close

Вход

Забыли?

вход по аккаунту

?

Interpolymer Complexes and Miscible Blends of Poly(N-vinyl-2-pyrrolidone) with Novolac Resin and the Effect of Crosslinking on Related Behaviour

код для вставкиСкачать
Polymer International 41 (1996) 283-292
Effects of Soft-segment Prepolymer
Functionality on the Thermal and
Mechanical Properties of R I M
Copolymers
John L. Stanford, Richard H. Still & Arthur N. Wilkinson"
Manchester Materials Science Centre, University of Manchester and UMIST, Grosvenor Street, Manchester M1 7HS, UK
(Received 31 January 1996; revised version received 28 March 1996; accepted 27 April 1996)
Abstract: Segmented copolyureas (PUr) and copoly(urethane-urea)s (PUU)
comprising 50% by weight of polyurea hard segments (HS) and polyether soft
segment (SS) with different functionalities, have been formed by reaction injection moulding (RIM). The HS were formed from 4,4'-diphenylmethane diisocyanate reacted with mixed isomers of 3,5-diethyltoluene diamine. The nominal functionality of the SS prepolymers used (either amino- or hydroxyl-functionalised
polyoxypropylenes with a constant molar mass per functional group of
-2000gmol-') was systematically increased from 2 to 4. RIM materials were
characterised using dynamic mechanical thermal analysis, differential scanning
calorimetry, tensile stress-strain and fracture mechanics studies. Generally, the
PUr exhibited far superior thermal-mechanical properties than equivalent PUU
materials but inferior fracture resistance, owing to morphological variations
resulting from differences in copolymerisation behaviour. For both systems,
tensile behaviour was shown to be dominated by the degree of phase separation,
whereas fracture properties showed a degree of dependence on SS functionality.
Key words: reaction injection moulding, polyurea, polyurethane-urea, thermomechanical properties, fracture behaviour.
INTRO DUCTlON
ing homopolymers, various AB-type block copolymers
and free monomers.' This solid mixture possesses a
non-equilibrium m ~ r p h o l o g y ~which
- ~ arises from the
direct competition between homo- and copolymerisation kinetics and thermodynamic changes occurring
during the RIM process. Extensive studies at
UMIST3,4,7-'1 have shown that such a morphology
comprises co-continuous, soft- and hard-segment
phases, the degree of phase separation of which depends
on hard-segment content, thermal history and hardsegment structure.
Soft-segment (SS) prepolymers utilised in RIM are
usually di- or tri-functional polyoxypropylenes with
molar masses per functional group (equivalent weights,
En)in the range 1000 to 2000gmol-' which, in order to
increase their reactivity, are either end-capped with 10
to 25% of polyoxyethylene to give primary hydroxyltipped polyols, or the terminal hydroxyl groups are converted to amine groups. The effects of SS prepolymer
Copoly(urethane-urea)s (PUU) and copolyureas (PUr)
formed by reaction injection moulding (RIM)' are used
extensively for large, complex components. Nominally,
these RIM copolymers are segmented block copolymers
formed via random-step reactions of mixtures of
(hydroxyl- or amine-) functionalised polyethers and aromatic diamines with aromatic diisocyanates.' However,
during reaction to form the copolymers, as the liquid
reactants are rapidly converted to a solid, a combination of spinodal decomposition-induced phase
~ e p a r a t i o n and
~ . ~ vitrification effectively quenches the
system to yield a mixture of reaction products compris-
* To whom all correspondence should be addressed. Present
address: Department of Materials Technology, Manchester
Metropolitan University, John Dalton Building, Chester
Street, Manchester M1 5GD, UK.
283
Polymer International 0959-8103/96/$09.00 0 1996 SCI. Printed in Great Britain
284
molecular structure on phase separation1’ are also
important in determining copolymer morphology and
properties. Varying the chemical structure of the SS
changes its solubility
and hence the
compatibility between soft and hard segments. In addition, increasing SS prepolymer molar mass at constant
functionality (i.e. increasing the molar mass per functional group or equivalent weight, EJ result^'^-^^ in a
higher degree of phase separation again, owing to
increased incompatibility between the two copolymer
segments. However, only limited studies have been
p~blished~’-’~
on the effects of varying the functionality of the SS prepolymer in RIM copolymer systems.
These papers concentrated mainly on the processing
and demoulding properties of RIM copolymers and
gave little, if any, insight into their thermal or mechanical properties. A previous paper25 reported on the
structural development of the RIM PUU and RIM PUr
copolymers used in the present study and showed that
significant variations in copolymer molar mass and
hard-segment (HS) sequence length occurred, both as a
result of the unequal reactivity of the monomers and of
the varying functionality of the SS prepolymer. Higher
reactivity and functionality of the SS resulted in a more
rapid development of copolymer molar mass. Consequently, higher SS functionality was shown to increase
significantly the incipient (or ‘green’) strength of both
PUU and PUr, but also to reduce the overall degree of
phase separation developed in these materials. The RIM
PUr exhibited relatively poor green strength and a
higher degree of phase mixing than equivalent PUU
materials. These differences were ascribed to the kinetic
competition between polymerisation and phase separation processes, in that PUU systems with relatively low
SS reactivity exhibited phase development more akin to
that of a blend, with HS homopolymerisation and rapid
macrophase separation of unconnected HS and SS
chains being the favoured processes. Thus, materials
formed under these conditions have relatively coarse,
highly phase-separated co-continuous morphologies.
However, in systems with high SS reactivity such as
PUr, the rapid reaction of the SS prepolymer results in
phase development more akin to that of a block copolymer system undergoing relatively slow microphase
separation. Consequently, these materials form finerscale, less phase-separated co-continuous morphologies.
This paper extends the previous
and reports on
the effects of the aforementioned variations in morphology on the mechanical and thermal properties of RIM
PUU and PUr materials.
J . L. Stanford, R . H . Still, A. N . Wilkinson
polymer with En z 2000gmol-’; (iii) an aromatic
diamine chain extender and (iv) a catalyst for the
hydroxyl-isocyanate reaction. The chemical structures
of the various reactants are shown in Scheme 1
The polyisocyanate (I), Isonate M340 (Dow
Chemical), is a mixed uretonimine/urethane-modified
variant of 4,4‘-diphenylmethane diisocyanate (MDI)
with a value of En= 143 f 2gmol-’ by isocyanate
titration, and a nominal functionality of 2.15. The five
polyethers are: amine-functionalised polyoxypropylenes
(IIa), D4000 diamine and T5000 triamine (both Texaco
Chemical); and hydroxy-functionalised poly(oxypr0pylene-oxyethy1ene)s (IIb), DS25 diol (ARCO), M1 11
trio1 (Lankro Chemical) and P1 tetrol (Union Carbide
Chemical). The En of each of the five prepolymers was
determined by acetylation:26 potentiometric amine
titration2’ was also used to characterise the polyamines.
Gel permeation chromatography (GPC) was used to
determine average molar masses for the three polyols,
relative to polyoxyethylene standards.28 The values of
En obtained from acetylation, which is quantitative
towards hydroxyl and amine groups, are shown in
Table 1 and are similar for the various prepolymers,
varying between
1900 and
2300 g mol- l . The
polyols used are block copolymers comprising 802085 mol% of oxypropylene end-capped with
15mol% of oxyethylene and are known1.29~30to
contain primary and secondary hydroxyl groups, with
80% being primary. The number-average functionalof the polyols were determined from the ratio of
ities
number-average molar mass (from GPC) to equivalent
weight, and are given in the last column of Table 1. In
-
-
--
- vn)
I
.l
f,=2or3
IIa
r
1
EXPERIMENTAL
Reactants
The materials formed in this study comprised up to four
components: (i) a polyisocyanate; (ii) a polyether pre-
111
Scheme 1. Idealised reactant structures.
POLYMER INTERNATIONAL VOL. 41, NO. 3, 1996
285
Efects of soft-segment prepolymer functionality
TABLE 1. Characterisation data for soft-segment prepolymers
Prepolymer
Enu (g mol-')
E , , ~(g mol-I)
MnC(g mol-')
fnd
DS25 diol
M111 triol
P1 tetrol
D4000 diamine
T5000 triamine
2046 f 19
2182 f 22
2287 f 23
21 67 20
1899f10
-
4206
6066
7982
2304 f 31
1964 f 20
-
2.06
2.78
3.49
2"
3"
a
Number-average molar mass per functional group, E n , from acetylation.26
E , from potentiometric t i t r a t i ~ n . ~ '
Number-averagemolar mass from GPC.
Number-averagefunctionality (f, =/
(GPC)/E,(acetylation)).
I?"
" Nominal functionality.
mixture of the 2,4- and 2,6-isomers of DETDA (Lonza
the case of the triol and tetrol, the values off, are lower
AG). The PUU systems contained dibutyltin dilaurate,
than the nominal values, being 3 and 4 respectively.
at a concentration of 1.05 mol/mol isocyanate groups,
These differences are due to side reactions that occur
during the base-catalysed p r o p ~ x y l a t i o n , ~yield~ * ~ ~ ~ ~to
' catalyse the hydroxyl-isocyanate reaction. The reactants were used as-received without further purification.
ing unreactive, unsaturated end groups. The main competing reaction involves the isomerisation of propylene
Reaction injection moulding
oxide into monofunctional ally1 alcohol; subsequent
reaction with propylene oxide results in the formation
RIM materials were moulded as rectangular plaques
of allyl-terminated poly(oxypropy1ene) monols which
(150 x 250 x 3.5mm) using in-house RIM equipment
reduce the overall functionality of the polyol. The polythat has been described in detail elsewhere." The foramines are known32933
to contain primary and secondmulations and processing data are given in Table 2,
ary amino groups and residual secondary hydroxyls,
where Qi and Q , are the machine throughputs used,
with at least 85% of the total being primary aminos.
respectively, for the polyisocyanate and the polyol/
The characterisation data in Table 1 show the respecpolyamine reactant streams, p, and Re, are the viscosity
tive reactive groups of D4000 and T5OOO to comprise 94
and the Reynolds number of the polyol/polyamine reacand 97% amino groups. End-group a n a l y ~ i s ' ~has
,~~
tant stream. It is well established that, to achieve good
found these groups to be almost exclusively primary
mixing in an impingement mixing device, nozzle Reyamines, in agreement with published data.33 GPC
nolds numbers must exceed a critical value which is in
analysis of phenyl-isocyanate-capped T500034 reported
the range of 300-500 for amine-based systems.' The
6000 g mol- indicating a functionalan R,value of
values of Re, for the polyol/polyamine reactant streams
ity of 3.1. This close agreement with the nominal value
were all >700, and the materials were observed to be
off, may reflect some functionalisation of unsaturated
well mixed with no visible striations or gel lines. All the
species under the conditions employed during aminamaterials were formulated to contain 50% by weight
t i ~ n . ~The
'
aromatic diamine chain extender, 3 3 DETDA/MDI hard segments and the stoichiometric
diethyltoluene diamine (DETDA) (111), is an 80 : 20
-
TABLE 2. Formulations and processing details used t o produce
R I M copolymers
~~~
M340"
Prepolymer"
DETDA"
Q , (CIS-')
Qp (gs-')
Ppvas)
Re,
DS25
MI11
PI
04000
T5000
208
265
100
70.1
123.3
0.30
1038
209
264
100
69.3
121.3
0.34
908
206
267
100
67.7
121.0
0.42
734
209
263
100
69.1
120.0
0.26
1182
207
265
100
67.7
119.1
0.28
1093
" The numbers in the table represent the parts by weight of reactants used
in formulations.
Viscosity values were obtained at 50°C which includes t h e temperature
rise (11 f 2°C) in the reactant stream arising from viscous dissipation.28
POLYMER INTERNATIONAL VOL. 41, NO. 3, 1996
286
ratio of isocyanate to total amine/hydroxyl groups used
was 1.03. Reactant pressures of 3000 k 50psi and initial
reactant temperatures of 35 and 40"C, for the polyisocyanate and polyol/polyamine respectively, were used
throughout. Typical35 mould temperatures of 70 and
120°C were used for the PUU and PUr, respectively,
and the plaques were demoulded after a 'cure time' of
100 s. The RIM copolymers produced are designated
using a two-character code referring to the SS prepolymer used. For example, DS is a copoly(urethaneurea) formed using the polyether diol DS25.
Differential scanning calorimetry
Differential scanning calorimetry (DSC) studies were
performed on a DuPont 990 Thermal Analyser fitted
with a DuPont 910 cell base, equipped with a 990 DSC
cell. Samples (12-15 mg) and an inert reference material,
10 mesh glass beads (13 mg), were encapsulated in aluminium pans and cooled rapidly to - 120°C in the cell.
The sample and reference were subjected to a
20"Cmin-' ramp rate in static air to 350°C. The glass
transition temperatures were obtained from the DSC
trace as the intersection of tangents drawn to the onset
baseline and the endothermic slope.
The degree of phase separation for each copolymer
studied was obtained using the method of Camberlin
and P a ~ c a u l tThus
. ~ ~ the heat capacity change, AC;, at
the SS glass transition, T:, was measured and compared to ACP, the heat capacity change of the pure SS
material. The HS fraction of the material was known
and the phase separation ratio (PSR) was established as
reported previo~sly.~-'
Dynamic mechanical thermal analysis
Dynamic mechanical thermal analysis (DMTA) data
were obtained in the range -100 to 300°C using a
Polymer Laboratories apparatus operating at a frequency of 1 Hz and a heating rate of 5"Cmin-'. A
double-cantilever bending geometry was used for beam
samples (3 x 10 x 45mm) to obtain dynamic flexural
moduli and mechanical damping as functions of temperature.
Tensile stressstrain
Tensile stress-strain data were obtained at 20 f 2°C
using an Instron 1122 universal testing machine, with
dumb-bell specimens having an overall length of
150mm, neck length of 50mm, width of 12.5mm and
thickness of 3.5 mm. The crosshead separation gauge
length was 75 mm and the extension rate 10 mm min-'.
The strain in the sample up to 10% was recorded using
a strain-gauge extensometer clamped directly to the
neck of the specimen. Typically, ten specimens were
J . L. Stanford, R . H . Still, A . N . Wilkinson
tested for each material and the derived tensile properties are reported as the mean of at least five tests.
Single edge -notch fracture measurements
The fracture properties of the RIM copolymers were
measured in a single edge-notch tensile (SENT)
geometry using rectangular specimens of length
150mm, width (d) lOmm and thickness (b) 3.5mm.
Tests were conducted at 20 k 2°C using an Instron
1122 universal testing machine at an extension rate of
10mm min- and a specimen gauge length of 75 mm. At
least 20 specimens were tested for each material .and the
nominal notch depth, a, was varied to give a/d ratios in
the range 0.04 5 a/d I0-5. Notches were cut approximately to predetermined depths using a saw and finished with a fresh razor blade to sharpen the notch.
During the tests, the onset of crack propagation was
determined visually and marked on the force-time
curve. The actual notch depths were measured, after
fracture, to within 0.001 mm using a travelling microscope.
RESULTS AND DISCUSSION
Differential scanning calorimetry
The DSC data presented in Table 3 show the trend in
T: values on progressing from isolated SS polyether
networks to phase-separated RIM copolymers. The
table also allows comparison of the effects on T i and
PSR data of the different chemical structure of the func-
TABLE 3. DSC and phase separation data for R I M
copolymers
~~
Copolymers
DS
MI
PI
D4
T5
-68
-63
-64
-68
-62
Networks
M I 11/M340"
PI /M340b
T5000/M340C
-66
-65
-61
81 * 2
72 f 4
66f3
74f2
68f3
Data from ref. 21.
Data from ref. 30.
Data from ref. 10.
derived from at least three traces for each material,
typical 95% confidence limits (CL) f 1"C.
PSR data are averaged from at least nine determinations
for each material and are shown *95% CL.
POLYMER INTERNATIONAL VOL. 41, NO. 3, 1996
287
Efects of soft-segment prepolymer functionality
tional group (i.e. -OH versus -NH2) and of the SS
prepolymer functionality.
The greater potential for hydrogen bonding exhibited
by urea groups compared to urethane groups results in
a greater degree of association in the T5000-based
network. This increased restriction to molecular mobility results in a higher value (-5°C) of T: for the T5000
network, relative to the analogous M111-based
network. Comparing the values of TZ of the networks
with those of the RIM copolymers, only small increases
(-2°C) are observed, indicating that the SS phase
responsible for the transition is relatively pure. The
value of T: for both the SS prepolymers and RIM
copolymers differed by only N 5°C upon increasing the
nominal SS prepolymer functionality from 2 to 4, a
figure which lies within experimental error. In addition,
replacing the poly(oxypropy1ene-oxyethylene) polyols
with analogous polyoxypropylene polyamines had no
effect on TE within experimental error. No further transitions were observed by DSC between T i and the onset
of thermal-oxidative degradation at temperatures
>250°C.
In Table 3, the value of PSR for each material is
determined as the ratio of the heat capacity change of a
RIM copolymer at TE compared to that of the pure SS
network, normalised for sample mass. The value of PSR
is a measure of the fraction of SS which contributes to
heat capacity at T: and therefore does not include SS
trapped either in the HS phase or in the interphase
regions. The use of this simplistic three-phase model,
plus the experimental difficulties associated with measuring accurately the heat capacity change in a segmented block copolymer, limits the use of PSR data to
qualitative comparisons between the RIM copolymers.
The data in Table 3 show only very small differences
between the copolymers, typically just outside experimental error. However, the data are in agreement with
general trends for a wide range of RIM copolymers
Thus, comparison of
reported previously.'-'
the PSR data for the PUU and PUr materials indicates
the latter to be more phase-mixed. This is in agreement
with previously reported4v8-'l PSR data for PUU and
PUr materials, but is contrary to thermodynamic
prediction^^*^^^^ based on calculations of xHS/&, the
ratio of the HS/SS interaction parameter (derived from
the solubility parameters14) to the critical interaction
parameter, xi, for the occurrence of microphase separation in a segmented block copolymer system.37938The
disagreement between predicted and experimental
values can be ascribed to the dynamic nature of the
RIM process and the kinetic competition between
copolymerisation and microphase separation. The reactions which occur during the formation of RIM PUr are
much more rapid than those involved in the formation
of RIM PUU.39 Vitrification therefore occurs more
rapidly in RIM PUr, effectively quenching the system at
a lower degree of microphase separation.
For both PUr and PUU, the value of PSR is
observed to decrease with increasing S S prepolymer
functionality, which may be ascribed to the more rapid
development of molar mass. The concomitant reduction
in molecular mobility within the system will slow the
rate of phase separation relative to the rate of polymerisation, and thus the system will be more phase-mixed
at the onset of HS vitrification.
Dynamic mechanical thermal analysis
The curves in Figs 1 and 2 show storage modulus (E')
and damping (tan 6) versus temperature data for the
PUU and PUr materials, respectively. These curves are
similar to those reported p r e v i o ~ s l y ~ - ~fo
~ r~ -analo"
gous phase-separated RIM copolymers, in that two
major transitions are observed, ascribed to the SS and
HS glass transitions at T i approximately -40°C and
in the range 180 to 250°C. In addition, the damping
curves show a broad, low-intensity transition between
50 and 150°C which is associated with the breakdown
of hydrogen bonding between HS polyurea groups and
the ether oxygens of the SS. Detailed differences in
values of T: and TF for the various PUU and PUr are
given in Table 4.
The mechanical damping data show the value of T:
for the PUU (-42 f 2°C) to be relatively insensitive to
variations in SS functionality, and in good agreement
with DMTA values of T: of -40 f 1°C for homogeneous polyurethane networks derived from M340 and
either M111" or P1.4' These observations indicate that
lo
4
121328934
POLYMER INTERNATIONAL VOL. 41, NO. 3, 1996
0
i
2
24
-150
-75
0
75
150
225
300
Temperature I "C
Fig. 1. Dynamic flexural modulus ( E ) and mechanical
damping (tan 6) at 1Hz versus temperature for
copoly(urethane-urea)s: 0 , DS; 0,
M1; A,P1.
288
J . L. Stanford, R. H . Still, A . N . Wilkinson
~ o " " ' " " " ' " ' " ' " ' '
9.5
higher T: and lower tan S values than for the equivalent
PUU, due to a greater degree of phase boundary mixing
in these rapidly reacting systems. The degree of phase
mixing is also observed to increase slightly with increasing SS functionality, as evidenced by a small increase in
T i and a broadening of the transition peak.
More significant differences between the two RIM
copolymer systems are observed for values of T:. The
HS glass transition in the PUU materials is visible only
as an ill-defined shoulder on a rising tan S curve,
whereas the PUr materials exhibit definite peaks of
higher intensity at much higher temperatures (- 240°C
compared to -190aC), suggesting the presence of
longer HS sequence lengths in these materials. This may
be expected since the increased (- 50°C) mould temperature required to achieve PUr materials, which can
be demoulded without cracking, will delay vitrification
of the HS and thus allow the development of longer
sequence lengths. However, the requirement for higher
-
h
2
.
%
-
9-
OD
- 0.4
8.5
-
8-
~
0.3
2
7.5 -
0.2
-
0.1
7 -
"
I
0
incorporated into RIM PUU systems may also promote
depolymerisation, and indeed added catalyst has deleterious effects on the high-temperature properties of
RIM PUr.41 Thus, it is also possible that the lower
values of TF observed for the PUU are due to the plasticking effects of low-molar-mass degradation products.
The modulus-temperature dependence of the various
RIM copolymers, expressed as the ratio of moduli at
specified temperatures, is also given in Table 4. The
ratio of moduli at -30 and 65°C is often quoted' as an
indicator of the low-temperature modulus-temperature
behaviour of RIM copolymers. The magnitude of this
ratio is dominated by the modulus change associated
with the S S glass transition, the proximity of T: to
-30°C and by the degree of phase mixing present in
the material. Thus, the effect of segmental mixing may
be observed in the variation of the ratio with SS functionality, such that DS < M1 < P1 and D4 < T5 (see
Table 4) and the more phase-mixed PUr materials
exhibit higher ratios than equivalent PUU materials.
Modulus-temperature behaviour at higher temperatures is quantified over an equal temperature inter-
the SS phases responsible for the low-temperature
damping peaks are relatively pure. However, in all three
materials the temperature range of the transition region
is broad (approximately -60 to +45"C compared to
-60 to 0°C for the M111- and P1-based network^''.^^),
indicating a significant degree of domain boundary
mixing. These broad transitions are therefore artefacts
of the rapid vitrification of the system at a relatively
early stage of the microphase-separation process. The
width of these transitions also results in differences in
the values of T: measured by DSC (Table 3) and
DMTA (Table 4), larger than the typical 10-15°C difference exhibited by a block copolymer synthesised in a
relatively slow copolymerisation. The more phasemixed M1 and P1 materials exhibit the broadest transitions, which is a result of the reduction in the rate of
phase separation (with increasing S S prepolymer
functionality) and the consequent onset of HS vitrification at a lower degree of phase separation. In general,
similar low-temperature damping behaviour was
observed for the PUr materials, albeit with slightly
TABLE 4. DMTA data for R I M copolymers
Copolymer
DS
MI
PI
D4
T5
("C)
-44
-40
-41
-37
-34
tan
a
(c)
0.1 38
0.1 55
0.1 68
0.1 36
0.1 20
T," ("C)
184
188
194
243
238
tan
a (T:)
0.255
0.268
0.266
0.368
0.339
E' (-3O0C)/€(65"C)
2.6
3.0
3.2
3.2
4.1
€'(65"C)/E'(160"C)
2.0
2.1
2.0
1.7
1.8
POLYMER INTERNATIONAL VOL. 41, NO. 3, 1996
289
Effects of soft-segment prepolymer functionality
Val by the modulus ratio E' (65"C)lE' (160"C), which,
when compared to the E' (-30"C)/E' (65°C) ratios, are
not only smaller but also show little dependence on SS
functionality. For the PUU, the proximity of T! to
160°C results in significantly higher values of this ratio
than equivalent PUr, which are known41 to exhibit
superior and prolonged, high-temperature dimensional
stability.
0
0
Tensile stress-strain
The tensile properties of the RIM copolymers in terms
of Young's modulus (E), yield and ultimate stresses (by
and cu),yield and ultimate strains ( E and
~
8,) and tensile
toughness ( U ,, the total area under a stress-strain
curve) have been derived from averaged (at least five
determinations for each material) stress-strain curves
and are summarised in Table 5. To aid discussion, the
stress-strain curves in Figs 3 and 4 are presented as
stress versus log extension ratio. This has the effect of
expanding the initial portion of the curves relative to
engineering stress-strain curves.
There are two characteristic regions to the stressstrain curves. At low strains the materials have a high
modulus and the slope of the stress-strain curve is very
0
B
.
20
m
-.-I
0
15
.
'
A
A
DS
M1
P1
5
0
0
0.2
0.1
0.3
h
0.4
0.5
h
0.4
0.6
-
Fig. 3. Tensile stress-strain curves for copoly(urethaneurea)s: 0 , DS; 0,
M1;A,P1.
TABLE 5. Tensile stress-strain properties' of R I M copolymers
DS
MI
PI
D4
T5
436f10
337 f18
296f10
389f13
496f14
15.1 f0.3
b
15fl
1,
b
b
1,
b
16.3f0.2
16f1
27.9f1.6
24.9f 0.6
24.2 f 1 .O
25.7 f 1.4
28.1 f1.2
Data are shown as mean values f95% confidence limits.
Materials exhibited an extrinsic yield point.
POLYMER INTERNATIONAL VOL. 41, NO. 3, 1996
0.5
0.6
steep. All the materials display a yield point, either
intrinsic (DS, T5) or extrinsic (Ml, P1, D4), followed by
extensive post-yield drawing with a much shallower
increase in the stress-strain curve. These two regions
have been i n t e r ~ r e t e das~ ~
being indicative of materials
with a bicontinuous morphology, and the initial part of
the curve is attributed to the deformation and yielding
of the continuous HS phase, whilst the extensive postyield drawing results from deformation of the rubbery
SS phase.
In general, the data in Table 5 show that increasing
SS prepolymer functionality reduces the tensile properties of the PUU, but improves the stiffness and strength
of the PUr materials. The differing tensile stress-strain
behaviour of the PUU and PUr reflect the differences in
copolymerisation conditions between the two systems.
The relatively slow PUU systems exhibit relatively
coarse, highly phase-separated co-continuous morphologies, with relatively little phase connectivity. In the
PUr systems, however, the rapid reaction of the SS prepolymer results in finer-scale, co-continuous morphologies with a relatively high degree of interphase mixing
and strong phase connectivity. Due to these morphological differences, the PUr exhibit higher values of T:
in DMTA (- -36°C compared to -42°C) than the
PUU, and the corresponding inflections in E' extend
v)
lo
0.3
0,T5.
A
0
0.2
Fig. 4. Tensile stress-strain curves for copolyureas: 0 , D4;
25
h
0.1
249f15
194 f 17
197 f 14
203 f 15
170f9
68.1 f2.1
48.3f 0.6
45.4f 2.1
59.4f 0.5
46.9f1.4
290
J. L. Stanford, R. H . Still, A. N . Wilkinson
over a much broader temperature range, reflecting the
higher degree of interphase mixing in the PUr. Thus, at
room temperature, the more phase-mixed PUr effectively contain a greater volume of glassy material, consisting of HS and mixed HS/SS, than equivalent PUU.
The tensile properties of both systems are dominated,
therefore, by the degree of phase separation which
occurs, but with different results. Increased phase
mixing in the PUU results in reduced tensile properties,
as a smaller volume of the HS is located in the glassy,
reinforcing hard phase. In addition, the presence of
phase-mixed HS will reduce the elastomeric nature of
the soft phase, with a concomitant decrease in elongation at break. In contrast, the greater degree of phase
mixing in the T5 PUr results in a greater volume of
glassy material at room temperature, giving higher
stiffness and strength but reduced elongation at break
compared to the D4 material.
Single edge -notch fracture measurements
All the RIM copolymers in the present study exhibited
bulk, non-linear elastic behaviour during the SENT
tests and their fracture behaviour was therefore
analysed using a tearing analysis43 to determine the
critical strain-energy release rate, GI, , or fracture
energy. However, SENT specimens of some RIM
copolymers were observed to undergo bulk plastic
yielding. In order to assess the values of GI, derived
under such conditions, the SENT data of several RIM
copolymers were a n a l y ~ e d to
~ ~determine
.~~
the critical
J-contour integral, JI,. If fracture initiation occurs in a
SENT specimen under conditions where only smallscale yielding occurs around the notch, G = J . However
if elastic-plastic conditions prevail, resulting in gross
specimen-yielding, G < J,46 thus the J,, analysis is more
appropriate and the values of GI, derived give only a
qualitative indication of the relative fracture behaviour
of the RIM copolymers.
The critical strain-energy release rate is defined for
the test geometry used by the relation:43
GI, = 2kW,a
which exhibit bulk non-linear elastic behaviour due to
extensive crack-tip plasticity. For these materials, J may
be simply defined (in terms of energy) as the rate of
decrease of potential energy, U (in the form of the
stored elastic energy in the specimen), with increasing
crack length, a, and the criterion for crack growth is
J 2 J , , where J, is a material property independent of
crack length and specimen geometry. Figure 5 shows
schematic load-displacement curves for specimens
which differ only in initial crack length, a. The solid
circles represent the crack initiation points and the line
A-C the locus of the initiation points. This crackinitiation locus line has been ~ t i l i s e d to
~ ~analyse
.~~
material fracture in terms of J, , using:
)
J c =1- (Area OBC
b a4 - a ,
(2)
Equation (2) indicates that Area OBC (see the hatched
region in Fig. 5) is the energy needed to propagate a
crack from a, to a 4 . By defining an energy function
U,(a) = Area OAC in Fig. 5, eqn (2) can be rewritten as:
(3)
In order to apply the J-analysis to the RIM copolymers, experimental SENT load-deflection curves were
replotted. The crack-initiation locus line was defined
and U , determined using a digital image analyser to
integrate the area elements. Linear least-squares plots of
U , per unit thickness versus a, gave slopes equal to
-JIc. For all the materials tested the plots were linear,
with correlation coefficients of 20.88 and typical 95%
confidence limits of k 10%.
In order to provide limiting values of the respective
fracture parameters, which are independent of the specimen geometry, both the GI, and JI, analyses require the
specimens to be deformed under conditions of plane
such that the strain in the specimen normal to
(1)
where a is the initial crack length, W , is the strainenergy density of the specimen (determined from the
integral of the stress-strain curve between zero deflection and the point of crack propagation) and k is a
strain-dependent constant. An approximate expression
for k has been derived47 as k = n/A:l2, where 1, is the
extension ratio of the specimen at the onset of crack
propagation. Hence, values of G,, were obtained from
linear least-squares plots of 2 k W, versus a-', all of
which were linear with correlation coefficients of 2 0-90,
and values of 95% confidence limits typically f10%.
The J-contour integral4* describes the flow of energy
into a crack-tip region during crack propagation, and
has been used to characterise the failure of materials
Fig. 5. Schematic load ( P ) versus displacement (A) curves of
specimens which differ only in initial crack length, a, in which
u4 > a3 > a, > a,. The line A-C is the crack-initiation locus
line.
POLYMER INTERNATIONAL VOL. 41, NO. 3, 1996
29 1
Effects of sof-segment prepolymer functionality
the applied stress is zero. Thus, thickness criteria must
be met, which for specimens of thickness b are given for
GI, and J , , ,respectively, by:
bmin =
2.5(G,,
(4)
E/C:)
and
bmin
(5)
=25(J,c/~y)
Thus, the thickness requirement to achieve plane strain
conditions in J-tests is not as great and valid plane
strain, fracture energy values can be obtained with
much smaller specimens.
The values of G,, and J,, derived from the SENT
tests, and the respective values of bmin, are given in
Table 6. The RIM mould used in these studies had a
constant thickness of 3.5mm and it can be seen that
material D4 is the only one with specimens whose
dimensions are close to fulfilling the minimum thickness
criteria. The values of GI, and J,, determined at a
sample thickness <bmin are not absolute, owing to a
significant degree of plane stress behaviour, which
results in an overestimation of the fracture parameters.
However, they may be used in internal comparisons to
show trends in RIM copolymer fracture behaviour at a
sample thickness of 3.5mm, which is a common thickness for RIM mouldings. The three PUU exhibit an
average GI, value of 3.4 & 0.3 kJ mP2,which is in close
agreement to that (3.1 kJ m-2) reported previously’ for
a 51% HS DETDA-based PUU. The values of G,, and
J,, determined for the PUU are essentially the same
within experimental error, indicating that variations in
copolymer molar mass and differences in the degree of
microphase separation, which in this case are relatively
small, have little influence on the fracture behaviour of
these materials. In contrast, the PUr materials show a
significant increase in GI, for those formed by substituting a SS diamine (D4000) with an equivalent SS triamine (T5000). As it is difficult to reconcile an increase
in fracture energy with an increase in phase mixing, the
improvement in fracture behaviour is attributed to the
increase in copolymer molar mass within a triamine
system; the use of triamines has been shown5325to
improve the incipient fracture resistance of RIM PUr
compared to diamine-based systems.
The fracture parameters of the PUr materials in
Table 6 are significantly lower than the equivalent
PUU, indicating that the fracture properties of
DETDA-based copolymers are very sensitive to changes
in morphology resulting from the differing reaction conditions. In fact, the fracture surfaces of the D4 and DS
materials, which have the lowest and highest values of
GI, or J,, , respectively, were very different. The fracture
surface of DS revealed gross yielding and tearing, with
the cross-sectional area ahead of the notch being
reduced by -30%. In contrast, the fracture surface of
D4 was smooth, almost featureless, and there was no
discernible draw-down of the specimen, indicative of a
more brittle mode of fracture under conditions approximate to plane strain.
SUMMARY AND CONCLUSIONS
During RIM copolymerisation, significant variations in
the development of molar mass and HS sequence length
occur. These variations are due either to unequal reactivity of the monomers or to changes in SS functionality,
which markedly affect the development of copolymer
structure as the system phase-separates via spinodal
decomposition. In the systems studied, both higher SS
reactivity and SS functionality result in a more rapid
development of copolymer molar mass. Consequently,
increasing SS functionally appears to slightly reduce the
overall degree of phase separation developed in these
materials (as determined by DSC and DMTA), due to
increased domain boundary mixing. Reactivity differences result in RIM PUr which, because of their much
greater SS reactivity than equivalent PUU systems,
exhibit a higher degree of phase-mixing.
RIM PUr were shown to exhibit thermal properties
superior to those of equivalent PUU; this behaviour
TABLE 6 . SENT fracture dataa of R I M copolymers
~~~~
Copolymer
DS
MI
PI
D4
T5
E
(MPa)
0,”
GC
bminc
JC
bmind
(MPa)
(kJm-*)
(mm)
(kJm-’)
(mm)
436
337
296
389
496
15.1*
12.9
11.6
15.0
16.3*
3.7 (10)
3.5 (11)
3.1 (11)
1.8 (10)
2.6 (13)
17.7
17.8
17.0
7.7
12.0
5.1 (12)
4.8 (14)
4.4(14)
2.1 (12)
,
-
8.4
9.3
9.5
3.5
-
The values in parentheses are 95% confidence limits expressed as a percentage
of the mean value.
Yield stresses determined either from intrinsic yield points (marked *) or extrinsic yield points using Considere’s construction.
Calculated from eqn (4).
Calculated from eqn (5).
POLYMER INTERNATIONAL VOL. 41, NO. 3, 1996
292
was attributed to increased interphase connectivity, and
the absence of tin catalysts which promote thermooxidative degradation of the HS. The tensile properties
of both systems were dominated by the degree of phase
separation which occurred, but with different results.
Increased phase mixing in the PUU results in reduced
tensile properties, as a smaller volume of the hard segments were located in the glassy, reinforcing hard phase,
whereas a greater degree of phase mixing in the PUr
results in a greater volume of glassy material at room
temperature, giving higher stiffness and strength but
reduced elongation at break. In addition, SENT fracture studies showed the RIM PUU to exhibit higher
values of GI,and J, than equivalent PUr, reflecting the
stiffer, more brittle nature of the latter materials.
ACKNOWLEDGEMENTS
The authors wish to acknowledge the kind donation of
the reactants used in this study from Lankro Chemical,
Union Carbide Chemical, ARCO, Texaco Chemical,
Lonza AG and Dow Chemical.
REFERENCES
1 Macosko, C. W., Fundamentals of Reaction Injection Moulding.
Hanser, Munich, 1989.
2 Ryan, A. J. & Stanford, J. L., in Comprehensive Polymer Science,
ed. G. Allen & J. C. Bevington. Pergamon Press, Oxford, 1988,
Vol. 5, Ch. 25, p. 427.
3 Ryan, A. J., Polymer, 31 (1990) 707.
4 Ryan, A. J., Still, R. H. & Stanford, J. L., Plast. Rubb. Proc. Appl.,
13 (1990) 99.
5 Willkomm, W. R., PhD Thesis. University of Minnesota, USA,
1990.
6 Willkomm, W. R., Chen, Z. S., Macosko, C. W., Gobran, D. A. &
Thomas, E. L., Polym. Eng. Sci., 28 (1988) 888.
7 Stanford, J. L., Wilkinson, A. N., Lee, D-K. & Ryan, A. J., Plast.
Rubb. Proc. Appl., 13 (1990) 111.
8 Ryan, A. J., Stanford, J. L. & Still, R. H., Polymer, 32 (1991) 1426.
9 Ryan, A. J., Stanford, J. L. & Birch, A. J., Polymer, 34 (1993) 4874.
10 Birch, A. J., Stanford, J. L. & Ryan, A. J., Polym. Bull., 22 (1989)
629.
11 Ryan, A. J., Stanford, J. L. & Still, R. H., Brit. Polym. J., 20 (1988)
77.
12 Macosko, C. W., Fundamentals of Reaction Injection Moulding.
Hanser, Munich, 1989, Section 2.7.
13 Camberlin, Y. & Pascault, J. P., J. Polym. Sci. (Polym. Phys. Edn),
22 (1984) 1835.
J . L. Stanford, R. H . Still, A. N . Wilkinson
14 Ryan, A. J., Stanford, J. L. & Still, R. H., Polym. Comm., 29 (1988)
196.
15 CritcMeld, F. E. & Gerkin, R. M., J . Elast. Plast., 8 (1976) 326.
16 Zdrahala, R. J. & Critchfield, F. E., in Reaction Injection Moulding
and Fast Polymerization Reactions, ed. J. E. Kresta. Polymer
Science and Technology, Vol. 18, Plenum, New York, 1982.
17 Camargo, R. E., PhD Thesis. University of Minnesota, USA, 1984.
18 Camargo, R. E., Macosko, C. W., TirreU, M. & Wellinghoff, S. T.,
Polymer, 26 (1985) 1145.
19 Camargo, R. E., Macosko, C. W., Tirrell, M. & WeUinghoff, S. T.,
in Reaction Injection Moulding, ed. J. E. Kresta. ACS Symposium
Series 270, American Chemical Society, Washington DC, 1985.
20 Nishimura, H., Kojima, H., Yarita, T. & Nishiro, M., Polym. Eng.
Sci., 26 (1986) 585.
21 Lee, D-K., PhD Thesis. Victoria University of Manchester, 1988.
22 Markovs, R. A., J. Cell. Plast., Sept./Oct. (1985) 326.
23 Burchell, D. J. & Porter, J. R., Proc. 32nd SPI Ann. Pu Tech.
Mark. Conz 1989, SPI, 144.
24 Morgan, R. E., Berg, J. W., & Klumb, G. A., Proc. 32nd SPI Ann.
PU Tech. Mark. Confil989, SPI, 262.
25 Stanford, J. L., Still, R. H. & Wilkinson, A. N., Polymer, 36 (1995)
3555.
26 Sorenson, W. R. & Campbell, T. W., Preparative Methods in
Polymer Chemistry. Wiley Interscience, New York, 1981.
27 Ogg, C. L., Porter, W. L. & Willits, C. O., Ind. Eng. Chem. Ann.
Edn, 17 (1945) 394.
28 Wilkinson, A. N., PhD Thesis. Victoria University of Manchester,
1990.
29 Saunders, J. H. & Frisch, K. C., Polyurethanes Chemistry And
Technology, Part 1. Interscience, New York, 1962, p. 32-36.
30 Woods, G., The ICI Polyurethanes Book, 2nd Edn. ICI/Wiley,
1990, p. 35-39.
31 Glasspool, J., Proc. of UTECH ’86, The Hague,
- The Netherlands,
1986, p. 14-19.
32 US Datent. 3 654 370.1972. Texaco Chemical Comuanv.
33 ‘Jeffknine; technical’ data’sheet SC-024, 1991, Tkaco Chemical
Company.
34 Birch, A. J., PhD Thesis. Victoria University of Manchester, 1991.
35 Kirkland, C., Plast. Tech., 32,3 (1986) 83.
36 Camberlin, Y. & Pascault, J. P., J. Polym. Sci. (Polym. Chem. Edn),
21 (1983) 415.
37 Leibler, L., Macromolecules, 13 (1980) 1602.
38 Benoit, H. & Hadziannou, G., Macromolecules, 21 (1988) 1449.
39 Pannone, M. C., M S Thesis. University of Minnesota, USA, 1985.
40 Lyne, J., PhD Thesis, Victoria University of Manchester, 1990.
41 Dominguez, R. J. G., Rice, D. M. & Grigsby, R. A., Plast. Eng.,
Nov. (1987) 41.
42 Bengtson, B., Feger, C., MacKnight, W. J. & Schneider, N. S.,
Polymer, 26 (1985) 895.
43 Rivlin, R. S. & Thomas, A. G., J. Polym. Sci., 10 (1953) 291.
44 Kim. B. H. & Joe, C. R., Int. J. Fract., 34 (1987) 57.
45 Kim, B. H. & Joe, C. R., Polym. Test. 7 (1988) 355.
46 Williams, J. G., Fracture Mechanics of Polymers. Ellis Horwood,
Chichester, 1984, Ch. 5.
47 Lake, G. J., Proceedings of Yield Deformation and Fracture in Polymers. Institute of Physics, London, 1970, p. 531.
48 Rice. J. R., J. Appl. Mech., 35 (1968) 379.
49 Williams, J. G., Fracture Mechanics of Polymers. Ellis Horwood,
Chichester, 1984, Ch. 3.
POLYMER INTERNATIONAL VOL. 41, NO. 3, 1996
Документ
Категория
Без категории
Просмотров
12
Размер файла
944 Кб
Теги
pyrrolidone, interpolymer, resins, miscible, complexes, novolac, effect, behaviour, blend, vinyl, poly, related, crosslinking
1/--страниц
Пожаловаться на содержимое документа