Professional Documents
Culture Documents
ORG
1Department of Chemical Engineering and CERMA, Université Laval, 1065 Avenue de la Médecine,
Quebec, G1V 0A6, Canada
2CRIQ, 333 rue Franquet, Quebec, G1P 4C7, Canada
3Institute for Chemical Technology and Polymer Chemistry, Karlsruhe Institute of Technology (KIT),
Abstract:
The rheological behavior of thermoplastic elastomers (TPE) based on 50/50 recycled ethylenepropylene-diene monomer
(r-EPDM)/polypropylene (PP) was studied to determine the effect of feeding strategy when preparing these blends using
twin-screw extrusion. In particular, small and large deformation characterizations have been performed to better understand
the relationships between sample preparation and final properties of the blends. It was found that small changes in blend
morphology (particle size and interfacial adhesion) are better distinguished in rheological properties (melt state) under large
deformation (LAOS and step shear) compared to small deformation (SAOS).
Key words:
Polypropylene, recycled EPDM, thermoplastic elastomers, large deformation.
to macroscopic properties. From our previous works on ity, and interfacial adhesion. To this end, small and large
PP/r-EPDM blends [18, 19], it was found that coupling deformation characterizations were performed to get a
agent addition is not necessary for this system due to complete understanding of the differences observed in
good interaction between both phases as confirmed by mechanical properties by other means than morpholog-
mechanical and morphological analyses. This was ob- ical analysis (optical, scanning electron, and transmis-
tained through optimization of the processing condi- sion microscopy) which is very tedious. In particular,
tions (flow rate, screw speed, temperature profile, screw nonlinear oscillatory and large deformation have been
design, etc.) and feeding strategy in a twin-screw ex- recently used to study the behavior of different materi-
truder. It was found that the optimum properties were als like silica (SiO2) suspensions [25], bituminous binders
obtained when r-EPDM was introduced in zone 1, while [26] and polypropylene/multiwall carbon nanotube
PP was introduced in zone 4 of a twin-screw extruder (MWCNT) composites [27].
having a total of 10 zones. In this case, feeding the r-
EPDM first led to smaller rubber particles (around 200
μm) and higher elongation at break (155 %) compared to 2 EXPERIMENTAL
the reverse case of feeding the PP first (300 μm and 86 %
elongation at break). 2.1 MATERIALS
In order to differentiate between the processing
conditions tested, it is proposed to use rheological mea- Polypropylene (PP) and recycled ethylene-propylene-di-
surements in the melt state instead of the usual me- ene monomer (r-EPDM) rubber were used in this work.
chanical properties in the solid state (tension, flexion, The polypropylene (melt flow rate = 2.80 g/10 min
impact, etc.). It is expected that the matrix contribution (230 °C and 2.16 kg), melt temperature = 160 °C and den-
is less important in the melt state leading to a higher sity = 0.9 g/cm3) was supplied by LyondellBasell (USA)
contribution (higher sensitivity) of the blend morphol- under the trade name of Montell PF814. The r-EPDM (av-
ogy in terms of the dispersed phase size, dispersion qual- erage initial particle size = 480 μm and density =
1.29 g/cm3) was supplied by Royal Mat Inc. (Canada).
More information on the materials, methods, and char-
acterizations (morphological and mechanical proper-
Sample code Main feed (zone 1) Side feed (zone 4)
PP PP – ties) can be found elsewhere [18, 19].
P-E PP r-EPDM
E-P r-EPDM PP 2.2 BLEND PREPARATION
Table 1: Sample codes and feeding sequences used for sample PP/r-EPDM (50/50) blends were produced in a co-rotat-
produced via twin-screw extrusion. The blends are 50/50 % wt. ing twin-screw extruder (Leistritz ZSE-27, Germany)
with a screw diameter of 27 mm and a L/D ratio of 40
Tensile Tensile Elongation at Flexural Charpy impact (total of 10 zones). The extruder was operated with a
Sample modulus strength break modulus strength flat temperature profile of 180 °C (all the zones) at 150
(MPa) (MPa) (%) (MPa) (J/m)
PP 412 ± 22 32.3 ± 0.4 792 ± 117 1500 ± 170 38 ± 4 rpm with a flow rate of 6 kg/h and a 3 mm circular die.
P-E 149 ± 9 15.8 ± 0.4 85 ± 12 687 ± 9 76 ± 7
E-P 147 ± 2 14.5 ± 0.5 154 ± 13 559 ± 12 162 ± 13
The main feed hopper was at the beginning of the ex-
truder (zone 1) and a secondary feeder (side-stuffer) was
Table 2: Mechanical properties of the samples presented in placed at zone 4. An injection molding machine (Nissei
Table 1. The blends are 50/50 % wt. PS60E9ASE, Japan) was used with a linear screw tem-
Figure 7: Stress-time curves for P-E at different shear rates Figure 8: Stress-time curves for E-P at different shear rates
and 180°C. and 180°C.
sented in Figure 3, the absolute value of the complex r-EPDM contribution. As reported in our previous study
viscosity |h*| at 180 °C gives similar trends. It is clear that [18], feeding the r-EPDM before PP in the extruder pro-
|h*| increased with r-EPDM addition since the recycled duced smaller particle sizes which also improved the
rubber particles are partially crosslinked. adhesion between the matrix and the dispersed phase
Figure 4 (strain sweep tests) shows the shear mod- leading to larger interfacial area. Comparing the results
uli (G’ and G”) for PP at different temperatures, while of Figures 7 and 8, different behaviors between both
the values at 180 °C for the different samples are shown blends can be seen, especially when the shear rate is
in Figure 5. It is observed that G’ and G’’ values for PP/r- higher than 0.1 s-1: P-E samples (Figure 7) have a behav-
EPDM blends are higher than PP over the whole fre- ior similar to neat PP, while E-P samples have a two-step
quency range because of higher interaction between deformation behavior where the second step occurs on-
both phases. Nevertheless, the values of the elastic lim- ly when the shear rate is above 0.1 s-1. In all cases, the
its ge of the blends (Table 3) indicate that a difference corresponding critical total deformation (shear rate
exists between small and large deformation behaviors multiply by the time) is about 3.5 and constant critical
as reported by Filipe et al. [31]. Another way to apply dif- deformation for the overshoot, independent of the
ferent deformation on a sample is via stress-time shear rate applied, has been reported before [32, 33].
curves (shear step) and the results for PP and PP/r-EPDM These trends point again to large deformation being
samples at different shear rates (180 °C) are presented more sensitive than small deformation to small differ-
in Figures 6 to 8. At low shear rates, the stress curves ence in morphology (particle size here) between both
did not show any overshoots, while at higher shear samples.
rates (above 0.6 s-1) all the samples had an overshoot To better quantify the differences between both
before reaching a steady state. Again, the stress values TPE samples at high deformation, LAOS was performed
for PP/r-EPDM blends are higher than neat PP due to to detect any increased nonlinearity by the incorpora-
tion of r-EPDM particles into the PP matrix [29, 34]. The 4 CONCLUSIONS
response was analyzed by FT-rheology which is a very
reliable technique for nonlinear stress data analysis The rheological properties of a polymer blend in the melt
[35]. It has been shown that the relative intensity of the state are controlling the final structure of these blends
third harmonic to its first one I3/1 from FT-rheology and therefore controlling their processability and perfor-
quantifies the transition from the linear to nonlinear mance. Since most polymer processing methods are per-
response [33]. The values of I3/1 can also be used to com- formed at high shear rate/deformation, it is important to
pare different samples and the trends can indicate mor- determine the rheological behavior of the systems under
phological differences in multiphase systems [36, 37]. similar non-linear conditions. Consequently, the rheolog-
For our samples, the values of I3/1 as a function of strain ical behavior must be investigated over a wide range of
amplitude at different excitation frequency (0.5 and deformation, frequency, and temperature. In this work,
1 Hz) are presented in Figures 9 and 10. According to the rheological behavior of PP and PP/r-EPDM blends was
these curves, not only blending PP with recycled rub- studied. Using rheological characterizations under small
bers increased the nonlinearity of the system (higher and large deformation, significant differences were ob-
I3/1 values for P-E and E-P blends), but also different served for TPE blends having similar composition (50/50),
trends in the curves were obtained. For P-E samples, a but processed differently (component feeding order in a
broad peak around g = 1 is usually related to the inter- twinscrew extruder) leading to small changes in their fi-
face deformation created between both phases in a nal morphology. Although their small deformation be-
multicomponent system. This bump seems to increase havior was similar (SAOS), significant differences were
in intensity with increasing frequency (comparison be- observed under large deformation (step shear tests and
tween Figures 9 and 10). As P-E samples have larger par- LAOS combined with FT-rheology in the non-linear
ticle sizes and lower adhesion with the matrix (lower regime). These differences were not only attributed to
mechanical properties) [18], this bump can be associat- small differences in the dispersed phase particle size, but
ed with lower compatibility between the phases in the also to the interaction level (compatibility) between both
blends indicating that LAOS combined with FT-rheolo- phases. Even if small differences were observed in the sol-
gy is a powerful tool to differentiate between samples id state [18], melt rheology was more sensitive to these
having similar rheological behavior at small deforma- differences. This is especially the case under large defor-
tion. As this is a first step, further investigations are mation as LAOS is more sensitive to particle size/geom-
needed in the future. etry, while step shear is more sensitive is more sensitive
to interfacial interactions [31]. Nevertheless, more work
Sample (°C) w c (rad/s) g e (%) needs to be done to completely understand the relation-
PP (180) 12 74 ships between all the parameters involved and the char-
PP (200) 23 76 acterization performed, as well as to get simple models
PP (220) 46 86
P-E (180) 6.3 8.1
for process design, control, and optimization [38].
P-E (200) 9.4 6.1
P-E (220) 36 3.3 Sample aT (200 °C) aT (220 °C)
E-P (180) 5.7 8.4 PP 0.65 0.44
E-P (200) 11 10 P-E 0.50 0.07
E-P (220) 25 11 E-P 0.60 0.27
Table 3: Cross-over frequency wc and elastic limit Table 4: Shift factor aT for the different samples
ge values for the samples at different temperatures. (TREF = 180 °C).