Você está na página 1de 13

Proceedings of the Institution of Mechanical

Engineers, Part G: Journal of Aerospace


Engineering
http://pig.sagepub.com/

Compressive and lamination strength of honeycomb sandwich panels with strain energy calculation
from ASTM standards
M K Khan
Proceedings of the Institution of Mechanical Engineers, Part G: Journal of Aerospace Engineering 2006 220: 375
DOI: 10.1243/09544100JAERO76

The online version of this article can be found at:


http://pig.sagepub.com/content/220/5/375

Published by:

http://www.sagepublications.com

On behalf of:

Institution of Mechanical Engineers

Additional services and information for Proceedings of the Institution of Mechanical Engineers, Part G: Journal of Aerospace Engineering
can be found at:

Email Alerts: http://pig.sagepub.com/cgi/alerts

Subscriptions: http://pig.sagepub.com/subscriptions

Reprints: http://www.sagepub.com/journalsReprints.nav

Permissions: http://www.sagepub.com/journalsPermissions.nav

Citations: http://pig.sagepub.com/content/220/5/375.refs.html

>> Version of Record - May 1, 2006

What is This?

Downloaded from pig.sagepub.com at Queen Mary, University of London on July 11, 2014
375

Compressive and lamination strength of honeycomb


sandwich panels with strain energy calculation
from ASTM standards
M K Khan
Space Structure Lab, Satellite Research and Development Center, Pakistan Space and Upper Atmosphere Research
Commission (SUPARCO), PO Box No. 8402, Karachi 75270, Pakistan. email: kashoo@engineer.com

The manuscript was received on 26 December 2005 and was accepted after revision for publication on 25 April 2006.

DOI: 10.1243/09544100JAERO76

Abstract: The experiments were designed and performed to find the material properties of
honeycomb sandwich panels and to find the difference in properties between aluminium and
of glass fibre facing honeycomb sandwich panels, so as to get a path for consideration of
panel containing maximum stiffness to weight ratio which is the main idea behind material
selection for a spacecraft/satellite structure. Flatwise and edgewise compressive tests were
performed on aluminium and glass fibre facing honeycomb sandwich panels of different core
thickness to investigate the out-of-plane and in-plane compressive strength. The strain
energy and the critical stress were calculated in one of the compression mode. Another set of
experiments was performed to find the lamination strength of bond used for joining facing
and core of honeycomb sandwich panels. Again the tests were performed on both facing
types of sandwich panels to check the difference in strength with the change in core thickness
and facing type of the honeycomb sandwich panel. Different ASTM standards for testing
honeycomb sandwich panels were used.

Keywords: honeycomb sandwich panel, facesheet wrinkling, shear crimping, intracell


buckling, bending stiffness

1 INTRODUCTION which is the strongest axis under different kind of


support and against edgewise configuration, has
Sandwich construction is an attractive structural been investigated by many researchers analytically
design concept, as by the proper choice of materials or experimentally [1 –10] for bare honeycomb core
and geometry, constructions having high ratios of and on honeycomb sandwich panels [11 – 17]. Hoff
stiffness-to-weight can be achieved. The sandwich and Mautner [1, 2] studied the edgewise com-
panel is the composition of a weak core material pression mode and investigated the buckling of
with strong and stiff faces bonded on the upper and sandwich panels. Birman [3] gave a solution based
lower sides. The facings provide practically all of on simultaneous wrinkling of both facings during
the overall bending and in-plane extensional rigidity edgewise compressive mode. Baker et al. [4] deve-
to the sandwich. Sandwich construction is com- loped a method for measuring mechanical properties
monly used in structures where strength, stiffness, of high-density honeycomb for quasi-static and
and weight efficiency are required. Low-density, dynamic loading and then compared the stresses
hexagonal honeycombs are preferred as the core for quasi-static and dynamic loading. Zhao and
material on a performance basis. Honeycomb Gerard [5] presented a new application of split
materials have been extensively used as energy Hopkinson pressure bar for testing honeycombs
absorbers to resist external loads due to their high and compared out-of-plane and in-plane crushing
energy absorption capacity. The bare core of honey- behaviour with loading rate. Goldsmith and Louie
comb specimen is shown in Fig. 1. [6] studied the axial perforation in aluminium honey-
The behaviour of honeycomb sandwich panels comb cores under quasi-static and ballistic con-
under compressive loads, against out-of-plane axis ditions and investigated in-plane and out-of-plane

JAERO76 # IMechE 2006 Proc. IMechE Vol. 220 Part G: J. Aerospace Engineering
Downloaded from pig.sagepub.com at Queen Mary, University of London on July 11, 2014
376 M K Khan

thickness ¼ 0.064 mm, facing thickness ¼


0.5 mm.
2. Sandwich panel of glass fiber facings
thickness ¼ 26.5 mm, cell size ¼ 6.3 mm, adhesive
(resin) ¼ phenolic, core ¼ Al 3003, core density ¼
0.083 g/cm3, cell wall thickness ¼ 0.06 mm,
facing thickness ¼ 0.55 mm.
Flatwise compressive tests were performed on
sandwich panel of each category along with on bare
core (without facings) according to ASTM standard
Fig. 1 Bare core of sandwich panels
[18]. Edgewise compressive tests were performed
according to ASTM standard [19] on three specimens
of each category. The critical values of stresses were
modes of deformation. Zhang and Ashby [7] predicted calculated to predict the possible mode of failure
the different failure modes under compression and and then compared with experimental data. Mid-
showed that out-of-plane strength (compressive and span loading flexure test was performed on both
shear) is independent of height of the honeycomb facing types of honeycomb sandwich panels accord-
and totally dependent on density of honeycomb. ing to ASTM standard [20]. Facing bending stresses
They also studied the stress–strain behaviour in were calculated according to ASTM standard [20].
different modes of failures. Goldsmith and Sackman To determine the strength of the bond, the flatwise
[8] experimentally investigated static and dynamic tensile test was done as specified in ASTM standard
loading on honeycomb cores and panels to determine [21]. The tensile loading is applied to the two oppo-
the energy dissipation and force level transmission site ends of the fixture until the failure of bond
characteristics. Wierzbicki [9] derived a formula between facing skin and honeycomb core occurs.
which is very useful in determining mean crushing From the lamination strength test, it is concluded
strength of metal honeycombs and considered a that all panels have the same lamination strength.
shear failure mechanism. Enboa and Jiang [10] Results of compressive loading on honeycomb
studied the behaviour of aluminium honeycomb sandwich panels indicate that the compressive
under static and dynamic loading. Hong et al. [11] strength of honeycomb sandwich panels primarily
studied the effects of dynamic crush tests on alu- depend upon the facing type and slightly on core
minium honeycomb specimens and concluded simi- thickness. From bare core compressive strength
lar crush behaviour for honeycomb under pure tests, it is concluded that the thicker core has slightly
compressive and inclined loads. Mouring et al. [13] more compressive strength when compared with
studied the effect of impact damage under edgewise thinner core. Results of compressive loading on hon-
compression tests on composite honeycomb sand- eycomb sandwich panels in edgewise mode indicate
wich panels. Ley et al. [15] performed several com- that edgewise compressive strength of aluminium
pressive tests to predict facesheet wrinkling in facing honeycomb sandwich panels is almost equal
sandwich panels. Walker [16] studied edgewise com- to that of glass fibre facing, as both have almost
pressive load behaviour at different temperatures to the same value of modulus of elasticity (for
honeycomb sandwich panels. aluminium, E ¼ 70  109 Pa and for glass fibre,
Different ASTM standards were followed to E ¼ 81.3  109 Pa). From flexure tests, bending stres-
perform tests on different facing types of honeycomb ses and deflection are calculated and comparison
sandwich panels which are most commonly used in between deflection values from graph and formulae
structures of aircraft, spacecraft, and satellites. The is presented. From lamination strength test, it is con-
sandwich panels used in the tests were procured cluded that all the sandwich panels have almost the
from M/S Good Fellow Corporation having the same bond strength between facing and core. The
following characteristics. experimental data can be used to select the more
light and stiff combination of sandwich panels
according to the environmental conditions of
1. Sandwich panel of aluminium facings (Al 5052)
aircraft, spacecraft, and satellites.
(a) thickness ¼ 26.4 mm, cell size ¼ 6.3 mm,
adhesive (Resin) ¼ phenolic, core ¼ Al 3003,
core density ¼ 0.083 g/cm3, cell wall thickness ¼
0.064 mm, facing thickness ¼ 0.5 mm; 2 THEORETICAL BACKGROUND
(b) thickness ¼ 14 mm, cell size ¼ 6.3 mm,
adhesive (resin) ¼ phenolic, core ¼ Al 3003, In edgewise compression [19], the following mode of
core density ¼ 0.083 g/cm3, cell wall failures can occur

Proc. IMechE Vol. 220 Part G: J. Aerospace Engineering JAERO76 # IMechE 2006
Downloaded from pig.sagepub.com at Queen Mary, University of London on July 11, 2014
Compressive and lamination strength of honeycomb sandwich panels 377

Panel buckling Therefore

P P
p2 D p2 D sb1 ¼ and sb2 ¼ (6)
P¼ 2 þ (1) tb tb
L Gc hb
For exact solution [22]
where D is the bending stiffness given by
   
d  t1 bEc
y ¼ bt1 þ tc ½t2 þ d  t1  t2 
Ef th2 b 2 E1
D¼ (2)    
2 (bE2 =E1 )t2 (t2 =2) bEc bE2
þ þ tc þ t2 (7)
bt1 E1 E1
Shear crimping
For same facing type of sandwich panels
P ¼ tc Gc b (3)
bt½d  t=2 þ bt 2 =2
y¼ (8)
Intracell buckling 2bt

and
tf
s ¼ 2Ef (4)
s M(d  y) My E2
fb1 ¼ , fb2 ¼  (9)
I I E1
Skin wrinkling
For flexural test [20] in bending mode
Shear stiffness
s ¼ 0:5{Gc Ec Ef }1=3 (5)
S ¼ bhGc (10)
The formula for skin wrinkling is valid for W equal
to or less than b/2. According to Hoff and Mautner Deflection
[1], reasonable safe values of buckling stress can be
obtained by using simple approximate formula. kb PL3 ks PL
Figure 2(a) shows the end-loading response and d¼ þ (11)
D S
Fig. 2(b) shows the coordinate geometry of honey-
comb sandwich panels. Facing bending stress
From the values of stresses obtained from these
expressions, the values of bending stresses obtained M
from different solution methods at failure load s¼ (12)
htf b
values obtained from the tests are compared. The
bending stresses produced due to edgewise where
compressive loading have two types of solution, i.e.
standard and exact. According to standard or PL
approximate solution [22] M¼ (13)
4

M M The bending coefficient kb and ks can be obtained


sb1 ¼ and sb2 ¼ where M ¼ Ph by boundary conditions of beam. For flexure test,
tbh tbh
values of kb and ks are 1/48 and 1/4, respectively.

2.1 Flatwise compressive test


Flatwise compressive tests were performed to find
difference in strength of the sandwich panel when
compressive loading is applied in out-of-plane
direction. The purpose of the test was to investigate
the difference in strength with respect to core
thickness and facing type of the sandwich panels.
Fig. 2 (a) Approximate bending stress and (b) Tests were performed according to ASTM standard
Cartesian coordinate system [18] on sandwich panel of aluminium facings

JAERO76 # IMechE 2006 Proc. IMechE Vol. 220 Part G: J. Aerospace Engineering
Downloaded from pig.sagepub.com at Queen Mary, University of London on July 11, 2014
378 M K Khan

Fig. 5 Thick and thin core specimen during testing

Three specimens of each category were tested


using this procedure for which description is given
in Table 1, whereas Fig. 6 is the comparison
between strength of different specimens. The
flatwise compressive strength was computed by
Fig. 3 Graphical representation of compressive dividing the failure load by the area of the
strength for flatwise compressive test specimen. Average nominal compressive strength
of aluminium (thickness 26.4 mm), glass fibre (thick-
ness 26.4 mm), and aluminium (thickness 14 mm)
was found to be 4.99, 3.55, and 4.2 MPa, respectively,
(thickness 26.4 and 14 mm) and glass fibre facings with standard deviation of 0.21, 0.19, and 0.18 MPa,
(thickness 26.4 mm) (Fig. 3). The compressive load- respectively.
ing is applied to the specimen until the failure of Bitzer [22] found that the compressive properties
core occurs. The tests were done with a rate of and shear moduli do not vary much with thickness
25 kg/s using a 117.6 kN load cell with a 23.5 kN changes, but shear strength becomes lower as thick-
range. After installing the specimens into the test- ness increases due to which slight difference in
ing machine, the force was ramped up until failure strength of thick and thin core aluminium facing
of the interface occurred. The specimens are shown sandwich panels are obtained.
in Figs 4 and 5. From experimental data, it is concluded that for
equal core thickness, flatwise compressive strength
of aluminium facing sandwich panels is higher
than that of glass fibre facing panel, which reflects
that for sandwich panels compressive strength
depends upon facing type. The thin core aluminium
facing panel showed greater compressive strength
when compared with thick core glass fibre facing
panel, which reflects that there is no such effect of
thickness of core on compressive strength. Similarly,
almost the same compressive strength of thick and
thin core aluminium facing honeycomb sandwich
Fig. 4 Specimen during and after compressive test panels is obtained.

Table 1 Results of flatwise compressive test on sandwich panels

Size of specimen Failure load Strength Mean Standard COV ¼ standard


(mm2) (kg) (MPa) (u) deviation deviation/mean (%)

Al facing (thickness 50  50 1270 4.97


26.4 mm) 70  70 2400 4.80 4.99 0.21 4.2
60  100 3200 5.22
Glass fibre facing 54.5  5.35 1055 3.41
(thickness 26.4 mm) 55  54 1072 3.47 3.55 0.19 5.35
55  57 1210 3.78
Al facing (thickness 55  53.5 1280 4.26
14 mm) 53  54.5 1285 4.35 4.20 0.18 4.28
54.5  54 1203 4.00

Proc. IMechE Vol. 220 Part G: J. Aerospace Engineering JAERO76 # IMechE 2006
Downloaded from pig.sagepub.com at Queen Mary, University of London on July 11, 2014
Compressive and lamination strength of honeycomb sandwich panels 379

data, it is concluded that flatwise compressive


strength of thick and thin core is almost the same,
giving again the same idea that compressive strength
is not a function of core thickness (Fig. 3).

2.3 Edgewise compressive test


Edgewise compressive tests were performed to find
difference in strength of the sandwich panel when
compressive loading is applied in in-plane direction.
The purpose of the test was to investigate the differ-
ence in strength with respect to core thickness and
facing type of the sandwich panels and to predict
the expected failure mode a sandwich panel can
suffer in this kind of compressive loading. The stres-
ses were calculated according to standard and exact
solutions, as described by Bitzer [22]. Edgewise com-
Fig. 6 Graphical representation of compressive pressive test consists of compressively loading the
strength for bare core panel to the facing plane. Sandwich panels in this
mode of loading usually fails by facing wrinkling
(dimpling), shear crimping, or overall buckling.
2.2 Bare compression method Therefore, it is rare to obtain the ultimate compres-
sive strength of the facing material, but results give
Flatwise compressive tests were performed to find a near value of compressive strength (Fig. 7).
difference in strength of the core of the sandwich Edgewise compressive tests were performed
panel when compressive loading is applied in according to ASTM standard [19] on sandwich
out-of-plane direction (Fig. 6). The purpose of the panels of each category. The compressive loading is
test was to investigate the difference in strength applied to the specimen until the failure of core
with respect to core thickness. Compressive loading occurs. The tests were done in load control with a
is applied to the specimen until the failure of core rate of 12.5 kg/s using a 250 kN load cell with a
occurs. The tests were done with a rate of 60 N/s 45 kN range. After installing the specimens into the
using a 117.6 kN load cell with a 23.5 kN range. testing machine, the force was ramped up until fail-
After installing the specimens into the testing ure occurred. To prevent buckling of facings, the core
machine, the force was ramped up until failure of of a small part around 10 mm is removed from both
the core occurred as shown in Fig. 5. the edges and then adhesive Araldite 2013 is inserted
Six specimens of different core thickness were in that empty region and the specimens allowed to
tested according to ASTM standard [18] on honey- cure at 50 8C. After curing, the facings of the part
comb core of aluminium for which details of results enclosed with adhesive is removed, so as to get
are given in Table 2. The compressive strength of alu- cured Araldite on both the edges of specimen. For
minium (thick and thin core) was found to be 2.67 proper surface finish at both the edges, the adhesive
and 2.42 MPa, respectively, with standard deviation part is properly machined to get a smooth surface as
of 0.12 and 0.37 MPa, respectively. shown in Fig. 8.
Compressive strength obtained in this case is less Nine specimens were tested (three of each
when compared with sandwich panels. According category) according to this procedure for which the
to Bitzer [22], adhesive fillet stabilizes the honey- mean strength is obtained as 1.45 MPa with standard
comb cell walls and produces slightly higher results deviation of 0.26 MPa for sandwich panel of alu-
than bare compression test. From experimental minium facing (thickness 26.4 mm), 1.44 MPa with

Table 2 Results of flatwise compressive test on bare core

Size of specimen (mm2) Core thickness (mm) Breaking load (kg) Strength (MPa) Mean (u) Standard deviation COV (%)

56  37 600 2.83
33.7  44.6 22.2 390 2.54 2.71 0.12 4.49
49.7  48.9 690 2.78
33  65 10.2 500 2.28
36.6  72.3 9.6 770 2.85 2.42 0.37 15.2
33.5  62.4 10.2 460 2.15

JAERO76 # IMechE 2006 Proc. IMechE Vol. 220 Part G: J. Aerospace Engineering
Downloaded from pig.sagepub.com at Queen Mary, University of London on July 11, 2014
380 M K Khan

The description of the test results is given in


Table 3. The load deflection curve of some speci-
mens is given in Fig. 9, and different specimens
after testing are shown in Fig. 10.
The critical stress values at which different failure
mode can occur for first specimen in Table 3
are D ¼ 657.39 N mm2, sbuck ¼ 0.31 MPa, scrimp ¼
124.46 MPa, swrinkle ¼ 1445 MPa, and sintra ¼
881 MPa.
Now, calculating the stress at which failure
occurred according to ASTM standard [15]

P
Facing stress ¼ sf ¼ ¼ 186:4 MPa
tf b

Fig. 7 Graphical representation of edgewise Comparing the value of facing stress with the
compressive strength above values of critical stress, it was found that
both panel buckling and shear crimping can be
obtained as a possible failure mode, as critical
stress values for these failure modes are lower than
critical values of the remaining failure mode. There-
fore, from this method, the failure mode of any
specific honeycomb sandwich panel can be
predicted.
Three specimens of each category are tested
according to ASTM standard [19] and it was con-
cluded that the edgewise compressive strength of
aluminium facing honeycomb sandwich panels is
equal to that of glass fibre facing honeycomb sand-
wich panels, as both have same facing thickness
Fig. 8 Specimen with cured Araldite 2013 at both ends and this kind of loading is primarily supported by
the core and not the facings. The strength of thick
core honeycomb panel is slightly higher than that
standard deviation of 0.14 MPa for sandwich panel of of thin core honeycomb panels, as there is some
glass fibre facing (thickness 26.4 mm), and 1.32 MPa account of core thickness in shear crimping as
with standard deviation of 0.29 MPa for sandwich reflected in equation (3). The comparison in
panel of aluminium facing (thickness 14 mm). edgewise compressive strength is given in Fig. 7.

Table 3 Results of edgewise compressive test on sandwich panels


Bending
stress

Size of Breaking load Strength Mean Standard COV D sbuck scrimp swrink sintra Standard Exact Predicted
specimen (kN) (MPa) (u) deviation (%) (N m2) (MPa) (MPa) (MPa) (MPa) (MPa) (MPa) failure mode

Aluminium facing (panel thickness ¼ 26.4 mm)


56  50 4.66 1.66 657.39 0.31 124.46 1445 881 166.4 177.0 Buck/cri
51  49 4.27 1.55 1.45 0.26 17.9 598.69 0.28 127 1445 881 167.4 182.4 Buck/cri
53  51 3.12 1.15 622.17 0.29 122.01 1445 881 115.4 133.3 Buckling
Glass fibre facing (panel thickness ¼ 26.4 mm)
55  50 4.15 1.50 824.86 0.315 124.4 1519 1239 137.1 141.8 Buck/cri
51  50 3.28 1.28 1.44 0.14 9.7 764.8 0.292 124.4 1519 1239 116.9 124.5 Buckling
55  50 4.27 1.55 824.86 0.315 124.4 1519 1239 141.1 145.9 Buck/cri
Aluminium facing (panel thickness ¼ 15.5 mm)
35  49 1.95 1.13 111.6 0.089 65 1445 881 111.4 184.2 Buck/cri
34  48 2.71 1.66 1.32 0.29 21.9 108.4 0.090 66.3 1445 881 159.4 256 Buck/cri
34  50 4.27 1.17 108.4 0.089 63.7 1445 881 251 403.4 Buck/cri

Proc. IMechE Vol. 220 Part G: J. Aerospace Engineering JAERO76 # IMechE 2006
Downloaded from pig.sagepub.com at Queen Mary, University of London on July 11, 2014
Compressive and lamination strength of honeycomb sandwich panels 381

Fig. 9 Load –deflection curve of thick core aluminium and glass fibre facing specimen for edgewise compressive
strength

2.4 Strain energy calculation and


The mode shape of wrinkles in facings is assumed txz ¼ Gc gxz (19)
sinusoidal, that is du
ex ¼ (20)
dx
px  dv
ez ¼ (21)
v ¼ W sin (14) dz
L
du dv
gxz ¼ þ (22)
dz dx
The basic equations of stresses can be written from
the theory of elasticity for plane stress edgewise According to Hoff and Mautner [1], vertical displa-
compression cement ‘u’ at buckling is very small and can be neg-
lected. Therefore

dsx dtxz dv
þ ¼0 (15) gxz ¼ (23)
dx dz dx
dsz dtxz
þ ¼0 (16)
dz dz Assume symmetric compression as shown in
Fig. 11 as described by Hoff and Mautner [1],

whereas

sx ¼ Qxx ex þ Qxz ez (17)


sz ¼ Qxz ex þ Qzz ez (18)

Fig. 10 (a) Shear crimping failure and (b) pure


edgewise compression Fig. 11 Assumed shape of symmetric compression

JAERO76 # IMechE 2006 Proc. IMechE Vol. 220 Part G: J. Aerospace Engineering
Downloaded from pig.sagepub.com at Queen Mary, University of London on July 11, 2014
382 M K Khan

edgewise compression mode involves bending of The strain energy of bending stored in one face is
face thickness and the elongation and compression
of horizontal fibres of the core that have initial  ð  2
Ef If L d2 vW
length ‘b’. According to Hoff and Mautner [1] displa- Uf ¼ dx
2 0 dx 2
cement occur only in marginal zones of depth ‘W ’.
The vertical displacement is negligibly small. The  2  4   3
a p t
horizontal displacement ‘v’ of the point originally ¼ Ef (31)
2 24 L
at x, y on one side of the marginal zone is given by
the equation
The shortening length DL of the distance between
the points that were originally at x ¼ 0 and x ¼ L can
z px  be calculated as
v¼a sin (24)
W L
ðL   2  2 
1 dvW 2 a p
DL ¼ dx ¼ (32)
2 0 d x 4 L
Therefore
The constant force is
dv  p  z  px 
¼a cos (25) P ¼ st
dx L W L
dv  a  px 
¼ sin (26) Now the work done by the compressive force is
dz W L
P(DL)

Put these values in equations (21) and (23) to get  2  2  


a p t
normal and shear strain Work ¼ P(DL) ¼ s (33)
2 2 L

a px  The equation is given by


ez ¼ sin (27)
W p  z L px  Work ¼ Ue þ Ug þ Uf
gxz ¼a cos (28)
L W L
From the above equation, s can be calculated
(buckling stress)
Now, work done by the compressive force during
the vertical displacement of its point of attack    2    
1 L 1 W
caused by the bending of the face layer be equal to s¼ Ec þ Gc
the strain energy of bending stored in the face p2 tW 3 t
material plus strain energy of extension and shear  2   2
p t
stored in the core. þ Ef (34)
The compressive load carried by the core is 12 L
neglected. Poisson’s ratio for the core material is
taken as zero. Because of symmetry, it is sufficient According to Hoff and Mautner [1] the actual
to calculate the work and the strain energy for one values of parameters W and L are those which
half of the cross-section. The width of the sandwich make critical stress a minimum causing ds/dW and
skin perpendicular to the plane of the drawing is ds/dL as zero
taken as unity. The strain energy can be obtained    2     
by integrating the respective strain over both ds 1 L 1 1
¼  2 Ec 2
þ Gc ¼0 (35)
horizontal and vertical displacements. dW p tW 3 t
The extensional strain energy and shear strain      2  2 
ds 2 L p t
energy within one half of the wavelength becomes ¼ Ec  Ef 3 ¼ 0 (36)
dL p2 tW 6 L

  ðL ðW  2    Solution of these equations gives


Ec a Ec L
Ue ¼ e2z dzdx ¼ (29)
2 2 2 W  
0 0
  ðL ðW  2  2    W Ef Ec 1=3
Gc a p W ¼ 0:91 (37)
Ug ¼ gxz dydx ¼
2
Gc t Gc2
2 0 0 2 6 L L
(30) ¼ 1:65(Ef2 Ec Gc )1=6 (38)
t

Proc. IMechE Vol. 220 Part G: J. Aerospace Engineering JAERO76 # IMechE 2006
Downloaded from pig.sagepub.com at Queen Mary, University of London on July 11, 2014
Compressive and lamination strength of honeycomb sandwich panels 383

Put these values in equation of s values of deflection are obtained from the load –
deflection curve. Three specimens of each category
s ¼ 0:91(Ef Ec Gc )1=3 (39) were tested according to this procedure for which
different values of ultimate load given in Table 4
The above formula is valid for W equal to or less than are obtained. The bending stresses were calculated
b/2. According to Hoff and Mautner [1], reasonable according to ASTM standard [20]. Compressive load-
safe values of buckling stress can be obtained by ing is applied to the specimen until the failure of core
using simple approximate formula occurs. The tests were done in load control with a
rate of 0.1 kN/s using a 250 kN load cell with a
s ¼ 0:5(Ef Ec Gc )1=3 (40) 50 kN range.
As thin core aluminium facing honeycomb sand-
In Table 3, the values of intracell buckling is calcu- wich panels have low bending stiffness, less values
lated with this simple formula rather than calculated of failure loads and larger values of bending stresses
from relation derived by strain energy calculation. are obtained. Glass fibre has a large value of modulus
For other values of W of elasticity (81.3 GPa), due to which glass fibre
facing honeycomb sandwich panel has more bend-
    ing stiffness when compared with aluminium
2z px
v¼a sin (41) facing honeycomb sandwich panel of same thick-
b L
ness, and similarly, glass fibre facing honeycomb
and sandwich panels exhibit more bending stresses.
Therefore, for a case of pure bending with stiffness
    to weight constraint, thin core sandwich panels
Ef Ec t b
s ¼ 0:817 þ 0:166Gc (42) and glass fibre facing sandwich panels are good
b t
choices.

2.5 Flexure test


2.6 Lamination strength test of honeycomb
Flexure tests were performed to find elastic deflec- sandwich panels
tion pattern and difference in strength of the sand-
Flatwise tensile tests were performed to find differ-
wich panel when compressive loading is applied in
ence in bonding strength of the resin used to join
mid-span of the specimen supported on two ends.
facing and core of the sandwich panel when tensile
The purpose of the test was to investigate the
loading is applied in out-of-plane direction. The pur-
nature of curvature with respect to core thickness
pose of the test was to investigate the difference in
and facing type of the sandwich panels. Mid-span
strength with respect to core thickness and facing
flexure tests were performed according to ASTM
type of the sandwich panels. Flatwise tensile tests
standard [20] on each sandwich panel of each cat-
provide a measure for interface strength against
egory. The sandwich specimens were placed in a
loading perpendicular to the plane of the sandwich
single-point bending apparatus which supported
panel. To determine the strength of the bond, nine
the specimen. Mid-span loading is applied to the
specimens were tested as specified in ASTM standard
specimens until a complete curvature is obtained
[21] for which description of the results is given in
as shown in Fig. 12. A little part of the cell configur-
Table 5. The tensile loading is applied to the two
ation of honeycomb sandwich panels was also
opposite ends of the fixture until the failure of
affected by loading due to which slightly larger
bond between facing skin and honeycomb core
occurs as shown in Figs 13 and 14. Transmission of
the loads to the sandwich coupon is achieved
through thick loading blocks that are bonded to the
face sheets through epoxy Araldite 2013. The tests
were done with a rate of 5 mm/min using a 6000 kg
load cell with a 12 000 kg range. The adhesive bond
strength was computed by dividing the failure load
by the bonded area. Araldite 2013 is used as bonding
of facing to fixture for aluminium facing honeycomb,
and Araldite 2015 is used as bonding of facing to
fixture for glass fibre facing honeycomb to ensure
that this bond strength is greater than the lamination
Fig. 12 Specimens after single point flexure test strength of honeycomb adhesive (Fig. 15). To ensure

JAERO76 # IMechE 2006 Proc. IMechE Vol. 220 Part G: J. Aerospace Engineering
Downloaded from pig.sagepub.com at Queen Mary, University of London on July 11, 2014
384 M K Khan

Table 4 Results of flexure test on sandwich panels


Bending stress
Size of specimen Breaking load D
(mm2) (kN) (N m2) S (MPa) sb (MPa) d (mm)

Al facing (26.4 mm) 140  70 2.60 821.74 444 185 100.3 0.385
140  70 2.60 821.74 444 185 100.3 0.385
140  70 2.73 821.74 444 185 105.4 0.404
Al facing (14 mm) 140  70 2.25 275.625 257 250 166.6 0.91
140  70 2.00 275.625 257 250 148.1 0.81
140  70 2.10 275.625 257 250 155.5 0.89
Glass fibre facing 140  70 2.80 1049.83 444 185 108.1 0.38
(26.4 mm) 140  70 2.60 1049.83 444 185 100.3 0.36
140  70 3.32 1049.83 444 185 128.1 0.46

Table 5 Results of lamination strength test

Type of facing Size of specimen Thickness Breaking load (kN) Strength (MPa) Mean (u) Standard deviation COV (%)

Al 50  50 1150 4.50
70  70 26.4 2535 4.47 4.40 0.2 4.54
60  100 2375 4.25
Glass fibre 54.5  5.35 2470 4.30
55  54 26.4 1100 4.31 4.45 0.25 5.6
55  57 1210 4.74
Al 55  53.5 1200 4.70
53  54.5 13.4 2700 4.70 4.74 0.08 1.68
54.5  54 2780 4.84

Fig. 13 (a) Fixture for testing and (b) specimen during Fig. 15 Load –deflection curve of thick core
testing aluminium and glass fibre facing specimen
for flexure test

sufficient bonding strength between fixture and alu-


minium facing of honeycomb, both surfaces were
treated with sand blasting.
Tests were performed on nine specimens (three of
each category) on fixture designed according to
ASTM standard [21]. Average failure load was found
to be 4.4, 4.45, and 4.75 MPa with standard deviation
of 0.2, 0.25, and 0.08 MPa for aluminium facing
(thick. 26.4 mm), glass fibre facing (thick. 26.4 mm),
Fig. 14 Aluminium and glass fibre facing specimen and aluminium facing (thick. 14 mm), respectively
after failure of bond Table 5.

Proc. IMechE Vol. 220 Part G: J. Aerospace Engineering JAERO76 # IMechE 2006
Downloaded from pig.sagepub.com at Queen Mary, University of London on July 11, 2014
Compressive and lamination strength of honeycomb sandwich panels 385

From the results, it has been concluded that there 2 Hoff, N. J. and Mautner, S. E. Bending and buckling of
is no difference in lamination strength of bond sandwich beams. J. Aeronaut. Sci., 1948, 15, 707 – 720.
between facing and core of sandwich panels, as all 3 Birman, V. Thermo mechanical wrinkling in composite
panels exhibit same kind of failure and strength. sandwich structures. AIAA J. Spacecraft Rockets, 2004,
42(7), 1474 –1479.
The lamination strength of glass fibre facing honey-
4 Baker, W. E., Togami, T. C. and Weydert, J. C. Static
comb sandwich panels is slightly towards higher
and dynamic properties of high density honeycomb.
side which shows that phenolic resin has slightly Int. J. Impact, 1998, 21(3), 149 –163.
strong adhesion with glass fibre facing. 5 Zhao, H. and Gerard, G. Crushing behavior of
aluminum honeycombs under impact loading. Int.
J. Impact Eng., 1998, 2(10), 827 – 836.
3. CONCLUSION
6 Goldsmith, W. and Louie, D. L. Axial perforation of
aluminum honeycombs by projectiles. Int. J. Solids
Effect of compressive and tensile loading in flatwise Struct., 1995, 32(8/9), 1017 – 1046.
and edgewise directions on honeycomb sandwich 7 Zhang, J. and Ashby, M. F. The out of plane
panels of different type of facings and core thick- properties of honeycombs. Int. J. Mech. Sci., 1992,
nesses is investigated experimentally and analytically. 34(6), 475 –489.
Experimental data can be used to select the more light 8 Goldsmith, W. and Sackman, L. J. An experimental
and stiff combination of sandwich panels according study of energy absorption in impact of sandwich
to the environmental conditions of aircraft, space- plates. Int. J. Impact Eng., 1992, 12(2), 241 –262.
craft, and satellites. The analytical model of strain 9 Wierzbicki, T. Crushing analysis of metal honeycombs.
Int. J. Impact Eng., 1983, 1(2), 157– 174.
energy can be used to predict possible failure mode
10 Enboa, W. and Jiang, W. Axial crush of metallic
in edgewise compression which is a more compli-
honeycombs. Int. J. Impact Eng., 1997, 19(5–6), 439–456.
cated pattern of loading. Results of compressive 11 Hong, S. T., Pan, J., Tyan, T., and Prasad, P. Effects
loading on honeycomb sandwich panels indicate of impact velocity on crush behavior of honeycomb
that the compressive strength of honeycomb sand- specimens. SAE technical paper series, 2004-01-0245,
wich panels primarily depend upon the facing type 2004 SAE World Congress, Michigan, 2004.
and not on the core thickness, as higher compressive 12 Rouse, M., Ambur, D. R., Dopker, B., and Shah, B.
strength of thin panel of aluminium facing when Response of composite fuselage sandwich side panels
compared with thick glass fibre facing panel is subjected pressure and axial tension. AIAA/ASME/
obtained. From bare core compressive strength ASCE/AHS/ASC, 35th Structures, Structural Dynamics
tests, it is concluded that the thicker core has more and Materials Conference, California, AIAA paper
no. 98-1708, 1998.
compressive strength when compared with thinner
13 Mouring, S. E., Barton, O., and Joyce, P. J.
core. From edgewise tests, it is concluded that edge-
Mechanical behavior of composite sandwich structures
wise compressive strength of aluminium facing hon- subjected to impact damage (US Naval Academy, Depart-
eycomb sandwich panels is equal to that of glass fibre ment of Mechanical Engineering, Annapolis, MD).
facing, as in this type of loading, facing is not involved. 14 Christian, H. L. Flatwise compression L-shear flatwise
From flexure test, bending stress and deflection have tensile characteristics of lightweight carbon honeycomb
been calculated. From lamination strength test, it is core, 2001 (Ultracor Incorporated, California).
concluded that there is no difference in lamination 15 Ley, R. P., Lin, W., and Mbanefo, U. Facesheet wrink-
strength of bond between facing and core of sandwich ling in sandwich structures. NASA/CR-1999-208994,
panels, as all panels exhibit same kind of failure and Northrop Grumman Corporation, El Segundo,
strength. California, 1999.
16 Walker, S. Evaluation of composite honeycomb
sandwich panels under compressive load at elevated
ACKNOWLEDGEMENT temperatures. NASA/TP-1998-207645, NASA Langley
Research Center, Virginia, 1998.
The author thanks Satellite Research and Develop- 17 Ani, U., Zehnder, A. T., and Ingraffea, A. R. Fracture
mechanics approach to facesheet delamination in
ment Center Karachi (SRDC-K) and QA&MR (Quality
honeycomb: measurement of energy release rate of
Assurance and Material Research) Division of
the adhesive bond. Cornell Fracture Group, Cornell
Suparco for their continuous support during the University, available from www.elsevier.com/locate/
whole project. engfracmech, New York 14583, 2002.
18 ASTM C-365-94. Standard test method for flatwise
compressive properties of sandwich cores, 1994 (American
REFERENCES Society for Testing and Materials, Philadelphia).
19 ASTM C-364-94. Standard test method for edgewise
1 Hoff, N. J. and Mautner, S. E. The buckling of compressive strength of sandwich constructions, 1994
sandwich-type panels. J. Aeronaut. Sci., 1945, 12(3), (American Society for Testing and Materials,
285 – 297. Philadelphia).

JAERO76 # IMechE 2006 Proc. IMechE Vol. 220 Part G: J. Aerospace Engineering
Downloaded from pig.sagepub.com at Queen Mary, University of London on July 11, 2014
386 M K Khan

20 ASTM C-393-94. Standard test method for flexural I moment of inertia


properties of sandwich constructions, 1994 (American kb bending coefficient
Society for Testing and Materials, Philadelphia). ks shear coefficient
21 ASTM C-297-94. Standard test method for flatwise ten- L length of sandwich panel
sile strength of sandwich constructions, 1994 (American
M moment
Society for Testing and Materials, Philadelphia).
P failure load
22 Bitzer, T. Honeycomb technology, 1997, chapters 2, 3, 7
(Chapman and Hall, California). Qij reduced stiffness
s cell size
S shear stiffness
BIBLIOGRAPHY t thickness of facing
tc thickness of core
Timoshenko, S. P. and Woinowsky-Krieger, S. Theory of tf facing thickness
plates and shells, 2nd edition, 1981, chapters 2, 5, 6 u displacement in x-direction
(McGraw Hill, Japan). U1 normal strain energy in core
Accessed from www.hexcel.com. Ug shear strain energy in core
Uf bending strain energy in facing
W depth
APPENDIX
y centroidal distance
Notation
a wrinkling depth
A area of specimen g shear strain
b sandwich panel width d deflection
COV coefficient of variance 1 axial strain
d sandwich panel thickness m mean
D bending stiffness s stress
Ec core bending modulus sb bending stress
Ef bending modulus of facing si stress on i-axis
E1 bending modulus of facing 1 sbuck buckling
E2 bending modulus of facing 2 scrimp shear crimping
Gl shear modulus in l-direction sintra intracell buckling
Gw shear modulus in w-direction swrink skin wrinkling
Gc shear modulus of core tij shear stress
h distance from centre top facing to centre v displacement in z-direction
bottom facing vW displacement in z-direction at z ¼ W

Proc. IMechE Vol. 220 Part G: J. Aerospace Engineering JAERO76 # IMechE 2006
Downloaded from pig.sagepub.com at Queen Mary, University of London on July 11, 2014

Você também pode gostar