatively large change in shape
mated that the transition between figures due to change in the other component or
4.56(b) and 4 . 5 6 u ) occurs at a/b=3.9. a/b is Xz(I).Deflection values t o be used a t
Pavlik (refs. 4.111 and 4.112) extended Ritz' x/a=O.5 in figure 4.57 for varying values of
work to nonsquare rectangular plates. Fre- Y,@) are given in table 4.73 for a/b=l.O.
quencies and mode shapes for three aspect Increasing n also increases the magnitude of
ratios are presented in tables 4.69 to 4.71 for the negative curvature in the range
v=0.25. The functions X, and Y, are as <
0.3 x/a< 0.5.
defined previously in equation (4.58). Variation in edge deflection of X2@) with
I n reference 4.13, extensive results are a/b ratio is shown in table 4.74 for Yt($.
obtained for a/b=W and 35 and Y=%. These Accurate upper and lower bounds for the
are listed in table 4.72. Values in parentheses doubly antisymmetric modes of a rectangle
me interpolated. (see discussion earlier in this section) are re-
Mode shapes in the form Wmn(Z, y)=X,(Z) ported in reference 4.118. These results are
Y,@) corresponding to w,, were found in ref- given in table 4.75 for v=O.3. Upper bounds
erence 4.13. The shape of the components from reference 4.78 for doubly antisymmetric
RECTANGULAR PLATES 109
TABLE 4.67.-Bounds on Frequency Parameters
w a 2 J p P f o r Modes of a Completely Free
Square Plate Which Are Antisymmetric About
the Coordinate Axes and Symmetric About the
Diagonals
Mode
II I
I Upper bounds
Lower
bounds
25 terms erms
(ref. 4.118) ( 4.76)
I v=0.225
i
1_ _ _ _ _ _ _ 13. 851
1
14. 119
I I
14.111
2_ _ _ _ _ _ _ 76. 245 77. 621 . 154
3_ _ _ _ _ _ _ 151. .26
4 _ _ _ _ _ -210.
- .29
* 94
I
v= 0.300
_ _ - __
__ 1..52
. __ 5. 1( 9.14 15.8 23.0 32.5 43 55.2 70 84 101 119 141
_____. 1 2. 71 5. 3( 10.3 15.8 23.9 32. 2 43 55.8 71 86.1 102 121 ___
1.94 2.71 4. 81 8.52 12.4 19.0 26.4 34 46.6 59 73 89 105 124 ___
5.10 6.0( 8. 52 11. 8 16.6 22.6 30.0 39.5 50.5 63.4 77.5 92.4 110 128 ___
9. 9 10. 3 13. 2 16. 6 21.5 28.7 35.5 45.4 55.9 69.7 82.9 99 116 132 ___
15. 8 16.6 19. 0 23. 3 28.7 35 43 52.1 64.5 75.9 90 106 122 136 ___
23. 8 23. 9 27. 1 30. 0 35.9 43 51 61.7 73 84 99 115 130 ____ ___
32. 5 32.4 34.0 39. 8 45.4 53 61.7 70.3 84 93 108 124 -___ ____ ___
43. 0 43.0 46.6 50. 5 57.2 64.5 73 84 94.4 106 120 136 ____ ____ _._
55. 2 55.8 59 63.4 69.7 76.2 84 93.2 106 120 133 ______ ____ ____ ___
70.0 71. 0 73 77. 5
84. 0 86. 1 89 92. 4
01 02 05 10
19 21 24 28
41 _-__. -__--
TABLE 4.69.-Frequencies and Mode Shapes for a Completely Free Rectangular Plate; a/b=P .O@;
I -I O I
1
1 l
I
‘ 2
5.593
1
I
3
15. 418
4
30. 223
5
49.965
6
74. 639
(17.61) 27. 032 (42. 25) 61. 628 (85. 56)
37. 585 (51. 70) 70. 007 (91. 78) (117.29)
(65. 17) (75.05) 91. 963 (111. 58) 135. 794 (162. 56)
(149. 57) 170. 974 (196. 56) (223. 50)
(226. 41) 248.876 274. 639 (303. 18)
(324. 72) (345.96) (372. 88) 402. 968
I
N 0 9.944 27. 410 53.735 88. 826 132. 691
9. 905 22. 245 40. 339 66.309 100. 928 (144. 5)
22.373 (30. 36) 46. 654 (68. 39) 97.822 (133. 40) 177.606
61. 673 (69. 56) 86. 028 111. 510 143. 532 182. 204 (226. 20)
4 120.903
199.860
298. 556
i
1
(127.7)
(205.
(302. 1)
(145. 2)
222. 088
(320. 4)
(160. 5)
(250. 0)
(347. 8)
204. 804
283. 715
(382. 6)
(245.9)
326.580
(425.6)
294. 258
(374. 8)
476.853
I
Edge deflection_ _ _ _ _ _ _ _ _ 1. 81 1. 72 1. 67
I I I l l 1
114 VIBRATION OF PLATES
Mode Lower bound Upper bouni Lower bound Upper bouni Lower bound Upper bound
1 bla= 2.00
-
Mode
~
1
o a 2 W D- - - - _ _ _ - -3. 2597
213
~
34. 622
5
_
53. 092
6
64. 080
_
7
71. 048
8
_
77. 232
9
84. 532
-_
-
10
102. 87
_ ~
4.4 ELASTIC, DISCONTINUOUS, AND POlNT tions along z = ~and z=a. The remaining
SUPPORTS boundary conditions are
.4.1 Elastic Edge Supporis
(z, O)=--K,d”(z, 8)
Consider first the rectangular plate simply k?l
supported (SS) along the sides z=O and x=u
and elastically restrained (ES) against both
translation and rotation along the other sides
bW
Mv(z,h)=K*- (z,6)
* (4.71)
I alb = 1.09
I
I 1 \
I I 1 1 1
I
I
a/b=2.0
.Fundamental frequency of ft 3.94- by 3.62- by 0.720-in. plate was 423 cps. For a 6.15- by 5.67- by 0.0906-in.
plate, it was 220 cps.
bFundamental frequency of a 9.81- by 6.38- by 0.934-in. plate was 134 cps.
.Fundamental frequency of a 2.36- by 1.172- by 0.0807-in. plate was 1730 cps. For a 5.55- by 2.78- by 0.1240-in.
plate, it was 482 cps.
FIGURE
4.59.-SS-ES-SS-ES plate.
RECTANGULAR PLATES 117
where K,,Z,,, are the stiffness coefficients of Das (ref. 4.10) showed that the characteristic
distributed translational and rotational springs equation for the caSe K,=Kz=K, K3sK4=m
acting along the edges y=Q and y=b. For becomes
simplicity it will be assumed that these co- CosXIb CoshXzb-I - Xt-X:
-
efficients do not vary with x. The constants -
sin X1b sinh Xzb 2X1Xz
K1 and K2 have mensions of moment/(unit
length) and K X d K4 have dimensions of
force/(unit length) 2. Substituting equation
(1.37) into equations (4.71) results in a char- (4.72)
acteristic determinant, the zeros of which with Xi and X2 as defined in ons (4.27)
the vibration frequencies. and that the mode shapes are by
FIGURE
$.6O.--FFe~penCy parameters w2aWp/rn2&D for a SS-ES-SS-ES rectangular plate with symmetrical slope
restraints.
In reference 4.12’7 the problem is also solved which gives A= (?rKb/$D)B. Formulating the
by the Rayleigh method. A mode shape is Rayleigh quotient yields the frequency pararn-
chosen as eter
h
d
v
h
U
v
-e
h
a
v
cI h
ro
v
.m
rD
E
c
0
-
rn E1
v
a
s
Ld
c h
II s
9k
.
P
B
a h
.s" ~
h
a
v
h
U
v
h
0
v
h
U
v
RECTANGULAR PLATES 121
h
e
__
h
v
U
h
c,
w
I I I
I
I
I I
3084337 0 - 7 0 4
VIBRATION OF PLATES
nowhere differs from the series solution by more square by using the same procedure as in refer-
than 0.7 percent. I t must be noted from equa- ence 4.128. These are shown i n figure 4.62.
tions (4.81) and (4.85) that choosing equal I n reference 4.130, the problem is also solved
values of and t b does not give equal slope by using the Rayleigh-Ritz method and alge-
restraint along all edges except for the case braic polynomials.
of the square. I n reference 4.131, the typical electronic
The case of uniform slope. restraint and chassis which is formed by bending the edges
of a plate down is treated as a plate with elastic
W=C along all edges was stu
edge supports. An eigenfunction is used to
et a]. (ref. 4.60), who used a solve the problem which is an average of the
t o obtain frequencies for the eigenfunctions for plates with simply supported
s of a square having variable edges and those having clamped edges. The
restraint. These results are shown in figure Rayleigh-Ritz method is employed. Theo-
4.61. Results for this problem were also pre- retical and experimental results are obtained
sented ip1 reference 4.129 for the case of khe for particular chassis.
RECTANGULAR PLATES 123
Ends Elastically Restrained Against Rotation
m, n = 4 rn, n=6
€4 Ad= -3 4 $4 65 $5
Hoppmann and Greenspon (ref. 4.132) pre- 4.4.2 Discontinuous Edge Conditions
sented a method for experimentally simulating Some interesting results are available for the
elastic edge supports by means of sharp V- case of a square plate which is simply supported
grooves machined along the edges of a clamped but clamped along segments of its edges.
plate, the degree of slope restraint being deter- Consider f i s t the square which is clamped
mined by the depth of the grooves. A curve along four symmetrically located segments of
showing the frequency parameter for a clamped length 11, and simply supported along the re-
square plate as a function of the notch ratio R mainder of the boundary as in figure 4.64. Ota
is shown in figure 4.63; R is the ratio of the and Hamada (refs. 4.133 and 4.134) solved tho
depth of the notch to the thickness of the plate. problem by assuming a deflection function
Experimentally determined points are shown which satisfies the simply supported boundary
conditions everywhere (ea. (4.19)), and applying
as circles. The curve was drawn through end-
distributed edge moments of the type, for
points determined by the theoretical results of
example,
Iguchi (ref. 4.9) and fitted t o the four experf
mental points.
124 VIBRATION OF PLATES
~ ~~
-
0.8 0 16. 19 35. 14 45.79 64. 74 66.72 96. 33
20 25.80 46. 02 59. 98 79. 06 79.24 111.2
(25. $6) (46.17) (60. 16) (79.32) (79.50) (111. 5)
03 29. 08 52. 52 68.52 89. 40 89.29 124. 5
(29. 18) (52. 76) (68. 80) (89.86) (89. 69) (125. 0)
-
0.6 0 13. 42 24.08 41.85 43.03 53.69
20 22.30 32.58 50.48 56.97 66.96
(22.34) (32. 68) (50. 63) (57.11) (67. 17)
03 25.90 37.28 56.93 65. 18 75.94
(57.20) (65.39) (76. 31)
30
51?
10
€1’112 ktl
TABLE
4.82.-Experimental Cyclk Frequencies
I
and Nodal Patterns for a Simply Supported
0 Experimental Data (h: 0.063")
Square Plate Clamped Along 4 ~ ~ ~ e t r ~ ~ l ~
Located Segments
0
0 $ ) ; i I
!?,/a
1
1
lengths of 6.50 inches and thicknesses of 0.063
Frequency, cps-.. - - 280 535 725
and 0.091 inch.
Experimental frequencies and nodal patterns
for the first three modes for an aluminum
plate 11.8 inches long, 0.012 inch thick, and
having &/a=% were obtained in reference 4.135 Y
w a z m for
~ values of tzla of-
Source
- ~
Ref. 4.133_ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _
Ref. 4.135 - - - _ _ - _ - - - _ - - - - - - _ _
_
19.74
19. 74
_ .
TABLE
4.84.-Experimental Cyclic Frequencies and Nodal Pafiernsf o r a Simply Supported Square
Plate Clamped Along 2 Symmetrically Located Segments of Opposite Edges
Nodal pattern
I
Frequency, c p s - - - _ _ - - _ - - - _ . 225 420 1 500 660 785 955
128 VIBRATION OF PLATES
s3
c
-.
I5 2
0 I 1 I
2
3 3
4 I 1
0 3 5 -2
P a/,.
lala 0
_ _ _ - ~ - -
34 34 x 1
FIGURE4.69.-Frequency parameters for SS-SS-SS-SS wa2Jpm_ _ _ _ _ ~ _ _19.74 23. 0 23.4 23. 6 23. 65
square plate clamped along two unsymmetrically I
located segments of opposite edges. (After ref. 4.133)
tained on the same plates as those described
earlier in this section.
The case when the plate is clamped along
one segment a t the end of one edge is shown in
figure 4.72. Nowacki (refs. 4.136 and 4.137)
expressed a unit moment acting a t a point
along the clamped interval in terms of a trigo-
\
c
-3 '
I
I
I3 '$perimytal Dot; ( h = Q.063")
FIGURE4.7O.-SS-SS-SS-SS square plate clamped e o (h=0.091"!
along one symmetrically located segment of an edge. C
0 a1 i. 1
s i 3 3 I
Q,/o
terns obtained in reference 4.135 are given in FIGURE 4.7P.--Frequency parameters for SS-SS-SS-SS
table 4.87 for &/a=?$. Experimental results square plate clamped along one symmetrically
shown in figure 4.71 and table 4.87 were ob- located segment of an edge. (After ref. 4.133)
RECTANGULAR PLATE8 129
Y over discrete segments of the interval, which
resulted in a system of equations, each term
of which is an infinite series of transcendental
functions containing the eigenvalues. Trun-
cating the series and solving the resulting
characteristic determinant yielded the vibration
frequencies.
Numerical resdts from reference 4.133 are
given in figure 4.73. Data from references
4.133 and 4.136 are also given in table 4.88.
By looking at the results of reference 4.136
in table 4.88, it is seen that they are clearly
inaccurate, the frequency parameter listed for
the case when li/a=% being greater than the
well-known result for the case when 15/a=1
(see discussion on SS-C-SS-SS plate, sec. 4.2.2).
The solution is also given in reference 4.136
for the case when the interval O<x<l, is
clamped along the edge y=O (fig. 4.72), the
interval Z5<x<a is free, and the remaining
FIGURE4.72.-SS-SS-SS-SS square plate clamped edges are supported' It was found
along one segment at the end of an edge. for Isla= % that w a 2 m D = 14.8.
The case obtained when the simply sup-
nometric series and formulated an integral Ported Portions Of the edges of the Plate shown
equation involving a Green's function along in figure 4-72 are replaced by clamped edge
the clamped interval. The integral equation conditions and the remaining portion has zero
was replaced by a finite summation carried out slope and shear is included in reference 4.138.
TABLE4.87.-Experimental Cyclic Frequencies and Nodal Patterns for a Simply Supported Square
Plate Clamped Along 1 Symmetrically Located Segment of an Edge
Nodal pattern
The necessary integral equations are completely where X=wa2.drD and the subscripts 4 and 6
formulated but no numerical results are identify the two meshes used.
obtained. The mode shapes W(x,y) corresponding to
the fundamental frequencies are given in
table 4.90, where the grid locations me those
roughout this section the term "point shown in figure 4.76.
support" will be used to denote a constraint Higher frequencies for the square supported
of zero deflection at a point. Unless otherwise a t the corners were also given in reference
stated, there d l be no constraint on the 4.139. These are listed in table 4.91 for two
slopes at such points. mesh widths. Extrapolated values using equa-
Consider first the problem of the rectangular tion (4.90) are also given.
plate free along all edges and supported a t Mode shapes corresponding to these frequen-
the four corner points (fig. 4.74). Cox and cies are shown in figure 4.77, and the amplitudes
Boxer (ref. 4.139) solved the problem by of W ( x ,y) a t the grid locations shown in figure
means of finite difference equations. Punda- 4.75 are listed in table 4.92 for v=Q.3. TWO
mental frequencies for a/b=1, 1.5, 2 , and 3 independent mode shapes correspmding to the
for ~ = 8 . 3are listed in table 4.89 and plotted second frequency were found. They are iden-
in 6guse 4.75. The mesh widths Aa and Ab tised as 2a and 2b. As ean be seen from figure
l I I 1
a Extrapolated value from eq. (4.90).
RECTANGULAR PLATES 131
variation of the frequency itself with Poisson’s
ratio to be seen. This is shown in figure 4.80.
Nishimura (ref. 4.14) used the finite-differ-
ence method and a relatively coarse grid (char-
acteristic determinants of order no larger than
six) to obtain the first 10 frequencies and nodal
patterns of a free square plate point supported
at the four corners. He also obtained experi-
mental results on a steel plate 10.1 by 10.1
inches by 0.087 inch. These results are shown
in figure 4.81, with experimental values given
in parentheses. It is noted that the third,
sixth, and ninth mode shapes and frequencies
otb also exist for the completely free square plate.
FIGURE4.75.-Frequency parameters w a
2m for a
Reed (ref. 4.140) obtained extensive analyti-
free rectangular plate point supported at the four cal and experimental results for the rectangular
corners; v=0.3. (After ref. 4.139) plate supported at its four corners. Analyti-
cal results were achieved by two methods-
4.77, the third mode shape and frequency are the Rayleigh-Ritz and series methods. The
identical to those of the fundamental mode of a deflection function
completely free square plate (sec. 4.3.15).
Variation in the frequency parameter wa2JplD
with Poisson’s ratio is shown in figure 4.79.
However, it must be remembered that D de-
pends upon v. Substituting equation (1.2)
for D into the frequency parameters permits the
i t l
2-7- -9l -8l -7l -2r 20-
8
l l l r r l i r l l r i l i l l l r l
21 -22-23 -24- 25-26- 27- 28-29 - 28- 27-26-25-24 -23- 22- 21-20
i
3-6
l -9-10I S -9 I -6-I 3! 30-31
I r r l l l r l r I l i i l I t I l l
-32-33-34-35 -36 -37-38 -39 -38 - 37-36-35-34 -33-32-31 -30
i l r r l I I l l l l I l i l l i l 1 1 1 ! 1 1 1 \
2-7- 9- 9 -8 -7- 2 20-21-22-23-24-25-26-27-28-29-28-2?-26-25~24-23~22~2~~20
I r l I I i i I 1 I l l l l i l l D s i r i o I 1 1
1-4-5-6-5-4- 1 10-1 l~~2~l3-i4-i5~~6-l?-l8-~9-~8-~~-i6-l5-l4-~3-12- I 1-10
-2-3-2-1
I I I
t-2-3-4-5-
I 1 r i I 1 I i
-7-8-9-8---?-6-5-4~3-2-I
I s s l I
a/b=i a/b=3
- 2- 3 -4 ~ 4 -3 -2- I
(a/b=1.5 a/b=2
FIGURE
4.76.-Finite diflerence meshes. (After ref. 4.139)
132 VIBRATION OF P L A T E S
Grid
location
’
Supported at the 4 Corners; u = O . 3
~
W ( z ,y) for values of alb of-
as half of the solution, the other half being a
similar series obtained by interchanging x and
y. Frequency parameters obtained in keeping
24 terms of the series are listed in parentheses
,1 1 . 0 1. 5 2. 0 3.0 in table4.93. I n table4.94 the theoretical cyclic
.95257
96707
~
-
w a z m for mode-
Mesh width
_______ 2 --___-____3 1 4 1 5
A a = a b = a / 5 _ _ ~ _ ~ _ _ _ . - - - - - - - - - - - - - - - . - - - - - 15.0541 i
16. 8311 35. 59.51 38. 7292
Aa=Ab=a/6___.---.-.--.------------------ ~
1 a
15. 2650
15. 73 a
17. 5659
19.13
l
I
I
40.2638
843.55
46
44
I I I I I
~ o i s m ' sRatio, v
TABLE4.92.--Higher Mode Shapes W(x, y) f o r a Free Square Plate Point Supported at the 4 Corners;
v=o.s
W ( z ,y) for mode-
Grid location
2a 2b 3 4 5
~
FIGURE4.81.-Theoretical and experimental frequency parameters and nodal patterns for a free steel square plate
point supported at the four corners. Experimental values are given in parentheses. (After ref. 4.14)
Mode 1:
-__
alb
Normalized mode-shape Normalized mode-shape
coefficients coefficients
~~
a02= -.I753
ao,= .0002
1.5 9.21 22.78 ~ 2 .1530
=
(8.92) (21.53) a24= - .0059
a42= .0012
a06= .0009
TABLE
4.93.-Frequency Parameters, Nodal Patterns, and Amplitude Coefiients for a Rectangular
Plate Supported aj Its 4 Corners; v=03--Continued
[Values i n parentheses are obtained by keeping 24 terms of the series]
Mode 3: Mode 4:
i- all=O. 7924
a13 = - .0877
boz= 1. 000
b04=-. 0808 56. 2
ao1= 1. 000
a03= -. 0726
bo1 = - 0.
bo3 =. 0903
742
2. 0 (32. 83) a3,=. 1496 bzz=. 1713 (52. 0) aZ1=. 2048 bzl=. 2277
a33=. 0426 b2a =. 0334 a23=. 0322 b23=. 0341
asl= - .0304 b42= -. 0272 u41=. 0011 bdl= -. 0277
were used in reference 4.2 to obtain approximate lem of a free square plate clamped a t one mid-
fundamental frequencies for general values of point as shown in figure 4.83. A deflection
a/b and v=0.25. The frequency may be com- function
puted from equation (4.17) with
(4.95)
(4.94)
was used to yield a ~ u n d a ~ e n frequency
ta~
parameter wa2mD=2.580. I n this case the
Cox (ref. 4.142) also used the finite-diff erence point clamp a t ( 0 , 0) permits rotation about
method to solve the problem of the free square the Z-axis, but not about the @axis.
plate supported at the midpoints of its sides The square plate having two adjacent edges
(see fig. 4.82). Frequencies obtained from two both either damped or simply supported and
mesh widths and from the extrapofation formula a point support at the opposite corner (see
equation (4.90) are listed in table 4.95 for fig. 4.84) was also malyzed by Cox (ref. 4.144).
v=o.3. The finite difference method and ~ = 0 . 3 was
Plass (ref. 4.143) used a variational method used. Frequency parameters for both prob-
described later in this section to solve the prob- lems are listed in table 4.96 for two mesh
RECTANGULAR PLATES 137
TABLE 4.93.-prequency Parameters, Nodal Patterns, and Amplitude Coe$icient.s for a Rectangular
Plate Supported at Its 4 Corners; v=O.S-Continued
I
Mode 6:
O I (38. 44)
41.
a32=. 2107
~ 3 4 0645
~ .
a52= - .0433
ais= - .0053
b32=. 2107
634 =.0645
bsz = - .0433
hie= -. 0053
51. 6
(50.3)
a3,= -. 2789
a33= -. 0007
a51=.0318
bzz=. 0337
bzh= -. 0492
b,z= -. 0000
alz= -.
0666 biz= 1. 000 all= 1. 000 boz= -0. 8072
all= -.
0342 bl4= -. 0797 aI3=-. 0446 bm=. 0477
56. 0 a32=. 1809 b32= -. 0153 57. 7 as1=-. 1487 b z z = . 1222
1. 5
(52. 7) a34=. 0119 b34=. 0326 (57. 7) a33=.0225 bz4z -. 0232
a52= -. 0049 b5z= -. 0190 a51 =.0084 b 4 2 -. ~ 0078
ala=. 0014 bla= -. 0101
a12= -. 1869 biz= 1. 000 a l l = 1. 000 boz= -0. 6479
a ] ( = - - . 0122 bl4= -. 0671 ais= -. 0597 bo4 =. 0324
67. 1 a33=. 1764 ba2= -. 0548 73. 0 asl= -. 0626 bZ2=. 1557
2. 0
(63.8) a34=. 0011 b31=. 0238 (71.3) ~ 3 ~ =0316. bzn= -. 0051
am=. 0068 b52=-. 0114 aS1=-. 0013 biz= -. 0151
ala=. 0015 bts= -. 0134
QZ= -. 2238 biz= 1. 000 all= 1. 000 boz= -0. 5687
all= -. 0021 b14= -. 0630 a13= - .0705 bM=. 0720
2. 5 78. 5 a32=. 1772 b32= -. 0717 97. 5 a 3 1 = -. 0104 b 2 2 = . 1772
a34= -. 0038 bra=. 0195 a 3 3 = . 0364 bzr=. 0027
aS2=.0134 b52= -, 0074 a s l = . 0059 biz= -. 0195
ale=. 0013 hie= -. 0160
Mode 7:
I1 Plate 1: 12- by
12- by 0.129-in. 2024 aluminum Plate 2: 10- by 20- by 0.173-in. 2024 aluminum
Mode ~ E= 10.6X lo5 psi (book value) E= 10.6X IO3 psi (book value)
Clamped Point
/- -X
a/2
Mesh width
-- Extrapolated
value
I
TABLE4.96.-Frequency Parameters u a 2 j m
for a Square Plate Simply Supported or
Clamped on 2 Adjacent Edges and Supported
at the Opposite Corner; v=O.3 Simply Supported /-
or Clomped
FIGURE4.84.-Square plate simply supported or
Adjacent edge clamped OR two adjacent edges and supported at the
CORditiORS Mesh width Extrap-
- olated opposite corner.
Simply supported---
Clamped _ _ _ _ _ _ _ _ _ _ _
for each section, and satisfying the boundary where a=mr/a. The roots h3 and X4 of
conditions along y=O and y=b and the con- equation (4.98) yield the frequencies.
tinuity conditions along y= 9 exactly. Con- The fundamental frequency parameters for
tinuity of transverse shear along y=q requires three alb ratios and with the point support at
expanding the point load at t, q into a Fourier the center (f=a/2, q = 15/29are listed in table 4.97
e
sine series. These conditions lead to the char- Frequencies were also determined (ref. 4.137)
acteristic equation for the case of the square when the support
point was allowed to relocate along the line
sinh X,q sinh h4(b-q) y=a/2=6/2. Results are given in table 4.98.
k4sinh Xqb It is noted that corresponding values (t/a= x,
a/b=P) of tables 4.97 and 4.98 show consider-
able disagreement.
140 VIBRATION OF PLATES
€la Q % % W %
---___-
FIGURE
4.86.-Square hub-pin plate.
RECTANGULAR PLATEsl 141
A --Expenmentol
0 -Theoretical
FIGURE 4.87.-Theoreticai and experimental mode shapes for a square hub-pin plate.
(=?/a and s=iJa. Further experimental re- man (ref. 4.148). I n reference 4.148, the
sults (ref. 4.86) for a thinner plate are shown Rayleigh-Ritz method is used with a funda-
in figure 4.88. More work on point-supported mental mode shape
plates is contained in reference 4.147.
4.5 ADDED MASS W ( z ,y)=sin -
(4.102)
Tbe problem of the rectangular plate, simply
supported on two opposite edges, free on the where A is an undetermined coefficient to be
other two, and carrying a rigid mass of finite found from the minimization process. An ex-
width I running across the center of the plate plicit formula for the frequency parameter is
(fig. 4.89) was studied by Cohen and HandeI- found t o be
(4.104)
142 VIBRATION O F PLATES
CO.80-
e o 70
p0.W
ROTATIONAL
CONSTRAINT POINT
C060--
+0.40-
-0.4 -0.2
E.0.l
xe 0 SHAKER PMITION-
NO DEFLECTION DATA
-0.m 2
2
/j
t.O.30
E.O.50
-0.80
E.0.70
-I 001
(=O 50 ------a
F=o70
[=0.90 -0.40
(=Os0
-0.60
e SHIiKER POSITION -
0 SHAKER POSITION - NO DEFLECTION DATA
NO DEFLECTION DATA
RECTANGULAR PLATES 143
+I.007-
€=O.WP * '"T
FIGURE 4.88.-Experimental node lines and normalized deflection of a square hub-pin plate;
material, 6061-T6 aluminum W inch thick. (a) Experimental node lines and data points.
(b) First mode;fl=58.8 cps. (c) Second mode;f2=119 eps. (d) Third mode; f3=339
eps. ( e ) Fourth mode; f4=462 cps. (f) Fifth mode; f5=570 cps.
FIGURE
4.89.-SS-F-SS-F plate carrying a rigid mass.
144 VIBRATION O F PLATES
and where p' is the mass density per unit area large values of alb, equation (4.104) can be
of the plate plus the additional mass in the simplified by retaining terms of order (b/u)2,
region <
( 4 2 )- (112) 1 x< (a121-/- (112) 1. For ut no higher powers, giving:
X2' (4.105)
(
tan0 1--
f) f
+/3-=0 (4.107)
15
5
2
0
0
lh jI0
E3=& (1 - a"> 7
bo
(4.110) FIGURE4.92.-Variation of the first antisymmetric
frequency for a SS-F-SS-F square plate carrying a
For large values of ajb, equation (4.108) rigid strip mass. (After ref. 4.149)
simplifies t o
FIGURE
4.94.-SS-SS-SS-SS rectangular plate with a
point mass M.
tanh:t'(2m+l)2-X
where
(4.113)
FIGURE4.93.-Variatien of the first antisymmetric
frequency for a SS-F-SS-F rectangular plate carrying The function .fl(X) is given in table 4.180 and
a rigid strip mass; a/b= 10. (After ref. 4.149) plotted in figure 4.95.
TABLE4.100.-Characteristie Functions for a SS-SS-SS-SS Square Plate Having a Mass at
the Center
0 -m -m
2,y)=&(-1)"
m=O
23. 4
17. 5
15. 0
13. 2
the eigenfunctions of the plate without masses, 12. 0
and the equations of motion of the plate-mass 11. 0
system are determined from Lagrange's
equation reference 4.155. The characteristic equation
for the simply supported square having a mass
and a spring a t its center is equation (4.112)
with the right-hand side modified to become
where T is the kinetic energy of the plate-
mass system, U is the potential energy of the
system, q, are the generalized coordinates cor-
responding to the eigenfunctions used, and t
is time. The resulting infinite set of ordinary
for doubly symmetric modes. Again, values
differential equations in the qr are solved for of X ( X ) may be taken directly from table 4.100
the frequencies of the system in the usual and figure 4.95. From equation (4.119) and
manner.
figure 4.95 it is seen that for w = J k m the
Numerical results were obtained for a simply
vibrations of the spring-mass system and the
supported aluminum plate 20 by 20 inches by
plate become uncoupled. As k + w , j , ( X ) +0
0.091 inch with a concentrated mass having
and the solution is that of a rigid point support
negligible rotational inertia at the center. By at the center.
assuming a specific weight of 0.0955 pound per
Consider next the simply supported square
cubic inch for aluminurn, this gives the weight
plate having four equal masses symmetrically
of the plate as 3.48 pounds. Theoretical and
located along its diagonals as shown in figure
experirr ental fundamental cyclic frequencies
4.96. For modes symmetric with respect to
were obtained and are given in table 4.102.
x=a/2 and antisymmetric with respect to
nly the first four eigenfunctions of the S%
y = a / 2 , the frequencies may be determined
SS-SS-SS plate were used in the calculation of
from the characteristic equation (ref. 4.150)
the frequencies. c
Table 4.103 (ref. 4.153) lists the results for
the effect of adding various numbers of cylin- Y
f21N
100
80
60
40
m
0 -A
m
40
60
80
100
150 VIBRATION O F PLATES
In references 4.156 and 4.157, Solecki gives The problem of the SS-SS-SS-C square plate
the fundamental frequency of a square plate having two point masses, one at (1=71=0.2a
clamped all around and having a point mass at and the other at [2=72=0.4a, was also solved
the center of twice its own mass. The fre- by Das and Navaratna (ref. 4.158). Frequency
quency is found t o be ratios are shown in figure 4.100.
A method for determining frequencies of
(4.122) rectangular plates having added masses and
elastic edge constraints is given in reference
The problem of the rectangular plate having 4.131. Theoretical and experimental fre-
Lhree sides simply supported and the ot,her quencies are given for specific plates used as
clamped and having a mass M and a spring of deetronic chassis.
st,iffness k attached at a given point (fig. 4.98) For a specific ease of a rectangular cantilever
was solved in reference 4.158. The method plate having added mass a t the tip (x=a>,
used was essentially that given in reference see the discussion under parallelogram plates
4.153 and discussed previously in this section. entitled “Other Supports and Conditions”
Ratios of the fundamental frequency of the (sec. 5 . 2 ) .
system to that of the plate alone as functions
of the stiffness ratio k / k c and the mass ratio
M/pab are shown in figure 4.99 for the case of
the square, and E=q=0.2a. The quantity
k, may be thought of as a generalized spring
constant corresponding to a uniformly loaded
SS-SS-SS-Cr square plate of negligible mass;
that is, k,=D/0.00279a2.
Frequency Ratio
Q
0
0
Frequency Ratio N
3
FIGURE 4,1QO.--Ratio of the fundamental frequency of
a SS-SS-SS-C square plate having point masses
M I and M z at E1=71=Q.2a and Ez=?z=Q.4a, respec-
tively, t o that of the plate alone.
Frequency ratio
Nodal pattern !
a
i
Clamped edge SS edge
-
0 1.000 1.000
a .2 . 986 .985
.4 1. 118 .965
0 1.000 1.000
.2 .916 . 913
.4 .
876 .go4
0 1.000 1.000
.2 1. 040 1. 024
.4 1. 195 1.228
--
308-337 0 - 70 - ii
154 VIBRATION OF PLATES
___
Coefficients retained
_______
Ai Az ' A3 A4 A1-43
~-___________~__-_
AIAZAO A1A3A4 AIA~AAJ
3. _
oaZJa. _ _ _ _ _ _ _ 189_ _
9. 478
_ _ 7.
_ 617
.__25.45
_ _ 3. 026 2. 914 2. 962 2. 896
II
I
7
(I X
REFERENCES
- x
4.1. WARBURTON, G. B.: The Vibration of Rectan-
gular Plates. Proc. fnst. Meeh. Eng., ser. A,
FIGURE4.106.--F-F-F-F square plate with a central voi. 168, no. 12, 1954, pp. 371-384.
square hole.
4.2. JANICE,R.: Die naherungsweise Berechnung der
Eigenfrequenzen von rechteckigen Platten bei
verschiedenen Randbedingungen. Die Bau-
tion (4.126) are listed in table 4.108. The technik, vol. 3, Mar. 1962, pp. 93-99.
lowest value is the most accurate. 4.3. HEARMON, R. F. S.: The Frequency of Vibra-
Consider next the rectangular plate simply tion of Rectangular Isotropic Plates. J. Appl.
supported on all external edges and having a Mech., vol. 19, 1952, pp. 402-403.
narrow slit of length e along one axis of sym- 4.4. VET, MAARTEN:Natural Frequencies Qf 'Thin
metry as shown in figure 4.107. This problem is Rectangular Plates. Mach. Design, vol. 37,
studied in reference 4.136. One numerical no. 13, June 1965, pp. 183-185.
4.5. NADAI, A,: Die Elastiche Flatten. Julius
result is given but it is highly inaccurate. 'The Springer (Berlin), 1925, pp. 164-173.
case when the slit is completely internal is 4.6. LECKIE,F. A.: The Application of Transfer
formulated in reference 4.161, but no numerical Matrices to Plate Vibrations. 1ngr.-Arch.,
results for vibration are given. V O ~ .32, no. 2, Feb. 1963, pp. 100-111.
RECTANGULAR PLATES 155
4.7. FLETCHER, H. J.; WOODFIELD, N.; A N D LARSEN, 4.22. UNGAR,E. E.: Free Oscillations of Edge-Con-
K.: Natural Frequencies of Plates With Op- nected Simply Supported Plate Systems.
posite Edges Supported. Contract DA-04- J. Eng. Ind., vol. 4, Nov. 1961, pp. 434-440.
495-ORD-560 (CFSTI NO. AD 107 224), 4.23. KIST’YAN,K. YA: Approximate Determination
Brigham Young Univ., Sept. 1956. of the Frequencies of Rectangular Plates.
4.8. FLETCHER, H. J.: The Frequency of Vibration Prikl. Mekh., vol. 8, no. 2, 1962, pp. 219-222.
of Rectangular Isotropic Plates. J. Appl. (In Ukrainian.)
Mech., vol. 26, no. 2, June 1959, p. 290. 4.24. DILL, E. H.; A N D PISTER,K. S.: Vibration of
4.9. IGUCHI,S.: Die Eigenwertprobleme fur die Rectangular Plates and Plate Systems. Proc.
elastische rechteckige Platte. Mem. Fac. 3d U.S. Natl. Congr. AppL Mech., 1958, pp.
Eng., Eiokkaido Univ., 1938, pp. 305-372. 123-132.
4.10. DAS, Y. C.: On the Transverse Vibrations of 4.25. VOLMIR, A. S.: Stability of Elastic Systems.
Rectangular Isotropic Plates. J. Aeron. SOC. Gos. Izd. Phys.-Mat. Lit. (Moscow), 1963.
India, vol. 13, no. 4, Nov. 1961, pp. 111-117. (In Russian.)
4.11. VELETSOS,A. S.; A N D NEWMARK, N. M.: Deter- 4.26. VOIGT, W.: Bemerkungen zu dem Problem der
mination of Natural Fre-uencies of Contin- transversalen Schwingungen rechteckiger Plat-
uous ?lates Hinged Along Two Opposite ten. Nachr. Ges. Wiss. (Gottingen), no. 6,
Edges. 3. Appl. Mech., vol. 23, no. 1, Mar. 1893, pp. 225-230.
1956, pp. 97-102. 4.27. JANKOVIC, V.: The Solution of the Frequency
4.12. LURIE, H.: Vibrations of Rectangular Plates. Equation of Plates Using Digital Computers.
J. Aeron. Sei., vol. 18, no. 2, Feb. 1951, pp. Stavebnicky Casopis, vol. 12, no. 6, 1964,
139-140. pp. 360-365. (In Czech.)
4.13. ODMAN,S. T. A.: Studies of Boundary Value 4.28. ZEISSIG, C.: Ein einfacher Fall der transversalen
Problems. Part 11. Characteristic Functions Schwingungen einer rechteckigen elastischen
of Rectangular Plates. Proc. Platte. Ann. Physik, Bd. 64, 1898, pp. 361-
Cement and Concrete Res. Inst., Roy. Inst. 397.
Tech. (Stockholm), 1955, pp. 7-62. 4.28. NARUOXA,M.: On Transverse Vibration of
4.14. NISHIMURA, T.: Studies on Vibration Rectangular Flat Plates Clamped at Four
of Flat Plates by Means of Difference Edges. Trans. JSME, vol. 17, no. 57, 1951,
Proc. 3d Jap. Natl. Congr. Appl. Mech., pp. 26-30. (In Japanese.)
1953, pp. 417-420.
4.30. TOMOTIKA, S.: The Transverse Vibration of a
4.15. HAMADA, M.: A Method for S
Square Plate Clamped at Four Edges. Phil.
Mag., ser. 7, vol. 21, no. 142, Apr. 1936, pp.
gular Plates With Clamped or Supported
745760.
Edges. Bull. JSME, vol. 2, no. 5, 1959, pp.
92-97. 4.31. SUZUBI,S.: On the Transverse Vibrations of
4.16. KANAZAWA, T.; A N D KAWAI,T.: On the Lateral Rectangular Flat Plates Clamped at Four
Vibration on Anisotropic Rectangular Plates. Edges. Trans. JSME, vol. 13, no. 44, pp.
?roc. 2d Jap. Natl. Congr. Appl. Mech., 1952, 1-50-1-58. (In Japanese.)
pp. 333-338. 4.32. HABATA, Y.: On the Lateral Vibration of a Rec-
4.1’7. TXMOSXENKO, S.; AND GERE, J. M.: Theory of tangular Plate Clamped a t
Elastic Stability. McGraw-Hill Book Go., JSME, vol. 13, no. 44, 1947, pp. 67-83. (In
h e . , 1961. Japanese.)
4.18. KIST’YAN,K . YA: An Approximate Determina- 4.33. MUNAKATA, IS.: On the Vibration and Elastic
tion d the Frequencies of Vibrations of Plates Stability of a Rectangular Plate Clamped a t
and Their Subsequent Improvement. Str-vo Its Four Edges. J. Math. & Phys., vol. 31, no.
Arkhitekt., no. 9, 1959, pp. 67-70. (In 1, Apr. 1952, pp. 69-74.
Russian.) 4.34. DEVERALL, L. I.: Modified Method of Colloca-
4.19. SONNEMAN, GEORGE: On Correlation of Buck- tions for Vibrating Plate Problems. Int.
ling and Vibration of Plates. Proc. 1st Conf. on Partial Differential Equations and
Midwest. Conf. Solid Mech., Eng. Exp. Sta., Continuum Mechanics (Univ. Wisconsin, 1960),
Univ. Michigan, Apr. 1953, pp. 124-126. Univ. Press, 1961.
4.20. KATO, H.: On the Bending and Vibration of 4.35. CLAASSEN, R. W.; A N D THORNE, C. J.: Vibrations
Rectangular Plates. J. SOC.Nav. Architect., of Thin Rectangular Isotropic Plates. J. Appl.
vol. 50, 1932, pp. 209-230. (In Japanese.) Mech., vol. 28, no. 2, June 1961, pp. 304-305.
4.21. SEZAWA,K.: On the Lateral Vibration of a 4.36. GEAAPSEN, R. W.; AND TEORNE, 6. J.: Trans-
Rectangular Plate Clamped a t Four Edges. verse Vibrations of Thin Rectangular Isotropic
Rept. No. 70, Aeron. Res. Inst., Tokyo Univ., Plates. NOTs Tech. Pub. 2379, NAVWEPS
1931, pp. 61-70, Rept. 7016 U.S. Naval Ordnance Test Sta.,
156 VIBRATION OF PLATES
China Lake, Calif., Aug. 1960. (Errata 4.52. SOROKIN, E. S.: Dynamics of Between-Floors
available from CFSTI as AD 245 000.) Ceilings. Stroiiedat (Moscow), 1941. (In
4.37. JOGA-RAO, C. V.; A N D PICKETT, GERALD: Vibra- Russian.)
tions of Plates of Irregular Shapes and Plates 4.53. ESCHLER, H.: Zur Ermittlung der Eigenschwin-
With Holes. J. Aeron. SOC. India, vol. 13, gungszahlen der in ihrer Mittelebene belasteten
no. 3, 1961, pp. 83-88. Rechteckplatte. 1ngr.-Arch., Bd. 18, 1950,
4.38. HIDAKA, K.: Vibration of a Square Plate Clamped pp. 330-337.
at Four Edges. Math. Jap., vol. 2, 1951, 4.54. COLLATZ, L.: Das Mehrstellverfahren bei Plat-
pp. 97-101. tenaufgaben. ZAMM, vol. 30, 1950, pp.
4.39. GREEN,A. E.: Double Fourier Series and Bound- 385-388.
ary Value Problems. Proc. Cambridge Phil. 4.55. GONTKEVICH, V. S.: The Lower Bounds of
SOC.,1944. Natural Frequencies of a Plate in Flexural
4.40. LEISSA,A. W.: A Method for Analyzing the Vibrations. Prikl. Mekh., AN UkrSSR, vol.
Vibration of Plates. J. Aerospace Sci., vol. 6, no. 3, 1960, pp. 346-350. (In Ukrainian.)
29, no. 4, Apr. 1962, p. 475. 4.56. SEZAWA,K.: Free Vibration of a Clamped
4.41. TOMOTIKA, S.: On the Transverse Vibration of Square Plate. J. Aeron. Res. Inst., no. 6,
a Square Piate With Clamped Edges. Aeron. 1924, pp. 29-42.
Res. Inst. Rept., Tokyo Univ., vol. 10, 1935, 4.57. BOLOTIN,V. V.: Dynamic Edge Effect in the
p. 301. Elastic Vibrations of Plates. Inshen. Sbornik.,
4.42. HERSCH,J.: Contribution A la MBthode des vol. 31, 1961, pp. 3-14. (In Russian.)
Gquations aux Diffbrences. Z IMP, vol. 9a, 4.58. BAZLEY,N. W.; Fox, D. W.; AND STADTER,
no. 9, 1958, pp. 129-180. J. T.: Upper and Lower Bounds for the
4.43. ABRAMOWITZ, M.; A N D CAHILL,W. F.: On the Frequencies of Rectangular Clamped Plates.
Vibration of a Square Clamped Plate. J. Tech. Memo. TG-626, Appl. Phys. Lab.,
Assoc. Comput. Mach., vol. 2, no. 3, 1955, The Johns Hopkins ‘C‘niv., May 1965.
pp. 162-168. 4.59. BAZLEY, N. W.; A N D Fox, D. W.: Methods for
4.44. AYNOLA,L. YA.: Methods of Investigating Lower Bounds to Frequencies of Continuous
Elastic Plate Vibrations. Inz. Zhurnal, vol. 3, Elastic Systems. Rept. TG-609, Appl. Phys.
no. 2, 1963, pp. 312-321. (In Russian.) Lab., The Johns Hopkins Univ., OGt. 1964.
4.45. GALIN,M. P.: On the Transverse Vibrations of 4.60. BOLOTIN,V. V.; MAKAROV, B. P.; MISEENKOV,
Plates. Prikl. Mat. Mekh., V Q ~ . 12, no. 3, G. V.; A N D SHVEIKO, Yu Yu: Asymptotic
1947, pp. 387-388. (In Russian.) Method of Investigating the Natural Fre-
4.46. STANISIC: M.: An Approximate Method Applied quency Spectrum of Elastic Plates. Raschet
to the Solution of the Problem of Vibrating na Prochnost, Mashgie (Moscow), no. 6,
Rectangular Plates. J. Aeron. Sci., vol. 24, no. 1960, pp. 231-253. (In Russian.)
2, Feb. 1957, pp. 159-160. 4.61. KAUL,R. K.; A N D CADAMBE, V.: The Natural
4.47. YOUNG,D.: Vibration of Rectangular Plates by Frequencies of Thin Skew Plates. Aeron.
the Ritz Method. J. Appl. Mech., vol. 17, Quart., vol. 7, 1956, pp. 337-352.
no. 4, Dec. 1950, pp. 448-453. 4.62. IWATO: Approximate Calculation for the F r e
quency of Natural Vibration of a Thin Rec-
4.48. NAGARAJA, J.; A N D RAO, S. S.: Vibration of
tangular Plate the Two Adjacent Edges of
Rectangular Plates. J. Aeron. Sci., vol. 20, no.
Which Are Clamped While the Other Two
12, Dee. 1953, pp. 855-856.
Edges Are Freely Supported. Trans. JSME,
4.49. ARONSZAJN,N.: The Rayleigh-Ritz and the
vol.17, no. 57, 1951, pp. 30-33. (In Japanese.)
Weinstein Methods for Approximation of 4.63. GLAASSEN,R. W.; A N D THORNE, C. J.: Vibrations
Eigenvalues-111: Application of Weinstein’s of a Rectangular Cantilever Plate. Tech.
Method With an Auxiliary Problem of Type I Rept. PMR-TR-61-1, Pacific Missile Range,
Tech. Rept. no. 3, Proj. NR 041, 090, Okla- Aug. 1962.
homa A. and M. College (Stillwater, Okla.), 4.64. DILL, E. H.: The Vibration of Plates and Plate
1950. Systems. Ph. D. Thesis, Univ. Calif. (Berkeley,
4.50. HOPKINS,H. 6.: The Solution of Small Dis- Calif.), Jan. 1957.
placement, Stability, or Vibration Problems 4.65. PLASS, H. J., JR.; GAINES, J. H.; A N D NEWSOM,
Concerning a Flat Rectangular Panel When C. D. : Application of Reissner’s Variational
the Edges Are Either Clamped or Simply Sup- Principle to Cantilever Plate Deflection and
ported. R.A.E. Rept. No. SME 3331, R. & Vibration Problems. J. Appl. Mech., vol. 29,
M. No. 2294, British A.R.C., June 1945. no. 1, Mar. 1962, pp. 127-135.
4.51. GONTKEVICH, V. S.: Natural Vibrations of Plates 4.66. AUSTIN, R. N.; CAUGHFIELD, D. A.; A N D PLASS,
and Shells. A. P. Filippov, ed., Nauk. Dumka fI. J., JR.: Application of Reissner’s Varia-
(Kiev), 1964. (Transl. by Lockheed Missiles tional principle to the Vibration Analysis of
& Space Co. (Sunnyvale, Calif.).) Square Flat Plates With Various Root Sup-
RECTANGULAR PLATES 157
port Conditions. Developments in Theo- Skew and Rectangular Cantilever Plates.
retical and Applied Mechanics. Vol. 1. Ple- Proc. SOC. Exp. Stress Anal., vol. 9, no. 2, 1952,
num Press (New York, N.Y.), 1963, pp. 1-24. pp. 51-66.
4.67. GAINES, J. H.: Application of the Modified 4.82. DALLEY, J. W.; AND RIPPERGER, E. A.: Experi-
Reissner Variational Principle to a Cantilever mental Values of Natural Frequencies for
Plate Problem. Def. Res. Lab. Rept. DRL- Skew and Rectangular Cantilever Plates. Def.
440, CF 2766, Univ. Texas, Mar. 1959. Also, Res. Lab. Rept. DRL-231, CF-1354, Univ.
M. Sc. Thesis, Univ. Texas, June 1959. Texas, Dec. 1949.
4.68. CAUGHFIELD, D. A.: Application of the Modified 4.83. HEIBA, A. E.: Vibration Characteristics of a
Reissner Variational Principle to a Cantilever Cantilever Plate With Sweptback Leading
Plate Problem Using the Techniques of Nu- Edge. Rept. No. 82, Cranfield College of
merical Integration and Finite Differences. Def. Aeron., Oct. 1954.
Res. Lab. Rept. DRL-457, CF 2902, Univ. 4.84. PLUNKETT, R. : Natural Frequencies of Uniform
Texas, Jan. 1961. Also, M. Sc. thesis, Univ. and Non-uniform Rectangular Cantilever
Texas, Jan. 1961. Plates. J. Mech. Eng. Sci., vol. 5, no. 2, 1963,
4.69. REISSNER,E.: On a Variational Theorem in pp. 146-156.
Elasticity. J. Math. Phys., v01. 29, 1950, 4.85. WILSON,R. E.; AND PLUNKETT, R.: Vibration of
pp. 90-95. Cantilever Plates With Rectangular and
4.70. MACNEAL, R. H.: The Solution of Elastic Plate Wedge-Shaped Cross-Sections. Rept. DF
Problems by Electrical Analogies. J. Appl. 53GL17, Gen. Elec. Co., Mar. 1953.
Mech., vol. 18, no. 1, Mar. 1951, pp. 59-67. 4.86. CRAIG,R. R.; PLASS, H. J., JR.;AND CAUGHFIELD,
4.71. GREENWOOD, DONALD: Some Difference Methods D. A.: Experimental Determination of Fre-
of Plate Vibration Analysis. NASA Grant quencies and Mode Shapes of Cantilever and
NsG-63-60; CFSTI NO. N62-14018. Hub-Pin Plates. Def. Res. Lab. Rept. DRL-
4.72. CRAIG,R. R.; A N D PLASS, H. J.: Vibration of 518, CR-13, Univ. Texas, June 1964.
Hub-Pin Plates. AIAA J., vol. 3, no. 6, June 4.87. GUSTAFSON, P. N.; STOKEY, W. F.; AND ZOROW-
SKI, C. F.: The Effect of Tip Removal on the
ion of Rectangular and Natural Vibrations of Uniform Cantilevered
Skew Cantilever Plates. J. Appl. hlech., Triangular Plates. J. Aeron. Sei., vol. 21,
vol. 18, no. 1, June 1951, pp. 129-134. no. 9, Sept. 1954, pp. 621-633.
Free Vibration Characteristics 4.88. WALTON,W. C., JR.: Applications of a General
Plates. Defense Res. Lab. Finite-Difference Method for Calculating Bend-
Rept. DRL-222, CM 570, Univ. Texas, Dee. ing Deformations of Solid Plates. NASA T N
1949. D-536, 1960.
4.75. BAZLEY, N. w . ; F O X , D. w; AND STADTER,
4.89. HOUBOLT, J. C.: A Study of Several Aerother-
J. T.: Upper and Lower Bounds for Fre- moelastic Problems of Aircraft Structures in
quencies of Rectangular Cantilever Plates. High Speed Flight. Ph. D. Thesis, Zurich
Tech. Memo. TG-705, Appl. Phys. Lab., E.T.H., 1958.
The Johns Hopkins Univ., July 1965.
4.90. DRAPER,K. J.; IRONS, B.; AND BAZELEY,G.:
4.76. SIGI LITO, V. 6.:Improved Upper Bounds for
Comment on Vibration of a 45" Right Tri-
Frequencies of Rectangular Free and Canti-
angular Cantilever Plate by a Gridwork
lever Plates. Eng. Memo. EM-4012, Appi.
Method. AIAA J., voi. 2, no. IO, Oct. 1964,
Phys. Lab., The Johns Hopkins Univ., Dec.
pp. 1870-1871.
1965.
4.77. CLAASSEN, .R. W.; A N D THORNE, C. J. : Vibrations 4.91. HALL,A. H.: The Nature and
of a Rectangular Cantilever Plate. J. Aero- Wing Deformations With Reference t o the
space Sci., vol. 29, no. 11, Nov. 1962, pp. Prediction of Normal Modes and Frequencies.
1300-1 305. Preprint no. 494, Can. Aeron. Inst. of Aeron.
4.78. MARTIN,A. I.: On the Vibration of a Cantiiever Sci., presented a t CAI-IAS Int. Meeting,
Plate. Quart. J. Mech. Appi. Math., v01. 9, Oct. 1954. I n CAS1 J., Sept. 1955.
pt. 1, 1956, pp. 94-102. 4.92. BARTON,M. V. : Free Vibration Characteristics
4.79. GRINSTED,B.: Nodal Pattern Analysis. Proc. of cantilever Plates. Def. Res. Lab. Rept.
Inst. Mech. Eng., ser. A, vol. 166, 1952, pp. DRG212, CF-1258, Univ. Texas, May 1949.
309-326. 4.93. DALLEY, J. W.: Experimental Studies on Vibra-
4.80. FORSYTH, E. M.; A N D WARBURTON, G. B.: Tran- tion Characteristics of Some Idealized Missile
sient Vibration of Rectangular Plates. J. Fins. Def. Res. Lab. Rept. DRL-402, CF-
Mech. Eng. Sei., vol. 2, no. 4, Dee. 1960, 2643, Univ. Texas, Sept. 1957.
pp. 325330. 4.94. REISSNER,E.; AND STEIN, M.: Torsion and
4.81. DALLEY, J. W.; A N D RIPPERGER, E. A.: Experi- Transverse Bending of Cantilever Plates.
mental Values of Natural Frequencies for NACA T N 2369, 1951.
158 VIBRATION OF PLATES
4.95. CHLADNI,E. F. F.: Entdeckungen iiber die 4.115. NAKATA,Y.; AND FUJITA, H.: On Upper and
Theorie des Klanges. Leipzig, 1787. Lower Bounds of the Eigenvalues of a Free
4.96. CHLADNI,E. F. F.: Die Akustik. Leipzig, 1802. Plate. J. Phys. SOC.Japan, vol. 10, 1955, pp.
4.97. CHLADNI,E. F. F.: Ann. Physik, Leipzig, 823-824.
vol. 5, 1825, p. 345. 4.116. PAYNE, L. E.: Inequalities for Eigenvalues of
4.98. CHLADNI,E. F. F.: Neue Beitrage zur Akustik. Supported and Free Plates. Quart. J. Appl.
Leipzig, 1817. Math., vol. 16, no. 2, 1958, pp. 111-120.
4.99. STREBLKE:Ann. Physik (Leipzig), vol. 4, 4.117. KATO,T.; FUJITA, H.; NAICATA, Y.; AND NEW-
1825, p. 205. MAN, M.: Estimation of the Frequencies of
4.100. STREBLKE:Ann. Physik (Leipzig), vol. 18, Thin Elastic Plates With Free Edges. J. Res.
1830, p. 198. Natl. Bur. Std., vol. 59, no. 3, Sept. 1957, pp.
4.101. STREHLKE:Ann. Physik (Leipzig), vol. 27, 169-186.
1833, p. 505. 4.118. BAZLEY,N. W.; Fox, D. W.; AND STADTER,
4.102. STREHLKE:Ann. Physik (Leipzig), vol. 95, J. T.: Upper and Lower Bounds for the Fre-
1855, p. 577. quencies of Rectangular Free Plates. Tech.
4.103. STREHLHE:Ann. Physik (Leipzig), vol. 146, Memo TG-707, Appl. Physics Lab., The
1872, p. 319. Johns Hopkins Univ., Aug. 1965.
4.104. KONIG,R.: Pogg. Ann., vol. 122, 1864. 4.119. WALLER,MARYD.: Vibrations of Free Square
4.105. TANAKA, S.: Ann. Physik (Leipzig), vol. 32 Plates. Proc. Phys. SOC. (London), vol. 51,
1887, p. 670. Jan. 1939, pp. 831-844.
4.106. WHEATSTONE, CHARLES:On the Figures Ob- 4.120. PAVLIK, B. : Biegungsschwingungen bei mag-
tained by Strewing Sand on Vibrating Surfaces, netostriktiv erregten Kreisplatten. Ann.
Commonly Called Acoustic Figures. Phil. Physik, Bd. 26, Heft 5, 1936, p. 625.
Trans. Roy. SOC.(London), 1833, pp. 593-633. 4.121. WALLER,MARYD.: Concerning Combined and
4.107. RAYLEIOH,Lord: On the Nodal Lines of Degenerate Vibrations of Plates. Acustica,
a Square Plate. Phil. Mag., ser. 4, vol. 46, vol. 3, 1953, pp. 370-374.
no. 304, Aug. 1873, pp. 166-171. 4.122. WALLER, MARYD.: Fundamental Vibration of a
4.108. RITZ,W.: Theorie der Transversalschwingungen, Rectangular Plate. Nature, vol. 143, no.
einer quadratischen Platte mit freien Randern. 3610, Jw. 1939, pp. 27-28.
Ann. Physik, Bd. 28, 1909, pp. 737-786. 4.123. WALLER,MARYD.: Vibration of Free Rectan-
4.109. GOLDMANN, ELLA:Anwendung der Ritzschen gular Plates. Proc. Phys. SOC.(London), ser.
B, vol. 62, no. 353, 1949, pp. 277-285.
Methode auf die Theorie der Transversal-
schwingungen freischwingender Platten von 4.124. PFEIFFER,F.: Ubergang zu Systemen mit
rechteckiger, rhombischer, dreieckiger und unendlich vielen Freiheitsgraden. Ch. 4 ,
elliptischer Begrenzung. Dissertation, Bres- Handbuch der Phys., Julius Springer (Berlin),
lau Univ., 1918. 1928, pp. 337-402.
4.125. RAYLEIGH, Lord: On the Calculation of Chladni
4.1 10. LEMKE, A. : Experimentelle Untersuchungen
Figures for a Square Plate. PGil. Mag., vol .
zur W. Ritzschen Theorie der Transversal-
22, 1911, pp. 225-229.
schwingungen Quadratischer Platten. Ann.
4.126. TANAKA, S.: Ann. Physik (Leipzig), vol. 32,
Physik, Bd. 4, ser. 86, 1928, pp. 717-750.
1887, p. 670.
4.111. PAVLPIC, B.: Beitrag zur Theoretischen und 4.127. EUNDQUIST, E. E.; A N D STOWELL, E.: Critical
Experimentellen Untersuchung der Biegungs- Compressive Stress for Flat Rectangular
schwingungen bei rechteckigen Platten mit Plates Supported Along All Edges and Elasti-
freien Randern. Ann. Physik, 1936, pp. cally Restrained Against Rotation Along the
532-542. Unloaded Edges. NACA Rept. 733, 1942.
4.112. PAVLIK,B.: Beitrag zur Untersuchung des 4.128. CARMICHAEL, T. E.: The Vibration of a Rec-
Zusarnmenhanges der bei Beigungsschwing- tangular Plate With Edges Elastically Re-
ungen an rechteckigen und quadratischen strained Against Rotation. Quart. J. Mech.
Platten beobachteten Staubfiguren. Ann. Appi. Math., vol. 12, pt. 1, 1959, pp. 29-42.
Physik, Bd. 28, Heft 5, ser. 632, 1937, pp. 4.129. JOGA-RAO, C. v.; A N D KPINTHAM, c. L.: Natural
632-648. Frequencies of Rectangular Plates With
4.113. IGUCHI, S. : Die Eigenschwingungen und Klang- Edges Elastically Restrained Against Rota-
figuren der vierseitig freien rechteckigen Platte. tion. J. Aeron. Sci., vol. 24, no. 4, Nov.
1ngr.-Arch., Bd. 21, ser. 303, Heft 5-6, 1953, 1957, pp. 855-856.
pp. 304-322. 4.130. CEULSY, S. J.: Vibration of Elasticaliy Re-
4.1 44. GRAUEWS,H. : Transversdschwingungen rech- strained Rectangular Plates. M.S. Thesis,
teckiger Platten mit besonderer Rucksicht Univ. Wisconsin, 1957.
der K~ickung.Ingeniorsvetenskapsakademien, 4.131. STOKEY, W. F.; ZOROWSKI, 6.E'.; A N D APPL,
Handingar, 98, Stockholm, 1929. F. C. : Prevention of Mechanical Vibrations
RECTANGULAR PLATES 159
in Electronic Chassis-Design Manual. Rept. 4.146. BECK,6.W.: An Excitation and Instrumenta-
to Rome Air Develop. Center, Contract tion System for Vibrating Plate Studies.
AF30(602)-913, Sept. 1955. Def. Res. Lab. Rept. DRL-467, CF-2930,
4.132. HOPPMANN, W. H., 11; A N D GREENSPON, J.: An Univ. Texas, June 1961.
Experimental Device for Obtaining Elastic 4.147. PLASS, H. J.: Theoretical Studies on Vibration
Rotational Constraints on the Boundary of a Characteristics of Several Missile Fin Models.
Plate. Proc. 2d U.S. Natl. Congr. Appl. Def. Res. Lab. Rept. DRL-401, CF-2642,
Mech., 1954, pp. 187-191. Univ. Texas, July 1957.
4.133. OTA, T.; AND HAMADA, M.: Fundamental Fre- 4.148. COHEN,H.; A N D HANDELMAN, G.: Vibrations of
quencies of Simply Supported but Partially a Rectangular Plate With Distributed Added
Clamped Square Plates. Bull. JSME, vol. 6, Mass. J. Franklin Inst., vol. 261, no. 3,
no. 23, Aug. 1963, pp. 397403. Mar. 1956, pp. 319-329.
4.134. OTA,T.; AND HAMADA, M.: Bending and Vibra- 4.149. SUBRAMANIAN, N. R.; AND KTJMARASWAMY,
tion of a Simply Supported but Partially M. P.: Antisymmetric Vibrations of a Rec-
Clamped Rectangular Plate. Proc. 8th Jap. tangular Plate With Distributed Added Mass.
Natl. Congr. Appl. Mech., 1958, pp. 103-106. J. Aeron. SOC. India, vol. 12, no. 3, Aug.
4.135. KURATA,M.; A N D OKAMURA, H.: Natural T'i- 1960, pp. 63-68.
brations of Partially Clamped Plates. J. Eng. 4.150. GERSHGORIN, S.: Vibrations of Plates Loaded
Mech. Div., Proc. Am. SOC.Civil Eng., June by Concentrated Masses. Prikl. Mat. Mekh.,
1963, pp. 169-186. vol. 1, no. 1, 1933, pp. 25-37. (In Russian.)
4.136. NOWACKI, W.: Free Vibrations and Buckling of 4.151. WAH, T.: Natural Frequencies of Plate-Mass
a Rectangular Plate With Discontinuous Systems. Proc. Indian Soc. Theor. and
Boundary Conditions. Bull. Acad. Pol. Sci., Appl. Mech., 1961, pp. 157-168.
cl. 4, vol. 3, no. 4, 1955, pp. 159-167. 4.152. AMBA-RAO, C. L.: On the Vibration of a Rec-
4.137. NOWACKI,W.: Dynamics of Elastic Systems. tangular Plate Carrying a Concentrated
John Wiiey & Sons, Inc., 1963. Mass. J. Appl. Mech., vol. 31, no. 3, Sept.
4.138. KURLANDZKK, J.: A Method for Solving Prob- 1964, pp. 550-551.
lems of Rectangular Plates With Mixed 4.153. STOKEY,W. F.; AND ZOROWSKI,C. F.: Normal
Boundary Conditions. Proc. Vibration Probl., Vibrations of a Uniform Plate Carrying
vol. 2, no. 4, 1961, pp. 377-396. Any Number of Finite Masses. J. Appl.
4.139. Cox, H. L.; AND BOXER,J.: Vibration of Rec- Mech., vol. 26, no. 2, June 1959, pp. 210-216.
tangular Plates Point-Supported at the 4.154. LEE, W. F. Z.: Free and Forced Vibrations of
Corners. Aeron. Quart., vol. 11, no. 1, Feb. Constrained Beams and Plates. Ph. D . thesis,
1960, pp. 41-50. Carnegie Inst. Tech., 1952.
4.140. REED, R. E., JR.: Comparison of Methods in 4.155. FILIPPOV, A. P.: Vibrations of Elastic Systems.
Calcuiating Frequencies of Corner-Supported AN UkrSSR Press, 1956. (In Russian.)
Rectangular Plates. NASA TN D-3030, 4.156. SOLECKI,R.: Vibration of Plates With Con-
1965. centrated Masses. Bull. Acad. Pol. Sei.,
4.141. KIKK, C . L.: A Note on the Lowest Natural Ser. Sci. Tech., vol. 9, no. 4, 1961, pp. 209-215.
Frequency of a Square Plate Point-Supported
4.157. SOLECKI, R.: Vibrations of Straight Bars and
at the Corners. J. Roy. Aeron. SOC., vol.
Plates With Concentrated Masses. Roz-
66, no. 616, Apr. 1962, pp. 240-241.
prawy Inzh. CC 11, vol. 9, no. 3, 1961, pp.
4.142. Cox, N. L.: Vibration of a Square Plate, Point
497-511. (In Polish.)
Supported a t Midpoints of Sides. J. Acoust.
Soe. Am., vol. 27, no. 1, 1955, pp. 791-792. 4.158. DAS, Y. C.; A N D NAVARATXA, D. R.: T'ibrations
4.143. PLASS,N. J., JR.: Application of Reissner's of a Rectangular Plate With Concentrated
Variational Principle to Cantilever Plate Mass, Spring, and Dashpot. J. Appl. Mech.,
Deflection and Vibration Problems. Def. vol. 30, no. 1, Mar. 1963, pp. 31-36.
Res. Lab. Rept. DRL-418, CM-921, Univ. 4.159. TAKAKASHI,S.: Vibration of Rectangular
Texas, Aug. 1958. Plates With Circular Holes. Bull. JSME,
4.144. COX,H. L.: Vibration of Certain Square Plates voI. 1, no. 4, 1958, 380-385.
Having Similar Adjacent Edges. Quart. J. 4.160. XUMAI, T.: The Flexural Vibrations of a
Mech. Appl. Math., voi. 8, pt. 4, 1955, Square Plate With a Central Circular Hole.
pp. 454-456. Proc. 2d Jap. Natl. Gongr. Appl. Mech.,
4.145. NOWACKI, W.: Vibration and Buckling of 1952, pp. 339-342.
Rectangular Plates Simply-Supported a t the 4.161. SOLECKI,2.:Bending and Vibration of an
Periphery and a t Several Points Inside. Isotropic Rectangular Plate With a Hinged
Arch. Mech. Stos., vol. 5, no. 3, 1953, p. 437. Slot. Acta Polytech. Scandinavica, no. 12,
(In Polish.) 1962, pp. 3-19.
Chapter 5
es
Because no exact solutions t o equation (1.4) where k,a is themth positive root of the tran-
expressed in skew coordinates by equation scendental equation
(1.39) are known to exist in variables separable
form, no significant exact solutions exist for tan (k,a/2)=(-1)m tanh ( k m a / 2 ) (5.3)
parallelogram plates. Even the case when all
The functions #%(q) are obtained by replacing
edges are simply supported requires an intricate
& a, and m in equation (5.2) by q , b, and n,
solution, unlike the case of the rectangle (sec.
4.1). Some solutions have been obtained by respectively.
approximate methods for a few of the many Results were obtained in reference 5.1 by
possible combinations of boundary conditions. using only one term of equation (5.1) and the
Rayleigh method t o obtain upper bounds for
Particular emphasis exists in the literature for
the case of the cantilevered parallelogram frequency parameters for the case of the rhom-
because of its importance as an aerodynamic bus (a= b ) . These results are given in table 5.1 ;
lifting or stabilizing surface. the notation mfn is used to indicate the number
of approximate half sine waves in the S/q
5.1 SIMPLE EDGE C ~ N D ~ ~ I ~ ~ S directions, respectively (at least for sma
Results for plates with clamped (C), simply values of a>. Combined modes of the form
supported (SS), and free (F) edges are given (m/nfnfm> having nearly equal frequencies
in the following subsections. exist, as in the case of the square. (See see.
5.1.1 C-C-C-C 4.3.1.)
Kaul and Cadambe (ref. 5.1) proposed 8 Lower bounds were obtained in reference 5.1
solution to the problem of the C-C-C-C for some of the modes by use of the Kato-
parallelogram plate which used the Rayleigh-
Ritz method and the products of characteristic
TABLE5.1.-Frequency Parameters aa2yiplD
beam functions; that is,
cos2 a for a 6'-C-6'-C Rhombic Plate
W & ~ cos2
D for values of skew
angle, deg, of-
(Y,
15 30 45
--
36.67 38. 15 40.08
74. 76 77. 48 81. 06
111.43 118. 19 126. 84
132. 90 135. 96 140. 02
133.71 138. 03 142.70
169. 56 179. 12 191. 41
226.76 242. 04 261. 46
246. 91 258. 02 272. 36
249.67 261. 40 276.64
m=l, 2,3,.
161
162 VIBRATION O F PLATES
Maximum
Skew angle, 0, deg Mode type possible
Mean value percentage
deviation from
mean value
-____
Temple method. These are given in table 5.2 and by Hasegawa (ref. 5.5) who used the
along with a mean value of frequency parameter Rayleigh-Rita method and deflection functions
determined from the lower and upper bounds (see fig. 5.1)
and a computation of the maximum possible
error which can arise from using the mean Wi-, 5)=E"~a/2)"2 ~;i"-(b/2>"3'~A~
value.
It is clear from table 5.2 that the accuracies
of the solutions decrease as (1) the mode
number increases and ( 2 ) the skew angle
increases. These results are summarized in table 5.3 for
Further results for this problem were obtained a/b=1 and a/b=0.5. The problem is also
by Hamada (refs. 5.2 and 5.3) who used the discussed in reference 5.6.
method of Trefltz (ref. 5.4) and deflection In references 5.2 and 5.3, experimental results
functions for the rhombic plate were also given. Mild
steel plates with a=b=2.36 inches and h=0.035
inch were used. Figure 5.2 shows the ratio of
the freqhency of the rhombic plate t o that of
+&,, cos -
mat
a
sin-
b
nz-g
t C,, sin-
mrt
a
nag
cos-
b
the square as a function of the skew angle.
The curve shorn is from the theoretical results.
points are experimental data.
(5.4) Conway and Farnham (ref. 5.7) analyzed
PARALLELOGRAM PLATES 163
TABLE5.4.-Frequeney Parametersfor a C-C-C-C Rhombic Plate
FIGURE
5.L-GGGC parallelogram plate.
FIQURE
5.3.-C-GGC rhombic plate.
TABLE5.5.-Frequeney Parameters w a 2 C D
cos2 o for a C-(2-C-Si3 Rhombic Plate
racies of the solutions decrease as (1) the mode and 6 may be considered as a perturbation
number increases and (2) the skew angle parameter. Solutions for W and are then
increases. assumed in the form
5.1.4 ss-ss-ss-ss -
wmn=w~~+Ew:;+E~w~,n,s
(5.9)
*
C
(5.12)
(5.13)
n a
FIGURE
5.6.--SS-SS-SS-SS parallelogram plate.
166 VIBRATION OF PLATES
1 3
Amplitude coefficients - - - - - - - Air= 1.00000 A ~ z =I. 00000 A,,= 1.00000 A**= 1.00000
Azz= -. 00963 Azl= 0 Aiz=O All=. 03850
Azc= -. 00058 A%= -. 09020 Air=. 00219 Ais=. 11540
Aza= -. 00019 A2&=-. 00126 Ale=. 00028 Ais=. 00269
-.
A42= 00058 A*&=.00219 A32 = - .09020 &=. 11540
A44= -. 00009 A43= -. 00173 Aar= -.
00173 A s = -. 02880
-.
A4a= 00004 Ads= -. 00025 A36 = - .00033 A35= -.
00274
Aa2= -. 00019 A B I = .00028 Aai= -. 00126 Ab,=. 00269
AM=-. 00004 A s = -. 00033 A s = -. 00025 -.
A ~ J = 00274
Am=-. 00001 A i = -. 00007 Ass= -. 00007 Ass = -,00048
(5.15)
FIGURE
B.?.--Fundamentaf frequency parameters %I=
wllub&jB/2a as a function of skew angle LL and
aspect ratio parameter @ / a ) COS a for a SS-SS-SS-
(5.16) SS parailelogram plate.
PARALLELOGRAM PLATES 167
and the mode shapes by
(am+n)rx nr&y
-2sin a cos -
a
M= 2 -.A,J,(kr)cos@
n=1,3,.
(5.18)
(5.19)
M= 2.. A,J,(kr)sinqB
R.=1,2,.
(5.20)
along the diagonal AC having the length c. Analogies which permit one to obtain fre-
Pointwise symmetry conditions employed were quencies for polygonal plates simply supported
all around from the problems of either (a)
I
w T= cl3= WI 7 =sei 3 membrane vibration or (b) plate buckling due
Wlr=2ci3=WIr=4ci3 to hydrostatic pressure are discussed in the
chapter entitled “PIatesof Other Shapes” (ch. 8).
(5.21)
5.1.5 C-F-F-F
Barton (ref. 5.10) obtained the first compre-
hensive set of results for the problem of C-F-
F-F parallelograms (fig. 5.10) by using the
Solutions of the resulting fourth-order character- Rayleigh-Ritz method with deflection functions
istic determinants are given in table 5.11 for which are products of characteristic beam
various angles p and a16 ratios.
functions; that is,
Accuracy of the results can be estimated by
comparing values for p=90° with the known
exact ones (section entitled “All Sides SS” under w(t,S)=k AAnb(t)+n(ql
m = l n=1
(5.221
“Rectangular Plates” (4.1)) and the parameters where
for a/6= 1 with those of table 5.10. em5 __ em 5
c$~= eosh - cos -
a a
TABLE
5.10.-Frequency Parameters wc2&?D
for -am(sinha ern4 - sin -
SS-SS-SS-SS Rhombic Plates a
W
Cm
aDfor determinant of
size-
3by3 1 6by6
Results were computed by using 18 terms in graphs of nodal patterns obtained when a=30°
equation (5.22), and frequency parameters, are shown in figure 5.12.
nodal lines, and mode shape amplitude coefi- Claassen (refs. 5.13 and 5.14) extended the
cients are given in table 5.12 for cr=15’, 30°, work of reference 5.10 by using the same ana-
lytical procedure. A detailed Fortran pro-
gram statement listing for the procedure is also
given in reference 5.13. The first nine fre-
quency parameters for a=Oo, 5’, IO’, . . .,
3 5 O , and a/b=l are given in table 5.14. I n
reference 5.13, extensive frequency and node
resu-lts include an app line data are given in the vicinity of “transition
curves”; i.e., the frequencies at which the basic
form of the nodal pattern changes into another.
This phenomenon is discussed in the section on
rectangular plates entitled “All Sides Clamped”
nodal patterns, is shown in figure 5.11. Photo- (4.3.1). I n this case the mode shapes vary
TABLE
5.12-Frequeney P Lines, and Amplitude Coe$icients for GF-F-F
s; a/b=l; v=O.3
a, deg ..................... 15 30 45
2 1 2 1 2
Mode
Corrected Theoretical Uncorrected Corrected
Test results test results results percent percent
difference difference
I
2
3
3. 38
8. 63
21.49
3. 44
8.68 1
_______-_____I______________
3. 60
8.87
6. 2
2. 7
4. 6
2. 1
_ _ _ - _ _ _ - - - _- _ _ _- _ - _ _ _ - - - - _
4 26.04
5 33.01
1 3. 82
2 9. 23
3 24. 51
4 25.54
5 40.64
1 4. 26 10. 3
2 11.07 18.5
3 26.52
4 30. 13
5 50.19
172 VIBRATION OF PLATES
the swept centerline (or normal to the <:axis). Craig, Plass, and Caughfield (ref. 5.19)
Cyclic frequency parameters fa2Jmo/EIcos a, measured the first four frequencies and mode
where mo is mass (Slugs) Per unit length meas- shapes on aluminum rhombic plates having
ured along the [-direction, are shown in figure sweep angles of 150,300, 450, and 600. cyclic
5.15. Translational and rotational mode shape frequencies,nodal patterns, and mode shapes
deflections are listed in reference 5.18 for 12
for these four configurations are shown in fig-
values of + and the sweep angle, o/a, and
An estimate
EIiGJvariations just described. The volume of 5.'6 to 5*19,
theseresults (47 pages of tables) is too great of the accuracy of the nodal Patterns can be
to be included here. obtained from figure 4.47.
TABLE
5.14.-Frequency Parameters u a 2 q D cos2 Q! f o r a C-F-F-F Parallelogram; a/b=l; v=O.S
I 2 3 4 5 6 7 8 9
-~~-~
3. 48 8. 52 21. 3 27. 2 31. 1 54. 3 61. 4 64. 3 71. 3
3. 46 8. 48 21. 3 26. 8 31. 2 53. 6 61. 3 64. 3 71. 6
3. 42 8. 36 21. 1 26. 0 31. 6 51. 6 60. 9 64. 0 72. 3
3. 36 8. 16 20. 8 24. 7 31. 9 48. 8 60. 3 63. 0 73. 4
3. 25 7. 91 20. 4 23. 1 32. 1 45. 6 59. 2 61. 4 74 0
3. 12 7. 60 19. 8 21. 4 32. 1 42. I 57. 7 59. 2 70. 5
2. 96 7. 24 19. 1 19. 6 31. 8 38. 7 55. 2 56. 3 66. 5
2. 76 6. 87 17. 8 18. 4 31. 2 35. 3 51. 2 53. 4 63. 7
TABLE
6.BG.-Theoretical and Experimental Cyclic Frequencies jor C-F- F-F Parallelogram Plates;
Material, 65 S Aluminum Alloy
PARALLELOGRAM PLATES 173
12
OB
04 2s
-04
FIRST MODE
6 0 15
08 008
SECOND MODE
. os
B
2 0
Theoretical Node Line D
-ow
-0a -0 08
-1 so I I I I i-016
Theoretical Node Line 02 04 06 08 10
10 Spn PDItM $1.
THIRD MODE
FIGURE5.14.--Theoretical and experimental mode
shapes for GF-F-F parallelogram plates; material,
65 S aluminum alloy. (a) Fundamental mode;
ru=30°. (b) First overtone mode; a=30°. ( e )
Second overtone mode; a= 30'. (d) Fundamental
mode; 0 ( = 4 5 O . (e) First overtone mode; a=45O.
FIGURE
5.13.-Mode shapes for a GF-F-F parallelo- cf) Second overtone mode; a=45'. ( 9 ) Fundamen-
gram plate; a=45O; a=b; v= 0 . 3 . (After refs. 5.15 tal mode; a=603. ( k ) First overtone mode; a=60°.
and 5.16) (i)Second overtone mode; a=60'.
174 VIBRATION OF PLATES
FIGURE
5.14-Concluded.
PARALLELOGRAM PLATES 175
a Dc!pees
60
55
50
45
37 5
W
I5
0
0 I 2 3 4 5 6
c /a
a Degrees
60
55
50
45
315
W
15
0
K K
0 I 2 3 4 5 6
c/o
a Degrees
a Oeqrees 60
:0 55
60
Jo
55 45
08 37.5
so
48 0
15
37 5
30
06
K 8 K
04
02
0
0 I 2 3 4 5 6
176 VIBRATION OF PLATES
CLAMPEDEDGE 7)
-0.4c -02 0 +0:2 +0.4
E4 3
SHAKER POSITION-
NO DEFLECTION DATA
+G.80
.4 0.60
+040
I
+020$
PARALLELOGRAM PLATES 177
FIGURE5.16.--Experimentally determined cyclic fre-
quencies, nodal patterns, and mode shapes for a
6-F-F-F rhombic plate; a= 15O; material, 6061-
T6 aluminum alloy W inch thick. (a) Experimental
node lines and data points. ( b ) Mode 1; fi=76.6
cps. (c) Mode 2; fz=179 cps. ( d ) Mode 3; fa=469
cps. (e) Mode 4; f4=566 cps.
@SHAKER POSITION-
NO DEFLECTION DATA
id)
O W A K E R POSITIW-
NO DEFLECTION DATA
178 VIBRATION OF PLATES
C W P E D EDGE +1.m
+O.K
+am
4 0.40
eo. I 2
C0.31
-020
B
p0.s M
-0.38 - 0.19 0 7) +0.19 1-0.38
k 0 SHAKER PDSITION-
E50.i2&
A
r
\ - 0.40 0 No DEFLECTION DATA
Y
N
p0.63
(e.31 -0.W
-0.80
(=0.5
(=0.69
0 SHAKER POSITION-
PARALLELOGRU PLATES 1179
0 SHAKER POSITION-
-ompI NO DEFLECTION DATA
C W E D EDGE
14
-as
e0 14
-018 0 ’ SO18 to 36
PO 32
PO 5
0 SHAKER POSITION-
NO DEFLECllON DATA
i
-080
PARALLELOGRAM PLATES 181
FIGURE
5.18.-ExperimentaUy determined cyclic fre-
quencies, nodal patterns, and mode shapes for a
GF-F-F rhombic plate; a= 45'; material, 6061-2'6
aluminum alloy % inch thick. (a) Experimental
node lines and data points. (b) Mode 1; jl=97.4
cps. (e) Mode 2; f2=231 cps. ( d ) Mode 3; f3=560
eps. (e) Mode 4; j4=669 eps.
182 VIBRATION OF PLATES
CLAMPEDEDGE
FIGURE
5.19.-Experimentally determined cyclic frequencies, nodal patterns, and mode shapes for a 6-F-F-F
rhombic plate; a=60'; material, 6061-3'6 aluminum alloy inch thick. (a) Experimental node lines and
data points. ( b ) Mode 1; .f1=97 cps. (c) Mode 2; f2=305 cps. (d) Mode 3; f3=570 cps.
PARALLELOGRAM PLATES 183
n
& I
TABLE
5.18.-De$ections for First 3 Modes of Plate 1
---
0.1 1 0.3 0.5 0.7 0.9 1.0
j 0.00
.25
0.316
.338
0. 547
.
569
0. 800
.817 0.952
1s
I .50 .048 . 185 .
360 569
~ .840 .966
.75 .053 .zoo .383 .608 . 856 .983
1. 00 . 056 .225 . 406 .
631 . 875 1.000
. 00 -.088 -. 361 -. 579 -.485 -.207 -. 055
. 25 -.080 -.289 -. 407 -. 297 -. 008 . 117
2 b . 50 -.069 -. 210 -. 260 -. 106 . 180 . 310
.75 -. 044 -. 135 -. 120 . 100 . 386 .524
1. 00 -. 014 -.062 .014 331
~ . 758 1.000
. 00 .099 .
106 . 162 401
e . 788 1.000
.25 .021 -. 035 -. 021 . 116 . 412 . 654
3c .50 -. 042 -. 190 -. 225 -. 085 ~ 190 .352
.75 -. 099 -. 345 -. 451 -. 338 -.099 . 085
1 1.00 - 1155 -. 556 -. 831 --.148 -. 313 -. 162
PARALLELOGRAM PLATES 185
TABLE
5.19.-DeJtections for First 3 Modes of Plate 2
1.
0.00 0.011
.015
.018
.025
I 0. 052
. 067
.080
. 098
0.472
. 509
. 635
.778
.030 . 118 1.000
TABLE
5.2O.-DeJlections for First 3 Modes of Plate 3
1"
I 0:;;
.50
0.012
.015
.022
0.040
.058
.091
0.085
.I16
.
171
0.380
.475
.620
.75 .040 .135 .272 .800
1. 00 .062 .211 .400 1.000
The case of a cantilevered parallelogram vibration for a=3Oo, 45", and 60' wit
with an added mass at the tip is discussed in without the tip mass.
reference 5.22. An aluminum-alloy plate hav-
ing dimensions a=30 inches, c=10 inches, TABLE5.22.-6yclic Frequencies for a 6-F-
b=% inch and having a total mass of 0.0468 F-F Parallelogram Plate With Added Tip
Ib-sec2/in. is loaded by a mass at the tip Mass; Material, 65 S Aluminum Alby
(,$=a, q=b/2, in terms of fig. 5.10) which has
the following inertial properties : mass= 0.0330 Gyciic frequency, cpe, for values of
Ib-sec2/in.,le= 6.483 lb-in.-sec2, &=0. 1242 lb-in.- skew angle, a, deg, of-
Mode
sec2. The mass moments of inertia IO and I+
are about axes in the #- and q-directions, re- 0 30 45 60
spectively. These axes pa& through ,$=a an - ~ -
s = b / 2 . The first three theoretical frequencies B _ _ _ _ _ _ _ _ _ _ _ _ _ _ 11. 33 12.40 10.91 15. 35
for a=Oo, 30°, 45", and 60" are given in table 2 _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ 23. 07 27. 39 40.83
5.22. I n figure 5.26 are shown the nodal lines 3 _ _ _ _ _ _ - _ - _ _101.7 _ _ _ 114.9 111.1 153.5
for the fundamental and second modes of
PARALLELOGRAM PLATES 187
-I"#
FIGURE5.22.--First three mode shapes and frequencies for a C-F-F-F plate; c~=45';
material, magnesium. (a) Mode 1; j c = 3 8 cps. (b) Mode 2; fz= 184 cps. (e) Mode 3 ;
j3=263 CPS.
190 VIBRATION OF PLATES
I .O
FIGURE
5.23.--First three mode shapes and frequencies for a C-F-F-F plate; a=60'; material, magne-
sium. (a) Mod? 1;f1=.47cps. (b) Mode 2: f2=207 eps. (c) Mode 3; j3=380 cps.
PARALLELOGRAM PLATES 19 1
- Fundomentol
---- First Overtone
FIGUEE
5.25.-Nodal patterns of F-F-F-F pardldo-
g a m plates. (After ref. 5.21)
192 VIBRATION OF PLATES
era
6.1 TRAPEZOIDS
6.1 .1 All Edges Simply Supported
The problem of the trapezoidal plate simply
supported all around ( S M S - S S ) (see
fig. 6.1) was solved by Klein (ref. 6.1) by using
the collocation method for the case a1=az=a.
A fnnction
a
- ~
10.11 10.09
10.96 11.177
13. 4 14. 311
19. 7 24. 7
6.4.-C-F-F-F
FIGURE trapezoidal plate.
FIGURE 6.3.-Fundamental frequency parameter
-A=-> W B J D -
where b = ( b , + b J / 2 , against blla for
+2
an isosceles SS-SS-SSSS trapezoidal plate. (After
ref. 6.1)
quency parameters of isosceles trapezoids (aI=
ff2=d:
10" 9
' 8' 75" 7" 6 5" 6"
E O 01 02 025 03 035 04
~~~ 1-0 I-I 9-2 1-3 1-4 1-5 1-6
Mode 1
FIQURE6.6.-C!--F-F-F trapezoidal plate configura-
Upper bound 1 Lower
tions. ( a ) Series I plates. ( b ) Series I1 plates.
(c) Series 111 plates. (After ref. 6.7)
Beam Polyno-
1 bound
functions mial TABLE
6.3.-Mate&l Properties for 3 Series of
Trapezoidal Cantilever Plates; v=O.29
7. 152 7. 163 6. 880 21. 209
8. 465 8. 150 8. 042 23. 996 Modulus of elasticity, psi
13.121 12. 291 18. 160 26. 625 Series Thickness,
In.
18. 397 30. 965 x-direction y-direction
TABLE
6.4.-Experimentally Measured Cyclic Frequencies for C-F-F-F Trapezoidal Plates; v=O.29
1 2 5 6
3 1 4
Nodal patterns corresponding to most of the (mln) identify the number of nodal lines ap-
frequencies of table 6.4 are shown in figures proximately parallel to the x- and y-directions,
6.7, 6.8, and 6.9. Plate designations are respectively. Modes having double labels (e.g.,
shown on the fundamental modes and refer to (0/1)+(2/0)) can be thought of as being the
those of figure 6.6. superposition of two simple modes, each of the
Weiba (ref. 6.8) experimentally determined designated label. The variation in frequency
frequencies and mode shapes for 12 trapezoidal with tan cyI is shown in Ggure 6.11. It is seen
plates of various aspect ratios and having that this choice of parameters yields small
cu1=150, 30°, and 45' and az=O (fig. 6.10). variations. Frequencies for a I = Qfor the rec-
Aspect ratios of 2.0, 1.6, 1.2, and 0.8 were used, tangle, as well as nodal patterns, are listed in
where the aspect rstio 3 4 u / ( b l+bz). The plates section 4.3.12.
were made of %-inch-thick steel. Cyclic fre-
quencies and nodal patterns for the first six .2 OTHER ~ ~ A ~ R ~ ~ AOF
~ EGE
R A ~ S
modes of each plate are shown in figure 6.10. ERAL SHAPE
Planform dimensions are given on the funda- No published results exist for quadrilaterals
mental mode in each case. The mode labels of general shape.
198 VIBRATION OF PLATES
FIQURE
6.8.-?+odal p a t t h n s for aeries 11 plates; v = O . ~ . (From ref. 6.7)
OTHER QUADRILATERAL PLATES 199
FIGURE 6.9.--Nodal patterns for series I11 plates; v=0.3. (From ref. 6.7)
I -
-__ --- -
/----
__c-
.- . ^ . _.^ . _ . _
3084.37Q-7-Id
202 VIBRATION O F PLATES
4
700
600
50 (310
40
30
100
(O/O)
0 0.2 0.4 0.6 0.8 1.0 tana,
(b)
RQURE
6.Il.-Variation of frequency (cps) with tangent of sweep angle for a trapezoidal GF-F-F plate; material,
steel. (a)Aspect ratio= 2.0. (b) Aspect ratio= 1.6. (e) Aspect ratio= 1.2. (d) Aspect ratio=0.8.
OTHER QUADRILATERAL PLATES 203
REFERENCES 6.5. HATO,T.: On the Upper and Lower Bounds of
Eigenvalues. J. Phys. SOC. Japan, vol. 4, no. 1,
6.1. KLEIN, B.: Vibration of Simply Supported Isos- 1949, pp. 334-339.
celes Trapezoidal Flat Plates. J. Acoust. SOC.
6.6. TEMPLE, G.: The Accuracy of Rayleigh’s Method
Am., vol. 27, no. 6, Nov. 1955, pp. 1059-1060.
6.2. TSYDZIK, P. V.: Application of the Perturbation of Calculating the Natural Frequencies of
Method To Solve Problems of the Natural Vibrating Systems. Proc. Roy. SOC. (London),
Vibrations of Nearly Rectangular Plates. Prikl. ser. A, vol. 211, 1952, pp. 204-224.
Mat. Mekh., vol. 16, no. 3, 1952. (In Russian.) 6.7. GWSTAFSON, P. N.; STOKEY, W. F.; A N D ZOROWSKI,
6.3. REIPERT, ZBIQNIEW: Application of Simple Func- 6. F.: The Effect of Tip Removal on the
tional Series to the Solution of Problems Con- Natural Vibrations of Uniform Cantilevered
cerning Statics, Stability, and Vibration of Triangular Plates. J. Aeron. Sci., vol. 21, no. 9,
Plates Having Non-Typical Forms. Arch.
Sept. 1954, pp. 621-633, 648.
Mech. Stos., vol. 6, no. 15, 1963, pp. 791-815.
6.4. NAOARAJA, J. V.: Effect of Tip Removal Upon the 6.8. HEIBA, A. E.: Vibration Characteristics of a
Frequency of Natural Vibrations of Triangular Cantilever R a t e With Sweptback Leading
Plates. J. Sci. Ind. Res.India, vol. ZOB, no. 5, Edge. Rept. No. 82, Cranfield College of
May 1961, pp. 193-197. Aeron., Oct. 1954.
Chapter 7
FIGURE
7.1.--C-C-C triangular plate.
Cox and Klein (ref. 7.1) took a deflection choice of coordinates, + can almost always be
function kept less than 25'. The mode shape compo-
nents arising from equation (7.2) are shown in
figure 7.3.
The results were also checked in reference
7.2 €or the case when cp=O and the triangle is
( m = I , 3 . . .) (7.2) equilateral by using the finite difference method.
he two triangular meshes shown in figure 7.4
where AI and Az are undetermined constants. ere used. For the fundamental mode, only
Equation (7.2) satisfies the boundary condi- one sextile of the triangle is required; this
tions exactly. Equation (7.1) was satisfied a t results in independent deflections of one point
the two points E=c/2 and 2eI3 and q = Q ; this in figure 7.4(a) and eight points in figure
yielded a second-order characteristic determi- 7.4(b). Results from using these two meshes
nant. Fundamental frequency parameters are and the extrapolation formula (eq. 4.90)) are
shown in figure 7.2 for +-0" and 25". As dis- given in table 7.1.
cussed later in this section, the h i t i r i g case ns I n reference 7.3 the solution for the rhombus
2 c / 6 4 0 is wc'%/qD=22.4, an exact solution, given in reference 7.4 (see discussion on the
which indicates a lack of accuracy for small C-6-@-C rhombic plate, see. 5.1.1) is extended
values of 2c/b in figure 7 . 2 . According to t o yield the solutior, for the isosceles triangle
reference '?.I, the results are not sufficiently damped di around. Fundamental frequency
accurate for use when 4 > 2 5 O , but, by suitable parameters wEZ3/ for C Y ~ = C where
Y~=C I Y
is ,
205
206 VIBRATION OF PLATES
(b)
IO0 500
400
300
200
100
2a
FIGURE?.B.--Theoretical and experimental funda- 0 30' 60' 90" 120' 150'
mental frequency parameters for C-6-C and 6-6-58 2Q
isosceles triangular plates. (After ref. 7.3) FIGURE7.&--Fundamental frequency parameterti
wP.\l= €or G C - 6 and C-C-SS isosceles triangular
plates. (After ref. 7.3)
ein (ref. ?.S> solved the problem
by using the collocation method and the de-
flection function moment Mz along x=a. It satisfies this con-
dition o d y at the midpoint of the side @e., a t
A,sin-+Azsin-++A3sin-
m 2m 3m y=O). The differential equation (eq. (1.4))
a a a was satisfied at the three points ( 4 2 , 0),
(2a/3,01, and (3a/4,O),thus giving a third-order
characteristic determinant for the frequencies.
Results for the fundamental frequency param-
on (7.3) satisfies all the eter obtained directly from the collocation pro-
exactly except that fo cedure are shown as the broken curve in figure
208 VIBRATION OF PLATES
250
200
150
100
90
5 ::
60
50
2a
45
FIGURE7.7.-Fundamental frequency parameters
W C ~ C-C-C and C-C-SS
Z for isosceles triangular
plates. (After ref. 7.3) 30
“ 0 0 5 10 15 20 25 30 35 40
2a/ b
2a, deg 30 60 90
______--____
120 ’ 150
FXGIJRE
7.12.--Fundamental frequency parameters for a
C-SS-SS isosceles triangular plate. (After ref. 7.8)
FIGURE
7.13.-GSS-F right triangular plate.
7.1.6 SS-SS-SS
Conway and Farnham (ref. 7.11) solved the FIGURE7.14.-SS-SS-SS triangular plate.
problem for the SS-SS-SS isosceles triangle
( a I = a 2 = a in fig. 7.14) by using the method
employed on the SS-SS-SS-SS rhombus (sec.
5.1.4). Functions given in equation (5.18)
were used and boundary conditions of zero
bending moment were satisfied at N points
along the edge x=a (fig. 7.14). Frequency
parameters arising from various Nth-order char-
acteristic determinants are displayed in table
7.4. For a first-order determinant, the single
point used was at x=a, y=O.
Cox and Klein (ref. 7.2) solved the case of
the isosceles triangle by the collocation method
using a deflection function
fl 2n-x
sin -+A, sin -
a a
w a 2 g D for .determinant of
€312-
deg relating the frequency parameters w c 2m
I
corresponding to the medians of the triangle
which have lengths cl, c2, and c3. Again, the
Ibyl 3by3 curves are inaccurate for small values of 2clb.
177. 69 Solecki (ref. '7.10) gave the frequencies and
97.93
66. 34
mode shapes for the 30'-60'-90' triangle shown
49.45 in figure 7.17. Mode shapes were taken as
39. 48
32. 87
28. 18
24.67
I I
FIGERE
? . ~ ~ . - ~ ~ ~ a ~a ~ e ~~ tparametem
a ~ e for~ ~ e ~ ~ ~
a SS-SS-SS triangular plate. (After ref. 7.12) FIGURE
?.1?.-3O0-6Q0-9O0 SS-SS-SS ~ r i ~ ~ plate.
g ~ l a ~
212 VIBRATION OF PLATES
in terms of figure 7.17. Corresponding fre- a vibrating membrane and a simply supported
quencies are found from substituting equation polygonal plate (see the chapter entitled
(7.7) into equation (1.4), giving “Plates of Other Shapes” (ch. 8)).
7.1.7 SS-SS-F
There are no specific solutions of the problem
of SS-SS-F triangular plates. Wes tmann (ref.
(m=2, 3 , . . .;n=l, 2 , 3 , . . ,; m>n) (7.8) 7.7) proposed obtaining bounds from SS-SS-F
sectorial plates. (See sec. 7.1.3.)
Thus the fundamental frequency is found from
7.1.8 C-F-F
equation (7.8) to be ~,~a~,&@=92.113.
Consider first the symmetric cantilevered
This was also found in reference 7.13 by triangle depicted in figure 7.19. Andersen
using the solution for the SS-SS-SS-SS rec-
(refs. 7.19 and 7.20) solved the problem by
tangle and the method of images. Nodal
using the Rayleigh-Ritz method and the
patterns for the first six modes are shown in
triangular u-v coordinates shown in figure 7.19
figure 7.18. The case of the 3O0-6Oo-9O0
(see also the discussion for the C-F-F-F
triangle is also discussed in reference 7.14.
Schaefer and Havers (ref. 7.15) found the trapezoidal plate, sec. 6.1.2). For symmetric
fundamental frequency of the equilateral tri- modes, the four-term series
angle of altitude a t o be wa2,i,lD=39.478.
The problem was also solved by Conway by
analogy in reference 7.16 and by the point-
matching method in reference 7.11. The was used, and for antisymmetric modes the
problem is also solved in references 7.17 and 7.18. series
The case when C Y I = C Y Z = ~ ~ ~(fig. 7.14) can be
deduced from the higher mode shapes of a
SS-SS-SS-SS square plate. The fundamental
frequency parameter is wa2WD=24.674.
The case when al=az=600 was examined by
Seth (ref. 7.171, who gave a fundamental fre-
quency parameter of wa2&@= 17.272.
Much more information is available for this
problem from an analogy that exists between
FIGURE
7.18.-Nodal patterns for a 3Q0-6Oo-9Q0SS-
SS-SS triangdar plate. FIGURE
?.1S.-Symmetric 6-F-F triangular plate.
TRIANGULAR PLATES 213
was used, where and c $ ~ represent the f i s t P
two modes of a cantilever beam free at u=O
and clamped a t u=l. (See discussion of the
rectangular cantilever beam, sec. 4.3.12.) The
functions $J3 and $J4 represent the first symmetric
and antisymmetric modes, respectively, of a
beam free at v = + I . The expression for the
strain energy in triangular coordinates is
given in equation (6.4). Integration was
performed numerically. Frequency param-
eters, nodal patterns, and amplitude coeBcients
for the first four modes and several a/b ratios
are given in table 7.5. Poisson’s ratio is 0.3.
Variation of frequency parameter with a/b
ratio for the two antisymmetric modes is
shown in figure 7.20. It is seen that the
frequency parameters increase linearly with
a/b, as was the case for the C-F-F-3’
rectangle (sec. 4.3.12). Frequency variations
for the first two modes are shown in figure 7.21
where the frequency parameters wa21/12p/Eh3
obtained from beam theory are also plotted 0
0 I 2 3 4 5 6
as horizontal broken lines. It must be re- o/b
membered that the plate and beam frequency
FIGURE7’.2O.--Variation in antisymmetric frequency
parameters differ by the factor l - v 2 . Thus, parameters with a/b for a C-F-F symmetric triangular
when Poisson’s ratio is considered, the plate plate; v=Q.3. (After ref. 1.20)
frequencies themselves are slightly higher than
those predicted by beam theory.
DuEn, Gustafson, and Warner (ref. 7.21)
also used the Rayleigh-Ritz method to analyze
the triangular plate of symmetric shape. A
partial summary of deflection functions used
and frequency parameters obtained is given
in table 7.6, where the notation used is that of
figure 7.19 and v=1/4. Because modes 1 and
2 are symmetric and antisymmetric, respec-
tively, the frequency parameters listed for
these modes are guaranteed to be upper bounds
OR the exact frequencies, and improvement in
bounds with the various functions used is
clearly indicated in the table. Further results
were obtained which showed the variation in
fundamental frequency parameter and mode
shape with a/b ratio and Poisson’s ratio
using the deflection function.
Amplitude
Nodal lines coefficient
30.803 II 30.718 1
I
30.654 1
I
30.638
2
-0.77460 -0.76682 -0.76427 -0.76368
1. 00000 1.00000 1.00000 1.00000
-. 02305 . 00527 .00208 .00073
. 04645 * 01022 .00241 .00077
oa%?D
3
1.64125 0.60941 0.33684 0.27432
1.00000 1.00000 1.00000 1.00008
.00581 ,00155 .00038 .00012
-. 00380 -. 00079 -. 00019 -. 00006
4
1.00000 1.00000 1.00000 1.00000
-. 32893 -. 31823 -. 31430 -. 31330
-. 00808 -. 00156 -. 00036 -. 00012
.00586 .00122 .00029 .00009
TRIANGULAR PLATES 215
TABLE
7,6.-De$ection Functions and Frequency Parametersfor a Rayleigh-Ritz Analysis of a C-F-F
Symmetric Triangdar Plate; v = l f 4
5- I)a(z+A,)
a a A3=0.462 23. 0
2-
a I)&4,) Ad=49/164 37. 1
z-
a I)'(:+As) Aa=5/3 7. 15
Kumaraswamy and Cadambe (ref. 7.22) Further experimental results from reference
experimentally determined the first 18 modes 1.23 for a/b=l are given later in this section.
and frequencies of a symmetric triangular Consider next the delta cantilever plate
cantilever plate made of commercial mild steel. depicted in figure 7.23. This problem was
Pertinent dimensions and physical constants solved in reference 7.20 for the first two modes
were: a=6.00 inches, b=6.00 inches, h=0.0895 by the method described earlier in this section.
inch, pg=O.282 pound per cubic inch, length- The following six-term series was used for the
wise E=29,83X IO6 psi, breadthwise E=29.18 deflection function :
X10' psi, and v=0.29 (assumed). Cyclic fre-
quencies and frequency parameters are given
in table 7.8. The disagreement in values ob
w a z d a between tables 7.5 and 7.8 for afb=l
is readily apparent. Nodal patterns are shown Frequency parameters, no al patterns, and
i~ figure 7.22. mnplitude coeEcients are listed in table 7.9 for
216 VIBRATION O F P L A T E S
6499 272. 4
6526 273. 5
6884 288. 5
7627 319.7
8498 356. 4.
9875 413. 8
TABLE
7.9.-Frequency Parameters, Nodal Patterns, and Amplitude Goeflcients for a G-F-F Right
Triangular Plate; v=0.3
w a 4 D
X
a
IO" IO"
Designation AI A2 A3 A4 A5
t
FIGURE
7.24.--C-F-F delta configurations. (After ref.
FPGURE
7.23.--C-F-F right triangular plate. 7.23)
3054337 0-70-15
218 VIBRATION OF PLATES
7. 15
6. 37
7. 16
6. 57
7. 05
28. 0
26. 9
57. 5
54. 3
TABLE
7.Il.-Frequency Parameters w a r n (M, Total Mass o j Plate) and Mode Shapes for Q
C-F-F 45' Delta Triangular Plate; v=O.S
2 1 3 1 4 1 5 8 7 1 8 9 I 10
-I i
4. 35 16.76 23.01 38.90 53.65 60.32 107. 1 1 148. 6
1 1 1 1 1 1 1 1
.65 .29 -. 27 .07 -.74 -1.01 -1.15 -1.02 -1.32 -1.60
..56 -. 94 .45 -3.62 18 1. 39 1. 86 -1.03 1. 64 3. 70
.33 -. 05 -. 81 .39 -- 06 .04 1. 36 1. 10 2. 32 5. 08
.28 -. 78 -. 07 -. 31 . 04 -. 002 -2.08 1.74 -1.65 -6. 98
.20 -1.33 .43 1. 88 . 03 -1.99 4.38 -. 37 -. 65 5. 54
.10 -. 05 -. 47 .44 .55 .90 3. 05 --I. 66 -1.05 -5.48
.08 -. 32 -. PI .49 .47 .27 -3.77 -.70 -. 43 8. 62
. 06 -. 45 . 12 1. 24 -. 19 .27 -1.92 -1.17 2.76 -6. 00
. 02 -. 31 . 14 1.78 -1. 28 2. 65 .84 1.18 -1.99 2. 89
-
TRIANGULAR PLATES 219
Picswa~7.25.--Experimentally observed cyclic frequencies, cps, and nodal patterns for C-F-F delta triangular
steel plates. (From ref. 7.23)
1300
1200
I100
1000
900
EO0
iL
8 700
=.
2u 600
E 500
400
300
200
100
0
0 2 4 6 8 10 12 14 16 18 20 22
blo
Kawashima (ref. 7.28) used the finite differ- The delta plate for the cases b/a=l and 2
ence method to obtain frequencies and mode was a h analyzed by Walton (ref. 7.29) by
shapes for delta plates (fig. 7.23) having using the method of reference 7.30 which re-
b/a=1 and 2. The 45' delta was analyzed by places the derivatives in the strain energy
integral by finite differences. Twenty-eight
free grid points were used in the analysis.
Frequencies were computed and compared
with experimental data for s
having the dimensions a= 1
inches and a=10 inches, a
more accurate Both plates were 0.061 inch thick and v and E
were taken as 0.025 and 30X106 psi, respec-
tively. The first six cyclic €re
plate are given in table 7.13.
clamped boundary, for the five higher modes o
The first gives the d depicted in figures 7.30 and 7.31.
Hanson and Tuovila (ref. 7.31) experimen-
FIGURE
7.28.--Nodai patterns for a GF-F 45' delta
triangular plate; material, steel. (After ref. 7.25)
0 -0459 0449 1595 2.992
TABLE
7.12.--Theoretieal Cyclic Frequencies .for SECOND MODE
GC-F Delta Triangular Steel Plates
Mode 1
2 (6 grid
points)
6 grid 10 grid
points points
I
0
.979
on three 6061-T6 duminum plates X inch P 32. 8 32. 8 1. 00
thick and having the dimensions a=7.5 inches, 2 89. 9 91. 0 .988
3 164 164 1. 00
b=7.5 inches; a=12.5 inches, b=7.5 inches; 4 175 181 . 967
and a=15 inches, b=7.5 inches. Cyclic he- 5 263 283 . 929
quencies, nodal patterns, and mode shapes me 6 328 348 . 943
given in figures 7.34, 7.35, and 7.36.
Mode 2 Mode 3
,
Mode 2 Mode 3 Mode 4
Mode 4 Mode 5
Experimenfoi
Calculated
- Experimental
A Calculoled
Mode 6
Mode 5 Mode 6
Fraom 7.30.-Nodal. patterns for a 6-F--F delta 7.3I.-Nodal patterns for a C-F-F delta tri-
FIGUEE
triangular steel plate, bja= 1. angular steel plate, b/a= 2.
222 VIBRATION OF PLATES
\
---- Heavy sold lines lndlcate
0
.
plate 01 no
FIGURE
7.32.-Experimental frequencies and mode shapes for a 60" delta cantilever plate; material, magnesium.
(u) Mode 1,Jl=50 cps. ( b ) Mode 2,f2=184 eps. ( c ) Mode 3,fa=258 cps.
TRIANGULAR PLATES 223
FIGWBE
7.33.--Experimental frequencies and node shapes for a 45" delta cantilever plate; material, magnesium.
(a) Mode 1,j1=66 cps. ( b ) Mode 2, f2= I85 cps. ( c ) Mode 3, f3=336 cps.
224 VIBRATION OF PLATES
4
CLAMPED
EDGE
1 SHAKER POSITION
FOR ALL MODES
-0.67 I
4
+ 0.4
I I
-0.4 -0.2 6 +0.2 f0.4
71
(C)
"Mi
+ I201
-0
+I 20
""T
+' "t
(f) -160
+0401
1
-Iml
FIGURE ?.36.--Experimental data for a 28.8' delta cantilever plate; material, BO61-TB
aluminum ?4 inch thick. ( a ) Experimental node lines and data points; jl=11.2 cp5;J~=300
cp5; f3=508 cps. ( b ) Normalized deflection; mode I;fl=91.2 cps. (6) Normalized deflection;
mode 2; fi= 300 cps. (d)Normalized deflection; mode 3; j3= 508 cps.
TRIANGULAR PLATES 227
Consider finally the triangular cantilever 101.3', and 126.6'. Results for the first six
plate of general shape as shown in figure 7.37. modes are shown in figure 7.39.
In reference 7.29 this problem was also (see mein (ref. 7.34) proposed a set of empirical
discussion earlier in this subsection) solved ana- formulas for the prediction of frequencies of the
lytically for a sheet steel plate having dimen- first three bending modes and the first torsional
sions a= 10 inches, b = l O inches, /32= 116.6', mode for arbitrarily shaped triangles. These
and h=0.061 inch. Material constants were formulas are given in table 7.15.
taken as ~=0.250a 30X loe psi. Thirty- The planform dimensions used on both sides
one grid points wer n the analysis. The- of the formulas in table 7.15 are those of
oretical frequencies are compared with experi- figure 7.40. Substantiation of the formulas of
mental ones in table 7.14. Nodal patterns for table 7.15 was given in reference 7.34 by com-
the five higher modes are depicted in figure parison with the experimental results of refer-
7.38. ence 7.23. These data are reproduced in table
Frequencies and nodal patterns were found 7.16. The plate designations used are those
experimentally in reference 7.23 for sheet shown in figures 7.25 and 7.39.
steel plates having a= 10.0 inches, b= 10.0 The vibration of C-F-F triangular plates is
inches, h=0.061 inch, and &=63.4', 78.7', go', also discussed in references 7.35 and 7.36.
27. 6 26. 3 I. 05
107 101 1. 06
173 171 1. 01
262 259 1. 01
352 346 1. 02
480 522 ~ 92
FIQURE
?.38.--Nodal patterns for a C-F--F triangular
steel plate.
228 VIBRATION O F P L A T E S
FIGURE7.39.-Experimentaily observed cyclic frequencies, cps! and nodal patterns for C-F-F triangular steel
plates. (From ref. 7.23)
Mode
i Frequency formula
!
35. 3 34. 5 140 136 325 325 192 190
s4- - -_-- - - - - - - _ -
I 31. 6 32. 4 121 120 293 293 179 182
S5 - - __-- ---_ - - --- 26.6 26. 3
98 I 101 255 259 166 171
where x and y are as shown in figure 7.41 and BRTS AND CONDIT
the beam functions apply to beams of length a. The problem of a, simply supported 30'-60'-
Cyclic frequencies were obtained for a brass 90' triangular plate with an internal point
230 VIBRATION OF PLATES
Cyclic frequency, cps, for values FIQURE7.42.-Nodal patterns for a F-F-F 45' right
of R of- triangular plate; material, brass. (From ref. 7.37)
ROTATIONAL
CONSTRAINT POINT
I
I
?*IN.
= 0.9
SHAKER POSITION
FOR ALL MODES
FIGURE
7.46.--Gyclie frequencies and nodal patterns
for a a isosceles right triangular plate with hub-pin
supports; material, 6061-T6 aluminum 36 inch
thick. fi= 76.9 cps; j2= 297 cps; f3= 390 eps; ,f4= 848
cps.
PHEURE
7.45.-SS-SS-SS 30°-600-900 triangular plate
with internal point support.
232 VIBRATION OF PLATES
4- 1.20
+OB0
(=0.6
- 2
&I - B .201
FKXTEE
7.4?.--Normdieed deflections of a 45" triangular hub-pin plate; material, 6061--T6aluminum. (a) Mode
f1=76.9 cps. (b) Mode 2;f2=297 cps. (c) Mode 3;f3=390 cps. (dj Mode 4; J,=841 eps.
TRIANGULAR PLATES 233
Is?
Z
VIJjKAllON
234
REFERENCES 7.15. SCHAEFER,H.; AND HAVERS,A.: Die Eigen-
schwingungen der in ihrer Ebene allseitig
7.1. Cox, H. L.; AND KLEIN, B.: Fundamental Fre- gleichmassig belasteten gleichseitigen Dreiecks-
quencies of Clamped Triangular Plates. J. platte. Ingr.-Arch., Bd. 7, 1936, pp. 83-87.
Acoust. SOC.Am., vol. 27, no. 2, Mar. 1955, 7.16. CONWAY, H. D.: Analogies Between the Buckling
pp. 266-268. and Vibration of Polygonal Plates and Mem-
branes. Can. Aeron. J., vol. 6, no. 7, Sept.
7.2. Cox, H.; AND KLEIN,€3.: Vibrations of Isosceles
Triangular Plates. ZAMP, vol. 6, 1955, pp. 1960, p. 263.
68-75. 7.27. SETH, B. R.: Transverse Vibrations of Recti-
linear Plates. Proc. Indian Acad. Sci., sec.
7.3. OTA, T.; HAMADA, M.; A N D TARUMOTO, T.:
A, vol. 25, Jan. 1947, pp. 25-29.
Fundamental Frequency of an Isosceles-
Triangular Plate. Bull. JSME, vol. 4, no. 15, 7.18. SEN, 3.: Note on Some Two-Dimensional Prob-
lems of Elasticity Connected With Plates
Aug. 1961, pp. 478-481.
Having Triangular Boundaries. Bull. Calcutta
7.4. HAMADA, M.: Compressive or Shearing Buckling Math. Soe., vol. 26, no. 2, 1934, pp. 65-72.
Load and Fundamental Frequency of a 7.19. ANDERSEN, 3. W.: Ph. D. thesis, Univ. Illinois,
Rhomboidal Plate With All Edges Clamped. 1953.
Bull. JSME, vol. 2, no. 8, Nov. 1959, pp. 7.20. ANDERSEN,B. W.: Vibration of Triangular
520-526. Cantilever Plates by the Ritz Method. J.
7.5. KERSCH,J.: Une Mbthode pour I'fivaluation par Appl. Mech., vol. 21, no. 4, Dec. 1954, pp.
Dbfaut de la Premihre Valeur Propre de la 365-376.
Vibration ou du Flambage des Plaques En- 7.21. DUFFIN,E. J.; GUSTAFSON, P. N.; AND WARNER,
castrbes. Bull. Acad. Sci. (Paris), SBanee du W. H.: Natural Vibrations of Cantilevered
13 June, 1960, pp. 3943-3945. Triangular Plates, I. Rept. CIT-AF8-TR22,
7.6. Cox, H. ; A N D KLEIN,3 . : Buckling and Vibration Carnegie Inst. Tech., Sept. 1952.
of Isosceles Triangular Plates Having the Two
7.22. KUMARASWAMY, M. P.; A N D CADAMBE, V.: Ex-
Equal Edges Clamped and the Other Edgc perimental Study of the Vibration of Canti-
Simply-Supported. &,J.Roy. Aeron. SQC.,gwo.. levered Isosceles Triangular PIBtks. J. ' Sei.
59, no. 130, Feb. 1955, pp. 151-152. Pnd. Res. (India), vol. 15B, no. 2, Feb. 1956,
7.7. WESTMANN, R. A.: A Note on Free Vibrations of pp, 54-60.
Triangular and Sector Plates. J. Aerospace 7.23. GUSTAFSON,P. N.; STOKEY, W. F.; and
Sci., vol. 29, no. 9, Sept. 1962, pp. 1139-1140. ZOROWSKI,C. F.: An Experimental Study of
7.8. Cox, H.; A N D KLEIN,B.: Vibrations of Isosceles Natural Vibrations of Cantilevered Triangular
Triangular Plates Having the Base Clamped Plates. J. Aeron. Sci., vol. 20, May 1953,
and Other Edges Simply-Supported. Aeron. pp. 331-337.
Quart., V O ~ . 7, no. 3, Aug. 1956, pp. 221-224. 7.24. NAGARAJA, J. V.; KUMARASWAMY, M. P.; A N D
7.9. YOUNG, D.: Vibration of Rectangular Plates by SUBRAMANXAN, N. R.: On the Vibration Fre-
the Ritz Method. . I .
Appl. Mech., vol. 17, quencies of Skew Cantilever Triangular Plates.
no. 4, Dec. 1950, pp. 448-453. J. Sci. Ind. Res. (India), voi. 203, no. 10, Oct.
7.10. SOLECKI,R.: Free and Forced Vibration of a 1961, pp. 479-482.
Triangular Plate. Rozprawy Inzh. 142, vol. 7.25. CHRISTENSEN, R. M.: Vibration of a 45" Right
8, no. 1, 1960, pp. 65-81. (In Polish.) Triangular Cantilever Plate by a Gridwork
7.11. CONWAY, H. D.; A N D F A R N H A M , K. A.: The Free Method. AIAA J., vof. 1, no. 8, Aug. 1963,
Flexural Vibrations of Triangular, Rhombic pp. 1790-1795.
and Parallelogram Plates and Some Analogies. 7.26. HRENNIKOFF, A.: Solution of Problems of Elas-
Hnt. J. Mech. Sci., voi. 7, 1965, pp. 811-816. ticity by the Framework Method. J. Appl.
7.12. KLEIN, B.: Fundamental Frequencies of Arbi- Mech., vol. 8, Dee. 1941, pp. A-169-A-175.
trarily Shaped Simply-Supported Triangular 7.27. LUBKIN, J. L.; A N D LUKE,Y. L.: Modes and Fre-
Plates. J. Roy. Aeron. SQC.,vol. 60, no. 541, quencies of Wings of Triangular Planform.
Jan. 1956, pp- 63-64. Rept. 56-335, Wright Air Develop. Center,
7.13. PAN, LIH-CWOW:Equilibrium, Buckling and June 1956, pp. 1-46.
Vibration of a 3O0-6O0-9O0 Triangular Plate 7.28. XAWASHIMA, S.: On the Vibration o f Eight
Simply Supported at Its Edges. Acta Phys. Triangular Cantilever Plates. Mem. Fac.
Sinica, vol. 3, no. 12, 1956, pp. 215-245. (In Eng., Kyushi Univ., vol. 18, no. 1, 1958, pp.
Chinese.) 9-21.
7.14. SOLECHI, R.: The General Solution of a Tri- 7.29. WALTON,W. C., JR.: Applications of a General
angular Plate 30"-60"-90" by Means of Eigen- Finite-Diff erence Method for Calculating 3end-
transform. 3ull. Aead. Pol. Sci., Ser. Sei. Tech., ing Deformations of Solid Plates. NASA T N
vol. 8, no. 7, 1960, pp. 325-331. (In English.) D-536, 1960.
TRIANGULAR PLATES 235
7.30. HOUBOLT, J. C.: A Study of Several Aerother- 7.34. KLEIN, B.: Natural Frequencies of Constant
moelastic Problems of Aircraft Structures in Thickness Cantilever Triangular Plates of
High Speed Flight. Ph.D. thesis, %rich Arbitrary Planform. J. Roy. Aeron. SOC.,v01.
E.T.H., 1958. 60, no. .544, Apr. 1956, pp. 281-282.
7.31. HANSON, P. W.; A N D TUOVILA, W.: Experi- 7.35. MARTIN,N. C.; A N D GURSAHANY, H. J.: On the
mentally Determined Natural Vibration Modes Deflection of Swept Cantilevered Surfaces.
of Some Cantilever Wing Flutter Models by J . Aeron. Sci., vol. 18, 1951.
Using an Acceleration Method. NACA T N 7.36. SUZUKI, S.: On the Natural Frequency of Swept
4010, 1957. Cantilevered Rectangular Plate. J. Jap. Soc.
7.32. CRAIG, R. R.; PLASS, I-r. J., JR.; A N D CAUGH- Aeron. Eng., vol. 5, 1957, pp. 50-58. (In
field, D. A.: Experimental Determination of Japanese.)
Frequencies and Mode Shapes of Cantilever 7.37. WALLER,MARYD.: Vibrations of Free Piates:
Isosceles Right-Angled Triangles. Proc. Pbys.
and Hub-Pin Plates. Def. Res. Lab. Rept.
SOC.,V O ~ . 53, pt. 1, 1941, pp. 35-39.
DRG.518, CR-13, Univ. Texas, June 1964. 7.38. WALLER,MARYD.: Vibrations of Free Plates,
7.33. CRAIG, R. R . ; A N D Puss, E.J.: Vibration of Line Symmetry, Corresponding hfodes. Proc.
Nub-Pin Plates. AIAA J., vol. 3, no. 6, June Roy. Soc. (London), ser. A, vol. 211, 1952,
1965, pp. 1177-1178. pp. 265-276.
Chapter 8
es o
FIGURE
8.2.-Nodal patterns of completely free regular pentagonal plates. (From ref. 8.4)
FIGURE
8.3.-Nodsl patterns of completely free regular hexagonal plates. (From ref. 8.4)
PLATES OF OTHER SHAPES 239
FIGURE 8.4.--Nodal patterns of completely free regular octagonal plates. (From ref. 8.4)
8.1.3 Other Polygonal Plates edge. I€ one takes solutions t o equation (1.4)
The fundamental frequency of a regular in the form of equation (2.1) with n = ~ / 2 a ,
octagonal plate of side length a and simply 3 ~ / 2 a., . . (fig. 8.5), satisfaction of the bound-
supported along all edges was computed to be ary conditions along the circular edge yields a
wa24rD=3.624 in reference 8.3. The method second-order characteristic determinant for the fre-
used was that described in the discussion of quencies of symmetric modes. Similarly, the
pentagons (sec. 8.1.1). antisymmetric modes are determined by re-
Experimentally observed nodal patterns for placing cos ne with sin ne where %=a/@,
completely free regular octagonal plates are 3a/a, . . . . I n spite of the relative simplicity
set forth in figure 8.4 (ref. 8.4). of this approach, the only known solutions of
this type are those €or which n is an integer
8.2 SECTORIAL PLATES and which correspond to the higher modes of
Coordinates and dimensions of a circular a circular plate.
sector are shown in figure 8.5. Westmann (ref. 8.7) solved the case when
the circular edge is free by using the Rayleigh
8.2.1 Radial Sides Simply Supported procedure, assuming a deflection function
An exact solution is obtainable for the case
when the two radial edges are simply sup- e)
W(T, =T2 cos n8 03.9)
ported, regardless of the homogeneous bound- and obtained the following approximate for-
ary conditions which exist along the circdar mula for the fundamental frequency parameter:
=3 (n4-2n2+ 8 )-V( Sn*- 8) (8.10)
w2a4p/D
For the case when n=3 (2a=60°), results for
w obtained from equation (8.10) for v=O and
v=1/3 are determined (ref. 8.7) to be 4.8 an
5.5 percent too high, respectively, when com-
pared with an exact solution obtained from the
threefold symmetric mode of a completely free
circular plate (see sec. 2.1.3).
8.2.2 Other Boundary Conditions
Ben-Amoz (ref. 8.8) used the Rayleigh-Ritz
method to solve the problem when all edges are
clamped. A deflection function
W ( T0), =t2(1-En)' COS^ Y ~ O S C O S726)
+C2(co~h7~O-cosr~e)+C~
(sinhrle
240 VIBRATION OF PLATES
2
uo= 3 (m-1)
2
bo= $(m+2) (2+3m+2m2) I (8.12)
+
m 6)(2m-1- 3)
^1
where m is taken as ?ria.
Waller (ref. 8.4) experimentally observed the
two nodal patterns shown in figure 8.7 for a
completely free semicircular plate.
8.3 OTHER PLATES
Grinsted (ref. 8.9) experimentally determined
the frequencies and mode shapes of a flat brass
plate designed to simulate an impeller blade.
The plate was 0.064 inch thick, and the remain-
ing dimensions are given in figure 8.8. Mode FIGURE8.7.-Nodal patterns for a completely free
shapes observed, along with the corresponding semicircular plate. (From ref. 8.4)
cyclic frequencies, are depicted in figure 8.9.
In reference 8.9, experimental results are ing mode shapes are shown in figure 8.11. The
also given for a cantilevered plate of irregular material is mild steel.
shape intended to simulate a marine propeller Ruscoe (ref. 8.10) experimentally found
blade. Dimensions of the plate and cyclic fre- several “complex modes” of a gat plate in the
quencies are given in figure 8.10. Correspond- shape of a turbine vane having a curved edge
PLATES OF OTHER SHAPES 241
Edge
T?<a 7
.,, ? : ,
FIGURE8.9.--Cyelic frequencies and mode shapes for a flat-plate model of an impeller blade. (From ref. 8.9)
242 VIBRATION OF PLATES
7,542
316
4 5 6
FIGURE
8.10.--Plate dimensions and cyclic frequencies for a gat-plate model of a marine propeller
blade. (After ref. 8.9)
PLATES OF OTHER SHAPES 243
5,558 2/3+0/5 6,098 1/5-3/2 4,934 2/3-0/5 6,245 /O 6,517 312 -8/5
es
No work in the literature has been found for The solution to equation (9.3) can be expressed
the case of general anisotropy. Results for as a power series
the special cases of polar and rectangular
orthotropy are summarized in the following (9.4)
sections.
9.1 P O L A R O R T H O T R O P Y as was shown first by Akasaka and Takagishi
The differential equation for the transverse (ref. 9.1) and later in references 9.2, 9.3, and
bending of a polar orthotropic plate is (see the 9.4. Substitution of equation (9.4) leads to a
appendix) : reeursion relationship among the coefficients ( e 3 .
Results exist for circular plates for only two
d4w . D,e b4w +--+22-
Deb4w D
D,- 2- cases of simple edge conditions-when the
W + r2 a r 2 a 2 r4 w r i~
edge is either completely clamped or simply
2Drs b3w Dsb'~ 2 aZW
supported.
r3 , 2 + $ D B + D T 8 ) ~
where
24 5
.-
and where f2=Dr/pa4and k2=Do/Dr. An ap- (9.4)) and obtained the following characteristic
proximate formula for the first two axisym- equat,ion for arbitrary values of n:
metric modes is obtained from equation (9.5) by
truncating the series and is given in reference
9.1 8 5
w2a4p/Dr=24(3+k)
~
5+k
[( 4 f k ) (7+ k)
(9.15)
where
-+(2)2'(2j)!(9-kZ)(49-k2).
(cV)*j . .[(4j-1)"k'] +...
A=rya,
a4=w2p/DI1 k2=Dg/Drl and v g is the elastic responding frequency parameters and nodal
cofistant in the axisymmetric relationship patterns are shown in figure 9.1 for the first
five axisymmetric modes and the first four
nonaxisymmetric modes.
In reference 9.4 the frequency equation is
written as
The primes indicate differentiation with respect
to r .
Axisymmetric frequency parameters for vari-
ous combinations of elastic constants are given
in table 9.1 taken from reference 9.3.
Experimental frequencies were also measured
in reference 9.3 for the plate of table 9.1 having
k=1.50, v,=O.5O1 and 0 , = 1 1 5 0 0 . The cor-
2 3 4 5
= 27.70 =64.93 =116.3 =183.1I
(a)
(b 1
FIGURE
9.l.-Experimentally determined frequency parameters 0 a 2 J n , and nodal patterns for a simply supported
circular plate having polar orthotropy. ( a ) Axisymmetric modes. ( b ) Nonaxisymmetric modes. (After ref. 9.3)
15
o/b
FIGURE
9.2.--Frequency parameters for a clamped-free
annular plate having polar orthotropy.
and where the remaining symbols are as defined earlier in this section. The characteristic deter-
minant giving a is
=o (9.21)
250 VIBRATION OF PLATES
Frequency parameters for varying ratios of alb If the orthotropic constants D:, Di, and
andDB/D, are depicted in figure 9.3. The figure D:, are known with respect to the a? and Y’
is not completely definitive for the same reason coordinate axes, it has been shown (refs. 9.6
8s that given in the preceding paragraph. and 9.7) that the orthotropic constants D,, D,,
and D,,can be determined from
9.2 RECTANGULAR ORTHOTROPY
The differential equation for the transverse D,=D: cos4++DI sin4++2D:, sin2 + cos2 4
bending of a plate having rectangular orthot- D,=D: sin4 ++D; c0s4 ++2D:, sin2+ c0s2+
ropy is (see the appendix) :
Dzu= (30:+30~-220:,) sin29 cos2+
+D:.(COS~ +-sin2 412
(9.25)
The moment-curvature relations are
When the angle + between the 2’- and the
x-axis is a multiple of 22.5’, equations (9.25)
can be used to obtain the equivalent elastic
constants for equation (9.22). For an angle 4
not equal to 22.5”, 45O, 67.5’, . . ., however,
equation (9.22) is transformed into an equation
having terms of the type d4wlbx and d4w/d23
ay as well.
The assumption of simple harmonic motion
Other useful equations are given in the appendix.
The elastic constants are related by (see the w=W(s, y) cos ot (9.26)
appendix)
Dz- E2h3 gives for equation (9.22)
12(1-v,Yy)
D,= Eh3
12 (1-v,v,) (9.24)
(9.27)
The strain energy of bending and twisting
H) Gh3 of a plate having rectangular orthotropy,
k- 12
expressed in rectangular coordinates, is
20 “=z 1
dA (9.28)
b
(9.31)
TABLE
9.2.-Frequency Coe$icients in Equation (9.2929)
Boundary conditions c
I 1 I ! 1
I 1, 2, 3,
: j I, 2, 3, . . .
. .
252 VIBRATION O F PLATES
(9.32)
0 2 3 0 I 2 3
This result was obtained by Wearmon (ref. 7-
9.10) and by many others.
The variation of frequency with a/b ratio
was determined in reference 9.8 for several
higher modes. This variation is depicted in
figure . 9.5 for a five-ply maple-plywood
plate having Dz/Dzy=1.543 and D,/D,,=4.810.
The accuracy of the Rayleigh-Ritz method s I
I t
0 I 2 3 0 1 2 3
as applied to orthotropic plates was studied in a/b
reference 9.10 by solving this problem using
a deflection function FIGURE 9.5.-Frequency parameter wabdplD,,/d for
SS-SS-SS-SS, C-C-SS-SS, and C-GC-C five-ply
W(x,y) =x(a- 2)y ( b-y) (a2 ax- 2 2 ) + maple-plywood rectangular orthotropic plates. (After
+ +
(b2 by -Y2)iA 1 A2z (a-x)y (6 -y) 1 (9.33; rei. 9.8)
where AI and Az are undetermined coefficients. A2=O) and both A, and Az are given in table
The results obtained by taking only A, (Le., 9.3 for five-ply plywood and veneer square
plates of birch with the orthotropic constants
Y determined experimentally.
Extensive experimental results are also given
a
in reference 9.10 for several types of wood ve-
neers and plywoods. In references 9.7 and
9.11, this experimental work is extended t o
study the effect on the frequencies when the
grain of the veneer or plywood is not parallel to
the sides of the plate.
oppmann, Wuffington, and Magness (ref.
9.12) simulated a stiffened plate by taking a
steel plate and milling longitudinal grooves into
FXGURE
9.4.-SS-SS-SS-SS plate. it. In one case, the grooves were on only one
Material
Properties I
-
Dz D, Dz, 1 term 2 terms Exact
-
PIywood - - _ _ _ _ _ _ _ - - - _ _ _ 19.
Veneer. . . . . . . . . . . . . . . . . . . .
_ _ 1x108
-
2. 97
1
7. 1X108
.2H . 69 I
4. 4X108 0. 592QX105 0. 5917X105 0.5916x105
. 2131 .2136 .2135
ANISOTROPIC PLATES 253
side of the plate, and, in the other, they were TABLE9.5.--Cyclic Frequencies for Grooved
on bath sides. The dimensions and spacing of SS-SS-SS-SS Square Plates
the grooves are given in figure 9.6. The plate ~~
tropic constants are set forth in table 9.4. Grooved on 1 side Grooved on both
Mode m/n sides
Nine experimentally measured cyclic frequen-
cies for each of the plates are listed in table 9.5,
along with theoretical results as determined
from equation (9.32) by using the data of
! Theoret-
ical
Experi-
mental
Thcoret-
ical
Experi-
mental
table 9.4.
366 294 302
This work mas further extended in reference
820 657 644
9.13 wherein an aluminum plate 11 by 11 inches 1620 4250 1216
by 0.275 inch thick had grooves 0.625 inch wide 870 799 810
and 0.210 inch deep milled into one side of it. 1330 1175 1152
A typical repeating section of the plate -was 2100 1782 1760
1700 1643 1580
0.75 inch wide, thereby giving 15 integral
2180 2022 2040
stiffeners each 0.125 inch wide across the width 2900 2645 2570
of the plate. Fifteen cyclic frequencies, both
theoretical and experimental, are exhibited in
table 9.6, where the grooves are assumed to run
in the y-direction (i.e., D,>Dz). The corre- TABLE 9.6.-Experamental and Theoretical Cy-
sponding measured mode shapes are depicted clic Frequencies Jor a Qrooved SS-SS-SS-SS
in figure 9.7. The problem was discussed Square Plate
further in reference 9.14.
[Theoretical values (from eq. (9.32)) are in parentheses]
TABLE
9.4.--0rthotropic Constants .{or Grooved
Plates n
1
l
Cyclic irequegcy, cps, for values of m of-
__
I Orthotropic constants, Ib-in
I
j D, j D, Dxv
__-__
Dk
25 210 8920
Both sides
18 030 6480 1 (1950); (2020)l (2150) (2349) (2638)
I i
0 2568"
0.1378"
L
Q)
E
Q)
rc
.-
v-
3i
tc
0
E
.-
c
0
0
L
.-
FIGURE9.7.-Experimentally observed nodal patterns for grooved SS-SS-SS-SS square plate. (After ref. 9.13)
Wah (ref. 9.15) made a study in which he theory as displayed in table 9.7. The ortho-
evaluated the accuracy of simulating the gross tropic constants used in the orthotropic-plate
vibration modes of a beam-plate system by idealization were DJD= 3.396, D,lD= 1, and
means of an orthotropic plate. The cross Dz,/D=1.08, where D is the flexural rigidity of
section of a plate having stiffeners of a particular the unstsened plate. The quantity p b is de-
size and spacing is shown in figure 9.8. The fined as the mass density per unit volume of
stiffeners are parallel to the s-direction. Both stiffener, and R is the number of stiffeners. I t
materials are assumed to be mild steel. First, would appear from table 9.7 that orthotropic-
an "exact" solution to the beam-plate structure plate theory gives frequencies that are a$proxi-
is feud by using classical isotropic $ate theory mately 3 percent too high regardless of the
for the plate and beam theory for the beams, stiffener spacing.
including twisting. Continuity conditions are A method for representing a simply sup-
enforced across the stiffeners. This solution is ported gridwork of beams as an orthotropic
compared M ith the results of orthotropic-plate
plate is discussed in reference 9.16. The
vibration of a S S - S S S - S S rectangular ortho-
tropic plate is also discussed in references
" H a l f " Stiffener
9.8 and 9.17 to 9.20.
9.2.2 Two Opposite Sides Simply Supported
Let a rectangular plate have its sides x = O ,
z=a simply supported as shown in figure 9.9.
Plote Stiffener or Beam Simple Support It is easily seen that the solution originally
FIGURE
9.8.--Cross section of a stigened plate. (After suggested by Voigt in 1893 (ref. 9.21) for the
ref. 9.15) vibration of an isotropic plate having t;wo
ANISOTROPIC P L A T E S 255
WbZ __
TABLE9.7.-Frequency Parameters pd~"lD
for a Stiflened SS-SS-SS-SS Rectangular Plate
Mode R
1.0 0.5 0.333
(9.35)
as a general solution
X
Ym=A,sin $,y+B, cos $,y
FIGURE
9.9.-Reetangular orthotropic plate having two
+@, sinh +,y+Dm cosh +,y (9.36)
where
opposite sides simply supported.
(9.34)
m=l (9.37)
256 VIBRATION OF PLATES
It is seen that equations (9.34) and (9.36) are investigation of this has been made for ortho-
of exactly the same form as equation (4.21) tropic plates.
for isotropic plates, the only difference being By using the Rayleigh method, Hearmon
in the definitions of the frequency parameters (ref. 9.20) gave an alternate form of equation
$ m and 4,. (9.29) for determining the fundamental fre-
The standard procedure for satisfying the quency parameters of rectangular orthotropic
boundary conditions along the sides y=O and plates having two opposite sides simply sup-
g= b, lvhatever they may be, is substitution of ported. Accordingly, the fundamental fre-
equation (9.36) into these conditions. The quency parameter is determined from
determinant of the resulting four homogeneous
equations in A,, B,, C,, and D, is then set
equal to zero for a nontrivial solution. This
yields an exact solution for the frequencies.
This procedure mas followed by Huffington
and Hoppmann (ref. 9.19), who presented (9.38)
frequency equations and mode shapes for all
six cases arising from the sides y=O, b being where J , K , and L are given in table 9.8 for
either clamped, simply supported, or free, and the various cases. Fundamental frequency
the case of the sides elastically supported. parameters for a five-ply maple-plywood plate
It is easily seen that the boundary conditions determined by equation (9.38) are also given
for simply supported or clamped sides are there.
identical to those of the isotropic case. I t For the SS-C-SS-C plate (fig. 4.4) the
was previously mentioned that the solutions boundary conditions are given by equation
to the governing differential equations also (4.25). The frequency equation is given in
take the same form. Thus, substitution of reference 9.19 as
the solution into the boundary conditions for
the three sets of boundary conditions ( S S S S - a25'sinh +b sin #b++#(1 -cosh 4b cos @I)=@
SS-SS, SS-C-SS-C, and SS-C-SS-SS) would Dv
yield the same , characteristic determinant (9.39)
in terms of J. and 4 as that for the isotropic
case. However, $ and 4 are related differently with $ and $I as given in equations (9.37).
than they are in the isotropic case; conse- The mode shapes are
quently, the eigenvalue results (uzp/a4D)ob-
tained for the isotropic problems in the dis-
cussion of SS-SS-SS-SS, SS-C-SS-C, and
SS-C-SS-SS rectangular plates (sees. 4.1, (9.40)
4.2.1, and 4.2.2) cannot be directly applied here.
It should be noted that the form of solution +
where $ and are the roots of equation (9.39).
given by equation (9.36) depends upon $ and The fundamental frequency parameters of a
CP being real, positive constants. However, by
five-ply ~ a p ~ e - ~ ~ plate
~ o determined
o d by
this method in reference 9.22 are given in
looking at equations (9.37) it is seen that,
table 9.8.
depending upon the ratios Dx/D, and D,,/D,, Kanazawa and Kawai (ref. 9.23) solved this
the constants #I and t#~ may also take on zero, problem by an integral equation approach and
imaginary, or complex values. I n these cases gave numerical results for the fundamental
the form of equation (9.36) must be modified. frequency parameters of a square having vari-
A careful study of this phenomenon was done ous ratios of Dx/D,an Du/Dzv. These are ex-
in the case of isotropic plates (see the discussion hibited in table 9.9. The values computed from
of rectangular plates with two opposite sides equation (9.29) are fslsnd in re€erence 9.8; these
simply supported (sec. 4.2)), but no systematic can be compared with the footnoted values in
ANISOTROPIC PLATES 257
TABLE
9.8.-Fundamentul Frequency Parameters for a 5-Ply Maple-Plywood Rectangular Orthotropic
Plate Hawing Various Boundary Conditions
I I I
a=2.0
500. 56 121. 5 129.5 94.57 94. 56 b
33.117
D
v "=a12
!
~ 97.41 1 97.41
I
91. 41 48. 65 48. 65
0 20. 65
I i
TABLE9.9.--Fundamental
~
Frequency Parame- table__9.9. The values from reference 9.8 are
ters ~~~ljp/D~. for SS-GSS-6 Square Ortho- wa24p/~,,=21.@, 26.5, 42.2, and 45.1.
tropic Plates Having the Sides x=O and x=a Frequencies for this problem may also be
Simply Supported determined from equation (9.29).
D.,
11 oaZdp/Dz, for values of D,/Dz, of-
For the SSC-SS-SS plate (fig. 4.8) the
boundary conditions are given by equation
(4.32). The frequency equation is given in
I
reference 9.19 as
24. 664 826. 595 6 tan $4=# tanh 6b (9.41)
26. 397 28. 226
30. 968 32. 507 with # and 4 as given in equations (9.37) ~ The
38. 384 39. 662 mode shapes are
44. 589 &45.696
where $ and 4 are the roots of equation (9.41). TABLE9.lO.-Freyuency Parameters wa2JpP,
The fundamental frequency of a five-ply maple- for a SS-C-SS-F 5-Ply Maple-Plywood
plywood plate determined by this method in Rectangular Orthotropic Plate
reference 9.22 is given in table 9.8. The case
when alb=10 was also analyzed for the same
material and gave ua2Jm 1546.68
=
equation (9.41) was used and 1546.96 from
when I -I Exact value
Wa2JplD,
Rayleigh method
equation (9.38). 1
~
sinh +b sin $b
'\Jt"-y4-4264)
+24$y2h2(~osh+b COS $6- 1 > = O (9.48)
(9.45)
with $, 4, y, and S given by equations (9.37)
Several roots of equation (9.43) were found
and (9.44). The mode shapes are
in reference 9.22 for a five-ply maple-plywood
plate having alb=2.0 and having the material 6 cosh 6ySy COS $y-
properties listed in table 9.8. The frequency Y(y'=y6(cosh +--cos $b)
parameters for this plate are given in table 9.10.
The corresponding values obtained by the - $7 sinh +y +46 sin 9.49)
Rayleigh method from equation (9.29) are aIso sinh4b-+h2
$q2 singb
given in reference 9.22 and are listed in table Some numerical results for this problem are
9.10. It should be noted that for m = I and given in table 9.8.
n=3 the "exact" value is not lower than that Naruoka and Yonezawa (ref. 9.24) rewrote
of the Rayleigh method; this indicates round-off the differential equation (eq. (9.27)) as
error in these calculations.
For the SS-SS-SS-F plate (see fig. 4.11) the
boundary conditions are given by equation
(4.40). The frequency equation is (ref. 9.19)
(9.50)
(9.46) where
I( =D,UlJrn, (9.51)
with #, 4, y , and 6 given by equations (9.37)
I n this form it is clear that equation (9.50) is
and (9.44). The mode shapes are
factorable if K is either I or 0, and these values
are used in reference 9.24. Furthermore,
symmetry is taken advantage of by using
ANISOTROPIC PLATES 259
the 5 and y-axes (see fig. 4.12;) through the TABLE9.11.-Ratio of Second and Third Fre-
plate centroid. Finally, the two cases quencies 20 the Fundamental for SS-F-SS-F
Rectangular Orthotropic Plates
I
_
1:4.0:17 ,
I
1:1.9:3.9
1:3.4:13
second symmetric mode, both taken with fundamental is set forth in table 9.11 for various
respect to i . These modes are shown in figure a/b and D,/D, ratios. Poisson’s ratio and K
9.10. Variation in frequency parameter with are not given in table 9.11 but are presumed
D,/D, ratio is shown in figure 9.11 for K=O, t o be 0 and 1, respectively.
K = I , and beam theory. Poisson’s ratio V = For the plate elastically supported on the
vzl/=vYz is taken as zero and alb=2. Further edges y=O and y=a (fig. 1.59) and simply
results for varying a/b ratios are given in figure supported on the other two edges, the boundary
9.12 for the second symmetric mode. Finally, conditions are given by equations (4.71). The
the ratio of second and third frequencies to the frequency equation is (ref. 9.19)
{$zyz(y2- K,K3a4)
-+262(62- +K&a4[42(62-K,K3a4)-$2(y2- K,K8a4)]
KlK3a4)
+a2(y+ 6)2(K1K2+2@-
KsK4a4))
sinh +b sin $b +++[2y26’- KlK3u4(y2+6’)
-KzK4a4(62+y2-2K,K3a4)
--a4(K2K3+K,K4)
(y+6)2] C Q S ~+b cos $b
+ a $ ( $ + 6 ) ~ - K ~ a 2 y 2 + K ~ ~ 2 6 2 - K ~ a 2 ~+K~+2(62-KlK3a4j
yz-K~K~a~~
sinh & cos $b+a+(y+6)~K~a282+K,$2y2
+K2K4a4(K3a2-KK,+2)]
+K4a2(62-K,K3u4)
+KZJ/2(y2-KlKsa4)
--KzK,jK3aZ+KlJ/2)]
cash +b sin ’$b
-2+$($ +K2K4a4)
-+K1K3cc4y(y& =0 (9.53)
with $, 4’y, and 6 given by equations (9.37) and (9.44) and the spring constants K,, . . .,
determined by equations (4.71). The mode shapes are
260
\
c
0
J t
0-4
6
5
4
6 4
3
3
N
.
t
2
-
NO
2
I
I
20
16
6 2 12
I \
N IN
1. "e 8
0 100 200 300 400
4
(b) DxIDy
16
The problem for the plate with all sides
clamped is described by figure 4.18. Frequency 0 100 200 300 400
Dx / Dy
parameters may be calculated from a formula (d
based upon the Rayleigh method given previ-
ously as equation (9.29). Plots of frequency F:GWRE 9.E-Frequency parameters for the second
symmetric mode (with respect to i j ) of SS-F-SS-F
parameter variation with a/b ratio for four rectangular orthotropic plates for various a/b ratios ;
modes were given previously in figure 9.5 for a K= 1. ( a ) a / b = l . ( b ) a/b=2. ( c ) a/b=4. (d)
particular maple-plywood plate. a/b=8. (After ref. 9.24)
ANISOTROPIC PLATES 26 1
Another Rayleigh solution is obtained in TABLE9.I2.-Frepuency Parameters wa2& for
references 9.10 and 9.17 by taking the deflection C-6C-C Square Orthotropic Plates Made of
function Birch
-[
W(2,5) 22- (4)2 p-(;y-J (9.55)
j ua2Jp for-
Method
which yields the fundamental frequency Isotropic case 5-ply plates Veneer
plate
I
W(Z,?)=(I +cos F) + 7)
(I cos (9.57)
is used, giving
&!!any experimentally determined funda-
mental frequencies are also given in reference
9.20 for plywood and veneer plates made of
various wooden materials. In references 9.7
by the Rayleigh method. Finally, reference 9.8
and 9.11 this experimental work is extended
gives the Rayleigh solution using beam func-
in order to study the effect on the frequencies
tions described previously as
when the grain of the veneer OF plywood is not
parallel t o the sides of the plate.
HuEngton (ref. 9.26) postulated the exist-
(9.59) ence of nonparallel node lines for clamped
orthotropic plates; this idea was based upon
this latter clearly being the best of the three his observations of the numerical behavior of a
results listed, because it gives the lowest upper two-term Ritz solution using beam functions.
bound unless D,, is considerably larger than This phenomenon is predicted by the curves
,and D,. of figure 9.13 which show frequency parameters v !
Further improvement of the theoretical fre- as functions of a/b ratio for the-:case when -,e
quencies was obtained in reference 9.10 by D,/D,,=1.543 and D,/D,,=4.810. The nu-
taking the two-term de6ection function Rnerical results show that the curves (each
associated with a mode shape) do not cross but
approach each other and veer away. I n the
s7icinity of the location where the curves ap-
proach each other, there is a rapid change in
(9.60) nodal patterns, as depicted in figure 9.14. It
must be remarked that this phenomenon has
and using the Rayleigh-Ritz procedure. The been observed elsewhere (see discussion of the
convergence of frequency parameters when C-C-6-C rectangular plate (sec. 4.3.1) and
equations (9.55) and (9.60) are used can be that of the 6-F-F-F rectangular plate (sec.
seen in table 9.12 for two types of square 4.3.12)) and the question exists of whether it
plates made of birch. Results are also in- is the result of numerical truncation.
cluded for the isotropic case for comparison Kanazawa and Kawai (ref. 9.23) solved
with Tomotika’s “exact” solution (ref. 9.25). this problem by an integral-equation approach
(See discussion of the 6-C-6-6 rectangular and gave numerical results for the funda-
plate (sec. 4.3.1).) mental frequency parameters of a square having
262 VIBRATION OF PLATES
"_l __
WCL~JZ
for values of D,/D,, of-
3/1+1/3 mode D,
D,,
34 % 1 1 2 1 3
---____--
b)
FIGWREQ.l5.-Fundamental frequency parameters
w2a4p/DZ,7r4 against D J D z u and D,/Ds, for G C -
FIGURE9.14.-Nodal patterns in the vicinity of a C-C orthotropic square plates. (After ref. 9.23)
transition point. (a) Nodal patterns for mode
3/1-1/3. (b) Nodal patterns for mode 3/1+1/3.
9.2.4 Other Boundary Conditions
(After ref. 9.26)
Frequency parameters for C-C-C-SS and
various ratios of Dz/Dz, and D,/Dz,. These are C-C-SS-SSrectangular orthotropic plates may
exhibited in table 9.13. An interesting plot of be determined from the RayIeigh formula given
the results of table 9.13 is given in figure 9.15. previously as equation (9.29). Plots of fre-
It would appear from this figure that the varia- quency parameter against a/b ratio for four
tion in the square of the frequency with either modes are given for the C - 6 - S S S S case in
or D,is linear. figure 9.5.
ANISOTROPIC PLATES 263
An integral-equation approach (ref. 9.23) where
gave numerical results for the fundamental
frequency parameters of square plates having
C-C-C-SS and C-C-SS-SS edges. These are
listed in tables 9.14 and 9.15.
D y '
Ds,
_--_______--
% I % 1 2 3
(see discussion of rectangular plate with two FIGURE 9.17.-Experimentally observed cyclic frequen-
cies and nodal patterns for a clamped circular plate
opposite sides simply supported (sec. 9.2.2)).
having stiffeners. (After ref. 9.27)
The identical result was obtained in reference
9.17 by using the Rayleigh-Ritz method.
9.2.6 EHipticai Plates aving Rectangular Orthot-
ropy
In reference 9.16 the Galerkin method is
used with the one-term deflection function
wClamped Boundary
Fundamental Mode
Frequency = 850 cps
Frequency-1150 cps
-------
Frequency= 1490cps Frequency = L960 cps
-------
Frequency = 2630 cps Frequency =3320 c p s
REFERENCES
9.1. AKASAKA, T.; A N D TAKAGISHI, T.: Vibration of
Corrugated Diaphragm. Bull. JSME, voi. i,
IIO. 3, 1958, pp. 215-221.
9.2. BORSUK,K.: Free Vibration of Rotations of a
Cylindrically Orthotropic Circular Pl:r tc.
Arch. Mech. Stos., vol. 12, no. 516, 1960,
pp. 649-665.
9.3. MINKARAX, I. A.; A N D HOPPMANN, W. H., PI:
Flexural Vibrations of Cylindrically Aeolo-
tropic Circular Plates. Jour. Acoust. Soe. Am.,
vol. 36, no. 3, Mar. 1964, pp. 470-475.
9.4. PANDALAI, K. A. V.; A N D PATEL, S. A.: Natural
Frequencies of Orthotropic Circular Plates.
AIAA J., voL 3, no. 4, Apr. 1965, pp. 780-781.
9.5. P Y E S Y E N N I K Q V A , N. K.; A N D SAKHARQV, I. E.:
Natural Vibrations Frequencies of the Funda-
mental of Annular Plates With B Cylindrical
FIGURE 9.18.-Elliptical plate with slots milled parallel
Anisotropy. Izv. An SSSR, OTN. Mekh. i
t o major and minor axes t o simulate a n orthotropic
Mashin., no. 6 , 1959, pp. 134-135. (In
plate. (From ref. 9.27) Russian.)
308-337 0-70-18
266 VIBRATION OF PLATES
es
I n this section the effects of forces acting in (see the appendix). It is emphasized that the
the plane of the undeformed middle surface of inplane forces are generally found by first
the plate d l be considered. The differential solving the plane elasticity problem for known
equation of motion expressed in rectangular boundary values of N,, N,, and ATzu. If these
coordinates in this case becomes (see the quantities are constant around the boundary,
appendix) : it is well known that they are also constant
throughout the plate, and equation (1O.I) is
a =+
b4W 2Dz, b4W +0,-+
b4w
by4
bt2
P b2w further simplified to the case of constant
coefficients. I n the special case of uniform
- + 2 i j ~ , , b2W
= ~ , -b2W -+~~- b2W (10.1) boundary tension (N,=N,=N; Nzy=O), the
6X2 bxby bf equation for the isotropic plate simplifies t o
where D,,D,,and D,, are the constants of b2W
rectangular orthotropy, as used extensively in DV~w--NV2w+p--=@ (10.2)
at2
the discussion of rectangular orthotropy of
anisotropic plates (sec. 9.2). Because no pub- Assuming sinusoidal time response, equation
lished results are known for plate vibrations (10.2) becomes
when both inplane forces and orthotropy are
present, only the isotropic constant 13 mill
appear in the remainder of this section.
The inplane force intensities N,, A',, and N,,
where W is solely a function of the spatial co-
are assumed t o be functions of on13 the spatial
ordinates. Furthermore, it can be seen Ghat
coordinates x, y or r , 8. That is, they do not
equation (10.3) can be factored into
epend upon time nor upon the transverse *.
deflection w. These assumptions ore required + w=
(v" 2 )(V2- p* 1 0 (10.4) ';
in order that-
(1) The vibration be free, not forced
(2) The equation of motion remains linear
lnplane forces not depending upon 20 can be
realized in one of the following two ways: (10.5)
(1) The boundary conditions provide no
fixity in the plane of the plate
( 2 ) Tbe deflection is sufficiently small relative
to the initial tension or compression in the
plate so that the inplane forces are not signifi- 10.1 CIRCULAR PLATES
eantly &ected. The main results arailable for circular plates
The normal forces N , an IV, aseFpositive in for the case of hydrostatic inplane force.
equation (10.1) if the pla is in tension; the en v2is expressed in terms of polar coordi-
shear force N,, is positive according to the nates by means of equation (1.10) and Fourier
accepted convention of the theory of elasticity components in e are assumed as in equation
267
268 VIBRATION O F PLATES
(1.15), equation (10.4) yields the two second- and 2 nodal circles and nodal diameters for a
order equations range of inplane forces varying from tension to
compression. These results are given in table
10.1. Herein the quantity 4 is used as a
multiple of the critical buckling load in com-
) (10.6) pression ; that is,
+=-----Na2 (10.11)
14.680
0>=2
~ ( r ,
n=O
IAnJn(m>+BnYn(m)+QnIn(or1
oaZJrD for values of
n of-
10 p = w a d p (10.12)
o a 2 m derived by-
nls
Naz
_-
D Exact method Southwell method Rayleigh method
(ref. 10.3) (ref. 10.5) (ref. 10.6)
1 0 39.772 39.772
1 40. 190 40. 152
4 41.419 41.272
25 49. 146 48.396
100 69.916 67.996
400 120.59 117.25
m 5.5151dNFD
this section, the modes having 3, 4, 5 , and 6 nodal diameters were also investigated, but the per-
turbation technique did not give accurate results. The Rayleigh and Southwell techniques were
also employed, thereby obtaining bounds. Resulting frequency parameters are given in table
10.4.
The problem was also discussed from a variational standpoint in reference 10.2. A method
for including translational and rotational springs acting at discrete points within the interior of
the plate was proposed and demonstrated for the case of a translational spring of stiffness k a t
the center. All terms applying to cos n6 are retained in equation (10.8). I n addition to the
boundary condition equations (2.2) , the conditions of transverse force equilibrium and null slope
at the center are enforced. For the axisymmetric modes, the resulting characteristic determinant
takes the form
(10.18)
20
w = R 9 Cz,(~)cosnO cos(&++)
n=o t=O
(10.25)
-I 0
(C)
(10.29)
and has nodal circles at r=0.4013a and
r=0.8472a and an amplitude at the boundary
where LY is the coefficient of thermal expansion. of WO2(a)=0.5336Cz.
The problem is solved by the Rayleigh- For the modes having two no
Ritz technique. Poisson's ratio is taken to a deflection function
be 0.3. For the axisymmetric modes a deflec-
tion function
1-2.6161
(3'
- $1.1090
(3-
i56.48
F65.54 (10.35)
1Q.i.4 Rotating Disk, Clamped at Center, Outer The differential equation of motion (eq. (10.1))
Edge Free becomes
Southwell (ref. 10.11) analyzed the problem N b2w N a2w N b2w
of a circular disk which is clamped at its center, v4W-k4W=--4:-+2"U-+>-
D bx' D bxby D by2
is free a t its outer edge, and is rotating with
constant angular velocity 8. He again used the (10.40)
method for finding lower bounds on the fre- duxe
quencies which was discussed in section 10.1.1. k"-% w2 (10.41)
The frequencies are given by
D N , are constants, say Nl and N z ,
W~=KIQ'+KZ
-
Pa4 respectively, and N,,=O, equation (10.40)
becomes
where Kl and K 2 are given in table 10.6 and V4W- k 4 W -NL -
a2w N a2w
+2 - (10.42)
v=0.3. D ax2 D by2
10.2 RECTANGULAR PLATES which is of a form particularly amenable to
As described in the chapter entitled "Rec- solution.
tangular Plates" (ch. 41, there exist 21 pos-
sible combinations of simple boundary con- 10.2.1 Plates Having Alf Sides Simply Supported
ditions for rectangular plates. Results were The boundary conditions for the problem of
found in the literature for all 21 cases for plates having all sides simply supported are
isotropic plates not having inplane forces. As defined by equations (4.18)and figure 10.5. In
be Seen in the fo]lowing discussion, pub- figure 10.5, the positive Senses Of the inplane
Bished results exist for very few cases when forces N,, N,, and N,, are shown for the special
inplane forces are present. Also, it will be seen case when each is constant throughout the plate.
Nodal diameters
Nodal circles ___
0 1 2
I 3
I
The influence of a body force is also con- The frequency ratio ( W / W ~ )is~ plotted in figure
sidered in reference 10.12. The body'force is 10.7 as a function of the ratio Nl/(Nl)cr and a
assumed to be acting in the x-direction and parameter E defined by
may be due to the weight of the plate (if it is
in a vertical position), or it may arise from (10.51 )
acceleration in the negative x-direction. Thus,
in this case, all the inplane forces are not for the particular aspect ratio alb=3. The
constant but are given by
quantity w, is defined by equation (10.48).
Frequency parameters for this problem were
N,= Ni -yb 2 computed in reference 10.13 for use in deter-
(10.49)
N, =N,, =0 mining lower bounds for completely clamped
square plates subjected to hydrostatic tension.
where Nl is the inplane tension at the end These are listed in tabIe 10.7.
x=O, and y is the body force (force per unit Some experimental results are reported in
area). The Rayleigh method was used to reference 10.14. A 24S-T duralumin plate, 12
solve the problem, with the first term (m=n= 1) inches by 12 inches by 0.040 inch thick, was
of the sine series expansion for deflection (eq. simply supported along all edges and subjected
(10.43)) being kept. This yielded the frequency to the constant inplane load N,=N, and
parameter Ny=Nzy=O. It was found that the experi-
mentally measured frequency does not decrease
as rapidly as that predicted by theory when the
compressive loading is increased. This is
(I 0.50) shown in figure 10.8. In reference 10.14 this
effect is attributed to the possibility of slight
initial curvature in the plate.
6 """"
5 2500
4 2000
$ 3
% 1500
2
1000
I
500
0
0 0 25 0 50 0 75
0 -50 -100 -150 -200 -300
0
FIGURE IO.?.-hfluence of end loading N1 and body FIGURE HO.B.--Deviation of experimentally measured
force ratio 5 on the fundamental frequency of a frequencies from those predicted by theory for a
SS-SS-SS-SS rectangular plate for aib= 3. wi= SS-SS-SS-SS square plate loaded in one direction.
4Lh4/paZba;E= rab2/9D. (After ref. 10.12) (After ref. 10.14)
PLATES WITH INPLANE FORCES 279
TABLE 10.7.-Frequeney Parameters for a Square where a=mz-/a. Substituting equation (10.52)
Plate Subjected to Hydrostatic Tension and into equation (10.42) yields
Hauing Clamped Boundaries Compared With
Those -for a Plate With Simply Supported
Boundaries
Frequency param- +(a4--k4+ga2)~,=~ ( m = l , 2 , . . .>
Frequency parameter?
for simply supported eter w a 2 d J B for (10.53)
Nu2 plate clamped plate
- __
which has a general solution
8 2 0
(2,y > = g Y , ( y )
m=l
sinm (10.52) v , = & , , + ~ +bn~ , , *(10.56)
280 VIBRATION OF PLATES
By looking at equation (10.55) , it can be seen where and are coordinates having their
that +,
and cpm can be positive real, zero, origin at the center of the plate. (See fig. 4.18.)
imaginary, or complex. The solution form of For purposes of comparison, the easily deter-
equation (10.54) is based upon the assumption mined frequency parameters when all sides are
that +m and cp, are positive real numbers; simply supported were computed in reference
otherwise, the form would change. No study 10.13 and are also given in table 10.7. Also, a
is known in which the character and range of plot was made which compares the frequencies
applicability of the separate forms of solution of a clamped square plate with those of clamped
have been investigated. circular plates having area and circumference
The Xayleigh method is used in reference equal to those of the given square plate. The
10.12 to obtain an approximation for the funda- circular-plate results mere obtained from refer-
mental frequency of a rectangular plate having ence 10.3, as discussed previously for clamped
the edge y=b free and the others simply circular plates (sec. 10.1.1). These curves are
supported. The loading is N,=N, and N , shown in figure 10.9.
=Nzy=O. A deflection function In reference 10.19 the Kato-Temple method
TX (refs. 10.20 and 10.21) was used to derive an
W(z,y)=ysin- ( 10.57) extremely accurate lower bound for the funda-
a
mental frequency of a clamped square plate
was used. The resulting expression for the
subjected to hydrostatic tension N= 10srzD/u2.
frequency is
Accurate upper bounds were obtained by using
the Rayleigh-Ritz method with beam func-
tions (see discussion of the C-C-C-C rec-
I n reference 10.18 the case is considered when tangular plate (sec. 4.3.1)), keeping both 6 and
three sides are simply supported, the other is 36 terms in the series. These results are com-
damped, and two concentrated, collinear, com- pared with those of reference 10.13 in table
pressive forces Po act upon the two opposite 10.8.
simply supported edges. No numerical results The perturbation technique is used in refer-
are given. ence 10.15 to obtain fundamental frequency
Experimental results are given in reference
10.14 for the case when two opposite edges are
damped. A disagreement with theoretical re-
sults was found, similar to that discussed
previously in the discussion of plates with all
sides simply supported (sec. 10.2.1).
10.1.3 Rectangular Pfates aving AH Sides
Clamped
The problem of plates with all sides clamped
is defined by figure 10.5 with boundary condi-
tions w=dwlbn=O on dl edges.
Weinstein and Chien (ref. 10.13) used a rari-
ational technique to obtain lower bounds for
the fundamental frequency of a square plate
under the hydrostatic tension N,= N,= N and
N,,=O. Results are listed in table 10.7 for
I I I I I
varying degrees of inplane tension. Upper 20 40 60 ao IC0 120
Ref. 10.19
Ref.
10.13
6 terms where a is the altitude of the triangle (see
fig. 7.15) and h t are the eigenvalues of the
59. 922
1 59. 98389
! 59.98488
i 59. 98498
l 60. 392 membrane vibration problem determined from
2
xi==- 7T2(Z2$ m2+ 2 )
3
parameters for the problem previously discussed. (10.62)
E+mSn=O
Results are summarized in table 10.9.
I,m,n=&t, &2,&3, ...
10.3 PLATES HAVING OTHER SHAPES
The first six values of (Z2+m2+n2)are given in
Lurie (ref. 10.14) showed that for a plate of table 10.10. A plot of the first six plate €re-
any polygonal shape, with all its boundaries quency parameters as functions of the inplane
simply supported and subjected t o hydrostatic compression appears as figure 10.10.
pressure N,=N,=N=-p rtnd Nz,=Ol the Kaczkowski (ref. 10.23) utilized the fact that
vibration mode shapes are independent of the the superposition of certain vibration modes
btensity of p . Hence, the mode shapes are (having the same frequency) of a simply sup-
identical to the buckling modes of the plate and ported square plate will give a combined mode
also identical to the vibration modes of a stretched which has a nodal line on the diagonal of the
membrane having the same shape. Further- square. In this way the frequencies and modo
more, the frequency of the loaded plate can be shapes of a plate in the form of an isosceles right
triangle with all edges simply supported caa be
expressed as
(10.60)
'HIABLE 1 ~ . 9 ~ - F ~ n ~ m~ er en ~~ ~~ e nh cmym -
eter w a 2 m Derived From the Perturbation
Method for a Clumped Square BEate Subjected
to Hydrostatic Tension
35.989
49. 62%
FIGXIRE IO.lO.-Frequency parameters w 2 a 4 p l D as func-
60. 019
68. 566 tions of inplane hydrostatic pressure for a simply
supported, equilateral, triangular plate. (After ref.
10.22)
282 VIBRATION OF PLATES
TABLE 10.10.-Temzs for Computing the First (1) The side x = O clamped, the others
Six Eigenvalues for the Equilateral Triangular simply supported
Membrane (2) The sides clamped, the hypotenuse
simply supported
i n P+m2 +n2 (3) The side x = O free, the others simply
supported
1 1 -2 6 (4) Two sides free, the hypotenuse simply
1 2 -3 14 supported, and the point (0, 0) sup-
2 2 -4 24 ported
1 3 -4 26 (5) One side clamped, one side free, the
2 3 -5 38
P 4 -5 42 hypotenuse simply supported
(6) Two sides simply supported, the hy-
potenuse clamped
(7) Two sides simply supported, the hy-
found. The frequencies for N,=N,=N, N,,=O potenuse free
are given by Pan (ref. 10.24) used the method of images
to show that the square of the fundamental
frequency of a 30"-60"-90" triangular plate
simply supported on all sides (see fig. 7.17)
and subjected to hydrostatic tension N is
( m ,n=1, 2, 3 . . .) (10.63)
and the fundamental frequency occurs when and the mode shape is
m=n= 1 :
l+m
-Ex(
W l I - a2 "> (10.64)
. ?Fx
~ ( xy),=sin ----sin
a43
. 3ry
-+sin
a
. 4*x
~
a&
.
sin22
a
The mode shapes of the triangular plate are
(in terms of fig. 10.5):
es Varia ess
I n the case of plates With variable thickness, where D=Eh3/12(l- 2) ; that is, D is a function
the governing differential equation of motion is of the thickness.
found to have variable coefficients, and this To obtain a fourth-order differential equation
fact increases the difficulty of solution. This corresponding to equation (1.11, it is only
added complexity will be demonstrated below necessary to substitute equations (11.2) into
in both polar and rectangular coordinates. the last two of equations (11.1) and, in turn,
Results are available only for isotropic plates substitute these into the first of equations
having no inplane forces. (11.1). However, if the thickness is a function
of r and/or e, the resulting differential equation
1i .i CIRCULAR PLATES
will be quite lengthy and will have variable
If inplane forces and rotary inertia are coefficients (i.e~,functions of r and/or e). This
disregarded, the equations of motion in polar expanded equation will not be presented here.
coordinates are Needless to say, very little has been done
toward obtaining solutions t o this differential
equation in all its generality.
Timoshenko and Woinowsky-Krieger (ref.
11.1) and Conway (ref. 11.2) showed that, for
the axisymmetric problem (no variation with
e), the equation of motion becomes
equation (11.4). For m=2, v=1/9, and a solid For m=3 and v = 1 / 3 , there is the important
plate, the solution takes the form case of linearly varying thickness, which is
discussed in reference 11.3 as well as in refer-
=T - 213 + 1
C J , (u) CJ, (u) I 1.7) ence 11.2. The characteristic equation for a
where solid circular plate is found to be
u =br2J3 1
(11.8) Jz(uo)Ii(~o)=Ji(uo)lz(~o) (11.14)
with
and J1 and Il are the regular and modified
Bessel functions of the first kind of order one.
Applying the boundary conditions (eqs. (2.2)) Equation (11.14) is also the characteristic
gives the characteristic equation equation for the transverse vibrations of a
cantilever beam having a circular cross section
(11.9) and linear taper. Thus, by analogy with
results for beams, the first three roots o€
where the primes indicate differentiation with equation (11.14) are found to be u,=5.906,
respect to the argument u and 9.197, and 12.402.
The ratios of the first three axisymmetric
ug= b d J 3 (11.IO) frequencies of the constant-thickness plate
to those of the variable-thickness plate having
where a is the boundary radius. The first
the same thickness a t the boundary are found
10 roots of equation (11.9) were given in
to be 1.17, 1.88, and 2.31.
table 2.1 (n=l), the lowest root being
The case when m=6 and v is arbitrary is
also discussed in reference 11.2, but no numer-
I ' 4 a2J3
= (21.2 6) /' ( 11.11) ical results are given.
uo=(x)
It is interesting to observe that in the case
Consider u clamped circular plate having a of variable-thickness plates the frequency pa-
constant thickness equal to the maximum rameter depends upon Poisson's ratio for
thickness (at the boundary) of the variable clamped as well as for other boundary conditions.
thickness plate previously described (m=2, In reference 11.4 the work just described
v=l/9). Then, according to reference 11.2, was extended to annular plates of linearly
Lhe ratio of the fundamental frequency of the varying thickness which are clamped on both
constant-thickness plate to that of the variable- the inner and outer boundaries (fig. 11.1).
thickness plate is 1.08. The solution €or the linearly tapered beam
For m=18/7 and v=5/21, the frequency again applies when Poisson's ratio for the
equation plate is 113. The characteristic determinant
yielding the frequencies is
uo(tanu,+tanhuo)=2 tanu, tanhuo (11.12)
was given, where now
T
H
1
(11.IS>
Y *-
288 VIBRATION O F P L A T E S
where
32 w27b4
+--
3 Eh2
(11.19)
K=a/b
12(l-v2)wZy ={ -(pp’-2a’X-2aXUx’)
2%;
PLATES WITH VARIABLE THICKNESS 289
where a, a', p, p', x, and x' are given by
- 1
b b2
k+6- k2+3
(a-ay ~
(a- b)3
a5-b5
_ _ - k3
_ a6-b6
_ _ ~
5 (a-b)3 6
b b2
3-k+3-
a-b (a--b)*
3 b
--k+6- (11.24)
a--b (a--b)*
5 5 + 7
2(a3-b3)b5-4ja5-b5)b3-(a6-b6)bz
15 25 6
2(aZ-bz)b6-4(u5-b5)b3
25
(14.25)
290 VIBRATION OF PLATES
Detailed calculations were made for an (see fig. 11.6). His primary work was a direct
aluminum disk having the following constants : attack upon the differential equation by assum-
a= 1.00 inch, b=0.375 inch, E= 10.6X lo6 psi, ing a series form of solution. Boundary con-
and v=0.33. These results are plotted in figures ditions led to an infinite characteristic determi-
11.4 and 11.5 for various tapers k and are nant, which was truncated for an approximate
compared with experimental results for k=4/5. solution. Detailed numerical results were given
In figure 11.4 the theoretical values are plotted for the special configuration where the boundary
directly as they arise in the computations. r=b=O.lro was clamped and the boundary
In figure 11.5 the values are adjusted t o account r=a=0.5ro was completely free. A Poisson's
for additional cement and a barium titanate ratio of 113 was used. By use of the series method
element used in the experiment. the lowest axisymmetric frequency parameter
Kovalenko (ref. 11.6) made a study of the was found to be
annular plate having thickness varying accord-
(1I .27)
ing t o the equation
"
0
c 12
U
When equations (11.27) and (11.28) are com-
c pared it can be observed that, as in the
- 8 case of certain b/a ratios for Constant-thickness
annular plates (see discussion for annular
4 0 Computed Valve plates (sec. 2.2.7)), the fundamental mode is
X Computed Valve for k =$ antisymmetric. In table 11.3 are given the
mode shapes corresponding t o these two fre-
0 10 0 15 0 20 0 25
h, in inches quencies and the ratios of bending moments.
Rayleigh-Ritz sohtions were also obtained
FIGURE11.4.--Uncorrected cyclic frequencies f for a in reference 11.6 by using the radial variation
linearly tapered, circular aluminum plate. (After in deflection
ref. 11.5)
W ( T=) A ~ ( T - ~ ) ' + A ~ (b)3
T - (11.29)
giving the frequencies
X
-
0 Computed Value
4
x Measured value for k = 3
h t---.O----l
t
0 IO 015 0 20 0 25
h, In tnches
0 1 I 0 1 1
'
. 362 .968 . 112 .319 ,713
. 362 . 149 .552 ,359 .I20 . 334
.677 j . 245 .025 , . 121
. 094 0 1 . 035
I
~2a*yh~/D~
a
-
b CY
da'rho/&
10
where the constants A, B, C, and D are given
08 in table 11.5 for the types of boundary condi-
tions depicted in the table. The thickness
-6 0 6
u parameter is defined by
-
2 04
02 (11.36)
0
0 02 04 06 08 IO where the thicknesses ho and hi are as shown
in table 11.5.
FIGURE 11.7.--Fundamental mode of a simply sup- Runkett and w&on (refs. 11.13 and 11.14)
ported square plate having linear thickness variation measured the frequencies of linearly tapered
in the z-direction; v=0.3. (After ref. 11.9)
steel cantilever plates, with the taper o ~ c u ~ ~ i n g
between the free edges as shown in figure 11.8.
Gumeniuk (ref. 11.11) used the finite-diff er- Figure 11.8 shows the variation in the frequency
ence method to derive a formula for the fun- parameter
damental frequency of a simply supported
I rectangular plate having linear thickness vari- (11.37)
ation. This work was extended by Gontkevich E
(ref. 11.12) to plates having other boundary
undamental frequencies ttre de- with the wedge angle 8, where ho is the greatest
termined from the formula thickness and a is the span of the plate ( 5 inches,
in fig. 11.8). The values shown for zero wedge
angle (constant thickness) were computed by
elementary beam theory. Fundamental fre-
(11.35) quency parameters for the various wedge
294 VIBRATION OF PLATES
PLATES WITH VARIABLE THE-SS 295
-$ 2
n
x im
Y
296 VIBRATION O F PLATES
400
I Complete Wedge
zoo$
C#I
1
Wedge angle, e, O 1.35
(eq. (11.37)). _ _
- 2.4
~
2. 52
_ 2. 57
1
1
- I 3.7
-
~ 5.9
2. 47 2. 32
11.8
~
2. 28
-
Mech., vol. 23, no. 1, Mar. 1956, pp. 109-115.
11.9, APPL, F. C.; A N D BYERS,N. R.: Fundamental
Frequency of Simply Supported Rectangular
Plates With Linearly Varying Thickness. J.
Appl. Mech., vol. 32, no. 1, Mar. 1965, pp.
163-167.
TABLE11.7.-Variation in Fundamental Fre- 11.10. APPL, F. C.; A N D ZOROWSKI,C. F.: Upper and
Lower Bounds for Special Eigenvalues. 9.
quency Parameter With Aspect Ratio f o r Line- Appl. Mech., vol. 26, 1959, pp. 246-250.
arly Tapered, Rectangular Cantilever Plates; 11.11. GUMENIUK, V. S.: Determination of the Free
Material, Steel Vibrations Frequencies of Variable Thickness
Plates. Dopov. AN UkrSSR, no. 2, 1956,
a/b 1 1 1
2.00 2.22 2.86 4.00 6.67
__-
1 pp. 130-133. (In Ukrainian.)
(b (eq. (11.37))_.__~ 1
2. 57 2. 57- - - I1 2. 71 2. 91 3. 15 18.12. GONTKEVICH, V. S.: Natural Vibrations of Plates
and Shells. A. P. Filippov, ed., Nauk.
Dumka (Kiev), 1964. (Transl. by Lockheed
11.3 OTHER SHAPES Missiles &. Space Go. (Sunnyvale, Calif.) .)
Except for the work in references 11.19 and 11.13. PEUNKETT, R.:Natural Frequencies of Uniform
and Non-Uniform Rectangular Cantilever
11.20, virtually nothing‘ has been done for Plates. J. Mech. Eng. Sci., vol. 5, no. 2, 1963,
variable-thickness plates when their shapes are pp. 146-156.
other than circular or rectangular. A method 11.14. WILSON,R. E.; A N D PLUNKETT, R.: Vibration of
is presented in reference 11.19 for analyzing Cantilever Plates With Rectangular and
Wedge-Shaped Cross-Sections. Rept. D F
cantilever variable-thickness plates having an 53GL17, Gen. Elec. Co., Mar. 1953.
arbitrary quadrilateral shape. Reference 112 0 11.15. MAZURKIEWICZ, Z.: The Problem of Bending
gives 8 method for analyzing clamped variable- and Free Vibration of a Simply Supported,‘
thickness plates of arbitrary shape. Isotropia, Non-HQmogeneous, Rectangular
Hate. Arch. Mech. Stos., vol. 12, 1960, pp.
499-52 1.
B 1.16. SOLECHI, R. : The Non-Homogeneous Isotropic
Rectangular Plate With Arbitrary Boundary
11.1. TIMOSHENXO, S.; A N D WOINOWSKY-KRIEGER, S.: Conditions. Bull. Acad. Po!. Sci., Ser. Sci.
Theory of Plates and Shells. Second ed., Tech., vol. 9, no. 6, 1961, pp. 329-335.
McGraw-Hill Book Co., Inc., 1959. 11.17. MAZURKIEWICZ, Z.: A Certain Solution of the
11.2. CONWAY, N. D.: Some Special Solutions €or the Dynamic Problem for the Disc Made of
Flexural Vibration of Discs of Varying Thick- Elastic Orthotropic and Non-Homogeneous
ness. 1ngr.-Arch., vol. 26, 1958, pp. 408-410. Material. Bull. Aoad. Pol. Sci., Ser. Sci.
11.3. CONWAY, M. D.: An Analogy Between the Tech., voB. 11, no. I, 1963, pp. 17-25.
Flexural Vibrations of a Cone and a Disc of 11.18. MAZURKIEWICZ, Z.: Free Vibration of an Iso-
Linearly Varying Thickness. ZAMM, vol. 37, tropic Non-Homogeneous Rectangular Plate.
no. 9/10, SeptJOct. 1957, pp. 406-407. Bull. Acad. Poi. Sci., Ser. Sei. Tech., V O ~ . 8,
11.4. CONWAY, H. D.; BECKER,E. C. 33.;A N D DUBIE, no. 2, 1960, pp. 63-68.
3 . F.: Vibration Frequencies of Tapered Bars 1I. 19. SIVEHIKOV, B. E.: Equilibrium Equations for
and Circular Plates. J. Appl. Mech., vol. 33, Static Bending and Free Vibration of Irregu-
no. 2, June 1964, pp. 329-331. larly Shaped Cantilever Plates. Izv. Vyssh.
11.5. THURSTON, E. G.; A N D TSUI, Y. T.: On the Uchebn. Zavedenii: Aviat. Tekh., no. 2, 1962,
Lowest Flexural Resonant Frequency of a pp. 86-94. (Transl. by Foreign Tech. Div.,
Circular Disk of Linearly Varying Thickness Air Force Systems Command, Wright-Patter-
Driven a t Its Center. J. ASA, vol. 27, no. 5, son Air Force Base, Ohio.)
Sept. 1955, pp. 926-929. 11.20. SLOBODIANSKI, M. 6.:Estimates of Natural
11.6. MOVALENKO,A. D.: Circular Variable-Thickness Frequencies of Vibration of Clamped Plates of
Plates. Fitzmatgiz. (Moscow), 1959. (In Constant and Variable Thickness. Seminar
Russian.) on Problems of Continuum Mechanics (Phila-
18.7. KAZANTSEVA, 6. E.: On the Vibrations of Cir- delphis, Pa.), Soc. Ind. and Appl. Math., 1961,
cular Plates of Variable Thickness. Priki. pp. 473-482.
308437 G-70-20
The effects of the folBowing complications will the present section some of the papers that
be considered in the present chapter: deal primarily with this problem %ill be sum-
(I) Surrounding media marized. The topic is generalized to include
(2) Large deflections other media in addition to air-notably, water.
(3) Shear deformation an6 r o ~ a r yinertia
I 2 . i -1 Circular P l ~ t e ~
(4) Sonhomogeneity
I n an early paper Lamb (ref. 12.1) considered
Generally, because of the complexity of the a damped circular plate which i s in contact; on
resulting theory, there are not many numerical one side with an infinite expanse of water. The
results showing the effects of these complica- Rayleigh method is used with EL deflection
tions. Indeed, in many cases the technical iit- 2+--
*LilGtiOr,
erature deals mainly with the development of w= Ci; - (r/a)232 012~:;
the needed theory~ Nevertheless, it w i l l not be
the purpose of this chapter t o repeat those deri- The kinetic energy is computed on the assramp-
vations ; the reader is referred to the references &ion that the water i s incompressible The
LheEselves for this. The primary purpose of resulting formula f o r the fundaniental fre-
this chapter, as of the preceding ones. is zhe quency parameter is
pmsentstioz 02 ii Uluericd resints, where avail-
able, with explanatory material as necessary for
ars understanding of their significance.
I t d l be assumed in this chapter that the
reader d l already be reasonably familiar with
the coordinate systems, notation, boundary where y,/y is the dimensiodaess ratio of the mass
conditions, and SO forth, used in the preceding density of water to t h a t of the plate and a h
chapters and so mucb tedions redefiiritigz d l E! iE,e radiils-thickness ratio.
be omitted. tion (12.2) can be applied ~ U Q any incompressible
f h i d If both sides sre ;a Sa exposed tc the
F suww@ru m3nite fluid, &en the 0.6689 in equation (12.2‘
geceral. :4 has been the prac;ice in his is replaced by 2XO.6689.
work to discuss plates in bending which are T h e frequeocy of she second mlodo, {hmiag
srL;cou$ei “0x1 0 Aer e1:IELstic St;ructllles havzng one sodal diameter) was also ~akcu1ated in
mass. 7r, t h i s way o d y a single differential :.efesence 32.1 -ql_tlathe zse of
equmior; of motion--that of the plate-is
involved. P e t it is apparent that practical
experiments are conducted in air, and that the
.;?ass cf the ar i k ~ moved
s has the effect sf
decreasing the vibration frequencies of the
system The diEerence S&rneeE experircental
arid theoretical results for this reason Las beeri
alluded to iii many places in the preceding
chapters and, indeed, corrections of one or %e Hescs. the eEeet ~i toe vir~:er’s zrrertia is less
ather tti obtain comparabie vaims were evez u p n the second mom than upon the &st
made in a few places (and so identified; in In order to check the accrasacy of she foregoing
299
300 VIBRATION OF PLATES
results, a two-term Ritz solution was carried is perfectly attached to a massless, elastic, i d -
out in reference 12.1 for the first and second nite half space. The Rayleigh-Ritz approach is
modes of a particular iron plate; this calcula- used, with the potential energy of the half space
tion yielded results which differed from those being added to that of the plate. Clamped,
calculated from equations (12.2) and (12.4) by free, and simply supported edge conditions are
less than I percent. The effects of damping considered for the plate. For the clamped case
due to the water are also discussed. a deflection function for the plate is taken in
Experimental results for the preceding prob- the form of equation (12.1). The square of the
lem are given in reference 12.2. fundamental frequency parameter is found to
McLachlan (ref. 12.3) extended Lamb's be :
work to the case of a circular plate having
a free boundary. For a plate having both sides w"a4p= 3$Jj.7+ 4.378( 1 -T') a3 (12.11)
D D
immersed in an infinite fluid, 'ne shows that the
where
ratio of the frequency of the system w to the G
frequency of the plate in a vacuum wo can be +2,- (12.12)
x+2G
determined from the formula
and X and G are Love's (ref. 12.6) elastic
(12.5) constants for the half space:
where, for the case of one nodal circle,
MI=-16 yfa3 (12.6) (12.13)
35
and, for the case of a point support a t the
center, For the free plate, a two-term solution function
(12.7) is assumed as a constant plus the first term
of a Dini series; that is,
and, in both cases,
w = A o + A ~J0h
Wa>3 (12.14)
7r
M , = ya2h (12.8)
whereXlis thefistroot of Jo(X)=O. By applying
where yf is now the mass density of the the Ritz method, the two resulting fqequencies
surrounding fluid. I n reference 12.3, equation are determined from
(12.5) is applied to the problems of an aluminum
plate vibrating in either air or water.
The previous work was extended further by
Peake and Thurston (ref. 12.4), who applied
the Rayleigh method to the problem of the
simply supported circular plate having water
loading on one side. A deflection function
20 =1-1. 245 (ria) '+ 0.245 (ria) (12.9) (12.16)
W2a4p- __ 21.70+
__ 11.25,,+ 7.4608 ( P 2.18)
D
72.1.2 Rectangular Plates
An interesting experimentd and theoretical
study of the vibration frequencies of rectangular
cantilever plates (see fig. 4.40) immersed in
water was reported by Lindholm, Kana, Chu,
and Abramson (ref. 12.7). cyclic frequencies
are listed in table 12.1 for 15 plates made of type
1080 cold-rolled steel having various aspect
ratios and ratios of thickness to width. Theo-
retical values are based upon Barton’s work
(see discussion of rectangular cantilever beams
(sec. 4.3.12)), where applicable, and elementary
beam theory. These pertain, of course, to the
case of a vacuum. Frequencies are measured
both in air and in water.
A correction formula of the form given in
equation (12.5) was derived in reference 12.7 by
means of hydrodynamic strip theory to account 8.6 8
h
Thickness Ratio 5
for the added “apparent mass” of the surround-
ing fluid. The ratios M J M , to be used in
equation (12.5) are given in table 12.2 for six FIGURE12.l.--Thickness correction factor of a rectan-
modes of the cantilever plate (see definition gular cantilever plate for modes 1, 5, and 6. (-4fter
ref. 12.7)
of modes in table 12.1).
A further correction is suggested in reference
12.7 to account for the effect of plate thickness The variation of node-line location in going
on the apparent mass of the air. I n this case from air to water is shown in figure 12.3.
equation (12.5) becomes Frequency variation with depth below the sur-
face is set forth in figure 12.4 for plate I1 of
w
_- table 12.1. Finally, the effect on frequency
- (12.19)
Wo due to partial immersion is shown in figure 12.5
for plate 8. It is stated in reference 12.7 that
where the angle of inclination of the plate to the sur-
face seems to have an effect only for very
( 12.20)
shallow angles.
Greenspon (refs. 12.8 and 12.9) has proposed
and R is obtained from figure 12.1 for modes I,
3, and 6.
a correction formula to account for the effects
of water on one side of rectangular plate for
A comparison of theoretical and experi-
all boundary conditions. The frequency ratio is
mental results for frequency parameters is
made in figure 12.2 for the six modes. The
eiects of corrections for aspect ratio AB, afb and
thickness ratio blh are clearly seen.
302 VIBRATION OF PLATES
?A
r(.
4
.
c
O T N E R CONSIDERATIOKS 303
TABLE12.2.-Mass Correction Faetors for Ea.
(12.6)
0.6904
0
-i Bii Aij
0. 4053
0
-l
B,i
0. 25
.25
.3023 . 1351 .25
. 1924 .0810 .25
.3023 . 1351 25
0 0 . 25
. 1324 .0450 . 25
rectangular plates (ch. 4)) ; that is, boundary conditions of the plate. Thus, for
example, the tern ‘(simply supported” is no
(12.23) longer completely definitive, for the degree of
and the dimensions of the plate in the x- and restraint placed upon the two inplane 60111-
ponents of displacement must also be specified.
y-directions are a and b, respectively. The
In deriving the equations of equilibrium in
integrals given by equations (12.22) are readily the appendix the assumption is made that the
evaluated by means of the tables of reference slope of the middle surface relative to its
12.10. The coefficients for seven modes of rec- undeflected plane remains small in order that
tangular plates having all edges clamped or the sines of the angles between the normals
spPnply supported are given in table 12.3. of the deformed and undeformed middle sur-
In reference 12.8, equation (12.21) was shown faces can be replaced by their tangents bw/bx
to become and bwfby and the cosines can be replaced by
‘ty. This assumption is usually retained in
0
-
large deflection theory of plates and
L-
( 12.24) gives equiTlibsiuan equations (AS), (A.6), and
WQ
4.0 22
3.5 20
0 .5
0 6
1 I
0 LO 2.0 3.0 4.0 5.0
0 I 2 3 4 5
a/b
(5) a/b
OTHER CONSIDERATIONS 305
65
60
55
50 2 3
L
0
30 45
c
E
25 40
u
0
g 20 35
lz
.)
15 30
4 5
25
20 Air
Water
-
-----
FIGURE12.3.--Comparison of node-line locations in air
15 and water. Plate 10; a/b=l; h/b=0.0131. (After
6
(e ) a/b ref. 12.7)
FIGWRE
12.2.-Experimental and theoretical frequency
parameters in water, air, and a vacuum. ( a ) hfode 1. Another equation is obtained from the equa-
( 6 ) Mode 2. ( c ) Mode 3. (d) Mode 5. ( e ) Modes tion of compatibility of strains for the middle
4 and 6. (A4ffterref. 12.7)
surface. By using equations (12.251, this is
found to be:
displacement equations (A.11) are generalized
to include terms of the next order; that is
- - o bv
*o- au b2w bwbw
22-+--
=u- ax by axby a~ay
(12.27)
Equations (12.25) are then substituted into
equations (A.18) or (A.19) and then into
equations (A.6). It is found that the additional
terms in equations (12.25) which are even in
z drop out in the bending moment integrations, is introduced. This guarantees that the inplane
namely, equations (A.20(d)), (A.2Q(e)), and equations of motion (eqs. (A.5)) are identica
(8.20(f)), leaving the fourth-order equilibrium satisfied. Substituting equations (12.2'5) into
equation (A.27) unchanged. equation (12.26), using equations (A.19))
306 VIBRATION OF PLATES
FJGURE
12.6.-Virtual mass function for rectangular
plates. (After ref. 12.8)
3.
b2W
-2- (12.29)
a x ay
It is observed that equations (12.28) and (12.29)
are both nonlinear.
Equations (12.28) and (12.29) were derived
for the static case by Von Ki8rm&n (ref. 12.11).
They were extended to the dynamic case and
generalized further by Rerrmann (ref. 12.12).
19.9.1 Circular Plates
60 Wah (ref. 12.13) used the Berger (ref. 12.14)
0 8.25 0.50 0.75 Loo0
Depth Rotio S/a
simplieation of the Von E&rm&nequations to
study the problem of the circular plate having
FIGURE12.5.-Frequency change for surface-piercing
plates. Plate 8; a/b=5; h/b=0.0238. (After ref. either a clamped or simply supported boundary.
12.7) The pfate is constrained against inplane dis-
OTHER CONSIDERATION'S 307
placement a t the boundary in both cases. The
differential equation to be solved is
a2w
DV4w-NV2w+p ----0 (12.30)
at2-
(12.31)
"
0.2 0.4 0.6 0.8 I .o
Amplitude
Thickness 0.2 0.4 0.6 0.8 I 0
Ampiitude
FIGURE 12.8.-Nondmensional bending strew in large- Thickness
amplitude vibrations of circular plates. (After ref.
12.13)
FIGURE
12.9.-Nondimensional membrane stress in
large-amplitude vibrations of circular pkates. (After
ref. 12.13)
Deflection functions were taken in the form
where C, is a constant determined from the
inplane boundary conditions of equations
where C, and C2 were chosen to satisfy the (12.36); that is,
transverse boundary conditions exactly; that is,
Cases I(a)and II(a): I
Case H : i
C,=-,(3C~+4Q~C~+2C~)
G,=- IS-v (12.38)
5+v Gases I(b) and HI(6):
Case 11: c;=--2
C,=l c,=- 24( 11-v) 13(3-Y > c:
Substituting equation (12.37) into the first of +4( 5- Y) c,C,+2( 7 -v)Q;1
equations (12.35) and letting (12.41)
4 = A T > .i-2@> (12.39)
give
" Finally, the Galerkin technique is applied to
approximate the second of equations (12.35) ;
the result is the ordinary differential equation
(12.48) e+&
+0 2
dt2
7 3= 0 (1 2.42)
OTHER CONSIDERATIONS 309
where and p2 are given in table 12.4 for the W(T, t>=R(r)+(t) (12.44)
four cases defined by equations (12.36). The
solution of equation (12.42) results in figure is taken for the first axisymmetric mode,
12.10, which shows the effect of amplitude- where R ( r ) is the mode shape of the linear
thickness ratio upon the ratio of linear-to- problem; that is,
nonlinear frequency for the four cases for v=0.3.
(12.45)
(12.43)
was examined by Massa (ref. 12.17) as an where a is the coefficient of thermal expansion,
extension of his previous work (see discussion and ( ~ T ~ a ~ / lish ~the
) ~ critical
,, value of the
of completely free circular plates (sec. 10.1.3)) parameter aT0a2/h2at which buckling occurs,
for the linear problem. A deflection function according to the linear theory; that is,
(aToa2/h2)
,,=3.62 (12.47) .,
I "
-1 ( 12.48)
%v, E I.0524h-JO.2759(aT0a2/h2)
0.2
0 0.4 0.8 1.2 1.6 2.0
Amplitude
(12.49)
Thickness
F'ncnm I2.lO.--Ratio of linear t o ::onlinear frequency
In figures 12.11 and 12.12 the square of the
as a function of amplitude/thiekness ratio for eircuiar
plates having b O U ~ d a T Y conditions defined by ratio of the nonlinear frequency to the iso-
equations (12.36); v=0.3. (After ref. 12.15) thermal linear frequency (Gi1=7.42?3Eh3/pd,
310 VIBRATION OF PLATES
is chosen, where
(12.51)
0 1 2 3 4 5 6 7
For aToa2/h2 6 (aT0a2/h2),,=3.62,the square of
-
aT,, o
h' the nonlinear frequency is approximated by
aT,d
FIGURE12.11.-Effect of temperature upon the non- 1+0.3772--3-0.3164
linear frequency of a completely free circular plate h2
for various amplitude/thickness ratios; Y = 0.3; one (12.52)
nodal circle.
and for crToa2/h2>(crToaz/hz)
cr there results
""i
b
a Too
J .
(12.57)
-
WIO
0.8 1.0 1.2
r%
(Y2)
+ $)+ (1 Y
0.95
7 5
w(x, y9t)=hr(t) cos2-a cos2Q
b (12.62)
0.65
B .o
0.9
8.%
0.7
0.6l i I
0 0.4 0.8 1.2 1.6 2. v .7
0 1.0 2.0
Amplitude Amplitude
Thickness
Thickness
FIGURE 12.17.--Effect of large amplitude on the FIGWRE12.19.-Eff ect of amplitude upon frequency €or
frequency of a GC-C-6 square plate for t h e e SS-GSS-SS rectangular plates; u= 0.3. (After ref.
cases of inplane contraint; u=O.3. (After ref. 12.25) 12.20)
308337 0 - 7 0 4 1
314 VIBRATION OF PLATES
(12.68)
y =-+-
b w bv
yL ay
(12.69)
Shear Deformation
where K~ is a constant which is introduced t o
account for the fact that the shear stresses
FIGURE12.21.--Modes of bending and shear T~~ and T~~ are clearly not constant over the
deformation.
thickness -h/2<z<h/2 as the simple kinematic
relationships, equations (l2.64), would lead
one to believe. I n Reissner’s static theory
displacement. equations (A.10), then into stress- (ref. 12.26) K~ was taken as 5/6. Mindlin
strain equations (A.191, and the bending mo- (ref. 12.28) chose K so as to make the dynamic
ments are integrated by means of equations theory consistent with the known exact fre-
(A.aO(d)), (A.20(e)), and (A.20(f)), giving quency for the fundamental “thickness shear”
mode of vibration. More will be said about
this in the fonowing discussion. {
The right-hand sides of -momeat equilibrium
equations (A.8) are made consistent with
the present theory; they become
bM!
_---___I
Ph23%
ax ay 12 at2
Substituting equations (12.65) and (12.69) It can be farther shown (ref. 12.28) that, for
into equations (12.70) and (12.71) yields the equations (12.72) t o give results consistent
fundamental set of equations for the system with equation (12.74), K~ must be chosen for
an isotropic plate to be
?=-2 (12.75)
12
u=f(z)eiot
(12.73)
v=w=o
It can be shown (ref. 12.29) that, for a plate
having infinite dimensions in the 2- and y-
directions, the exact frequency of the first ‘
0 4 8 12 16 20 24 28 32
antisymmetric mode of thickness-shear vibra- 2a/h
tion is
FIGURE 12.22.--Ratio of plate frequency ‘Go thickness-
(12.74) shear frequency for a clamped circular plate derived
from classical theory; v=0.312. (After ref. 12.35)
OTHER CONSIDE RATIONS 317
tions (see refs. 12.33, 12.34, and 12.62), yields (1) Lowering the fundamental frequency for
a set of characteristic equations for the a given diameter-thickness ratio
frequencies. (2) Rendering more frequencies in a given
Results for the axisymmetric modes were range of w j Z for a particular plate
presented by Deresiewicz (ref. 12.35). The fre- (3) Completely altering the curves in the
quency ratio w / Z derived from the classical high-frequency range WE>1
theory of plates is plotted in figure 12.22. With The case when the circular boundary is
the use of the notation of the chapter entitled simply supported was attempted by Tomar (ref.
“Circular Plates” (ch. 2), the circular fre- 12.63). Again, when only the axisymmetric
quencies of the plate can be obtained from modes are sought, only two boundary condi-
tions are required; namely,
W & ~ = A ? ( i = l , 2,. . . ) (12.77)
w(a)=M,(u)=O (12.79)
where Xi are the eigenvalues determined from
the characteristic equation. By using equa- In reference 12.63 the equations of motion (eqs.
tions (12.74) and (12.77), it is easily seen that (12.72)) are retained in rectangular coordinates,
the ratio of the plate flexural frequency to the and their finite-difference equivalents are writ-
thickness-shear frequency E is ten. Because of the choice of coordinate
system, a rectangular finite-difference grid must
be fitted t o a sector of the circular plate. This
(12.78)
W is accomplished by using nine mesh points
within one octant obtained from a square grid
where the subscript i on w and X has been having elements of dimension 4 4 . Fundamental
dropped but is implied. Figure 12.22 is conse- frequency parameters 4w2a2p/Eh for various
quently a plot of equation (12.78) for a par- thickness-radius ratios given in table 12.5 and
ticular value of Poisson’s ratio v=0.312. figwe 12.24 for v=O.3 are taken directly from
Figure 12.23 is a corresponding plot with the reference 12.63. In addition, the frequency
plate frequencies w obtained by the theory of
parameter w u 2 J m is presented in table 12.5
this section, although this figure is plotted over
a smaller range of w / Z , thereby emphasizing for direct comparison with the classical result
the region in the vicinity of w / Z = 1. In com- wa2JplD=4.977 (see sec. 2.1.2) which applied
paring figures 12.22 and 12.23, it is obvious that for very small values of h/a. From this com-
consideration of shear deformation and rotary parison it appears that the accuracy of the
inertia has the effects of- results given in table 12.5 and figwre 12.24 is
highly questionable.
Numerical results for the completely free cir-
cular plate were found by Mindlin and Deresie-
0 4 8 12 16 20 24 28 32
0. 2 0.43365 5. 4403
2dh
.4 I. 44326 4.96242
FIGURE 12.23.-Ratio of plate frequency t o thickness- .6 2.53474 4. 38426
shear frequency for a clamped circular plate derived .8 3. 49852 3. 86308
from the MindIin theory; v=0.312. (After ref. 12.35) I
318 VIBRATION OF PLATES
w=wl+wz f
where
A,=2[1$g- (- I)'B,]--l ( j = h 2)
B 1=[ (1-g) +49 (;/a,) 2]1'2 (j =1, 2)
g=K y 1 -v)/2
0 0.1 0.2 0.3 0.4
-h and where afZ is the ratio of plate frequency
20 to thickness-shear frequency used earlier in this
chapter (W is defined by eq. (12.74)), K is
FIGUREP2.24.-Fundamental frequency parameters
for a simply supported circular plate; ~'0.3. (After
given by equation (12.75), and v is Poisson's
ref. 12.63) ratio. Substituting equations (12.81) and
(12.82) into equations (12.72) results in the
three uncoupled equations
wicz (refs. 12.33 and 12.34). In this case the (V2+6gw1=0
boundary conditions are
(V2+6t)wz=o (12.83)
M, (a)=M,e (a)=&, (a)=0 (12.80) (V24-,/2)H= 0
*r
The twisting-moment condition is identically
satisfied by symmetry for the axi
modes. In reference 12.34 frequency param-
eters for axisymmetric modes were deduced
when v=0.312. Plots of the frequency ratios
w/W discussed earlier in this section are depicted
in figures 12.25 and 12.26 for the classical
theory and the Mindlin theory, respectively.
Results for the antisymmetric modes (having
one nodal diameter) were computed h refer-
ence 12.33 and are presented in figure 12.27,
again for v=0.312. 2 o/h
In reference 12.64, C ~ ~ ~ usedh a the
n MindIin
theory t o derive characteristic determinants FIGURE12.25.-Ratio of plate frequency t o thickness-
shear frequency for the axisymmetric modes of a
corresponding to eight separate sets of con- completely free circular plate derived from classical
tinuous boundary conditions for circular plates. theory; u=Q,312. (After ref. 12.34)
OTHER CONSIDERATIONS 319
2o/h
FmuRE 12.26.-Ratio of plate frequency to thickness-shear frequency for the axisymmetric modes of a completely
free circular plate derived from the Mindlin theory; v=O.312. (After ref. 12.34)
2 a/h
FIGURE
12.27.-Ratio of plate frequency to thickness-shear frequency for the antisymmetric modes of a completely
free circular plate derived from the Mindlin theory; v=0.312. (After ref. 12.33)
means of the approach just given. In terms of In figure 12.28 (taken from ref. 12.36) the
a coordinate system Z , ?/having its origin at the three sets of frequency ratios given by equations
center of the plate (cf. fig. 4.4), the boundary (12.89) are plotted against the length-thickness
conditions are : ratio as a function of the parameter r&, where
w=Mz=I),,=O (on 5 = f a / 2 ) ( s ~ ~ ~ / b ~ ) ](j=
~$~=[ry+ ” ~1, 2, 3) (12.91)
(12.85)
W=M,,=J/~=O (on y= fb/2) and where v=0.312. From figure 12.28 it can
It is easily seen that be seen that for a given plate and for a given
mode number j the frequencies are ordered
according to w1<wa<w2 and that w2 and w 3 are
much greater than w1 except for very thick
plates.
In figure 12.29 (taken from ref. 12.36) a
more detailed plot of the frequency ratios is
indicated in the vicinity of w / Z = 1 for a fixed
ratio s,h/b=0.2 and for v=0.312. This cor-
responds to the particular case when the dis-
tance in the y-direction between node lines
(including the boundaries) is five times the
Substituting equations (12.86) into equations plate thickness. In this figure ~ ~ = mTa=n, ,
(12.85) gives and r3=p; that is, the curves m,n, q=constant
give the frequencies of the wl,w2,and H modes,
cuj=rjn/a respectively. In this case, each mode has a
(12.88)
&=sp/b (j=l, 2, 3) low-frequency cutoff given by
2
=I+@
(12.89)
where
d h
(12.90)
FIGGRE 12.28.--Ratio of plate frequency t o thickness-
shear frequency for a SSSS-SS-SS rectangular plate
derived from the MindHin theory; v=0.312. (After
ref. 12.36)
OTHER CONSIDERATIONS 321
1.20 -n=m
1.15
e3
1.10
1.0704
i.05
1.0198
0 2 4 6 8 IO 12 14 16 18 20 22 24 26 28 30 32 34 36
O/h
FIGURE12.29.-Ratio of plate frequency to thickness-shear frequency for a SS-SS-SSSS rectangular plate when
the distance between nodes along the width is five times the thickness; ~ = 0 . 3 1 2 .(After ref. 12.56)
where +: and Q: are given by equations with classical theory and will not be repeated
(12.90) with r,h/a=O. These formulas give here.
the values 1.0198 and 1.0704 shown in figure A stiffened plate was treated as an ortho-
12.29. tropic plate for purposes of analysis in refer-
The mode shapes corresponding to wl, wz, ence 12.67. The effects of rotary inertia were
and H are depicted in figure 12.30 (taken from considered, but shear deformation was ignored.
ref. 12.36). The mode shape corresponding to In this case the system of governing differential ~ .
classical theory is shown in figure 12.31, which equations remains fourth order. Equation
is also from reference 12.36. Of the three (9.22) is generalized t o
modes, the w1mode most closely resembles the
classical mode; heme, it is called a “ffexuraJ”
mode. As a/h+m, this mode approaches the
classical mode, and its frequency approaches
the classical frequency given by equation
(4.20). For the wzmode the thickness-shear
where 1, and I, are the moments of inertia of
deformation predominates. The N mode shape
(fig. 12.30(c)) contains no average deflection, the stiffened plate about axes parallel to the
y- and z-directions, respectively.
but twists the plate; hence, it is called a “thick-
ness twist” mode. For a rectangular plate simply supported
I n references 12.65 and 12.66 the problem of along the edges x=0, a and y=O, b, it is appar-
the simply supported plate is attacked by the ent that the boundary conditions will be
finite-difference method. Mindlin’s equations satisfied by the deflection function
are the basis for this method in reference 12.65, W ( Z , ~ J ) = Sa2~ ,@y (12.94)
whereas in reference 12.66 an alternate set is
used. Numerical results for frequencies are where ( ~ = m ~ / Pa =
, n ~ l b , and m and n are
given in both papers, but they are inconsistent integers. Substituting equation (12.94) into
322 VIBRATION OF PLATES
o=b=Il"
hs=0.275''
-l' y e -
h< 0.065"
a; 0.063"
;a 0.625''
7
c FIGURE12.32.-Dimensions of stiffened plate. (After
A ref. 12.67)
1 1 1 2 3 4 5
Q /h
FIGURE12.33.-Ratio of plate frequency to thickness-shear frequency for a SS-F-SS-F rectangular plate when the
distance between nodes along the direction parallel t o the free edges is five time the thickness; v=0.312.
(After ref. 12.36)
324 VIBRATION OF PLATES
Mathieu functions. The boundary conditions The necessary modifications of the theory are
are satisfied by finding the roots of an infinite discussed in the section of the appendix entitled
determinant, each element of the determinant “Force and Moment Integrals” (sec. A.5). This
being an infinite series of Mathieu functions is the type of nonhomogeneity discussed in
containing the frequency within their argu- reference 12.69.
ments. In reference 12.68 the infinite charac- Bolotin (ref. 12.70) generalized the model
teristic determinants are displayed for eight for the layered plate by assuming that the plate
types of boundary conditions, but no numerical is composed of both “hard” and “soft” layers.
results are given. The hard layers obey the Kirchhoff hypothesis
while slippage occurs in the soft layers. I n the
12.4 EFFECTS OF NONHOMOGENEITY
soft layers the inplane stresses oz, uu, and T , ~
A brief survey of the literature dealing with are assumed to be zero, while the transverse
the vibration of nonhomogeneous plates will shear stresses ruz and rz2 are constant within
now be given. Nonhomogeneity may arise the layer. On the basis of these assumptions,
in many ways. Overall material properties a complete plate theory is developed in refer-
themselves may vary in a continuous manner ence 12.70. Another formulation, based upon
(e.g., a continuum representation of a fibrous the three-dimensional equations of elasticity, is
composite plate). Inclusions or holes may given in reference 12.71.
occur within the plate. As can be seen from The theoretical work of Yu on layered
earlier chapters, the effect of a “classical1’ @.e., (refs. 12.72 through 12.83) is particula
cylindrical) hole, even if small, can cause a nificant. This effort is primarily devoted to the
significant effect upon the vibration frequencies incorporation of shear deformation and rotary
of a plate. inertia effects into the layered-plate theory.
Some practical and commonly used types of It is shown that these effects, particularly shear
sandwich plates deformation, are especially important when one
de& with conventional sandwich plates com-
posed of a relatively e material con-
tained between two r
sheets. The statemen
shear-deformation effects
tant for a sandwich pl
which may be only 1 pe
corresponding solid, hom
shown furthermore that,
plates, the shear effe
inertia of the faces a
and the flexural rigidity o
beyond the scope of this work. Even with ble; of importance
these assumptions, the complexity of the core, the rotary and translatory inertias of the
results and the number of parameters required core, the translatory inertia of the faces (in-
to describe the sandwich make it impractical cluding the rotary effect of the faces about the
to report detailed numerical results in this midplane of the sandwich plate), and the flex-
seetion. ural and extensional rigidities of the faces (ref.
I n the most simple case, a layered plate is 12.75).
made up of several layers bonded together, each A one-dimensional theory was developed in
layer being homogeneous and isotropic, and the references 12.72 to 12.76, which is applicable
KirchhoB hypothesis of normds to the mid to the vibration of plates in modes of plane
surface remaining straight and normal is as- strain. The transverse displacement w, as in
sumed valid. I n this case the mathematical the Mindlin theory, was assumed to be con-
complication of the plate theory is minimal. stant through the plate thickness. The dis-
OTHER CONSIDERATIONS 325
placements in the plane of the plate are assumed theoretical values obtained by a simple analysis
to vary linearly through the thickness, with the by using the Rayleigh method.
slope in the face sheets not necessarily the I n reference 12.89 radial nonhomogeneity in
same as the variation in the core. circular plates is accommodated by treating the
The theory is generalized t o a two-dimen- plate as a composite of homogeneous, isotropic
sional variation in 20 in references 12.77 and annuli and enforcing continuity conditions
12.78 and is applied to the problem of a rec- across the internal junctions.
tangular plate simply supported on all edges. The plate consisting of a thin face sheet
In references 12.78 and 12.83, sets of formulas stiffened by corrugated sheet (see fig. 12.34) is
are presented for the calculation of natural analyzed in reference 12.90. It is shown that
frequencies of the simply supported rectangular this configuration cannot be treated as ortho-
plate. Those formulas will not be reproduced tropic plate because the twisting-moment rela-
here because of their inherent complexity tion M,,=M,, is no longer applicable. A
(arising from the relatively complicated geom- theory for this case is derived.
etry and material properties of the sandwich In reference 12.91 an inJEatable plate is
plate) and the amount of explanation which analyzed. This plate consists of two woven
would be required. cover membranes joined to each other by
The theory is extended to the nonlinear closely spaced perpendicular filaments. The
(large-deformation) domain in references 12.80 space between the covers is pressurized, and the
to 12.82. It is shown that the basic behavior filaments hold the cover membranes together
is the same as that for homogeneous plates; (see fig. 12.35). A variable-thickness plate is
that is, the membrane stiffening due to large obtained by using variable-length connecting
deformations causes the overall stiffness of the filaments. The theory developed in reference
system to be like a “hard” spring, thus causing 12.91 was applied in reference 12.92 t o obtain
an increase in frequency with increase in ampli- natural frequencies of square plates havbg
tude. (See section 12.2 of this work for back- simply supported edges. Results were com-
ground information.) I n particular, the non- pared with experimental ones.
linear theory is applied to a rectangular plate
having immovable, hinged edges.
Further theoretical derivations of equations
for the vibrational behavior of layered plates
are made in references 12.84 to 12.86. I n
both references 12.84 and 12.85 the analyses
are generalized to include orthotropic core
ma terids, and explicit frequency equations
are developed for the case of a plate simply
supported on all edges.
Experimental results for sandwich plates
Having honeycomb and Styrofoam cores are
FIGURE
12.34.--A corrugation-stiffened plate.
given in reference 12.87. Experiments were
conducted in a vacuum and data were compared WOVEN C O V E R S 7
with analytical frequencies obtained from a
finite-difference solution of the classical plate CONNECTING
equations. It was found that the classical theory FILAMENTS
12.62. REISSNER,E.: On Axi-Symmetrical Vibrations 1082 (CFSTI No. AD 234 221), Dept. Mech.
of Circular Plates of Uniform Thickness; Eng., Poly. Inst. Brooklyn, Aug. 1959.
Including the Effects of Transverse Shear Also, J. Aeron. Sci., Dec. 1960, pp. 8 9 6 9 0 0 .
Deformation and Rotary Inertia. J. Acoust. 12.77, YE, Y. Y.: A New Theory of Sandwich Plates:
SOC. Am., vol. 26, no. 2, Mar. 1954, pp. General Case. Tech. Note No. 6 (CFSTI
252-253. No. AD 234 232), Dept. Mech. Eng., Poly.
12.63. TOMAR, J. S.: On Flexural Vibrations of Iso- Inst. Brooklyn, Nov. 1959.
tropic Elastic Thin Circular Plates According 12.78. Yu, Y. Y.: Flexural Vibrations of Rectangular
to Mindlin’s Theory. Proc. Nat. Inst. Sci. Sandwich Plates. Tech. Note No. 8 (CFSTI
India, vol. 29, no. 5, 1963, pp. 552-560. No. AD 248 073), Dept. Mech. Eng., Poly.
12.64. CALLAHAN, W. R.: On the Flexural Vibrations Inst. Brooklyn, Aug. 1960.
of Circular and Elliptical Plates. Quart. 12.79. Yu, Y. Y.: Extensional Vibrations of Elastic
Appl. Math., vol. 13, no. 4, Jan. 1956, pp. Sandwich Plates. Tech. Note No. 9 (CFSTI
371-380. No. AD 248 301): Dept. Mech. Eng., Poly.
12.65. TOMAR,J* S.: On Flexural Vibrations of ISO- Inst. Brooklyn, Oct. 1960. Also, Proc. 4th U.S.
tropic Elastic Thin Square Plates According Nat. Congr. Appl. Mech., 1962, vol. 1, pp.
to Mindlin’s Theory. Proc. Nat. Inst. Sci. 441-448.
India, v01. 29, no. 2, 1963, pp. 169-179. 12.80. Yn, Y. Y.: Nonlinear Flexural Vibrations of
12.66. TOMAR,J. S.: On Flexural Vibrations of Iso- Sandwich Plates. Tech. Note No. 12
tropic Elastic Thin Square Plates. Bull. (CFSTI No. AFOSR 1310), Dept. Mech.
Calcutta Math. SOC., vol. 55, no. 1, 1963, Eng., Poly. Inst. Brooklyn, Feb. 1962. Also,
pp. 1-10. Acoust. SOC. Amer., vol. 34, 1962, pp. 1176-
12.67. THORKILDSEN, R. L., A N D HOPPMANN, W. H.: 1183.
Effect of Rotatory Inertia on the Frequencies 12.81. Yu, Y. Y.: Effect of Thickness Deformations
of Vibration of Stiffened Plates. J. Appl. on Vibrations of Sandwich Plates. Tech.
Mech., vol. 26, no. 2, June 1959, pp. 298-300. Note No. 13 (CFSTI No. AD 279 509),
12.68. CALLAHAN, W. R.: Flexural Vibrations of Ellip- Dept. Mech. Eng., Poly. Inst. Brooklyn,
tical Plates When Transverse Shear and Feb. 1962.
Rotary Inertia Are Considered. J. Aeoust. 12.82. Yu, Y. Y.: Application of Variational Equation
SOC. Am., vol. 36, no. 5, May 1964, pp. of Motion to the Nonlinear Vibrational
823-829. Analysis of Homogeneous and Layered Plates
12.69. PISTER,K. S.: Flexural Vibration of Thin and Shells. Tech. Note NO. 14 (CFSTI No.
Laminated Plates. J. Acoust. SOC. Am., AD 289 868), Dept. Mech. Eng., Poly. Inst.
vol. 31, no. 2, Feb. 1959, pp. 233-234. Brooklyn, Feb. 1962. Also, J. Appl. Mech.
12.70. BOLOTIN,V. V.: Vibration of Layered Elastic V O ~ . 30, 1963, pp. 79-86.
Plates. Proc. Vibration ProbI.. vol. 4, no. 4,
12.83. Yu, Y. Y.: Damping of Flexural Vibrations of
1963, pp. 331-346.
Sandwich Plates. J. Aeron. Sci., July 1962,
12.71. MOSKALENKO, V. N.: Free Vibrations of a
pp. 790-803.
Three-Layered Plate. Izv. Akad. Nauk
SSSR, Otd. Tekh. Nauk, Mekh. i Mashin., 12.84. CHANG, C. C.; A N D FANG, B. T.: Transient and
no. 4, July 1962, pp. 125-129. (In Russian.) Periodic Response of a Loaded Sandwich
12.72. Yu, Y. Y.: A New Theory of Sandwich Plates: Panel. J. Aeron. Sci., May 1961, pp. 382-396.
One Dimensional Case. J. Appl. Mech., 12.85. BIENIEK,M. P.; A N D FREUDENTHAL, A. M.:
V O ~ .26, 1959, pp. 415-421. Frequency-Response Functions of Qrtho-
12.73. Yn, Y. Y.: Simple Thickness-Shear Modes of tropic Sandwich Plates. J. Aeron. Sei.,
Vibration in Infinite Sandwich Plates. Sept. 1961, p p ~ 732-735, 752.
AFOSR TN 58-897 (CFSTI NO. AD 204 132), 12.86. MEAD, D. J.; A N D PRETLOVE, A. J.: On the
Dept. Mech. Eng., Poly. Inst. Brooklyn, Vibrations of Cylindrically Curved Elastic
Oct. 1958. Also, J. Appl. Mech., vol. 26, Sandwich Plates: Part I, The Solution for
Dec. 1959, pp. 679-680. Flat Plates, and Part 11, The Solution for
12.74. Uu, Y. Y.: Flexural Vibrations of Elastic Cylindrical Plates. R. & M. No. 3363,
Sandwich Plates. J. Aeron. Sci., Apr. 1960, British A.R.C., 1964.
pp. 273-282, 290 (CFSTI NO. AD 211 219). 12.87. POWELL, C. A., JR.; A N D STEPHENS,D. 6.:
12.75. Yn, Y. Y.: Forced Flexural Vibrations of Vibration of Sandwich Panels in a Vacuum.
Sandwich Plates in Plane Strain. AFOSR Thirty-fifth Shock and Vibration Symp.
T N 59-567, Dept. Mech. Eng., Poly. Inst. (New Orleans, La.), Qct. 25-28, 1965.
Brooklyn, July 1959. Also, J. Appl. Mech., 12.88. THOMPSON, W. M., JR.; A N D CLARY,R. R.: An
vol. 27, no. 3, Sept. 1960, pp. 535-540. Investigation of the Natural Frequencies
12.76. Yu, Y. Y.: Simplified Vibration Analysis of and Mode Shapes of Double Conical Sandwich
Elastic Sandwich Plates. AFOSR TN 59- Disks. NASA TN D-1940, 1963.
OTRER CONSIDERATIONS 329
12.89. V O D I ~ K A
V.:
, Free Vibrations of a Composite 12.91. MCCOMB,H. G., JR.: A Linear Theory for
Circular Plate. Acta Phys. Austriaca, vol. Inflatable Plates of Arbitrary Shape. NASA
17, no. 4, Apr. 1964, pp. 319-332. (In T N D-930, 1961.
English.) 12.92. STROUD, W. 9:. Experimental and Theoretical
12.90. FUNG,Y. C.: On Corrugation-Stiffened Panels. Deflections and Natural Frequencies of an
GALCIT SM 62-33 (CFSTI NO. AD 429 Inflatable Fabric Plate. NASA T N D-931,
7701, Calif. Inst. Tech., June 1962. 1961.
,3084370--70---22
APPENDIX
e ions
The purpose of this appendix is to present moment equilibrium equations 7zy= TU3? ryr=
the notation, conventions, assumptions, and T , ~ ,and T,= T,, (neglecting couple stresses) have
fundamental equations upon which the main already been introduced in figure A. 1.
past of this work is based. The effects of Figure A.2 shows a plate element of thickness
(1) Anisotropy h and incremental dimensions dx and dy. The
(2) Inplane forces x- and y-axes are chosen to contain the unde-
(3) Variable thickness formed middle surface of the plate. This plane
will be explicitly included. Where other com- is called the “neutral plane.” More will be
plicating effects (e.g., large deflections) enter said later about its location when layered
the formulation, they will be pointed out. plates are discussed. For a plate homogeneous
Basic derivations are, for the sake of simplicity, through its thickness, the neutral plane lies
carried out in rectangular coordinates. midway through its thickness. The z-axis is
normal to the undeformed middle surface.
A.l NOTATION The z-axis is shown, for convenience only,
A notation will be developed which is con- as acting along one edge of the element. Thus,
sistent with that of elasticity theory; that is, a t it is noted that the xyz coordinate system i s
a point the directions of positive stress space $xed. The transverse shearing force
taken as shown on the element of figure A.1. intensities Qz and Q,, the inplane normal
Positive normal stresses are tensile. Positive and shearing force intensities N,, N,, and
shear stresses are directed in the positive x-,y-> N,, and their incremental changes are shown
and z-directions if they lie on “positive faces” acting on the sides of the element, with positive
of the element; that is, those faces of the three forces acting in positive directions on positive
parallel sets whose 2-, y-, and z-coordinates are faces. These quantities have dimensions of
the largest. The three well-known (ref. A.1) force per unit length. As will be seen later,
fhese forces arise from the integrals of the
even componen!~of positive normal and shear-
FIGURE
A.2.-For0ea (intensities) acting on a plate element.
FIQURE
A.3.-Moments (intensities) acting on a plate element.
right-hand vectors in the figure. These quan- because the shear stresses causing them are
tities have dimensions of moment per unit identical.
length. As it will be seen later, these moments The middle surface of the element after
arise from the integrals of the odd components deformation is shown in figure A.4. The origin
of positive normal and shearing stresses. These of the space-fixed coordinate system is taken
stress variations are depicted typically on at one corner of the element for convenience
two faces of the element. The twisting moments only. The displacemedt in the z-direction i s
Mz,are identical on the faces x=O and y=O taken as w. Slopes, along with their incre-
APPENDIX-PLATE EQUATIONS 333
Z,W
ay [ab
y +
a x da y dx] d y
FIGURE
A.4.--Deformed middle surface of a plate element showing slopes and their changes.
mental changes, are shown at all corners of where p is mass density per unit area and
the element, with positive changes assumed in b2w/dt2 is the acceleration in the z-direction.
positive directions. For small displacements The technique of generalizing the above equa-
it will be assumed later that the slope (tangent tion to account for large deformations (slopes)
of the angle) and the sine of the angle are is self-evident. Expanding the terms involving
equivalent.
products, discarding resulting third-order differ-
A.2 EQUILIBR~UMEQUATIONS ential terms, dividing through the equation by
Considering small deflections (or, more the area dx dy, and simplifying yield:
precisely, small slopes), summing forces in the
z-direction yield the equation (refer to figs.
A.2 and A.4)
where u,v, and w are displacements in the order than the others and the equations
5-, y-, and z-directions, respectively, and p* is simplify t o
mass density per unit volume. When the
inplane inertia forces within the plite are
neglected and the transverse shearing stresses
T , ~ and T,, are small relative to the other
stresses, the first two of equations (A.3) become:
where the terms on the right-hand sides ac-
count for the rotary inertia of the plate element
and are customarily considered small relative
t o the remaining terms in the equations.
Because these equations must be satisfied for The moment equation about the z-axis is
every infinitesimal thickness (dz) of the plate identically satisfied.
element, their integrals over the thickness must A.3 KINEMATICS OF DEFORMATION
also be satisfied. That is,
The assumption of elementary beam theory
that “plane cross sections remain plane” is
generalized to apply t o a plate as follows:
Normals to the midplane of the undeformed plate
remain straight and normal t o the midplane during
deformation.
By use of equations (A.51, equation (A.2) now
simplifies to An edge View of a portion of a plate is shown
in figure A.5. The undeformed position of the
plate is shown in solid lines, while the deformed
shapeis shownin broken lines. The longitudinal
elastic displacement (due to inplane forces) of
a point‘I on the midplane is depicted as uo.
Points such as 0 not falling on the midplane
If one were to sum forces in the 2- and y- will also have, in general, displacement due t o
directions, he would arrive a t the following rotation of the normal Thus, the longitudinal
equations : components of displacement of points within
the plate will be characterized by
bA-7)
Sane inertia forces will be considered t o
small, as before. If the transverse shearing
forces are small relative to the inplane forces,
and the slopes are d s o considerably less than
unity, then terns of the t y p e Q,(dwfdz) cac
certainly be considered negligible compared
with terms of the type N,, for example. Equa- E
e --
av (A.10)
v-ay
av au
?xv=&+G
where yzv is engineering strain as differentiated
from the tensorial strain required €or tensorial
manipulations. Substitu ting equations (A .9)
into equations (A.lO) gives
bu b=W
e -LzL.-
"-ax a22 and where la/ is the determinant
(A.11)
(A. 16)
where
Eh3
D = 12(1-v2)
(A.21)
PIGUEE
A.6.---layered plate.
APPENDIX-PLATE EQUATIONS 337
the upper surface is denoted by c. As before, where
the transverse coordinate z w i l l be measured D,,=v@z+2D, (A.26)
from this neutral plane.
The force and moment integrals will be Combining equations (8.25) with equation
formulated for the layered plate of figure A.6, (A.6) gives the equation of plate bending
which has three layers. Extension of this
a4w iYW a4w a2w
formulation t o other numbers of layers is Dx baP +2Dzu h x g +O,-+p*a$
~
a plate normal t o the x-direction is shown in which K,=O and K,J,approaches infinity. This
figure A.7. Translational and rotational last condition is physically possible but receives
springs having stiff nesses K , and K+, virtually no treatment in the literature on
respectively, are attached t o the edge. The plates.
force K,w required to deflect the translational The meaning of the “edge reaction” will now
spring in the positive direction and the moment be discussed. I t would appear that for a free
K*&w/dxrequired t o cause a positive rotation edge normal to the y-direction all three quanti-
are shown, along with their reactions on the ties Mu,M,,, and &, would be zero. However,
edge of the plate. The “edge reaction” V , as discussed previously, only two boundary
and the bending moment M , occur at an in- conditions are admissible per edge. It is found
finitesimal distance within the plate (the “edge that Q, and M,, combine into a single edge
reaction” is discussed later). By summing condition as wiyill be described now. Figure A.8
forces and moments on the infinitesimal ele- depicts a free edge parallel to the 2-direction.
ment and neglecting higher order terms such The twisting moment Mzu=MZu(x) along the
as those arising from forces and moments edge can be represented by pairs of vertical
acting on the two planes parallel to the plane forces having intensities M,, and infinitesimal
of the paper, the following equations are found changes, as shown. The vertical force resultant
to hold on the boundary: from the opposing forces is t,hus bM,,/dx in
intensity. When this is added to the trans-
V,=-K,w>
versing shearing force, the total edge reaction is
(A.29)
(A.30)
The inplane force component N , does not enter
this equation, for it was defined t o be taken In terms of arbitrary directions normal and
always in; the deformed neutral plane. The tangent to the boundary (n and t ) , equation
generalization of equations (A.29) to arbitrary (A.30) is generalized to
edge directions is accomplished by using n in
place of x, where n is the direction of the outer
normal t o the edge.
Special cases arise when the spring constants
K, and/or K+ are zero or infinity. When
K,=K+=O, the edge is completely free. When
both K , and K+ approach infinity, the edge
becomes clamped. When K#=O and K , ap-
proaches infinity, the edge becomes simply
supported. The last possible ease is %hat in
7
h
iMX
FIGURE
A.T.-Elastic edge constraints. FXGURE A.g.--Twisting moments along an edge.
APPENDIX-PLATE EQUATIONS 339
For further discussion of the free edge con- where the flexural rigidities are defined by
dition, seereferencesA.1 (p. 84) and A.2 (p. 17).
A,$ POLAR ORTHOTROPY
A development parallel t o that of the pre-
ceding sections may be carried out for the case
of polar orthotropy. That is, if the stresses
associated with plane polar coordinates (see
fig. 1.1) are ur7ue, and T,@ and the corresponding
strains are er, Ee, and Yre, the stress-strain
relations are given by the equae' mns When moment equilibrium equations equiv-
alent to equations (A.8) are used and rotary
1 inertia is neglected, the transverse shearing
t -- (ur-Yru@)
-- Er forces are found to be (ref. A.3)
(A.32)
Yro=Tle/G
where zd and v now identify the radial and Finally, the transverse force equilibrium
circumferential displacements. The strain- equation gives the governing differential equa-
displacement equations become $ion of motion
dU
e --
'- br
where the integral is taken over the volume of where the remaining integral is yet to be taken
the body. Restatement of anearlier assumption over the plate area, and where D,, D,,D,,,and
that the transverse stresses u , , r y r , and I,, are D, are as defined previously in equations (A.22)
small relative to the others in the case of a and (A.26). For an isotropic plate, equation
plate allows equation (8.40) to reduce to (A.42) simplifies t o
Habata, Y., 58 Kumaraswamy, M. P., 144, 145, 146, 215, 216, 217
Hall, A. H., 87, 171, 172, 184 Kurata, M., 125, 126, 127, 128
Wamada, M., 47, 51, 58, 69, 71, 123, 125, 126, 127, 129, Kurlandzki, J., 129
162, 163, 205,206,208
Handelman, G., 32, 33, 141, 144 Lagrange, 148
Hanson, P. W., 184, 220 Lamb, H., 269, 274, 275, 299, 300
Hardy, H. C., 10, 11, 13, 16 Larsen, K., 45, 47, 48, 50, 51, 53, 54, 58
Hasegawa, M., 162, 163 Lauricella, G., 8
Havers, A., 212, 281 Leckie, F. A., 45
Wearmon, R. F. S., 43, 250, 252, 254, 256, 257, 261 Lee, P. C. Y., 314
Heiba, A. E., 83, 84, 196 Lee, W. F. Z., 147
Hencky, H., 316 Legendre, 104
Herrmann, G., 277, 278, 280, 306, 310, 311, 312, 313, Leissa, A. W., 58
316 Eekhnitski, S. T., 254, 261, 264, 327
Hersch, J., 58, 206, 238 Lemke, A., 87, 104
Widaka, K., 58 Lindholm, U. S., 301, 305, 306
Hilbert, D., 8 Love, A. E. H., 250, 300
Hopkins, H. G., 58 Lubkin, J. L., 217, 218
Hoppmann, W. H., 11, 123, 125, 245, 246, 247, 248, 252, Luke, Y. L., 217, 218
253, 254, 256, 257, 258, 259, 263, 264, 265, 266, 321, Lundquist, E. E., 117, 118, 119
322 Lurie, H., 46, 278, 280, 281
Hort, W., 20
MacNeal, R. H., 77
Houbolt, J. C., 86, 220
McComb, H. G., Jr., 325
HrennikofT, A., 217
McLachlan, N., 2, 3, 7
Wufington, N. J., Jr., 252, 253, 254, 256, 257, 258, 259,
McLachlan, N. W., 38, 300
261, 262
McNitt, R. P., 38
Magness, L. S., 252, 253, 254
Iguchi, S., 46, 47, 51, 58, 59, 60, 62, 63, 89, 91, 102, 103,
Makarov, B. P., 62, 64, 122
104, 123
Mansfield, E. H., 337
Irons, B., 87
Martin, A. I.,79, 80, 108
Pwato, 69, 71, 125
Martin, C. J., 271
Martin, H. C., 227
Janich, R., 42, 43, 58, 61, 136 Massa, N. E., 275, 309, 310
Jankovic, V., 54 Massonnet, C., 379
Joga Rao, C. V., 20, 23, 58, 122, 151, 152 Mathieu, 324
Mazurkiewicz, Z., 296
Xaczkowski, Z., 237, 238, 239, 281, 282, 316 Mead, D. J., 325
Kana, D. D., 301, 305, 306 Mengotti-Marzolla, e., 38
Xanazawa, T., 47, 48, 51, 58, 62, 63, 65, 69, 71, 256, Mindlin, R. D., 314, 315, 316, 317, 318, 319, 320, 321,
261, 262, 263 322, 323
Kantham, 6. L., 14, 122 Minkwah, I. A, 245, 246, 247, 248
Kato, W., 50, 58, 103 Mishenkov, 6. V., 62, 64, 122
Kato, T., 195, 280 Modeer, J. E., 313
Kaul, R. X., 65, 77, 161, 164, 280, 314 Molotkov, L. A., 316
Kawai, T., 47, 48, 57, 58, 62, 63, 65, 69, 71, 256, 261, Morley, L. S. D., 5
262, 263 Moskalenko, V. N., 324
Hawashima, S., 220 Munakata, K., 58
Kazantseva, G. E., 291
Khachatryan, A. A., 254 Nadai, A., 45
Kirchoff, G., 13, 14, 324 Nagaraja, J. V., 58, 73, 77, 194, 195, 216
Kirk, C. L., 133 Nakata, Y., 103
Kist’yan, K. Ya., 50, 51, 58 Narsimhamurthy, E’., 3 16
Klein, B., 193, 194, 205, 206, 208, 209, 210, 211, 227, Naruoka, M., 57, 58, 258, 259, 260
228 Nash, W. A., 2, 279, 280, 313
Koenig, M., 20 Navaratna, D. E., 150
Kondo, H., 162, 163 Newman, E. G., 103, 314
Konig, R., 87 Newmark, N. M., 46
Kovalenko, A. D., 290 Newsom, 6.D., 77, 171, 173
Bumai, T., 20, 21, 26, 151, 153 Nishimura, T., 47, 58, 69, 71, 131, 135
INDEX
Wah, T., 147, 254, 268, 269, 272, 273, 274, 306, 307, Woinowsky-Krieger, S., 14, 285, 329
308, 313, 314 Wood, A. B., 13
Waller, M. D., 11, 13, 38, 39, 104, 109, 110, 183, 184, Woodfield, N., 45, 47, 48, 50, 51, 53, 54, 58
222, 223, 229, 231, 233, 234, 238, 239, 240
Wallisch, W., 316 Yamaki, N., 307, 309, 311, 312, 313
Walton, W. C., Jr., 86, 141, 220, 227 Yonezawa, H., 258, 259, 260
Warburton, 6. B., 41, 42, 43, 58, 61, 81, 250 Young, D., 58, 59, 60, 65, 72, 76, 77
Warner, W. H., 213, 216
Yu, Y. Y., 324, 325
Weinstein, A., 58, 61, 278, 280
Westmann, R. A., 208, 212, 239,240
Wheatstone, C., 87 Zeissig, C., 54, 58
Willers, F. A,, 268, 269, 272, 273 Zorowski, C. F., 86, 122, 125, 147, 148, 150, 195, 197,
Wilson, R. E., 83, 293 198, 199, 215, 217, 218, 219, 220, 227, 228, 291
ex
345
308-337 0--7&23
346 VIBRATION O F PLATES
Polar orthotropic plate, transverse bending, 245 Rectangular orthotropy, 250, 266
Polar orthotropy, 245-250, 337 circular plates having, 263-264
flexural rigidities, 337 elliptical plates having, 264-265
kinematic relationships, 337 plates with inplane forces, 267
moment curvature, 337 rectangular plates having, 250-263
rotary inertia, 337 Rectangular plates, 41-154
strain displacement, 337 added mass, 141-151
transverse force equilibrium, 337 admissible functions, 77
transverse shearing forces, 337 aluminum, 83, 86, 89, 126, 133, 143, 148
Polygonal plates, 237-239 anisotropic, 250-266
buckling analogy, 237 anticlastic bending effects, 89
hexagons, 238 area integrals replaced by double summations, 86
membrane vibration analogy, 237 beam functions, 58, 71, 76, 81, 87, 104
octagons, 238 behavior like beam, 54, 86
parallelogram plates, 161-192 biharmonic singular function, 151
pentagons, 237-238 boundaries as nodal lines, 42
point-matching method, 238 boundary Conditions, possible combinations, 41
rectangular plates, 41-154 brass, 108, 116
simply supported, all edges, 237, 238 buckling, 45, 46, 117
trapezoidal plates, 193-196 cantilever, 76-87
triangular plates, 205-235 C-C beam, 60
Positive faces, notation, 6-C-C-C, 58-65, 280-281
Positive shear stresses, notation, 329 C-C-C-C square plate, 60
Prestressed boundary, planes with inplane forces, 273 C-C-C-F, 65
Prestretched membrane, polygonal plates, 237 C-C-C-SS, 65
Propeller blades, marine, 240, 242 C-C-F-F, 72
C-C-SS-F, 71
Quadrilateral plates 6-6-SS-SS, 65-71
of general shape, 196 C-F-C-F, 74-75
parallelogram, 161-192 C-F-F-F, 76-87
rectangular, 41-154 C-F-SS-F, 75-76
circular holes, 152
Radial sides simply supported, sectorial plates, 239 coarse finite difference grids, 47
Rate of taper, variable thickness, 291 constraint of zero deflection, 130
Rayleigh method continuity condition for transverse shear, 145
anisotropic plates, 250, 256, 258, 259, 261, 262 C-SS-C-F, 73
elliptical plates, 37, 38 C-SS-F-F, 74
inplane forces, 269, 270, 272 6-SS-SS-F, 74
parallelogram plates, 164 cutouts, other, 152
rectangular plates, 41, 43, 58, 118, 132 cylindrical masses, 148
sectorial plates, 239, 240 deflection functions, 77, 81, 119, 131, 136, 140
surrounding media, 299, 300 Dirac delta function, 147
Rayleigh-Ritz method, 20 discontinuous edge conditions, 123-130
anisotropic plates, 252, 261, 264 double-precision arithmetic, 77
circular plates, 20 elastic edge supports, 114-123
inplane forces, 275-280 electrical analogies, development of, 77
parallelogram plates, 161, 162, 168 electronic analog computer, 77
rectangular plates, 58, 59, 61, 65, 69, 72, 73, 76, equal slope restraint, 122
77,79,81,86,103,119,122,131-133,141,151,152 extrapolation formula for finite difference method,
surrounding media, 300 130, 136
trapezoidal plates, 194, 195 F-F-F-F, 87-115
triangiilar plates, 212, 213, 215, 216 finite-difference equations, 71, 86, 130
variable thickness, 288, 290 finite-difference mesh, 86
Rectangular cantilever plates, 301 finite-difference method, 58, 131, 136, 220
Rectangular coordinates, 4 finite differences replace derivatives, 86
bending and twisting moments, 4 finite summation replaces integral equation, 129
edge reactions, 4 flexural rigidity, 86
Laplacian operator, 4 Fourier sine series, 139, 145
shearing forces, transverse, 4 Galerkin method, 61, 72, 88
strain energy, 4 general rectangle, 89
INDEX 351
Rectangular plates-Continued Rectangular plates-Continued
Green’s function, 129 “veering away” phenomenon, 63, 74
half-sine waves, 47 V-groove simulation of simply supported edge, 124
high-frequency parameters, 51 Warburton’s formula, 86
hub-pin plate, 140 weight density, 86
inplane forces, 276-281 Weinstein method, 58, 61
integral equation, replaced by finite summation, Recursion formulas, 14, 20, 32
129 Regularity conditions, 17
internal cutouts, 151-154 Reissner’s static theory, shear deformation, 315
Lagrange’s equation, 148 Reissner’s variational method, rectangular plates, 140
large deflections, 310-314 Rhombic plates. See Parallelogram plates.
large error, frequency of, 73 compared to square, 163
Legendre functions, 77, 104 parallelogram, aluminum, 172
mesh widths, 130 triangular, 205
modulus of elasticity, 86 Rigid body translation, 31
narrow internal slit, 154 Rigidity, flexural, 1
nonsquare, 47 Rigid strip mass, rectangular plates, 141-145
nonsquare cantilever, 77 Ritz method, anisotropic plates, 261
orthotropic, 250-266 Rotary inertia, 314-324
plates with inplane forces, 276 AT-cut quartz crystal plates, 314
principle of stationary total energy, 123 boundary conditions, 324
point masses, 145-151 circular plates, 316
point-matching method, 151 effects of, 314, 317, 323
point supports, 130-141 elliptical plates, 322
Poisson’s ratio, 41, 54, 74, 79, 87, 89, 131, 132, 133 finite difference method, 314
Rayleigh method, 41, 43, 58, 118, 132 inplane forces, 316
Rayleigh-Ritz method, 58, 59, 61, 65, 69, 72, 73, large deflections, 316
76, 77, 79, 81, 86, 103, 119, 122, 131, 132, 133, low frequency cutoff, 320
141, 151, 152 Mathieu functions, 324
Reissner’s variational method, 140 Mindlin’s equations, 323
rigid strip mass, 141, 145 Mindlin theory, 318-319
rotary inertia, 318-323 rectangular orthotropy, 321
series method, 58, 60, 63, 74, 79, 102, 131 rectangular plates, 318-323
shear deformation, 318-323 synthesis of equations, 335
simple edge conditions, other, 58 thermal effects, 316
soap powder, 83 thickness-shear mode, 315, 316, 317, 321
Southwell’s method, 78 thickness-twist mode, 321
spring-mass system, 148 variable thickness, 285
SS-C-SS-C, 46-50 Rotating disk, clamped a t center, outer edge free, 27%
SS-6-SS-F, 51-52 Rotation
SS-6-SS-SS, 50-51 edge, 13
SS-ES-SS-ES, 116, 120 modes, 31
SS-F-F-F, 87
SS-F-SS-F, 53-58 Sectorial plates, 239-240
SS-SS-F-F, 87 all edges clamped, 239
SS-SS-SS-F, 52-53 boundary conditions, other, 239
SS-SS-SS-SS,43-45, 276-279 circular edge, 239
steel, 79, 83, 86 completely clamped, 240
stepwise superposition of modes, 57 exact solution, 239
strain energy, 119 radial sides simply supported, 239
surrounding media, effects of, 301-303 Rayleigh method, 239, 240
symmetrical slope restraints, 120 semicircular, 240
transcendental functions, 129 Semicircular plates, sectorial, 240
transition points, 54, 65, 75, 79, 109 Series method, rectangular plates, 60, 63, 78, 131
translational spring, 148 Shear de€ormation, 314-324
transverse shear, continuity of, 139 anisotropic material, 314
two opposite sides SS, 45-46, 279-280 AT-cut quartz crystal plates, 314
uniform slope restraint, 122 bending moments, 315
variable thickness, 291-297 circular plates, 31E-318
variational method, 47, 51, 58, 65, 79, 136, 140 effects of, 314, 317
352 VIBRATION OF PLATES