Available online at www.sciencedirect.

com

Procedia Engineering 10 (2011) 2324–2329

ICM11

Mechanical behaviors of carbon fiber reinforced composite

truss core sandwich structures
Li MA*, Ming LI, Bing WANG and Lin-Zhi WU
Center for Composite Materials, Harbin Institute of Technology, Harbin 150001, China

Abstract

Sandwich panels are widely used in lightweight construction especially in aerospace industries because of their high
specific strength and stiffness. Analytical and experimental investigations are presented to study the response and
failure of pyramidal truss core sandwich panel made of carbon fiber composite under out-of-plane, axial compression
and three-point-bending. Pyramidal truss core sandwich panels were fabricated using a hot-press technique. The
responses of the sandwich panels were measured up to failure. The failure mechanisms of such sandwich structure
were also considered in this study.
© 2011 Published by Elsevier Ltd. Open access under CC BY-NC-ND license.
Selection and peer-review under responsibility of ICM11

Key Words: Composites; sandwich structure; mechanical behavior; failure

1. Introduction

Sandwich panels are widely used in lightweight construction especially in aerospace industries because
of their high specific strength and stiffness. Unlike traditionally made using stochastic cores such as
aluminum alloy foam or micro-architecture lattice materials such as hexagonal honeycombs, recently
three-dimensional periodic cores, such as pyramidal [1], tetrahedral [2] and octet-truss core [3] are
recognized as attractive candidates for building multifunctional ultra-light structures due to their open-cell
structures and high nodal connectivity. Metal lattice truss core sandwich structures have been fabricated
earlier and demonstrate superior mechanical response. As known, composite materials of continuous
carbon fiber reinforced resin, with a low density and a tensile strength far superior than lightweight metals,
have a high uniaxial specific stiffness and strength. Thus the trusses or webs of a lattice structure made

*
Corresponding author. Tel.: +86-451-86402739; fax: +86-451-86402386
E-mail address: mali@hit.edu.cn.

1877-7058 © 2011 Published by Elsevier Ltd. Open access under CC BY-NC-ND license.
Selection and peer-review under responsibility of ICM11 doi:10.1016/j.proeng.2011.04.383
 Li MA et al. / Procedia Engineering 10 (2011) 2324–2329 2325

from composite will be lighter and stronger than the existing lattice structures made of metals [4~6]. In
this study, carbon fiber composite pyramidal truss core sandwich structure was fabricated by using the
molding hot-press technique. Analytical and experimental investigation has been carried out to explore the
mechanical response of carbon fiber composite truss core sandwich panel. The failure mechanisms of
such sandwich structure were also considered in this study.

2. Materials and Manufacturing

A hot compression molding method was adopted to fabricate the pyramidal truss core sandwich
structure. The detailed fabricating process is as follows [5]: (1) The mold surfaces are cleaned with
acetone and coated with a mold release agent, and then the molds are assembled in order. (2) The upper
face of the molds is covered with m plies of carbon/epoxy (T700/3234) prepreg, where the number m is
determined in terms of different aims. (3) The truss is rolled into the circular cross section from a sheet of
prepreg with the rod axis along the fiber direction so that its stiffness and strength can be utilized
efficiently, and then inserted into the circular hole in the molds through the m plies of pre-drilled prepreg.
The ends of the trusses are split into many parts, which are then pressed upon the prepreg tightly. (4)
Additional m plies of prepreg are laid on the prepreg laid earlier. Thus, the truss ends are embedded in
the mid-plane of the face-sheet. The lower face-sheet is fabricated in the same way. (5) The preformed
sandwich structures are cured at 125ć in an autoclave under a pressure of 0.5 MPa for an hour. The resin
is melted and redistributed in the curing process, joining the trusses with the face-sheets. The sandwich
structures are formed after the curing of resin. (6) The molds are removed, and the fabricated truss core
sandwich structures are cut into the required dimensions. The detailed fabrication process can be found in
our previous published paper. The photograph of fabricated sandwich structure is shown in Fig. 1. The
radius and the length of truss member were r 1.25mm and l 21.2mm , respectively.

Fig. 1. Photograph of a fabricated pyramidal lattice truss core sandwich structure

3. Out-of-Plane Compression Response

The pyramidal truss core sandwich structures are tested at ambient temperature in accordance with
ASTM standard C365 [7]. Out-of-plane compression tests are performed using Instron 5569 test machine.
The load is applied in a rate of 1.0mm / min and measured by the load cell of the test machine. The
relative displacement of two face-sheets is measured via a clip gauge used to define nominal strain of the
specimen. In the test, the specimen is placed between two parallel steel platens without being bonded to
the platens. The pyramidal truss core sandwich specimen is 120mm in length and 120mm in width,
accommodating 3×3 unit cells. The thickness of face-sheets is 1.76mm, and the total thickness of the
specimen is 18.52mm. The typical compressive nominal stress versus strain curves is plotted in Fig.2
In all cases, bedding-in effect during the early stage of deformation is detected and an initial linear
response is observed which is consistent with the parent material properties. Typically, the peak stress
occurs at the point where node rupture but not core member crushing is first observed and then the stress
decreases suddenly. Subsequently, continued loading results in node rupture successively and followed by
a stress plateau. The plateau stress is about half of the peak stress.
2326 Li MA et al. / Procedia Engineering 10 (2011) 2324–2329

4.0
experiment
3.5 analysis
Compressive stress (MPa)

3.0 core member crushing

2.5 node rupture

2.0

1.5

1.0

0.5 bedding-in

0.0
0.000 0.005 0.010 0.015 0.020 0.025 0.030 0.035
Compressive strain

Fig. 2. Analytical predictions and experimental results for the out-of-plane compressive response of sandwich structures

4. Axial Compression Response

Consider a pyramidal truss core sandwich column with clamped ends and subjected to a compressive
end load P on the top of the column. Four failure modes exit for such columns under end compressive
loading: (1) macro Euler (bending) buckling (MEB), (2) macro shear buckling (MSB), (3) face-sheet
wrinkling (FW) and (4) face-sheet crushing (FC).
It is noted that the differences between the macro bending buckling and shear buckling mode is subtle,
as both modes are always simultaneously active and the final deformed shapes are somewhat similar. In
order to illustrate the differences between the two modes it is necessary to observe the deformation of the
core. In bending governed buckling, cross-sections of the core rotate and remain approximately
perpendicular to neutral axis of the column. On the other hand, in shear dominated buckling, the core
shears and cross-sections of the core will not undergo significant rotations [8]. Only if the critical load of
one is far smaller than that of the other, the differences between the two deformed shapes is discernible.
Otherwise, the coupled effect is significant and we can not distinguish the two deformation mode optically.
The limit loads corresponding to the four failure modes are given by
k12S 2 Deq
Pb for macro buckling (1)
L2
2k22S 2 Deq
Pfw for face-sheet wrinkling (2)
( 2l cos Z  t ) 2
Pfc 2 Bt f V fc for face-sheet crushing (3)
where, the factor k 2 depends on the end-constraints set by the pyramidal truss core, Deq is the bending
rigidity of the column, V fc is the crushing strength of face-sheet.
Results of sandwich columns failure in compression expressed in terms of maps facilitate visualization
during the design stage. In constructing such a map, it is assumed that the operative failure mode is the
one associated with the lowest failure load. Failure mechanism maps are constructed for a prescribed
aspect ratio r / l and inclined angle Z of core member. These maps are developed as a function of the
non-dimensional geometrical parameters t f / l and L / l , and the boundaries of failure modes are
 Li MA et al. / Procedia Engineering 10 (2011) 2324–2329 2327

obtained by equating the critical loads for different failure modes. The aspect ratio r / l and inclined
angle Z of core member are fixed on 0.058 and 45°, respectively.
Note that the choice of material properties for the parent material significantly influences the locations
of the boundaries between failure modes. For the laminate face-sheets with equal thickness of every single
ply (for each), the equivalent mechanical properties will change with different stack sequences; for the
lattice truss core, the equivalent shear modulus will change with different geometrical configurations. In
order to explore this response, failure mechanism maps are constructed based on different laminate stack
sequences and core configurations. For the a typical stack sequences of laminate face-sheets and
geometrical parameter of truss core, the relevant failure mechanism maps are plotted in Fig. 3.
80 80
(a ) Eeq
f = 110 GPa (b ) Eeq
f = 76.1 GPa
Vfc = 870 MPa Vfc = 646 MPa
60 60
G13 = 463 MPa G13 = 463 MPa
L/l

L/l

40 40
MEB MEB
* B C
20 20 p
MSB A* * MSB
FW FC FW FC *D
0 0
0.04 0.06 0.08 0.1 0.12 0.14 0.16 0.04 0.06 0.08 0.1 0.12 0.14 0.16
tf / l tf / l

80 80
(c ) Eeq
f = 76.1 GPa (d ) Eeq
f = 59.2 GPa
Vfc = 646 MPa Vfc = 534 MPa
60 60
G13 = 168 MPa G13 = 168 MPa
L/l

L/l

40 40
MEB MEB

20 20
MSB MSB
FW FW
FC
0 0
0.04 0.06 0.08 0.1 0.12 0.14 0.16 0.04 0.06 0.08 0.1 0.12 0.14 0.16
tf / l tf / l

Fig. 3. Failure mechanism maps of the pyramidal truss core sandwich column with different laminate stack sequences and core
configuration (MEB: macro Euler buckling; MSB: macro shear buckling; FW: face-sheet wrinkling; FC: face-sheet crushing).

From Fig. 3, it is found that locations of boundaries between failure modes change obviously with
different laminate stack sequences and core configurations. The face-sheets crushing mode may not even
occur in Fig. 3c. This reminds us that with changing the laminate stack sequence and truss core
configuration the most weight efficient sandwich structure can be reached. And the designable
characteristic is just the advantageous of composite laminate and truss core.

5. Three-Point-Bending Response

Consider a simply supported sandwich beam loaded in three-point bending. The mid-point of the beam
deflects by a transverse displacement u due to the applied load P of the mid-roller. Let l be the beam
length between the supports, w the width of the beam, h the core thickness, and t f the face thickness.
Collapse of the beam occurs by one of several competing mechanisms, the operative failure mode is
dictated by the geometry of the beam and the mechanical properties of the face and core materials.
When the pyramidal truss core sandwich beam is loaded in 3-point bending, four main failure modes
have been identified [9]: (1) core member buckling (CB), (2) core member crushing (CC), (3) face
wrinkling (FW) and (4) face crushing (FC) (It is noted that if the materials comprising sandwich beam is
elastic-brittle, such as fiber reinforced composite, it will crush in external load; if the materials is elastic-
plastic, such as metal, it will yield. We take crush as the failure mode). In the current implementation, the
2328 Li MA et al. / Procedia Engineering 10 (2011) 2324–2329

core is assumed to carry all of the shear and the faces all of the moment, the mutual coupling effect
between the core and face sheets wasn’t considered in the analytical models.
The limit loads corresponding to the four failure modes are given by
4 B sin Z
Pcb V cbSr 2 for core member buckling (4)
2l cos Z
4 B sin Z
Pcc V ccSr 2 for core member crushing (5)
2l cos Z
4H c
Pfw V fw Bt f for face wrinkling (6)
L
4H c
Pfc V fc Bt f for face crushing (7)
L
where V cb and V cb are the buckling and crushing strength of core member, respectively, and V fw and
V fc are the wrinkling and crushing strength of the faces, respectively.
The weight efficiency K is defined as
Pcr
K (8)
W
where Pcr is the failure load of the pyramidal lattice truss core sandwich beam loaded in 3-point bending,
W is the weight of the sandwich beam.
Table 1 The predicted and measured failure loads and weight efficiencies

Thickness of face-sheet (mm) 0.84 1.73 2.6

Predicted failure mode FW FCC CC
failure load (kN) 1.57 13.73 13.78
weight (g) 74.5 151.9 227.6
weight efficiency 2107 9038 6054
Measured failure mode FW FCC CC
failure load (kN) 1.55 11.63 12.87
weight (g) 85 156 235
weight efficiency 1823 7455 5476

FW: face wrinkling, FCC: face and core member crushing simultaneously, CC: core member crushing; for simplicity, the
acceleration of gravity is taken as 10N/kg in the computation of weight efficiency.

Ideally, we should have wished to design specimen geometries in order to probe different failure
modes. However, in the current study, it was not feasible to make a range of hot compression molding
method to manufacture pyramidal truss core sandwich beam with different truss dimensions. To
investigate different failure modes, three types of specimens whose face thicknesses are taken as 0.84mm,
1.73mm and 2.6mm, respectively. The length and width of the specimen are 290 mm and 100,
respectively. And the span of the sandwich beam is taken as 248 mm.
 Li MA et al. / Procedia Engineering 10 (2011) 2324–2329 2329

The specimen is tested by 3-point bending using Inston 5569 machine, applying loading through a
10mm diameter cylindrical roller in accordance with the ASTM standard C393. The load is applied at a
rate of 1 mm min-1 and recorded with the load cell of the testing machines.
The failure loads and weight efficiencies of the three types of specimens are listed in the Table 1. From
Table 1, it is found that the agreement between measured failure loads and predicted ones is good, and
measured failure loads are somewhat lower than predicted ones. This discrepancy is attributed to
imperfections in the manufactured specimens, while the analytical predictions pertain to the perfect
geometry. The predicted and measured weight efficiency of specimen with face thickness 1.73 mm is
lager than those of specimens with face thickness 0.84 mm and 2.6 mm, which validates the effectiveness
of the structural design method.

6. Conclusions

The mechanical response and failure of composite sandwich panels with pyramidal truss cores under
out-of-plane, axial compression and three-point loading were studied. Our investigation complements the
previous studies on the response and performance of lightweight sandwich panels with complex core
construction and provides insight into the failure mechanisms of composite sandwich panels. the results
indicated that the choice of different laminate for the face sheet and relative densities of pyramidal truss
cores significantly influence the locations of the boundaries between failure modes.

Acknowledgements

The present work is supported by NSFC under grant No. 10872059, the 973 Program under grant No.
2011CB610303, the program of Excellent for New Century Excellent Talents in University under grant
No. NCET-08-0152, the Fundamental Research Funds for the Central Universities under grant No.
KIT.KLOF.2010029.

References

[1] Wadley HNG. Cellular metals manufacturing. Adv Eng Mater. 2002; 4(10)726-733
[2] Kooistra GW, Deshpande VS, Wadley HNG. Coressive behavior of age hardenable tetrahedral lattice truss structures made
from aluminium. ACTA Mater. 2004, 52(14):4229-4237
[3] Deshpande VS, Fleck NA, Ashby MF. Effective properties of the octet-truss lattice material. J Mech Phys Solids. 2001;
49(8):1747-1769
[4] Fan HL, Meng FH, Yang W. Sandwich panels with Kagome lattice cores reinforced by carbon fibers. Compos Struct. 2007;
81(4): 533-539
[5] Wang B. Manufacturing and mechanical behavior of the sandwich structure with carbon fiber reinforced pyramidal lattice
truss cores. Mate. Design. 2010; 31(5):2659-2663
[6] Xiong J, Ma L, Wu LZ, Wang B, Vaziri A. Fabrication and crushing behavior of low density carbon fiber conposite
pyramidal truss structures. Compos Struct. 2010; 92(11):2695-2702
[7] ASTM: C365/C 365M-06. Standard test method for flat wise compressive properties of sandwich cores, ASTM Int. West
Conshohocken ,PA, 2006.
[8] Cote F, Fleck NA, Deshpande VS. Fatigue performance of sandwich beams with a pyramidal core. Int J Fatigue
2007;29:1402-1412.
[9] Wicks N, Hutchinson JW. Optimal truss panels. Int J Solids Struct. 2001;38:5165–5183