https://ntrs.nasa.gov/search.jsp?

R=19710023855 2020-04-23T14:27:11+00:00Z

NASA NASA SP-8068

SPACE VEHICLE
DESICN CRITERIA
(STRUCTURES)

BUCKLING STRENGTH
OF STRUCTURAL PLATES

JUNE 1971

NATIONAL AERONAUTICS A N D SPACE ADMINISTRATION

 FOREWORD

NASA experience has indicated a need for uniform criteria for the design of space
vehicles. Accordingly, criteria are being developed in the following areas of technology:

Environment

Structures

Guidance and Control

Chemical Propulsion

Individual components of this work will be issued as separate monographs as soon as

they are completed. A list of all published monographs in this series can be found at
the end of this document.

These monographs are to be regarded asguides to the formulation of design require-

ments and specifications by NASA Centers and project offices.

This monograph was prepared under the cognizance of the Langley Research Center.
The Task Manager was A.L. Braslow. The author was R.H. Gallagher of Cornell
University. A number of other individuals assisted in developing the material and
reviewing the drafts. In particular, the significant contributions made by H.P. Adam
and R.R. Meyer of McDonnell Douglas Corporation; M.F. Card of NASA Langley
Research Center; W.J.Crichlow and R.E. Hubka of Lockheed Aircraft Corporation;
S.B. Dong of the University of California, Los Angeles; N.F. Dow of General Electric
Company; L.E. Hackman of North American Rockwell Corporation; R.N. Hadcock of
Grumman Aerospace Corporation; and P. Seide of the University of Southern
California are hereby acknowledged.

NASA plans to update this monograph when need is established. Comments and
recommended changes in the technical content are invited and should be forwarded to
the attention of the Design Criteria Office, Langley Research Center, Hampton,
Virginia 23365.

June 1971
For sale by the National Technical Information Service, Springfield, Virginia 22151 - Price $3.00
 GUIDE TO THE USE OF THIS MONOGRAPH

The purpose of this monograph is t o provide a uniform basis for design of flightworthy
structure. It summarizes for use in space vehicle development the significant experience
and knowledge accumulated in research, development, and operational programs to
date. It can be used to improve consistency in design, efficiency of the design effort,
and confidence in the structure. A11 monographs in this series employ the same basic
format three major sections preceded by a brief INTRODUCTION, Section 1. and
~

complemented by a list of REFERENCES.

The STATE OF THE ART. Section 2. reviews and assesses current design practices and
identifies important aspects of the present state of technology. Selected references are
cited to supply supporting information. This section serves as a survey of the subject
that provides background material and prepares a proper technological base for the
CRITERIA and RECOMMENDED PRACTICES.

The CRITERIA, Section 3, state what rules, guides, or limitations must be imposed
t o ensure flightworthiness. The criteria can serve as a checklist for guiding a design
or assessing its adequacy.

The RECOMMENDED PRACTICES, Section 4. state how t o satisfy the criteria.

Whenever possible, the best procedure is described: when this cannot be done,
appropriate references are suggested. These practices, in conjunction with the criteria,
provide guidance to the formulation of requirements for vehicle design and evaluation.

...
111
 1. INTRODUCTION . . . . . . . . . . . . . . . . . . . . . . . 1

2. STATEOFTHEART . . . . . . . . . . . . . . . . . . . . . . 2

2.1 Analytical Methods for Stability Evaluation . . . . . . . . . . . 2

2.2 Stability Evaluation of Plates of Typical Construction Forms . . . . . 4
2.2.1 Unstiffened . . . . . . . . . . . . . . . . . . . . . 4
2.2.2 Corrugated . . . . . . . . . . . . . . . . . . . . . 7
2.2.3 Stringer Stiffened . . . . . . . . . . . . . . . . . . 8
2.2.4 Waffle Stiffened . . . . . . . . . . . . . . . . . . 9
2.2.5 Sandwich . . . . . . . . . . . . . . . . . . . . . 10
2.2.6 Fiber Reinforced . . . . . . . . . . . . . . . . . . . 13

3. CRITERIA . . . . . . . . . . . . . . . . . . . . . . . . . . 14

3.1 Analysis Input . . .. . . . . . . . . . . . . . . . . . . . 14

3.2 Analysis Procedures .
. . . . . . . . . . . . . . . . . . . . 14
3.3 Tests . . . . . . . . . . . . . . . . . . . . . . . . . . 15
3.3.1 Materials . . . . . . . . . . . . . . . . . . . . . . 15
3.3.2 Plates . . . . . . . . . . . . . . . . . . . . . . . 15

4. RECOMMENDED PRACTICES . . . . . . . . . . . . . . . . . -15

4.1 Analysis Input . . . . . . . . . . . . . . . . . . . . . . . 15

4.2 Analysis Procedures . . . . . . . . . . . . . . . . . . . . . 16
4.2.1 Geometric Representation . . . . . . . . . . . . . . . 16
4.2.2 Material and Structural Properties . . . . . . . . . . . . 17
4.2.3 Classical Methods . . . . . . . . . . . . . . . . . . 17
4.2.3.1 Unstiffened (Isotropic) . . . . . . . . . . . - 1 8
i 4.2.3.2 Fiber-Reinforced Composite . . . . . . . . . . 21
4.2.3.3 Corrugated . . . . . . . . . . . . . . . . . 22
4.2.3.4 Waffle Stiffened . . . . . . . . . . . . . . . 22
4.2.3.5 Stringer Stiffened . . . . . . . . . . . . . . 22
4.2.3.6 Sandwich . . . . . . . . . . . . . . . . . 23

V
 4.2.4 Numerical Methods . . . . . . . . . . . . . . . . . . 24
4.3 Tests . . . . . . . . . . . . . . . . . . . . . . . . . . 26
4.3.1 Materials . . . . . . . . . . . . . . . . . . . . . . 26
4.3.2 Plate Structures . . . . . . . . . . . . . . . . . . . 26

REFERENCES . . . . . . . . . . . . . . . . . . . . . . . . 31

NASA SPACE VEHICLE DESIGN CRITERIA

MONOGRAPHS ISSUED TO DATE . . . . . . . . . . . . . . . . 43

vi
 BUCKLING STRENGTH
OF STRUCTURAL PLATES

1. INTRODUCTION
A plate is a planar body whose thickness is small compared with its other dimensions.
Flat or slightly curved plates are frequently used elements in space-vehicle structure. A
plate structure, as defined herein, may be as simple as the flat web of a stiffener or as
complex as an integrally stiffened plate supported by heavy frames and rings.

In the behavior of these plate structures under inplane compression and shear loads, a
critical point exists where an infinitesimal increase in load can cause the plate surface
to buckle. The load at this critical point defines the buckling strength of the plate.
Increases in load beyond the load at the initiation of buckling increase the buckling
deformations until collapse occurs. Thus, the load at collapse defines the postbuckling
or crippling strength of the plate. The behavior of plate structures in this regard differs
markedly from the behavior of columns and many thin curved shell structures for
which the buckling load corresponds closely to the collapse load.

A fundamental problem in space-vehicle design is to size plate elements so that plate

buckling will not induce detrimental deformations a t any expected service load up to
limit load (the maximum load expected in service), and that the design ultimate load
(the limit load multiplied by the ultimate design factor) will not exceed the plate-
crippling strength.

Buckling of a plate structure can cause an unacceptable degradation in the

aerodynamic profile of a space vehicle. It can trigger general buckling of a larger
structure because of a redistribution of loads; it can also affect the response to the
structure sufficiently to cause failure from excessive displacement or fatigue or
aeroelastic phenomena.

This monograph presents criteria and recommends practices for determining the
buckling and crippling strengths of structural plates under various types of static
loading, both mechanical and thermal. The document is concerned primarily with flat
plates, but is also applicable to plates with shallow curvature. It explicitly treats the
buckling and crippling of plates of unstiffened, corrugated, stringer-stiffened, waffle-
stiffened, sandwich, and fiber-reinforced composite construction, but it does not treat
beam-column effects, which are considered to be a stress-analysis problem.
 The basic approach t o determining the buckling strength of plate structures is through
analytical or numerical solution of the linearized equations governing the transition
from a flat form to a slightly buckled form. However, the analysis is different in some
details for the various forms of plate construction, such as stringer-stiffened, waffle-
I
stiffened, and sandwich panels. Crippling strength is generally determined semiempir-
ically. A test program is usually required when analytical procedures or existing data
are inadequate for determining buckling strength, as in panels with cutouts, local
loadings, new forms of stiffening, uncertain support conditions, or new fastening
methods.

Three other monographs in this series relate directly to the buckling of plate
structures: those on the buckling of cylinders (ref. I ) , of cones (ref. 2), and of doubly
curved shells (ref. 3), which describe the instability analysis of complete structural
systems of restricted geometric form. Other monographs are planned on design factors
and on problems of estimating the thermal environment and dynamic response of
structural panels.

2. STATE OF THE ART

For well-constructed flat plates, agreement between buckling theory and experiment is
good enough to warrant determining buckling loads directly by analysis without
empirical correlation factors. For curved plates, empirical corrections must be applied
to analytically determined buckling loads; the crippling load is usually determined
semiempirically. For unstiffened and stiffened plates of isotropic materials (plates
whose material elastic properties are the same in all directions), the processes of
determining buckling and crippling loads are well known and have been adequately
summarized in references 4 to 11. For lightweight plates, such as those reinforced by
sandwich construction and those of nonmetallic materials reinforced by fibers,
orthotropic-plate-buckling theory has been applied (as used here, the term orthotropic
denotes an analytical model in which the directional stiffness properties of reinforce-
ment are accounted for). A number of buckling solutions for orthotropic plates can be
found in references 12 to 14.

2.1 Analytical Methods for Stability Evaluation

The methods predominant in evaluating plate stability are discussed here under the
categories of classical and numerical.

Classical methods, as defined here, represent plate-behavior parameters by math-

ematical functions that are continuous throughout the entire panel. The parameters
used are generally the transverse displacements of the plate’s middle surface. Classical

2
methods offer both closed-form and series solutions, which are derived by a direct
attack on the governing differential equations or from the Rayleigh-Ritz or Galerkin
procedures for manipulating these equations. References 4 to 6 can be consulted for
descriptions of these classical-solution procedures.

Closed-form solutions fully determine panel-behavior parameters under all equilibrium,

continuity, and boundary conditions. In practice, however, the cases amenable t o
closed-form solutions are limited, and methods leading to series-form solutions are
more generally applicable. The form of the series for a particular problem must be
chosen so it can be truncated at a finite number of terms while approaching exact
satisfaction of the governing equations in the limit.

In theory, classical methods can deal with all the phenomena of flat-plate stability,
including inelastic buckling and postbuckling (crippling) strength. Providing the proper
treatment of these phenomena by classical methods is a matter of being able t o
represent parameters in the basic equilibrium, constitutive, and strain-displacement
relationships. This representation can best be accomplished for rectangular plates with
simple boundary conditions. Irregular geometry or nonuniformly distributed applied
loads or temperatures are extremely difficult to represent in parametric form, and the
flat-plate stability problem can therefore be solved by classical methods in only a few
special cases when such conditions exist.

By contrast, numerical methods characterize the behavior of a structure at points or

within regions of the structure and result in large-order systems of equations whose
coefficients are numerically evaluated functions of the material, geometry, and
applied-load parameters at these points or regions. As a group, these methods furnish
wide latitude in the treatment of nonuniformly distributed values of the design
parameters and nonlinear behavior, but solutions cannot be represented parametrically
as readily as they can with classical solutions. The finitedifference and finite-element
procedures are the foremost numerical methods.

The finite-difference method (ref. 15 ) represents the traditional approach to numerical

analysis; its well-established principles lead to definition of the correct- solution and
assure convergence t o it as the gridwork of points is refined. Moreover, there is an
extensive body of experience in applying this method to practical problems. The
resulting algebraic equations are relatively simple, permitting the use of highly efficient
solution techniques. On the other hand, the finitedifference method in its usual form
can evaluate only that stability behavior which can be described by one or more
differential equations, and it is difficult to apply if it must deal with irregular
gridworks or boundaries.

3
At present, the finite-element method is well established as a procedure for the static
and dynamic analysis of complex aerospace structures under stable conditions (refs. 16
to 18). Application of this method to instability problems is a more recent develop-
ment (refs. 19 to 21), and verification of the techniques for instability problems for
plates is limited. If convergence of the finite-element model has not been well estab-
lished when a complex plate structure is analyzed, the solution is usually verified by
experiment. Progress in finite-element instability analysis is surveyed in reference 22.

Numerical methods and many applications of the classical methods based upon series
representations require the computation of the buckling-load intensity and the
associated buckling-mode shapes from the coefficients of large-order systems of linear
equations. The mathematical description of buckling-load intensity is the eigenvalue,
and the mathematical description of mode shape is the eigenvector. Reference 23 is a
detailed treatise on the numerous procedures available for the computation of
eigenvalues and eigenvectors. There is a great need to improve computer programs to
the point where they can solve large-order eigenvalue problems efficiently and
accurately.

2.2 Stability Evaluation of Plates of Typical

Construction Forms
A relatively few construction forms of plates predominate in space vehicles; these are
largely unstiffened, corrugated, stringer-stiffened. waffle-stiffened, sandwich, and
fiber-reinforced composite plates. In practice, some combination of analytical and
empirical techniques is used t o evaluate the buckling and crippling behavior of each
class of plate construction, as discussed in the following sections.

2.2.1 Unst iffened

Formulas for buckling stress for flat, rectangular, unstiffened plates have been
determined for many support and load conditions, and have been substantiated by test.
Buckling stresses for geometrically irregular plates, especially those containing cutouts,
are not easily determined, however, and evaluation of stresses in such plates requires
use of numerical methods.

If the buckling stress of an unstiffened plate exceeds the material’s elastic limit (Le., is
in the plastic range of the material), buckling strength can be calculated in essentially
the same manner as the elastic-buckling strength of orthotropic plates, with coefficients
for terms in the governing equations drawn from functions of the moduli in the non-
linear region of the material’s stress-strain diagram (ref. 24).

4
 Formulas derived from this approach to evaluating plastic buckling for simple plate
I
geometries and load conditions can be expressed as the familiar elastic-buckling
equation multiplied by a plasticity-reduction factor, qp.

!
I
I
Two major approaches have been taken in determining q p analytically, one based on
the flow theory (ref. 25) and the other on the more simplified deformation theory
I (refs. 26 and 27). Incremental or flow theories work with incremental stress-strain laws
I in which increments of plastic strain are expressed in terms of the stress. Analysis of
the structure involves an integration of the plastic strains over the complete history of
loading. Deformation theory of plasticity is defined by a stress-strain law in which the
total strain components are related to the current stress and are therefore assumed to
I be independent of the history of loading.

Experimental evidence (e.g., ref. 28) has confirmed the suitability of deformation
theory in flat-plate inelastic-buckling analysis. Vol’mir (ref. 6) has suggested that
without a linear relationship between applied load and stress up to the level of the load
that causes buckling, the deformation theory in plastic buckling may be limited for
plates with significant initial imperfections. No experimental studies which could
confirm or deny this suggestion have yet been reported.

Although experimental data in substantiation of plastic-buckling theory pertain largely

to uniaxial compression, the analogy drawn in reference 24 to the applicability of
orthotropic-plate theory has been used successfully for biaxial compression. Refer-
ence 29 describes the basic theoretical relationships for more general stress states,
including shear, but does not give detailed solutions for such cases.

Both the postbuckling stiffness and the crippling strength of isotropic flat plates have
been studied extensively. The new analytical procedures for treatment of nonlinear
phenomena of postbuckling behavior for all forms of structures under compression
were motivated by Koiter (ref. 30) and considerably developed by other theorists
(refs. 31 and 32).

Flat-plate postbuckling behavior under axial compression has been described in

references 33 and 34. Reference 35 extends this work and develops the design charts; it
also contains the computer program listing from which the plotted data of the design
charts were obtained. The shear stiffness of a plate buckled in axial compression is
given in references 36 and 37; the special case of postbuckling stiffness under shear
loading is treated in references 36 and 38.

To date, most theoretical approaches to evaluating crippling strength (e.g., ref. 39)
assume that the flatness of the plate is ideal and that the shape of the postbuckling
displacements can be defined on the basis of the initial buckling mode. Recent studies

5
(refs. 37, 40, and 41) have demonstrated the inadequacy of these assumptions at the
onset of buckling failure, a point reached when yielding first occurs in some highly
deformed region of the middle surface.

The complexity of a nonlinear analysis that would account for uncertainties in

evaluating crippling strength has led to the adoption of various simplified procedures.
A common approach is based upon the concept of effective width; that is, the
assumption that the entire collapse load is taken by uniformly stressed widths of
material adjacent to the supports. The stress for crippling is assumed t o be the
material’s yield stress. The many formulas proposed for the determination of so-called
“effective” widths are discussed and compared in reference 42, which demonstrates
that no single formula is valid for all types of loading and support conditions.

An alternative approach to crippling-strength analysis which has been correlated with

extensive test data is advocated in reference 43. This approach relies upon a material
property, the “crippling-strength parameter,” defined as a function of the secant
modulus of elasticity and the compressive-yield strength. Thus, crippling-strength-
parameter data obtained from tests of one material can be used in calculation of the
crippling strength of a plate of different material. Although intended to cope with
elevated-temperature design problems, the approach is suitable for determining the
crippling strength of plates in structures under other conditions.

Theory regarding the determination of crippling strength of plates that buckle

inelastically is essentially underdeveloped. However, one indication from experiment is
that only small differences exist between inelastic buckling- and crippling-load levels.

Thermal buckling, initial imperfections, and plate curvature deserve special comment.
Thermal buckling occurs when constraint against thermal expansion in the middle
surface of a plate produces enough compressive stresses to cause buckling. The
constraint may be imposed by the plate-edge members or may be within the plate as a
result of the variation of thermal strain from point to point.

Since thermal-stress conditions often result in nonuniformly distributed compressive

stress, it is difficult to evaluate thermal buckling. Furthermore, it may be necessary t o
account for the thermal effects on material properties in this evaluation. Thus,
relatively few approaches for defining thermal buckling have been published; these are
cited in the annotated bibliographies of references 4 4 and 45, and surveyed in refer-
ence 46. In general, these references deal merely with elastic buckling under
nonuniform stress and are not uniquely associated with phenomena resulting from
elevated-temperature conditions. Moreover, empirical data on postbuckling behavior
and crippling strength under thermal stress are scarce. The problem is complicated
because the stress system causing buckling (which arises as a result of the constraint of

6
midplane thermal expansion in the unbuckled state) is modified and, in general,
aileviated by the transverse (flexural) displacements in the postbuckling phase. The
governing nonlinear equations for this behavior have been presented and solved for
specific cases in references 47 and 48, but these solutions are limited in application. A
simple, semiempirical approach t o determining the onset of permanent buckles in the
presence of thermal stress is proposed in references 49 and 50.

Initial imperfections (i.e., those deviations from flatness due t o fabricational

inaccuracy which all plates possess t o some degree) cannot be described adequately by
simple functions. It is nevertheless possible to draw general conclusions concerning the
influence of such imperfections on buckling through a criterion for the relationship
between applied load and deformation at the instant of buckling embodied in Koiter’s
previously cited approach to nonlinear analysis (ref. 3 0 ) . Applying this criterion,
reference 3 2 demonstrates theoretically that initial imperfections do not significantly
influence the buckling loads of flat plates. This is in accord with long-standing
experimental evidence.

When discussing the influence of curvature on plate stability, it is useful t o distinguish

between plates with shallow deviations and those with deep deviations from flatness.
Theoretically, when considered as a small deflection, curvature has the effect of
increasing the initial buckling level beyond that of a corresponding flat plate. From an
empirical point of view, however, it is evident that the imperfection-sensitivity of
curved shells grows with the increase of curvature and reduces the actual buckling level
below that predicted by a linear theory which includes curvature. Early design data
(e.g., refs. 5 1 and 5 2 ) therefore represent a correlation of theory and experiment. More
recently, reference 53 has shown that solutions of the shallow-shell differential
equations of reference 54 for the limiting case of zero curvature correspond with
flat-plate solutions only when special types of edge conditions exist. N o parametric
representation of the results of these solutions is feasible because of the complexity of
the problem and the many relevant parameters. The survey of the work given in
reference 53 attests to the paucity of prior experimental and theoretical efforts.

Reference 5 5 conjectures that the behavior of a buckled singly curved plate asymptot-
ically approaches that of a buckled flat plate. In the absence of experimental data on
the crippling strength of singly curved plates, this conjecture leads t o the use of
flat-plate theory for crippling analysis. There is no experimental evidence concerning
the crippling strength of doubly curved plates.

2.2.2Corrugated
Corrugated plates were used to form the surfaces of early metal aircraft t o meet a
design requirement which called for withstanding only shear loads, although these

7
plates have significant uniaxial buckling strength. At present, design interest in
corrugated plates has been renewed because of their ability to minimize thermal stress.
A corrugated plate with a single face sheet is treated as a stringer-stiffened plate
(Sec. 2.2.3); with two face sheets, it is treated as a sandwich plate (Sec. 2.2.5).

Analytical methods for calculating initial corrugated-plate buckling have been based
almost exclusively upon conventional orthotropic-plate theory (refs. 56 to 58). Such
calculation is often in poor agreement with the test data because of the great flexibility
existing in the plate in planes perpendicular to the corrugations, the practice of
attaching edges at discrete points in real structures (i.e., riveting), and inconsistencies in
the procedures for determining the effective twisting stiffness for use in equivalent
orthotropic-plate theory.

The effect of flexibility in planes perpendicular to the corrugations is analogous to the

effect of transverse shear deformation. Reference 59 accounts for this effect by use of
shell theory, but the complexity of the solution limits application of this approach to
plates with a shallow sinusoidal form of corrugation. Recent studies (refs. 60 and 6 1 )
indicate that the theory of thin-walled beams with deformable cross section promises
to be applicable for more realistic evaluation of corrugated structures.

References 56, 62, and 63 consider the intermittent attachment at corrugated-plate

edges in dealing with initial buckling under pure shear. In contradiction to experi-
mental data and St. Venant’s principle, these references employ assumptions implying
that edge effects propagate throughout the plate. References 59 and 60 also report
studies directed toward realistic intermittent-edge attachment.

For corrugated plates, analytical representation of local buckling, postbuckling

strength (ref. 59), and plastic buckling is in its infancy. Available information
demonstrates significant discrepancies between buckling theory and test data for
corrugated plates.

2.2.3 Stringer Stiffened

As shown in figure 1 , a stringer-stiffened plate consists of a sheet with attached
stiffening members along the direction in which the predominantly uniaxial loads act.
Stringers, either integral with the sheet or attached t o the sheet by rivets, spot welds,
or adhesives, efficiently enhance panel stiffness. A large number of stringer cross-
sectional shapes are possible; however, the economics of fabrication favor the forms
illustrated in figure 1 .

Early methods for stiffened-panel buckling analysis, as summarized in references 10

and 1 1 , treated initial buckling as synonymous with the instability of the sheet

8
 Figure 1. - Types of stiffeners used in stringer-stiffened plate construction.

between stringers, and computed crippling strength by analysis of a stringer and a

portion of the adjacent sheet in isolation. This model adequately represented the
behavior of structures within the design requirements of the time and was substan-
tiated by extensive test data.

Requirements for retaining a smooth structural profile up to the level of the initial
buckling load (required today for many space-vehicle designs) and the increased
emphasis on structural efficiency under failure loads motivated the development of
different analytical models for stringerstiffened plates. Both the coupled action of the
sheet and stringers (ref. 64) and the stiffener eccentricity (ref. 65) are therefore
recognized in current formulations for determination of initial buckling loads. For
example, a general method exploiting the capabilities of the computer was recently
devised for analyses of the coupled modes of buckling for stiffened-plate construction
(ref. 66); an extension of this method has been developed for composite-reinforced
stiffened plates (ref. 67). General methods using the finiteelement approach have also
been developed. Similar advances have been made in formulations for determining the
crippling strength of stiffened plates (refs. 68 to 70).

Excellent summaries of practice in the analysis of stiffened plates up through the

mid-1950s are presented in references 10 and 11. Since 1960, less attention has been
given stiffened-plate instability analyses. This is due in part to satisfaction with existing
formulations, reduced emphasis on this form of construction, and anticipation that
emerging computer procedures will improve the efficiency and accuracy of currently
available analytical methods.

2.2.4 Waffle Stiffened

A waffle-stiffened plate consists of a grid of stiffeners integral with each other and with
the skin of the plate (fig. 2), and derives its structural efficiency from geometric
uniformity and the absence of attachments.

Overall analysis of the buckling of waffle-stiffened plates involves two procedures:

( 1) the transforma tion of the basic geometric proportions into the coefficients, or

9
 lsogrid Longitudinal and transverse

Skewed

Figure 2. - Types of waffle stiffening.

elastic constants, relating the stress resultants to the strain parameters; and (2) the
solution of the governing buckling equations. Published studies have dealt principally
with the former procedure (refs. 71 and 72); more recent studies (ref. 73) have
included the effects of eccentric stiffening. Since waffle-stiffened structures have been
mainly in the shape of monocoque cylinders, the buckling of waffle-stiffened flat
plates has not been directly evaluated. There is little experimental information on the
plastic-buckling or postbuckling performance of waffle-stiffened plates.

2.2.5 Sandwich
Sandwich plates are formed by bonding two face sheets to a core (fig. 3). The face
sheets generally provide all of the plate’s flexural rigidity; the core separates the faces
and transmits shear. Sandwich plates may fail in either an overall mode of buckling
similar t o the buckling of unstiffened plates or in a variety of local modes, as illustrated
in figure 4.

Many theories of varying degrees of complexity have been formulated to analyze the
overall mode (fig. 4a) of buckling of sandwich plates (e.g., ref. 74). These theories
differ in their approaches to idealizing the core and face deformation. Theories must
usually account for transverse-shear deformation of the core. Unified treatments from
which both the overall and local modes of buckling can be calculated have also been
developed (e.g., ref. 75). In most practical situations, the results obtained for overall
buckling by application of either particular or unified theories are essentially the same.

10
 Truss core Corrugated Rigid foam, balsa,
or honeycomb

'FS

Figure 3. - Sandwich-plate construction.

(a) (b) (C) (d) (e)

++++ + + + I+ + + I ++++ ++++

Faci

tttt tttt t ttt

Overall buckling Dimpling Wrinkling of facings Shear
of facings crimping

Figure 4. -Typical failure modes of sandwich plates.

Moreover, unified approaches to the description of local buckling phenomena have

thus far demonstrated no special advantage over treatments based upon selective
analytical models because empirical considerations predominate in the description of
such phenomena.

11
Plastic buckling of sandwich plates has been treated either (1) by replacement of the
elastic modulus of the faces with properties of the stress-strain law in the plastic range
(ref. 75); or (2) to account for plasticity, by correction of the flexural stiffness in a
plate-buckling theory that includes transverse-shear deformation. All theories assume
elastic behavior of the core and only isotropic faces have been considered in these
descriptions, so experimental confirmation is limited. Moreover, there is no verification
that the more complicated methods yield more reliable results than the simplest
approximations. Theoretical understanding and experimental evidence on postbuckling
strength are even more limited than on plastic buckling.

The principal modes of local buckling for homogeneous and cellular-core sandwich
plates are dimpling (fig. 4b) and face wrinkling (figs. 4c and 4d); local buckling in this
form of construction, as in truss-core and corrugated-core sandwich construction, is
synonymous with failure. Shear crimping (fig. 4e) is not a local but a general instability
mode in which the buckle wavelength is very short due to a low value of the
transverse-shear modulus of the core. Shear crimping is accurately characterized by an
easily derived theoretical formulation (ref. 76).

In the face-wrinkling mode of local failure, the sandwich facing buckles as a plate on an
elastic foundation. Intracellular buckliiig (also called dimpling or monocell buckling) is
the buckling of the face into the spaces found in cellular- and corrugated-core construc-
tions. Since the occurrence of wrinkling strongly depends upon difficult-to-de fine plate
characteristics such as waviness of the face, the elasticity and strength of the core, and
the bond between the face and core, theoretical descriptions (e.g., refs. 77 and 78)
have generally proven unsatisfactory, and useful formulas are based upon experimental
data. Monocell-buckling formulas (refs. 79 and 80) are of semiempirical form and have
been adequately verified by test.

Evaluation of sandwich plates with corrugated cores requires that core properties be
represented orthotropically and assumes infinite transverse-shear stiffness in planes
parallel t o the corrugations. Corrugated-core sandwich plates have significant
postbuckling strength; in common with other forms of sandwich construction,
corrugated-core plate is, however, subject to failure in local buckling and inelastic
buckling modes. Reliable methods are available for the analysis of each of these modes
of behavior (ref. 8 I ).

The historical development of sandwich-plate analysis is critically reviewed in refer-

ence 82; detailed examination of related theory is presented in references 81 and 83. A
comprehensive review and compilation of analysis data, which serve as a detailed
state-of-the-art assessment, are contained in reference 84. References 81 t o 84 also
report on isolated published studies of the effects of such factors as thermal stress,
panel curvature, and initial imperfections on sandwich-plate buckling.

12
2.2.6 Fiber Reinforced

Fiber-reinforced composite plates are made of high-strength, high-stiffness fibers

immersed in a matrix, or filler material, of lower strength and stiffness. The conven-
tional fiber-reinforced composite plate consists of several plies, with long fibers or
filaments oriented in a single direction in each ply. Reference 14 is an authoritative
compendium of detailed analysis and design information on fiber-reinforced composite
plates. It is continually updated as a result of on-going research.

During the early period of the development of fiber-reinforced composite structures, it

was hoped that designs could be analyzed solely on the basis of the properties of the
constituent materials (i-e., the fibers and the matrix). It has become apparent, however,
that the most reliable fundamental material-property data are obtained from tests of
the individual ply or lamina. The stiffness properties of the composite plates in turn are
usually determined by integration of the individual ply properties with the formulas
given in reference 14.

Basic theoretical procedures for calculating composite-plate stiffness from the

properties of the individual laminas (ref. 14) depend on the coupling between inplane-
membrane and plate-flexural behavior. The governing equations for evaluating
fiber-reinforced composite-plate buckling are those of an orthotropic plate if no
coupling exists (e.g., when the laminas are properly arranged about the plate’s middle
surface) or is insignificant enough to be eliminated from consideration (e.g., when the
plate is composed of a large number of cross-plied elements). References 12 and 13
present design data for orthotropic plates applicable to fiber-reinforced composite
plates with negligible coupling. Under certain conditions, the buckling analysis of
fiber-reinforced composite plates may even be reduced to the form used for evaluation
of isotropic plates. Experiments and comparative theoretical studies of fiber-reinforced
composite-plate buckling (refs. 85 and 86) have generally been based on membrane-
flexure coupling. These studies have confirmed the validity of using approximate
procedures when the coupling is weak.

As yet, there is no complete understanding of all local failure modes in composite

plates. Local yielding prior to buckling is possible, and may invalidate an elastic-
buckling computation. A special criterion necessary for determining yielding in fiber
composites has been formulated by Tsai, and is discussed in reference 87. The criterion
requires careful engineering judgment in determining allowable yield-stress input,
however, and has been exercised with only limited success for shear and compression
loadings. Other criteria are discussed in reference 88. Experiments have shown that the
delamination mode of local failure represents an important limitation on the post-
buckling strength of fiber-reinforced composite plates (ref. 89). Because of the many
geometric and material factors involved, extensive work is being undertaken to develop

13
suitable descriptions of the local failure modes and the postbuckling and inelastic
strengths of fiber-reinforced composite plates (e.g., ref. 90).

3. CRITERIA
The buckling and crippling strengths of the plates of space-vehicle structures shall be
determined with sufficient accuracy to ensure that ( 1 ) any buckling deformation
resulting from any expected service load up t o limit loads will not impair the function
of the plate itself or any other system, nor produce any unanticipated changes in
stiffness or load distribution that are not accounted for in the overall structure; and
(2) plate crippling will not occur under application of design ultimate loads. Realistic
representations of the applied load and temperature conditions, the material
properties, and the structural geometry, details, and support conditions shall be incor-
porated in appropriate analytical and test models. Elastic or inelastic analyses of plate
buckling and crippling strengths shall be conducted, as appropriate, and shall be
substantiated by experimental data, when necessary.

3.1 Analysis Input

The values for applied stresses employed in determining plate buckling and crippling
strengths shall represent conservative estimates of stress fields resulting from the
combination of mechanical, thermal, and residual stress loading under all expected
service conditions. Any significant dependence of the mechanical and physical
properties of the material upon environmental conditions shall be accounted for. The
plate stresses shall be derived from stress analyses of overall structure, and the individ-
ual plates shall be realistically represented in the overall structure model used in these
analyses.

3.2 Analysis Procedures

The buckling and crippling strengths of structural plates shall be determined by
calculation. Plate-support conditions shall be defined by realistic determinations of
elastic fixity or, where this is not feasible, by conservative assumptions.

The following, as a minimum, shall be taken into account in the analysis when
applicable:

Significant variation of the material’s constitutive relationships with

geometry, time, and load level

0 Joints, attachments, and cutouts

14
, 0 Local failure modes, such as:
(a) Interrivet buckling, torsional rolling, and stringer crippling of stringer-
stiffened plates

(b) Face wrinkling, face dimpling, shear crippling, and general buckling of
sandwich plates

(c) Yielding or delamination of composite laminates prior to buckling

0 Combined loading effects

3.3 Tests

3.3.1 Materials

When adequate data on material properties are not available, they shall be obtained
from standard material-property tests.

3.3.2 Plates

When the determination of plate buckling and crippling strengths by calculation cannot
be shown to be adequate, representative tests shall be conducted to confirm the design.
Pretest documentation shall demonstrate that the test’s boundary conditions are valid
and that the test loads, specimen, fixtures, instrumentation, and procedures are appro-
priate.

4. RECOMMENDED PRACTICES
Analysis t o determine the buckling and crippling strengths of structural plates begins
preferably with the definition of the loads applied t o the plate, the geometric represen-
tation of the plate, and the values for the material properties of the plate. The analyst
should then specify and solve the analytical model, review the problem for critical
design details, and define the need for supporting test programs. It is usually necessary
to repeat this process several times because of changes and refinement in the total
vehicle design.

4.1 Analysis Input

The definition of loads applied t o the plate should be based upon proper analytical
modeling of the entire vehicle structure or of the major component of which the plate

15
forms a part. Such features as cutouts, concentrated loads, and shear-deformation
effects in the skin of stiffened-shell construction should be represented in the model.
The elastic-edge fixity of individual plates should be determined realistically from the
results of a numerical-methods analysis of overall structure, from a detailed study of
the design of individual elements, or from tests.

The plate analyst and the personnel responsible for the analysis of the complete space-
vehicle structure should coordinate closely the determination of individual plate loads.
With few exceptions, proper analysis of the complete structure depends upon the
relative elastic characteristics of the components of the system, characteristics that are
in part a function of the dimensions of the component plates. The plate analyst should
therefore make sure that any changes in plate proportions motivated by his analysis are
reflected in values for elastic characteristics used in the analysis of the complete vehicle
structure. To ensure consistency of plate and vehicle design, the correlation of plate
analysis with vehicle analysis should be repeated until plate loadings in successive
iterations agree to within 5 percent or less.

Thermal stresses should be determined by realistic assessment of temperature changes

expected from the environments, and the constraints on individual plates; residual
stresses should be determined by realistic assessment of fabricational conditions (e.g.,
ref. 91). Since the effect of postbuckling behavior on these stresses is uncertain, stress
intensity should be assumed t o be unchanged in calculating crippling strength unless
empirical data allow a different assumption. Flight conditions should be realistically
represented when externally applied loads are combined with loads due to thermal
stresses. Effects of temperature on material properties should also be considered in
evaluating combined loads.

4.2 Analysis Procedures

4.2.1 Geometric Rep resenta t ion

Basic geometric data on plate size, curvature, and support conditions are required for
the definition of an analytical model of a plate. Overall geometric data should normally
be obtained from layout drawings of the space vehicle; occasionally, geometric data
must be experimentally determined.

Preferably, support conditions should be determined from analyses of the major

component of which the plate forms a part, or of the complete space-vehicle structure,
as recommended in Section 4.1. Otherwise, support conditions should be determined
from elementary analyses of the adjacent plates and stiffening members; in the absence
of such analyses, simple support conditions should be assumed for edges possessing
essentially complete constraint against transverse displacement.

16
4.22 Material and Structural Properties
Values for the properties of plate materials should be obtained from MIL-HDBK SA
(ref. 92) for metals, and values for the properties of materials and plate rigidities of
sandwich-plate construction should be obtained from reference 93. Reference 14
should be consulted for the basic properties of fiber-composite materials. Values for a
wide range of elevated-temperature and cryogenic properties for metallic materials are
given in references 94 and 95.

In determining plate stiffness, it should be kept in mind that fiber-reinforced

composite and waffle-stiffened plates present special problems. Stability analyses of
these forms of construction are generally conducted with orthotropic-plate theory.
Coupling between the flexural and membrane stiffnesses exists in an orthotropic plate,
but the consequences of this coupling associated with direct stress are generally more
important in evaluating buckling than are the consequences of coupling of stiffnesses
associated with shear stress.

The coupling between membrane and flexural stiffnesses is usually weak for
laminated-fiber composite plates, and the stability of these plates can be determined
with sufficient accuracy by reducing the stiffness relationships to the form of
orthotropic-plate bending properties with the use of the approximation described in
references 8 5 , 86, and 96. The inadequacy of existing procedures for computing indi-
vidual layer-stiffness coefficients from the filament and matrix properties makes it
necessary that these coefficients be derived from tests which establish the biaxial
extensional stiffnesses and the shear stiffness in the plane of each individual layer.
Stiffness properties should then be calculated for the complete plate by using the
integration procedures given in Section 3 of reference 14.

Plate-stiffness coefficients for waffle-stiffened construction should be computed from

formulas presented in reference 7 1. Although an approximate reduction of this form of
stiffness representation t o orthotropic-plate flexure is feasible by using the approach
cited for fibercomposite plates (refs. 85,86, and 96), simpler and therefore preferable
representations for specialized types of waffle stiffening are given directly in terms of
orthotropic-plate properties in references 73, 97, and 98. Available references present
theory only; no comprehensive design charts can be recommended.

4.2.3 Classical Methods

Classical formulations and supporting tabular data have been substantiated for many
cases of plate analysis and are available in the published literature. These data. should
be used in determining the buckling and crippling strengths of structural plates when-
ever possible, especially for the conventional forms of plate construction (unstiffened,

17
stringer stiffened, and sandwich) which have a rectangular planform and are subjected
to uniformly distributed loads.

Plates with special features should also be analyzed by classical methods if such
features are amenable to representation in simple functional form, as described in
Section 2.1. Examples of classical methods applied to thermal-stress, residual-stress,
initial-imperfection, and tapered-thickness analyses are given in references 99, 9 1, 40,
and 100, respectively. Time-shared, remote-access computer facilities with substantial
capacity make these types of analyses available to most designers. Care should be
exercised, however, in determining the number of terms used in the functional repre-
sentations of the problem (refs. 4 to 6). If the circumstances of a problem are too
complex for adequate functional representation, the numerical methods described in
Sections 2.2 and 4.2.2 should be employed.

Recommendations for the use of classical methods in determining the buckling and
crippling strengths of unstiffened (isotropic), fiber-reinforced composite, corrugated,
waffle-stiffened, stringer-stiffened, and sandwich plates follow. Since there is an
immense body of published data for these forms of construction, emphasis in the
following discussion is placed on references to sources for these data, rather than on
the presentation of extensive formulas and detailed design guides.

4.2.3.1 Unstiffened (Isotropic)

As the basis for calculating the buckling strength of unstiffened isotropic plates of
conventional materials and for evaluating local failure modes of other forms of plate
construction, the following characteristic formula for critical stress should be used:

where uc. is the value of the reference stress (i.e., the stress at a point selected to
characterize the overall distribution of stress in the middle surface of the plate); E and
p are the elastic modulus and Poisson’s ratio, respectively; q c and 7) are the reduction
P
factors for cladding and plasticity, respectively; k is the buckling coefficient; and D is
the plate’s flexural stiffness per unit width. The geometry of the plate is in part
represented by the plate thickness, t, and a characteristic width dimension, b. Under
uniaxial compression, b is the width of the plate.

Reference 7 should be used for information on cladding-reduction factors. Plasticity-

reduction factors should be computed from deformation theory (refs. 24 and 26) or
should be obtained from charts derived from this theory (ref. 101).

18
The buckling coefficient is a function of the loading distribution and of the plate
geometry and boundary conditions. Equation (1) applies directly to conditions where a
uniform uniaxial or constant shear stress prevails [although the equation can be
employed in more complex situations by defining ucr as a reference value, as in
eq. ( l ) ] . Buckling coefficients for a wide range of nonuniform uniaxial-load conditions
and support arrangements can be found in standard texts (refs. 4 to 6).

The range of possible combinations in multiaxial-stress states is too large to permit

efficient data representations through the device of a single reference value. Therefore,
the stress-ratio concept in the form of interaction formulas should be used for
multiaxial-stress conditions. A wide range of interaction formulas covering combined
inplane-bending stress and shear is presented in references 7, 8, and 92.

Buckling strengths of uniaxially compressed, unstiffened isotropic plates with curva-

ture in the direction normal to the loading should be determined by using the empirical
method originally given in reference 5 1, and more recently described in reference 70.
This approach is based on a comparison of a curvature parameter, Z, with a parameter
describing the severity of initial imperfections, y, where

and

(3)
2O J l + 6flf

in which b is the characteristic plate dimension, r is the radius of curvature, k is the

buckling coefficient for the corresponding flat plate, and 0 defines the amplitude of the
initial imperfections. For Z < 7 , the curved plate is treated as a flat plate with modified
buckling coefficient, while for Z > y, the governing stability equation is of the general
form of the cylinder-stability equation. (See ref. 1 for discussion of the latter.) Esti-
mates of the parameter 0 and the detailed forms of the governing equations for this
method may be obtained from reference 70, but preferably, the parameter fl should be
determined by testing plates of the same fabricational quality as the plates being
analyzed.

Initial buckling due to thermal gradients should be determined by procedures described

in references 46 and 48. These procedures are directed mainly toward the problem of
instability under the condition of a nonuniform distribution of middle-surface stress
caused by thermal gradients. References 4 4 and 45 should be consulted for sources of

19
the limited available design data based on these procedures. Since available classical
procedures and design data are largely applicable only to one particular circumstance,
it should be expected that numerical approaches will generally be necessary to
determine temperature changes that cause buckling. I

In general, if temperature change causes no greater than a 10-percent difference in

material-property values within a plate, then the approximation of material-property
values at average temperature should be satisfactory in determining thermal buckling.
However, this approach is not likely to reduce effort significantly when numerical
methods are applied. If the temperature gradient is highly localized, use of average-
temperature property values is also permissible even when property values vary more
than 10 percent.

If the postbuckling stiffness of a plate must be known for analyses of the overall
structure of which the plate forms a part, it is recommended that the design charts
presented in Appendix A of reference 35 be used for biaxially applied stress states.
Reference 35 also describes a computer program capable of performing such stiffness
determinations. An empirically based formula (ref. 102) is recommended as a simple
alternative for determining postbuckling stiffness in the special case of uniaxial
compression. For determining the shear stiffness of a plate buckled in axial compres-
sion, reference 36 should be employed. For plates buckled under pure shear loadings,
the formulations derivcd in reference 36 or 38 should be used; both of these references
contain simple design formulas and graphical data.

Plastic buckling and crippling should be regarded as synonymous. The crippling

strength of the flat isotropic plate that buckles elastically may be substantial. In
theory, this strength is a complicated function of the support conditions, the loading
distribution, and the material’s mechanical properties. Crippling strength can be
determined by completely theoretical means using digital computers. For uniaxial
compression, however, simplified expressions based upon effective-width concepts
should be adequate for design purposes. Effective-width formulas for plates whose
middle-surface displacements in the direction normal to the unloaded edges are
constrained should be obtained from references 8 and 10. The empirical effective-width
formula of reference 103 is recommended for plates with edges free to move in the
plane of the plate. This formula has been substantiated by recent theoretical studies
(ref. 39) and extensive experimental data.

N o theoretical methods to determine crippling in the presence of thermal strain can be

recommended. A simple approximate method for determining the onset of permanent
buckling due to thermal- and compressive-load stresses developed in reference 60 can
be recommended although this procedure is substantiated by only limited test data.

20
Reference 43 presents a method which is recommended for determining the crippling
strength of a plate from data available for a plate of a different material. The correla-
tion of the crippling strength of plates of different materials depends upon a material
parameter that is a function of the compressive-yield strength and the secant modulus
of elasticity. Recent test data (ref. 104) suggest that this method may be unsatisfactory
for application to complex titanium-plate structures.

For determining the crippling strength of singly curved plates under complex loads,
flat-plate theory should be used in conformance with the conjecture of reference 55.
For doubly curved plates, the absence of data also makes it necessary to use flat-plate
theory to determine crippling strength.

42.3.2 Fiber-Reinforced Composite

As noted in Section 2.2.6, orthotropic models are generally used in the determination
of elastic-buckling stresses of fiber-reinforced composite plates. Reference 13 is
recommended as an authoritative source of theory, background information, and
formulas for stability analysis of orthotropic plates. Reference 14 should also be
consulted as a means of maintaining awareness of current work. Since the equations
and associated design data for fiber-composite plates cover only a limited range of the
governing parameters, it may be necessary to employ numerical methods or series-
solution procedures with classical methods.

The reduced bending-stiffness concept and other simplifications intended t o make

fiber-composite-plate analysis amenable to simple orthotropic-plate theory are usually
acceptable in design practice within the limits discussed in references 85 and 86. These
simplifications, however, err on the unconservative side and should be checked by an
examination of simple test cases.

Because of delamination and the absence of substantiated analysis procedures, either

plastic buckling and postbuckling situations should be avoided in the design of fiber-
reinforced plates or expected behavior in these situations should be verified by test.
Limited experimental evidence (ref. 89) shows that when plates have a sufficient
number of plies to be isotropic in behavior, the effective-width concept for determin-
ing crippling strength is valid. Even in such cases, however, delamination and other
local failure mechanisms can be expected to limit the plate strength to that of initial
buckling.

For laminatedcomposite plates, it is considerably more difficult to check for yielding

under direct compression before buckling than it is for isotropic plates. The yield
condition should be determined on the basis of the yield strength of each orthotropic
lamina in reference t o the lamina’s principal axes since the yield strength is established

21
experimentally with respect t o these axes. Depending upon the nature of the plate
structure, the procedures described in references 87 and 88 are acceptable.

Uncertainties in quality control and tolerances in fabrication, along with the lack of
understanding of critical failure mechanisms at the level of the individual fiber, should
be considered as potential problem sources in the practical analysis of the elastic
stability and buckling strength of fiber-reinforced composite plates.

4.2.3.3 Corrugated

Orthotropic-plate theories which account for shear deformation in the plane perpendi-
cular to the corrugations are preferred for initial buckling analyses of corrugated plates.
Shear stiffnesses for use in such analyses should be obtained from references 60
and 61, or computed from the types of procedures they describe. Reference 6 1 should
also be consulted for appropriate stiffness parameters for plates with intermittent
attachments at the corrugation ends.

If shear-stiffness effects can be shown to be negligible by analysis or test, solutions for

the initial buckling of corrugated plates should be taken from references 4, 13, 57,
and 58. Available data and experi1;iental correlations are insufficient t o permit
recommendation of specific procedures for determining postbuckling strength, local
buckling, and inelastic buckling of corrugated plates.

4.2.3.4 Waffle Stiffened

Calculations for determining the instability characteristics of waffle-stiffened plates are

based on orthotropic-plate theory so that the foregoing comments regarding fiber-
reinforced composite plates are applicable here as well. Local instability in the form of
independent buckling of the ribs and sheet plates is a possible failure mode for waffle
plates (ref. 105). Although no experimental studies have dealt specifically with plastic
buckling and postbuckling strengths of waffle plates, it appears reasonable to use the
procedures recommended for unstiffened and stringer-stiffened plates to determine
these strengths.

4.2.3.5 Stringer Stiffened

Stringer-stiffened plates are subject to failure as (1) long plates in Euler (column)
buckling: (2) short plates in a variety of local buckling or crippling modes; and
( 3 ) intermediate-length plates in a combined crippling and column mode. Because of
the many possible failure modes and the wide range of stiffener shapes and spacings, it
is not feasible to recommend specific formulations for this type of construction.

22
Reference 10 should be consulted for details on available computational procedures
substantiated by much experimental data accumulated over many years; reference 7 is
especially recommended as a source of graphical design data. References 4, 5 , and 11
should be consulted concerning other failure modes (e.g., the torsional rolling mode of
open-section stiffener buckling).

In addition to the modes of instability cited, attention should be given to instability

modes associated with skin-to-stringer attachments. Interrivet buckling should be
examined with the use of the procedures and data given in references 7, 10, and 106.
Another attachment-failure mode that should be checked for is forced crippling. This
occurs when a thick skin buckles and the stiffener is forced to follow the skin
d e f o r m tions. Formulations for analysis of this condition are summarized in
reference 10.

Advances made after the publication of references 7 and 10 in defining local buckling
modes of stringer-stiffened plates are described in reference 64; advances in the
treatment of inelastic behavior and shear loadings, in references 24 and 107; and
advances in the representation of stiffener eccentricity, in reference 65. Reference 65 is
particularly helpful for identifying some of the inadequacies of orthotropic-plate
theory as an approximate representation of overall buckling.

In the stability analysis of stringer-stiffened plates, attention should also be given to

emerging theoretical procedures founded in classical analysis concepts but which
require digital-computer capabilities for application (ref. 66). These procedures employ
a general representation of structural geometry and behavior, and should eliminate the
need for independent study of numerous estimated buckling modes.

Zee-shaped stiffeners are used more often than other shaped stiffeners o n curved
plates. The buckling of singly curved, Zee-section-stiffened plates should be treated by
application of the procedure given in reference 70, (discussed in the preceding sections
in connection with isotropic unstiffened plates). Reference 70 also contains an exten-
sion of available solutions for the buckling coefficient of flat Zee-stiffened plates. For
stringer-stiffened plates, the recommended procedure again involves a comparison of
curvature- and imperfection-sensitivity parameters, Z and 7 , respectively.

4.2.3.6 Sandwich

Sandwich plates should be checked for three possible modes of instability failure:
(1) general failure, with core and facings acting together; (2) local failure in the form of
dimpling, wrinkling, or delamination of a face, or shear buckling of the core; and
(3) local buckling of a plate element of the core, as in truss-core or corrugated-core
sandwich construction.

23
Reference 84 should be consulted for formulas used in analyzing both local and general
instability. This document critically reviews, correlates, and summarizes recommended
formulas for nearly all aspects of sandwich-plate analysis for which solutions have been
published. Reference 108 should be consulted for formulas used in analyzing sandwich
plates with fiber-composite face sheets. Reference 93 should be used as an authorita-
tive source of sandwich-plate design practice and data, but it does not discuss
underlying theory; it is oriented toward the selection of plate proportions to meet
specified buckling strengths for highly regular conditions. References 82 and 83 should
be consulted for a detailed discussion of theoretical approaches to general and local
instability.

Solutions for the thermal buckling of sandwich plates are given in graphical form in
reference 109. Because no data can be recommended for determining the influence of
attachments on sandwich-plate instability, test data should be obtained for each case.

From a design standpoint, experimental evidence (chap. 6 of ref. 81 and Sec. 9 of

ref. 83) indicates that it is acceptable to apply the plasticity-reduction factor for
unstiffened plates to determine both general and local instability phenomena in
sandwich plates. Reference 84 should be consulted for those plasticity-reduction-factor
formulas appropriate to general conditions and those appropriate to local conditions.
An incomplete understanding of local failure mechanisms and core-to-face bond
strength, however, requires that no allowance be made in design for postbuckling
Strength.

Formulas for local instability are not reliable for the complete range of parameters in
practical sandwich construction. For heavy cores (i.e., where the ratio of core density
to face-sheet density exceeds approximately 0.03), the approximate equations for local
instability should be adequate. Heavy cores do not impose a significant weight penalty
upon the total plate design and are therefore recommended as a means of effecting
large margins of safety against failure in local buckling modes.

For corrugated-core sandwich plates, references 1 10, 1 1 1 , and 1 12 should be consulted

for design data and formulations for overall buckling, for local buckling and post-
buckling strength, and for plastic buckling, respectively.

4.2.4 Numerical Methods

Numerical methods are recommended for analyses that are too complex for the more
economical classical approaches. Advantage should be taken of the ability of numerical
methods to cope with flat or curved plates that ( 1 ) are formed in irregular shapes and
varying proportions; (2) have stiffeners, reinforced cutouts, and unusual support condi-
tions; and/or (3) are subjected to nonuniformly distributed loads and temperatures.

24
Such methods are also effective in accounting for initial imperfections, residual stresses
inelastic- and anisotropic-material properties, and the influence of temperature vari-
ations on mechanical and physical properties of materials.

The accuracy of a numerical solution can be compromised by data-preparation errors,

the format of the mathematical model, and the arithmetical procedures of the
computational system. Reference 1 13 should be consulted for safeguards against these
sources of error.

When experimental or analytical verification is lacking, numerical solutions should be

substantiated by one of two alternative procedures. In one, solutions should be
obtained from analytical models that are successively more refined and the calculated
critical loads of the successive analyses compared to determine convergence trends.
Rules for extrapolation to estimate the convergent solution from these comparisons
have been developed for finitedifference analysis (ref. 15) and can be applied to
finite-element calculations, as well.

The second substantiation procedure uses two mathematical models, both constructed
on the same level of grid refinement. Upper and lower bounds for the convergent
solution are established by constructing one of the two models to ensure that the
calculated buckling load will be higher than the convergent solution, while the other
analytical model is constructed to yield a buckling load lower than the convergent
solution. The upper-bound model, formulated in conformity with the principle of
stationary potential energy, requires the definition of assumed displacements in an
appropriate form. Lower-bound evaluation procedures have not yet been explored in
finite-element buckling analysis, although their role in analyses for stable conditions is
well explored. References 114 and 115 should be consulted for theoretical and
practical aspects of evaluating lower bounds for convergent solutions. When a
converged solution has been obtained, the buckling-mode shape should be computed to
assist in identification of the critical element for stability.

System-reliability requirements, fabricational tolerances, and the degree of uncertainty

associated with analysis-input data should be assessed in determining the accuracies
required in solutions. In general, however, calculated results should correspond to the
convergent solution for the mathematical model within 5 percent.

Numerical methods are not well suited for incorporating simplifications (such as the
effective-width concept) which are used in elementary analytical theory for
determining postbuckling strength. On the other hand, numerical methods can be
directly applied to more rigorous and realistic exercises in computing postbuckling
phenomena, particularly in situations with unusual stress-strain laws, and in comnuting

25
combined inelastic- and finite-displacement behavior. Reference 1 16 and 1 17 should be
examined for descriptions of the procedures to follow in these computations.

It is important t o note that current consideration of the finite-element method for

evaluating plate buckling may require extension of existing element formulations for
plate flexure. A great many alternative representations of these relationships have been
published; these are summarized in references 1 18 and 1 19. Any valid mathematical
model is an acceptable choice for extension of the element formulation to the elastic-
instability case, whether it satisfies either equilibrium or compatibility conditions, or
neither, throughout the element and system. It is desirable to retain the functional
representations of element behavior as originally used in the basic flexural representa-
tion in the development of the additional terms required for instability analysis.
Numerical-me thod formulations should be validated by analyses for which solutions
are known.

4.3 Tests

4.3.1 Materials

Compression tests or shear tests should be performed, as appropriate, to establish

mechanical-property values for materials not covered by tabulated data o r where
special conditions are anticipated. For example, reference 120 demonstrates that the
reduction in the proportional-limi t stress due to fatigue will reduce buckling strength.
A simulation of service history, followed by performance of the standard material-
properties tests, may therefore be necessary for a structure designed for long life. Tests
should conform to the standards given in references 12 1 and 122.

4.3.2 Plate Structures

Experimental data on the buckling and crippling strengths of structural plates are
especially needed for designs with irregular geometry, cutouts, and combined thermal
and mechanical loads.

Specimens should possess the initial-imperfection and residual-stress qualities of the

actual structure, and a sufficient number of specimens should be tested to provide a
valid statistical sample. Whenever possible, tests should be conducted with full-scale
specimens and with realistic simulation of the continuous adjacent structure. When
single plates are tested in isolation, it is difficult to simulate any adjacent structure
conditions other than the extremes of fully fixed, free, or simple support. Refer-
ences 40, 123, and 124 should be consulted for descriptions of plate-buckling test-
support conditions. Care should be exercised to interpret test data properly for flat-end

26
 tests, for which the assumption of full or nearly complete fixity is often made but not
always realized. Also, in tests of postbuckling behavior and crippling strength,
particular attention should be given to the support conditions along the unloaded edges
due to their major influence on behavior.

1 Care should be taken that test loads are introduced into the plate in the same manner
as anticipated for the real structure, and that the elastic response of the test fixture
t
does not influence the buckling and crippling loads. Thermal loads can be imposed by
means of radiant- or induction-heating facilities (ref. 125). Thermal testing usually
involves the imposition of a specified variation of temperature throughout the panel,
accompanied by a gradual increase of load until buckling or crippling occurs. It is often
impractical t o maintain the specified temperature by control of the thermal-input
devices and the laboratory environment alone. Coolant tubes, attached t o or
incorporated in the test specimens, have been successfully used in providing more
practical thermal control (refs. 126 and 127).

Because of its initial imperfections, a plate being tested will seldom experience
buckling at a sharply defined load level. Various methods for identifying the buckling
load from measured test data are widely used (fig. 5). The method of Southwell
(refs. 128 t o 130) interprets the critical stress as the slope of the straight line passing
through points which give the maximum normal displacement ( 6 ) versus the ratio of
this displacement and the applied stress (6 /u) (fig. 5a). Particular attention should be
given to the interpretation of these points; the measured deflection must be small
enough so that the small-deflection theory remains applicable and yet large enough so
that the amplitude of the buckling mode, rather than the initial displacement,
predominates.

Another approach to the identification of buckling loads is the “top-of-the-knee”

method (refs. 89 and 123), illustrated in figure 5b. According to this method, the
buckling load is essentially the load corresponding to the top of the knee of a curve of
stress versus the lateral deflection. Reference 123 also describes the “strain-reversal”
method, in which the buckling load is taken as the load at which the strain at a point
on the crest of the buckle stops increasing and starts decreasing (fig. 5c).

In another approach for interpreting test data to identify the buckling load (ref. 123),
the average stress on the plate is plotted against the unit-compressive strain (the total
I
shortening of the plate divided by the plate length), as illustrated in figure 5d. The
buckling stress is then considered as the point where a sharp reduction in slope appears.

Still another method (ref. 131) is based upon the interpretation of a plot of the unit
shortening versus the square of the maximum deflection of the plate.

27
 b
d
E
FJ

Lateral displacement, -
6 Lateral displacement, 6
Stress U
Southwell DlOt T o w f - t he-knee method

Strain, E Strain, E

Strain-reversal method Average -stress/unit-shortening method

Figure 5. - Methods for identifying buckling stress from experimental data.

Each of the foregoing procedures has been employed with success in both plastic- and
elastic-buckling load evaluations (refs. 123, 130, and 131). It is recommended that the
strain-reversal method be used in conjunction with at least one of the other methods
described above. The tests should reveal both the buckling and crippling strengths of
the plate structure.

28
 Test instrumentation should include strain gages and deflection transducers. Since
instability represents a transition from compressive t o flexural behavior and the
I
interpretation of test data can depend upon assessment of net compressive strain, care
should be taken to locate an adequate number of strain gages on both faces of the plate
structure (back to back) to ensure identification of the onset of buckling.

References 84, 124, and 132 should be consulted for modem practice in the instability
testing of special forms of plate construction, such as fiber-reinforced composite,
sandwich, and stringer-stiffened plates.

29
 REFERENCES

1. Anon.: Buckling of Thin-Walled Circular Cylinders. NASA Space Vehicle Design

Criteria (Structures), NASA SP-8007, Revised 1968.

2. Anon.: Buckling of Thin-Walled Truncated Cones, NASA Space Vehicle Design

Criteria (Structures), NASA SP-80 19, 1968.

3. Anon.: Buckling of Thin-Walled Doubly Curved Shells. NASA Space Vehicle

Design Criteria (Structures), NASA SP-8032, 1969.

4. Timoshenko, S.; and Gere, J.M.: Theory of Elastic Stability. Second ed., McGraw-
Hill Book Co., Inc. 1961.

5. Bleich, F.: Buckling Strength of Metal Structures. McGraw-Hill Book Co., Inc.,
1952.

6. Vol’mir, A.: Stability of Elastic Systems. Translation of U.S. Dept. of Commerce,

Nov. 15, 1965. (Available from DDC as AD 628508).

7. Anon.: Structures Data Sheets. Vol. 11, Roy. Aeronaut. SOC.(London) (contin-
ually updated).

8. Gerard. G.; and Becker, H.: Buckling of Flat Plates. Pt. I of Handbook of Struc-
tural Stability, NACA TN 3781, 1957.

I 9. Bulson, P.S.: The Stability of Flat Plates. American Elsevier, 1969

I 10. Gerard, G.: Compressive Strength of Flat Stiffened Panels. Pt. V of Handbook of
Structural Stability, NACA TN 3785, 1957.
1
11. Argyris, J.H.; and Dunne, P.C.: Structural Analysis. Pt. 2 of Structural Principles
and Data, Sir Isaac Pitman and Sons, Ltd. (London), 1962.
I
12. Hearrnon, R.F.S.: An Introduction to Applied Anistropic Elasticity. Oxford
University Press, 1961.

31
13. Lekhnitski, S.G.: Anisotropic Plates. Second Russian ed., Gordon and Breach
(New York), 1968.

14. Arvin, G., ed.: Structural Design Guide for Advanced Composite Applications.
F i r s t edition, prepared under Contract No. F33615-69-C-1368, Advanced
Composite Division Air Force Materials Lab., (Wright-Patterson AFB, Ohio), Aug.
1969.

15. Collatz, L.: The Numerical Treatment of Differential Equations. Third ed.,
Springer-Verlag (Berlin), 1960.

16. Przemieniecki, J.S.; et al., eds.: Proc. Conf. on Matrix Methods in Structural
Mechanics. AFFDL TR 66-80, Oct. 1965.

17. Przemieniecki, J.S.: Theory of Matrix Structural Analysis. McGraw-Hill Book Co.,
Inc., 1968, ch. 15.

18. Zienkiewicz, O.C.: The Finite Element Method in Structural and Continuum
Mechanics. Second ed., McGraw-Hill Book Co., Inc., 1971.

19. Gallagher, R.H.; and Padlog, J.: Discrete Element Approach to Structural Instab-
ility Analysis. AIAA J. vol. 1, no. 6, June 1963, pp. 1437-1439.

20. Kapur, K.; and Hartz, B.: Stability of Plates Using the Finite Element Methods.
Proc. ASCE, J . Eng. Mech. Div., vol. 92, no. EM2, Apr. 1966, pp. 177-195.

21. Gallagher, R.H.; Gellatly, R.A.; Padlog, J.; and Mallett, R.H.: A Discrete Element
Procedure for Thin-Shell Instability Analysis. AIAA J., vol. 5, no. 1, Jan. 1967,
pp. 138-145.

22. Gallagher, R.H.: The Finite Element Method in Elastic Instability Analysis. Proc.
International Ship Structure Congress. Symp. on Finite Element Method
(Stuttgart, W. Germany), June 1969.

23. Wilkinson, J.H.: The Algebraic Eigenvalue Problem. Clarendon Press (Oxford),
1965.

24. Peterson, J.P.: Plastic Buckling of Plates and Shells Under Biaxial Loading. NASA
TN D-4706, 1968.

25. Prager, W.: An Introduction to Plasticity. Addison-Wesley Publ. Co. (Reading,

Mass.), 1959.

32
26. Stowell, E.Z.: A Unified Theory of Plastic Buckling of Columns and Plates.
NACA Rept. 898,1948.

27. Bijlaard, P.P.: Theory and Tests on the Plastic Stability of Plates and Shells.
J. Aeron. Sci., vol. 16, no. 9, Sept. 1949, pp. 529-554.

28. Pride, R.A.: Plastic Buckling of a Simply Supported Plate in Compression.

J. Aeron. Sci., vol. 19, no. 1, Jan. 1952, pp. 69-70.

29. Jones, R.M.: Plastic Buckling of Eccentrically Stiffened Circular Cylindrical

Shells. AIAA J., vol. 5, no. 6, June 1967, pp. 1 147-1 151.

30. Koiter, W.T.: Elastic Stability and Post-Buckling Behavior in Nonlinear Problems.
R.E. Langer, ed., Univ. of Wisconsin Press, 1962.

31. Budiansky, B.; and Hutchinson, J.W.: A Survey of Some Buckling Problems.
AIAA J., vol. 4, no. 9, Sept. 1966, pp. 1505-1510.

32. Pope, G.G.: On the Bifurcational Buckling of Elastic Beams, Plates and Shallow
Shells. Aeron. Quarterly, vol. XIX, no. 1, Feb. 1968, pp. 20-30.

33. Van der Neut, A.: The Post-Buckling Stiffness of Rectangular Simply Supported
Plates. Rept. VTH-I 13, Technische Hogeschool (Delft, Netherlands), Oct. 1962.

34. Mayers, J.; and Nelson, E. : Maximum Strength Analysis of Postbuckled Rectan-
gular Plates. AIAA Sixth Aerospace Science Meeting (New York), Jan. 1968.

35. Dickson, J.N.; and Brolliar, R.H.: The General Instability of Eccentrically
Stiffened Cylindrical Shells Under Axial Compression and Lateral Pressure. NASA
CR-1280, 1969.

36. Boley, B.A.: The Shearing Rigidity of Buckled Sheet Panels. J. Aeron. Sci.,
vol. 17, no. 6, June 1950, pp. 356-362.

37. Stein, M.: Behavior of Buckled Rectangular Plates. Proc. ASCE, J. Eng. Mech.
Div., vol. 86, no. EM2, Apr. 1960.

38. Kuhn, P.: Stresses in Aircraft and Shell Structures. McGraw-Hill Book Co., Inc.,
1956.

39. Abdel-Sayed, G.: Effective Width of Thin Plates in Compression. Proc. ASCE, J.
Structural Div., vol. 95, no. STIO,Oct. 19W, pp. ? I 83-2203.

33
40. Dwight, J.B.; and Ratcliffe, A.T.: The Strength of Thin Plates in Compression, In
Thin-Walled Steel Structures. K.C. Rockey and H.V. Hill, eds., Gordon and
Breach (New York), 1968.

41. Yamaki, N.: Posbuckling Behavior of Rectangular Plates with Small Initial
Curvature Loaded in Edge Compression. Trans. ASME, J. Appl. Mech., vol. 26,
Series E, no. 3, Sept. 1959, pp. 407-414.

42. Jombock, J.R.; and Clark, J.W.: Postbuckling Behavior of Flat Plates. Trans.
. Pt. 11, 1962, pp. 227-243.
ASCE, V O ~127,

43. Anderson, R.A.; and Anderson, M.S.: Correlation of Crippling Strength of Plate
Structures with Material Properties. NACA TN 3600, 1956.

44. Boley, B.A.; Weiner, J.H.; and Tolins, I.S.: Bibliography. Pt. I1 of Thermal Stress
Analysis of Aircraft, WADC TR 56-1 02, 1956.

45. Gallagher, R.H.; and Huff, R.D.: A Bibliography of Thermal Stress Analysis
References, 1955-1962. Pt. I of Thermal Stress Determination Techniques in
Supersonic Transport Aircraft Structures. ASD-TDR-63-783, Jan. 1964.

46. Johns, D.J.: Thermal Stress Analyses. Pergamon Press, Ltd. (Oxford), 1965.

47. Gossard, M.L.; Seide, P.; and Roberts, W.M.: Thermal Buckling of Plates.
NACA TN 277 1 , 1952.

48. Van der Neut, A.: Buckling Caused by Thermal Stresses. Ch. 1 1 of High Tempera-
ture Effects in Aircraft Structures, N.J. Hoff, ed., Pergamon Press, Ltd. (Oxford),
1958.

49. Zender, G.W.; and Hall, J.B., Jr.: Combinations of Shear, Compressive Thermal
and Compressive Load Stresses for the Onset of Permanent Buckles in Plates.
NASA TN D-384, 1960.

50. Zender, G.W.; and Pride, R.A.: The Combinations of Thermal and Load Stresses
for the Onset of Permanent Buckling in Plates. NACA TN 4053, June 1957.

51. Stowell, E.Z.: Critical Compressive Stress for Curved Sheet Supported Along All
Edges and Elastically Restrained Against Rotation Along the Unloaded Edges.
NACA RB 3 107 (WR L-253), 1946.

34
52. Schildcrout, M.; and Stein, M.: Critical Combinations of Shear and Direct Axial
Stress for Curved Rectangular Panels. NACA TN 1928, 1950.

53. Pope, G.G.: On the Axial Compression of Long, Slightly Curved Panels. R. & M.
no. 3392, Aeron. Res. Council, Great Britain Ministry of Aviation (London), 1965.

54. Mushtari, J.M.; and Galimov, K.Z.: Non-Linear Theory of Thin Elastic Shells.
NASA TJT-F62,1961.

55. Koiter, W.T.: Buckling and Post-Buckling Behavior of a Cylindrical Panel Under
Axial Compression. Rept. S.476, National Aeron. Res. Inst., (Amsterdam), May
1956.

56. Easley, John T.; and McFarland, D.E.: Buckling of Light-Gage Corrugated Metal
Shear Diaphragms. Proc. A X E , J. Structural Div., vol. 95, no. ST 7, July 1969,
pp. 1497-1 5 16.

57. Peterson, J.P.; and Card, M.F.: Investigation of the Buckling Strength of
Corrugated Webs in Shear. NASA TN D424, 1960.

58. Rothwell, A.: The Buckling of Shallow Corrugated Webs in Shear. Aeron. J.,
vol. 72, no. 694, Oct. 1968, pp. 883-886.

59. Green, A.E.: The Elastic Stability of a Corrugated Plate Under a Shearing Force.
Proc. Cambridge Phil. SOC.,vol. 33, 1937, pp. 459473.

60. Lin, C.J.; and Libove, C.: Theoretical Study of Corrugated Plates: Shearing of
Trapezoidally Corrugated Plate with Trough Lines Held Straight. Rept.
MAE 1833-T1, Syracuse Univ. Res. Inst., Dept. of Mech. and Aerospace Eng.,
May 1970.

61. Lin, C.J.; and Libove, C.: Theoretical Study of Corrugated Plates: Shearing of a
Trapezoidally Corrugated Plate with Trough Lines Permitted to Curve. Rept.
MAE 1833-T2, Syracuse Univ. Res. Inst., Dept. of Mech. and Aerospace Eng.,
June 1970.

62. McKenzie, K.I.: The Shear Stiffness of a Corrugated Web. R. & M. no. 3342,
Aeron. Res. Council, Great Britain Ministry of Aviation (London), 1963.

63. Bryan, E.R.; and El-Dakhakhni, W.M.: Shear Flexibility of Corrugated Decks.
Proc. ASCE, J. Structural Div., vol. 94, no. ST 1 1 , Nov. 1968, pp. 2549-2580.

35
64. Rothwell, D.C.: Coupled Modes of Buckling of Panels with Z-Section Stiffeners.
Aeron. J., vol. 72, no. 686, Feb. 1968, pp. 159-163.

65. McElman, J.; Mikulas, M.; and Stein, M.: Static and Dynamic Effects of Eccentric
Stiffening of Plates and Cylindrical Shells. AIAA J., vol. 4, no. 5, May 1966,
pp. 887-894.

66. Wittrick, W.H.: A Unified Approach t o the Initial Buckling of Stiffened Panels in
Compression. Aeron. Quarterly, vol. XIX, no. 3, Aug. 1968, pp. 265-283.

67. Viswanathan, A.V.; Soong, Tsai-chen; and Miller, R.E., Jr.: Buckling Analysis of
Axially Compressed Flat Plates, Structural Sections, and Stiffened Plates
Reinforced with Laminated Composites. Prepared by The Boeing Co. under
Contract NAS 1-8858. (To be published as NASA CR 1859)

68. Bijlaard, P.P.: On the Buckling of Stringer Panels Including Forced Crippling.
J. Aeron. Sci., vol. 22, no. 7, July 1955, pp. 491-501.

69. Bulson, P.: The Local Instability of Structural Sections with Flange Reinforce-
ments. Thin-Walled Steel Structures, K.C. Rockey and H.V. Hill, eds., Gordon and
Breach (New York), 1969.

70. Peterson, J.P.; and Whitley, R.O.: Local Buckling of Longitudinally Stiffened
Curved Plates. NASA TN D-750, 1961.

71. Dow, N.F.; Libove, C.; and Hubka, R.: Formulas for the Elastic Constants of
Plates with Integral, Waffle-Like Stiffening. NACA TR 1195, 1954.

72. Crawford, R.F.; and Libove, C.: Shearing Effectiveness of Integral Stiffening:
NACA TN 3443, June 1955.

73. Meyer, R.R.: Buckling of 45" Eccentric-Stiffened Waffle Cylinders. J. Roy.

Aeron. SOC.,vol. 71, no. 679, July 1967, pp. 516-520.

74. Libove, C.; and Batdorf, S.B.: A General Small-Deflection Theory For Flat
Sandwich Plates. NACA Rept. 899, 1948.

75. Benson, A.S.; and Mayers, J.: General Instability and Face Wrinkling of Sandwich
Plates-Unified Theory and Applications. AIAA J., vol. 5, no. 4, Apr. 1967,
pp. 729-739.

36
76. Zahn, J.J.; and Kuenzi, E.W.: Classical Buckling of Cylinders of Sandwich
Construction-Orthotropic Cores. Res. Note FPL-018, Forest Products Lab.
(Madison, Wis.), Nov. 1963.

77. Williams, D. : Sandwich Construction-A Practical Approach for the Use of

Designers. R. & M. no. 2466, Aeron. Res. Council, Great Britain Ministry of
Aviation (London), 1947.

78. Yusuff, S.: Face Wrinkling and Core Strength in Sandwich Construction. J. Roy.
Aeron. SOC.,vol. 64, no. 591, Mar. 1960, pp. 164-167.

79. Kuenzi, E.W.: Flexure of Structural Sandwich Construction. Rept. 1827, Forest
Products Lab. (Madison, Wis.), 195 1.

80. Weikel, R.C.; and Kobayashi, A.S.: On the Local Elastic Stability of Honeycomb
Face Plate Subjected to Uniaxial Compression. J. Aerospace Sci., vol. 26, no. 10,
Oct. 1959, pp. 672-674.

81. Plantema, F.J.: Sandwich Construction, the Bending and Buckling of Sandwich
Beams, Plates, and Shells. John Wiley & Sons, Inc., 1966.

82. Habip, L.M.: A Survey of Modern Developments in the Analysis of Sandwich

Structures. Appl. Mech. Rev., vol. 18, no. 2, Feb. 1965, pp. 93-98.

83. Allen, H.G.: Analysis and Design of Structural Sandwich Panels. Pergamon Press,
Ltd. (London), 1969.

84. Sullins, R.T.; Smith, G.W.; and Spier, E.E.: Manual for Structural Stability
Analysis of Sandwich Plates and Shells. NASA CR-1457, 1969.

85. Kicher, T.; and Mandell, J.P.: A Study of the Buckling of Laminated Composite
Plates. AIAA J., vol. 9, no. 4, Apr. 1971, pp. 605-613.

86. Ashton, J.E.; and Love, T.S.: Experimental Study of the Stability of Composite
Plates. J. Composite Materials, vol. 3, no. 2, Apr. 1969, pp. 230-242.

87. Ashton, J.E.; Halpin, J.C.; and Petit, P.B.: Primer on Composite Materials:
Analysis. Technomic Pub]. Co. (Stamford, Conn.), 1969.

88. Calcote, L.R.: The Analysis of Laminated Composite Structures. Van Nostrand-
Reinhold Co. (New York), 1969.

37
89. Davis, J.C., Jr.; and Zender, G.W.: Compressive Behavior of Plates Fabricated
from Glass Filaments and Epoxy Resin. NASA TN D-3918, 1967.

90. Hadcock, R., et al. : Advanced Composite Wing Structures Boron-Epoxy Design
Data, vol. 11-Analytical Data. Prepared for Air Force Material Lab, under
Contract F 3361 5-68-C-1301, Grumman Tech. Rept. AC-SM-ST-8085, Nov. 1969.

91. Nishino, F.; Ueda, Y.; and Tall, L.: Experimental Investigation of the Buckling of
Plates with Residual Stresses. Paper in Test Methods for Compression Members.
ASTM STP 419, 1967.

92. Anon.: Metallic Materials and Elements for Aerospace Vehicle Structures.
MIL-HDBK-SA, Department of Defense, Washington, D.C., Feb. 8, 1966.

93. Anon.: Structural Sandwich Composites. MIL-HDBK-23A, Department of

Defense, Washington, D.C., Dec. 30, 1968.

94. Anon.: Aerospace Structural Handbook. AFML TR 68-1 15, Wright-Patterson

AFB, 1969.

95. Anon.: Cryogenic Materials Data Handbook. AFML TDR 64-280, Wright-
Patterson AFB, Aug. 1964.

96. Whitney, J.M.; and Leissa, A.: Analysis of Simply Supported Laminated
Anistropic Rectangular Plates. AIAA J., vol. 8, no. 1, Jan. 1970, pp. 28-33.

97. Flugge, W.: Stresses in Shells. Springer-Verlag (Berlin), 1960.

98. Peterson, J.P.: Buckling of Stiffened Cylinders in Axial Compression and

Bending-A Review of Test Data. NASA TN D-5561, 1969.

99. Miura, K.: Thermal Buckling of Rectangular Plates. J. Aerospace Sci., vol. 28,
no. 4, Apr. 1961, pp. 341-342.

100. Pope, G.G.: The Buckling of Plates Tapered in Thickness. Rept. Struct. 272,
Royal Aircraft Estab. (London) Oct. 196 1.

101. Krivetsky, A.: Plasticity Coefficients for Plastic Buckling of Plates and Shells. J.
Aeron. Sci. (Readers Forum), vol. 22, no. 6, June 1955, pp. 432-435.

38
102. Peterson, J.R.; Whitley, R.O.; and Deaton, J.W.: Structural Behavior and
Compressive Strength of Circular Cylinders with Longitudinal Stiffening. NASA
TN D-125 1. 1962.

103. Winter, G.: Strength of Thin Steel Compression Flanges. Trans. ASCE, Vol. 1 12,
1947, p. 527.

104. Musgrove, M.D.; and Mortensen, R.E.: Development of Titanium Structural

Element Allowables for the Boeing SST. Bound Volume of Technical Papers on
Structures, AIAAlASTM Eleventh Structures, Structural Dynamics, and Materials
Conference (Denver), Apr. 1970.

105. Schmit, L.A.; and Kicher, T.P.: Structural Synthesis of a Symmetric Waffle Plate.
NASA TN D-1691,1962.

106. Semonian, J.W.; and Peterson, J.P.: An Analysis of the Stability and Ultimate
Compressive Strength of Short Sheet-Stringer Panels with Special Reference to
the Influence of the Riveted Connection Between Sheet and Stringer. NACA
Rept. 1255, 1956.

107. Rockey, K.C.: Shear Buckling of Thin-Walled Sections. Thin-Walled Structures,

A.H. Chilver, ed., John Wiley & Sons, Inc., 1967.

108. Suarez, J.B.: Advanced Composite Wing Structures, Stability Analysis of

Advanced Filamentary Composite Panels. Prepared for Advanced Composite
Division, Air Force Materials Lab, under Contract F33615-68-C-1301,
Grumman Tech. Rept. AC-SM-8087, May 1970.

109. Gellatly, R.A.; .Bijlaard, P.P.; and Gallagher, R.H.: Buckling of Sandwich Panels
Under Nonuniform Stress. J. Aircraft, vol. 2, no. 3, May-June, 1965, pp. 224-227.

110. Harris, L.A.; and Auelmann, R.R.: Stability of Flat, Simply Supported,
Corrugated-Core Sandwich Plates Under Combined Loads. J. Aerospace Sci.,
vol. 27, no. 7, July 1960, pp. 525-534.

11 1. Anderson, M.S.: Local Instability of the Elements of a Truss-Core Sandwich Plate.

NACA TN 4292,1958.

112. Mathauser, E.E.; and Pride, R.A.: Compressive Strength of Stainless Steel
Sandwiches at Elevated Temperature. NASA Memo 6-2-59L, 1959.

39
113. Melosh, R.J.; and Palacol, E.L.: Manipulation Errors in Computer Solution of
Structural Equations. Philco-Ford (Palo Alto) Document EM 12-69, Dec. 1969,
NASA CR 1784, Mar. 1971.

114. Fraeijs de Veubeke, B.F.: Displacement and Equilibrium Models in the Finite
Element Method. Stress Analysis, O.C. Zienkiewicz and G. Holister, eds., John
Wiley & Sons, Inc., 1965.

1 15. Pian, T.H.H.; and Tong, P.: Basis of Finite Element Methods for Solid Continua.
Int. J. Num. Meth. Engng., vol 1, no. 1 , Jan. 1969, pp. 3-28.

1 16. Harris, H.G.; and Pifko, A.B.: Finite Element Application to Elastic, Plastic, and
Anistropic Buckling of Stiffened Rectangular Plates. Proc. Symp. on Application
of Finite Element Methods in Civil Engineering (Vanderbilt Univ.), Nov. 1969.

117. Marcal, P.V.: Large Deflection Analysis of Elastic-Plastic Plates and Shells. Proc.
ASME First International Conf. on Pressure Vessels (Delft, Netherlands), Sept.
1969.

118. Holand, I.; and Bell, K., eds.: Finite Element Methods in Stress Analysis. TAPIR,
Technical Univ. of Norway Press (Trondheim, Norway), 1969.

119. Gallagher, R.H.: Finite Element Analysis of Plate and Shell Structures. Proc.
Symp. on Application of Finite Element Methods in Civil Engineering (Vanderbilt
Univ.), Nov. 1969.

120. Barrois, W.: Possible Effects of Service Changes in Compression Stress-Strain

Curves on Buckling. AGARD Rept. 564, Apr. 1968.

12 I . Anon.: Metals: Test Methods. Federal Test Method Standard No. 15 I . General
Services Administration (Washington, D.C.), J d y 1956.

122. Anon.: General Test Methods, Pt. 30 of ASTM Standards, May 1967

123. Stowell, E.Z.; Heimerl, G.J.; Libove, C.; and Lundquist, E.E.: Buckling Stresses
for Flat Plates and Sections. Trans. ASCE, vol. 117, 1952, pp. 545-575.

124. Peterson, J.P.: Influence of Specimen Design and Test Procedure on Results of
Buckling Tests of Shell Structures. Test Methods for Compression Members.
ASTM STP 4 19, 1907.

40
125. Taylor, J.: Experimental Methods in High-Temperature Structural Research. High
Temperature Effects in Aircraft Structures. N.J. Hoff, ed., Pergamon Press Ltd.
(London), 1958, ch. 15.

126. Heldenfels, R.R.; and Roberts, W.M.: Experimental and Theoretical Determina-
tion of Thermal Stresses in a Plate. NASA TN D-2769, 1962.

127. Gallagher, R.H.; Quinn, J.F.; and Turrentine, D.: Techniques for Testing
Thermally Affected Complex Structures. Experimental Mechanics, vol. 1 no. 8,
Aug. 1961, pp. 41-49.

128. Southwell, R.V.: On the Analysis of Experimental Observations in Problems of

Elastic Stability. Proc. Roy. SOC.(London), Series A, vol. 135, 1932, p. 601.

129. Roorda, J.G.: Some Thoughts on the Southwell Plot. Proc. ASCE, J. Eng. Mech.
Div., vol. 93, no. EM6, Dec. 1967, pp. 37-48.

130. Wang, C.T.: inelastic Column Theories and an Analysis of Experimental Observa-
tions. J. Aeron. Sci., vol. 15, no. 5, May, 1948, pp. 283-292.

131. Yoshiki, M.; and Fujita, Y.:Determination of Plastic Buckling Load for Axially
Compressed Plates from Experimental Data. Test Methods for Compression
Members, ASTM STP 419,1967.

132. Kuenzi, E.W.; Erickson, W.S.; and Zuhn, J.J.: Shear Stability of Flat Panels of
Sandwich Construction. Rept. 1560 Forest Products Lab. (Madison, Wisc.),
revised May, 1962.

41
 NASA SPACE VEHICLE DESIGN CRITERIA
MONOGRAPHS ISSUED TO DATE

SP-800 1 (Structures) Buffeting During Atmospheric Ascent, May 1964 -

Revised November 1970
SP-8002 ( S tructures) Flight-Loads Measurements During Launch and
Exit, December 1964
SP-8003 (Structures) Flutter, Buzz, and Divergence, July 1964
SP-8004 (Structures) Panel Flutter, July 1964
SP-8005 (Environment) Solar Electromagnetic Radiation, June 1965 -
Revised May 197 1
SP-8006 (Structures) Local Steady Aerodynamic Loads During Launch
and Exit, May 1965
SP-8007 (Structures) Buckling of Thin-Walled Circular Cylinders, Sep-
tember 1965 - Revised August 1968
SP-8008 (Structures) Prelaunch Ground Wind Loads, November 1965
SP-8009 (Structures) Propellant Slosh Loads, August 1968
SP-80 10 (Environment) Models of Mars Atmosphere (1967), May 1968
SP-80 1 1 (Environment ) Models of Venus Atmosphere (1 968), December
1968
SP-80 12 (Structures) Natural Vibration Modal Analysis, September 1968
SP-80 13 (Environment) Meteoroid Environment Model - 1969 [Near
Earth to Lunar Surface], March 1969
SP-80 14 (Structures) Entry Thermal Protection, August 1968
SP-80 1 5 (Guid an ce Guidance and Navigation for Entry Vehicles, No-
and Control) vember 1968
SP-80 16 (Guidance Effects of Structural Flexibility on Spacecraft
and Control) Control Systems, April 1969
SP-80 17 (Environment) Magnetic Fields - Earth and Extraterrestrial,
March 1969
SP-80 18 (Guidance Spacecraft Magnetic Torques, March 1969
and Control)
SP-80 19 (Structures) Buckling of Thin-Walled Truncated Cones, Sep-
tember 1968
SP-8020 (Environment) Mars Surface Models (1 968), May 1969
SP-802 1 (Environment) Models of Earth’s Atmosphere (1 20 to 1000 km),
May 1969

43
SP-8022 (Structures) Staging Loads, February 1969
SP-8023 (Environment) Lunar Surface Models, May 1969
SP-8024 (Guid an ce Spacecraft Gravitational Torques, May 1969
and Control)
SP-8025 (Chemical Solid Rocket Motor Metal Cases, April 1970
Propulsion)
SP-8026 (Guidance Spacecraft Star Trackers, July 1970
and Control)
SP-8027 (Guidance Spacecraft Radiation Torques, October 1969
and Control)
SP-8028 (Guidance Entry Vehicle Control, November 1969
and Control)
SP-8029 (Structures) Aerodynamic and Rocket-Exhaust Heating During
Launch and Ascent, May 1969
SP-8030 (Structures) Transient Loads from Thrust Excitation, February
1969
SP-803 1 (Structures) Slosh Suppression, May 1969
SP-8032 (Structures) Buckling of Thin-Walled Doubly Curved Shells,
August 1969
SP-8033 (Guidance Spacecraft Earth Horizon Sensors, December 1969
and Control)
SP-8034 (Guidance Spacecraft Mass Expulsion Torques, December
and Control) 1969
SP-8035 (Structures) Wind Loads During Ascent, June 1970
SP-8036 (Guidance Effects of Structural Flexibility on Launch Vehicle
and Control) Control Systems, February 1970
SP-8037 (Environment) Assessment and Control of Spacecraft Magnetic
Fields, September 1970
SP-8038 (Environment) Meteoroid Environment Model - 1970 (Interplane-
tary and Planetary), October 1970
SP-8040 (Structures) Fracture Control of Metallic Pressurc Vessels, May
1970
SP-804 1 (Chemical Captive-Fired Testing of Solid Rocket Motors,
Propulsion) March 197 1
SP-804 2 (Structures) Meteoroid Damage Assessment, May 1970
SP-8043 (Structures) Design-Development Testing, May 1970
SP-8044 (Structures) Qualification Testing, May 1970
SP-8045 (Structures) Acceptance Testing, April 1970
SP-8040 (Structures) Landing Impact Attenuation for Non-Surface-
Planing Landers, April 1970
SP-8047 (Guidance Spacecraft Sun Sensors, June 1970
and Control)

44
 SP-8048 (Chemical Liquid Rocket Engine Turbopump Bearings, March
Propulsion) 1971
SP-8050 (Structures) Structural Vibration Prediction, June 1970
SP-805 1 (Chemical Solid Rocket Motor Igniters, March 197 1
Propulsion)
SP-805 3 (Structures) Nuclear and Space Radiation Effects on Materials,
June 1970
SP-8054 (Structures) Space Radiation Protection, June 1970
SP-8055 (Structures) Prevention of Coupled Structure-Propulsion Insta-
bility (Pogo), October 1970
SP-8056 (Structures) Flight Separation Mechanisms, October 1970
SP-805 7 (Structures) Structural Design Criteria Applicable to a Space
Shuttle, January 1971
SP-8058 (Guidance Spacecraft Aerodynamic Torques, January 197 1
and Control)
SP-8059 (Guidance Spacecraft Attitude Control During Thrusting
and Control) Maneuvers, February 197 1
SP-8060 (Structures) Compartment Venting, November 1970
SP-806 1 (Structures) Interaction with Umbilicals and Launch Stand
August 1970
SP-8062 (Structures) Entry Gasdynamic Heating, January 197 1
SP-8063 (Structures) Lubrication, Friction, and Wear, June 197 1
SP-8066 (Structures) Deployable Aerodynamic Deceleration Systems,
June 1971
SP-8068 (Structures) Buckling Strength of Structural Plates, June 197 1
SP-8072 (Structures) Acoustic Loads Generated by the Propulsion System,
June 1971

NASA-Langley,1971 -3 2 45
NATIONAL AND SPACE ADMINISTRAI
AERONAUTICS ION
D. C. 20546
WASHINGTON,
FIRST CLASS MAIL
~

OFFICIAL BUSINESS
PENALTY FOR PRIVATE USE $300
POSTAGE A N D FEES PAID
NATIONAL AERONAUTICS ANI
SPACE ADMINISTRATION

If Undeliverable (Section 158

Porn1 Manual ) Do Not h r n