Sie sind auf Seite 1von 4

J O U R N A L O F M A T E R I A L S S C I E N C E L E T T E R S 1 9, 2 0 0 0, 2205 2208

The fracture properties of novel aluminum foam sandwich structures


W. J. CANTWELL, P. COMPSTON, G. REYES
Department of Engineering, Materials Science and Engineering, University of Liverpool, Liverpool L69 3GH, UK
E-mail: msedept@liv.ac.uk

Sandwich structures consisting of strong stiff skins


bonded to a low density core are currently finding
widespread use in a range of marine and aerospace components where high strength and stiffness combined
with low weight are fundamental design requirements.
Many of these foams are based on polymeric materials such a polyvinyl chloride, polyetherimide and
polyurethane [1]. Although these systems are widely
used in conjunction with low temperature curing
thermoset-based composites, their temperature sensitivity preclude them from use with high temperature
systems such as fiber-reinforced thermoplastics. This is
unfortunate since many thermoplastic-matrix composites, as well as being extremely tough, can be shaped
and molded in a very short manufacturing operation.
At present, there is a growing interest in the potential
offered by a number of lightweight metallic foams for
use in high performance structural applications. Initial
testing has shown that aluminum foams offer excellent sound-proofing properties, electromagnetic wave
shielding, low thermal conductivity low toxicity and excellent toughness characteristics [2]. Recently, a range
of aluminum foams have been developed that retain
many key mechanical properties at temperatures in excess of 300 C [2]. The introduction of these metallic
foams opens up many exciting new avenues for developing novel lightweight cost-effective sandwich structures. One possibility is to manufacture lightweight
aluminum foam sandwich structures based on fiber reinforced thermoplastic skins. Here, the thermoplastic
plies and the metallic core material can, in principle,
be heated to the processing temperature of the composite and stamped in a cold press. The manufacturing
cycle can therefore be completed in a few minutes, a
timescale significantly shorter than that associated with
the co-curing of thermosetting composites skins.
The aim of this study is to investigate the fracture
properties and impact response of a glass fiber reinforced polypropylene sandwich structure based on a
high performance closed cell aluminum foam. Initially,
the skin-core interfacial fracture properties are investigated through a simple fracture mechanics test using
the three point bend sandwich (TPBS) geometry [3].
Following this, a number of impact tests have been conducted on simple beam-type samples to investigate the
failure processes under dynamic loading conditions.
Prior to manufacturing the sandwich structures, the
mode I interlaminar fracture properties of the glass fiber
reinforced polypropylene (GFPP) were characterized
by conducting double cantilever beam tests on 5 mm
thick (10 ply), 20 mm wide unidirectional samples. DeC 2000 Kluwer Academic Publishers
02618028

tails of the manufacturing procedure used these laminates are given in Ref. [4]. The DCB tests were undertaken at a crosshead displacement rate of 1 mm/min
on an Instron 4505 universal test machine. The length
of the aluminum starter defect in each sample was approximately 40 mm and crack advance along the edge
of the sample was monitored with the aid of a graduated scale marked on the edge of the sample. The mode
I interlaminar fracture energy of the composite, G Ic ,
was computed using Berrys experimental compliance
method for which:
G Ic =

n Pd
2Ba

(1)

where B = specimen width, a = crack length, P =


applied force and d = load-point displacement. The
Berry constant n was obtained from the slope of the
plot of log crack length vs log compliance [5].
Sandwich panels with dimensions 120 100 mm
were manufactured via a cold stamping process similar to that used in the manufacture of glass mat thermoplastic components [6]. Here, sheets of glass fiber
reinforced polypropylene prepreg were stacked either
side of a 10 mm thick closed cell aluminum foam block
(Alporas foam from the Shinko Wire Company, Japan)
and placed in a picture frame mold. The nominal density of the foam was 250 kg/m3 and the average cell
size was approximately 4 mm. A 0.5 mm thick layer of
maleic anhydride modified polypropylene (Fusabond
from DuPont de Nemours) was incorporated at the
composite-foam interface to ensure optimal adhesion
[4]. The mold was placed in an air-circulating oven,
heated to 185 C and then removed and stamped in a
cold press. Once the mold had cooled to below 60 C
(typically within five minutes), the panel was removed
from the mold and visually inspected for defects.
The skin-core interfacial fracture properties of the
sandwich structure were characterized using the three
point bend sandwich (TPBS) specimen shown in Fig. 1.

Figure 1 Schematic of the three point bend sandwich (TPBS) specimen


used to characterize skin-core adhesion in the sandwich structures.

2205

Here, a folded aluminum foil starter defect was incorporated at one end of the panel during the manufacturing operation. In preparation for testing, part of
the core and lower skin directly under the aluminum
starter defect were removed to leave the uppermost
skin extending as shown in the figure. Beams with
dimensions 120 mm 12 mm were tested on an Instron 4505 universal testing machine. The specimens
were supported on rollers positioned 100 mm apart and
tested at a crosshead displacement rate of 1 mm/min.
During the test, crack propagation was monitored with
the aid of a painted grid applied to the skin-core interfacial region. The specimen was loaded until the
crack approached the central loading point and then
unloaded.
The interfacial fracture energy was calculated using
an experimental compliance method where the interfacial fracture energy, G c is given by:
P 2 dC
Gc =
2B da

(2)

where P is the applied force, B the specimen width,


C the specimen compliance and a the crack length.
The specimen compliance was determined as a function
of crack length and a curve fit of the following form
applied:
C = Co + K a 3

(3)

where Co and K are constants for a given specimen. In


this study, K was determined by measuring the slope
of the graph obtained by plotting the compliance C
against a 3 .
The impact resistance of the aluminum foam sandwich structures was investigated by conducting low velocity impact tests on beams based on 1 mm thick skins
(2 plies of GFPP). Here, an instrumented impact carriage with a mass of 1.976 kg and a 10 mm diameter
hemispherical head was released from varying heights
to generate impact energies between 1 and 4 Joules.
After impact, the edges of the samples were ground
and polished in order to elucidate the dynamic failure
mechanisms occurring under impact loading.
Fig. 2 shows a typical resistance curve for a mode I
DCB test on the plain glass fiber reinforced polypropylene composite. From the figure, it is clear that the
mode I interlaminar fracture toughness, G Ic , increases
rapidly from 1700 J/m2 to reach a plateau value of approximately 2200 J/m2 . An optical examination of the
edges of the samples highlighted the existence of extensive fiber bridging between the upper and lower crack
flanks. The presence of this bridging mechanism may
well explain the stable mode of crack propagation observed in all tests as well as the initial increase in G Ic
with crack length apparent in Fig. 2. A subsequent scanning electron microscope study of the mode I fracture
surfaces highlighted the presence of significant ductility and plastic flow in the polypropylene matrix. From
this evidence, it is clear that this semi-crystalline fiber
reinforced thermoplastic offers an excellent resistance
2206

Figure 2 Resistance curve for a mode I DCB test on the plain GFPP
composite. Crosshead displacement rate = 1 mm/min.

to interlaminar fracture and should represent an ideal


skin system for use in metal foam sandwich structures.
An optical examination of the sandwich panels indicated that the composite skins had successfully bonded
to the aluminum foam during the processing cycle.
The properties of the interface were then characterized using the TPBS geometry shown in Fig. 1. Fig. 3
shows a typical load-displacement curve for a test on
a GFPP/aluminum foam sandwich specimen. From the
curve it is clear that crack propagation is again stable
throughout the duration of the test. Marked on the loaddisplacement trace are the crack lengths observed on the
edge of the sample during the test. It is interesting to
note that initial crack propagation occurs at relatively
low loads on the linear portion of the load-displacement
trace. As the test continues, the load begins to plateau
reaching a value of approximately 300 Newtons prior
to unloading. The crack lengths and associated loads
and displacements were then used to determine the interfacial fracture energy according to the compliance
calibration method. Fig. 4 shows a resistance curve corresponding to a TPBS test at a crosshead displacement
rate of 1 mm/min. An examination of the R-curve indicates that initial crack propagation occurs at values of
G c below 400 J/m2 . The resistance curve then climbs
very rapidly approaching a value of approximately
2000 J/m2 at the end of the test. It is interesting to note
that this value is very close to the mode I interlaminar
fracture energy of the plain GFPP composite reported

Figure 3 Typical load-displacement trace for a test conducted at a


crosshead displacement rate of 1 mm/min.

Figure 4 Resistance curve for a TPBS test at a crosshead displacement


rate of 1 mm/min.

in Fig. 2 suggesting that adhesion between the skin and


core materials in these sandwich structures is excellent.
An examination of the edges of the TPBS samples indicated that initial crack propagation occurred at the
interface between the maleic anhydride interlayer and
the glass fiber composite. Further investigation showed
that this initial mode of failure resulted in the modest
values of G c in the early part of the R-curve, Fig. 4.
With continued loading, the primary crack propagated
into the lower plies of the skin material leading to significant amounts of fiber bridging within the glass fiber
reinforced thermoplastic and a rapidly rising resistance
curve.
A limited number of TPBS tests were undertaken at
impact rates of strain. Unfortunately, it was not possible to monitor the advance of the interfacial crack under
these dynamic loading conditions and attempts to establish an R-curve proved fruitless. Nevertheless, the maximum force measured during each test was similar to
those measured at quasi-static rates of loading suggesting that the fracture properties of the skin-core interface
remained impressive at high strain rates. An examination of the edges of these samples indicated that the
primary crack occasionally deviated into the aluminum
foam before returning to the uppermost composite skin.
Once again, significant amounts of fiber bridging were
in evidence in the fractured samples.
The impact response of these lightweight sandwich
structures was investigated at energies up to 4 Joules.
A detailed examination of the sample subjected to an
impact energy of 1 Joule failed to highlight any damage in either the skin or core materials. After 2 Joules,
a small indent was observed in the vicinity of the point
of impact, however, no damage was apparent in the
core material. At impact energies of 3 and 4 Joules,
there was significant damage to the top surface composite skin and the aluminum foam core. A microscopic examination of the damaged region highlighted
the energy-absorbing fracture mechanisms. The optical micrograph in Fig. 5a, from a sample impacted at
4 Joules, shows buckling of the uppermost composite skin, and crushing and fracture of the cells in the
foam core directly under the point of impact. Closer

Figure 5 Micrographs showing the edge of a sandwich beam subjected


to an impact energy of 4 Joules, highlighting (a) buckling of the composite skin and foam cell fracture and (b) foam cell collapse and interlaminar
separation.

inspection of the sample, Fig. 5b, highlights the collapse of cells during fracture. It is also interesting to
note from Fig. 5b that interlaminar separation has occurred within the composite skin rather than along
the bi-material interface or within the aluminum foam
core.
The evidence presented in this paper suggests that
a new range of energy-absorbing sandwich structures
based on thermoplastic matrix composite skins can be
manufactured in a simple and time-efficient stamping operation. Fracture mechanics tests using the three
point bend sandwich geometry have highlighted the
high degree of adhesion between the tough composite skins and the ductile aluminum foam core. In addition, a series of low velocity impact tests on small
beam-like samples have shown that these structures
offer enormous potential for use in a wide range of
energy-absorbing applications.

Acknowledgments
The authors are grateful to the Mexican Government
(CONACYT) and the Engineering and Physical Sciences Research Council for supporting this work. The
donations of glass fiber reinforced polypropylene by
Sandra Coburn of BI Composites Ltd., the Fusabond
M613-05 by Dr. Karlheinz Hausmann of DuPont de
Nemours and the aluminum foam by Dr. Jerry Lord
of the National Physical Laboratory are also gratefully
acknowledged.

References
1. D . Z E N K E R T ,

An Introduction to Sandwich Construction


(EMAS, Cradley Heath, 1997).
2. T . M I Y O S H I , M . I T O H , S . A K I Y A M A and A . K I T A H A R A ,
in Proc. of the Symp. on Porous and Cellular Materials for Structural
Applications, edited by D. S. Schwartz, D. S. Shih, A. G. Evans and
N. H. G. Wadley (MRS, California, 1998) Vol. 521, p. 133.

2207

3. W . J . C A N T W E L L , J . R A T C L I F F E , R . S C U D A M O R E
and P . D A V I E S , Composites Science and Technology 59 (1999)
2079.
4. G . R E Y E S and W . J . C A N T W E L L , ibid., to appear.
5. Protocol for mode I interlaminar fracture testing, European Structural
Integrity Society (ESIS), 1997.

2208

6. W . J . C A N T W E L L , W . T A T O , H . H . K A U S C H and R .
J A C Q U E M E T , Journal of Thermoplastic Composites 5 (1992)
304.

Received 25 April
and accepted 8 June 2000

Das könnte Ihnen auch gefallen