Composite Structures 94 (2012) 2337–2342

Contents lists available at SciVerse ScienceDirect

Composite Structures
journal homepage: www.elsevier.com/locate/compstruct

Crashworthiness characteristics investigation of silk/epoxy composite square tubes

S.A. Oshkovr a,⇑, R.A. Eshkoor a, S.T. Taher b, A.K. Ariffin a, C.H. Azhari a
a
Department of Mechanical & Materials Engineering, University Kebangsaan, 43600 Bangi, Malaysia
b
Department of Mechanical & Manufacturing Engineering, Fibigerstraed 16, DK-9100 Aalborg, Denmark

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

Article history: This research concentrates on the evaluation of crashworthiness characteristics of natural silk/epoxy
Available online 5 April 2012 composite square tubes energy-absorbers. Composite laminate specimens were subjected to static axial
compression load and experimental evaluation of the energy absorption capability of silk/epoxy compos-
Keywords: ite. Specimens were in the form of square cross-sections with the dimension of 80 mm  80 mm and a
A. Fibres radius curvature of 5 mm. The variables in the experiment were the length of the tubes built 50 mm,
A. Polymer–matrix composites 80 mm and 120 mm. Meanwhile, the thickness of the walls, consisting of laminates of silk/epoxy of 12,
B. Mechanical properties
24 and 30 plies, correspond to equivalent wall thickness of 1.7 mm, 3.4 mm and 4.2 mm, respectively.
The parameters measured were the total absorbed energy (Etotal), and the crash force efficiency (CFE).
Etotal is the measure of the amount of energy that the structure can withstand without failure and thus
is a measure of its strength, while CFE gives a quantitative indication of the mode of failure of the com-
posites. The mode of failure was observed using photography.
Ó 2012 Elsevier Ltd. All rights reserved.

1. Introduction load-carrying structures. Furthermore, axially crushed thin-walled

square tubes have been widely used as energy absorbers because
Recently together with structural efficiency, passenger safety is of their high specific energy absorption capacity and long stroke
also an important issue in application of material to transportation [7]. On the other hand, based on some studies remarkably few
industries. Hence, the crashworthiness parameters are introducing works has been done on natural composite materials in the auto-
to predict the capability of structure to prevent the massive damage motive industry during the last century hence limited researches
and protect the passenger in the event of a crash. Crashworthiness for thin shell structures made of natural fibres have been published
parameters for various thin-walled tubes made from metal or fibre/ [8]. Nowadays, natural fibres are receiving more interest due to
resin composites in different geometries have been studied. A crit- increase attention to use ecofriendly materials.
ical difference of tubular composites failure modes compared with This research focused on using silk due to its superior strength,
metallic is the brittle collapse. In addition, in composites, tubular durability, high failure elongation, higher specific energy absorption
failure modes are involved with micro-cracking development, also it is notable that silk has better resistance to moisture when
delamination, fibre breakage, etc., instead of plastic deformation. compared to other natural fibres. Previous paper [8] presented re-
Implementation of composite materials in the field of crashworthi- sults for 12 layers silk/epoxy composite square tubes also a compar-
ness is attributed to Hull, who in 80s and 90s of the last century ison table of mechanical properties of some natural fibres including
studied extensively the crushing behaviour of fibre reinforced com- silk and synthetic fibres. In order to improve the silk/epoxy square
posite material. He found that the composite materials absorbed tubes crashworthiness parameters, in the current work crushing
high energy in the face of the fracture surface energy mechanism response of silk/epoxy composite in the form of 24 and 30 layers
rather than plastic deformation as observed for metals [1]. This of silk/epoxy square tubes experimentally investigated. It analyses
observation has inspired others to further investigation about the crushing load–displacement profile of silk/epoxy square tubes
crashworthiness characteristics of composite materials. Studies determine the crashworthiness parameters and overall energy
have examined the axial crushing behaviour of fibre-reinforced absorption on composite tubes. The findings might be used as the
tubes [2], fibreglass tubes [3,4], PVC tubes [5] and carbon fibre basis of design guidelines for using silk in crashworthy systems
reinforced plastic (CFRP) tubes [6]. and safety applications.
In real applications, for example, in automobiles, the square
tubes are not only energy-absorbing members, but also serve as 2. Materials and methods

⇑ Corresponding author. Tel./fax: + 45 9815 1675. Silk/epoxy square tubes made of silk woven fabric in weft (90°)
E-mail address: [email protected] (S.A. Oshkovr). and wrap (0°) direction and weft fibre direction was parallel to the

0263-8223/$ - see front matter Ó 2012 Elsevier Ltd. All rights reserved.
http://dx.doi.org/10.1016/j.compstruct.2012.03.031
2338 S.A. Oshkovr et al. / Composite Structures 94 (2012) 2337–2342

2 (a)
3

(b)

Fig. 1. Axial compression test of square tube with 50 mm length and 30 layers silk/epoxy: (a) photographs of different steps of compression test and (b) combined diagram of
load P, and crash energy absorption E, variation during the test. The sequential number of each picture marks the point of the two curves corresponding to the photographs.

tubes’ axis. Specimens were compressed in INSTRON MTS 810 uni- regarding the morphology of the tubes in steps of the tests and infor-
versal testing machines utilised with 250 kN load capacity. Load– mation obtained from load displacement curves have been dis-
displacement curves obtained from quasi-static compression test cussed as following.
for silk/epoxy square tubes is presented in the following sections.
Results are divided into three groups namely short length, medium
length and long length according to their lengths (50 mm, 80 mm 3. Results
and 120 mm). Meanwhile, each section presents crushing history
as load–displacement curves, which are describing the different Load–displacement curves of silk/epoxy composite square
type of square tubes used for the tests according to their number tubes resulted from compression test with 50 mm length and
of silk/epoxy layers (12, 24 and 30 layers). For easy comparison different numbers of silk/epoxy layers (24 and 30 layers) presented
in the current paper, final results for 12 layers [8] are added to from Figs. 1–3. Combined diagrams of the load–displacement curve
the results for 24 and 30 silk/epoxy laminates square tubes. and energy are indicated from Figs. 1b–3b. According to the load–
(Figs. 1–3) consist of two parts. Part (a) depicts photographs of displacement curves, tubes behaving linearly in the beginning of
different stages of the tests, and part (b) indicates the load– the test. Fig. 1b presents the load drops down by compression of
displacement curve which presenting the energy-absorbing capa- the square tube. This remarkable decrease in load presents mid
bility of tubes. The sequential number of each picture marks the length buckling with failure starting in the composite square tube
point of the two curves corresponding to the photographs. Results from half of the length of the tube. Photographs present the steps

1
2 3
(a)
4

(b)

Fig. 2. Axial compression test of square tube with 80 mm length and 30 silk/epoxy layers: (a) photographs of different steps of compression test and (b) combined diagram of
load P, and crash energy absorption E, variation during the test. The sequential number of each picture marks the point of the two curves corresponding to the photographs.
 S.A. Oshkovr et al. / Composite Structures 94 (2012) 2337–2342 2339

1
2
3 (a)
4

(b)

Fig. 3. Axial compression test of square tube with 120 mm length and 30 silk/epoxy layers: (a) photographs of different steps of compression test and (b) combined diagram
of load P, and crash energy absorption E, variation during the test. The sequential number of each picture marks the point of the two curves corresponding to the photographs.

of deformation in the tubes and are taken at various steps of the and unstable collapse of the compressed tube, which commences
compression during the testing of the square tubes. with a circumferential fracture of the composite laminate at a
distance approximately equal to its half the tube length this phe-
nomena is clearly presented in Fig. 2a. This figure performs a sig-
4. Discussions
nificant reduction of the compressive load after reaching the
peak load followed by tube catastrophic failure.
4.1. Failure mechanism
Some of the silk/epoxy composite tubes, for example, in Figs. 1b
and 2b diagrams demonstrate first and second peak loads. These
Silk/epoxy tubes were found to crush in modes slightly different
tubes performed significant reduction of the compressive load
than those observed in synthetic composite structures. For brittle
immediately after the load–displacement curve reached the maxi-
composite materials such as glass and carbon composites based
mum load in the region of the crush zone. Away from this zone rest
on epoxy resin, the crushing process is a cyclic process of interlam-
length of the tube has remained with less damaged, and almost it
inar cracks propagating between layers in the crushed region of the
looks like as a shorter square tube. Hence with continuing the
tube forming lamina bundles [9]. For the metallic tubes, the
deformation second peak load was observed when the load
crushing process is a process of plastic deformation. Silk/epoxy
reached the second maximum, thereafter load dropped to the
composite square tubes in compression test presented a ductile
lower value until the compaction zone.
deformation without debris splitting after being crushed.
In general, the failure modes observed during compression tests
were unstable local buckling and mid length buckling and affected 4.2. Crashworthiness characteristics
by the geometry of composite square tubes in terms of a number of
silk/epoxy laminates and lengths of composite square tubes. In Figs. 4–7, the crash load/energy absorption characteristics of
According to previous researches the nonlinear energy absorption the tested composite square tubes namely peak compressive load
response suggests that care must be taken in selecting specimen (Pmax), average compressive load (p), absorbed crash energy (E),
geometry for energy absorption characterisation studies [11]. specific absorption energy (SAE), maximum compressive strength
In the initial stages of compression, load–displacement curves (rmax), crush force efficiency (CFE) and stroke efficiency (SE) are
showed a significant drop in load after reaching the peak load. High presented. Error bars denote the standard deviations of specimens
peak crush failure load leads to a catastrophic failure mechanism at with same geometrical characteristics. Errors indicate the variety
initial failure crush stage, which results in low crush force effi- of data for specimens in the compression tests. In order to have a
ciency (CFE) at post-crush stage as well as unstable load–displace- general graphical comparison of tubes according to their length,
ment behaviour. However, significant energy can also be absorbed two sets of crashworthiness characters were combined in each
when structural elements catastrophically fail [2]. diagram.
Utilising compression test photographs helped to understand
the modes of failure. Tubes with 24 and 30 number of silk/epoxy 4.3. Peak and average load
composite laminates square tubes in all lengths (50 mm, 80 mm
and 120 mm) exhibited mid length buckling, which was failure ini- Each load–displacement curve shows the history of crushing for
tiated at the mid of the tube length which then proceed to overall related tube. Crushing started at a critical value which is called peak
tube buckling followed by catastrophic failure. This failure clearly load, and then the load traveled along the tube. As it has been re-
has been shown in Figs. 1–3. In the other word, mid length buck- ported before by some researchers, during catastrophic failure, the
ling in the compressed specimens is featured by ductile fracture peak load is remarkably high and drops off quickly, so the average
2340 S.A. Oshkovr et al. / Composite Structures 94 (2012) 2337–2342

Fig. 4. Peak and average load characteristics for short length (a), medium length (b)
and long length (c) composite square tubes (error bars denote standard deviations). Fig. 5. Total absorbed crash energy E for short length (a), medium length (b) and
long length (c) square tube (error bars denote standard deviations).

load is low [10]. Information from the load displacement curves in Meanwhile, results presented the critical thickness in long length
Figs. 1b–3b directly showed the maximum loads (Pmax) in silk/epoxy tubes with 24 layers of silk/epoxy composite. It is proven by previ-
composite tubes during compression tests. Meanwhile, average ous researches, mentioned that peak load is depended on the
compressive load (p) for each tube was calculated from the load–dis- geometric and material characteristics of the tested specimen [1,2].
placement curve related to the tube. Results have been indicated in
Fig. 4 for silk/epoxy composite tubes with different length and 4.4. Crash energy absorption capability
thickness.
Increase peak load and average load associated with the rising The value of energy absorption capability of a structure is
number of composite laminates in Fig. 4a and b. It clearly indicated depended on the area under the load–displacement curve. Conse-
that, number of composite plies in the tube wall of the tested spec- quently, square tubes with higher value of crushing load showed
imens influence significantly on peak load value. It is notable in the higher value of energy absorption. Fig. 5c presents the critical
Fig. 4a there is a significant higher value of peak load and average thickness in the square composite tubes which is indicating by a
load in tubes with 30 plies of silk/epoxy laminates to compare with constant value of energy absorption in long length square tubes
24 plies of silk/epoxy laminates. This difference is slightly in Fig 4b, with 24 and 30 layers plies of silk/epoxy laminates.
and finally in Fig. 4c peak load and average load are almost con- Total absorbed crash energy (E) is depicted in Fig. 5 for short,
stant in 24 and 30 silk/epoxy composite plies. This constant medium and long lengths square tubes. The effect of decreasing
amount is presenting the critical thickness of silk/epoxy composite thickness in energy absorption reduction calculated in the form of
square tube, which is 24 layers of silk/epoxy laminates in long a percentage. Obtained results indicated that the energy absorption
length tubes (120 mm). capability decreased (13.1%) when wall thickness reduced from
As conclusion, results showed specimens with higher numbers 4.2 mm (30 layers silk/epoxy) to 3.4 mm (24 layers silk/epoxy)
of plies and shorter length performed more resistance in compres- and decreased (76.5%) when wall thickness reduced from 4.2 mm
sion test with performing higher value of peak and average load. to 1.7 mm (12 layers silk/epoxy). These figures also show that while
 S.A. Oshkovr et al. / Composite Structures 94 (2012) 2337–2342 2341

Fig. 6. Specific energy, SAE and maximum compressive strength, rmax for short
length (a), medium length (b) and long length (c) specimen (error bars indicate
standard deviations).
Fig. 7. Crush force efficiency CFE, and stroke efficiency SE, characteristics for short
length (a), medium length (b) and long length (c) square tubes (error bars denote
the highest length (120 mm) decreased to 80 mm and 50 mm, standard deviations).
energy absorption capability reduced 35.6% and 37.6%, respectively.
Results remarked that the capability of energy absorption of the strength. Results cleared increasing the number of plies has a
silk/epoxy square tube is depended on the geometry of tubes. direct effect on the increase the maximum compressive strength
Tubes with higher length and thickness are able to absorb more and specific energy. As it mentioned earlier maximum strength is
energy, which is due to absorbing more energy by their long length depend on the peak load to the cross section area of the specimen.
until the end of crushing. This phenomenon is the same for tubes Specific energy in short and mid lengths tubes increase with the in-
with the higher number of laminate plies resulted in higher value crease the number of layers of specimens. On the contrary, the
of absorbing energy caused by resistance of more layers. Overall specific energy for long length tube with 30 layers decrease due
length and thickness tubes effect on the load hence peak load to increase of the weight of specimens.
and average load. In the meantime, the energy is the area under
the load–deformation curve hence it is depended on peak and 4.5. Stroke and crush force efficiency
average load. The following figure also performs the critical thick-
ness in tested silk/epoxy square tube composite. Another important measure of crush performance is the crush
Fig. 6 indicates the specific absorbed energy SAE and maximum force efficiency, which is defined as the ratio of the mean crush aver-
compressive strength rmax values according to specimen length. age load to the initial crush failure load. In order to evaluate the
Specific energy absorption of the tubes is defined as the absorbed crashworthiness of energy absorber device, attention should be
energy per unit mass of tube and maximum compressive strength directed to its crush force efficiency (CFE) the stroke efficiency (SE).
of the tubes is the maximum carried load to the cross sectional The diagrams depicted in Fig. 7 include the crush force effi-
area of the specimen. This figure revealed that the length of spec- ciency (CFE) and the stroke efficiency (SE). As mentioned crush
imens has the reverse effect on the specific energy and maximum force efficiency (CFE) is an important crushing characteristic of
2342 S.A. Oshkovr et al. / Composite Structures 94 (2012) 2337–2342

the load–displacement curve. In silk/epoxy composite square tubes failure initiated at the mid of the tubes which then proceed to be
once the crush started load dropped significantly and collapse con- overall. Results showed that the highest values of Etotal (1.03 kJ)
tinues with low value load. The ratio of average load and peak load were obtained for the specimens with 30 silk/epoxy laminates
is a considerable value to presents the failure mechanism of com- and 120 mm depth, while the maximum value for CFE (0.30) was
posite namely catastrophic or progressive failure. obtained in the specimen with 12 layer silk/epoxy laminates and
Small values of CFE proved that catastrophic failure happened 120 mm length. Meanwhile, a remarkable increase in peak load
in all silk/epoxy composite tubes. Catastrophic failure is affected was performed from square tubes with 24 and 30 layers silk/epoxy
by geometry of a specimen. Data comparison between diagrams composite laminate to compare with square tubes with 12 layers
in Fig. 7 performed long length tubes (120 mm) obtained slightly silk/epoxy composite in short length tubes. Increasing the length
higher value of crush force efficiency. These figures also show of tubes showed the critical thickness in a long length tube which
silk/epoxy tubes’ stroke efficiency. The desire energy absorber de- is 24 layers of silk/epoxy composite, which caused a constant peak
vice has crushable structures, which can be defined as SE. and average load in square tubes with 24 and 30 plies of silk/epoxy
laminates.
4.6. Comparison with similar works
References
Silk/epoxy has high elongation at failure which causes high en-
ergy absorption capability. Moreover low density of silk decreases [1] Hull D. A unified approach to progressive crushing of fibre reinforced tubes.
Compos Sci Technol 1991;40:377–421.
weight of tubes and improves the specific energy absorption. But [2] Farley G, Jones R. Crushing characteristics of continuous fibre-reinforced
comparison between silk/epoxy and similar carbon/epoxy tubes composite tubes. Int J Compos Mater 1992;26:37–50.
previously reported by Mamalis et al. [11] shows much higher [3] Mamalis A, Manolakos D, Demosthenous G, Ioannidis M. Analysis of failure
mechanisms observed in axial collapse of thinwalled circular fibreglass
specific energy absorption by carbon/epoxy tubes. This fact is composite tubes. Thin-walled Struct 1996;24:335–52.
due to the lower Young’s modulus of silk, which results decreasing [4] Mamalis AG, Robinson M, Manolakos DE, Demosthenous GA, Ioannidis MB,
in buckling load in the silk/epoxy composite tubes. This phenome- Carruthers J. Crashworthy capability of composite material structures. Compos
Struct 1997;37:109–34.
non is more significant when length of tubes increases or thickness [5] Mamalis A, Manolakos D, Viegelahn G. Axial crushing of thin PVC tubes and
decreases. To avoid early failure in the tube, controlling the buck- frusta of square cross section. Int J Impact Eng 1989;8:241–64.
ling using different means are suggested, e.g., using a hybrid of silk [6] Mamalis AG, Manolakos DE, Ioannidis MB, Papapostolou DP. On the response
of thin-walled CFRP composite tubular components subjected to static and
and carbon fibres based composite, implement thick sandwich in-
dynamic axial compressive loading: experimental. Compos Struct
stead of thin laminate walls and controlling the buckling load using 2005;69(4):407–20.
a proper trigger mechanism. [7] Zhang XW, Su H, Yu TX. Energy absorption of an axially crushed square tube
with a buckling initiator. Int J Impact Eng 2009;36:402–17.
[8] Ataollahi S, Taher ST, Eshkoor RA, Ariffin AK, Azhari CH. Energy absorption and
5. Conclusions failure response of silk/epoxy composite square tubes: experimental. Compos
Part B: Eng 2012;43(2):542–8.
[9] Harte A, Michael MNF, Ashby F. Energy absorption of foamed filled circular
This study focused on natural silk/epoxy square tubes energy tubes with braided composite walls. Eur J Mech 2000;19(A):31–51.
absorber. Square tubes made of a different number of silk/epoxy [10] Schultz MR. Energy absorption capacity of graphite–epoxy composite tubes
layers, which are 12, 24 and 30 layers. According to the tests, engineering mechanics. Virginia Polytechnic Institute and State University;
1998. p. 1–129.
crashworthiness parameters were defined and analysed. In the ax- [11] Mamalis AG, Manolakos DE, Ioannidis MB, Papapostolou DP. Crashworthy
ial compression tests, all the specimens exhibited catastrophic characteristics of axially statically compressed thin-walled square CFRP
failure. Specimens failed with mid length buckling, which was composite tubes: experimental. Compos Struct 2004;63:347–60.