Missouri University of Science and Technology

Scholars' Mine

International Specialty Conference on Cold- (1971) - 1st International Specialty Conference

Formed Steel Structures on Cold-Formed Steel Structures

Aug 20th, 12:00 AM

Cold-formed Folded Plate Structures

Wilbur C. Schoeller

Richard H. J. Pian

Harry R. Lundgren

Follow this and additional works at: https://scholarsmine.mst.edu/isccss

Part of the Structural Engineering Commons

Recommended Citation
Schoeller, Wilbur C.; Pian, Richard H. J.; and Lundgren, Harry R., "Cold-formed Folded Plate Structures"
(1971). International Specialty Conference on Cold-Formed Steel Structures. 1.
https://scholarsmine.mst.edu/isccss/1iccfss/1iccfss-session5/1

This Article - Conference proceedings is brought to you for free and open access by Scholars' Mine. It has been
accepted for inclusion in International Specialty Conference on Cold-Formed Steel Structures by an authorized
administrator of Scholars' Mine. This work is protected by U. S. Copyright Law. Unauthorized use including
reproduction for redistribution requires the permission of the copyright holder. For more information, please
contact scholarsmine@mst.edu.
 COLD-FORMED FOLDED PLATE STRUCTURES

by

Wilbur C. Schoeller, 1 M. ASCE

Richard H. J. Pian, 1 F. ASCE
and Harry R. Lundgren, 2 M. ASCE

INTRODUCTION span between a hi gh-leve 1 compression ring near the center and a 1ower tension

ring near the perimeter (Figure 2). Adjacent plates are attached to a common
The object of the research work on steel folded plates conducted at
fold line member (bent plate) at the ridge or valley by screws, bolts, or
Arizona State University has been to investigate the behavior of cold-formed
welds. If the secondary effects due to the deformations of the fold 1 ine
steel folded plate structures including both parallel and tapered chords of
members, the tapered plates (or pane 1s), and the inner and outer rings are
rectangular, square, and circular plans (Figures 1 and 2). The studies have
assumed to be negligible, the following procedure may be used for analysis:
included (1) the formulation of theoretical methods to determine stresses and
(1) First consider each inclined tapered plate to be a simple trans-
displacements based on simpllfied analysis, and (2) the development of stress,
verse (tangential) span supported at fold lines. The load normal to this
displacement, and stability analyses employing finite element techniques.
surface is carried by the flexural strength of the individual light gage steel
Fabrication and material characteristics peculiar to cold-formed steel necessi-
panels. The panel span length varies from a maximum at the perimeter to a
tate analysis procedures that are different from those rmployerl for other
minimum near the center.
mdterials.
(2) The reaction of the tangential panel strip is applied as a 1 ine
Folded plate systems of cold-formed steel in long span roof may be
loading along the fold lines. The resulting loads are resolved into plate loads
considered as deep girders tilted from the usual vertical position so that the
that are carried to the support at the perimeter by the inclined plate (or
flanqes of adjacent girders are in contact either at the ridge or valley line
panel) acting as a girder with its top and bottom flanges each being half of a
(Figure 3). Effective use of material is provided since the longitudinal sloping
fold line member and the web being the steel panels. The center ring supplies
plates (or panels) not only carry the web shear of the tilted girders but also
a couple and a horizontal force to the girder, the outer rin£ provides a hori-
serve as thP. transverse structural roof deck. system. The various components
zontal force, and a web stiffener at the outer ring transmits the shear
can be used in many different combinations to provide attractive variation in
reaction from the girder to the ring.
appearance. In addition to rerforming a structural function, the roof deck may
( 3) Us i nq a simp 1ifi ed procedure it is pass i b 1e to determine the
c;erve as the finished interior ceiling.
forces (or stresses) of various parts of the structure neglecting the deforma-
SIMPLIFIED ANALYSIS
tions of the inner and outer rings, the axial defon11ations of the flanqes, and

The following :1eneral assumptions are used in the simplified analy~is the shear deformation of the web.

of cold-formed steel folded plates: Jn add1tion to the consideration of strength, the deflection of t.hr>

(a) Material is homogeneous and elastic. folded plate structure<; should be investigated. particularly for lonf]-':.pan

(b) Longltudindl edrws (at ridge and valley) are simply supported structures. The plate elements may be assumed as temporarily separated at the

(not continuous). fold lines, and the in-plane deflections of the individual plates can be

(() rrinciple of superposit1on is applicable. computed. The true deflection of the fold line can be found if the plates are

(d) Individual plates posse'->s negliqible torsional resistance. brought back. into coincidence. This can be done analytically or graphically

(e) The supporting members (diaphragms, frames, walls, columns, etc.) (Figure 5).

do not provide restraint against rotation of the ends of the plates.

RECTANGULAR PLAN FORM FOLDED PLATES USING FINITE ELEMENTS
(f) The effects of deformations are neglected for stress analysis

of the system. :r-neral Method:>[ Anal-!Jsia.

1 A more precise method of analysis than that

The cold-fonned steel folded plate system with parallel chords and previously described has been developed in which the method of finite elements

rectanqular plans may be considered as divided into two parts: (1) a simple is utilized to obtain stresses and displacements for cold-formed steel folded

trdnsverse ~pdn o;,upported at folded lines, and {2) a series of simple qirders plate structures. It applies to structures that are rectangular in plan in

'>rann1nq lon11iturhnally between end ()Upports. The load normal to the surface which the plates can be orthotropic and the properties can vary from plate to

1~, carried by the flexural strength of the individual cold-formed ster.l panels plate.

dnrl thP rE'act1on of the transverse panel strip is i'lpplied d'> a line loadinq Support restraints can be varied, although restraints must be ri q; d;

alonq the fold lines. The action of the folded plate units in resisting this that is, elastic supports cannot be accommodated. Loads are defined dS vertical

loarl is simildr to that of inclined deep girders braced by adjacent plates and horizontal projection uniformly distributed loads and vertical and horizon-

with top flanqe (bent plate) in compression, bottom flange in tension, and web tal line loads on fold lines.

rlements in shear (figure 4). If the web contribution for flexural strenqth The solution is formulated using the basic stiffness method, wherein

1<; neqlected, the lonqitudinal flanqe force in fold line members is obtained the plate element stiffnesses are obtained by employing the finite element

tJy dividing the bending moment of the deep girder by its depth. The shear techniques based upon an assumed displacement function. fold line member

force is resisted by the shear diaphragm action of the steel panels. strains are compatible with the extensional strains of the plate elements.

For tapered folded plates (or pleated dome) with circular or square Displacements are obtained for a specific loading and then back substituted to

plans, a similar approach can be followed. Here, radial tapered plate elements yield the stresses.

Nate Element Stiffneea. A stiffness matrix is developed for each panel.

'Prof. of Engrg., Arizona State Univ., Tempe, Az. The panel is first divided into a specified number of elements each of which
1Assoc. Prof. of Engrg., Arizona State Univ., Tempe, Az. has five degrees of freedom at each corner, consisting of the two rotations
140
 (a) Sawtooth Shape (b) Hat Shape

tension
ring

PLAN

(c) Tapered Pleated Dome

ELEVATION

FIGURE 1 FOLDED PLATE SHAPES FIGURE 2 TAPERED FOLDED PLATE

Ridge Member
Valley Member

Longitudinal Plate Action

Steel Roof Panels

~~
\

Co 1~mns End Frame

Members~

Steel Folded Plate Section

FIGURE 3 FOLDED PLATE BUILDING FIGURE 4

141
about the axes in the plane of the plate and the three linear displacements The same procedure is followed for the remaining fold lines resulting

coincidental with the orthogonal axes. A vector representation of the degrees in the stiffness matrix for the entire structure being formed by superimposing

of freedom is shown in Figure 6. w, ex, and ey represent the bending displace- the transformed sti ffnesses of each element.
ments and are related to the "bending element." u and v represent the in-plane
P,.ogl'am Utilization. The application of the program is limited to
displacements and are related to the "plane stress element." Stiffnesses of
folded plate structures comprised of rectangular panels as well as having a
the "bending element" and of the "plane stress element" are generated indepen-
rectangular plan. The number of panels that can be specified will vary with
dent ly and superimposed.
the storage capability of the computer on which the program is run. This is
The bending clement stiffness is formulated for the basic rectangular
not considered a serious limitation, however, since solutions can be performed
element (Figure 7) using the displacement function presented by Zienkiewicz
for typical interior and end panels that would probably provide the required
and Cheung ( 4). •
design information. The present edition of the program has the supports
Plane stress triangular elements based on linear strain (3) are used
located at the ends of the panels to simplify the input data. This could,
to provide the in-plane behavioral characteristics of the plate. Linear strain
however, be altered to allow for supports at interior positions. Figure 1~
is selected in preference to constant strain in order to describe the plate
shows end elevations of some structures that can be analyzed by this program.
strains with as few elements as necessary. In addition, the elements are
A common problem was solved using both the Simplified Method and the
overlapped to minimize the direction a 1 effects inherently caused by the
Finite Element program. A multiple sawtooth folded plate structure with a
qeometric proportions of the individual elements.
span of 40 feet was analyzed subjected to a uniformly distributed gravity loid.
The linear strain triangle requires six nodal points; therefore the
Critical results were as follows:
three corner nodes are supplemented by nodes at the mid-length of each side,
Simplified
thereby providing six per element.
Maximum Fold Line Force 22 ,500# 23 ,400#
Since the elemental pattern for rectangular plates most naturally leads
Maximum Plate Shear 1 ,024#/ft 958#/ft
to a rectilinear nodal pattern, the triangular elements and nodes are arranged
Since the simplified method yields only typical interior panel results,
no comparison could be made for the end panels. The deflection pattern

yielded in the computer analysis was consistent with physical restraints and
symmetry and appeared reasonab 1e.
ha 1f the actua 1 e 1ement thickness, are superimposed to provide the tota 1 thick-
ness (Figure 9}. For example, one element has nodes a, b, d, and e at its CIRCULAR PLANFORM FOLDED PLATES
USING FINITE ELEMENTS
corners. Another has nodes b. c, e, and f.

The combination of these two elements provides the five deqrees of Baokgrvund. The initial studies that were made on tapered folded plate
freedom per node shown in Figure 6. The nonnal finite element procedure is structures with circular plan form were based upon certain simplifying assump-
deviated from at this po1nt to acco~~~nodate a special problem caused by the tions with regard to the action of the structure. The loads in the fold line
lack of moment continuity across the fold lines. This problem has been resolved members were determined on the assumption that each trapezoidal segment was
by the formulation of special bending elements to be used at the fold lines rigid. These computed loads were then applied to the individual trapezoidal
that incorporate a hinged edge condition into the formulation. panels and the stress distribution in the fold line members and the webs were
determined by ~ethod of finite elements.
,1::;.~r·"!l·h '- / i:.:l(Jmcnts. The panel is fanned by assembling the appropriate
basic elements in such a way that the panel is constrained to conform to the The latest computer program to be developed considers the entire dome,

physical support system. rather than a segment of it, permits a more refined displacement analysis and
thus allows a closer study of behavioral characteristics of pleated domes
In orde• to reduce the degrees of freedom (and thereby conserve core
under various load conditions. This program assumes a uniform gravity load
space) internal degrees of freedom are eliminated. That is, the stiffness of
the panel is formulated in terms of the nodes on the panel boundary (Figure 10). condition but can be readily adapted to partial loading as well as wind loading.

In the reduction process, all nodes designated "•" in Figure 10 will

Method o[ Analysis. The stiffness method of analysis is used in which
be eliminated from the formulation. This results in the stiffness being
stiffness coefficients are obtai ned for nodal points that are assumed a long
expressed in terms of the "o" nodes and hence eliminates 18 of the original 44
the fold 11ne members. Displacements are determined for load systems that are
nodes--a considerable conservation of core space. It should be noted that this
assumed to be applied only at these nodal points. Stresses in each panel are
process yields the same accuracy of results as if the nodes had not been then determined from the displacements.
eliminated.
P/.atc u...,eni Stiffnesn. Each trapezoidal panel is assumed to be
''""""' /.y I ''""..t". After the stiffness of each panel has been obtained subdivided into 28 trapezoidal segments (Figure 13) and each of these into
and stored (in auxiliary storage) the panel stiffnesses are combined to form
four triangular segments (Figure 14) yielding in all ll2 trianqular seqments.
the structure stiffness matrix.
Using the finite element method, each triangle is assumed to be in a state of
Consider an example comprised of six panels in a "sawtooth"-type folded
uniform plane stress throughout and stiffness coefficients are obtained in
plate structure (see Figure ll for the structure plan and elevation).
terms of the nodes at the vertices of the triangles. Bending of the elements
First panel I is transformed from its own axes system to the structure
is not considered. A force at each node can have two components in the plane
axes system (shown in Figure 11). Then panel II is also transformed to the
of the panel and the node has only two degrees of freedom.
structure axes. Now it can be seen that the two panels share co111non nodes

along fold 11ne B. Panel Sti[fness Matrio:, In all, there are 68 nodes of this type in a
 a b c
/
/
/
/
e /
lJPt :P,~terl Plate Position /
/
d f
c'

FIGURE 5 PLATE DEFLECTION DIAGRAM h

g 1
z FIGURE 9

plate
boundary
w I I I
• • • • •
• • • • •
y~:
I
u
X

FIGURE 10

y
FIGURE 6 y
lrx
VI

1 2
X

A c E G

3
ZL X
B D F

FIGURE 11

y
FIGURE 7

a b c a b c
I

~
e
f d
h

FIGURE 8 FIGURE 12
143
panel. Assuming two degrees of freedom for each node results in a total of determined by comparing such an analysis to existing test results.

136 degrees of freedom for the panel. Providing for the necessary constraint The problem of obtaining the critical buckling stresses of thin plates

conditions for equilibrium results in a total of 133 unknown displacements, has been solved classically for only a relatively few cases. The large

requiring a 1ike number of equilibrium equations. Considering the extremely majority of these solutions are for rectangular plates with either simply

large number of equations that there would be to solve for a complete roof supported or clamped edges. If stiffeners or orthotropy are considered the

structure, it ,.,. decided to reduce the stiffness matrix for each panel to availability of solutions is drastically reduced. Recent innovations in

include only nodal points along the fold lines. The interior nodal points are cold-formed steel folded plates have created the need for calculating buckling

"condensed out" by matrix algebra. loads for thin plates of not only the standard rectangular shape, but also

such shapes as triangles, trapezoids, parallelograms, and even more exotic

Sub:::truC'tu:re AnalltBio of r:orrrelete Roo[. The general stiffness matrix
irregular shapes. In addition, in most of the plate buckling literature now
for the nodal points along the fold lines is built using three degrees of
available, a constant stress distribution is assumed, the notable exception
freedom per node. The additional degree of freedom is possible because of the
being the solution for a rectangular plate under combined bending and compres-
angle made between the adjacent plates, so that although each plate may be
sion. Obviously, more complicated stress distributions exist and should be
considered to be loaded only in its plane, two plates together can carry
taken into account along with a variety of boundary conditions.
components of load in any direction along their common boundary.
In order to accollJilodate the variation in loading and to provide the
As this analysis was developed with 8 nodal points along each fold line,
necessary flexibility in support conditions, a finite element solution has
each with 3 degrees of freedom, the total number of equations that it would be
been formu 1a ted and programmed.
necessary to solve for an actual roof would number several hundred. For this
General Method of Analttsie. The finite element method is used to
reason, it was decided to employ the method of substructure analysis in the
obtain the lowest elastic buckling load for a plate, either isotropic or
formulation of the solution.
orthotropic, with the in-plane stress distribution varying in a known manner
In the use of substructure analysis as applied herein, the structure
i throughout the p1ate.
is considered to be initially restrained completely. The constraints are
Almost any shape of plate can be reasonably well approximated by a
released and the loads applied progressively until the structure is free of
suitable combination of triangular and trapezoidal (of which a rectangle is a
artificial constraint. This is accomplished by releasing one fold line first,
special case) elements. The elemental stiffness matrix for the trapezoidal-
arbitrarily starting with a ridge (Figure 15). Next, the two adjacent valleys
shaped element is based upon the derivation by Zienkiewicz (4). Tocher's (3)
are released. Then, two ridges are released, and so on. When the final fold
T-10 element is used to obtain the elemental stiffness of the triangular
line is released the deflections of the nodal points along it are calculated,
elements. The stiffness modifying matrices for both e 1ements are derived
ne<t the displacements of nodal points along the two fold line members adjacent
following the method suggested by Kapur and Hartz (1), using the same displace-
to it, and so on, progressively determining fold line deflections until the
ment functions assumed in deriving the elemental stiffness matrices.
ridge that was initially released is reached.
The stiffness modifying matrices are derived for only the case of a
This method of analysis permits the analysis of very large structures
constant stress within the element (the stress may vary from element to
on computers that are limited in capacity. In its present form, it is capable
element, however) although it is possible to extend the derivation to allow
of work1ng with up to 2-i fold lines (ridges and valleys combined) on the
for various other stress distributions.
CDC 3400. It would be a simple matter to extend it by altering a dimension
Si nee the finite element method, as do many other numerica 1 techniques,
statement, but there might be some difficulty with roundoff error if larger
depends somewhat upon the fineness of the element grid pattern for close
numbers of fold lines were used.
convergence, an iterative technique which takes advantage of the symmetry and
STABILITY ANALYSIS BY FINITE ELEMENTS bandedness of the resulting system of equations is used.
An additional aspect of this study is the determination of critical The problem of calculating the buckling stresses for stiffened plates
stre~ses in elements of cold-formed folded plate structures. (including corrugated shear panels) is managed by consideration of an equiva-
It has 1ong been understood that the load carrying capabi 1ity of lent orthotropic flat plate.
structures fabricated of thin plates, stiffened and unstiffened, may be a The procedure used to obtain the critical stresses from the finite
function of their buckling characteristics. The large flat surfaces inherent element formulations is a variation of the decomposition procedure. The value
in folded plate structures, therefore, must be examined for this consideration. of the load that causes the modified stiffness matrix to become non-positive
The behavior of this type of structural element is complex. It is definite is determined, indicating an unstable structure.
anticipated that very early in the loading history the plate will buckle The computer analysis assumes that fasteners retain their integrity
elastically and change its load carrying system to one consisting of a tension throughout the loading. Also, the program is not capable of imposing non-

field simulating plate girder behavior. The post-buckling strength may then rigid support conditions. Therefore, comparative analyses with test results

extend to many times its pre-buckling strength. A mathematical analysis are made with work performed by MacFarland (2) since his test reports indica-

through the post-buckling range presents many problems such as determining th• ted a failure mode primarily free of fastener effect.

point of change of configuration and the buckled shape. These are particularly In Figure 16, MacFarland's experimental results for 20 gage material

difficult to determine for plates fabricated of cold-formed material because are represented by solid lines for each of the two identical test specimens.

of variations in materia 1 properties caused by manufacturing processes as well Computer analysis results reflecting the allowable upper and lower material

as the critical problem of fastening. After this introduction the question thicknesses are represented by dashed lines. Both fixed edges and simply

might well be posed--is an elastic critical stress analysis that does not supported edges were simulated in computer solutions.

consider the post-buckling configuration of any value? The worth can only be The edge details of the tested panels are designed to restrict rotati"on
144
resulting in a theoretical fixed edge, although local bending may have some
m
effect. It can be seen that the computer solution for fixed edges does
j
predict a failure load reasonably close to the test results.
Figure 17 shows simi 1ar results for 24 gage materia 1. Again, the
results agree most closely with the fixed edge solution although the lighter
materia 1 might be expected to allow more rotation at the supports than 1s
experienced by the 20 gage sample.

It does appear that a correlation may exist between the elastic buckling
stress computations and the actual ultimate strength of a panel fabricated of
cold-formed components and subjected to shear loading.
n
k FIGURE 14
11 13 15
3 60 ~---T

a d
50 ~~----

b 'e
£
c u
"' 40 Fixed Edge
1-
Q)
Solution
Q_

4 Vl
.0
30
""'
FIGURE 13 0
....J
----} Simply
Supported
"Q)
20 Solution
Q_

~
Q_
Ex peri menta 1 Results
I
10
r-~-L-~
R I

I 0
0 .2 .4 .6 .8
Max. Lateral Deflection (inches)
FIGURE 15 COMPARISON OF EXPERIMENTAL AND ANALYTICAL RESULTS
C20 GAGEl
50.------,-------.·~---.------~--~

40~-----4------~
support line for
columns at valleys £
u
<:

1-
Cl!
0. 30 r-------t--- -··-·
V1
.0

Fixed Edr,e
Solution

X
~ Simply
Sur ported

!~~--·-i
Solution

{~·
I --.
) I ' I

I' _•__
0 .2
Max. Lateral Deflection (inches)
FIGURE 15 PLAN VIEW OF ROOF FIGURE 17 COMPARISON OF EXPERIMENTAL AND ANALYTICAL RESULTS
145 <24 GAGE>
 ~l.t!MARY · (2) MacFarland D. E. "An Investigation of the Static Stability of
I Corrugated'Rectangular Plates Loaded in Pure Shear" (unpublished
The results of the theoretical portion of research on cold-formed steel Ph.D. dissertation, University of Kansas, January 1967).

folded plates are presented. The studies Include stress, deflection and stabil- (3) Tocher, J. L., "Analysis of Plate Bending Using Triangular Elements"
(unpublished Ph.D. dissertation, University of California (Berkeley),
ity analyses of systems with rectangular as well as circular planforms. 1962).

(4) Zienkiewicz and Cheung, "The Finite Element Method for Analysis of
CONCLUSIONS Elastic Isotropic and Orthotropic Slabs," Proc. Inst. Civ. Eng.,
28 (1964), 471-BB.
This research provides design-oriented material that can be used to

faci 1 i tate the design of folded plate structures fabricated of cold-formed BIBLIOGRAPHY
elements.
Campbell, C. H. "Simplified Analysis of Tapered Light Gage Steel Folded Plates:O•
The Simplified Method provides a direct technique that will suffice for Unpublished Master's thesis, Arizona State University, Tempe, Arizona, 1967.
use in fi na 1 designs for many structures as we 11 as for preliminary design of "Design of Light Gage Steel Diaphragms." American Iron and Steel Institute,
First Edition, 1967.
others.
"Folded Plate Roof Design." Design procedure for light gage steel folded
The Finite Element stress analysis permits a more detailed analysis that plate, Fenestra Company.
allows additional freedvm of loading, supports. and material variation. Kapur, K. K., and 8. J. Hartz. "Stability of Plates Using the Finite Element
Method." Proc. Amer. Soc. Ci v. Engrs., Engr. Mech. Jour., (April 1g66),
The Buck 1i ng Analysis can utilize as input the results of one of the pp. 177-195.
aforementioned stress analyses to compute the elastic buckling characteristics Luttrell, L. D. "Structural Performance of Light Gage Steel Diaphragms."
Cornell University, Report No. 319, 1965.
of the panels.
MacFarland, D. E. "An Investigation of the Static Stability of Corrugated
Rectangular Plates Loaded in Pure Shear." Unpublish~d Ph.D. dissertation,
ACKNOWLEDGEMENTS University of Kansas, January 1967.
This paper is based on work carried out by the Engineedng Research Nilson, A. H. "Folded Plate Structures of Light Gage Steel." ASCE Trans.,
Vol. 128, 1963, Part ll.
Center 9 Arizona State University, Tempe, Ar1 zona, under a contract with the
Nilson, A. H. "Diaphragm Action in Light Gage Steel Construction." AISI
American Iron and Steel Institute of New York, New York, to perform the inves- Regional Technical Paper, 1960.
tigation of "Light Gage Steel Folded Plates." The writers wish to express their Nilson, A. H. "Shear Diaphragms of Light Gage Steel." Proc. ASCE, J. Struc. Div.,
Yo l. 86, No. ST 11 , November 1960.
appreciation to William G. Kirkland and Or. Albert L. Johnson of the A.I.S.I,,
"Specifications for the Design of Cold-Formed Steel Structural Members." 1968
and the A. I.S. I. Committee on Building Research and Technology (D. S. Wolford, Edition. American Iron and Steel Institute.
Chairman of the Research and Specifications Subconmittee, and Dr. J. B. Scalzi, Tocher, J. L. "Analysis of Plate Bending Using Triangular Elements." Unpub-
lished Ph.D. dissertation, University of California (Berkeley). 1962.
Chairman of the Task Group on Shell Roof Structures) for their helpful sugges-

tions and continuing interest. They are also deeply indebted to ,John J. Glancy Weaver, loln. Jr. Computer Prorams for Structural Analysis. Princeton, N.J.:
D. Van Nostrand Co., Inc., 967.
for h1s valuable as~1stance.
Winter, George. "Co111nentary on Light Gage Cold-Formed Steel Design Manual."
For American Iron and Steel Institute, 1962.
REFERENCES
Yu, Wei-Wen. "Design of Liqht Gage Cold-Formed Steel Structures." Engineering
Experiment Station, We't Virginia University, 1965.
(J) Kapur, K. K., and B. J. Hartz, "Stability of Plates Using the Finite
Element Method," Proc. Amer. Soc. Civ. Engrs., [ngr. Mech. Jour., Zienkiewicz and Cheung. "The Finite Element Method for Analysis of Elastic
(April l9oG). pp. 177-195. Isotropic and Orthotropic Slabs." Proc. Inst. Civ. Eng., 28 (1964), 471-88.

146