You are on page 1of 7

Manufacture and compression

properties of syntactic foams


P. BUNN and J.T. MOTTRAM"
(Westland Aerospace/" University of Warwick, UK)
Received 3 November 1992; revised 17 November 1992

A study has been made of the processability, bulk density and uniaxial compression properties of a syntactic foam system with varying volume fractions of
phenolic microballoons. Short-term compression tests showed that the compression yield strength and initial tangent modulus of elasticity were linearly dependent on the bulk density (and the volume fraction of microballoons). The microballoon concentration and resin binder composition was found to be crucial to the
ease of manufacturing syntactic foams.
Key w o r d s : syntactic foams; composition; microballoon volume fraction; bulk
density; uniaxial compression properties
In the aerospace industry structures are often fabricated
using composite sandwich panel constructionL Where
the sandwich panel shape is complex (i.e., a helicopter
rotor blade), the core is often foam (e.g., Rohacell). In
other cases where the structure is flatter the core may be a
honeycomb2. The compressive strength of conventional
stabilized sandwich core is less than l0 MPa 2. The low
value means that there is a design problem in those
regions of an aircraft sandwich panel which experience
high compressive loading (e.g., where door hinges are
mounted). To overcome this problem, the weak conventional honeycomb or foam core may be replaced by a
plug of a much stronger syntactic foam with a compressive yield strength between 30 and 100 MPa. The compressive modulus will be similar to that of conventional
cores (i.e., 1 to 2 GPa). The shear strength of a syntactic
foam plug is also a relevant property of the material as it
determines the size of the plug required to transmit the
applied compressive loading. Although not reported in
this paper, the shear strength of the foams was determined using the short beam method.
The word 'syntactic' simply indicates that the foam is
made by mixing microballoons, ceramic spheres, or
other lighweight aggregate, with a resin system3. Unlike
most other foams, a syntactic material is one whose
density before cure is the same as that after cure. This
property is advantageous in the manufacture of aerospace plugs. Other typical applications34 for syntactic
foams are as automotive tooling compounds, ablative
heat shields for re-entry vehicles, underwater buoyancy
aids, and structural components in submarines and missiles.
Westland Aerospace, Isle of Wight, have used syntactic
foam plugs for a number of years. Their material7 consists of a mixture of small, hollow, phenolic-type spheres

(average diameter 90 ~tm), called microballoons, with a


thermosetting resin binder. The latter consists of a
mineral-filled paste epoxy, a plasticized liquid epoxy and
an appropriate hardener. The specified foam 7 has a phenolic microballoons volume fraction of about 40%. This
foam should have a minimum compressive yield strength
of not less than 49 MPa 7. Syntactic foams made with
phenolic microballoons feature reduced weight,
increased stiffness and improved crack resistance over
standard foams5.
Westland Aerospace have found that their specified syntactic foam7 has a too high compressive yield strength for
a number of applications. It has been suggested that a
lower strength may be obtained by increasing the volume
fraction of microballoons in the standard mixture. This
procedure would have the additional benefit of weight
saving.
The study described here aimed to establish how the bulk
density and short-term compressive properties of a syntactic foam system varied with microballoon concentration. This study gave additional insight into the effect
of microballoon content on the processability of the
foam.

MA TERIAL SPEClFICATION
The phenolic microballoons, UCAR Thermoset Microballoons type BJO-0930, were manufactured by Union
Carbide. They have an average particle diameter of 90
lam and a maximum bulk density of 0.104 g cm-3.
The cold-setting thermoset epoxy binder comprised three
components all manufactured by Ciba-Geigy. Part I was
Araldite AV1218. This is a mineral-filled epoxy resin
paste. The mineral, chalk (density 2.6 g cm-3), is added

0010-4361/93/070565-07 1993 Butterworth-Heinemann Ltd

COMPOSITES. VOLUME 24. NUMBER 7. 1993 565

T a b l e 1. C o m p o s i t i o n o f s y n t a c t i c f o a m s
Composition number

1
2
3
4
5
6
7
8
9
10

Volume % of
microballoons*

% Weight of three-part resin binder


Part 1 AV121

Part 2, AY103

Part 3, HY951

71
71
71
71
71
71
71
38
38
37

24
24
24
24
24
24
24
58
57
56

5
5
5
5
5
5
5
4
5
7

0
14
25
34
41
47
53
52
52
51

*Error is + 4%, exceptcomposition1

to increase the viscosity of the base resin and as an


inexpensive method of increasing the bulk and strength
of the cured material. The chalk content (volume fraction 35 to 40%) also prevents excessive thermal expansion and shrinkage during curing. AV121, when mixed
with about 4% by weight of Part 3, the hardener
HY9518,9, provides a solvent-free adhesive with excellent
gap-filling properties 8. It has a density after curing of 1.7
g cm -3. Part 2 was Araldite AY1039. This is a plasticized
liquid epoxy resin with a viscosity less than that of
AV121. When mixed with about 9% by weight of the
hardener HY951, an excellent bond strength9 can be
developed. AY103 has a density of 1.1 to 1.2 g cm -3. The
hardener HY951 has a density of 0.98 g cm -3. The
recommended8,9 curing time at 25"C for both epoxies
AV121 and AY103 with the hardener HY951 is 24 h.
The composition of the syntactic foam specified by Westland Aerospace is7: Araldite AV121, 75 parts by weight;
Araldite AY103, 25 parts by weight; hardener HY951, 56 parts by weight; and phenolic microballoons BJO0930, 10 parts by weight of binder. Compared with conventional syntactic foams 3-5, the high proportion of the
mineral-filled epoxy AV121 in this composition means
the foam is much stronger and denser.
MATERIAL PROCESSING AND TEST SPECIMEN
PREPARATION
The syntactic foams were produced following the
method specified for the manufacture of plugs at Westland Aerospace. The procedure is detailed in the Westland Group Specification Sheet for pre-moulded light
inserts7. Material processing and specimen preparation
was carried out by the first author in the laboratories at
Westland Aerospace. The temperature, 19 + 2C, and
the relative humidity, 45 + 5%, recorded in the laboratory as environmental conditions were believed to
influence material properties significantly.
The moulds for the compression test specimens were 200
mm lengths of mild steel tube with an inner diameter of
27 mm and a wall thickness of 3 mm. To prevent the
foam from adhering during curing, the release agent
Ambersil DP 200 was sprayed onto the moulds.
The composition of the 10 syntactic foams tested is given
in Table 1. Each foam has been given a composition

566

COMPOSITES. NUMBER 7 . 1993

number for identification in the text. Because extraction


equipment had to be used when measuring and mixing
microballoons and resin, some of the former may have
been lost, thus reducing the microballoon concentration
calculated from the weight measurements (see Table 1).
Compositions 1 to 7 had the same resin binder composition and a volume fraction content of microballoons
ranging from 0 to 53%. The maximum practical fraction
content of microballoons was found to be 53% by
volume. The resin binder was prepared as specified in
Reference 7. This binder composition will be referred to
in the text as the standard binder.
To reduce viscosity of the uncured foam mixture with the
maximum microballoon concentration ( ~ 53%), the
proportions in the three-part binder mix were changed.
Compositions 8 to 10 (see Table 1) had a much higher
proportion of the low viscosity liquid epoxy AY103.
The three-part resin binder was mixed thoroughly before
the micr0balloons were gradually added. It was found to
be important to mix the microballoons into the binder
slowly to minimize loss due to the extraction equipment.
Once thoroughly mixed, the uncured syntactic foam mixture was heated to 30"C. This reduced the viscosity and
aided further mixing in v a c u o . Finally, the mixture was
reheated to 30C to make it more fluid.
Depending on the viscosity of the uncured syntactic
foam mix, three methods were used to fill a mould I.
1)

2)

For volume fractions ofmicroballoons in the range 0


to 14%, the foams (compositions 1 to 2) had low
viscosity so that the mixture could be poured into
the mould.
For volume fractions of microballoons in the range
15 to 47%, the viscosity was such that the mould had
to be filled by extrusion. A circular disc with a
central hole of diameter 15 mm was pushed down on
top of the foam mixture. The disc had an outer
diameter just less than the 180 mm inner diameter of
the mixing pot. When the disc was pushed down, the
trapped mixture could only flow upwards through
the hole at the centre of the disc. An empty mould,
coated with releasing agent, was placed directly
above the hole and the foam mixture was forced to
flow up into the mould. When the mould was filled, a
plate was placed over the open end to prevent the

Load

mixture escaping. For a microballoon volume content of <41% (compositions 3 to 5) the force could
be applied manually, but a pneumatic ram was
necessary to provide sufficient force when the microballoon content was 47% (composition 6).
3) For volume fractions of microballoons above 47%,
the foams (compositions 7 to 10) became too dry to
extrude and high pressure moulding, possibly with
manual compaction, was used to produce the samples.

IIIJLIII Ol
'_I ~ ~~

,,

Sillien~iieiiaiiai~n

After a mould had been filled it was allowed to cold-set at


a room temperature of 19+2C for 12 h. The foam was
then post-cured at 170 4- 5"C for 2 h, as recommended
by Ciba-Geigy8,9.
Following curing, the cylindrical sample (200 mm long
by 27 mm diameter) of foam was removed from the
mould. The sample always had a thin, hard, crust of resin
on the surface. Each sample was machined to make the
ASTM D695 cylindrical compression specimens"
(nominal dimensions 50.8 mm long by 25.4 mm diameter). The removal of the resin crust improved the
homogeneity of the specimens, though it made the true
volume fraction of microballoons slightly higher than the
figure quoted in Table 1. Five compression test specimens were prepared for each foam composition in
accordance with ASTM D695".

scale

To

@
Die set

COMPRESS~ON TEST METHOD


A 250 kN Dartec 9500 servo-hydraulic testing machine
was used for compression testing. The salient features of
the compression test rig are shown in Fig. 1. A rocker
fixture, with graphite lubrication, was located centrally
between the Dartec and test rig to ensure pure axial
loading. The load from the testing machine was transferred to the specimen by means of a high precision die
set which minimizes frictional effects and guarantees
axial loading. The bearing surfaces for end loading were
flat and parallel platens of silver grade steel with a yield
stress much higher than that of the syntactic foam.

Fig. 1 Compressiontestrig salientfeatures

Point selected to calculate modulus of elasticity

50F I / I
%'

Specimens were compressed at a constant rate of 0.02


mm s-l, giving a strain rate of 394 ~te per second. Each
test lasted for at least 7 min. The short-term tests were
carried out in accordance with the standard ASTM
D6951] for testing rigid plastics. For each specimen the
compressive yield strength and modulus of elasticity was
measured. There were a number of deviations from the
conditions of the test standard: the relative humidity was
in the range 49 + 7%, the ambient laboratory temperature was high at 24 4- 1.5C, and the stroke rate was 1.2
mm rain -] as opposed to 1.3 mm min-L
Fig. 2 shows a typical compression stress/engineering
strain diagram for a phenolic microballoon syntactic
foam (composition 5). Engineering strain was determined by dividing the change in stroke (from onset of
loading) by the original specimen length. The compression stress, o'c, was calculated on the basis of the original
average cross-sectional area. The average area was used
here because it was very similar to the minimum called
for in the ASTM standard. The compressive yield
strength, O%y,is defined as the stress at the yield point,
being the first point on the stress/strain diagram at which

/ 1/
45[- T

/Compressive
yield
s t r e n g t ~

40 ~ _ _ _ / . . _ / r . _ ~

Plateau

35
~- 30

~'~

~- Densification
completed

20

15
[..)

Section

~-

I
15

I
20

Section

I
5

I
10

I
25

I
30

i
35

40

Engineering strain (%)

Fig. 2 Average compressive stress against engineering strain for


syntactic foam having41%microballoonvolumefraction

an increase in strain occurs without an increase in stress.


The value of the compressive yield strength is shown in
Fig. 2. The modulus of elasticity was calculated from the
tangent to the initial linear portion of the load/deformation characteristic.

COMPOSITES . NUMBER 7. 1993

567

Fig. 3 Microstructure of the standard binder composition. The dark


regions are the chalk filler

Fig. 4 Microstructure of the standard syntactic foam having 41%


microballoon volume fraction

RESULTS AND DISCUSSION

remained intact following the stirring and the mould


filling processes. Microscopy also revealed that, despite
the thorough mixing of the constituents, the syntactic
foams often were not homogeneous and contained
regions which were binder-rich. From X-ray photographs it was noted that the regions of lower microballoon content were at the centre of the samples. This
suggested that, when the foam flows through the mould,
surface friction produces a velocity gradient in the flow
causing the binder to flow more quickly than the microballoons. When the method of filling a mould was high
pressure moulding (compositions 7 to 10), the microstructure of the foam was more homogeneous.

The resin binder for compositions 1 to 7 (see Table 1)


had a composition very similar to that specified in the
Westland Group Process Specification7. In an attempt to
reduce the viscosity of the uncured foam, compositions 8
to 10 (foams with >50% microballoon volume fractions), the percentage weight of the mineral-filled AV121
was reduced whilst that of the plasticized liquid AY103
was increased (see Table 1). Compositions 8 to 10 had
percentage weights of the hardener HY951 equal to 4, 5
and 7, respectively. The change in viscosity achieved with
the new binder mix was not adequate for the method of
mould filling to be changed to the extrusion route.
Cured samples of compositions 8 and 9, having 4 and 5%
hardener, had many circumferential splits along their
length. When all other parameters remained constant,
the samples of composition 10, having 7% hardener, did
not possess these splits. It appeared that the splitting was
due to combination of a high microballoon content and a
lack of hardener HY951. The amount of hardener in
compositions 8 and 9 (4 and 5%) was not sufficient to
provide full cold-set curing of both epoxy components.
This is supported by the Ciba-Geigy instruction sheets8,9,
which state that for full cure between 6 and 8% by
weight of the hardener is required. As a consequence of
insufficient hardener, the microballoons were not constrained by the partially cured binder and could expand
considerably on heating to the post-cure temperature of
170 5"C. The splitting in compositions 8 and 9 was thus
due to thermal shrinkage on cooling down to room temperature. Increasing the hardener content to 7% in composition 10 provided sufficient hardener for the binder to
cold cure completely, thus preventing the thermal expansion problem.
Figs 3 and 4 show photographs of the polished surfaces
for material compositions 1 and 5. Fig. 3 shows the
microstructure of the standard epoxy binder without any
microballoons. The two phases of the foam can be
clearly seen. The darker area is the chalk filler. The chalk
particles range in size and shape, the largest granules
being about 100 ~tm across. Fig. 4 shows a section
through the syntactic foam with 41% microballoon
volume content (composition 5). It can be seen that the
microbaUoons are spherical in shape. A majority
(> 95%) of the microballoons was found to have

568

COMPOSITES.

NUMBER

7 . 1993

When preparing the foam samples it was impractical to


eliminate all porosity. X-ray photography was used at
Westland Aerospace to provide an indication of the
amount, size and shape of the porosity. A general observation from the X-ray photographs was that porosity
was a function of foam mixture and method used to fill a
mould. The pure binder (composition 1) contained a lot
of bubbles of diameter < 2 mm. As the microballoon
volume content increased to 41% (compositions 2 to 5),
the number of bubbles decreased, but the diameter grew
to between 2 and 3 mm. The decrease in number of
bubbles may be attributed to the increase in viscosity
which reduced turbulence in the processing. Increasing
microballoon content above 41% (compositions 6 to 10)
gave foams with varying amounts of porosity depending
mainly on how well the samples had been consolidated in
the moulds.
The bulk density of the foam compositions 1 to 7 was
measured using the compression test specimens. Theoretical predictions for the bulk density could be made using
the constituent weights and densities and the rule of
mixtures (ROM) relationship. The linear relationship
between the average bulk density and the average
percentage volume fraction of microballoons is plotted
in Fig. 5. Error bars on the foam properties have been
included.
The average bulk density of the foams ranged from 0.78
to 1.5 g cm -3 as the volume fraction of phenolic microballoons decreased from the maximum of 53 to 0%. The
least-squares method was used to fit a straight line to the
results. The equation is
p = 1.470 - 0.012 Vmi

(1)

.5 t

'

\-.~2

Composition

1.2
I

cl

-r q

1.0

-r 5

u 0.9
co

---

Theoretical

- -

Measured

no.

3
, ~ J -

q j~J-

, ~.i.~ /
'

Equation

(2)

i
1.q

J5
1.

'

30

m
o_
E

lt0

],

~ 2o

0.8
I

~u~ 4o
"~
>.

60

E
~ 5o

Composition no.

E
~

70

(1)
1.3

80

15

20

25

30

35

40

45

50

Volume f r a c t i o n o f microballoons,

i
0 8

J
0.9

55

~0
1.

i
1.1

i
1.2

Bulk

density,

J
1.3

p (g cm -3)

Vmi (~)

Fig. 5 Bulk d e n s i t y a g a i n s t p e r c e n t a g e m i c r o b a l l o o n v o l u m e fract i o n f o r the s t a n d a r d syntactic f o a m s

where p is the bulk density (g cm-3), and Vmi is the


percentage volume fraction of phenolic BJO-0930 microballoons. Equation (1) is valid for foams manufactured
with the standard binder mixture and a microballoon
volume fraction between 0 and 53%.
The theoretical density (see dashed error bars in Fig. 5)
was generally slightly higher than that measured for two
reasons: (1) some porosity is always present; and (2) the
resin-rich crust on the foam samples, having a higher
density due to the chalk filler, had been machined off
when preparing the compression test specimens.
Fig. 2 shows a typical compressive stress/engineering
strain curve for a phenolic microballoon foam (composition 5). The curve can be sub-divided into three
sections. During section 1 the foam exhibits elastic
behaviour. Providing the yield point is not exceeded, the
foam will revert back to its original shape after the load
has been released. Elastic behaviour is exhibited by a
foam to strain of 4 4- 1%. The yield point, corresponding with the compression yield strength (O-~y), occurs at
the transition between the elastic section and the collapse
plateau, which is section 2 of the curve. The compressive
load remains fairly constant during the plateau section.
In this region of the compression test the foam can
withstand 95 % of its compressive yield strength (as given
in Fig. 6). The engineering strain at the end of the plateau
section is large, and can be 25%. It is associated with the
foam undergoing densification as the microballoons collapse. The third, and last, section of the curve occurs
after full densification has been achieved and the compression sample can resist higher loads. The prominence
of the third section was found to be dependent on the
microballoon volume fraction and the onset of surface
cracks ~. In the majority of the compression tests for this
study the stroke was terminated after the plateau had
been established and before the third section, if it existed,
had been reached.
The frictional constraint at the loaded ends combined
with the Poisson's expansion effect caused the cylindrical
specimen to change shape considerably during test. The
barrelling of a specimen became very visible by the time
it had undergone full densification. The mode of failure
was shear. The first visible signs of material failure were

Fig. 6 Compressive yield strength against bulk density for the


standard syntactic foams

3.0

,~Ti '

~2.5

2DL3
~4

~.uation"

3
2.0

Composition5no.

13)

LI

~i~ 1.5
c~ m
E
' ~ 1 . 0

0.5

0 8

0t
.9

I0
1.

I
1.1

J
1.2

Bulk density,p

i3
1.

t
1.q

i
1.5

( g c m -3)

Fig. 7 Initial tangent modulus of elasticity against bulk density for


the standard syntactic foams

at large strain ( > > 10%). A surface crack at 45* to the


load direction appeared at mid-depth of the column and
grew gradually as the compression continued.
Figs 6 and 7 present the compressive yield strength and
initial tangent modulus of elasticity against bulk density
for foam compositions 1 to 7. Fig. 6 shows that the
average compressive yield strength decreased from a
maximum of 71 MPa for the pure binder (composition 1)
to 28 MPa for the foam (composition 7) having the
maximum volume fraction of microballoons ( ~ 5 3 % ) .
The variation of strength with bulk density is practically
linear suggesting that the ROM relationship is appropriate. The relationship is
Crcy = 60.74 p - 19.42

(2)

where O-cyis the compressive yield strength (MPa) and p


is the bulk density (g cm-3).
Fig. 7 shows that the average initial tangent modulus of
elasticity decreased from the resin binder value of 2.7 to
0.81 GPa when the microballoon volume fraction was
,-, 53%. The variation of modulus with density is fairly
linear for bulk densities >0.83 g cm -3 (microballoon
volume fractions < 41%), suggesting that the ROM relationship was again appropriate. This was a surprising
result as the syntactic foam phase distribution of a dispersion of spheres in the continuous matrix does not
satisfy the modulus ROM one. For the latter to be appro-

C O M P O S I T E S . N U M B E R 7 . 1993

569

priate, the two phases must be distributed as continuous


parallel slabs aligned with the loading direction. However, the same observation has been made by other
researcherslL The ROM relationship between modulus
and density for foams with bulk density in the range 0.83
to 1.49 g cm -3 is
E = 2.11 p - 0.37

(3)

where E is the initial tangent modulus of elasticity (GPa),


and p is the bulk density (g cm-3).
The modulus of elasticity when the microballoon concentration was a maximum (53%), and the density was
lowest (0.78 g cm-3), was found to be considerably lower
than Equation (3) predicts. As this modulus measurement was from a single specimen no definite conclusions
can be drawn.
In an attempt to reduce the viscosity of the syntactic
foam having maximum microballoon concentration
( ~ 53 %), the binder composition was changed in compositions 8 to 10 (see Table 1). Because the new binder
compositions had less mineral-filled AV121 there was a
reduction in the bulk density of about 10%. The splitting
of the samples during the curing process for compositions 8 and 9 meant that compression specimens could
not be prepared. The increased proportion of the hardener HY951 in composition 10 meant that compression
specimens could be prepared and tested. The average
compressive yield strength for composition 10 was 30.0
MPa. This new average strength was similar to the value
of 28.8 MPa for composition 7, the foam manufactured
with the standard binder mixture. The average modulus
of elasticity for the composition 10 foam was 0.98 GPa.

CONCLUSIONS
1) Syntactic foams having the standard Westland
Aerospace binder composition and a phenolic microballoon (UCAR type BJO-0930) concentration up
to 53% have been manufactured. As the volume
fraction of microballoons was increased there was
an accompanying viscosity increase in the uncured
foam mixture. This change in viscosity with volume
fraction of microballoons meant that the method for
filling a sample mould changed from simple casting
(<14%), to extrusion (>14% to _<47%), and
finally to high pressure moulding (>47%). The
latter method requires expensive equipment for component manufacture.
2) The bulk density of seven syntactic foam compositions using the standard resin binder composition
ranged from 0.78 to 1.5 g cm -3, as the volume fraction of microballoons decreased from 53 to 0%.
There was a linear relation between these quantities
as expected from the rule of mixtures.
3) Short-term compression tests to ASTM D695 have
been carried out on five specimens of each of seven
foams with the standard binder. The average compressive yield strength ranged from 28 to 71 MPa,
and the average initial tangent compressive modulus
of elasticity ranged from 0.81 to 2.7 GPa. The lowest
compression properties were for foams having the
highest microballoon concentration (-,~53%). Both
compressive properties gave a fairly linear relation-

570

COMPOSITES

. NUMBER

7 . 1993

ship with bulk density (or volume fraction of microballoons). These rule-of-mixtures relationships were
a surprising result as the particulate composite
microstructure of syntactic foams would suggest
otherwise.
4) The composition of the three-part binder was
changed from the standard mixture in an attempt to
decrease the viscosity of the uncured foam containing maximum microballoon concentration ( ~ 53 %).
It was found that it was still difficult to fill the mould
as the new binder mix did not significantly reduce
the viscosity. After the curing process had been completed it was further found that the quality of this
new syntactic foam composition was poorer than the
standard foam composition if there was insufficient
hardener for complete cold curing.
5) The average compressive properties of the syntactic
foam prepared with the new binder composition
having sufficient hardener were slightly above those
for the equivalent foam manufactured with t h e
standard binder.
Further experimental work is needed to develop additional insight into the compressive properties of syntactic
foams with long-term loading (creep and fatigue resistance), in typical aerospace environments, and with
different compositions of the binder mixture. The objective of these tests would be to provide a comprehensive
property data base for aerospace plug design.

ACKNOWLEDGEMENTS
P. Bunn carried out this study for his final year undergraduate project in the Engineering Department, University
of Warwick. He and J.T. Mottram are grateful for the
provision of materials, laboratory facilities and technical
assistance from R. Spencer and R. Perry, Westland
Aerospace, Isle of Wight. The authors thank the reviewer
for helpful comments during the preparation of the
manuscript.

REFERENCES
1 Structural Sandwich Composites, M1L-HDBK-23 (Government
Printing Office, Washington, DC)
2 Marshall, A.C. 'Core composites and sandwich structures' in
International Encyclopedia o f Composites, Vol 1 edited by S.M. Lee
(VCH Publishers, New York, 1990) pp 488-506
3 Malloy, R.A. and Hudson, J.A. 'Foamed composites' in International Encyclopedia of Composites, Vol 2 edited by S.M. Lee (VCH
Publishers, New York, 1990) pp 257-267
4 Shutov, F.A. 'Syntactic polymer foams' in Handbook of Polymeric
Foams and Foam Technology edited by D. Klemper and K.C.
Frisch (Hanser, Munich, 1991) pp 355-374
5 Ruhno, R.A. and Sands, B.W. 'Hollow spheres' in Handbook oJ
Fillers and Reinforcements for Plastics edited by H.S. Katz and
J.V. Milewski (Van Nostrand Reinhold, New York, 1978) pp 301310
6 Lyle, A.R. and Collins, M.H. 'Syntactic foam as a structural material in a marine environment' Proc 2nd lnt Conf Polymers in a
Marine Environment, London, 14-16 October 1987 (Marine Management, 1989) pp 167-170
7 Westland Aerns0ace 'The manufacture of pre-moulded lightweigh!
inserts for honeycomb sandwich panels' Westland Group Proces~
Specification 369 Issue 2 (1987)
8 Ciha-Geigy 'Araldite AV121 with hardener HY951 or HY965!
Instruction Sheet No A.30d (May 1990)

9 Ciba-Geigy'Araldite AV103and AYI05 with hardener HY951 or


HY965' Instruction Sheet No A.26j (August 1985)
10 Buan, P. 'Manufacture and compression properties of syntactic
foams' Final Year Undergraduate Project (Department of Engineering, University of Warwick, 1992)p 63
11 'Standardtest method for compressive properties of rigid plastics'
A S T M Standard D695-69

12 Luxmoore,A.R. and Owen, D.R. 'Syntacticfoams' in Mechanics o f


Cellular Plastics edited by N. C Hilyard (Macmillan, New York,
1982) pp 359 39l

AUTHORS
P. Bunn is now a Graduate Engineer with Westland
Aerospace, East Cowes, Isle of Wight PO32 6RH, UK.
Correspondence should be addressed to J.T. Mottram,
Lecturer in Structural Mechanics, at the Department of
Engineering, University of Warwick, Gibbett Hill Road,
Coventry CV4 7AL, UK.

COMPOSITES. NUMBER 7. 1993

571

You might also like