404 Chapter 11.
Flat Plates
at the edges if the edges are held or by circumferential compression if the edges are not
horizontally restrained. In thin plates this circumferential compression may cause buckling.
When this condition of large deflection exists, the plate is stiffer than indicated by the
ordinary theory and the load-deflection and load-stress relations are nonlinear. Stresses
for a given load are less and stresses for a given deflection are generally greater than the
ordinary theory indicates.
Circular Plates
Formulas for stress and deflection when middle surface stresses are taken into account are
given below. These formulas should be used whenever the maximum deflection exceeds
half the thickness if accurate results are desired. The table following gives the necessary
constants for the several loadings and support conditions listed.
Let t = thickness of plate; a = outer radius of plate; q = unit lateral pres.rure; y = maximum
deflection; <Jh =bending stress; <Jd =diaphragm stress; <J = <J0 + <Jd = maximum stress due to
flexure and diaphragm tension combined. Then the following formulas apply:
(11.11-1)
(11.11-2)
First solve for yin Eq. (11.11-1) and then obtain the stresses from Eq. (11.11-2).
EXAMPLE For the plate of Example 1 of Sec. 11.2, it is desired to determine the maximum deflection
and maximum stress under a load of 10 lb/in2•
SOLUTION If the linear theory held, the stresses and deflections would be directly proportional
to the load, which would indicate a maximum stress of 9240(10)/3=30,800lb/in2 and a maximum
deflection of 0.0883(10)/3 = 0.294 in. Since this deflection is much more than half the thickness,
Eqs. (11.11-3) and (11.11-2) with the constants from case 1 in the table will be used to solve for the
deflection and stress. From Eq. (11.11-1), we obtain
10(104 )
30(106){0.2')
1.016 y
0 376
1-o.3T+ ·
(y)T3
2.0833 = l.4514f +0.376(f J
Starting with a trial value for y somewhat less than 0.294 in, a solution is found when y = 0.219 in.
From Eq. (11.11-2) the maximum stress is found to be 27,500 lb/in2•
Warshawsky (Ref. 3) fitted Eqs. (11.11-1) and (11.11-2) to the data presented by Mah in Ref. 71,
and cases 5-9 in the following table give these results. Chia in Ref. 91 has a chapter on nonlinear
bending of isotropic nonrectangular plates in which he covers in great detail the derivations, plotted
results, and formulas similar to Eqs. (11.11-1) and (11.11-2) for distributed loadings, concentrated
center loads, applied edge moments, and combined loadings for circular plates with various boundary
conditions. The uniformly loaded circular plate on an elastic foundation is discussed and results
presented for several boundary conditions. He also treats annular plates, elliptical plates, and skew
plates under uniform loading. Reference 54 presents the results of a study of the large deflections of
clamped annular sector plates for sector angles from 30-90° in steps of 30° and for ratios of inner to
outer radii from 0-0.6 in steps of 0.2.
11.11 Effect of Large Deflection; Diaphragm Stresses 405
Orcular Plates under Distributed Load Predudng Large Defledlons
Case No., Edge Condition Constants
1. Simply supported K = 1.016 K =0.376
(neither fixed nor held). I 1-V2
Uniform pressure q over
238
entire plate. K = 1. K 4 = 0.294 (Ref. 5)
' 1-v
2. Fixed but not held (no K = 5.33
edge tension).Uniform I 1-V2 2
K =0.857
pressure q over entire
2
plate. (Atcenter) K3=- K4 =0.50
1-v
4
(At edge) K3= l 2 K4 =o.o
-v (Ref. 5)
3. Fixed and held. Uniform K = 5.33 2.6
pressure q over entire • l-v2 Kl=ll
-v
plate. 2
(Atcenter) K3=- K4 =0.976
1-v
4
(At edge) K3=ll K4 =0.476
-v
(Refs. 15 and 16)
4. Diaphragm without K1 =0.0 K 2 =3.44
flexural stiffness, edge
(Atcenter) K3 =0.0 K4 =0.965
held. Uniform pressure q
over entire plate. (At edge) K3=0.0 K4 =0.748
(
(At r from the center) y = Ymn: 1-0.97-0.l a 5 r2 r5)
(Refs. 18 and 29)
5. Fixed and held. Uniform Atedge Atcenter
pressure q over a central
area of radius r0 . v = 0.3 r/a K1 K2 ~ K4 K3 K4
1.00 5.86 3.32 4.40 1.73
0.75 6.26 3.45 3.80 1.32
0.50 9.17 5.50 3.38 0.76
0.25 27.1 13.9 4.62 1.18
(Ref. 3)
6. Simply supported and Atcenter
held radially. Uniform
pressure q over a central r/a K1 K2 Kl K•
area of radius r•. v = 0.3 0.75 1.71 3.21 1.84 0.81
0.50 2.95 5.07 2.06 0.95
0.25 9.95 13.8 2.60 1.31 (Ref. 3)
7. Fixed and held with a Ymu at r=0.45a
central support. Uniform
K1 =36.4 K2 =20.0 (Ref. 3)
pressure q over entire
plate. v = 0.3
(Continued)
406 Chapter 11. Flat Plates
Orcul1r Plltes under Distributed Load Producing Larp Defledlons (Continued)
Case No., Edge Condition Constants
8. Annular plate fixed and For inner edge radius= 02a, max deflection y
held at both inner and
outer edges. Uniform atr=0.576a
pressure q over entire K1 =84.0 K2 =63.5
annular plate. v = 0.3
For stress at r = 0.2a,
K 3 =36.0 K4 =25.8 {Ref. 3)
9. Annular plate simply For inner edge radius= 0.2a, max deflection y
supported and held
radially at both inner and atr=0.576a
outer edges. Uniform K1 =20.3 K 2 =51.8
pressure q over entire
annular plate. v = 0.3 For stress at r = 0.2a,
K 3 =12.14 K4 =2.41
For inner edge radius= 0.4a, max deflection y
atr=0.688a
K 1 =57.0 K2 =159
For stress at r = 0.664a,
K3 =14.52 K4 =6.89 (Ref. 3)
Elliptical Plates
Nash and Cooley {Ref. 72) present graphically the results of a uniform pressure on a
clamped elliptical plate for alb = 2. Their method of solution is presented in detail, and
the numerical solution is compared with experimental results and with previous solu-
tions they have referenced. Ng {Ref. 73) has tabulated the values of center deflection for
clamped elliptical plates on elastic foundations for ratios of al b from 1 to 2 and for a wide
range of foundation moduli. Large deflections are also graphed for two ratios alb (l.5 and 2)
for the same range of foundation moduli.
Rectangular Plates
Analytical solutions for uniformly loaded rectangular plates with large deflections are
given in Refs. 30-34, where the relations among load, deflection, and stress are expressed
by numerical values of the dimensionless coefficients y It, qb4 IEt4 , and ab 2 IEt2 • The val-
ues of these coefficients given in the table following are taken from these references and
are for v =0.316. In this table, a, b, q, E, y, and t have the same meaning as in Table 11.4,
C1 d is the diaphragm stress, and C1 is the total stress found by adding the diaphragm stress
and the bending stress. See also Ref. 17.
In Ref. 35 experimentally determined deflections are given and compared with those
predicted by theory. In Ref. 74 a numerical solution for uniformly loaded rectangular
11.13 Ultlmate Strength 407
plates with simply supported edges is discussed, and the results for a square plate are
compared with previous approximate solutions. Graphs are presented to show how
stresses and deflections vary across a square plate.
Chia in Ref. 91 includes a chapter on moderately large deflections ofisotropic rect-
angular plates. Not only are the derivations presented but the results of most cases are
presented in the form of graphs usable for engineering calculations. Cases of initially
deflected plates are included, and the comprehensive list of references is useful. Aalami
and Williams in Ref. 92 present 42 tables oflarge-deflection reduction coefficients over
a range of length ratios a!b and for a variety-three bending and four membrane-of
symmetric and nonsymmetric boundary conditions. Loadings include overall uniform
and linearly varying pressures as well as pressures over limited areas centered on the
plates.
Parallelogram Plates
Kennedy and Ng (Ref. 53) present several graphs oflarge elastic deflections and the accom-
panying stresses for uniformly loaded skew plates with clamped edges. Several aspect ratios
and skew angles are represented.
11.12 PLASTIC ANALYSIS OF PLATES
The onset of yielding in plates may occur before the development of appreciable dia-
phragm stress if the plate is relatively thick. For thinner plates, the nonlinear increase
in stiffness due to diaphragm stresses is counteracted by the decrease in stiffness
which occurs when the material starts to yield (Refs. 52 and 80). Save and Massonnet
(Ref. 81) discuss the effect of the several yield criteria on the response of circular
and rectangular plates under various loadings and give an extensive list of references.
They also compare the results of theory with referenced experiments which have been
performed. Orthotropy in plates can be caused by cold-forming the material or by the
positioning of stiffeners. The effect of this orthotropic behavior on the yielding of
circular plates is discussed by Save and Massonnet (Ref. 81) as well as by Markowitz
and Hu (Ref. 82).
Crose and Ang (Ref. 83) describe an iterative solution scheme which first solves the
elastic case and then increments the loading upward to allow a slow expansion of the
yielded volume after it forms. The results of a test on a clamped plate are compared favor-
ably with a theoretical solution.
11.13 ULTIMATE STRENGTH
Plates of brittle material fracture when the actual maximum tensile stress reaches the
ultimate tensile strength ofthe material. A flat-plate modulus of rupture, analogous to the
modulus of rupture of a beam, may be determined by calculating the (fictitious) maxi-
mum stress corresponding to the breaking load, using for this purpose the appropriate
formula for elastic stress. This flat-plate modulus of rupture is usually greater than the
modulus of rupture determined by testing a beam of rectangular section.
408 Chapter 11. Flat Plates
Redilngullr Plates Under Unlfonn I.Old Producing Larp Deflection
qb 4/Et 4
Edges and Point
a/b of Max er Coef. 0 12.5 25 50 75 100 125 150 175 200 250
1 Held, not fixed ylt 0 0.430 0.650 0.930 1.13 1.26 1.37 1.47 1.56 1.63 1.77
a 4 b2 /Et2
Atcenterof
plate { ab2/Et 2
0
0
0.70
3.80
1.60
5.80
3.00 4.00
8.70 10.90
5.00
12.80
6.10
14.30
7.00
15.60
7.95
17.00
8.60
18.20
10.20
20.50
Heldand ylt 0 0.406 0.600 0.840 1.00 1.13 1.23 1.31 1.40 1.46 1.58
2 2
riveted { a 4 b /Et 0 0.609 1.380 2.68 3.80 4.78 5.75 6.54 7.55 8.10 9.53
1
Atcenterof ab2/Et 2 0 3.19 5.18 7.77 9.72 11.34 12.80 14.10 15.40 16.40 18.40
plate
Held and fixed ylt 0 0.165 0.32 0.59 0.80 0.95 1.08 1.19 1.28 1.38 1.54
Atcenterof { cr4 b2 /Et2 0 0.070 0.22 0.75 1.35 2.00 2.70 3.30 4.00 4.60 5.90
1
long edges ab2/Et 2 0 3.80 6.90 14.70 21.0 26.50 31.50 36.20 40.70 45.00 53.50
Atcenterof 2 2 0 0.075 0.30 0.95 1.65 2.40 3.10 3.80 4.50 5.20 6.50
plate {aabb/Et/Et
4
2 2 0 1.80 3.50 6.60 9.20 11.60 13.0 14.50 15.80 17.10 19.40
Held, not fixed y/t 0 0.625 0.879 1.18 1.37 1.53 1.68 1.77 1.88 1.96 2.12
1.5 Atcenterof { adb 2/Et 2 0 1.06 2.11 3.78 5.18 6.41 7.65 8.60 9.55 10.60 12.30
plate ab2 /Et2 0 4.48 6.81 9.92 12.25 14.22 16.0 17.50 18.90 20.30 22.80
Held, not fixed y/t 0 0.696 0.946 1.24 1.44 1.60 1.72 1.84 1.94 2.03 2.20
2to Atcenterof
00
{ a 4 b2 /Et2 0 1.29 2.40 4.15 5.61 6.91 8.10 9.21 10.10 10.90 12.20
plate ab2/Et 2 0 4.87 7.16 10.30 12.60 14.60 16.40 18.00 19.40 20.90 23.60
1.5 Held and fixed y/t 0 0.28 0.51 0.825 1.07 1.24 1.40 1.50 1.63 1.72 1.86
to Atcenterof { a 4 b2 /Et2 0 0.20 0.66 1.90 3.20 4.35 5.40 6.50 7.50 8.50 10.30
00 long edges ab2/Et 2 0 5.75 11.12 20.30 27.8 35.0 41.0 47.0 52.50 57.60 67.00
Plates of ductile material fail by excessive plastic deflection, as do beams of similar
material. For a number of cases the load required to produce collapse has been determined
analytically, and the results for some of the simple loadings are summarized as follows.
I. Circular plate; uniform load, edges simply supported
wu =CJ y (1.1tt
2
2)
(Ref. 43)
2. Circular plate; uniform load, fixed edges
Wu =cs1 (2.8141tt 2 ) (Ref. 43)
(For collapse loads on partially loaded orthotropic annular plates see Refs. 81 and 82.)
3. Rectangular plate, length a, width b; uniform load, edges supported
t' y t
W.. =RCJ 2
where ~ depends on the ratio of b to a and has the following values (Ref. 44):
b/a 1 0.9 0.8 0.7 0.6 0.5 0.4 0.3 0.2
5.48 5.50 5.58 5.64 5.89 6.15 6.70 7.68 9.69
4. Plate of any shape and size, any type ofedge support, concentrated load at any point
W., =CJ1 (t1tt2 ) (Ref. 45)
In each of the above cases W., denotes the total load required to collapse the plate,
t the thickness of the plate, and CJ1 the yield point of the material. Accurate prediction of W"
is hardly to be expected; the theoretical error in some of the formulas may range up to 30%,
and few experimental data seem to be available.
