You are on page 1of 13

See

discussions, stats, and author profiles for this publication at: https://www.researchgate.net/publication/250353420

Effects of Relative Density and Material


Distribution on the Elastic Properties and Yield
Strength of Metallic Honeycombs

Article in Materials Science Forum · January 2010


DOI: 10.4028/www.scientific.net/MSF.638-642.1003

CITATIONS READS

0 24

1 author:

H.X. Zhu
Cardiff University
71 PUBLICATIONS 1,643 CITATIONS

SEE PROFILE

Some of the authors of this publication are also working on these related projects:

Nano-structured composites with enhanced strength, stiffness, fracture toughness and conductivities
View project

Facial Biomechanical Modelling View project

All content following this page was uploaded by H.X. Zhu on 23 April 2015.

The user has requested enhancement of the downloaded file.


(This is a sample cover image for this issue. The actual cover is not yet available at this time.)

This article appeared in a journal published by Elsevier. The attached


copy is furnished to the author for internal non-commercial research
and education use, including for instruction at the authors institution
and sharing with colleagues.
Other uses, including reproduction and distribution, or selling or
licensing copies, or posting to personal, institutional or third party
websites are prohibited.
In most cases authors are permitted to post their version of the
article (e.g. in Word or Tex form) to their personal website or
institutional repository. Authors requiring further information
regarding Elsevier’s archiving and manuscript policies are
encouraged to visit:
http://www.elsevier.com/copyright
Author's personal copy

Mechanics of Materials 43 (2011) 276–286

Contents lists available at ScienceDirect

Mechanics of Materials
journal homepage: www.elsevier.com/locate/mechmat

Combined effects of relative density and material distribution


on the mechanical properties of metallic honeycombs
H.X. Zhu a,⇑, C.Y. Chen b
a
School of Engineering, Cardiff University, Cardiff CF24 3AA, UK
b
Department of Mechanics, Huazhong University of Science and Technology, Wuhan, PR China

a r t i c l e i n f o a b s t r a c t

Article history: This paper presents the combined effects of relative density and material distribution on
Received 14 May 2009 the elastic constants and the yield strengths of metallic honeycombs. Periodic regular hex-
Received in revised form 10 February 2011 agonal cell is employed as the structural model. Cell wall bending, transverse shear and
Available online 5 March 2011
axial stretching/compression are taken as the deformation mechanisms in the analysis.
Closed-form solutions for the yield strengths and all the five independent elastic constants
Keywords: are obtained for honeycombs with cell walls of uniform thickness. For honeycombs with
Metallic honeycombs
cell walls of non-uniform thickness, the closed-form solutions would be too lengthy to
Elastic properties
Yield strength
use in practical applications. We instead provide numerical results to show the combined
Relative density effects of relative density and material distribution on the initial and full yield strengths
Material distribution and all the five independent elastic constants of metallic honeycombs. The results can
serve as a guide for the optimal design of metallic honeycombs.
Crown Copyright Ó 2011 Published by Elsevier Ltd. All rights reserved.

1. Introduction honeycombs with cell walls of uniform thickness. Their


normalised results are consequently independent of the
Honeycombs have been widely used as the core materi- honeycomb relative density. Taking bending and stretch-
als of sandwich panels in aircraft constructions and many ing as the main deformation mechanisms, Warren and
other engineering applications. In the design of sandwich Kraynik (1987) have also developed a model to study the
panels, the honeycomb cores are generally treated as con- in-plane elastic properties of honeycombs. Considering
tinuum materials and the effective mechanical properties the transverse shear as an additional deformation mecha-
are used. By optimising the design of the core honeycomb nism, Silva et al. (1995) and Zhu (2010) have obtained
materials, not only sandwich panel structures can achieve the closed form in-plane elastic constants of regular hon-
high strength, stiffness and energy absorption capacity, but eycombs. Taking bending, stretching and hinging the
also great saving in weight. Therefore, precise predictions deformation mechanisms, Masters and Evans (1996) have
for the strength and stiffness of honeycombs are critical theoretically analysed the in-plane elastic constants of
in enhancing the reliability of designs and ensuring the honeycombs. The normalised in-plane elastic constants of
functions of the designed structures. honeycombs generally depend on the honeycomb relative
Many people have studied the elastic properties and the density (Silva et al., 1995; Zhu et al., 2001; Zhu, 2010) be-
yield strength of honeycombs. Taking bending as the sole cause the axial stretching/compression and transverse
deformation mechanism, Gibson et al. (1982) and Gibson shear play an important role in the deformation. Material
and Ashby (1997) have derived the in-plane elastic con- distribution along the cell wall also greatly affects the in-
stants and plastic collapse strength for regular plane elastic properties (Simone and Gibson, 1998) and
the yield strength (Chen et al., 1999) of honeycombs. In
⇑ Corresponding author. Tel.: +44 29 20874824. addition to the aforementioned deformation mechanisms,
E-mail address: zhuh3@cf.ac.uk (H.X. Zhu). cell irregularity and imperfection are of great influence

0167-6636/$ - see front matter Crown Copyright Ó 2011 Published by Elsevier Ltd. All rights reserved.
doi:10.1016/j.mechmat.2011.02.010
Author's personal copy

H.X. Zhu, C.Y. Chen / Mechanics of Materials 43 (2011) 276–286 277

on the in-plane mechanical properties of honeycombs effects of material distribution and honeycomb relative
(Silva et al., 1995; Chen et al., 1999; Zhu et al., 2001, 2006). density on the in-plane and out-of-plane initial yield
By theoretical analysis and experimental measure- strength, full yield strength and all the five independent
ments, Kelsey et al. (1958) have studied the out-of-plane elastic constants of hexagonal honeycombs with cell walls
shear modulus of aluminium hexagonal cell honeycomb of either uniform or non-uniform thickness. In cases when
sheet used as the core of sandwich panels in aircraft. the analytical solutions are likely to be too lengthy to use
Grediac (1993) has also conducted similar analyses using in practice, we instead provide precise numerical solu-
finite element simulations. Kim and Christensen (2000) tions. The results obtained in this paper are expected to
have obtained the closed-form solutions for the out-of- serve as a guide for the optimal design of metallic honey-
plane mechanical properties of honeycombs with different comb core materials.
types of cell geometry and uniform cell wall thickness, in
addition to the in-plane elastic properties. 2. Elastic properties
As can be seen in Fig. 1a, the width dimension (or the z
direction) of the honeycomb cell walls is much larger than 2.1. Independent elastic constants
their thickness. Although cell wall bending is generally re-
garded as the dominant deformation mechanism, to the Regular hexagonal honeycombs have three orthogonal
best of our knowledge, it has usually been treated as planes with one of isotropy and therefore have only five
plane-stress deformation in research publications. It is in-dependent elastic constants (Nye, 1985; Gibson and
therefore suspected that some of the well known results Ashby, 1997; Kim and Christensen, 2000). Fig. 1a shows a
may contain errors because the in-plane deformation of perfect regular honeycomb and the chosen coordinate sys-
honeycombs is actually a combination of cell wall bend- tem. The linear relationship between the elastic stress and
ing (which is plane-strain deformation), transverse shear strain in a regular honeycomb material is given by (Gibson
and axial stretching/compression. Although the out-of- and Ashby, 1997)
plane mechanical properties of honeycombs have been 2 3
0 1
 m12  m31 0 0 0 0
studied by many people (e.g. Shi and Tong, 1995; e1 1 E
6 m1
E1 E3
7 r1
1
Meraghni et al., 1999; Xu and Qiao, 2002; Pan et al., 6 1 m31 7
B e2 C 6  E1 E1  E3
12
0 0 0 7B r C
2006), as far as we are aware, no publication has consid-
B C 6 m13 7B 2 C
B C 6  mE131 E13 0 0 07 B C
B e3 C 7B r 3 C
ered the combined effect of honeycomb relative density B C ¼ ei ¼ Sij rj ¼ 6
6
E1
B
7B C
Bc C 6 0 0 0 1
0 07 r C
and material distribution along the cell wall on the out- B 23 C 6 G13 B
7B 23 C
B C 6 7 C
of-plane shear modulus and shear yield strength of @ c31 A 6 0 0 0 0 1
0 7@ r31 A
honeycombs. Taking the axial stretching/compression, 4 G13 5
c12 0 0 0 0 0 1 r12
transverse shear and plane-strain bending of the cell walls G12

as the deformation mechanisms, this paper aims to pro- ð1Þ


vide closed-form analytical solutions for the combined

Fig. 1. A regular honeycomb and the unit cells.


Author's personal copy

278 H.X. Zhu, C.Y. Chen / Mechanics of Materials 43 (2011) 276–286

where mij = ej/ei. The in-plane (i.e. x–y plane) isotropy 2 t0


q ¼ pffiffiffi ð4Þ
requires 3 l
E1 To study the effect of the material distribution on the
G12 ¼ ð2Þ
2ð1 þ v 12 Þ mechanical properties of honeycombs, a material distribu-
tion parameter b is introduced and defined as
and the symmetry of the compliance matrix Sij requires
b ¼ t 1 =t2 ð5Þ
m13 m31 Obviously, as it is well known that the bending moment in
¼ ð3Þ
E1 E3 the cell walls increases when approaching the cell wall
junctions, allocating more material near the cell wall junc-
Therefore, to fully determine the elastic relationship be-
tion would make better use of the material. Therefore, it is
tween the stress and the strain in a perfect regular honey-
logical to assume 0 < b 6 1 and b = 1 corresponds to a hon-
comb material, only five independent elastic constants,
eycomb with uniform cell wall thickness. A local coordi-
namely E1, m12, E3, v31, and G13, need to be determined
nate s is chosen as shown in Fig. 1c, with s = 0 at the cell
(Kim and Christensen, 2000). It is noted that, if the thick-
wall junction and s = l/2 at the middle of the cell wall.
ness of the cell walls in the Y direction of Fig. 1a is two
According to the assumption of the linear cell wall thick-
times that of the inclined cell walls, the xy plane will
ness distribution, the relationship between the thickness
not be isotropic and the compliance matrix, which relates
t and the location s is given by
the strain to stress, will have eight independent elastic
constants instead of five as mentioned above. This is be- tðsÞ ¼ t 2  2ð1  bÞt2 s=l; ð0 6 s 6 l=2Þ ð6Þ
cause the honeycomb has three orthogonal planes of sym-
metry and may have up to nine independent elastic
2.3. In-plane elastic constants E1 and m12
constants in general. These independent constants are
S11, S22, S33, S44, S55, S66, S12, S13 and S23. However, if
In order to find the in-plane elastic constants E1 and m12
applying a uniaxial compressive stress in the z direction
for a regular hexagonal honeycomb, we only need to look
(and remaining all other stress components as 0), we
at the deformation of a unit of the connected three half cell
e1 = e2 = S13r3 = S23r3 because the cell walls of the honey-
walls (i.e. a representative unit cell) as shown in Fig. 1c.
comb are uniform and made of the same isotropic mate-
When a global in-plane compressive stress rxx is applied
rial. Therefore S13 = S23 and the number of the
to the honeycomb of Fig. 1a in the x direction, the bending
independent elastic constants reduces to eight (from
moment at the middle of the inclined cell walls is zero be-
nine).
cause of the symmetry of the honeycomb structure and the
applied load. The horizontal force P acting at point B of the
2.2. Cell wall material and cell wall thickness distribution
inclined half cell wall BO is

The honeycomb is assumed to be made of an isotropic, 3


P¼ lrxx ð7Þ
elastic and perfectly plastic material with Young’s modulus 2
Es, shear modulus Gs, Poisson’s ratio vs and yield strength As there is no junction rotation because of the symmetry,
rys. It has been generally recognised that material distribu- we only need to look at the deformation of the inclined half
tion can greatly affect the in-plane mechanical properties cell wall BO as shown in Fig. 1c. The deformation mecha-
of honeycombs (Simone and Gibson, 1998; Chen et al., nisms of bending, transverse shear and axial compression
1999). To simplify the analysis and to obtain closed form are taken into account in the analysis.
analytic results, in this paper the cell wall thickness is as- Now we first look at the bending deformation. The
sumed to increase linearly from t1 at the middle of the cell bending moment in the half cell wall OB is a function of
walls to t2 at the cell wall junction, as shown in Fig. 1c. The the location s, and given by
assumption of linear material distribution is often adopted  
in research papers, such as Chen et al. (1999), because not l
MðsÞ ¼  s P cos 60 ð8Þ
only is it convenient for analysis, but also easy for manu- 2
facture. In practice, the most popular way to produce hon- Obviously, the inclined cell walls undergo plane-strain
eycombs is to stick the corrugated (half-hexagonal profile) bending because the dimension of the cell walls in the z
sheets together. Therefore, the thickness of the cell walls in direction (see Fig. 1b) is much larger than the wall thick-
the Y direction of Fig. 1a is often two times that of the in- ness. In the following analysis, the dimension of the cell
clined cell walls. In the strength or economic point of view, walls in the z direction is treated as a unit (which is sup-
the double thickness of the cell walls in the Y direction is a posed to be much larger than the call wall thickness t).
waste of material. It is easy to choose and control the right Assuming that the displacements and the rotation of point
parameters for the corrugation process of honeycomb O (i.e. the junction of the connected three half cell walls)
manufacture and hence to produce honeycombs with all are zero, the displacement of point B in the x direction
the cell walls of the same linear thickness distribution. due to plane-strain bending is given by
The length of the cell walls is l and the mean cell wall l  !
Z Z
thickness is t0 = (t1 + t2)/2. The honeycomb relative density  2
l=2 s
ð1  v 2s Þ  s Pds
Dbx ¼ ðcos 60 Þ 2
ds ð9Þ
is the ratio of the volume of the solid material to the vol- 0 0
1
Es 12 ½t 2  2ð1  bÞt2 s=l3
ume of the honeycomb, given by
Author's personal copy

H.X. Zhu, C.Y. Chen / Mechanics of Materials 43 (2011) 276–286 279

the displacement in the x direction due to transverse shear


is given by
Z l Z l=2
2 1:2Pds 1:2P ds
Dsx ¼ ðcos 60 Þ2 ¼
0 Gs tðsÞ 4Gs 0 t 2  2ð1  bÞt 2 s=l
3Pl
¼ ln b ð10Þ
20ð1  bÞGs t 2
and the displacement in the x direction due to axial com-
pression is given by
Z l Z l=2
2 Pds 3P ds
Dcx ¼ ðsin 60 Þ2 ¼
0 Es tðsÞ 4Es 0 t2  2ð1  bÞt 2 s=l
3Pl
¼ ln b ð11Þ
8ð1  bÞGs t 2
The total displacement of point B in the x direction is
therefore
Fig. 2. Dimensionless in-plane Young’s modulus vs. material distribution
parameter b for honeycombs with different values of relative density.
Dx ¼ Dbx þ Dsx þ Dcx ð12Þ
It is noted that in Eq. (10) a correction coefficient of 1.2 is
introduced for the shear deformation of a beam or a plate distribution parameter b. Considering both the out-of-
with a rectangular cross-section (Timoshenko and Goodier, plane shear modulus (to be discussed later) and in-plane
1982). Young’s modulus, the range of the material distribution
The compressive strain (which is assumed to be posi- parameter should be within 0.2 6 b 6 0.5.
tive) of the honeycomb in the x direction is given by For regular hexagonal honeycombs with cell walls of
uniform thickness, b = 1 and t2 = t0. Eq. (12) reduces to
Dx 4 Dx
ex ¼ ¼ pffiffiffi ð13Þ
ð1  m2s ÞPl
3
3ð1 þ v s ÞPl 3Pl
ðl=2Þ cos 30 3l Dx ¼ þ þ ð16Þ
8Es t30 10Es t 0 8Es t0
Using Eqs. (7) and (13), the effective in-plane Young’s mod-
ulus of the honeycomb can be obtained as and the normalised (or dimensionless) in-plane Young’s
rxx modulus E1 is obtained as
E1 ¼ ð14Þ
ex E1 ð1  m2s ÞE1 1
E1 ¼ 1:5Es ¼ ¼   ð17Þ
q
3 1:5Es q 3 4:05 1:8v s
v2 q
To make the results more useful, the Young’s modulus E1 is 1 þ 1 þ 1 2
1m2s m2
1:5Es 3 s s
normalised by 1 v 2 q and given bys
Eq. (17) is the same as that of Zhu (2010) and slightly dif-
E1 ð1  m2s ÞE1 ferent from Silva et al.’s (1995) closed-form result because
E1 ¼ 1:5Es ¼ ð15Þ
1m2s
q3 1:5Es q3 they treat all the deformation mechanisms (i.e. cell wall
bending, axial compression and transverse shear) as
where the honeycomb relative density q is given in Eq. (4). plane-stress deformation, while we treat cell wall bending
It is possible to obtain the closed-form analytic result of as plane-strain deformation but both axial compression
E1 for regular honeycombs with different values of relative and transverse shear as plane-stress deformation. If the
density and cell walls of non-uniform thickness. However, relative density of a honeycomb is very small, the dimen-
as it is likely to be too lengthy to be of practical application, sionless in-plane Young’s modulus given in Eq. (17) is very
we just show the precise numerical results. For different close to Gibson et al.’s (1982) result except that their result
fixed values of honeycomb relative density q, the numerical is normalised by a different constant.
results for the relationship between the dimensionless (or To obtain the in-plane Poisson ratio m12, we need to find
normalised) in-plane Young’s modulus E1 and the material the displacement of point B in the y direction and hence
distribution parameter b are shown in Fig. 2. It should be find the strain ey. The displacement components of point
noted that in obtaining the results shown in Fig. 2, the Pois- B in the y direction due to the combination of plane-strain
son ratio of the solid material is assumed to be ms = 0.3. bending, plane-stress transverse shearing and axial com-
Fig. 2 shows that both the relative density q and the mate- pression are given respectively by
rial distribution b can greatly affect the dimensionless in-
plane Young’s modulus E1 . The larger the relative density, Dby ¼ Dbx tan 60
pffiffiffi Z l=2 Z s !
the smaller the dimensionless Young’s modulus E1 . This is ð1  v 2s Þð2l  sÞPds
3
consistent with the findings by Silva et al. (1995) and Zhu ¼ 3
ds ð18Þ
4 0 1
0 Es 12 ½t 2  2ð1  bÞt 2 s=l
et al. (2001). However, material distribution parameter b
can either increase the in-plane Young’s modulus (Simone Dsy ¼ Dsx tan 60
and Gibson, 1998) or reduce it. For different values of fixed pffiffiffi Z l pffiffiffi
3 2 1:2Pds 3 3Pl
honeycomb relative density q, the peak dimensionless ¼ ¼ ln b ð19Þ
4 0 Gs tðsÞ 20ð1  bÞGs t2
Young’s modulus appears at different values of the material
Author's personal copy

280 H.X. Zhu, C.Y. Chen / Mechanics of Materials 43 (2011) 276–286

pffiffiffi Z l=2
Dcx 3P ds density can greatly affect the Poisson ratio; the larger the
Dcy ¼ ¼ honeycomb relative density, the smaller the Poisson ratio.
tan 60 4Es 0 t 2  2ð1  bÞt2 s=l
pffiffiffi
3Pl
¼ ln b ð20Þ 2.4. Out-of-plane Young’s modulus E3, shear modulus G13 and
8ð1  bÞGs t2
Poisson ratio m31
and
Obviously, the Young’s modulus of a honeycomb in the
Dy ¼ Dby þ Dsy þ Dcy ð21Þ z direction is proportional to the Young’s modulus Es of the
solid material from which the honeycomb is made and
Again, it is noted that in Eq. (19) a correction coefficient of proportional to the honeycomb relative density q (Gibson
1.2 is introduced for the shear deformation of a beam or a and Ashby, 1997; Kim and Christensen, 2000). To simplify
plate with a rectangular cross-section (Timoshenko and the results, the out-of-plane Young’s modulus is norma-
Goodier, 1982). The strain of the honeycomb in the y direc- lised by Esq and hence given by
tion due to the horizontal compressive stress rxx is
E3
therefore E3 ¼ ¼1 ð25Þ
Es q
Dy 4 Dy
ey ¼ ¼ ð22Þ The out-of-plane Poisson ratio m31 of a honeycomb is obvi-
l=2 þ ðl=2Þ sin 30 3l ously the same as ms of the solid material (Gibson and
The Poisson ratio m12 is obtained by the definition Ashby, 1997; Kim and Christensen, 2000).
Equation (1) shows clearly that G32 is equal to G31. In
ey the following, we are going to derive G32 instead of G31.
m12 ¼  ð23Þ
ex To derive the out-of-plane shear modulus of the honey-
comb core material, the size of the honeycomb core is as-
where the strains ey and ex are given in Eqs. (23) and (14),
sumed to be much larger than the hexagonal cells and
respectively.
hence much larger than the cell wall length l in order to
For different fixed values of honeycomb relative density
eliminate boundary effects. Fig. 4 shows the shear loads
q, the numerical results of the relationship between the
on the area of a representative three-half-cell-wall unit.
in-plane Poisson ratio m12 and the material distribution
As the size of the honeycomb core is assumed to be much
parameter b are shown in Fig. 3. When obtaining the
larger than the cell wall length l, the product of the cell
results shown in Fig. 3, the Poisson ratio of the cell wall
wall thickness t(s) and the shear stress s(s) in the cell wall
material is assumed to be ms = 0.3.
remains approximately constant for each of the cell walls,
For regular hexagonal honeycombs with cell walls of
i.e.
uniform thickness, i.e. b = 1, the closed-form result of the
Poisson ratio m12 is obtained as C 1 ¼ ½sðsÞ  tðsÞ1 ¼ ½Gs cðsÞtðsÞ1 and
C 2 ¼ ½sðsÞ  tðsÞ2 ¼ ½Gs cðsÞtðsÞ2 ð26Þ
1 þ ð1:05 þ 1:8v s Þq2 =ð1  v 2s Þ
m12 ¼ ð24Þ
1 þ ð4:05 þ 1:8ms Þq2 =ð1  m2s Þ the equilibrium conditions require

Similar to the case of the in-plane Young’s modulus, Eq. C 1 ¼ 2C 2 ð27Þ


pffiffiffi
(24) for the Poisson ratio m12 is the same as that of Zhu l l 3 3 2 

(2010), but slightly different from the results of Gibson et C 1 þ 2C 2 cos 60 ¼ l s32 ð28Þ
2 2 4
al. (1982) and Silva et al.’s (1995) due to the reasons men-
tioned above. Fig. 3 shows clearly that honeycomb relative

Fig. 3. Poisson ratio vs. material distribution parameter b for honeycombs Fig. 4. Shear stresses in the connected three half-cell-walls of a repre-
with different values of relative density. sentative unit cell.
Author's personal copy

H.X. Zhu, C.Y. Chen / Mechanics of Materials 43 (2011) 276–286 281

where s32 is the


pffiffi
equivalent (or effective) shear stress acting
2
on the area 3 4 3 l of the representative unit of the honey-
comb shown in Fig. 4.
The shear strains in the connected three half-cell-walls
of the representative unit cell can be solved from Equations
(6), (26)–(28) and given by
pffiffiffi
C1 3ls32 1
½cðsÞ1 ¼ ¼  ð29aÞ
Gs ½tðsÞ1 Gs t2 1  2ð1  bÞs=l

and
pffiffiffi
C2 3ls32 1
½cðsÞ2 ¼ ¼  ð29bÞ
Gs ½tðsÞ2 2Gs t2 1  2ð1  bÞs=l

The equivalent (or effective) shear strain of the honeycomb


representative unit is
"Z Z # Fig. 5. Dimensionless shear modulus G32 vs. material distribution
l=2 l=2
 1 parameter b.
c32 ¼3 ½cðsÞ1 ds þ ½cðsÞ2 ds
4
l 0 0
pffiffiffi
3ls32 ln b ð1 þ bÞs32 ln b When a regular metallic honeycomb is compressed by a
¼ ¼ ð30Þ
ð1  bÞGs t 2 ð1  bÞGs q uniform stress rxx in the x direction, the maximum com-
pressive stress locates in the surface of the inclined cell
The equivalent out-of-plane shear modulus of the honey- walls, which consists of two parts: bending stress and axial
comb is therefore compressive stress. The amplitude is a function of the loca-
s32 ð1  bÞGs q tion s (see Fig. 1c) and given by
G32 ¼ G31 ¼ ¼ ; ð0 < b  1Þ ð31Þ
c32 ð1 þ bÞ ln b 6MðsÞ P sin 60
rðsÞ ¼ þ
Eq. (31) shows clearly that for a given value of the material ½tðsÞ2 tðsÞ
l  pffiffiffi
distribution parameter b, the honeycomb effective out-of- 9 2  s lrxx 3 3lrxx
¼
2 þ
ð34Þ
plane shear modulus is proportional to the honeycomb rel- 4t 2 1  2ð1  bÞ sl
2t 22 1  2ð1  bÞ sl
ative density q and proportional to the shear modulus Gs of
the solid material. To simplify the results, the out-of-plane For honeycombs with relative density q and material dis-
shear modulus is normalised by Gsq and hence given as tribution parameter b, letting
G32 ð1  bÞ drðsÞ
G32 ¼ ¼ ð32Þ ¼ 0;
Gs q ð1 þ bÞ ln b ds
The relationship between normalised shear honeycomb the location of the maximum compressive stress rmax can
modulus G32 and the material distribution parameter b is be determined as
plotted in Fig. 5. Obviously, it is independent of the honey- ð1  2bÞð1 þ bÞ þ qð1  bÞ l
comb relative density q. When b = 1, Eq. (32) reduces to sm ¼  ; ð0 6 sm 6 l=2Þ ð35Þ
ð1  b2 Þ þ qð1  bÞ2 2
G32 1
G32 ¼ ¼ ð33Þ where q and b (0 < b 6 1) are given in Eqs. (4) and (5). To
Gs q 2 use Eq. (35), the condition sm P 0 should be satisfied,
which is the same as that obtained by Kim and Christensen which subsequently requires
(2000) for regular hexagonal honeycombs with cell walls pffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffi
t1 9 þ 10q þ q2  1  q
of uniform thickness. Fig. 5 shows clearly that honeycombs b¼ 6 ð36Þ
t2 4
with cell wall of uniform thickness have the largest out-of-
plane shear modulus. The smaller the material distribution Substituting (35) into (34), the maximum compressive
parameter b, the smaller the shear modulus G32 . stress rmax in the cell wall can be obtained as

3½1 þ b þ qð1  bÞ2 rxx


3. Plastic properties rcmax ¼  2 ð37Þ
16bð1  bÞ q
3.1. In-plane initial yield strength As the plane-strain bending of the cell walls results in a
stress component in the direction normal to the xy plane
When a metallic honeycomb is just about to yield, the (i.e. in the z direction) and the axial compression of the cell
deformation is small. In the same way as that employed walls does not, the exact closed-form result of the maxi-
by Gibson and Ashby (1997), to simplify the analysis, we mum von Mises effective stress within the cell wall will
use the undeformed configuration of the honeycomb to be very lengthy (although it is not too difficult to obtain)
compute the internal forces in the cell walls and hence to and it is not convenient to use in practical applications.
calculate the initial yield strengths. To simplify the calculation and to make the final result
Author's personal copy

282 H.X. Zhu, C.Y. Chen / Mechanics of Materials 43 (2011) 276–286

more convenient to use, it is assumed that both the bend-


ing and the axial compression of the cell wall are in plane-
strain deformation state and the maximum effective von
Mises stress in the cell wall surface is therefore
qffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffi
3½1 þ b þ qð1  bÞ2 rxx
reff
max ¼  2 1  v s þ v 2s 6 rys ð38Þ
16bð1  bÞ q
where, ms is the Poisson ratio for elastic deformation of the
solid material from which the honeycomb is made.
For a low density honeycomb (i.e. when q ¼ p2ffiffi3 tl0 is
small), the amplitude of the axial compressive stress with-
in the cell walls is usually much smaller than the maxi-
mum bending stress. The error of Eq. (38) resulted from
the plane-strain deformation assumption is expected to
be small. When a honeycomb is compressed in the x direc-
tion, the initial yield stress, which is normalised by Fig. 6a. Predicted dimensionless initial yield strength vs. material
q2 rys
pffiffiffiffiffiffiffiffiffiffiffiffiffiffi, can be obtained as distribution parameter b.
1v s þv s 2
pffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffi
riyxx 1  v s þ v 2s 16bð1  bÞ
riyxx ¼ ¼ ð39Þ
q2 rys 3½1 þ b þ qð1  bÞ2
It is noted that Eq. (39) applies only when the condition Eq.
(36) is satisfied. Otherwise, if Eq. (36) is not satisfied and
hence sm is smaller than 0, it should be treated as 0 because
0 6 s 6 l/2. This means initial yielding takes place in the
surface of the cell walls at the junction. In this case, the
normalised initial yield stress can be obtained as
pffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffi
riyxx 1  v s þ v 2s 4
riyxx ¼ ¼ ð40Þ
q2 rys 3½ð1 þ bÞ2 þ qð1 þ bÞ
For different fixed values of honeycomb relative density
q, the relationship between the normalised initial yield
strength riy xx (which is predicted by the combination of
Eqs. (39) and (40)) and the material distribution parameter
b is plotted in Fig. 6a for b from 0.2 to 1.0.
Taking cell wall bending as plane-strain deformation Fig. 6b. Comparison between the predicted dimensionless initial yield
and both axial compression and transverse shear as strength riyxx and the precise result for honeycombs with different values

plane-stress deformation, we have precisely performed of relative density and material distribution parameter b.
numerical calculations and verified that for regular honey-
combs with relative density q 6 0.3 and material distribu- pffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffi
riy 1  v þ v 2s 48bð1  bÞ
tion parameter 1.0 P b = t1/t2 P 0.26 or for honeycombs r ¼ yy 2 s
iy
yy ¼ ð41Þ
with q 6 0.2and 1.0 P b = t1/t2 P 0.2, initial yielding al- q rys ½3ð1 þ bÞ þ qð1  bÞ2
ways takes place in the surface of the cell walls. Therefore,
To use Eq. (41), the material distribution parameter b
Eqs. (39) and (40) serves as the approximate limit for initial
should satisfy the condition
yielding when the honeycomb is compressed in the x
pffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffi
direction. The precisely obtained numerical results for 81 þ 30q þ q2  3  q
the initial yield strength of honeycombs with relative den- b6 ð42Þ
12
sities of q = 0.3 and q 6 0.2 are compared with those pre-
Otherwise, if sm 6 0, it should be treated as 0 and initial
dicted by the combination of Eqs. (39) and (40) as shown
yielding takes place in the surface of the inclined cell walls
in Fig. 6b . Fig. 6b shows that although the combination
at the junction. The corresponding normalised initial yield
of Eqs. (39) and (40) slightly overestimates the initial yield
stress is therefore given by
strength in the x direction of metallic honeycombs, it pro-
pffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffi
vides a very good prediction; the smaller the honeycomb riy 1  v þ v 2s 4
relative density, the more precise the prediction. It is noted r ¼ yy 2 s
iy
yy ¼ ð43Þ
q rys 2
3ð1 þ bÞ þ qð1 þ bÞ
that to obtain the exact results in Fig. 6b, the Poisson ratio
of the cell wall material has been taken as ms = 0.3. Comparing the predicted results of Eqs. (39) and (40) to
In a similar manner, when the honeycomb is com- those of (41) and (43), the normalised initial yield stress
pressed by a uniform stress ryy in the y direction, the initial in the x direction riy xx is found to be slightly smaller than
q2 rys
yield stress, which is normalised by pffiffiffiffiffiffiffiffiffiffiffiffiffiffi
1v s þv 2s
, can be ob- riyyy in the y direction. Therefore, Eqs. (39) and (40) can be
tained as used as the approximate limits of the initial yield strengths
of regular metallic honeycombs.
Author's personal copy

H.X. Zhu, C.Y. Chen / Mechanics of Materials 43 (2011) 276–286 283

Fig. 6a shows clearly that for honeycombs with differ-


ent values of relative density, the maximum in-plane ini-
tial yield strength appears when the material distribution
parameter b = t1/t2 is in the range from 0.3 to 0.4. It should
be noted that if b = t1/t2 is too small (say smaller than 0.2),
initial yielding may take place at the middle of the cell
walls (i.e. s = l/2) because of the combination of the trans-

verse shear stress smax ¼ 32  P cost160 and the axial compres-

sive stress rc ¼ P cost130 , as shown by the broken curves in
Fig. 6b.

3.2. In-plane full yield strength

As the honeycomb is assumed to be made of an elastic


and perfectly plastic metallic material, with the increase
Fig. 7. Stress distributions over the critical cell wall cross-section which
of the applied compressive stress rxx in the x direction, first undergoes fully plastic yielding.
the inclined cell walls gradually undergo elastic deforma-
tion, elastic-plastic deformation, and at a critical level of nation of axial compression and transverse shear (note that
the applied load rfyxx , full plastic collapse (or full yield) takes the bending moment is zero in this cross-section).
place in a particular cross-section of s ¼ h 2l while in other In the analysis to follow, the solid metallic material of
cross-sections the material undergoes partly plastic and the cell walls is assumed to be incompressible in plastic
partly elastic deformation. Here, a new parameter h is intro- state and hence the Poisson ratio is taken as 0.5. As men-
duced to specify the location where full plastic collapse first tioned above, plane-strain bending of the inclined cell walls
takes place and the value of h (0 6 h 6 1) is to be deter- is the dominant deformation mechanism. To simplify the
mined. h = 0 corresponds to the location at the cell wall analysis, we assume that the bending, axial compression
junction while h = 1 implies the middle of the cell wall. and transverse shear of the cell walls are all in plane-strain
In this part, we focus on the analysis of the amplitudes deformation state. This assumption is expected to be more
and the distributions of the stresses over the critical cell reasonable than the plane-stress deformation assumption,
wall cross-section where full plastic collapse takes place as employed by Gibson et al. (1982) and Gibson and Ashby
first. As the solid metallic material is assumed to be elastic (1997). The exact numerical results for the initial yielding
and perfectly plastic, when the critical cell wall cross- shown in Fig. 6b have already provided a good justification
section is just about to fully plastically yield, the shear for the adoption of the plane-strain deformation assump-
stress is assumed to be distributed uniformly over the tion. Based on the plane-strain deformation assumption,
cross-section (as shown in Fig. 7) in order to simplify the the von Mises yield criterion requires
analysis. It is noted that the assumption of a uniform shear qffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffi
rys 2rys
stress distribution over the plate cross-section is widely ½rc ðhl=2Þ2 þ 3½sðhl=2Þ2 ¼ pffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffi ¼ pffiffiffi
2 3
adopted for the plastic strength analysis of plate or shell 1  0:5 þ 0:5
structures in research papers. As the von Mises effective Substituting (44) and (45) into the above equation leads to
stress is a constant rys over the critical cross-section, if  
the shear stress is uniform, the amplitude of the normal l 2t 2 ½1  hð1  bÞ
rc h ¼ pffiffiffiqffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffi rys ð46Þ
stress, including both bending and axial compression, 2
3 t 22 ½1  hð1  bÞ2 þ a2 t 21
should also be a constant, as shown in Fig. 7. Based on
the assumption of uniform distributions of the shear and and
the normal stresses, the full yield load (or stress) rfy xx ap-
4at
plied in the x direction of the honeycomb is related to rfyxx ¼ pffiffiffi1 rc ðhl=2Þ
the amplitudes of the shear stress sðh 2l Þ and the axial com- 3 3l
pressive stress rc ðh 2l Þ in the critical cross-section of the in- 8at 1 t 2 ½1  hð1  bÞ
¼ qffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffi rys ð47Þ
clined cell walls by
9l t 22 ½1  hð1  bÞ2 þ a2 t21
 
l 3lrfy
xx 3lrfy
xx
s h ¼ ¼ ð44Þ Normalising the full yield stress in the x direction by
2 4t 4t 2 ½1  hð1  bÞ q2 rys
pffiffiffiffiffiffiffiffiffiffiffiffiffiffi ¼ p2ffiffi3 q2 rys , where ms = 0.5, Eq. (47) reduces to
1v s þv s 2

  pffiffiffi
l 3 3lrfy rfy
rc h ¼ xx
ð45Þ rfyxx ¼ pxx
ffiffiffi
2 4at 1 ð2= 3Þq2ys

where parameters a (0 < a 6 1) and h (0 6 h ¼ 2sl 6 1) are to 4ab½1  hð1  bÞ


¼ qffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffi ð48Þ
be determined. at1 is the area of the central part of the cell 3qð1 þ bÞ ½1  hð1  bÞ2 þ a2 b2
wall cross-section which resists the axial compression in
the direction of the inclined cell walls (as shown in According to the distribution of the bending stress over the
Fig. 7). If a = 1, it implies that the cross-section at the mid- critical cross-section of the inclined cell walls as shown in
dle of the inclined cell walls is fully yielded by the combi- Fig. 7, the moment of the bending stress to the neutral axis
Author's personal copy

284 H.X. Zhu, C.Y. Chen / Mechanics of Materials 43 (2011) 276–286

of the cell wall cross-section should be equal to that re- By varying the value of parameter h within the range of
sulted from the application of the external load (or stress) 0 6 h ¼ 2sl 6 1, the amplitude of the dimensionless full
rfyxx , which requires yield strength of the honeycomb in the y direction rfy yy
can be found as the minimum value of either Eq. (52) or
rc ðhl=2Þ
Mðhl=2Þ ¼ ½ðtðhl=2ÞÞ2  ðat 1 Þ2  (53), and the corresponding value hy specifies the location
4
2
of the critical cell wall cross-section where fully plastic
3l yield takes place first when the honeycomb is compressed
¼ ð1  hÞrfy
xx ð49Þ
8 in the y direction.
Substituting Eqs. (46) into (49) and normalising the full For regular metallic honeycombs made of an elastic and
yield stress in the x direction rfy perfectly plastic material and with different values of rela-
xx by pffiffi q2 rys , one has
2
3
tive density q, when it is uniaxially compressed in either
rfy the x direction or the y direction, the relationships between
rfyxx ¼ pxx
ffiffiffi
ð2= 3Þq2ys the dimensionless full plastic yield strength rfy fy
xx or ryy

2½1  hð1  bÞ3  2a2 b2 ½1  hð1  bÞ and the material distribution parameter b are presented
¼ qffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffi ð50Þ in Figs. 8a and 8b. They show clearly that material distribu-
ð1 þ bÞ2 ð1  hÞ ½1  hð1  bÞ2 þ a2 b2 tion b greatly affects the full yield strengths; and for differ-
ent values of the honeycomb relative density, the peak full
Equating Eqs. (48) and (50), parameter a can be deter-
yield strength appears when the material distribution
mined as
parameter is in the range from 0.3 to 0.45. It can be seen
qffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffi
from Fig. 8a that the dimensionless full yield strength of
ð1 þ b2 Þ1  h2 þ 9q2 ½1  hð1  bÞ2  ð1 þ bÞð1  hÞ
ax ¼ in the x direction rfy
xx depends significantly on the relative
3qb
density; the larger the relative density, the smaller the
ð51Þ
rfyxx . In contrast, Fig. 8b shows that the dimensionless full
Eq. (51) shows that ax depends upon the honeycomb rela-
yield strength in the y direction rfy yy is almost independent
tive density q and the cell wall material distribution
of the honeycomb relative density, and always larger than
parameter b. By varying the value of parameter h within
the range of 0 6 h ¼ 2sl 6 1, the amplitude of the dimen- the dimensionless full yield strength in the x direction rfy
xx .

sionless full yield strength of the honeycomb in the x direc- Therefore rfyxx shown in Fig. 8a can be taken as the dimen-
tion rfy
xx can be identified as the minimum value of either sionless full yield strength of regular metallic honeycombs.
Eq. (48) or (50) because they are equated by the applica- The dimensionless full yield strengths are not very sensi-
tion of (51) and the corresponding value hx specifies the tive to the honeycomb relative density q when b P 0.7
location of the critical cell wall cross-section where fully and the relative density is q 6 0.3. They drop sharply with
plastic yield takes place first when the honeycomb is com- the decrease of b when b 6 0.2. For honeycombs with uni-
pressed in the x direction.
form cell wall thickness (i.e. b = 1), both rfy fy
xx and ryy are
In a similar manner, if a regular honeycomb shown in
very close to the results of Gibson and Ashby (1997) except
Fig. 1a is compressed in the y direction, the full yield stress,
that their full yield strength is normalised by q2rys rather
which is again normalised by p2ffiffi3 q2 rys , of the cell wall 2
q rys
should satisfy the following two equations: than by pffiffiffiffiffiffiffiffiffiffiffiffiffiffi ¼ p2ffiffi3 q2 rys . Zhu (2007) has shown that
1v s þv s
2

rfy when a metallic honeycomb is in-plane crushed, 90% of


rfyyy ¼ pyy
ffiffiffi the external energy is absorbed by the plastic hinges of
ð2= 3Þq2ys the cell walls. Therefore, the energy absorption capacity
4ab½1  hð1  bÞ
¼ qffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffi ð52Þ
qð1 þ bÞ ½1  hð1  bÞ2 þ 9a2 b2

rfy
rfyyy ¼ pyy
ffiffiffi
ð2= 3Þq2ys
2½1  hð1  bÞ3  2a2 b2 ½1  hð1  bÞ
¼ qffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffi ð53Þ
ð1 þ bÞ2 ð1  hÞ ½1  hð1  bÞ2 þ 9a2 b2

Equating Eqs. (52) and (53), the parameter a is determined


as
qffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffi
ð1 þ bÞ2 ð1  hÞ2 þ q2 ½1  hð1  bÞ2  hð1 þ bÞ
ay ¼
qb
ð54Þ
Obviously, ay depends upon the honeycomb relative den- Fig. 8a. Dimensionless full yield compressive stress in the x direction rfy
xx
sity q and the cell wall material distribution parameter b. vs. material distribution parameter b.
Author's personal copy

H.X. Zhu, C.Y. Chen / Mechanics of Materials 43 (2011) 276–286 285

In consistency with Fig. 9a, Fig. 10a shows that when a


honeycomb is compressed in the x direction, ax is always
equal to 1 if the honeycomb relative density is q = 0.3
and the material distribution parameter b = 0.26 or smal-
ler. That means the middle of the inclined cell walls is
yielded by the combined effect of transverse shear and ax-
ial compression (note that the bending moment is zero in
this cross-section). Fig. 10b shows that ay is always smaller
than 1 if the honeycomb relative density is 0.3 or smaller. It
is noted that the material distribution parameter b = t1/t2
should be always larger than zero. When a regular honey-
comb is in-plane compressed in the y direction, elastic
buckling with junction rotation may take place if the yield
strain of the solid material, such as a polymer, is large (Zhu
and Mills, 2000). Comparing the predicted in-plane elastic
buckling strength in the y direction (Zhu and Mills, 2000)
with the plastic collapse strength of a metallic honeycomb
Fig. 8b. Dimensionless full yield compressive stress in the y direction rfy
yy
as shown in Fig. 8, it is concluded that elastic buckling is
vs. material distribution parameter b.
unlikely to happen before the occurrence of plastic collapse
of a metallic honeycomb depends greatly on the location for a honeycomb made of an elastic and perfectly plastic
and hence the thickness of the plastic hinges. metallic material if its relative density is larger than 0.01.
Fig. 9a shows that when a honeycomb is compressed in
the x direction, full plastic yielding always initiate at mid- 3.3. Out-of-plane yield strength
dle of the inclined cell walls (i.e. hx = 1) if the honeycomb
relative density is q = 0.3 and the material distribution As the honeycomb is treated as a material rather than a
parameter b = 0.26 or smaller. In contrast, if the material structure, we do not consider the buckling strength of the
distribution parameter b = 0.54 or larger, full yield always honeycomb in this paper. Obviously, the compressive (or
takes place at the cell wall junction (i.e. hx = 0) irrespective tensile) yield strength in the z direction is proportional to
of the honeycomb relative density. Fig. 8a suggests that for the honeycomb relative density q and to the yield strength
honeycombs with fixed relative density of 0.3 or smaller, rys of the solid material from which the honeycomb is
the largest full yield strength always appears when the made. The dimensionless yield strength, which is norma-
material distribution parameter b is in the range from 0.3 lised by qrys, is given by
to 0.45, and full plastic collapse always takes place at a ryzz
location s = hxl/2 where 0 < hx < 1, which is neither at the ryzz ¼ ¼1 ð55Þ
qrys
middle of the inclined cell wall nor at the cell wall junction.
Fig. 9b shows that when a regular honeycomb with a rela- The shear yield strength of the honeycomb in the yz plane,
tive density of 0.3 or smaller is compressed in the y direc- which is normalised by qsys, can be easily obtained as
tion, full yield collapse never takes place at the middle of sy32 b
the cell walls or at the cell wall junction if the material dis- sy32 ¼ ¼ ð56Þ
qsys 1 þ b
tribution parameter b 6 0.49, but always takes place at the
cell wall junction (i.e. hy = 0) if b P 0.51.

Fig. 9a. Relationship between the location of the critical cross-section hx Fig. 9b. Relationship between the location of the critical cross-section hy
(i.e. hx = 2s/l) and material distribution parameter b. (i.e. hy = 2s/l) and material distribution parameter b.
Author's personal copy

286 H.X. Zhu, C.Y. Chen / Mechanics of Materials 43 (2011) 276–286

4. Conclusion

The combined effects of honeycomb relative density


and material distribution on all the five independent elas-
tic constants and the initial and full yield strengths of a
regular metallic honeycomb have been obtained in this pa-
per. It has been demonstrated that both honeycomb rela-
tive density q and the material distribution parameter b
can significantly affect the mechanical properties of a reg-
ular honeycomb. In applications, a honeycomb is de-
manded to have a good combination of the in-plane and
out-of-plane strength and stiffness. Based on the obtained
results in this paper, it is recommended that for a regular
metallic honeycomb with a relative density of 0.3 or smal-
ler, the material distribution parameter b = t1/t2 should be
in the range from 0.3 to 0.45.
Fig. 10a. Relationship between ax and material distribution parameter b.

References

Chen, C., Lu, T.J., Fleck, N.A., 1999. Effect of imperfections on the yielding
of two-dimensional foams. J. Mech. Phys. Solids 49, 2245–2271.
Gibson, L.J., Ashby, M.F., 1997. Cellular Solids: Structure and Properties,
second ed. Cambridge University Press, Cambridge, UK.
Gibson, L.J., Ashby, M.F., Schajer, G.S., Robertson, C.I., 1982. The mechanics
of two-dimensional cellular materials. Proc. Roy. Soc. Lond. A 382, 25–
42.
Grediac, M., 1993. A finite element study of the transverse shear in
honeycomb cores. Int. J. Solids Struct. 30, 1777–1788.
Kelsey, S., Gellatly, R.A., Clark, B.W., 1958. The shear modulus of foil
honeycomb cores. Aircraft Eng. 30, 294–302.
Kim, B., Christensen, R.M., 2000. Basic two-dimensional core types for
sandwich structures. Int. J. Mech. Sci. 42, 657–676.
Masters, I.G., Evans, K.E., 1996. Models for the elastic deformation of
honeycombs. Compos. Struct. 35, 403–422.
Meraghni, F., Desrumaux, F., Benzeggagh, M.L., 1999. Mechanical
behaviour of cellular core for structural sandwich panels. Composite
A 30, 767–779.
Nye, J.F., 1985. Physical Properties of Crystals. Clarendon Press, Oxford.
Pan, S.D., Wu, L.Z., Sun, Y.G., Zhou, Z.G., Qu, J.L., 2006. Longitudinal shear
Fig. 10b. Relationship between ay and material distribution parameter b. strength and failure process of honeycomb cores. Compos. Struct. 72,
42–46.
Shi, G., Tong, P., 1995. Equivalent transverse shear stiffness of honeycomb
cores. Int. J. Solids Struct. 32, 1383–1393.
where sys is the shear yield strength of the solid material. Silva, M.J., Hayes, W.C., Gibson, L.J., 1995. The effects of non-periodic
Similarly, the normalised shear yield strength of the hon- microstructure on the elastic properties of two-dimensional cellular
eycomb in the xz plane is given by solids. Int. J. Mech. Sci. 37, 1161–1177.
Simone, A.E., Gibson, L.J., 1998. Effects of solid distribution on the stiffness
and strength of metallic foams. Acta Mater. 46, 2139–2150.
Timoshenko, S.P., Goodier, J.N., 1982. Theory of Elasticity. McGraw-Hill
sy31 2 b International Book Company.
sy31 ¼ ¼ pffiffiffi  ð57Þ
qsys 3 1 þ b Warren, W.E., Kraynik, A.M., 1987. Foam mechanics: the linear elastic
response of two-dimensional spatially periodic cellular materials.
Mech. Mater. 6, 27–37.
Xu, X.F., Qiao, P., 2002. Homogenized elastic properties of honeycomb
Eqs. (56) and (57) show clearly that the shear yield sandwich with skin effect. Int. J. Solids Struct. 39, 2153–2188.
strength in the xz plane is larger than that in the yz plane. Zhu, H.X., 2007. Large deformation pure bending of an elastic plastic
Therefore, Eq. (56) can be taken as the out-of-plane shear power-law-hardening wide plate: analysis and application. Int. J.
Mech. Sci. 49, 500–514.
yield strength of a regular honeycomb with cell walls of Zhu, H.X., 2010. Size-dependent elastic properties of micro- and nano-
linear thickness distribution. honeycombs. J. Mech. Phys. Solids 58, 679–696.
Eq. (56) shows that the smaller the material distribu- Zhu, H.X., Mills, N.J., 2000. The in-plane non-linear compression of regular
honeycombs. Int. J. Solids Struct. 37, 1931–1949.
tion parameter b, the smaller the out-of-plane yield Zhu, H.X., Hobdell, J.R., Windle, A.H., 2001. Effects of cell irregularity on
strength. When b tends to zero, the out-of-plane yield the elastic properties of 2D Voronoi honeycombs. J. Mech. Phys. Solids
strength also tends to zero. Therefore b cannot be too 49, 857–870.
Zhu, H.X., Thorpe, S.M., Windle, A.H., 2006. The effect of cell irregularity
small. When b = 1, the dimensionless out-of-plane shear
on the high strain compression of 2D Voronoi honeycombs. Int. J.
yield strength sy32 ¼ 12, which is the same as the result of Solids Struct. 43, 1061–1078.
(Kim and Christensen, 2000).

View publication stats

You might also like