close

Вход

Забыли?

вход по аккаунту

?

715

код для вставкиСкачать
Polymer International 42 (1997) 241È244
The Mechanical Properties of
Particulate-filled Aramid and
Polyethylene Laminates
Z. P. Wang,* J. S. Ghotra,¤ G. Pritchard
School of Applied Chemistry, Kingston University, Penrhyn Road, Kingston upon Thames, Surrey, KT1 2EE, UK
& R. G. Rose
School of Mechanical, Aeronautical and Production Engineering, Kingston University, Friars Avenue, Roehampton Vale, London
SW15 3DW, UK
(Received 2 September 1996 ; accepted 22 September 1996)
Abstract : Some mechanical properties of particulate-Ðlled polyethylene/epoxy
and aramid/epoxy laminates are reported, following earlier work with
particulate-Ðlled glass/epoxy laminates. The behaviour of the organic Ðbre laminates was di†erent from that of glass Ðbre ones, investigated and reported earlier.
There was an increase in compression strength with increasing Ðller content, with
both kinds of organic Ðbre reinforcement. The interlaminar shear strength values
were signiÐcantly lower for the polyethylene laminates than the aramid ones at
all Ðller concentrations, and they fell to little more than 10 MPa at high Ðller
levels. However, the impact damage zone in drop weight tests was generally
smaller for the polyethylene laminates, and the visible crack damage was less
apparent than in the aramid ones. The Ñexural strength and modulus values are
also reported.
Key words : Ðller, aramid, polyethylene, alumina trihydrate, impact.
relating to Ðllers used in conjunction with organic Ðbre
(aramid and polyethylene) reinforced laminates. Such
systems combine an unusually brittle Ðlled matrix, possessing an elongation at break of perhaps less than
0É5%, with Ðbres of high energy absorption capacity.
This brief communication is intended simply to draw
attention to the main experimental observations from a
preliminary investigation of the subject, without
detailed analysis at this stage. The Ðller employed in
this work was the widely used Ðre retardant and smoke
suppressant, alumina trihydrate (ATH).
INTRODUCTION
It is often necessary for technical or economic reasons
to add particulate Ðllers to resins and glass Ðbre laminates. In most cases, the tensile strength of the unreinforced resins is reduced by Ðllers,1,2 although the
fracture toughness is increased for certain Ðller volume
fractions, exhibiting a maximum, or else levelling o†, as
the Ðller content increases.3 Some Ðllers, especially
those with a high aspect ratio (such as Ñy-ash4) cause
increases in tensile or Ñexural strength. The e†ect of
Ðllers on the mechanical properties of Ðbrous laminates
is less well documented, and there are very few reports
EXPERIMENTAL
* Current address : Polymer Department, Beijing Research
Institute of the Chemical Industry, PO Box 1442, Beijing
100013, PeopleÏs Republic of China
¤ To whom all correspondence should be addressed.
Eight-ply woven roving laminates were prepared by
hand lay-up using a modiÐed bisphenol A/epichlor241
Polymer International 0959-8103/97/$09.00 ( 1997 SCI. Printed in Great Britain
Z. P. W ang et al.
242
TABLE 1. Composition (by weight) and thickness of the aramid and
polyethylene fibre laminates
Material
Designation
K1
Kevlar
Dyneema
Epoxy
ATH
Thickness (mm)
V a
f
fb
K2
36
0
3·0
0·3
0
100
33
3·8
0·3
0·19
K3
36
67
50
4·4
0·3
0·31
D1
D2
D3
36
27
50
0
4·6
0·3
0
27
100
33
4·7
0·3
0·19
27
67
50
5·1
0·3
0·31
a Fibre volume fraction.
b Filler volume fraction in matrix only.
hydrin epoxy resin system, cured at room temperature
with an aliphatic mixed amine hardener. The particulate Ðller was alumina trihydrate (grade FRF20, surface
treated, with a mean particle size of 17 km, obtained from Alcan Chemicals, UK). The Ðller was mixed
with the epoxy resin before Ðbre impregnation, in the
proportions 1 : 1 and 2 : 1 by weight. We shall refer to
the ATH content as 0 (control), 50 and 100 parts per
100 parts resin by weight (phr). A dispersing agent,
Hypermer FP4, from ICI was added at 1 part per 100
parts Ðller by weight. The above materials, together
with the lay-up and curing procedures, have already
been described more fully in connection with the work
on glass Ðbre laminates1 and (with only slight
di†erences) in previous reports on the microscopy of
impact damage5 and on compressive strength retention
after impact.6 The di†erence between the present and
the previously reported work was the use of two
organic Ðbre reinforcements : (a) aramid, Kevlar' 49,
supplied by Plastic Reinforcement Fabrics Ltd, Bournemouth, UK, areal density 170 g m~2, warp/weft ratio
127/127 and (b) ultra-high molecular weight polyethylene, Dyneema' SK60, areal density 180 g m~2, warp/
weft ratio 132/132, from the same supplier.
The laminates were subjected to drop-weight impact
tests with three di†erent input energies, using the simple
apparatus previously described5 for glass laminates,
with a hemispherical steel indentor, and an impact
velocity of 4É4 m s~1. The damaged zones were examined visually and by optical microscopy (previous
attempts to use ultrasonic C-scans had been hindered
by the Ðller particles). The extent of the damage zones
in several specimens was evident from visual inspection
and optical microscopy and was carefully measured.
Other laminate samples were subjected to interlaminar
shear strength (ILSS) tests, using the short beam
method, with a 5 : 1 span to depth ratio and a crosshead
movement rate of 1 mm min~1, and Ðnally to Ñexural
and compressive measurements. The Ñexural test
method was ASTM D-790M, strain rate approximately
1% min~1 ; the compression tests were made using an
IITRI compression test jig1 with the calculation procedure following ASTM D 695-80, and crosshead speed
of 0É5 mm min~1.
RESULTS AND DISCUSSION
Impact
Table 1 identiÐes the composition of the laminates, and
shows the e†ect of the Ðller on the thickness.
The impact damage zones were examined, and their
volumes calculated from estimated dimensions. The
cross-sections of the zones always increased in area with
increasing distance from the impact site. The polyethylene zones were almost conical, having roughly circular
cross-sections, but the aramid cross-sections tended
towards irregular quadrilaterals with well deÐned
corners. As Table 2 shows, the volume of the damage
zone was much the same for both Ðbre types at the
lowest impact energy, but as the impacts became more
severe, the aramid zones were decidedly larger. This
may be because of the higher strain to failure of the
polyethylene Ðbres, with a correspondingly greater
capacity to absorb energy. The impact damage zone
increased progressively in all specimens as the impact
energy increased, but the largest increase in the case of
all the aramid laminates and also those polyethylene
samples with the highest Ðller content, was between 29
TABLE 2. Volume (in mm3) of the damage zone in
laminates as a function of impact energy
Energy (J)
K1
K2
K3
D1
D2
D3
14
29
39
169
600
4351
288
685
3089
472
2267
3047
182
592
640
150
388
557
357
1126
1507
POLYMER INTERNATIONAL VOL. 42, NO. 3, 1997
Mechanical properties of laminates
243
and 39 J. This has previously been observed with glass
laminates as well, and was believed in that case to correspond with the observed onset of extensive glass Ðbre
fracture.6 The aramid laminates sometimes showed
crosses, i.e. two intersecting surface cracks, on their
reverse faces, and there were also several fractured Ðbres
visible, in contrast to the polyethylene reinforced laminates. Increasing the ATH content of the aramid laminates resulted in a larger damage zone at 14 J and
especially 29 J, but there was a reduction in the zone
size at the 39 J level. The increased ease of crack initiation in the highly Ðlled epoxy resin is less relevant to
the damage process once Ðbre fracture has become
extensive. Complete penetration would probably have
left a much smaller damage zone. The correlation
between damage zone size and mechanical property
retention after impact has not yet been established for
organic Ðbre laminates. Polyethylene reinforced laminates always showed their largest damage zones at the
highest Ðller content, i.e. 100 phr ATH, but the 50 phr
ATH samples showed consistently smaller zones than
those without Ðller. So far, no reason has become
apparent for this ; the same trend was not observed with
aramid Ðbres. As might be expected, delamination
occurred more readily at high impact energies, particularly when Ðller content increased and the mismatch
between the elasticity of matrix and Ðbres increased.
ILSS and flexure
Table 3 shows the remaining mechanical properties.
The reproducibility of these results was quite good. The
ILSS values for the polyethylene laminates fell steadily
with increasing Ðller content. There was a similar but
less well deÐned e†ect with the aramid samples. The
unÐlled polyethylene laminates failed by a single shear
crack on each side of the specimens, but the Ðlled ones
and all the aramid specimens failed in multiple shear.
The polyethylene ILSS values were lower than those of
the aramid laminates, probably because of relatively
poor adhesion between polyethylene and epoxy. The
larger mismatch between the low extensibility of the
particulate-Ðlled matrix (elongation at break about
0É5%) and the high elongation of the Ðbres (3É3%) could
also contribute.
Table 3 gives some Ñexural strength and modulus
results. The strength was lowest for the 50 phr Ðller level
with both Ðbre types. The table also compares the
experimental Ñexural moduli obtained here with crude
estimates calculated from the semi-empirical Narkis7,8
model (intended for spherical particles in the same
range of Ðller volume fractions used here) in conjunction with the rule of mixtures, neglecting the transverse
Ðbres and ignoring any anelastic character in the
reinforcement. (The calculation assumes inter alia that
the measured moduli are tensile rather than Ñexural,
but experimentally the values of tensile and Ñexural
moduli for Ðbrous laminates are usually similar). Errors
arising from the inappropriateness of the rule of mixtures, especially with polyethylene, prevented any
detailed agreement between calculated and experimental values, except in the case of the Ðlled aramid laminates, where there was some agreement. The average of
all the calculated values of the Ñexural modulus for
aramid laminates was greater than the experimental
values by a factor of 1É066, whereas for polyethylene the
factor was 1É303.
Compression
Both Ðbre types showed an increase in compressive
strength with Ðller content (Table 3). This is despite the
reduction in epoxy resin tensile and Ñexural strength
caused by addition of ATH. It can be explained by the
fact that Ñaw initiation and crack opening are more difÐcult under compression than under tension, and so the
particles which facilitate fracture in tension do not
cause problems under compression. They suppress the
ductility of the matrix and are likely to delay failure,
which was always by yielding rather than cracking.
Interestingly, previous work with E-glass Ðbres, while
consistent with the trends shown in the present data in
respect of ILSS and Ñexural properties, did not give an
increase in compressive strength with increasing Ðller
TABLE 3. Interlaminar shear strength, flexural strength (FS), flexural
modulus (FM) and compressive strength (CS) values (average of five
specimens ; standard deviations in parentheses)
Laminate
ILSS (MPa)
FS (MPa)
FM (GPa)
Expl
K1
K2
K3
D1
D2
D3
20·9
16·8
17·3
15·7
10·6
10·2
(1·7)
(0·5)
(1·6)
(1·1)
(0·8)
(0·8)
230
178
187
139
90
126
(4·8)
(6·7)
(7·0)
(6·1)
(2·1)
(10·0)
POLYMER INTERNATIONAL VOL. 42, NO. 3, 1997
18·4
19·7
22·1
11·2
13·5
13·9
(0·9)
(0·9)
(1·5)
(1·2)
(1·1)
(1·3)
CS (MPa)
Calcd
20·2
21·3
22·5
15·6
16·7
17·8
57
65
79
46
47
54
(3·6)
(3·2)
(4·0)
(2·9)
(2·1)
(2·6)
Z. P. W ang et al.
244
content. The compressive strength of the corresponding
8-ply glass laminates was 159 MPa without ATH, and
148 MPa with 100 phr ATH. This is a small reduction
compared with that normally observed in tension,
where crack opening is facilitated around particulate
stress concentrations. (Much higher absolute values
were obtained for the compressive strength of the glass
laminates, both before and after impact, with the same
resin, when using a hot cure system to ensure a lower
viscosity resin mix. The fracture toughness of the
unÐlled resin was approximately the same for both cure
systems. The increase was attributed to lower porosity
in the low viscosity system.)
for both Ðbre types, in contrast to the observations
made previously with glass laminates. Further work is
necessary with Ðlled organic Ðbre laminates. This is
essentially a preliminary report.
ACKNOWLEDGEMENT
One of us (Z.P.W.) acknowledges the Ðnancial support
of Kingston University during a period as a visiting
scientist.
REFERENCES
CONCLUSIONS
The Dyneema (polyethylene) reinforced samples showed
a low ILSS, Ñexural strength and compressive strength,
but a high impact strength judged by damage zone size ;
the Kevlar (aramid) ones tended towards the opposite
position. The Ðller makes the matrix extremely brittle,
with a low elongation at break. It was conÐrmed that
compressive strength rose with increasing Ðller content,
1 Wainwright, R. W., Pritchard, G., Phipps, M. A., Yang, Q. &
Parsley, R., Paper 14, Proceedings of Conference on Polymers in a
Marine Environment. Inst. Marine Engineers, London, October
1991, p. 115.
2 Ramsteiner, F. & Theysohn, R., Composites, 15 (1984) 121.
3 Moloney, A. C., Kausch, H. H. & Steiger, H. R., J. Mater. Sci., 18
(1983) 208.
4 Srivastava, V. K. & Shembekar, P. S., J. Mater. Sci., 25 (1990) 3513.
5 Pritchard, G. & Yang, Q., J. Mater. Sci., 29 (1994) 5047.
6 Pritchard, G. & Yang, Q., Polym. Polym. Comp., 2 : 4 (1994) 233.
7 Narkis, M., J. Appl. Polym. Sci., 20 (1976) 1597.
8 Narkis, M., J. Appl. Polym. Sci., 22 (1978) 2391.
POLYMER INTERNATIONAL VOL. 42, NO. 3, 1997
Документ
Категория
Без категории
Просмотров
3
Размер файла
144 Кб
Теги
715
1/--страниц
Пожаловаться на содержимое документа