Escolar Documentos
Profissional Documentos
Cultura Documentos
C_
/_,_c,9
7//2._
_'7
ELASTIC
CONSTANTS SANDWICH
FOR
CORRUGATED-CORN PLATES
By
Charles Langtey
and
RM.ph
E.
I--Iubka
Laboratory Va.
)
Field,
,Washington ebruary
REPRODUCED BY "
i_-_
:j
T]-IIS BFST
DOCUMENT COPY
HAS FURNISHED
BEEN US IT IS
REPRODUCED BY THE
THE
AGI,__,_CY, i/_I_
RECOGNIZED IT OF MAKING
CERRE-
ILLEGIBLE,
INTEREST AS
INFORMATION
POSSIBLE,
r
U.S. DEPARTMENT OF COMMERCE Service
NACA-TN-2289
ELASTIC SANDWICH
CONSTANTS PLATES
FOR
CORRUGATED-CORE
Langley Langley
Aeronautical Field, VA
Laboratory
Feb
51
NATIONAL DVISORY OMMITTEE A C FORAERONAUTICS TECHNICAL NOTE2289 ELASTICCONSTANTS FORCORRUGATED-CORE SANDWICH PLATES By Charles Libove and Ralph E. Hubka SUMMARY The sandwich plate consisting of corrugated sheet fastened between two face sheets is considered. Application of existing theories to the analysis of such a sandwich plate requires the knowledge of certain elastic constants. Formulas and charts are presented for the evaluation of these constants. The formulas for three of these constants were checked experimentally and found to give values in close agreement with the experimental values. INTRODUCTION A type of sandwich plate for which practical use has recently been found in airplane-wing construction consists of a corrugated metal sheet fastened, at its crests and troughs, to two ordinary metal sheets (see_ for example, fig. i). The main advantage of this t_e of sandwich is that the corrugated-sheet core not only serves to separate the faces and, ther@by, to achieve high flexural stiffness, but it also carries a share of any compressive loading applied parallel to the corrugations and any edgewise shear loading. This type of sandwich has been called cardboard-box construction (reference i) and also double-skin construction. It is referred to herein as corrugated-core sandwich plate. Plate theories applicable to the symmetrical type of corrugated-core sandwich, illustrated in figure l(a), have been developed in reference 2 for flat plates and in reference 3 for curved plates, a These theories are essentially homogeneous orthotropic-plate theories extended to include deflections due to transverse shear, which can be significant for the corrugated-core sandwich plate because of the relatively flexible
core.
plates
reference 4 of referring to sandwich l(a) as symmetrical is adhered to in this figure is also called
shown
NACA
TN 2289
Application of the general sandwich-plate theories of references 2 and 3 to any particular type of sandwich requires a knowledge of certain elastic constants for that type of sandwich plate. These constants describe the distortions associated with simple loadings. They include two Dx and transverse and Ey_ Dy_ a shear stiffnesses stiffness Gxy3 two DQx and DQy, two bending moduli _x x and and _' y stiffnesses Ex _y
a twisting
modulus and
Poisson's ratios
associated associated
bending_ stretching.
Poisson's
The purpose of the present paper is to present formulas for evaluating these elastic constants for the corrugated-core type of sandwich plate. For the sake of completeness, formulas are also developed for evaluating the additional elastic constants that would needed for a unsymmetrical by Cxx_ Cxy_ rigorous type of Cy x, by extension sandwich. Cyy, and of the These T_ sandwich-plate theories to the additional constants_ denoted coupling The - for example, and
be
the
curvatures
produced
derivation
formulas
same as between
corrugation flats and face sheets is neglected_ but are extended slightly to include the effects of stretching of the corrugation (in addition to bending) and the prevention of anticlastic curvature in the elements of the sandwich plate. The former effect can be important when the sandwich cross section approaches a truss_ the latter_ because the length of the sandwich plate parallel to the corrugation axis is several times the corrugation pitch. The results obtained for the bending and twisting stlffnesses Dx, Dy, and Dxy for the symmetrical sandwich correspond to the slightly less precise formulas of reference 5. (Transverse shear stiffness was not evaluated in this reference. A slight difference in definition the present Because of the paper.) the formulas developed are generally rather involved_ charts symbols Dx and Dy exists between reference 5 and
are presented for one of them_ the transverse shear stiffness DQy, and approximations are given for several of the others_ together with the results of numerical investigations of the accuracy of these approximations. In calculating the charts and in investigating the accuracy of approximate formulas_ a family of corrugation shapes consisting of straight lines and circular arcs was considered. The bend radii of the corrugation_ measured to the center line, were generally taken as 0.18 times the corrugation depth hc_ but departures from this value were also considered_ as were departures from symmetry.
NACA 2289 TN
As a check on the formulas, bending and twisting tests were run on samples of a corrugated-core sandwich plate. Experimental values of bending stiffness Dy, transverse shear stiffness DQy_and twisting stiffness Dxy were obtained and comparedwith the theoretical values. The function of the elastic constants in a sandwich-plate theory is first briefly described. A section follows in which the formulas for the elastic constants for the corrugated-core sandwich are summarized. The tests and comparison between theory and experiment are then described_ a discussion section follows, and a section of concluding remarks ends the body of the paper. The symbols used in the body of the paper are listed and defined in appendix A. A number of them are also defined in the text where they first appear. Appendixes B to E contain the theoretical derivations. THEFUNCTION OFTHEELASTICCONSTANTS IN SANDWICH-PIATE THEORY The sandwich-plate theories of references 2 and 3 are based on a structural idealization of the sandwich as a plate of continuous construction with material which is orthotropic with respect to the mutually perpendicular x-, y-, and z-directions. The modulus of elasticity in the z, or thickness, direction is assumedto be infinite_ that is, local buckling of the faces is not considered and the over-all thickness is assumedto remain constant. Straight material lines normal to the middle surface are assumedto remain straight, but not necessarily normal to the middle surface, during distortion of the plate. This idealized structure can adequately represent a corrugated-core sandwich plate of either the symmetrical or unsymmetrical type for many practical purposes, provided the core has sufficient stiffness to keep the over-all thickness of the plate essentially constant and provided the plate width (perpendicular to the corrugation axis] is manytimes the corrugation pitch. If the symmetrical type of sandwich (fig. iCa)) is to be represented, then the elastic properties of the idealized-plate material may be regarded as varying symmetrically about the middle surface through the thickness. In order to represent the behavior of the unsymmetrical type of sandwich (fig. l(b)), the elastic properties of the idealized-plate material must be thought of as varying nonsymmetrically with respect to the middle surface. The behavior of a differential element of the idealized sandwich plate under load can be described by a set of force-distortion relationships. For an element of the symmetrical type of idealized sandwich
NACA 2289 TN
(fig. l(a)), subjected to forces and momentsas shown in figure 2(a), these relationships, as developed in references 2 and 3, are
_2w Mx _! 1 _Qx _x --Dx + _ My + _ x_2
(1) (2)
_2w _x
_y2 - D x Mx
_
- Dy
i--_
+ DQy _y
_x -
Nx
_'y
Ex
F 7
Ny
(3)
_'x Nx + Ny
_Y = Ex Ey
(4)
_2w -_+
_x _y Dxy
_Qx _)y
1 +---
1 _x
(5)
2 DQx
2 DQy
-- _
7_ 82w where surface quantities the loading distortion are amount zero of 82w ---_3 and 82w Gxy
(6)
_x
2'
_y
_x,
_x _y
and _x, the Fxy
are are
the the
and of in the
twist middle
of
the
middle The of a
and
surface.
then, in the
according x-direction
_ _1 Dx of
more
of the unsymmetrical type of sandwich (fig. that of the symmetrical type. In particular, among the distortions may in general produce
l(b)) is a certain
The same type of coupling can be expected in a symmetrical sandwich subjected to unsymmetrical loading. In setting up force-distortion relationships for an element of the unsymmetrical type of sandwich, loading on the element will be generalized as shown in figure 2(b). The the forces middle Nx, plane; Ny, and has Nxy an are no longer plane assumed of to be applied
the in "
each
arbitrary
application,
denoted
NACA
TN 2289
by I, II_ and Ill_ respectively. measured in these same respective ships for equations the (i) element to (6): are then
given
following
_2w
_x-7 - o-_ Dy = +
Mx
_Y My
ICxxNx
c_:
+ DQx-
_Ov
(i')
_2w
_x
My+,
'
',%xNx + C_y,,
+ D% _y
(2')
(3')
Nx - Cyx_, I + -- - _'Y Ny Ex Ey
% : ',Cx_x- c_: i I
- _'___x _z Nx +
Ex %
(4')
8x 8y Dxy
'TNxy' ,
+ , 2 DQx ()y
+-2 DQy
(5') 8x
7xy
, = ' 2TMxy ,
(6')
The
boxed
terms
are
the The
terms
that
have
been
added
to
express so on
the boxed
coupling
behavior.
coefficients
Cxx _
Cxy _ and
terms are the coupling elastic constants. coupling elastic constant in two equations reciprocity reciprocity theorem theorem for are elastic that _x -Dx structures. = _y Dy and
consequences
of
the
some chosen
be
T may
three special
6 cases, proper choice of locations in still further uncoupling. For choosing these planes to coincide causes all the coupling constants
NACA 2289 TN of planes I, II, and III will result the symmetrical sandwich_ of course, at the middle surface of the plate to vanish.
THEORETICAL RESULTS Elastic Constants for Symmetrical Sandwich In appendixes B to E, derivations are madeof formulas for the elastic constants for the general corrugated-core sandwich plate. The formulas obtained are nowgiven in reduced form for use in conjunction with the force-distortion equations (1) to (6) for the symmetrical sandwich plate. Generally, the subscript C denotes the core, and the subscripts 1 and 2 denote the lower and upper faces, respectively. In this section, however, only symmetrical sandwiches are considered and the subscript 1 is used for both faces. It should be kept in mindj therefore, that the definitions of manyof the terms appearing in the following formulas for the elastic constants apply only to the symmetrical type of sandwich.
Bending bending stiffnesses.Dx The and formulas Dy are obtained in appendix B for the
stiffnesses
Dx
= EYx
(7)
1where i + _ El tl h2
2 1 - 7x]
_x
= EC_C
1 E-Iy = _ Eltl h2
ratio
of
face of of
E1 EC
elasticity elasticity
material,
_CA TN 2289
I
7
moment of inertia, area of each per about face middle unit middle sheet, width, plane, inches of face sheets, inches of corrugation inches3 cross-
IC
between
surfaces
For core
practical is often to
the
of
inertia of
_C
by
the
compared sections
moment
inertia
faces In
to the the
corrugations.
is very may be
nearly
following
approximation
equation
made
(8,)
This approximation implies a neglect corrugation on the Poisson expansion Results of a numerical survey of the given type in table at I the for top the of symmetrical the table. of the restraining effect of the or contraction of the face sheets. accuracy of this apDroximatio_ are sandwich The table = 1.0 gives the of the of common the
shown
ratio
as computed from equation (8'), to the exact from equation (8). The error in the approxiover a large part of the range of conextreme cases, no more than 6 percent. with bending..bending The _x formulas and _y obtained are
Poisson's
ratios
associated
with
_x
= _i
(9)
:
Extensional tne case, extensional are stiffnesses.stiffnesses Ex The
Dx
(lO)
obtained reduced to in appendix the B for
formulas Ey,
and
symmetrical
Ex = EAx
(ll)
NACA
TN
2289
%
Ey =
(12)
i
where EA--- : EcA C x + 2Elt I
12(l 1
E-Ay = 2Elt I
B
area, per unit perpendicular If, once again, or zero ion: the restraining
expansion taken as
contraction and,
of the
therefore,
approximat
Ey ~
The error in this in the _x approximation to larger is somewhat Dy, since its larger the than the error of
(12')
obtained core to
approximation
contribution to of the
is relatively
than
contribution values
El x. ratio
The with
formulas extension
obtained _'x
x=_l
_z
_'y = _'x Ex
(14)
formula Dxy:
was
obtained
in
Oxy = ea-7
(15)
NACA _N 2289
where i _-y : _ Gltl h2
GI The stiffness
shear Dxy
modulus is
of
of
face
sheet of
psi since be
properties
symmetry zero.
requires
that
in the
corrugated-core
sheet
shear appendix
stiffness.C) by
The
horizontal
shear
stiffness
Gxy
a_
where Gctc GA= 2 + 2Glt I
= G--Z
(16)
Xc
GC tC
shear
modulus of
of
elasticity
of
core
material, inches
psi
corrugated-core
sheet,
axis.-
stiffness in planes perpendicular to corrugation shear stiffness in planes perpendicular to the is given (see appendix D) by the formula
corrugation
= Sh _
where
(17)
depth of corrugation, measured line at crest to center line appendix PC S Poisson's D), ratio inches of core material
vertically at trough
of
coefficient
depending
upon
shape
of
i0
NACA 2289 TN
Formulas for evaluating S are derived in appendix D. Because of the complexity of these formulas, a number of charts were computed which give S directly for the common type of sandwich with corrugation crosssectional shape consisting of straight lines and circular arcs. The charts of figure 3 are for the case in which the core and faces have the samematerial properties. They give S for a wide range of RC I geometric proportions but are restricted to the value 0.18 for --, hC where RCI is the corrugation center-llne bend radius. This restriction was madeprimarily for computational convenience, but it is generally consistent with corrugation shapes that have been considered for sandwich construction. The effect on S of departing from the value 0.18 for RC I can be estimated from figure 4(a), where a number of curves of S are hc RCI given for values of -of 0.12 and 0.24 as well as 0.18. Cross plots hC based on the charts of figure 3 would indicate that S becomesrelatively insensitive to the ratio
was not
For
(17).
included
than
are 4.30
plotted (steel 3,
for
values
of
_ii
of
0.23 along of
core, basic to
steel curves, be
faces) from
and
core, Ec for E1
magnesium - 1.00.
faces) The
figure
value in
seen
insensitive to large and the face sheets. If bot,h departures taneously be obtained determined <that is,
differences
elastic
modulus
between
from
the
conditions and
of
figure
3 occur effect
simulon S as may
RCI
_ 0.18h C by 4(a)
effects
figure
of
NACA 2289 TN
Ii
appendix D, used in conjunction with the auxiliary equations (D20) and (DI5), with ky and kz taken as i. If, besides being symmetrical, the corrugation center line consists of straight lines and circular arcs, then equations (D22) and (D23) or (D24) may be used instead of equations (DIS). This system of equations was used to compute the charts previously described.
A to Transverse shear general formula for the corrugation stiffness in planes parallel to corrugation axis.the transverse shear stiffness in planes parallel DQx , as derived in appendix E_ is
axis
Dqx
Gc It c h Z Pro Q ds
( 18 )
where I moment of inertia of to yz-plane_ taken y-axis, inches 4 corrugation pitch, width about 2p of cross section parallel centroidal axis parallel to
2p
length of one corrugation line, inches (see fig. coordinate measured inches (see fig. The quantity Q is the static plane for symmetrical sandwich) along E-3)
center
line
of
corrugation
leg,
If materials having different moduli of elasticity are used for the and faces_ a transformed cross section should be used in computing and Q. An approximate in the derivation, by the face corrugation, formula, a bending which moment is more practicable_ Mx is assumed to constant is thus is obtained be resisted
ifj only
in the
Gctc h2
_ GCfC 2
(18')
of
equation
(18')
NACA 2289 TN
The general formulas for the elastic constants derived in appendixes B to E are now to be discussed. These formulas, used in conjunction with the force-distortion equations (i') to (6'), describe the distortions of an element of either the symmetrical or unsymmetrical sandwich plate loaded as shown in figure 2(b). The symbols appearing in the formu/as are defined in appendix A.
Elastic formulas Dy, given _x, by for _y, constants the associated with flexure with Cxx_ and extension.and Cyx, extension and Cyy General Dx, are apply
flexure Cxy, B.
Ex,
equations
(B36)
appendix
These
formulas
to a sandwich with arbitrarily shaped corrugation, in and lower face sheets may differ in thickness, modulus Poisson's ratio and in which the loading planes I and chosen. Appreciable practical face then sheet apply. It is (B25') evident from both sets of to (B36') that the values these equations of the constants simplification that are the equal of the Poisson's (_2 formulas ratios results of the
from upper(B25')
assumption materials
= _i)"
Equations
and
Ey, _'x, _'y) and the coupling constants dependent upon the location of planes I and forces at Nx the and Ny, respectively, of the are
applied. cross
forces
applied
centroids
transformed
JJ
sections
_<t ot
simplification of the formulas reduce to equations (B25") to takes p1_ce. (B36"). The sandwich the approximations by equations to (8') Dy and when used, and (12') kll Ey may given also
Equations
(B25')
to
(B36')
unsymmetrical
sandwich are
= k_y
and Ely
= _i" E-Ay
expressions
however,
from their general formulas (B20) of appendix B. Table of the the _ccuracy unssqnmetrical
appendix results
NACA TN
2289
13
from order
use for
expressions sandwich as
constants associated with twisting and the constants associated with twisting Gxy , and C. The T are of given Gxy by and equations T depend
(C35), upon
values
location
is applied. Locating the the cross section (that by equation constant formulas (C31) T for to this or in
the
coupling The
vanish and simplifies the case are equations (C35'), As for the constants
and
extension_
a is
formula for Dxy occurs if the corrugation Equation (C35) then gives the following
Dxy
= 2G--_
h2
(19)
The results of a numerical are given in table I. The mate formula is seen to be case = i , no error at
survey of the accuracy error incurred through generally quite small. all results from neglect
of this approximation the use of the approxiFor the symmetrical of the core since
symmetry
requires
flow
to
be
Transverse axis.Equation
perpendicular shear
transverse
stiffness
for the symmetrical sandwich also applies to the unsymmetrical sandwich provided the coefficient S is obtained from formulas or charts which apply specifically to the unsymmetrical sandwich. Figure 3 gives extensive charts for evaluating S for a symmetrical sandwich with faces and core of the same material and with the corrugation center line consisting of straight lines and circular arcs_ the latter having a radius of curvature of 0.i$_. Figure 4(a) shows the effect of using a radius of curvature other than 0.18hc, and figure 4(b), the effect ol_ using core material different from that of the faces. The rest of figure 4 is devoted to showing separately the effects on S of two departures from symmetry for a sandwich that is otherwise the saF_e as that c:onsidered in figure J. Figure 4(c) is f'or a case in which the nonsymmetry is due to the core and consists in the lower and upper flats
14
NACA 9289 TN
being of unequal width; figure 4(d) applies when the core is symmetrical but the faces are of unequal thickness. No chart is given for the case in which the core is symmetrical and the face thicknesses equal but in which the nonsymmetry arises from the use of a different material for the lower face than for the upper face. However, for nonsymmetry of this type, S can generally be cbtained quite accurately by assuming, first, that both faces are of the upper-face material and_ next, that both faces are of the lower-face material and averaging the two values S1 and S2 thus obtained in the following manner:
s-7-or
1.26 S =
SIS 2
SI 3 +
$23 from the lower face, either determined approximately by the two values obtained by as the upper face and next that The error in such an approxima-
In general, when the upper face is different in thickness or material or both, S can be averaging in the previously described manner first assuming that both faces are the same both faces are the same as the lower face. tion will generally be less than 3 percent.
be
(Dlg)j
itself ky
simplification equations.
from
taking
= kz = 1
If the corrugation center llne circular arcs_ then equations (D21) instead of equations (D15). If, in symmetrical and if ky and kz then equations (D22) may replace Transverse Equations (18) stiffness sandwich. applied to DQx The the
consists of straight lines and and (D23) or (D2_) may be used addition, the corrugation is as 1 in (D21). equations (D18),
shear stiffness in planes parallel and (18') for the evaluation of the for a symmetrical sandwich also
apply
15
General Summary The elastic-constant formulas that were thought to need experimental verification were those which dependedto a large extent in their derivation upon the assumption that the thickness of the core remains essentially constant or that the corrugation cross section is undistorted. Amongthese, the formulas for Dy, DQy, and Dxy were selected for checking because these constants could be experimentally evaluated through simple bending and twisting tests on sandwich beamsand panels as described schematically in appendix A of reference 2. The test sandwich was of the symmetrical type. The core consisted of a readily available Alclad 24S-T36 aluminum-alloy standard circularly corrugated sheet having a nominal thickness of 0.032 inch and a nominal over-all depth of 3/4 inch. The faces were of 24S-T3 aluminum-alloy sheet having a nominal thickness of 0.064 inch. Two test specimens were used: A beamfor the evaluation of Dy and DQy and a panel for the evaluation of Dxy. Although blind riveting was necessary only on one side of the panel, it was used on both sides in order to maintain symmetry. On the beamdriven rivets were used in both faces since the beamwas relatively narrow. The results in the following properties were EC = i0_300,000 of the tests and comparisons with theory ar@ wsummarized table. In computing the theoretical values the following assumed: E1 = E2 = 10,500,000 pounds per square inch_ pounds per square inch, and _I = _2 = 7"
D%
( in. -ib ) Range of experimental values Theoretical value
aComputed beam tested was curvatures were with ( ib/in. 4010 )
182,000 4310
a4300
177,000
= _C
= 0
because
the
relatively therefore
narrow assumed
NACATN 2289
are
shown
in
figure 5(a). The beam was supported on two knife edges as shown in figures 5(b) and 5(c) and loaded near the ends so as to obtain a region of pure bending moment between supports. The supports were placed 19.05 inches apart for one test (the test which yielded the value of 221,000 in.-lb 5 pounds up to increments; in for Dy) a maximum a second and loads P were applied in increments of of 25 pounds and then removed in the same test (which yielded the value of 2243000 in.-lb
for Dy), the supports were placed 24.56 inches apart and loads P were applied in increments of 5 pounds up to a maximum load of 30 pounds and removed in the same increments. Deflections of the beam were measured at the locations of O.O001 inch. shown in figure 5(b) with gages having a sensitivity
Despite the fact that spacer blocks were inserted in the sandwich at the supports to prevent local distortion, downward displacements of the upper face were observed immediately above the supports. These displacements, on the order of 2 to 4 percent of the maximum deflections at the center of the beam, were probably caused primarily by thickness change of the beam, since gages placed directly on the supports showed no support displacements. In correcting for the upper-face sheet displacements above the supports, the vertical displacement of the "middle surface" of the sandwich at each support was taken as one-half of the face-sheet displacement. The deflections at points away from the supports _ere then referred to the straight line connecting the middlesurface points immediately above the supports. Away from the supports, gages placed in contact with the lower-face sheet showed that no thickness change occurred in the beam and that the deflection of the upper face could therefore be taken as the deflection of the middle surface. The deflections The above the varied described supports linearly manner resulted of with applied load. the deflection of Dy for the distortions were was
values station
which
independent
of the
whose
deflection
values of Dy were obtained the corrected deflections at the was assumption used:
Dy = (Pd)Ys(S - Ys)
2bws
(20)
17
fig. 5(b))
Ys
from at
left
support Ys,
to
any
station,
inches
w s
station
L b
between
of beam,
This
formula
was
applied
stations,
Ys
and
The
three values thus obtained 2 percent in any test; the the true value of of DQy 6(b)). Dy. DQy.were
from one another by no more than of the three values was taken as
The the
beam same
test as
specimen used to
and in
span
lengths Dy
those
evaluating
The
beam
was
subjected
several
different
a type to produce transverse shear. These schematically in figure 6(a) and the experi-
mental values of DQy obtained from each test are also given. A photograph of a typical test setup is shown in figure 6(b). Deflections were measured between the supports at six stations for the shorter span and at eight stations for the longer span and also immediately above the supports. As in the tests for Dy, slight downward displacements of the upper face were observed immediately above the supports. These displacements were generally of the order of i to 2 percent of the maximum deflection at the center of the beam but in two cases were as high as 3 and 5 percent, respectively, at the right support. The measured deflections were corrected for the distortions above the supports in the manner described load. The curves The for Dy. The deflections varied linearly with the applied
measured as
were which
used
to of
plot DQv
for
a whole, based on
values of
following
formula_
assumption
a uniform
18
NACA
TN 2289
used to spaced
- Ys
(21)
3 _ ys2(2L _ Ys]
12Dy
load applied at each crest pounds (see fig. 6(a)) number of loads P applied
or
each
trough
of
corrugation,
to
the
beam
(see
fig.
6(a))
Dy
bending stiffness per unit width (taken as 221,000 in.-ib when 224,000 in.-ib when L = 24.56 formula was used for the case of
1 DQy = 2wsb i PYs where P The deflections of Ys load on the beam_ pounds formulas and were the Thus_ deflections the formulas 2_ (3L2 4ys 2 )
(22)
substituted of 0.2L 3
at
0.8L.
for each test. These values differed from one by ll percent and their average was taken as the for the sandwich.
value
of wide_
Dxy.was
panel,
inches The
long faces
by of the
twisted the of
bent and
up
along end
sides
one and
panel,
nominal somewhat
thickness wider
3-inch plate
figs. the
7(a)
7(b).) and
steel
remaining
end
NACA TN
2289
19
in turn_ bolted to a rigid backstop with sufficient clearance to permit warping of the plate out of its plane. To the steel plate at the opposite end of the panel was bolted an aluminum-alloy loading plate (not shown) to which the torque was applied. The steel plates were bolted to the sides of the panel in order to help achieve a state of pure twist in the panel. Strain gages were placed back-to-back on the faces and corrugation legs across the width at the midlength of the panel in order to determine to what extent a state of pure twist (that is, constant face shear stress and zero corrugation shear stress) had been achieved. The dimensions of the panel are shown in figure 7(c). Loads were applied in increments of 2000 inch-pounds up to a maximum of i0_000 inch-pounds and removed in the same increments. Deflections of the panel were measured at seven stations across the width at each of four stations along the length (see fig. 7(d)), the stations starting approximately 12 inches from the supported end and spaced approximately 12 inches apart. The measured deflections varied linearly to the both applied across load. The the width the and along the length and the were proportional
_2w
From measured deflections_ Dxy was then twist from the was computed. formula; twisting stiffness obtained
_x _y
1 F_(T T' )
_2w _x _y
where b T T' width applied torque of panel torque, required (21.11 in.)
(23)
,ooo
(see reference 6, equation (156))
20
NACA TN
2289
The linearity of the deflections across the width and along the length indicated that a state of nearly pure twist was being achieved. The strain-gage measurements tended to confirm the existence of this state of pure twist. They showed that, except in the first two cells near each edge of the panel, the face shear stresses were very nearly uniform across the width, with only one value departing as much as 8 percent from the average. In the same region_ the corrugation-leg shear stresses were generally less than I. 5 percent of the face shear stresses. In to the order to investigate evaluation whether of the use of test side was plates repeated was necessary the
experimental
Dxy , the
with
side plates removed. The shear-stress distribution across the width became considerably nonuniform} the deflections were still linear across the width but departed slightly from linearity along the length. The test value of Dxy _ based on the twist in the central portion of the panel, was only about 0.85 as large as the experimental value obtained with the side plates on. This result indicates that side plates are desirable in order to minimize edge effects and achieve a state of pure twist when testing for Dxy.
DISCUSSION
Formulas have been presented for evaluating the elastic constants of a corrugated-core sandwich plate of either the symmetrical or unsymmetrical type. The formulas are rather comprehensive and precise, but reductions to several important special cases have been made and practical approximations to a number of the formulas have been given. Tests have been run to verify the formulas for three of the more important constants and, indirectly, the basic assumptions in their derivations.
The formulas given are limited to plates stressed in the elastic range and not subject to local buckling. Engineering adaptation of the results to cases involving plasticity and local buckling can probably be made; however, attempts at such an adaptation were beyond the scope of the present study. Each component of the sandwich (face sheet or core sheet) is assumed to be composed of homogeneous isotropic material. In actual construction this assumption may be violated by the presence of perforations in one sheet to facilitate the driving of rivets in the other sheet. In evaluating the elastic constants the presence of the perforations can be accounted for approximately by assuming a homogeneous face sheet of reduced modulus.
NACA 2289 TN
21
Whenvalues of the elastic constants for a given corrugated-core sandwich plate are substituted in equations (i) to (6) or (i') to (6'), the resulting equations describe approximately the distortions of an element of the plate under load. The distortions are described only approximately, because the actual plate does not behave in quite the manner assumedfor the idealized plate. In particular, straight material lines in the thickness direction will not remain straight under the presence of shear but will tend to warp. In evaluating the transverse shear stiffness DQx or DQy theoretically, therefore, the problem arises of choosing an average straight line through the warped one in order to define a transverse shear strain for the cross section. Fortunately, for most sandwiches the plausible range for choosing this straight line is small and causes only a slight ambiguity in extending the definition of DQx or DQy to an actual plate. For the corrugatedcore sandwich as analyzed in appendixes D and E, the average straight line was taken as the one passing through corresponding material points in the middle surfaces of the face sheets. This line has the minimum deviation from the true warped line (as determined by least squares) provided the core is ignored and is probably satisfactory whenever the effective contribution of the core to the total cross-sectional moment of inertia is small. The tendency of the originally straight lines to warp introduces a further complication inasmuch as any restraint against such warping (due to the mutual interference of adjacent parts of the plate) will tend to increase the transverse shear stiffness. Such restraint will be small except in the region of concentrated loads. In the theoretical derivations, the conservative assumption was therefore madethat there is no restraint at all against warping. Since the tendency of originally straight lines in the thickness directions to warp is a function of the type of loading, experimental values of DQx or DQy, as determined through beamtests, should, in principle, vary according to the type of spanwise loading distribution used. The variations observed in the tests to determine DQv however, (see , fig. 6(a)) seemedto be caused more by scatter and other factors than by the type of load distribution. Since the primary application of the elastic constants will probably be to sandwich-plate theory, it should be mentioned that the forcedistortion equations (i) to (6) or (i') to (6') represent one component of such a theory. If to these equations are added the differential equations of equilibrium of the element shown in figure 2 and equations relating strains and displacements, the combination of equations will constitute a complete formulation of a sandwich-plate theory. The force-distortion equations (i) to (6) have been presented before in references 2 and 3, but the generalized equations (i') to (6'), which include coupling terms, are believed to be new. The relative importance
22
NACA 2289 TN
of the coupling terms for the corrugated-core sandwich has not been rigorously evaluated; it would dependupon the degree of nonsymmetry of the cross section and the type of problem under consideration. There is reason to believe, however_ that in most cases the effect of coupling will be slight. For a sandwich having faces of the samePoisson's ratio but different thicknesses and having a core momentof inertia and area which approach zero, locating the loading planes I, II, and III at the centroidal plane between the two faces will cause all the coupling constants to vanish. Since the core of practical corrugated-core sandwiches will probably contribute only a small part to the total area of the cross section and a smaller part to the momentof inertia, the coupling constants will very likely be unimportant for properly chosen locations of planes I, !I, and III. In such cases and for someproblems neglecting the coupling terms in equations (i') to (6') maybe sufficiently accurate. CONCLUDING REMARKS In order to facilitate application of an existing sandwich-plate theory to the corrugated-core type of sandwich, formulas and charts have been presented for the evaluation of the necessary elastic constants. Both the symmetrical and unsymmetrical types of corrugated-core sandwich have been considered, and the extensions of the existing sandwich-plate theory required to make it strictly applicable to the unsymmetrical type are indicated. The formulas and charts presented are limited to plates stressed in the elastic range, which are not subject to local buckling. The formulas are rather comprehensive and precise, but reductions to several important special cases have been made. Practical approximations to a number of the formulas have been investigated numerically and found to be sufficiently accurate for most practical cases. The formulas for three of the elastic constants were checked experimentally and found to give values in close agreement with experiment. Langley Aeronautical Laboratory National Advisory Committee for Aeronautics Langley Field, Va. 3 November203 1950
NACATN 2289
23
APPENDIX
SYMBOLS
AND
DEFINITIONS
plane
plane
in which to
Nx faces
acts
and
in which
_x
is measured
parallel plane II
acts
and
in which
Ey
is measured,
plane
III
acts
and
in which
7xy
is measured_
CxDc
produced in x-direction
applied; -Mx3
Cxy
elastic
constant 82w
curvature of unit Ny of
in
8x 2
produced
applied; -Mx_
in y-direction
elastic
constant
representing
curvature
in
_y2
produced
per
unit
of unit
Ny of
applied; -My_
in y-dlrection
_y,. per
elastic
constant
curvature of unit Nx of
in
produced in x-direction
applied_ -My,
DQx,DQy
shear
stiffnesses_ xand
per
unit
width_
of
a beam
y-directions_
respectively_
24 Dx,Dy Dxy
NACATN 2289
bending plate
width 3 of a beam cut from respectively, inch-pounds unit-length to x- and element y-axes,
of per in
plate inch
in x-
and
y-directions,
xy-plane_
pounds in x-
per and
inch
resultant bending-moment intensities y-directionsj respectively_ pounds resultant and twisting-moment pounds normal per inch force intensity
with
regard
to x-
y-directions,
intensity in plane
of resultant I, pounds
acting
in
x-direction
Ny
acting
in
y-direction
Nxy
intensity of resultant shear y-directions in plane III, intensities sections tively, of transverse
and
Qx,Qy
resultant and
on
cross
xz-plane,
respec-
82w
T coupling duced shear elastic constant Nxy per y-, representing applied; unit of and twist _-_ the proper unit of strain 7xy in x-, also one-half Mxy , pound -I
U_V,W
z-directions,
respectively,
measured
parallel
to
corrugation
direction,
and
perpendicular
coordinate,
measured
perpendicular
to
faces,
inches
7x,7y
shear
strains
associated
with
Qx
and
%,
respectively
NiCl TN 2289 _xy _x,_y _x_by _'x_'y shear strain_ with respect to x- and y-directions_ plane III strains of plane I in x-direction y-direction_ respectively of
25
and of plane II in
Poisson's ratios associated with bending in x- and y-directions; respectively Poisson's ratios associated with extension in x- and y-dlrections; respectively Corrugated-Core Sandwich Symbols area per unit width of corrugation to yz-plane, inches cross section parallel
AI
area_ in width 2p, lying between corrugation center line and lower-skin center line (see fig. C4 of appendix C); square inches area_ in width 2p; lying between corrugation center line and upper-skin center llne (see fig. C4 of appendix C), square inches
A2
al_a 2 b13b2 ci_c2 dl_d2 el_e 2 fl;f2 gl_g2 Jl,J2 kl_k 2 RCI_RC 2 Ril_Ri 2 _I_2 _i,_2 b
dimensions of corrugation cross section consisting of straight lines and circular arcs (see fig. D5 of appendix D)
26
NACA
TN 2289
B3,B4,B6,B
nondimensional parameters in formula for S (equation (DI9)) for a symmetrical corrugated-core sandwich, defined by equations (D20)
parameters in formula for a corrugated-core (D18) load and for support lower and of
between
test
beam, faces,
inches
upper
EC
of
corrugated-core
sheet
material,
E' C
stretching sheet
modulus
of used
elasticity in
of
material,
derivation
EA x
extensional stiffness of corrugated-core in x-direction (bendi_ng in x-dlrectlon per inch (Elt I + EcA C + E2t2)
EAy
extensional stiffness of corrugated-core in y-directlon (restraining effect of ignored; bending in y-dlrection inch (Elt I + E2t2)
prevented),
EI x
per unit width, of a beam cut sandwich plate in x-direction, _itlk_x2 + Ec_(k_ - k_2 +
from inch-
bending stiffness, per unit width, of a beam cut from corrugated-core sandwich plate in y-directlon (restraining effect of corrugation ignored), inch-
GI,G2,G
moduli of elasticity of lower-face, upper-face, corrugated-core sheet materials, respectively, shear stiffness to inch xof and corrugated-core y-directions + G2t sandwich (twist
psi
GA
plate
with pounds
respect per
prevented),
2)
NACA 2289 TN GJ
27
torsional stiffness, per unit width, of a beam corrugated-core sandwich plate in x-direction,
pounds
itlkGJ-j2 +
o
AC
(kG-_ - kC)2
+ G2t2(l-
kG-_)
distance
between
middle
surfaces
of
face
sheets_
inches
depth of corrugation, measured line at crest to center line of appendix D), inches
vertically at trough
hEC
core
thickness D),
of
sandwich
plate
(see
fig.
D5
of
appendix moment of
inertia
centroidal
axis
parallel
moment of inertia, per unit width, of corrugation cross section parallel to yz-plane, taken about centroidal axis of corrugation cross section, inches 3
KAy,KAz
KI y ,K I z
integral C1,
in
equations of
for
B3,
functions
corru-
KL KLy,KL z
KLyz
gation cross-section geometry, defined by equations (D15) for general case and by equations (D21) and (D22) for corrugation having a cross-sectlonal center line consisting of straight lines and circular arcs
ky,kz
D)
klh,kllh,kll
between
planes I, II, and III, C1 of appendixes B and kch distance which shear
figs. BI inches
between middle surface of lower face and cuts corrugation into lobes of equal area center of corrugation), inches
plane (also
28
NACA
TN
2289
k_h
distance to
between axis
surface
of
lower
face
and parallel
centroidal yz-plane,
corrugation
cross
section
distance
between axis
surface
of
lower EIx,
face inches
and
centroidal k_Ec_
q
associated
with
+ E2t 2
distance
between axis
middle
surface
of
lower EL,_ J
face inches
and
centroidal
associated
with
kG-_h
distance shear
between plane"
middle associated
surface with
of
lower
face
and
"zero-
GJ 3 inches
kGj
L 2p P
between pitch to
supports (see
of
test
beam,
sketches pounds
in figs.
applied
test
beam,
about centroidal axis of cross-hatched cross section shown in figure EI_ inches 3 coefficient EC I/tcl 3 in formula for
D%,
D% = Sh(1 C2]\hcj _
coordinate measured along cross sections parallel figures C2, D3, and E3, center line to yz-plane; inches of corrugation see, for example,
29 thicknesses of lower-face, upper-face, and corrugatedcore sheets, respectively_ inches angle between face sheets and straight diagonal portion of corrugation leg (see sketches in figs. 3 and 4)
_i,_2,_C
Polsson's ratios for lower-face, upper-face, and corrugation materials, respectively angle between face sheets and tangent to corrugation center line (see fig. D3) Subscript
approx
approximate value
NACA 2289 TN
_x_ _y_
EX_
Ey_
_'y_
of the
Cxx,
formulas
with bending and stretching an element plate is considered which is subjected Mx and My and to horizontal
intensity Nx the
and Ny at arbitrary distances klh and klih , respectively, middle surfaces of the lower face. (See following fig. )
Plane'ff
Figure Equations are derived relating the forces and moments producing them; to are Dx_ Dy_ evident. _x, The _y_ Ex_ Ey, general formulas
BI
distortions of this element to the in these equations terms corresponding _'x, thus _'y, Cxx, Cxy' Cyx_ and Cyy obtained are reduced for special
applications. The the stiffnesses assumed sheets. moment and of Mx the and face force sheets; only by Nx the the are assumed of the My to core and be resisted the Ny of by are the face both
bending
extensional
stiffnesses
and
extensional
moment extensional
force
to be
resisted
stiffnesses
NACA
TN
2289
31
lines of
drawn the
between
points assumed
in to
the remain
upper perpenthe
undistorted
unchanged in length during of the element as a whole and _2w. _2 and y-directions; planes between in the it which upperis The middle
of
_2w _x 2
faces
will the
be
strained of linear by
in
the
x-
to may
imagine be
existence surfaces.
other
horizontal
obtained middle
interpolation
lower-face
as by
the the in
moment extensional
My
and
the
force of
Ny the
are face of
to
be
the and
surfaces
lowerand are
sheets respectively,
determinate
(B1)
qYl = tl h tl
_Y2
My. + NYkl - t2 h t2
(B2)
in
addition, in of the
the
in
the
x-direction were
lower-
fixed. that
determined the
from
conditions: Nx and
thrust the
intensity
x-direction I is
intensity
x-direction
plane
N x
cXltl
Cx2t2
_xcAC
(B3)
Mx
qXltlkih
- ax2t2(1
- kIlh
(B4)
_x 2
32
NACA
TN
2289
whe re
_Xl
direct
stress
in
the
x-direction
in
the
middle
surface
of
lower
face
qx2
direct
stress
in
the
x-direction
in
the
middle
surface
of
upper
face
_Xc
in corrugation of corrugation)
(also
direct
The
terms
qXl , in
_x2 , terms of
be
replaced
by
the
following
expressions
Ox I = Elex I + _l(_yI
= Elex I
Ox 2 = E2_x2
+ _2_Y2
(B6) = E2Ex2 + _2 _ + t2 /
x C
82___w _x2 = 8x 2
" _Xl h
(B8)
33
Exl[Eltl
+ EcAc(I-
k_ll
+ gx2(E2t2 + EcAck_I =
(B9)
h2 ]
Ex21E2t2(I-
kI)-
BII)
Mx
_x2 = -_xx2 _ + #2_xY2
My
T + t;xx2 Nx P'2"'xy2Ny
( B12 )
34
NACA TN
2289
where
k_xh2 _XX 1 = EI x
(m3)
(k_x_
kli) k_xh2
E-Yx +
kli
_Y2
_x ( _x-_)(_ _x)_
i
NACA TN 2289
35
and
_x
= ECI C +
_Itlk_x2
+ EcAc(k
_ - kE-Ix) 2
+ E2t 2 (i -
E1
the
strains the
in the strains
x-direction in the
and
the
stresses _Yl
known,
y-direction faces,
respectively,
(B15)
_Yl = El J Yl - _icxl
or
after
of
YI'
Y2' and
eXl , and
Ex2
by
means
of
equations
(Bll),
(B12),
=1-- - l XXl + Mx T
_lNY
- _l_x_iNx
(BI7)
_Y2
CYY2
My _
+ _2xx2
Mx + "'Yy2Ny T
_2"xx2Nx
(BI8)
36
NACA
TN
2289
where
_yyl
= (i-
+ _12_XYl
+ t.,u2 2_xy 2
(B19)
(%-_,)(, -%),q
_TY + _22_/2 and
EAy
= Elt I + E2t 2
E2t2
(_2o)
and
_XXl , in
_xx2 , equations
,XXl, (BI3).
*xx2,
_xYl,
_x72'
*xYl,
and
*xy 2
are
defined
NACA 2289 TN
37
With the strains EXl, _x2, _YI' and _Y2 known and the assumption madethat lines normal to the faces remain normal, the distortions of the element are completely defined. The curvatures can nowbe written as: 82w
8x 2 6x2 - 6Xl h
x_2 + _0 Mx + = _(
+ l xy:j_
(_XX 2
-1/XXl)"_-
(_2_xy
_l_XYl)_"_
(B21)
82w
_Y2
_Yl
_2
(B22)
The
strain
in the
x-direction
in the
plane
of
Nx
is
_ _ X _ _Xl _ kI _ X2
X_
(B23)
38
NACA TN
2289
and
the
strain
in the
y-direction
in
the
plane
of
Ny
is
Ey
= _yl
+ kii(_y 2
- _y_
of
equations
(B21)
to
(B24) of the
with
equations
(I')
to
(4'), for
permits constants:
identification
following
expressions
Dx
= El x
(B25)
Dy
= E-_yll
- C2 2<I
I
_x: _2 + (_I"
Dy _y = _x Yxx
EI x
EIy
(B26)
(B27) _2)k_x
(B28)
(B29)
NACA
TN
2289
39
Hi
H !
cJ
I I r-_ !
H + I
I
+ I
_
+ _L I + I--4 b--t
X
C,J I C_
-t I
3. !
,' I'%1_ I I
r-t
_i_
I ,--t I
-I II
-_.-_
H H
I
I
+ ,M
H
b-I
'}1,
_
0d
H H ! I
I--I H
_
!
H
OJ + N I I
r--4
"-I
+
II ,M
40
NACA
TN
2289
(k_
Cxx = _
x
m
- kI)h (B33)
EI x
- ii)h
Cxy = EI x + (_i_2)( I - kll) k_'_xhEl x
(B34)
(B35)
(_y
Cyy = -
( kE-Ix
klI)h]
(k_y
kII)k_ _y
1 E-Axh
(B36)
the face
case (that
in which _2
the
Poisson's
ratios
for
is,
expressions
are
elastic
constants
become
Dx : E-f x
(B25')
_x
= _i
(B27')
NACA 2289 TN
41
(B28')
-1
(B29')
Ey
(B30')
B'X
= _i i + (kE-ixkl) 2 _-_xh2Elx-
(B31')
(B32')
k_x
CXX = D
- ki) h El x
B33')
Cxy = EI x
B34')
Cy x
= -_iCxx
B35')
42
NACA 2289 TN
Cyy = -(i
- _12)(k_y
- kll)h Ely
- _lCxy
(B36')
It is evident from the preceding two sets of formulas that the values of the constants associated with stretching and also the values of the coupling constants depend upon the location of planes I and II in which the stretching forces Nx and Ny, respectively, are applied. Choosing
planes I and II at the centroids of the transformed, cross sections parallel to the yz- and xz-planes, respectively, Qthat is, letting k I = k_x the formulas and (B25') to (B36') and klI = k_-_y) results in reduces two of the coupling become further simplification of constants to zero. Equations
Dx -_E-f x
(B25")
=
Dy
l_12
-_
EI_]]
(B26")
_x
Ul
B27"
B28"
Ex
= EA x -1
B29"
E_
EA x
,fhT&-k
2
Elx ]J
B30"
_' X = _i
B31"
NACA TN 2289
43
Ey _'y = _'x Ex
B32")
Cxx=
B33")
_(_x-_)_
Cxy = Elx
B34")
Cyx
= 0
B35")
Cyy : -_ICxy
B36")
44
NACA
TN
2289
APPENDIX
DERIVATION
OF FORMULAS
FOR
Dxy ,
Gxy , AND
In ment Jected of
the
derivation
of
the
formulas
for
Dxy ,
T, is
an
ele-
plate qc
subof the
surfaces (See
lower-face, fig.)
upper-face,
and
sheet,
respectively.
following
Trr
Z ,W
average as plane
acting
arbitrarily
chosen
a twisting flows 71 ,
moment
of the face and core sheets. By linear interpolation (or extrapolation) between the middle surfaces of the face sheets, a shear strain for every horizontal plane can be defined. In this appendix equations are derived 82w relating the twist _y and the shear strain 7xy of plane III to
NACA 2289 TN
45
the resultant forces of intensities Mxy and Nxy which produce them. From these equations general formulas for Dxy, Gxy, and T are obtained. These general formulas are then reduced to special forms for particular applications. The orthogonal x- and y-axes are taken in the as yet undetermined plane of zero shear strain, as shown in the figure.
Assumptions.Vertical lines drawn between middle-surface points in the upper and lower faces before twist are assumed to remain perpendicular to the faces and unchanged in length during twist. The shape of the corrugation in planes parallel to the yz-plane is assumed to be rigidly maintained, whereas displacements in x-direction of the corrugation between lines In order to eliminate ment (x = 0, y = 0) vertical line at coincidence with tained only the the of attachment to the faces are freely permitted. rigid-body displacements, the corner of the eleis assumed to be fixed in space, and the originally corner z-axis. the is assumed to remain vertical, The distortion of the element shear flows ql and q2 that is, in is mainin the faces
through
constant
and qc in the corrugation) that is, the face are assumed to be so thin that twisting moments negligible.
_2 w Displacements.of the zontal face In terms of the twist _ and the height hI
xy-plane above the middle surface of the lower face, displacements of points in the middle surface of the uI and vI may be written as
the horilower
_2 w vI = The upper horizontal face u2 displacements and v2 are of -hlX points _-_ in the middle surface of the (C2)
_2 w
u2 : y
_2w v2 = h2x _
(c3)
(C4)
46
NACA
TN
2289
in the is
x-direction
of the
corrugation
middle-surface
82 w um and that of the trough line = h4p is un : 0 Vertical displacements are given by (C6) _-_ (C5)
nn'
82w
w = xy_-_-_ Shear the strains in the shear faces.strains In terms in the of the 71 foregoing and 72 (C7)
displacements_ can be
middle-surface as
faces
written
_i
_Vl
_2w
71 =
;_2
72 = _
+
;_2
+ x_
= -2hl
82w
: 2h2
(C8)
(09)
Shear strain in the corrugation.The shear strain in the corrugation can be determined by considering the portion between a crest and the adjacent trough as a beam which is being twisted about the x-axis
82w
at a constant rate to the _-_ x-axis with the shape of the corrugation (See the in planes following
perpendicular fig.)
rigidly
maintained.
NACATN 2289
47
Figure C2 The constant shear strain 7C in the corrugation must be such that continuity of displacements in the x-direction is maintained between the corrugation and the face sheets. With u and v' denoting axial and tangential displacements_ respectively, of the corrugation middle surface and s denoting the distance from nn' measured along the corrugation center line 3 the shear strain in the corrugation at any point P may be written as
7c
= _ss r _-_ (CIO)
where r is the perpendicular distance from the axis of twist Ox the tangent at point P and is considered positive if the tangent passes below point 0 (as in fig.) and negative if it passes above.
to
48
NACA
TN
2289
Integration gives
of
(CI0)
with
respect
to
between
points
and
Jo
where line, Z or is the length um
ds
= 7c
ds
+ _-_
J0'
r ds leg, measured along center
of
one
corrugation
- un =
Z7 C + _ 82w
_0 Z r ds
(ell)
The integral in equation (CII) represents twice the net area swept out by the radius vector 0 in going from n to m, or, as can be seen from the following figure, it equals twice area I minus twice area II. Area 1"1" 7
_-m_
\;_,a l .. z
Figure C3
With Area equation (CII) becomes 82 w um Continuity equation Therefore between (C13) 3 be core equal - un = faces um Z_' C + 2 2_-_ that by um - un as given and (C13) I - Area II=Z_A
(c12)
and to
requires as
by (C6).
- un
given
equations
(C5)
_2w
NACA TN
2289
49
or
7C
(h4P
i - 2 ZkA)T
_2w
PJ c
where width. _ AC equals Zt C P , the corrugation cross-sectlonal area per
(Cl4)
unit
area (C12) of
ZkA
which
appears
in
(C14)
and
by of
depends and
vertical be related
twist
corrugation.
Through other
geometrical
considerations,
areas, one of which ph 3 depends only on the vertical location of the axis of twist and the other of which depends only on the geometry of the cross section. The relationship is
ZkA =
_ h3P
- _
1 - A2
- P(tl
- t2
(C15)
where AI is the area, in width 2p, lying between the corrugation center line and the lower-skin center line, and A2 is similarly the area lying between the corrugation center line and the upper-skin center line. (See the following fig.)
t2
L__
CI
'qY_f_
2p Figure C4 through equation (C15), -
ZkA
in
equation 7C
(C14)
eliminated
for
becomes
7C =
lh2 - h I
A21tc
_2w
(C16)
50
NACA 2289 TN
Shear flows.With the shear strains (C9), and (C16), the following expressions flows:
(C8), shear
Al
" A2-)tC2
_2w
(cz9)
_2 w
These expressions give/the shear\ flows in terms of the twist _x--_ and the vertical location _hl, h_ of the plane of zero shear strain. In order to determine the elastic constants, the shear flows must be _2 w expressed in terms of the twist _-_ and the shear strain 7xy of
plane III. The shear strain of any with the distance from the xy-plane twist; hence,
or
h I = kiilh
- _
_2w
(020)
and
h2
= h
- hI
(i - kii_h
i 7x__ +- 2 _ _2 w
(c2l)
NACA
TN
2289
51
(C20) and (C21) to eliminate hI and h2 from equa(C19) gives the following expressions for the shear
ql
-2Gltl(kllIh
_-_ __2w
- 1 7_
(022)
q2
2G2t2
kii
h _
+ _ 7
(C23)
qC = -_----' Gctc2{_
- 2kII_h I)
+ AI
2pA2-__
+ 7 xy 1
(C24)
of
the
shear
flows_
namely
Nxy
and
Mxy _ may
now
be
of
.- The
shear
flows of
x7
horizontal shear flow
ql'
q2'
and
qc
combine
to
Nxy
= ql
+ q2
+ _
(C25)
where
ql'
q2'
and
qc
are
given
by
equations
(C22)
to
(C24).
Evaluation by taking
of
Mx7.in the
The
average 3 of
value qi,
of q2_
Mxy and
can the
be
determined
moments, of qc
yz-plane to
horizontal in this
components
with
respect
plane
III.
Use
is made
section of a horizontal plane which cuts the corrugation into lobes of equal area. This plane_ which is shown as the following sketch at a distance kch above the middle lower face, flows. is the centroid (or shear center) of the
corrugation
N J_.CA
Figure
C5
52
NACA TN
2289
Taking
moments
with
respect
to plane
Ill
gives (C26)
where
ql'
q2'
and
qc
are
given
by
equations
(C22)
to
(C24).
through
the
purely
geometrical
relationship
=
gives
(c27)
--
_%
_(_- _i)_
(c28)
Mxy
: 7xyGA(k_- _
GJ
= tlk_ 2 + Gctc 2/ )_
(C32)
NACA TN 2289
53
Solution
of
equations
(C28)
and
(C29)
gives
(kG7 - klli)
(C33) 2GJ
kln)h
GJ Comparison of equations (C33) permits the identification of and the
(c34) x + _V _]
(6')
Dxy
= 2GJ
(C35)
kiii) h
2GJ Choosing ki11 equal to kG_ reduces the foregoing equations to
(C37)
Dxy
= 2GJ
(c35')
(C36')
Gxy T=0
= GA
(c37')
54
NACA
TN
2289
APPENDIX
DERIVATION
OF
FORMULA
FOR
D%
In this
appendix
a formula
for
the
transverse
shear
stiffness
DQy
is derived which is fundamentally the same as that given in reference 4 for the case of interference of flats neglected but extended slightly to include the effects of stretching of the corrugation and the prevention of anticlastic curvature. The general formula is reduced to special forms for specific applications. The element has unit of a corrugated-core width normal to the sandwich shown in the following page and is in equilibrium under
figure
a Y
small transverse shear of unit intensity (Qy = l) and horizontal of magnitude p/h. The corrugation is assumed to be fastened to skins through rigid joints at its crests and troughs.
forces the
r - t2
distortions
to Qy. following
distortions
and
Figure
D2
NACA 2289 TN
55
yy may be taken as
then given by the
_Y h
- mSz p of
and
the
transverse to
ratio
shear
intensity
DQy
The minate tion tion in of sandwich plate to element is the a
_
h now
1
p analyzed as 6y a statically and for of of iz" the the a
(D1)
indeterSubstitucalculaunit-width
determine then
displacements general
(DI) any
gives
expression analysis
particular is made
In the element
element
assumption
is part
sandwich
page equal to infinity. taken as beams in which which factors amounts of the
completely
restrained, by
flexural
stiffnesses
more consistent with experiments in which used_ the Poisson's ratios _ may be set figure skins. are shown free-body are diagrams
following
These
elements
represented
z2pZ/I
I_2Z2
k.__
---5 (_ z, xl
[ _('c+',) z,
NACt,
Figure
D3
56
NACA 2289 TN
The distortions of the elements, assumedsmall, are shown in the following figure:
Figure It should considered be noted that the as acting in the forces Y midplanes
D4 on the corrugation elements of the skins and transmitted the and are to are
the corrugation through short rigid projections. Similarly, moments MI and M2 are taken about points in these planes not the actual moments in the corrugation sheets at the
Joints.
all are
the undeformed structure is symmetrical about any plane BE, and deformations in the two corrugation elements EA and EC as likewise are those in the two skin elements ED and EF two are or skin elements zero, and each Z2 and its BA and BC. skin element Zlp or Then is in Z2p the skin moments at B, equilibrium under its at one end.
moment
Since a shear of unity is assumed relation between the shear carried by carried by the two skins ZI and Z2
to the is
sandwich, X and
- Z I - Z2
= i requires that
(D2)
corrugation
elements
M2 - MI +_
-Xp
= 0
(n3)
NACA
TN
2289
57
Equilibrium
of moments
at
Joint
requires
that
M I + ZIp
or
+ M I + Zip
= 0
ML = -Zlp
Similarlyj at an upper joint M2 = Z2P
(D4)
(D5)
Finally, is given
the by
internal
moment
at
any
point
in
the
corrugation
sheet
M = _2
(t2 + tc
* kzhc)+
z]
static to the may be
- X -_-
+ y)
(D6)
The
five
equations of the
are
all
the
relations
needed.
at lower
rotation as , skins,
respectively,
(D7)
Z2p3
(D8)
in the corrugation sheet are due to both The three components of the displacement to the tangent at E are
bending at A
and or C
(D9) E C IC/(i - _C
98
NACA 2289 TN
ECI C/(1
_C 2)
(t2
+ tc
+ kzhC)
ds
1 E'ctc/(I
_(Y - _C 2)
cos
+ X
sin
9)cos
I/ ds
(DiO)
1 E'ctc/(1 -
2)
* ds
(Dii)
where the integrals are taken over one corrugation leg, as from A to E or C to E (excluding the short rigid projections), s is the distance measured along the corrugation center line, and @ is the angle between the tangent to the corrugation and the horizontal (see fig. D3). In equations (DlO) and (Dll) E' C denotes the stretching modulus of elasticity of the core. It has been distinguished from the bending modulus EC in order to permit identification of the terms representing the stretching contribution in the derivation. The Poisson's ratios associated with bending and with stretching of the core have, however, been assumed per through equal. unit the width In the I1, rest I2, of and the IC derivation, will the moments cases be of elimiinertia nated in most
relations 1 = _
Ii
t13
i 12 = _-_ t23
(DI2)
tc 3
(DII) can
contain be
ten reduced
unknowns to the
for
which
they four
equations
first
following
+ +
al2Z 2
+ a13
5z
+ 0
a22Z 2 + a23
5z + 0
5 z + 0 = r 3I 5y = r
+ 0 + a44
59
a21= _
_c%
z+ 2 e
i
12
(to12
c\%/ .
+
a41=
Ec_
a12
= -i
3 _2_
p h C KL
_c2kVd
=_ a32 C z
+ky 2
a42
i
a = m
23 =a a33
p =r 44 i
J
=i
Equation
(DI4)
continued
on
next
page
6o
NACA 2289 TN
r2 : _
l 311- C2) l tc t2
ECIc + _\_ + _ + k
-_-hc
r3 : h
tlc
%z + 2\_ + _ + _
K
_Az
(DI4)
Concluded
2 hc KAy 12 E, c
.2_.
r4
= -h-
Iy
+ _
+ k
KAY
_\_ + _ + kI l/t2 tc
The quantities K_,
KL_
+ 12E, (_)2 1 EC c
and
K__
Klz ,
so on
are
nondimensional
functions
of the corrugation shape and the origin the following integrals taken along one llne from the crest to the trough_ that figure D3:
location. They are defined by leg of the corrugation center is, from m to n in
if
Kiy z
=l 3 S yz hc
13 J - hc
ds
(DI5) z2 ds Kly
KAz
_l7
2
y ds
KAY = _2_z
as
Equation
NACATN 2289
61
K_ = _
KLy z : _
if oos2,ds
(mS)
KLz
: _
1j lj
solved the
Concluded 9 cos 9 ds
sin
sin2@
ds
(DI3) then
may
be
for
8y
and
8z"
in
furnishes
following
expression
D% -- sh _ _c2/khc/
where 3- F
S ____
(D16)
_Cc7(022cic3) c3 _
_(2c2- _Cl)
12,
(DI7)
62 and
NACATN 2289
l Eci1- 221
CI=K L + -_ E2(I I._C 2)
2_
he
C2
: KAz
+ h'P-_KL 2
+ 12
ETC\_/
KLz
C 4 : Kiy z + _ IEZ
+ ( 1 + t___)_(
KAz
+ ky 2
_ p
KL)
+ ky 2
_ p
KAy
(DI8)
C5 = KAy + _
Itcl
E1 1 _c2(tll 3
c7 - _'c 1
_12_cj
Special
Cases
corrugation.so on in the
of
the
terms upon of ky
the the
coordinatesj encountered
frequently
= k z _ l, leg.
at the and
midpoint KAz
corrugation
a result
parameters
vanish.
NACA 2289 TN
63
S_mmetrical
sandwich.-
For
the
case
of
the
symmetrical
corrugated-
core sandwich, elimination of KAy and KAz by choosing ky = kz = 1 is again advantageous. In addition, however, the numerator and denominator in equation (D17) for S contain a common factor. Cancellation of this factor yields the following simplified expression for S to replace equation (D17):
(D19)
where
B 3 = KIz
B 4 : Kly z
(D20)
B6
+l
= KIy
/tc 2
KLy
12 E'C \hc/
E1
1 _ kC 21t I h3
B'-r= c7 = ]_c 1
Corrugation center line consisting arcs.The center line of a corrugation three straight-line segments (two flats separated from each other by two
_12\Cc,/
of straight lines and circular leg in many cases consists of and one diagonal element) arcs. The following figure
circular
64
NACA 2289 TN
.........
_y
' r'
I
Rc,2 Ri2_
,
_hC I hG
0 -__',\
e _y
', ', rt I
//
Rcl--"
/ L__
Ri I
.
NACA
Figure
D_
NACA 2289 TN
65
shows such a corrugation leg with its dimensions. If the integrals in equations (DI5) are evaluated for this shape, with the origin of x and y chosen along the straight diagonal portion as shown in the figure, the results are
:_i 2_:i_!2_]
i :_Rc2) T
hc
KIyz
= _I jl kl dl 3\h C hc hc
i
e_+_ + 2
(o21 )
*_L_
#*_
* _-_
e_+2
_"
g:-_]]
_+
_I_ ( _
1Q_c2/_
g2 Equation e_l _ continued on next page (D21)
66
NACA 2289 TN
RCl(eb el he _+hc
e2 _+hc
KAy -- c h 2\h c
hc
- -_j_ - kz
e_+
(D21)
Concluded dl d2 RCI
ELy = 1
cos2O
+ _
(e + sin
O cos
e)
ELy z =
sin
@ cos
@ + 2\hc
sin2e
KLz
sin28
+ 2\hc
If, in addition to consisting of straight lines corrugation is symmetrical (that is, RC 2 = RC1, the leg origin (that of is_ y ky and z is chosen then at the
and b2
midpoint (D21)
= k z = 1),
equations
become
NACA 2289 TN
67
KTz=-_\_/ _ + 3_t_)
2 kll2dl 2 Jl kl dl _E_'_2
he
K_yz: 3 he _ _ + 212t,_1
2_ L_(e_-+heKIy: 7t,_l _ + C_ +
+_t_
2#
2(Jl/2dl 1 fl
(D22)
__t_
KAz = KAy = 0
_-+ __- +_
+_
dl
RCI
fl
K, : 2_ + 20_---+_
t':l_ dz _:l. KLy = _ + 2 _ cos2e + ._----(e+ sin e cos e)
dI KLy z = 2 _ sin a cos RC 1 a + _-sin2e
KLz
dl = 2 _
sin2e
RCl + _--(@
sin
8 cos
e)
The equations
that (D22)
have can
to be be
inserted
in the a few
right-hand basic
sides
of
obtained
from
dimensions
68
NACA 2289 TN
fl_
f2,
and
tc)
through
the
following
sequence
hC
= hEc
- tc tC
tc RCl = Ril + -_
Rc2 = Ri2 + Y
a2
= kz
hC 2
- RC 2
b2
=l
RC 2 c2
(D23)
dl
= (Cl 2
- RC12)I/2
d2
= (c22
- RC22)I/2
e = cu 1
+ 81
= cu + 92 2
eI
= RCI
cos sin
e e
e 2 = RC2
cos
gl
= IRC1
g2
= RC 2
sin
Jl
= al
+ el
J2 = a2
+ e2
kl
= bl
- gl
k2
= b2
- g2
NACA 2289 TN These dimensions required in equations (D21) and (D22) can also be obtained from a different set of basic dimensions (p_ hEC Ril , , 83 and tc) through the following sequenceof computations: hC = hEC - t C tC RCz = Ril + T tC +7 RC2 = _i2 he
a 2 = k z _- RC2
69
Ri 2,
e I = RCI
cos
e 2 = RC2
cos
gl
= RCI
sin
g2
= RC 2
sin
8 (D24)
Jl
= al
+ el
J2
= a2
+ e2
kl
Jl
cot
k2
J2 cot
dl
Jl csc
d2
J2 csc
bl
= kl
+ gl
b2
= k2
+ g2
,2:
b2)
70
NACA
TN
2289
APPENDIX
DERIVATION
OF
FORMULA
FOR
DQx
In the
derivation
of
the
transverse
shear
ele2p
axis
of
faces
combination
--
y////////,
"////
!\ //.....
Figure
E1
The transverse shear is equilibrated by a change in bending moment dM from one end of the element to the other. From the equation relating the distortions of this element to the shear V 3 a general formula for DQx is obtained. by assuming A more that the practicable core carries approximate no direct formula stress. is then
obtained
General derivation.The direct stresses produced in the element by the bending moment dM are assumed to vary linearly through the thickness. Assuming the only flexibility to be that of the corrugation in shear gives the following picture of the relative distortions of the element:
i
i
8x L- .......
dx E2
Figure
NACA
TN
2289
The
angle
5x -_and
is taken
as
an
average
shear
strain DQx
Yx
for
the
cross by
section,
the
transverse
shear
stiffness
is then
given
DR : Qx =
An expression is now stitution in equation Elementary a point such derived (El). for 5x as a linear function of V
(m)
for sub-
at
the as
shear
stress
in
the
corrugation
VQ = 21tc
(E2)
area
about of
neutral 2p
axis, about
inches 3 centroidal
section
width
(If faces and core are not all of the section should be used in calculating the corrugation sheet is
TC T C = GC
VQ
2Gc Itc (E3)
TC
along
one
corrugation-leg
center
line
(see
the
Figure
E3
5x = 0 Z -_x ds
- 2Gcitc
Vfo
in GcItch Z P_0
Q ds
(E4)
of DQx_
expression
(E4)
(El)
gives
the
following
general
DQx
(ES) Q ds
small to
If, as bending by
is usual, the corrugation carries only a moment M, then an accurate approximation assuming therefore, formula of that that for the the the DQx entire bending moment carries the same in h as calcuis, is no
DQx
and,
corrugation will be
stress. (ES)
but Q
with ds,
effect the
corrugation location
omitted
lating
I,
_0 Z
centroidal-axis
) that
I _ 2ptlIk_Yxh
2 ) 2 + _i 2p)t2(l
(E6)
_l
Q ds _
E___(2p)t2(l I
- k_-_x lh _
(E7a)
or
/0
Q ds
2pt I k_x
hZ
(ETb)
"
I0
NACA
TN
2289
73
E2 _ii t2 E2 tl + _II t2
(ES)
of
the
expressions
(E6), to
(E7), DQx:
and
(ES)
in
gives
approximation
Gctc2(h] _ "\P/
(E9)
74
REFERENCES
NACA 2289 TN
Jour.
Aero.
Sci.,
2. Libove_ Charles_ and Batdorf_ S. B.: A General Small-Deflection Theory for Flat Sandwich Plates. NACA Rep. 899, 1948. 3. Stein, Manuel_ and Sandwich Plates. Mayers, J.: A NACA TN 2017_ Small-Deflection 1950. Theory for Curved
Foster_ H.: Chordwise Shear Theory_ DoubleRep. No. 2326, Lockheed Aircraft Corp._ Sept.
4,
Kennedy_ W. B._ Jr._ and Troxell, W. W.: Study of Compression Panel_ Supported on Four Edges_ Formed of Corrugated Sheet with Flat Skin on Both Sides. NACA ARR 5B03, 1945. Strength of Materials. Part I - Elementary Theory Second ed. 3 D. Van Nostrand Co. 3 Inc., 1940,
NACA
TN
2289
75
TABLE
I,-
RESULTS
OF
_J_RICAL
SURVEY
OF
ACCURACY
OF
APPROX]S_ATE
FOI_MUIAS
FOR
Dy,
Ey_
Dxy,
AND
DQx.
(kll
k_y)
[El
=E2
=Ec]
GI
=G2
=GC;
_i
:_2:_;
fl
f2;
and
RCI
RC2
O.18hC_
Zp
"I_,
RcJ
k tI
(dSeg)
60
90
60
90
60
90
6o
9o
t2
o.811.4 o.8 1.4 0.8 1.4 o.8 1.4 o.8 _.4 o.811.4 o.8
o.99 .98 .97 _.99 .97 .97 o.99 .97 .96 0.99 -97 .96 o.98 -95 .94 o.98 .95 .95 o.97 .94 .94 o.98 .95 .94 Z.Ol 1.ol 1.Ol 1.oo .97 .96 1.ol .97 .96 1.oo'2 .9_*:i .93 1 1 i 1
1.4
0.8
1.4
(_)
1 1 1
(a)
1 1 1
(_)
1 1 ]
1 i 1
1 l 1
1 i 1
30 40 1.oo 1.25
I i 1
.30
10 1.00 1.25
-99
.97 .97
.99
.97 -97
.99
.96 .96
.99
.97 .96
.98
-95 .94
.98
.95 .95
.97
.94 .93
.98 1.Ol
-95 .94 1.01 1.01
Loo
.97 .96
1.ol
.96 .95
.99_1
.92
oo
.941.O0_ 1.00_
.8o
25
.97 .9o
.88_1.oo
40
30 mOO 1.25
10
.30
.5o
29
1.00 L25
.98! .941
aApproximate
and
exact
values
are
identical
for
symmetrical
sandwich
76
NACA
TN
2289
(a)
Symmetrical.
NACA 2289 TN
77
dy
_:
z,w dx/
___Qx
Mxydy MxdY
Middle surface
Nxydy dy
(a)
Symmetrical
loading. dy _-/ /
Z ,W
Mxdy Mxydy Y
dx
/
(b) Figure 2.Forces and moments General loading. infinitesimal sandwich-plate
acting on element.
78
20.0
I 1 __ l 1 I 1 I
NACA TN 2289
-j_ I !
15.0 10.0
60
h(;
.5.0
8.0 6.0 e o 1'0 _ -
1 7.5 700
60
4.0 3,0
\ \\,,\
S
2.0 1.5
"\',,.x
- \\._\_
1.0 .8 .6 RC_
.4
20.0 15.0
' k
Ioo /
1
6o
_-
_/oO\ k
S
2.0 1.5 Z 1.0 .8 .6
NAgA ::....
I 1 I I l I I I l
.4 .2
,4
.6
,8
1,0
1,2
1.4
.4
.6
.8
I 1.0
: 12
1.4
__P
hc hc
(a)
tc - o.3o.
tl
Figure
3.- Charts
for evaluating
coefficient
in formula
for
DQy section
for homogeneous symmetrical sandwich with corrugatlon cross composed of straight lines and circular arcs (E 2 = EC = El; P2 = PC = Pl; t2 = tl; f2 = fl; EC 2 = RC 1 = O'18hc)"
_CA
TN
2289
79
20.0 15.0
--
-I I | I ] O I I -
60
- -_.=2o.o
I0.0 8.0 6.0
60
_oO \\\
80
__
4.0 3.0
o,oo\ ,_\ \ \
\'\\\\\,\
- ,o\
\,, \\\\\
s
2.0 1.5
\,\\_\\\
\ \',\\\_ :
\ \\\\\\'
\ \\\\\?
1.0 .8 .6
-.@
I I I I I I I 1 1 I I I I I I I I --I I I I I I
.4
60o
70.0
6o
_oo\ \\
700 _ -
4.0 _LO
,oo__o\_ , \\\
\ \\\,\_ \\,\\,\
,o \\\
5
2.0 1,5
, '\ \\\\& \_
\'\\,\\_
1.0 .8 .6
",___
t *8 [.0 I |._ I |4 o_ I .4
.4 .2
4
I .6
I 6
I .8
l l,,O
I 1.2
I 1.4
p
hc hc
(a)
Concluded.
Figure
3.-
Continued.
80
NACA
TN
2289
20,0
15.0
--
hG
: 5.0
I0.0 8.0 1 60 o
7.5
soo
70
'
80
,o \\\\ o
,.0-e.9o ,. \ 3.0 \\_\
s
2.0
_,,
\\_\
'.0 _-_l
.6 --
i hc
\\_ t, r'//
Rc_
" ,, .....
,2
_-
_-..:
_"_
s.( Ec If'el'
J 1 i I I J I J i J t _
.4
20.0 15.0
__ _
_
l I [ I
Io.o
eo
150
60
,o.0
:: ,oO \
4.0
_o \ o
.oO\\ \
\
\\\_\\\'\\\"
\'\\
Z
s
2.0 1.5
\\\;
,,4 .2
I .4
i .6
i .8
i 1.0
I 1.2
i 1.4 .2
I .4
J .8
J .8
I 1.0
J 1.2
I 1.4
P hG
P ho
(b)
tc tI
0.40.
Figure
3.-
Continued.
11
NACA
TN
2289
81
20.0 15.0
J ] I I I ] 0 0 I !
600
30.0
4.0 5.0--
,o\\
800 \ 800 \ _
\ '_\\_\
S
2.0 1.5
\ \\\_\\_
\ \
1.0 .8 .6
I 1 I I I I I I I I I
.4
-I l I I I I
"2
- / .oo.\\
soo \_
4.0 3.0
__1
S
2.0 1.5
\\'\',
\,\
\\\
\_\'\
,o;\_\, \\'\\\
800
_ ,oo\_
\ ?, \ \"\
\ \\
hO .8 .6
"__
N_C_^a
I I I I I I I I I I I I 1 t
.4 .2
.6
.8
1.0
1.2
1.4
.2
.4
.6
.8 P hc
I.O
L2
L4
__p
hc
(b)
Concluded.
Figure
3.-
Continued.
82
NACA
TN
2289
20.0 15.5
_ _ h
_c =5.0
I0.0
7.5
60
80 e.o 70\\
soo -
70 \ eo \\ \
3.0
oo .
fl
__
I rlC'
_ .8
.4 I
__-
o_-s_lhi(' /_
' I I y-,_2cJ_hc] J 1 I I l I I I I I I I
20.0 15.0
,
I I t I ]
I
60
I0.0
6oo
15.0
,
6.0 800 k
_oO\
,o\\
s
e \
\ \
..o.00 \, 2,
"'
.4 ,2 I .4 l .6 I .8
\\\\)
[
I 1.0 J 1.2
, ? NkCk-_,_" ,
I 1.4 .2 I .4 I .6 i I .8 i I i 1.2 1.0
-_
1.4
P he
p
hc
(c)
tc - 0.50.
tl
Figure
3.-
Continued.
NACA
TN
2289
83
20.0
I I
--
60
70 \
30.0
so
4.0 3.0
.,oo \\\
_oO\\ \ \\\-
70o
\ \ , \\,
$
2.0 1.5
\'\\,\
1.0 .8 .6
.4
20.0 15.0
--
40.0 [
600
7 0.0
so o
70
4.0 5.0
--
s
2.0 1.5
.oo\\, \
\'\\\'\'g
I I I I I I
80\\\ 70 \
,oO \\\
\\\
\
1.0 .8 .6
"<'-,b
I 1.4 .2 .4 I 6
I I I I I I
.'_ .2
.6
.8
1.0
1.2
.8
1.0
1.2
i.4
he
he
(c)
Concluded.
Figmre
3.-
Continued.
84
NACA
TN
2289
_0,0
----
1 5.0
ho
_o.o
_' =5"0
i 7.5
60
,o
8.0
,oO
600
- - oo,,,\,\ \\ _ \, \ \,\ \\
,oO \\\
70
hc-
\\_'i fl
_l'/"
"
tl
DQY = ,4
_-F_}_hcJ
20.0 15,0
I
I0.0
60
I 5.0
60
4.0 3.0
,oO\ \\\
,o \\
_,o \ _ o
,
\ \-
,0o\\,\ \\_ \ _
\ \, \\\, \\-
.o , \
\ \, \ \'\
S
2.0 1.5
\,_\\\\
\'-
\\'\\,\\\
\\\\\\,=
\,\\,\\_
1.0 .8 .6 "_ .4 .2
l I I J I I I I I I I
NACA
I
,f"
I I
_ _
.6
.8
1.0
1.2
1.4
.2
.4
.6
.8
1.0
1.2
1.4
_.P
ho
hc
(d)
tc tl
0.60.
Figure
3.-
Continued.
m_cA TN 2289
85
20.0 15.0
I J 1 I I I
_-c = 20.0
so
30.0
60
80<
$
2.0 1.5
\\
: \'\\ \'\N
\\_',..\ --,..
L I I I i L I I I L i r
1.0 .8 .6
.4
E
40.0
600
70.0
60 ....
o
6.0 4.0 --
900
,oo , \,,,,
_:[ ,.o
_
.4 .2 i .4 i .6 I .8 i
\\\<\,.< .<@
\.\'-.'x
-,_
I t.0 1.2 _ 1.4 .2
I I
I .6
1 .8
I |.0
i 1.2 1.4
.4
P_
hG
hc (d) Concluded.
Figure 3.Continued.
86
20.0 15.0 I0.0 8.0 600 6.0 hc _ ,
I I I I I I I
NACA 2289 TN
1 I Z
-_c=5,o
600----
7_ .o \ _\ N", N
N\ \ -
r_ 2p f-ic
_ \" \\ >,,,?
._
o
.6
'9 _/
_DQ = Sh/
TI -_
, ;'
/ ....
ECI{tC
.4
20.0 15.0
J I I I 1 I -
I,
--
eo
15.0 __ 60 _
4.0 3.0
,oo,
.\i
-.<.-,.-_
oo \\\
S
2.0 1.5
L-
1.0 .8 .6 .4 .2
I 4 I .6 L ,8 t
"\\_\
I I.O I,_
i 1.4 2 .4
t .6
I .0
1 1.0
I 1.2
t 1.4
P he
p
hc
(e)
tc t I
0.80.
Figure
3.-
Continued.
NACA
TN
2289
87
,
60 o
, _ 30.0
60 o
700
6.0
80 o
__
s '
\\
,oo,_\\' \\\,\ ;
80
- \\',,',C\\
\\ \,\\
"-._,.,.__
.4
20.0 15.0
--
40.0
I0.0 8.0 600
70.0
600
.o:.oo
6,0 800
700
- =:d\\
,.o-1,0
\\ \\
\ \\
.,,
.4
.2
.4
.6
.8
1.0
1.2
1.4
.2
.4
.6
.8
1.0
1.2
1.4
__p
hc hc
88
NACA
TN
2289
, h_ 5.0
--
--
7.5 i 60 o-
60 6.0
4.0 3.0
.oo\\,_ _
_..,0o\ \'\\ _ _,c_-__/ ,,
70
\'\ \\\_
_ _
,oo .K\\_
i .
70 \ L
$
2.0 1.5
'\_
\\.__-
1,0 .8 .6
.4
20.0 15.0
I 1
I 0.0
Io,o
I 600
15.0 60
8.0
4.0 -e.o 900
' \ L
_-o \ o\
__ 90
\ \ \ '\ \ \ ',
?o 0oO \ _ \\\
\
s_.o_ ,,5,.o
.8 .6 .4 .2 I .4
I 6 I .6
\,, \? \N
\ \\x<\\,x<
:
\," _.___ _
t 1.0 I 1.2 I
I 1.4 .2
.6
.8
ID
1.2
1.4
_p
hc ho
(f)
tc tI
1.00.
Figure
3.-
Continued.
12
NACA
T_ 2289
89
--
I I I
30.0
--60
4.0 3,.0 - 8- 90
800
-70
_k,
S
2.0 1.5
\\\
\\
,\\
\\\\
1.0 .8 .6
I I I i I ]
.4
20.0
] I I I
-t 1 i I I I I -
I
40.0
60
80 o
800 7
----
800
70 _
S
2.0 1.5
\'
,
\\\\
\\\\_
\'\\\\,\'\
,\\\\
\'\\\\
\_
1.0 .8 .6 .4 .2
1 I I I I
-.._
] I I I J I I I I
.4
.6
.8
1.0
1.2
1.4
.2
.4
.6
.8
1.0
1.2
1.4
_E
h
_.p h
(f)
Concluded.
Figure
3.-
Continued.
_0
20.0
t I '1 I I t Z I I I I
NACATN 2289
I I
_-5.o eoo
70o __X
hc_
7.5
600
4.0 3.0
_,oo \\\ f
\ ;r ,\\\\ \
e-9o \
80_
S
2.0 1.5
\ \'\\ \"
: _-\\.\-\_ -..._
sh(Eal('_ t3
DQ_= _ _U\hc/
I I 1 ] I
\\ x.\ ._:
1.0 .8 .6 .4
20.0 15.0
,5.0
60
60
700
4.0 3.0
,,\\, ,,\\ ,,
90
.oo\\\\
k
S
2.0 1.5
\ \',,\
\',,,\ \',,
- \_-,
T , NACA I I I I
\\\ \'x\\:
\ .\i
\_
.4 .2 4
i .6
I .8
I 1.0
i 1.2
l 1.4
.6
.8 P hc
1.0
hc
(g)
tc tl
1.25.
Figure
3.-
Continued.
NACATN 2289
20.0 15.0
I I I I ! I I I I I I
91
30.0
600
4.0 3.0
_ _.,oo_\ \\
80 \ 70
S
2.0 1.5
1.0 .8 .6
I I I l
\\
._"
_'\\
\\
-.._
I I I I 1 I I I-I I I I I I I I I I --
,4
40.0
70.0
60
6.0
,o- ,oO_\
3.0 \
8oo
_
60 -
70
S
2.0
,._,o
\\\\
_
\'\_\\\\
-\-,<._
l .41. I I I I I I I .4 .6 I .e t I I |.0 t 1.2 1.4
.4
.6
.8
1.0
1.2
1.4
.2
_.p
hc
p_
hc
(g)
Concluded.
Figure
3.-
Concluded.
92
NACA
_fN 2289
20.0
_ -
i h(;
I i
s.o
-
_ =Io
tG
60o
_.
30--
600
.40
,o.o
6.0
_=o.4o - _o
{
,%
To_
,o _-' ,
S
,
,'_\
k
,_
90
3.0 ---
\\
- !
__
,A
'
1/'
,6
DQ
c;
= Sh{ Ec %{to%3
.i z
,I 8 --
.4
20.0 15.0 I r i i i i J I ! i
[
30
,
..... _
T I
--
I0 1.00
i
i .L 600
i.oo
____ _
.....
-T_60
-----[
O ......
.... ;
-90
_oo_ L-
,oo _k
'\
o o
,_
4.0 3,0
,,,\ \'%, _,
S
2.0 1.5
_,Z.,_
"_
I _
1.0 .8 .6
.4 .2
.4
.6
,8
1.0
1.2
1.4
.4
,6
,8
1.0
1.2
1,4
__p
hc
_.p
hc
_i
= RC2
on S of four of figure 3.
NACA
TN
2289
93
' I'
F'
I'
I'
=,oi i,oo
:_.o,oo i ,oO ....
O= 900
4.0 3.0
s
2.0 1.5
F I-
1.00_ 4.30
.4
20.0 15.0
i
I0 I 1.00
--, 1 T--
--
30 1,00
......... I
ft
,
j
.oo.
700
!\-
4.0 3.0
s
2.0 1.5
\\
,6
_ "_. a\ \)2'_
1.0 .8 .6
-,w_, ._-_
.8 P
%,L
1,0
1.00_
-,
.2 4
,_ ,_:_o/ ,-_-,
.6 .8
,4
_p
hc
94
NACA
TN
2289
20.0
]5: i
15.0 hG
i,
r
i
1
......
I--
_-E='O
I0.0 8.0 6.0
_
|
F
70
6o
I\\\ -
r-
,=
4.0 3.0
F,.,oo
. . --
k \k-
S
2.0 1.5
1.0 .8 .6
-h I'--_-
,c .... t2 Re"
--,,_
,4
20.0 15.0
'l'
1.00
_
I
Tt'I'
i
1.00
.....
4.0 3.0
,oo
8o
,_
_........
S
2.0 1.5
1.0 .8 .6
.4
I,
.2 4
:?o : ,oo_
j
I .4 i
4
l _i .... ] .S .8 p hc i
"__.'-_\ \_,
I I_) I I ,,.I ,.2 4 L4
.6
.8
1.0
1.2
1.4
.2
__p
hc
(o)
Figure
_ / _.
4.Continued.
NACA TN 2289
20.0
]
95
t5.0
e o
_r .i0, 17!_ ,!i0_! i r !
,,.o
%
o;'s:_l_)(_) _ ,:_o_
%.
I
I I I L i
-%,
J
.4
20,0 t5.O
_ -
q I0 1.00
T .......
I0.0
--60
6O _.o
8.0 _ -
eo
7o`> .,'X
kX
800
o 7O
,o _oo_ _
__
--
\
.50 1.00I J .4
:::! ,o
.4 2 h .4 I .6 t .8
-a \.\. %
I 1.0 i 1.2 _ 1.4 .2
__p
hc
(d)
t 2
t 1.
Figure
4.-
Concluded.
96
NACA
TN
2289
/_
- Rivets
,t
54 60
.........
b t.92
(a)
Dimensions
of beam
test
specimen.
Dial indicators
Loadings test
and setup of
and
5 , - Concluded,
NACA TN 2289
Type of load'ing
A/\/\/\/\A/v\A-\/\/\/\fif\/\/\/\A
I ii4.i.E] 'L
P P P P P P P P P
=
P P P P P P P P P
A/\/\/\/\/\/\/
I
l\)\j\)\)\j\)\/\/\, a
\/\/\A
P P P P P P P
D%*
"%*
NACA TN 2289
101
Figure
6,-
Concluded ,
NACA TN
2289
103
L-66801
(a) Photograph of test specimen and steel side and end plates assembled.
(b) Photograph of test specimen and steel side and end plates disassembled,
I 1
--
7.-
NACA TN 2289
( d ) Photograph of t e s t setup,
K$@zj7
L-58058
Figure
7. - Concluded.