Professional Documents
Culture Documents
In Plane Compressive Response and Crushing of Foam Filled Aluminum Honeycombs
In Plane Compressive Response and Crushing of Foam Filled Aluminum Honeycombs
COMPOSITE
Article M AT E R I A L S
Journal of Composite Materials
0(0) 1–14
! The Author(s) 2014
In plane compressive response and Reprints and permissions:
sagepub.co.uk/journalsPermissions.nav
crushing of foam filled aluminum DOI: 10.1177/0021998314561069
jcm.sagepub.com
honeycombs
Abstract
In this paper, the influence of foam filling of aluminum honeycomb core on its in-plane crushing properties is investigated.
An aluminum honeycomb core and a polyurethane foam with densities of 65, 90, and 145 kg/m3 were used to produce
foam filled honeycomb panels, and then experimental quasi-static compression tests were performed. Moreover, finite
element model, based on the conducted tests, was developed. In the finite element analyses, three different polyurethane
foams were used to fill three different honeycomb cores. The effects of foam filling of aluminum honeycomb core on its
in-plane mechanical properties (such as mean crushing strength, absorbed energy, and specific absorbed energy) were
analyzed experimentally and numerically. The results showed that the foam filling of honeycomb core can increase the in
plane crushing strength up to 208 times, and its specific absorbed energy up to 20 times. However, it was found that the
effect of foam filling decreases in heavier honeycombs, producing an increment of the above mentioned properties only
up to 36 and 6 times, respectively.
Keywords
Polyurethane foam, crushing strength, specific absorbed energy, finite element method, foam filled honeycomb
Downloaded from jcm.sagepub.com at Gebze Yuksek Teknoloji Enstitu on December 20, 2014
2 Journal of Composite Materials 0(0)
syntactic foams with different volume fractions of filled honeycomb structures as energy-absorbing mater-
micro balloons. Uniaxial compression tests were car- ials. Implementing rigid inclusions in honeycomb struc-
ried out under quasi-static conditions. They have ture decreases significantly its densification strain.
found that filling aluminum honeycomb by syntactic Moreover, the randomly filling honeycombs with rigid
foam can improve elastic modulus and plateau stress inclusions results in severe oscillation in stress–strain
values by 26–31% and 36–39% when compared to the response, especially at the plateau stress region. On
syntactic foam with the same volume fraction of micro the other hand, in the present study, honeycomb cells
balloons. This research gave useful information of were filled with rigid polyurethane foam. Uniformly
foam-filling technique in order to enhance the required filled honeycombs lead to stable variation of stress–
mechanical properties of honeycomb. Syntactic foams strain response and avoid abrupt changes. Moreover,
are multifunctional structural foams with closed-cell due to the cellular structure of rigid polyurethane
microstructure. These foams are relatively heavier in foams, the filled honeycomb cells deform to absorb
comparison with other polymeric foams. If the honey- energy in spite of rigid inclusions which do not partici-
comb core is filled with heavy foams, this produces an pate in energy absorption.
undesirable overweight of the all structure. Therefore,
some researchers have worked on the effects of filling
honeycomb cores with lightweight polyurethane foams. Experimental tests
However, their researches mostly heed to the determin-
ation of out of plane crushing properties of honeycomb
Materials
cores.25 AlaviNia and Sadeghi26 investigated the
Honeycomb core. Commercially available aluminum
effects of strain (in low strain rate regime) on the out
honeycomb core (made of Al 5052-H32; manufactured
of plane compressive mechanical behavior of bare
by Hexcel Corporation, USA) was used for the speci-
and polyurethane foam filled honeycomb structures.
mens. The specifications of the honeycomb sample are
They have found that bare panels are more sensitive
listed in Table 1 and shown in Figure 1.
to the strain rate than the foam filled ones. The incre-
ment of the plateau stress, induced by an increase in the
strain rate, is much more considerable in bare panels PU foam. The apparent density of polyurethane foam
than that of foam filled ones. Mozafari et al.27 studied which was selected for filling of honeycomb cores was
the application of polyurethane foam filled honeycomb
sandwich panels in Korean tilting train express (TTX).
They investigated the out of plane crushing behavior Table 1. Specifications of honeycomb sample.
and local stiffness of bare and foam filled honeycombs.
Honeycomb
They found that the local stiffness and out of plane
material S (mm) h (mm) Panel size (mm) b (mm)
crushing behavior could increase significantly, using
the polyurethane foam filled honeycomb as core of Al 5052-H32 9 0.14 50 50 35
TTX sandwich panel.
In this paper, the in plane crushing behavior of foam
filled honeycomb is investigated experimentally and
numerically. An aluminum honeycomb core and a
polyurethane foam were used to prepare the foam
filled honeycomb, then experimental quasi-static com-
pression tests were carried out. A finite element model,
based on conducted tests, was developed, then three
different polyurethane foams were used to fill three
types of honeycomb cores with different ratios between
the thickness and the length of the cell wall. Finally, the
effects of foam filling of aluminum honeycomb core on
its in-plane mechanical properties (such as mean crush-
ing strength, absorbed energy, and specific absorbed
energy) were studied. Nakamoto et al.28 studied the
in plane impact behavior of honeycomb structures
randomly filled with inclusions. They used FE method
to clarify the effect of rigid inclusions on the deform-
ation process, mean stress, densification strain, and
absorbed energy. They finally proposed randomly Figure 1. Aluminum honeycomb core with its dimension in mm.
Downloaded from jcm.sagepub.com at Gebze Yuksek Teknoloji Enstitu on December 20, 2014
Mozafari et al. 3
Downloaded from jcm.sagepub.com at Gebze Yuksek Teknoloji Enstitu on December 20, 2014
4 Journal of Composite Materials 0(0)
According to this material model, the yield where B defines the size of the yield ellipse, p and q are
function is: the pressure and the Von Mises stress, respectively. The
yield surface represents the Von Mises circle in the
pffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffi
f¼ q2 þ 2 p2 B ð1Þ deviatoric stress plane and the flow potential is an
ellipse centered at the origin. The typical yield surface
pffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffi
g¼ q2 þ 2 p2 ð2Þ and flow potential are shown in Figure 7. The shape
factor can be computed using the initial yield stress in
Downloaded from jcm.sagepub.com at Gebze Yuksek Teknoloji Enstitu on December 20, 2014
Mozafari et al. 5
uniaxial compression, 0C and the initial yield stress in close to a circle in the p q stress plane, which indicates
hydrostatic compression, pC0 using the relation: that the value of is approximately equal to unity.
Parameter (the shape of the flow potential ellipse on
2k 0C the p q stress plane) in equation (4) for the isotropic
¼ pffiffiffiffiffiffiffiffiffiffiffi with k ¼ ð3Þ hardening model is defined as:
9k pC
0
sffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffi
where k is the yield stress ratio. For a valid surface, the 3 1 2#p
¼ pffiffiffi ð4Þ
choice of the yield stress ratio must be in range 0–3.31 2 1 þ #p
For many low-density foams, the initial yield surface is
Figure 4. SANTAM (STM-150) apparatus that is used for quasi- 68,000 80 0.3 2680
static compression tests.
Figure 5. Finite element model of compression test. (a) PU foam, (b) bare honeycomb, and (c) foam filled honeycomb core.
Downloaded from jcm.sagepub.com at Gebze Yuksek Teknoloji Enstitu on December 20, 2014
6 Journal of Composite Materials 0(0)
These could be the initial yield stress in uniaxial com- movable plate was set just 1 mm above the honeycomb
pression, 0C initial yield stress in hydrostatic compres- core to have a better stability. Movable plate was con-
sion, pC C
0 . Parameter 0 is the stress of point B in straint to move with a constant velocity to crush the
Figure 2. Here the value of k is assuming to unity in honeycomb core under quasi-static conditions. The
equation (3). The corresponding value of can be analysis duration was defined according to the time
determined and then pC 0 is estimated. The line B-C of required for the compressed panel to reach its densifi-
uniaxial compression test is sufficient to define the evo- cation point. Then analyses were carried out. By con-
lution of yield surface (Figure 2). The hardening law sidering effective contact area and initial distance,
defines the value of the axial plastic strain.31 The stress–strain curves were obtained.
obtained data, used in this study, are shown in Table 3.
The general surface-to-surface contact algorithm
was defined on the unfilled honeycomb model in
order to avoid probable penetration between cell Table 3. Crushable foam mechanical properties used in this
simulation.
walls. Abaqus/Explicit automatically defines an all-
inclusive surface that is convenient for prescribing the Elasticity Plasticity
contact domain. In the case of foam filled honeycomb,
tie constraint was used in order to stick foam cell faces f (kg/m3) Foam type E (MPa) # k #p
to the adjacent honeycomb cell walls. 65 F1 5.5 0 1 0
The movable plate was constrained to move only in 90 F2 10.9 0 1 0
the Y direction. The stationary plate was constrained in
145 F3 27.2 0 1 0
its degrees of freedom. The initial position of the
Figure 6. Mesh sensitivity analysis of honeycomb core (HC1) and PU foam (F1).
Figure 7. Crushable foam model with isotropic hardening: yield surface and flow potential in the p–q stress plane.28
Downloaded from jcm.sagepub.com at Gebze Yuksek Teknoloji Enstitu on December 20, 2014
Mozafari et al. 7
m (MPa)
Downloaded from jcm.sagepub.com at Gebze Yuksek Teknoloji Enstitu on December 20, 2014
8 Journal of Composite Materials 0(0)
nonlinear portions are approximately the same. In foam filled cores, the effect of the foam filled
Furthermore, the calculated mean crushing strength panels on the mean crushing strength reduces with the
in Table 4 shows that FE model is in good agreement decreasing the foam density. Furthermore, the lower
with experimental results; the largest difference is for the foam density, the greater the differences between
foam filled case and is <6%. changes in crushing strength of different cores. It is
The mean crushing strength values of bare honey- seen from Table 6 that for all cores with any foam
combs which are obtained from FE model are listed in density, the mean crushing strength of foam filled
Table 5. In this table, the FE results are compared with core is always greater than the sum of mean crushing
Gibson and Ashby7 formulation which is written as strengths of bare honeycomb and foam when con-
sidered separately. Difference between these quantities
2
2 h is proportional to the percentage of increase of mean
m,HC ¼ ys ð6Þ crushing strength of core due to foam filling.
3 l
Increase of m is calculated by using equation (7)
m,HCF m,HCþF
Mean crushing strength values for foam filled cores increase of m ð %Þ ¼ 100 ð7Þ
m,HCþF
are listed in Table 6. It is seen that when a honeycomb
is filled with specific foam, the lower honeycomb dens- where m,HCþF ¼ m,HC þ m,F.
ity, the greater the increase in mean crushing strength
of core. For example, when the honeycombs are filled
with F1 foam, it is seen that for HC1F1 core, whose
density is equal to 41.46 kg/m3, the mean crushing
strength increases by about 5900%, whereas this
increase for HC2F1 (with 83.24 kg/m3 density) and
HC3F1 (with 117.59 kg/m3 density) is about 2011%
and 1877%, respectively.
m,HC (MPa)
Table 6. Comparison of mean crushing strengths of foam filled samples with bare honeycombs and bare foams:
finite element results.
Downloaded from jcm.sagepub.com at Gebze Yuksek Teknoloji Enstitu on December 20, 2014
Mozafari et al. 9
This difference could be attributed to the interaction It can be seen that there is a linear relation between
between cell walls and foam.25 Another interesting the mean crushing strength of foam filled core and the
result is in the relation between mean crushing strength sum of the mean crushing strengths of bare honeycomb
of foam filled cores and the sum of mean crushing and foam. Approximate ratio of these two quantities in
strengths of bare core and foam (Figure 10). all of the cores was determined by calculating the slope
Figure 11. In-plane crushing of HC2 and HC2F1 cores: experimental test.
Downloaded from jcm.sagepub.com at Gebze Yuksek Teknoloji Enstitu on December 20, 2014
10 Journal of Composite Materials 0(0)
of each trend line. It can be seen that the ratio is nearly foam-filled cores is approximately the same, and this
equal to 1.2 in all foam filled cores. feature is independent of foam type (Figure 13).
Increase in
Core f Eabsorbed absorbed SAE
type (kg/m3) (kN mm) energy (%) (kJ/kg)
Downloaded from jcm.sagepub.com at Gebze Yuksek Teknoloji Enstitu on December 20, 2014
Mozafari et al. 11
value is 980% and 683% for HC2F3 and HC3F3, was expected, the use of heavier PU foam as a filler of
respectively (Figure 14). honeycomb core results in higher absorbed energy. For
Figures 14 and 15 show the effect of foam filling of example, absorbed energy of HC1F1 is 17.215 kNmm,
bare honeycomb cores associated with PU foam type whereas this value increases up to 77.63 kNmm in
on their absorbed energy and SAE, respectively. As it HC1F3, which means that the absorbed energy of
Downloaded from jcm.sagepub.com at Gebze Yuksek Teknoloji Enstitu on December 20, 2014
12 Journal of Composite Materials 0(0)
HC3F3
140 avoid premature failure. Foam filling is considered as
an effective method to strengthening of honeycomb
HC2F3
120 cores. However, previous studies are mostly associated
E absorbed (KNmm)
HC1F3
100
HC3F2
study revealed the benefits of foam filling of honey-
HC3F1
80 comb core to reinforcing the core under in plane crush-
HC2F2
60 ing load. It was concluded that using lightweight
HC1F2
HC2F1
0
Conclusions
F1 F2 F3
Foam type As reported in literature, honeycomb cores are widely
used as core of sandwich panels in lightweight struc-
Figure 14. Absorbed energy of foam filled cores. tures. These cores are exposed to crushing and failure
under in plane loading conditions. Therefore, reinforce-
ment of honeycomb cores is necessary to avoid prema-
HC3F3
HC2F2
HC1F2
3.5
comb core to reinforcing the core under in plane crush-
HC2F1
3
HC1F1
Downloaded from jcm.sagepub.com at Gebze Yuksek Teknoloji Enstitu on December 20, 2014
Mozafari et al. 13
. foam filling of honeycomb core has no effect on its 14. Kim JS, Jeong JC and Lee SJ. Numerical and experimen-
deformation pattern. However, in foam filled cores, tal studies on the deformational behavior a composite
local densification of honeycomb cells decreases; train carbody of the Korean tilting train. Compos Struct
. calculated mean crushing strength shows that FE 2007; 81: 168–175.
model is in good agreement with experimental 15. Kim JS, Jeong JC, Lee SJ, et al. Structural safety evalu-
ation of the hybrid composite bodyshell for Korean tilt-
results and the largest difference is for foam filled
ing train by a whole body test. Adv Compos Mater Struct
case and is <6%.
2007; 334–335: 313–316.
16. Koronis G, Silva A and Fontul M. Green composites: a
Conflict of interest review of adequate materials for automotive applications.
Composites: Part B 2013; 44: 120–127.
None declared. 17. Liu Q, Lin Y, Zong Z, et al. Lightweight design of carbon
twill weave fabric composite body structure for electric
Funding vehicle. Compos Struct 2013; 97: 231–238.
This research received no specific grant from any funding 18. He M and Hu W. A study on composite honeycomb
agency in the public, commercial, or not-for-profit sectors. sandwich panel structure. Mater Des 2008; 29: 709–713.
19. Stocchia A, Colabellab L, Cisilinob A, et al.
References Manufacturing and testing of a sandwich panel honey-
comb core reinforced with natural-fiber fabrics. Mater
1. Bai Z, Guo H, Jiang B, et al. A study on the mean crush- Des 2014; 55: 394–403.
ing strength of hexagonal multi-cell thin-walled struc- 20. Crupi V, Epasto G and Guglielmino E. Comparison of
tures. Thin-Walled Struct 2014; 80: 38–45.
aluminum sandwiches for lightweight ship structures:
2. Wierzbicki T. Crushing analysis of metal honeycombs.
honeycomb vs. foam. Mar Struct 2013; 1: 74–96.
Int J Impact Eng 1983; 2: 157–174.
21. Hua L, Youa F and Yub T. Analyses on the dynamic
3. Santosa S and Wierzbicki T. Crash behavior of box col-
strength of honeycombs under the y-directional crushing.
umns filled with aluminum honeycomb or foam. Comput
Mater Des 2014; 53: 293–301.
Struct 1998; 68: 343–367.
22. Hua L, Youa F and Yu T. Effect of cell-wall angle on the
4. Crupi V, Epasto G, Guglielmino E, et al. Computed
in plane crushing behavior of hexagonal honeycombs.
tomography-based reconstruction and finite element
Mater Des 2013; 46: 511–523.
modelling of honeycomb sandwiches under low-velocity
23. Bitzer T. Honeycomb technology, materials, design, man-
impacts. J Sandwich Struct Mater 2014; 16: 377–397.
ufacturing, application and testing. Dordrecht: Springer
5. Aaron Jeyasingh V. Finite element analysis of drop test
Science þ Business Media, 1997.
equipment for nose landing gear configuration and its appli-
24. Jhaver R and Tippur H. Characterization and
cation to aircraft crashworthiness and occupant safety.
modeling of compression behavior of syntactic foam
Master Thesis. Wichita State University, Wichita KS,
USA, 2001. filled honeycombs. J Reinf Plast Compos 2010; 29:
6. Wierzbicki T and Mohr D. Crush response of double- 3185–3196.
walled sandwich columns with a honeycomb core. 25. AlaviNia A and Sadeghi MZ. The effects of foam filling
Sandwich Constr 2000; 5: 607–616. on compressive response of hexagonal cell aluminum
7. Gibson L and Ashby M. Cellular solids, structures and honeycombs under axial loading-experimental study.
properties. Pergamon Press, 1988, pp.110–112. Mater Des 2010; 31: 1216–1230.
8. Paik J, Thayamballi AK and Kim GS. The strength char- 26. AlaviNia A and Sadeghi MZ. An experimental investiga-
acteristics of aluminum honeycomb sandwich panels. tion on the effect of strain rate on the behaviour of bare
Thin-Walled Struct 1999; 35: 205–231. and foam filled aluminium honeycombs. Mater Des 2013;
9. Chou C. Honeycomb materials models for simulating 52: 748–756.
responses of foams. Report, Safety and Biomechanics 27. Mozafari H, Khatami S and Molatefi H. Out of plane
CAE Department, MD-48, AEC, Ford Motor crushing and local stiffness determination of proposed
Company, USA. foam filled sandwich panel for Korean Tilting Train
10. We E and Jiang W. Axial crush of metallic honeycombs. eXpress – numerical study. J Mater Des 2014.
Int J Impact Eng 1997; 19: 439–456. 28. Nakamoto H, Adachi T and Wakako Araki W. In plane
11. Goldsmith W and Sackman J. An experimental study of impact behavior of honeycomb structures filled with lin-
energy absorption in impact on sandwich plates. Int J early arranged inclusions. Int J Impact Eng 2009; 36:
Impact Eng 1992; 12: 241–262. 1019–1026.
12. Bandak M and Bitzer T. Honeycomb, a lightweight 29. Ali Ghamarian A, Zarei HR and Abadi MT.
energy absorbing material. 22nd International Society Experimental and numerical crashworthiness investiga-
for the Advancement of Material and Process tion of empty and foam filled end-capped conical tubes.
Engineering Technical Conference, pp. 1250–1262. Thin-Walled Struct 2011; 49: 1312–1319.
13. Jang BW, Lee JR, Park SO, et al. A health management 30. Manual of ABAQUS finite element analysis software
algorithm for composite train carbody based on FEM/ package. Analysis User’s Guide. Explicit dynamic
FBG hybrid method. Compos Struct 2010; 92: 1019–1026. analysis, version 6.11, 2011.
Downloaded from jcm.sagepub.com at Gebze Yuksek Teknoloji Enstitu on December 20, 2014
14 Journal of Composite Materials 0(0)
31. Deshpandeh VS and Fleck NA. Multi-axial yield behav- S honeycomb cell size (mm)
ior of polymer foams. Acta Materialia 2001; 49: SAE specific absorbed energy (kJ/kg)
1859–1866. m core mean crushing strength (MPa)
32. Tong-xi YU, Zi-yang G and Ling-ling H. Dynamic col- m,HC honeycomb mean crushing strength
lapse of honeycombs under in plane compressions.
(MPa)
J Ningbo Univ 2012; 25: 53–59.
m,HCF foam filled honeycomb mean crushing
strength (MPa)
m,HCþF sum of mean crushing strengths of
Appendix honeycomb and foam (MPa)
"D densification strain
Notation ys material yield stress (MPa)
C core density (kg/m3)
f foam density (kg/m3)
b panel thickness (mm)
HC honeycomb core density (kg/m3)
E elasticity modulus (MPa)
s material density (kg/m3)
Eabsorbed absorbed energy of core (kN mm)
h honeycomb cell wall thickness (mm)
Downloaded from jcm.sagepub.com at Gebze Yuksek Teknoloji Enstitu on December 20, 2014