Effect of Degree of Polymerization on Gelation and Flow

Processability of Poly(vinyl chloride)

Mitsuyoshi Fujiyama, Manabu Kondou

Tokuyama Corporation, 1-1, Harumi-cho, Shunan-shi, Yamaguchi-ken, 745-0024, Japan

Received 16 July 2003; accepted 4 November 2003

ABSTRACT: The effect of degree of polymerization (DP) extrusion temperatures and hence with increasing the gela-
on the gelation and flow processability of poly(vinyl chlo- tion. A sample with a lower DP tended to show a larger die
ride) (PVC) was studied. Sheets with adjusted degree of swell and this tendency was even more pronounced at the
gelation were prepared by rolling rigid pipe formulation higher extrusion temperature. The melt fracture easily oc-
suspension PVC compounds with DPs of 800, 1050 and 1300 curred when a sample with advanced gelation was extruded
by changing the milling temperature. Their degrees of gela- at low temperature. Whereas at low milling temperatures a
tion were measured with DSC and their capillary flow prop- sample with the lower DP showed a lower critical shear rate
erties were measured with a capillary rheometer at 150, 170 at onset of melt fracture, and thus easily generating melt
and 190°C and the effect of DP on the relation between fracture, at high milling temperatures it showed a higher
gelation and flow processabilities was studied. Because of critical shear rate and hence scarcely generated melt frac-
the higher shearing heat during milling, the sample with the ture. These experimental results were explained by the fact
higher DP had a higher history temperature and thus tended and concept that a sample with a lower DP shows a higher
to show a higher degree of gelation. The viscosity increased increase in the gelation during extrusion and/or the slighter
as the gelation increased. The dependency of viscosity on DP feature of particle flow as against the uniform molecular
was higher at higher milling and extrusion temperatures flow at the same gelation level. © 2004 Wiley Periodicals, Inc.
and thus at a higher degree of gelation and a lower shear J Appl Polym Sci 92: 1915–1938, 2004
rate. This was assumed to be attributed to the more promi-
nent uniform molecular flow as against the particle flow. Key words: poly(vinyl chloride) (PVC); degree of polymer-
The die swell increased with increasing the milling and ization; gelation; rheology; extrusion

INTRODUCTION ties of PVC are largely varied by the gelation. On the

other hand in an article flow-processed from PVC
A suspension poly(vinyl chloride) (PVC) powder
powder or compound pellets, lamellar crystallites in
has a hierarchical particle structure with lamellar crys-
the primary particles partly melt, diffuse, and recrys-
tallites in the primary particles. At flow processing,
tallize to form fringed micelle crystallites connecting
the particle structure collapses by the action of heat
and shear and approaches a unified homogeneous the primary particles. Accordingly, as the melt-unifi-
structure. Because PVC has a melting point near the cation proceeds and the gelation is advanced, the me-
decomposition temperature, it is generally melt pro- chanical properties are generally enhanced.
cessed at temperatures below the melting point and The material factors that affect the flow processabil-
thus the crystallites partly melt, perfect gelation does ity of PVC are the primary structures, such as molec-
not occur, and the particle structure partly remains. ular weight [degree of polymerization (DP)], its dis-
Accordingly at melt processing, the flow of PVC is tribution, tacticity, copolymerization, and crosslink-
caused not only by the “uniform flow” by the mutual ing; the formulations, such as thermal stabilizers,
movement of molecular chains but also by the “parti- lubricants, reinforcing agents, and fillers; and the par-
cle flow” by the slipping and rolling of particles (pri- ticle structures, such as particle size, its distribution,
mary particles).1–3 The proportion of uniform and par- porosity, and bulk density, for example. These factors
ticle flow is changed by the degree of gelation and affect the gelation characteristics and affect directly or
flow conditions. In general, the higher gelation and indirectly the flow processability through the change
flow temperature promote uniform flow. Accordingly, of the gelation. In a previous study,4 the degree of
as shown in a previous study,4 the flow processabili- gelation of a rigid pipe formulation PVC with a DP of
1050 was adjusted by changing the roll-milling tem-
perature and time, the capillary flow characteristics
Correspondence to: M. Fujiyama (m-fujiyama@mx51. were measured on the samples well-characterized in
tiki.ne.jp).
the gelation, and the effects of the gelation on the flow
Journal of Applied Polymer Science, Vol. 92, 1915–1938 (2004) processabilities were systematically studied. In the
© 2004 Wiley Periodicals, Inc. present article, DP (molecular weight) was chosen and
1916 FUJIYAMA AND KONDOU

its effect on the relation between gelation and flow loss, increases with increasing shear rate. It is higher
processabilities was studied with respect to flow pro- as DP is higher at high extrusion temperatures of
cessabilities, and capillary flow properties, such as about 200°C; its dependency on DP becomes weak and
viscosity, flow activation energy, die swell, melt frac- complex with lower extrusion temperatures, and be-
ture, and entrance pressure loss were studied. Because comes independent of DP at low extrusion tempera-
the meanings of these capillary flow properties in tures of about 150°C.9,12
processability were described in the previous article,4 The critical shear rate at onset of melt fracture in-
they are not repeated here. creases with increasing the extrusion temperature12
As for the effect of DP on the gelation characteris- and decreases with increasing DP.12,13 The critical
tics, it is reported that the gelation time5 and temper- shear stress scarcely depends on the extrusion temper-
ature6 at the milling of suspension PVCs by use of a ature and decreases with increasing DP.12
Brabender plastograph (Brabender Instruments, South The critical draw ratio, which is a measure of melt
Hackensack, NJ) linearly increase with increasing DP. extensibility, decreases with increasing DP18 and a
It is also shown that the DSC low-temperature endo- sample with higher DP shows inferior melt extensibil-
therm of an extrudate, which is a measure of gelation, ity.
is concomitantly lower as DP is higher at the same As mentioned above, a few studies have reported
processing temperature and, conversely speaking, a on the effect of DP on the gelation characteristics and
sample with a higher DP must be processed at a numerous studies have reported on the effect of DP on
higher temperature to obtain the same gelation level.7 the rheological characteristics. However, few studies
Next, experimental results reported so far on the have reported on the latter from the viewpoint of the
effect of DP on flow processabilities (rheological prop- former, that is, on the effect of DP on the relation
erties) are described. between the gelation and rheological characteristics.
Viscosity increases with increasing DP2,8 –12 and the The present study focused on just this aspect: that is,
dependency on DP is stronger at a higher extrusion capillary flow properties were measured on sheets
temperature at a lower shear rate for the lower syn- with adjusted degree of gelation by rolling rigid pipe
diotactic PVC. There are cases in extrusions at low formulation PVC compounds with various DPs at var-
temperatures and at high rates where the viscosity ious milling temperatures and the effect of DP on the
does not depend on DP8,9,11,12 and it inversely de- relation between gelation and flow processabilities
creases with increasing DP.8 As for the non-Newto- was studied.
nian nature of flow, it generally becomes more notable
with increasing DP and with lower extrusion temper-
atures.8 –10,12,13 The non-Newtonian dependency on EXPERIMENTAL
the DP of a milled sample is weaker than that of a Samples
powder sample.8 With respect to the flow activation
energy, one at a constant shear rate decreases with Straight suspension PVC powders, grade ZEST 800Z
increasing DP and shear rate13–16 and its dependency (DP ⫽ 800), 1000Z (DP ⫽ 1050), and 1300Z (DP
on DP diminishes with increasing DP13,15 and be- ⫽ 1300) produced by Shin Dai-ichi Vinyl Corp. (Japan)
comes independent of DP above about 800.10,13 The were used as raw materials. The DP of PVC is gener-
flow activation energy at a constant shear stress in- ally controlled by changing the polymerization tem-
creases with increasing DP and its dependency on DP perature. DP increases as the polymerization temper-
diminishes with increasing DP.15 ature decreases. The stereoregularity of syndiotacticity
Die swell is generally larger when a polymer with a also increases as the polymerization temperature de-
lower DP is extruded at a higher tempera- creases. Accordingly, the crystallinity of PVC in-
ture.2,3,9,12,17–19 A log–log plot of (die swell ratio – 1) creases with increasing DP.21 We measured the crys-
versus viscosity for extrudates obtained at various tallinity indices of three sample powders using the
extrusion temperatures gives a linear relation with X-ray diffraction method proposed by Mitani et al.22
negative slope and the absolute value of the slope Their values were 0.29, 0.28, and 0.27 for 800Z, 1000Z,
increases with increasing DP.13 and 1300Z, respectively, which are almost the same
Entrance pressure loss is higher as DP and the mill- independent of DP in the range of the present inves-
ing temperature are higher7,9,20 and the entrance pres- tigation.
sure loss of extrudate milled at low extrusion temper-
atures below 180°C shows only a scarce dependency
Formulation
on DP at reextrusion at low temperatures below about
150°C.20 The dependency of the entrance pressure loss A rigid pipe formulation was prescribed. The follow-
on the DP of extrudate is weaker than that of the ing additives were added to 100 weight parts (phr) of
powder.7 The end correction coefficient, which is a the PVC powder (in phr): tribasic lead sulfate, mono-
nondimensional measure of the entrance pressure hydrous, 1.5; lead stearate, 1.0; stearic acid, 0.2; cal-
EFFECT OF DEGREE OF POLYMERIZATION ON PVC 1917

Figure 1 Measurement of degree of gelation by DSC.

cium stearate, 0.2; and acrylic processing aid, 0.5. ated by the history temperature Tc, mentioned in the
These components were cold blended without heating next section.
in a 20-L Henschel mixer at high speed for 1 min and
a powder compound was obtained.
Measurement of degree of gelation
To check the thermal degradation by DSC measure-
Roll-milling
ment, thermogravimetric analysis (TGA) was mea-
A 150 g sample of the powder compound was milled sured with the same apparatus on the same sample as
by use of a 6-in. test roll at 150, 160, 170, 180, 190, and used in DSC measurement as preliminary experiment.
200°C for 1000Z and at 160, 180, and 200°C for 800Z Because it became clear that the thermal degradation
and 1300Z for 5 min after the material rounded the roll began to occur above about 250°C, the DSC measure-
and sheets of 0.6 – 0.7 mm thickness were prepared. ment was carried out below 240°C.
The roll temperatures are the set points of roll surface A circular specimen of about 10 mg was cut from
temperatures. The rotation speeds of the front and the rolled sheet and packed in a sealing Al sample
back rolls were 17 and 21 rpm, respectively, and the pan. It was put in a DSC apparatus (DSC6200R-type
mill gap was 0.4 mm. A rolling bank of material was manufactured by Seiko Instruments Inc., Tokyo, Ja-
observed. The folding back was carried out about 50 pan) and the temperature was raised from 30 to 240°C
times during 5 min of milling. The basis of using a at a rate of 20°C/min and a DSC thermogram was
constant 5 min after the material rounded the roll is as obtained. To prevent the thermal decomposition of
follows: given that the degree of gelation and the PVC, the temperature was suddenly decreased to 30°C
appearance of the rolled sheet changed only slightly at a rate of 80°C/min as soon as it reached 240°C. The
when the milling time was changed between 5 and 15 effect of annealing on gelation was studied as follows.
min as shown in the previous article,4 the shortest The powder compound of about 10 mg was packed in
time was chosen to suppress thermal degradation. The a sealing Al sample pan and placed in a DSC furnace.
purpose of rolling in the present experiment was to The temperature was raised from 30°C to annealing
adjust the degree of gelation—a shorter rolling time is temperatures (160, 170, 180, 190, and 200°C) at a rate of
better if the rolled sheet is uniform in gelation. We did 40°C/min, held at the annealing temperatures for 5
not measure the resin temperature during rolling. The min, and decreased to 30°C at a rate of 40°C/min.
resin temperature was assumed to change with time in After this annealing process, the DSC thermogram
a complex manner according to the roll temperature was measured by the same manner as described
and the DP of the resin. However, the resin tempera- above.
ture (maximum temperature reached) can be evalu- Figure 1 shows an example DSC thermogram of
1918 FUJIYAMA AND KONDOU

1000Z sheet roll-milled at 180°C. The low-temperature

endotherm ⌬Hm(A) was obtained as an area envel-
oped by the thermogram and a base line drawn so as
to touch the shoulder around 120°C and the history
temperature Tc. The high-temperature endotherm
⌬Hm(B) was obtained as an area enveloped by the
thermogram and a base line drawn so as to connect Tc
and 230°C (fixed). The degree of gelation was obtained
as the absolute value of ⌬Hm(A) or as the ratio G
⫽ ⌬Hm(A)/[⌬Hm(A) ⫹ ⌬Hm(B)]. It was assumed that
the low-temperature endotherm ⌬Hm(A) originated
from the fusion of fringed micelle crystals that melt- Figure 2 Measurement of entrance pressure loss ⌬Pent.
recrystallized during processing, that the high-tem-
perature endotherm ⌬Hm(B) originated from the fu- After the extrudate solidified, its diameter D at the
sion of lamellar crystals that did not melt during pro- part 5 mm from the front was measured with a mi-
cessing, and that the history temperature Tc was the crometer and the ratio D/D0, where D0 is the capillary
maximum temperature experienced by the resin.23,24 diameter, was designated the die swell ratio and used
DSC measurement was carried out five times on five as a measure of die swell. The standard deviation of
specimens taken from various parts of the rolled sheet die swell measurement was below about 3%.
for each sample and the averaged value was adopted. The appearance of the extrudate was observed with
The maximum standard deviation of ⌬Hm(A) was the naked eye and the occurrence of melt fracture was
about 0.5 J/g at a roll temperature of 190°C and that of checked. The photograph of the extrudate was taken
G was about 0.1 at a roll temperature of 170°C. at a magnification of ⫻7.5 under a real-image micro-
scope (SZH-type manufactured by Olympus Optical
Co., Tokyo, Japan). The standard deviation of mea-
Measurement of capillary flow properties
surement of critical shear rate at onset of melt fracture
The relation between the volumetric flow rate Q and was below about 50%.
the extrusion pressure P was measured with a con- As shown in the top part of Figure 2, when the PVC
stant-extrusion speed plunger-type capillary rheome- melt enters from a reservoir with a larger diameter
ter Capirograph 1B type (Toyo Measurement Instru- into a capillary with a smaller diameter (D0 ⫽ 2R) and
ments Co., Japan) at 150, 170, and 190°C by use of a flat length L, an entrance pressure loss ⌬Pent occurs as a
die with a capillary length L of 20 mm, a capillary result of the change of flow lines, as shown in the
diameter (D0 ⫽ 2R where R is the radius) of 1 mm, and bottom part of the figure. The entrance pressure loss
L/R ⫽ 40. The barrel diameter was 9.55 mm. The ⌬Pent mainly originates from the melt elasticity and in
sample amount was 20 g and the preheating time was the case of PVC, it shows the degree of unification
5 min. The measurement was performed with increas- (homogenization) of resin and can be used as a mea-
ing flow rate. One run of measurements was finished sure of gelation. Although ⌬Pent can be obtained as the
in 30 min. extrusion pressure measured by use of a die with L of
The apparent shear stress ␶, the apparent shear rate 0, the strength of the die with zero length cannot
␥˙ at the wall, and the apparent viscosity ␩ in capillary endure the extrusion force. Therefore, in the present
flow are given respectively in the following equations: experiment, the extrusion pressure was measured by
use of a flat die with a small length L (⫽1 mm) as
PR against the diameter (D0 ⫽ 2 mm) and the pressure
␶⫽ (1) was approximately equated with the entrance pres-
2L
sure loss. To prevent the effect of thermal history at
4Q measurement, the measurement was carried out at
␥˙ ⫽ (2) 140°C, which is 10°C lower than the lowest roll-mill-
␲R3
ing temperature of 150°C. The shear rate at measure-
ment of ⌬Pent was a single point of 7.6 s⫺1. Because the
␶ extrusion pressure changed slightly by the amount of
␩⫽ ˙ (3)
␥ resin in the reservoir, it was measured at a piston end
position 110 mm above the die. The sample amount
From these, the flow curve (␶–␥˙ or ␩–␥˙ ) can be ob- was 15 g and the preheating time was 5 min. The
tained. Although the pressure includes the entrance standard deviation of entrance pressure loss measure-
pressure loss, it was neglected because the die with a ment was below about 5%.
large L/R of 40 was used. The standard deviation of Because the thermal degradation of PVC compound
viscosity measurement was below about 5%. was not negligible at temperatures above 200°C, the
EFFECT OF DEGREE OF POLYMERIZATION ON PVC 1919

Figure 3 DSC curves of 1000Z sheets rolled at various temperatures.

capillary flow measurements were carried out at tem- minates around 200°C. 800Z seems to show slightly
peratures not exceeding 190°C. Although there might lower gelation than that of 1000Z and 1300Z. Figure
be a possibility that PVC melt slips at the die wall 6(a) shows the dependency, on the roll-milling tem-
during flow, the data were treated as no slip. Further- perature, of the low-temperature endotherm ⌬Hm(A),
more, although shear heating might occur at high which has been used as a measure of gelation since the
shear rates, it was neglected. earliest studies on gelation. Although a large differ-
ence is not observed, it tends to decrease with decreas-
RESULTS AND DISCUSSION ing DP except for a single point of 1300Z milled at
200°C. From Figures 5(a) and 6(a), it may be said that
Gelation a sample with lower DP tends to show lower gelation
Figure 3 shows the change of DSC thermogram with at the same milling temperature as a significant ten-
the roll-milling temperature for 1000Z. With increas- dency. This is assumed to be because a sample with
ing temperature, the low-temperature endotherm lower DP generates less shearing heat during milling
⌬Hm(A) appears, gradually increases, and the high- and shows lower history temperature Tc, as shown in
temperature endotherm ⌬Hm(B) gradually decreases. Figure 4(a). Although this tendency diminishes when
At the same time, the history temperature Tc in- the degree of gelation is plotted against the history
creases. temperature Tc, this tendency (the gelation decreases
Figure 4(a) shows the change of history temperature with decreasing DP) still remains. Obande and Gil-
Tc with the roll-milling temperature. Tc increases lin- bert7 reported that the degree of gelation evaluated by
early with increasing milling temperature and is about the low-temperature endotherm ⌬Hm(A) of a suspen-
10°C higher than the roll temperature. Tc tends to be sion PVC extruded by use of a screw extruder de-
lower for lower DP. This is assumed to be because a creases with increasing DP, compared with that at the
sample with lower DP generates less shearing heat same history (extrusion) temperature. This experimen-
during milling and attains a lower maximum temper- tal result is an inverse tendency of the present one for
ature because of the lower viscosity. Figure 4(b) shows roll milling. This difference originates in the milling
the change of Tc with the annealing temperature of machines: whereas they used a screw extruder, we
powder compounds. Tc increases linearly with in- used a roll. Because the milling time at the melt mill-
creasing annealing temperature and changes only ing part in the screw extruder is shorter than that in
slightly by DP. This means that the temperature expe- the roll and the shearing force of the former is weaker
rienced during annealing in DSC scarcely changes by than the latter, a large portion is supposed to gelate by
DP. the action of only heat and not to attain an equilibrium
Figure 5(a) shows the dependency, on the roll-mill- degree of gelation at that temperature. Because the
ing temperature, of the degrees of gelation G gelation time and temperature are longer and higher,
⫽ ⌬Hm(A)/[⌬Hm(A) ⫹ ⌬Hm(B)] of samples with var- respectively, for a sample with higher DP (as shown
ious DPs. The gelation begins around 160°C and ter- by Suzuki5 and Gonze6), it turns out that a sample
1920 FUJIYAMA AND KONDOU

Figure 4 Dependency of history temperature Tc on (a) roll temperature and (b) annealing temperature.

with higher DP gives an extrudate with a lower degree lower gelation for a sample with higher DP. Contrary
of gelation, assuming that the gelation in a screw to this, because it is considered that the gelation
extruder is still in a state of progression. progresses sufficiently to the equilibrium state in the
Figures 5(b) and 6(b) show the dependencies, on the roll milling, a sample with higher DP, which has a
annealing temperature, of the degree of gelation G and higher viscosity and thus is better milled, is assumed
the low-temperature endotherm ⌬Hm(A) of powder to show a higher degree of gelation. Based on the data
compounds, respectively, which may be regarded as of Figures 5(b) and 6(b) of powder compounds, the
degrees of gelation caused by the action of only heat fact that the dependencies of G [Fig. 5(a)] and ⌬Hm(A)
and of no shear. They increase with increasing the [Fig. 6(a)] of rolled sheets on DP are not so fine and
annealing temperature and decrease with increasing somewhat complex, although they tend to increase
DP; that is, they show that the gelation by the action of with DP, can be explained by a balance of the contra-
heat alone occurs only with great difficulty with in- dictory effects of shearing heat and natural gelation
creasing DP, which is also assumed to be a cause of characteristic.
EFFECT OF DEGREE OF POLYMERIZATION ON PVC 1921

Figure 5 Dependency of degree of gelation G on (a) roll temperature and (b) annealing temperature.

Flow curve (viscosity) reciprocal absolute temperature T. Straight lines were

obtained and the Arrhenius equation represented by
Figure 7 exemplifies the change, with the extrusion
the following form holds:
temperature, of a flow curve represented in the form
of the change of viscosity with shear rate for 1000Z
sheet milled at 180°C. The viscosity decreases with ⌬Ha
␩ ⫽ A exp (4)
increasing the extrusion temperature and shear rate, RT
showing non-Newtonian behaviors. The degree of the
decrease of viscosity with increasing the shear rate where A is a constant independent of temperature, R
(i.e., its non-Newtonian nature) changes at shear rates is the gas constant, and ⌬Ha is the flow activation
around 200 s⫺1, shows breaking points, and the non- energy.
Newtonian behavior of the high-shear rate side was Figure 8(a), (b), and (c) show the variation of ⌬Ha
more pronounced than that of low-shear rate side. obtained from the Arrhenius plots with shear rate and
To study the temperature change of viscosity, the milling temperature for 800Z, 1000Z, and 1300Z, re-
viscosity ␩ at each shear rate was plotted against the spectively. ⌬Ha shows maxima at shear rates around
1922 FUJIYAMA AND KONDOU

Figure 6 Dependency of low-temperature endotherm ⌬Hm(A) on (a) roll temperature and (b) annealing temperature.

102 s⫺1 and then decreases; 800Z shows weaker max- range of 0.3–150 s⫺1 and evaluated the changes of the
ima than those of the others. Samples milled at high flow activation energy with shear rate and stress. The
temperatures and having enhanced gelations show activation energies at low- and high-temperature re-
high ⌬Ha values at high shear rates around 104 s⫺1. As gions differ from each other and the former is lower
for the effect of gelation, samples milled at high tem- than the latter; whereas the latter decreases with shear
peratures and having enhanced gelations tend to rate, the former shows a maximum at a shear rate
show high ⌬Ha values. This tendency is particularly around 15 s⫺1. The temperature at transition from the
notable at high milling temperatures of 190 and 200°C. low-temperature region to the high-temperature re-
Figure 9 shows the comparison of ⌬Ha among samples gion increases with increasing shear rate. They regard
in the case of milling at 160°C. ⌬Ha scarcely depends the flow at the low-temperature region with low acti-
on DP. The same results were obtained in the cases of vation energies as a flow of domain or particulate and
other milling temperatures. that at the high-temperature region with high flow
Collins and Krier25 measured the viscosity of an Sn activation energies as a flow of molecular chains. They
formulation rigid suspension PVC compound in a assume that the fact that the transition temperature
temperature range of 160 –240°C and a shear rate from the particulate to molecular flows increases with
EFFECT OF DEGREE OF POLYMERIZATION ON PVC 1923

Figure 7 Temperature change of flow curve.

shear rate is because, with increasing shear rate, the with milling temperature, of flow curve of 1000Z mea-
molecular orientation increases and the dissolution to sured at 150, 170, and 190°C, respectively. At shear
molecular chains becomes more difficult. The extru- rates below the breaking points, the viscosity tends to
sion temperature range was 150 –190°C and the shear increase with increasing milling temperature and thus
rate range was 10 –10,000 s⫺1 in the present investiga- the gelation. This tendency is even more pronounced
tion, which enters in the low-temperature region of at lower extrusion temperatures and gradually be-
the experiment by Collins and Krier. The change of the comes virtually unrecognizable with increasing extru-
flow activation energy with shear rate in the present sion temperature. Similar results were also obtained
study shows maxima like those in their experiment. for other samples.
The shear rate at the maxima was around 100 s⫺1 in The fact that the viscosity increases with increas-
the present study, whereas it was around 15 s⫺1 in ing roll-milling temperature and thus with en-
their result, the former being slightly higher. The ab- hanced gelation is assumed to be because with en-
solute values of the activation energy in the present hancing the gelation the particle structure collapses
experiment were between those of their high- and and the particle flow shifts to a uniform flow. From
low-temperature regions. These facts are assumed to the facts that the change of gelation by the milling
be attributed to the differences in DP and formulation. temperature strongly affects the flow curve at low
The fact that samples milled at high temperatures and extrusion temperatures and scarcely affects it with
having enhanced gelations tend to show high flow increasing extrusion temperature, it is assumed that
activation energy in the present experiment was as- gelation proceeds during the extrusion (measure-
sumed to be because, with increasing gelation, the ment): Because gelation scarcely proceeds at low
particle structure collapses and approaches a unified extrusion temperatures, gelation of the original
homogeneous structure, which leads to the molecular sample affects the flow curve. On the other hand,
flow at the high-temperature region of Collins and because the gelation proceeds well at high extrusion
Krier. This inference is consistent with the results of temperatures, the difference in gelation of the orig-
viscosity, die swell, melt fracture, and entrance pres- inal samples becomes slight and thus the effect of
sure loss described below. As for the dependency of gelation of the original sample becomes less.
the flow activation energy on DP, it is reported that Figure 11(a), (b), and (c) show the changes, with DP,
the flow activation energy at the constant shear rate of flow curves of sheets milled at roll temperatures of
decreases with increasing DP and shear rate13–16 and 160, 180, and 200°C, respectively. Figure 12(a), (b), (c),
its dependency on DP diminishes with increasing and (d) show the dependencies, on DP, of viscosities at
DP13,15 and becomes independent of DP above about shear rates of 12.2, 122, 1220, and 12,200 s⫺1, respec-
800.10,13 Because DPs of the present study were above tively, using the roll-extrusion temperatures as param-
800, it is reasonable that the flow activation energy eters. The viscosity increases with increasing DP and
only slightly depends on DP. its dependency on DP is stronger at higher roll tem-
Next, the change of flow curve with gelation will be perature and lower shear rate. This is the direction of
examined. Figure 10(a), (b), and (c) show the changes, transition from the particle to uniform molecular flow
1924 FUJIYAMA AND KONDOU

Figure 9 Comparison of change of flow activation energy

⌬Ha with shear rate among samples rolled at 160°C.

as shown in the previous study.4 Because the particle

flow is a flow caused by the slipping and rolling of
particles (primary particles), molecular structures
such as DP do not affect it and thus DP scarcely
influences the viscosity when the feature of particle
flow is prominent. On the other hand, in the case of
uniform molecular flow, because molecular chains
are well entangled and the flow is caused by the
mutual movement of molecular chains, the viscosity
is higher as DP is higher, like the usual resins. In
this way, it can be explained that the dependency of
viscosity on DP becomes more notable when the
proportion of the uniform flow, as against the par-
ticle flow, increases. It is clear from Figure 12(a),
where the most prominent uniform molecular flow
is assumed to occur, that the dependency of viscos-
ity on the roll-milling temperature is higher as DP is
higher and the extrusion temperature is lower. In
the case of an extrusion temperature of 190°C, the
viscosity of 1300Z with the highest DP is higher as
the milling temperature is higher, and thus the de-
gree of gelation is greater as in the case of extrusion
temperatures below 170°C, although the degree is
low. 1000Z shows similar viscosities independent of
the milling temperature; 800Z shows a reverse ten-
dency, where the viscosity is lower as the milling
temperature is higher and thus the degree of gela-
tion is higher. This means that when 800Z with the
lowest DP is extruded at low shear rates such as 12.2
s⫺1, the viscosity decreases with increasing gelation;
that is, it is said that the dependency of viscosity of
PVC on DP differs depending on DP, extrusion
temperature, and extrusion rate. Although antipo-
dal results, where, with increasing gelation, the viscos-
Figure 8 Change of flow activation energy ⌬Ha with roll tem-
perature and shear rate for (a) 800Z, (b) 1000Z, and (c) 1300Z. ity increased1– 4,26,27 and decreased,8,28,29 are reported as
EFFECT OF DEGREE OF POLYMERIZATION ON PVC 1925

Figure 10 Flow curves of 1000Z sheets milled at various roll temperatures measured at (a) 150°C, (b) 170°C, and (c) 190°C.
1926 FUJIYAMA AND KONDOU

Figure 11 Comparison of flow curve among samples rolled at (a) 160°C, (b) 180°C, and (c) 200°C.
EFFECT OF DEGREE OF POLYMERIZATION ON PVC 1927

Figure 12 Dependency of viscosity on degree of polymerization at (a) ␥˙ ⫽ 12.2 s⫺1, (b) ␥˙ ⫽ 122 s⫺1, (c) ␥˙ ⫽ 1220 s⫺1, and
(d) ␥˙ ⫽ 12,200 s⫺1.
1928 FUJIYAMA AND KONDOU

Figure 12 (Continued from the previous page)

shown in the introduction of the previous article,4 this effects of DP and gelation on the viscosity robustly ap-
difference is assumed to originate from the differences in pear at low shear rates in extrusion processing, whereas
DP, extrusion temperature, and extrusion rate. Further- these effects decrease at high shear rates in injection
more, it must be noted that in actual processing that the molding as shown in Figures 11 and 12.
EFFECT OF DEGREE OF POLYMERIZATION ON PVC 1929

Die swell unification proceeds, PVC melt becomes elastic be-

cause of the well-entangled molecular chains, and
Figure 13(a), (b), and (c) show the changes, with shear
stores large elastic recoverable strain at the entrance of
rate, of die swell ratio D/D0 of 1000Z sheets milled at
capillary. Accordingly, the elastic recovery after
various temperatures and extruded at 150, 170, and
emerging from the capillary is substantial, leading to a
190°C, respectively. In the cases of extrusions at 150°C
large die swell. The effect of milling temperature on
[Fig. 13(a)] and 170°C [Fig. 13(b)], melt fractures as
the die swell is more prominent than that on the
mentioned later occurred and smooth extrudates
viscosity mentioned in the preceding section. It can
could not be obtained for the samples milled at 190
thus be said that the die swell, which is a melt elastic
and 200°C and hence their die swells could not be
property, is more significantly affected by the gelation
evaluated. For the extrusion at 150°C [Fig. 13(a)], the and in its turn by the flow mode than the viscosity,
samples milled at 150 and 160°C show a similar and which is a melt viscous property.
small die swell and the die swell increases with in- Figure 14(a), (b), and (c) show the changes, with
creasing the milling temperature from 170 to 180°C. shear rate, of die swell ratio D/D0 of each sample
For the samples milled at 150 and 160°C, because the sheet milled at 160°C and extruded at 150, 170, and
degrees of gelation are both 0, as shown in Figure 5(a), 190°C, respectively. The die swell increases with in-
the die swell curves show very similar shapes and creasing extrusion temperature. This increase is as-
values. For the extrusion at 170°C [Fig. 13(b)], the sumed to originate from the increase in gelation dur-
samples milled at 150 and 160°C also show similar and ing extrusion. With respect to the dependency of die
small die swell ratios that are slightly larger than those swell on DP, the die swell is higher as DP is lower and
of extrusion at 150°C. The die swell increases with the dependency becomes more notable with elevating
raising the milling temperature from 170 to 180°C. For the extrusion temperature. From this fact, it is as-
the extrusion at 190°C [Fig. 13(c)], the samples milled sumed that the increase in gelation during extrusion is
at temperatures below 180°C show similar die swell even more pronounced as DP is lower and/or that the
curves and the samples milled at 190 and 200°C show feature of the particle flow is weaker as DP is lower.
higher values. The gelations of all samples milled at Figure 15 shows results for the case of a milling tem-
temperatures below 180°C are assumed to be in- perature of 180°C, showing similar tendencies to the
creased to that at 190°C during extrusion at 190°C. case of the milling temperature of 160°C, with the die
Other experiments showed that the resin temperature swell generally being enhanced. Figure 16 shows a
increases by about 10°C higher than the roll tempera- result for the case of a milling temperature of 200°C
ture. It may be observed from Figure 13(a)–(c) that the and an extrusion temperature of 190°C. The die swell
die swell ratios of samples milled at temperatures further increases and becomes almost independent of
greater than 10°C lower than the extrusion tempera- DP. In this case, because both the milling and extru-
ture increased to that of a sample milled at a temper- sion temperatures are high, gelation is sufficiently en-
ature about 10°C lower than the extrusion tempera- hanced as shown in Figure 5(a), and thus the flow
ture: The die swell ratio of the sample milled at 150°C there approaches a uniform molecular flow; the flow
is increased to that at 160°C in the extrusion at 170°C behavior is assumed to approach that for the usual
and the die swell ratios of the samples milled at 150, resins where the die swell increases with increasing
160, and 170°C are increased to that at 180°C in the molecular weight. It is assumed that when the milling
extrusion at 190°C. The die swells of the samples and extrusion temperatures further increase and gela-
milled at 190 and 200°C could be measured in the tion is further enhanced, the die swell increases with
extrusion at 190°C [Fig. 13(c)] and there is almost no increasing DP (i.e., Fig. 16 is considered to depict the
difference between them. The reason for this is be- ongoing change). Figure 17 shows the dependency of
cause the degrees of gelation G of both samples are the die swell ratio at a shear rate of 122 s⫺1 on DP by
nearly 1.0; they are sufficiently gelated and thus show use of the milling– extrusion temperatures as param-
similar die swells. eters. Because the changes of die swell ratio with shear
The fact that the die swell ratio increases with in- rate show complex tendencies, with valleys and peaks
creasing milling temperature is assumed to be be- as shown in Figures 14 –16, such plots may be mean-
cause, with increased gelation, the particulate struc- ingless. However, they have a merit of being able to
ture collapses and approaches the unified homoge- see the dependency of die swell ratio on DP at a
neous structure. In the case of insufficient gelation, glance. Figure 17 shows that the die swell decreases
because PVC flows in a slippery manner by sliding with increasing DP in almost all conditions. This ten-
and rolling in the particulate state, its flow is a plastic dency diminishes at DPs above 1000Z and a reverse
flow and the storage of elastic recoverable strain at the tendency can be seen at maximum milling– extrusion
entrance of capillary is small, and thus the elastic temperatures of 200 –190°C.
recovery at the capillary exit is weak, leading to a As mentioned in the introductory remarks, it has
small die swell. With increasing gelation, the melt- thus far been reported that the die swell is larger as a
1930 FUJIYAMA AND KONDOU

Figure 13 Change of die swell ratio D/D0 with roll temperature and shear rate measured for 1000Z at (a) 150°C, (b) 170°C,
and (c) 190°C.
EFFECT OF DEGREE OF POLYMERIZATION ON PVC 1931

Figure 14 Comparison of change of die swell ratio D/D0 with shear rate among samples (a) rolled at 160°C, extruded at
150°C; (b) rolled at 160°C, extruded at 170°C; and (c) rolled at 160°C, extruded at 190°C.
1932 FUJIYAMA AND KONDOU

Figure 15 Comparison of change of die swell ratio D/D0 with shear rate among samples (a) rolled at 180°C, extruded at
150°C; (b) rolled at 180°C, extruded at 170°C; and (c) rolled at 180°C, extruded at 190°C.
EFFECT OF DEGREE OF POLYMERIZATION ON PVC 1933

Figure 16 Comparison of change of die swell ratio D/D0 with shear rate among samples rolled at 200°C, extruded at 190°C.

resin with a lower DP is extruded at a higher temper- ficult as DP is higher and/or that the feature of the
ature.2,3,9,12,17–19 These studies lack in providing a per- particle flow as against the uniform molecular flow is
spective on gelation; however, viewed from the stronger as DP is higher, at the same level of gelation,
knowledge of the present study, both conditions are in are thus elucidated.
the direction of advancing the gelation and of shifting
from particle flow to uniform molecular flow, which
Melt fracture
can explain the effect of the degree of gelation on the
die swell from the concept of the particle and uniform Figure 18 shows as an example of the changes, with
flows. The phenomena and hypothesis that the in- shear rate, of photographs of extrudates obtained by
crease in gelation during extrusion is made more dif- extruding sheets of various DPs milled at 180°C at an

Figure 17 Dependency of die swell ratio D/D0 at ␥˙ ⫽ 122 s⫺1 on degree of polymerization.
1934 FUJIYAMA AND KONDOU

From this it may be said that a resin with a higher DP

generates melt fracture only with a greater degree of
difficulty.
Figure 19 shows the change of the critical shear rate
␥˙ c with the milling temperature using DP and the
extrusion temperature as parameters. The arrows in-
dicate the critical shear rates over or under the mea-
suring limits of shear rate of the apparatus used. Fig-
ure 20 shows the dependency of the critical shear rate
on DP using the milling– extrusion temperatures as
parameters. The critical shear rate is lower as the
extrusion temperature is lower. It also decreases with
increasing milling temperature and thus the gelation.
As for the effect of DP, compared at the same milling–
extrusion temperatures, although a sample with a
higher DP shows the higher critical shear rate and
hardly generates the melt fracture at milling temper-
atures below 180°C, it shows the lower critical shear
rate and easily generates the melt fracture in the cases
of milling– extrusion temperatures of 200 –190°C. (In
the cases of extrusion temperatures below 170°C, the
tendency is not obvious because the critical shear rates
are below the measuring limit.)
The fact that the critical shear rate at onset of the
melt fracture ␥˙ c increases with increasing extrusion
Figure 18 Shape of extrudate rolled at 180°C, extruded at temperature is general for thermoplastic resins and is
170°C. reported also for PVCs.12,13,30 The fact that with in-
creasing gelation the melt fracture occurs easily is
fragmentarily reported by Berens and Folt,1 Shina-
extrusion temperature of 170°C. At low shear rates the gawa,26 and Fujiyama and Nagou.27 According to
extrudates are smooth; they begin to distort with in- Paradis,31 the elastic energy that the PVC melt has
creasing shear rate to around 60.8, 122, and 243 s⫺1 for stored when entering into the capillary from the res-
800Z, 1000Z, and 1300Z, respectively; and the distor- ervoir is consumed by the elastic recovery of die swell.
tion increases with increasing shear rate. Accordingly, However, when the extrusion rate increases and the
the critical shear rates ␥˙ c are 60.8, 122, and 243 s⫺1 for stored energy becomes excessive, it cannot maintain
800Z, 1000Z, and 1300Z, respectively, in this case. itself and a part of it is lost in the form of melt fracture.

Figure 19 Comparison of dependency of critical shear rate ␥˙ c at onset of melt fracture on roll temperature among samples.
EFFECT OF DEGREE OF POLYMERIZATION ON PVC 1935

Figure 20 Dependency of critical shear rate ␥˙ c at onset of melt fracture on degree of polymerization.

If the gelation of PVC is insufficient, because it flows ture more easily occurs with increasing DP, also like
in a slippery manner by the particle flow, the stored the usual resins. In the same way as in the case of die
elastic energy at the capillary entrance is small and swell in the previous section, the phenomena and
thus the melt fracture occurs only with difficulty. With hypothesis that the increase in gelation during extru-
increasing gelation, the PVC melt becomes more elas- sion is made more difficult as DP is higher and/or that
tic by the entanglement of molecular chains and stores the feature of the particle flow as against the uniform
more elastic energy and thus the melt fracture occurs molecular flow is stronger as DP is higher, at the same
easily. level of gelation, are thus elucidated.
As for the dependency of the critical shear rate at
onset of melt fracture on DP, a resin with a higher DP
Entrance pressure loss
shows a lower critical shear rate and more easily gen-
erates melt fracture for the usual thermoplastic resins. Figure 21 shows the dependency of the entrance pres-
The same tendency was also reported for PVC.12,13 sure loss ⌬Pent of each sample on the milling temper-
However these studies lack in providing a perspective ature. The entrance pressure loss increases with mill-
on gelation. Examining the dependency of the critical ing temperature. The entrance pressure loss of 1000Z
shear rate on DP in Figures 19 and 20 from the per- sheet milled at a low temperature of 150°C is lower
spective of gelation, samples milled at roll tempera- than that of unmilled powder. This is assumed to be
tures below 180°C have low degrees of gelation and because linking structures between the primary parti-
are thus assumed to robustly show the feature of cles in the polymerized powder are collapsed mechan-
particle flow. Accordingly, a sample with a higher DP, ically by milling at a low temperature. Although the
in which the increase in gelation during extrusion degree of gelation G reaches a maximum around
scarcely occurs or the feature of the particle flow is 200°C, as shown in Figure 5(a), the low-temperature
assumed to be even stronger at the same gelation endotherm ⌬Hm(A) continues to increase with the
level, shows a higher critical shear rate and can milling temperature, like the entrance pressure loss as
scarcely generate melt fracture. On the contrary, in the shown in Figure 6(a). Terselius et al.,32 Krzewski and
case of high milling (200°C) and extrusion tempera- Collins,33–35 Patel and Gilbert,36 Summer and Rabino-
tures, because gelation proceeds sufficiently and the vitch,20 and Obande and Gilbert37 showed that with
material approaches a unified homogeneous structure, increasing the processing temperature and hence the
like the usual thermoplastic resins, it is assumed that gelation of rigid formulation suspension PVC, the en-
the critical shear rate is decreased and the melt frac- trance pressure loss in the capillary flow increases in
1936 FUJIYAMA AND KONDOU

Figure 21 Comparison of dependency of entrance pressure loss ⌬Pent on roll temperature among samples.

an S-shaped fashion. In the present experimental re- The effect of transition from the particle flow to the
sults, although the degree of gelation G shows an uniform flow caused by the increased gelation appears
S-shaped change [as shown in Fig. 5(a)], the sign of more notably in the elasticity (melt fracture, die swell,
leveling off at the higher side was not observed in the entrance pressure loss) than in the viscosity.
entrance pressure loss ⌬Pent as shown in Figure 21.
The leveling off may be observed at higher milling
CONCLUSIONS
temperatures. The dependency of ⌬Pent on DP is
slight, although 1300Z with the highest DP seems to Sheets with adjusted degree of gelation were prepared
show slightly higher values. by rolling rigid pipe formulation suspension PVC
Figure 22(a) and (b) show the relationships between compounds with DPs of 800, 1050, and 1300 by chang-
the entrance pressure loss ⌬Pent and the degree of ing the roll milling temperature. Their degrees of ge-
gelation G and the low-temperature endotherm lation were measured with DSC and their capillary
⌬Hm(A), respectively. ⌬Pent shows good, nearly linear flow properties were measured with a capillary rheo-
correlations with G and Hm(A) except for one point at meter and the effect of DP on the relation between
G and ⌬Hm(A) of 0 for the sample milled at 150°C. The gelation and flow processabilities was studied.
correlations are only slightly dependent on DP. From
these facts, it may be said that the entrance pressure 1. Because of the difference in shearing heat at roll
loss measured at a lower temperature below the pro- milling, a sample with a lower DP showed a
cessing temperature can be used as a measure of the lower history temperature and hence a lower
degree of gelation regardless of DP. Because the en- degree of gelation.
trance pressure loss is also a physical quantity con- 2. The flow activation energy at constant shear rate
cerned with melt elasticity, like the die swell and melt increased with advancing gelation and de-
fracture, it is assumed that its value is low when the pended only slightly on DP in the range of the
gelation is insufficient and the particle flow occurs and present study.
is high when the gelation and molecular entanglement 3. The viscosity increased by elevating the milling
proceed sufficiently and a uniform flow occurs. temperature and this tendency was more sub-
The entrance pressure loss is also a measure of melt stantial at the lower extrusion temperature. The
elasticity33 and increases with increasing the gelation viscosity increased with DP and the depen-
caused by the increased milling temperature as shown dency on DP was even more pronounced at the
above. Accordingly the entrance pressure loss is used higher milling and extrusion temperatures and
as a measure of gelation. The degree of dependency of hence with advancing the gelation. The depen-
the entrance pressure loss on the gelation is similar to dency was also more notable at the lower shear
that of the die swell. The dependency of the rheologi- rate. The viscosity was constant, independent of
cal properties on gelation is in the order of melt frac- DP at low milling (160°C) and extrusion (150°C)
ture ⬎ die swell ⱌ entrance pressure loss ⬎ viscosity. temperatures. It scarcely depended on DP, in-
EFFECT OF DEGREE OF POLYMERIZATION ON PVC 1937

Figure 22 Relation between entrance pressure loss ⌬Pent and (a) degree of gelation G and (b) low-temperature endotherm
⌬Hm(A).

dependent of the milling and extrusion temper- or the feature of the particle flow is the weaker
atures at high shear rates. These were assumed for a resin with the lower DP. In the case of high
to be because, by shifting from low milling and milling (200°C) and extrusion (190°C) tempera-
extrusion temperatures to high ones and hence tures and hence of advanced gelation, the effect
with advancing gelation, and by decreasing the of DP becomes slight. In this case, it is assumed
shear rate, the particle flow shifted to a uniform that the particle flow is small and the uniform
molecular flow. flow is prominent.
4. The die swell increased with elevating the mill- 5. The melt fracture is the more prominent when a
ing and extrusion temperatures and hence with sheet milled at the higher roll temperature is
advancing gelation. It increased with decreasing extruded at the lower extrusion temperature.
DP and this tendency was even more pro- While at low milling temperatures (⬍ 180°C) a
nounced at the higher extrusion temperature. sample with the lower DP shows the lower crit-
From this it is assumed that the increase in the ical shear rate at onset of melt fracture and
gelation during extrusion is the more prominent hence easily generates melt fracture, at high
1938 FUJIYAMA AND KONDOU

milling temperatures (200°C) it shows the 13. Sieglaff, C. L. SPE Trans 1964, 4, 129.
higher critical shear rate and hence hardly gen- 14. Knyazeva, V. A. Int Polym Sci Technol 1974, 1, T/59.
15. Collins, E. A.; Metzger, A. P. Polym Eng Sci 1970, 10, 57.
erates melt fracture. This is assumed to be be-
16. Collins, E. A. Pure Appl Chem 1977, 49, 581.
cause the particle flow becomes little and the 17. Bowerman, H. H.; McKelvey, J. M. Polym Eng Sci 1968, 8, 310.
uniform flow becomes prominent as in the case 18. Khanna, R. SPE J 1973, 29, 48.
of usual thermoplastic resins. 19. Münstedt, H. Angew Makromol Chem 1975, 47, 229.
6. The entrance pressure loss measured at a low 20. Summer, J. W.; Rabinovitch, E. B.; Booth, P. C. J Vinyl Technol
temperature of 140°C increases with the milling 1986, 8, 2.
temperature and hence with the degree of gela- 21. Iobst, S. A. Ph.D. Thesis, Lehigh University, Bethlehem, PA,
1970.
tion. Because it is scarcely affected by DP, it can
22. Mitani, K.; Ogata, T.; Awaya, H. J Polym Sci Polym Chem Ed
be used as a measure of gelation even for resins 1973, 11, 2653.
with different DPs. 23. Potente, H.; Schultheis, S. M. Kunststoffe 1987, 77, 401.
24. Teh, J. W.; Rudin, A.; Cooper, A. A.; Batiste, J. L. H. Makromol
The authors thank Tokuyama Corp. for permission to pub- Chem Macromol Symp 1989, 29, 123.
lish this article. 25. Collins, E. A.; Krier, C. A. Trans Soc Rheol 1967, 11, 225.
26. Shinagawa, Y. Mitsubishi Plast Tech Rep 1971, 1, 47.
27. Fujiyama, M.; Nagou, S. Nihon Reoroji Gakkaishi 1974, 2, 470.
References 28. Minoshima, N.; Kobayashi, S.; Shimura, M.; Kinoshita, Y.
Kobunshi Kagaku 1971, 28, 953.
1. Berens, A. R.; Folt, V. L. Trans Soc Rheol 1967, 11, 95. 29. Fujiyama M.; Nagou, S. Kobunshi Ronbunshu 1975, 32, 378.
2. Berens, A. R.; Folt, V. L. Polym Eng Sci 1968, 8, 5.
30. Fujiyama, M.; Oba, S.; Awaya, H. Kobunshi Ronbunshu 1975,
3. Berens, A. R.; Folt, V. L. Polym Eng Sci 1969, 9, 27.
32, 418.
4. Fujiyama, M.; Kondou, M. J Appl Polym Sci 2003, 90, 1808.
31. Paradis, R. A. Chem Eng Prog 1966, 62, 68.
5. Suzuki, T. Porimā no Tomo (Japanese) 1970, 3, 161.
32. Terselius, B.; Jansson, J. F.; Bystedt, J. J Macromol Sci Phys 1981,
6. Gonze, A. Plastica 1971, 24, 49.
7. Obande, O. P.; Gilbert, M. Plast Rubber Compos Process Appl B20, 403.
1988, 10, 231. 33. Krzewski, R. J.; Collins, E. A. J Macromol Sci Phys 1981, B20,
8. Uryu, T.; Katagiri, K.; Tsuchida, M. Kogyo Kagaku Zasshi 1958, 443.
61, 1414. 34. Krzewski, R. J.; Collins, E. A. J Macromol Sci Phys 1981, B20,
9. Fukazawa, Y. Kogyo Kagaku Zasshi 1960, 63, 459. 465.
10. Mayrick, R. A.; Sieglaff, C. L. J Appl Polym Sci 1965, 9, 3917. 35. Krzewski, R. J.; Collins, E. A. J Vinyl Technol 1981, 3, 116.
11. Collins, E. A.; Daniels, C. A. Polym Eng Sci 1974, 14, 357. 36. Patel, S. V.; Gilbert, M. Plast Rubber Compos Process Appl 1985,
12. Fujiyama, M.; Wakino, T. Nihon Reoroji Gakkaishi 1991, 19, 64. 5, 85.