Nippon Gomu Kyokaishi, 88, No. 4, 2015, pp.

130–135

Mill Behaviour of Rubber on Two Roll Mill with

Temperature

Hiroyuki Ontsuka1 and Michiharu Toh2

1
Chemicals Evaluation and Research Institute (Inc.) Nagoya Establishment, 1-4-chome, Orito-cho, Showa-ku,
Nagoya, Aichi 466-0858, Japan
2
R&D Department Kurume Research Park Co., Ltd., 1-1 Hyakunen Kouen, Kurume, Fukuoka 839-0864, Japan

Selected from International Polymer Science and Technology, 42, No. 7, 2015, reference NG 15/04/130; transl. serial no. 17321

Translated by K. Halpin

INTRODUCTION [2] that open rolls, which were then the dominant form
of rubber milling machinery, increased their output two-
Open rolls, invented by Chaffee in 1833 and since
fold when the milling length of the rolls and speed of
improved, remain in use today. However, few reports are
rotation were doubled, and gave an account of open
to be found on the thermal behaviour of rubber on open
rolls using various modes of heat transfer in which steam
rolls, which has been looked at purely in the context of
or coolant was passed either to raise the temperature or
rubber milling and mixing operations. This review outlines
prevent rubber scorch due to self-heating during milling.
what has hitherto been reported on rubber behaviour in
milling on open rolls in relation to temperature. Drilled rolls reviewed are the Cowen-Bragg cooling
roll, which has eight linear channels in the axial direction
(i.e. lengthwise) close to the roll surface; the Bragg
HEAT TRANSFER CHANNELS IN OPEN ROLLS cooling roll, which is provided with a helical channel
close to the roll surface; and the Bragg built-up roll, an
Open rolls are of two kinds. The bored roll has a hollow improved form of these rolls. Bored rolls reviewed are
interior as in Figure 1 and transfers heat via a thick wall. the Brewster cooling roll, whose internal wall is provided
The drilled roll has heat transfer channels extending close with ribs between which eight perforated heat-transfer
to the roll surface as in Figure 1 and presents a relatively tubes are mounted; an improved form of this; and the
uniform surface temperature [1]. A century ago, a year Norris cooling roll, provided with channels formed by
before Banbury’s invention of the Banbury internal mixer subdividing the hollow cylinder with six lengthwise
in 1916, Pearson noted in the book “Rubber Machinery” partitions.

CHANGES IN RUBBER TEMPERATURE ON ROLLS

Heat transfer between rubber and rolls, etc.

In 1930 the journal Gomu carried a Japanese translation
by Ishiguro of a report by Young [3] entitled “Rubber
mixing machinery and mixing processes” [4]. The
report notes that the mean temperature in a two-roll
mill (since termed an open roll mill) is lower than in a
closed mixer (since termed an internal mixer) but that
Figure 1. Drilled roll and bored roll there is a temperature distribution, the temperature

© 2015 Smithers Information Ltd. T/1

varying as shown in Figure 2. The gauge expressed as in contact with air, is attributed to radiative transfer via
the nip (the minimum roll gap) N was 7/32 inch (5.5 the air and convective exchange with the air.
mm) and the calender sheet had a band thickness of The original paper does not record the milling
12/32 inch (9.5 mm). The rubber temperature on the conditions in Figure 2, but appends the principal
front roll was 105°C at the nip point N and 86°C at the dimensions of five different open rolls. The rolls in Figure 2
exit 1 from the nip, thereafter reaching 82°C at the roll are very large, with a 36-40 inch diameter and 84 inch
bottom point 2, 79°C at the level point 3, and 76°C at milling length; a back roll speed of around 15 rpm and a
the top point 4. Where the front roll is exposed to the friction ratio of about 2 may be inferred. With the speed
air, namely between the 90° (2) and 270° (4) positions, set at 16 rpm, for example, the peripheral speed of a
the rubber temperature falls 3°C every 90° owing to 40 inch back roll works out to 51 m/min (0.85 m/s).
heat radiation; in contrast the temperature rises 30°C, Although questions surround rate of incorporation of the
almost 10 times as much, from the roll top point 4 to N batch and compounding agents, amounts loaded, etc,
and falls by about 20°C from N to 1. it may be inferred from the scorch temperature that the
The temperature rise observed approaching the nip rubber is natural rubber.
is attributable to evolution of heat due to mixing with
compounding agents in the rotating bank, heat due to
friction between the rubber band and mill bank as the Critical temperature in the nip
nip is approached, and heat due to rapid stretching of
The most obvious difference between open rolls and
the rubber.
internal mixer for milling rubber is said to be that
The account notes that the critical temperature in the whereas in open rolls the rubber is passed through the
nip (the temperature limit for scorching) is held down nip subject to a critical temperature, this is not an issue in
by cooling on the back roll, and the rubber should thus the case of an internal mixer. The following observations
remain below the temperature at which scorch occurs. In are also made. In open rolls, only the rubber layer in
this case, the back roll that pulls the rubber in contact with the nip comes into contact with compounding agents,
the mill bank through the nip would be constantly cooled whereas contact in an internal mixer is wholesale. The
by heat convection and radiation to the air from the roll mean rubber temperature during mixing on open rolls is
surface and by heat conduction from the roll surface to 90-95°C, whereas in an internal mixer the mean rubber
the internal wall surface, heat travelling from the internal temperature is generally 100-105°C, occasionally
wall surface to the cooling water by convection. 110°C. It must not be forgotten that milling time is a major
The account states that the band rubber is cooled factor in scorch. In an internal mixer, the rubber is only
on the front roll by the coolant inside the roll, and also exposed to temperatures from ambient to at most 105°C
by convection and radiation. The heat in the band in and milling is complete in 6-7 min. In the case of open
contact with the roll is transferred by conduction from rolls, the rubber temperature averages up to 95°C, but in
the external roll surface to the internal wall surface, from milling that takes 35 min or more the rubber passes 500
which heat then travels by convective transfer in the times or more through the nip at up to the 120°C critical
coolant. Cooling from the rotating rubber band surface, temperature. From the standpoint of scorch, therefore,
the risk would seem to lie with open rolls.

Decrease in rubber temperature and back pressure

after passage through the nip
After the rubber passing through the bank emerges from
the nip, the rubber peels from the surface of the cooling
back roll surface, heated through contact with the hot
bank, and is exposed to the air. The pressure immediately
the two rolls part beyond the nip would be expected to
drop below atmospheric pressure, constituting a back
pressure and resulting in cavitation. Any low boiling
point matter incorporated would therefore volatilise into
the atmosphere from the rubber band on the front roll,
with a substantial decrease in the surface temperature
of the band rubber by about 20°C in the interval from
N in Figure 2 to the exit zone 1. The reports reviewed
below, though not relating to open rolls, substantiate the
Figure 2. Temperature distribution of rubber in a two-roll mill view that back pressure arises.

T/2 International Polymer Science and Technology, Vol. 42, No. 12, 2015
 Using oil as a Newtonian fluid, Banks et al. [5] and the roll temperature and gap and other operations are
Myers et al. [6,7] made detailed observations of flow in appropriately adjusted, the compounding chemicals will
the nip of a roll coater and divided flow into four regions: consolidate into small lumps instead of being completely
a bank region exhibiting secondary flow, a shear region distributed through the rubber. No explanation is offered
at the nip, a cavitation region due to pressure drop for a, b, c, d, e and f in the diagram. However, Macbeth
beyond the nip, and a filamentation region where the [13] presented the same drawing for the rubber mixing
two rolls part. Chung [8] measured the pressure change process in 1935, and although the rubber temperature
at the nip (gap 70-780 µm) in a stainless steel two-roll is not given, a-f are explained as follows:
coating mill (diameter 16.4 cm, roll length 20 cm) in the a) A zone in which the rubber mass (the bank) promptly
mixing of polymerised linseed oil, a Newtonian fluid. The begins to move, drawn towards the nip.
lower speed roll was immobilised, pressure sensors were
attached from the interior close to the roll surface at the b) A zone into which the back roll draws the added
nip, and the pressure change in the nip was measured compounding agents along with the rubber in zone a.
as the higher speed roll was rotated. The pressure was
found to shift from positive to negative across the nip,
consistent with theory.
Toh et al. [9] made a video recording of the motion
of the rotor, rubber and voids in a two-dimensional
model mixer and presented the results in Figure 3a.
A void-free sickle-shaped zone filled with rubber is
formed ahead of the rotor, while a void lies behind
the rotor. Rubber in sheet form emerges into the void
zone beyond the clearance between the rotor tip and
inner wall of the vessel. Figures 3b and c are enlarged
views of the clearance: the rubber on the rotor tip side Figure 3a. Behaviour of rubber and void in rotors (e) and (f)
is promptly peeled away on the trailing edge of the
clearance, developing surface undulations, and this
peeling invariably seems to occur ahead of peeling from
the internal wall of the vessel, behaviour also observed
in the actual video pictures.
Gondoh [10] attached a pressure sensor to the vessel
wall of a miniature mixer and measured the internal
wall pressure of the tank during mixing; a back pressure
was reported to develop directly after the rotor tip had
Figure 3bc. (b) Magnified view of clearance zone in rotor (e)
passed the sensor, resembling the change in pressure (c) Magnified view of clearance zone in rotor (f)
across the scraper blade in Figure 3d [11].
Figure 3e is an enlarged view of the nip section in
Figure 2. It is supposed that in open rolls back pressure
develops when the back roll separates from the band
rubber on the front roll directly beyond the nip, and just
like the above rubber peeled on the rotor tip side, the
band rubber surface peeling from the back roll undergoes
undulating shrinkage.

Figure 3d. Internal wall pressure distribution in front and

THE RUBBER MIXING PROCESS AND BANK behind scraper blade
FORMATION
In 1943 Mori [12] represented the state of mixing of
rubber in the mixing operation (rubber milling) as shown
in Figure 4, and observed that, generally speaking,
care should be taken in two respects: (1) the various
compounding materials must be blended (mixed) evenly
across the roll in the shortest possible time; and (2)
the rubber must be prevented from reaching a semi-
vulcanised state, known as scorching. Moreover, unless Figure 3e. Magnified view of nip environment

© 2015 Smithers Information Ltd. T/3

c) A zone in which the rubber reaches its greatest Taking a 10 inch diameter, 20 inch long, double-
speed, a large temperature rise occurs, and heat geared variable speed open roll as an example, the
passes to the roll. front roll was set at 18 rpm and the temperature at
70±1°C, and milling was commenced when the back
d-e) A zone in which the temperature of the rubber
roll reached 70±5°C. The friction ratio was varied from
gradually decreases.
1:1 to 1:1.6, and a 900 g initial charge was milled for
f) A zone in which the rubber temperature starts to rise
in contact with the bank of compounding agents.
Figure 4 seems to reference this drawing.

ROLL TEMPERATURE AND MILL BEHAVIOUR

In 1966, Tokita et al. [14, 15] and White [16] observed
and filmed the milling of BR, SBR, etc, on 3 inch open
rolls at different roll temperatures and discovered that
mill behaviour can be divided into four regions, Region
1 to Region 4, according to temperature, as shown in
Figure 5. They also made a theoretical analysis of the
four regions. The roll speed was set at 16.2 rpm for the
front roll and 22.7 rpm for the back roll, with a nip of
0.055 inch (1.4 mm). In Region 1 the rubber is at room
temperature, elasticity is dominant, and the material does
not penetrate the nip. In Region 2 the temperature is 66- Figure 4. Mixing of rubber
71°C and rubber that has passed through the nip winds
onto the front roll, forming a transparent viscoelastic
band; as milling proceeds, a rotating bank forms above
the nip. In Region 3, where the roll temperature has
risen to 100°C, the rubber becomes turbid and hangs
sack-like from the roll. In Region 4, the roll temperature
has risen further to 121°C, viscosity is dominant and
the rubber clings to the roll as a transparent liquid. The
Mooney viscosity of the rubber (ML-4) in the experiments
was 35-50.
Figure 6 corresponds to Region 2; it appears in
texts published by the Society of Rubber Science and
Technology Japan and others as a model of milling.

BANK TEMPERATURE AND RUBBER PROPERTIES

Figure 5. Open roll temperature and milling behaviour

Bank temperature and Mooney viscosity in milling

In an evaluation of carbon black filler conducted by
the Rubber Industry Technologists Association (Gomu
Kougyou Gijutsuin-Kai) Committee 4 (Carbon Black),
Aoe [18] reported the relation between the conditions
in milling on open rolls and the physical properties of
the rubber. The Committee member organisations used
open rolls of different diameter, namely 3.5 inch, 6
inch, 7.5 inch, 8.9 inch, 10 inch, 12 inch and so on;
dimensions such as roll length were also different. The
Mooney viscosity in the milling of natural rubber (NR,
RSS1) was therefore measured with the temperature of
the bank rubber set constant. Figure 6. Model of mixing on open rolls

T/4 International Polymer Science and Technology, Vol. 42, No. 12, 2015
5, 15, 25 and 40 min at a roll guide spacing of 36 cm A further milling experiment was carried out with
and gap at the nip of 1.4 mm. To measure the Mooney natural rubber using 8 inch and 10 inch rolls of different
viscosity, 100 g was collected after each milling time, roll dimensions and speeds in which the bank temperature
and to maintain the time taken to pass between the and number of passes (milling time) were controlled,
rolls the same as the initial time, the guide spacing was affording the results in Figure 9a and b. Despite variation
narrowed 4 cm at a time. The temperature of the bank of the 10 inch and 8 inch roll sets and various other
was measured by thermistor. conditions, the curves for the two sets agree within the
The results in Figure 7a and b show that, with the rolls range of error in measurement at both 85°C and 65°C. It
rotating at the same speed, the bank rubber temperature is concluded that, regardless of difference in roll diameter
decreased with milling time from 92°C (5 min) to 82°C
(40 min) while the Mooney viscosity decreased greatly
from 63.5 (5 min) to 32 (40 min); at the high friction ratio
of 1:1.6, however, while the bank temperature again
decreased from 98°C (5 min) to 94°C (40 min) and the
Mooney viscosity decreased from 66.0 (5 min) to 40.0
(40 min), the decreases were much smaller than at equal
roll speed. Referring to Figure 8, in a similar experiment
with the rolls cooled to hold the bank rubber temperature
at 85±2°C, the Mooney viscosity ratio decreased from
59.0 (5 min) to 33.5 (40 min) at 1:1 and from 62.0
(5 min) to 32.0 (40 min) at 1:1.6, showing agreement
regardless of friction ratio within a range of variation
attributable primarily to experimental error. Given that
bank rotation and the size of the rotating bank were
essentially kept constant in the experiments, it is evident
that the behaviour of the rotating bank is important.

Figure 8. Bank rubber temperature and Mooney viscosity at

different milling times

Figure 7a. Rubber temperature in mill bank and back roll

temperature

Figure 9a. Change in Mooney viscosity with mill time on 10

inch and 8 inch rolls at fixed bank rubber temperature

Figure 9b. Change in Mooney viscosity with mill time for

Figure 7b. Roll friction ratio and Mooney viscosity two different roll sets at fixed bank rubber temperature

© 2015 Smithers Information Ltd. T/5

and conditions, the same milling is possible provided REFERENCES
allowance is made for the other conditions, the main
aim being to maintain a consistent bank temperature. 1. Fujitani I., In “Gomu Kougyou Binran (4th ed.)”.
Nippon Gomu Kyoukai Ed. Nippon Gomu
Kyoukai, Tokyo, p.l070 (1994).
Bank temperature in roll mixing and vulcanisate 2. Pearson H.C., In “Rubber Machinery”, The
properties India Rubber World, New York, p.59 (1915).
3. Young H.C., India-Rubber Journal, 78, 621
Eight members of the Carbon Black Committee conducted
(1929).
roll mixing experiments with carbon black using 6, 8
and 10 inch open rolls. Based on the ASTM mixing roll 4. Young H.C., Isiguro K., Trans.: Gomu, 2, 113
standard of roll diameter 6±0.1 inch (150-155 cm), roll (1929-1931).
guide spacing (band width) 10-11 inch (23-28 cm), front 5. Banks W.H., Mill C.C., Proc. Roy. Soc.
roll speed 24±0.5 rpm and friction ratio 1:1.4, the rubber (London), A223, 414 (1954).
was mixed at a bank temperature of 80±5°C with the 6. Myers R.R., Miller J.C., Trans. Soc. Rheol., 2,
number of passes through the nip kept the same. Mixing 77 (1958).
was carried out 3 times under the same mill conditions: 7. Myers R.R., Hoffman R.D., Trans. Soc. Rheol.,
the nip of 1.4 mm at commencement of carbon black 5, 317 (1961).
addition was opened as appropriate during mixing, the
8. Chung G.: “ Roll Mill no Sousa Tokusei to
mixing rate, etc, being set by adjustment of the nip to
Bunsan Tokusei ni Kansuru Kenkyu “, Doctoral
1.9 mm by the time mixing was complete (20.5 min).
Thesis. Kyushu University (1986).
The compounding recipe was: constant viscosity NR
9. Toh M., Gondoh T., Murakami Y., Nishimi H.,
100.0 parts, carbon black (HAF) 50.0 parts, sulphur 2.5
Mishima M., In “Handbook of Applied Polymer
parts, stearic acid 3.0 parts, DM accelerator 0.6 part.
Processing Technology”, Marcel Dekker Inc.,
The compound was press-cured for 30 min at 145°C
New York, p.523 (1996).
and the physical properties of the vulcanisate, namely
tensile strength TB [kg/cm2], elongation EB [%], 300% 10. Gondoh T., 112th Gomuneri Kenkyu Bunkakai
modulus M300 [kg/cm2], and hardness HS [JIS], were Siryou (2014).
measured. 11. Toh M., Murakami Y., Kagaku Kogaku
The bank temperatures recorded were 76-89°C (4 Runbunshu, 6, 416 (1980).
min) and 75-84°C (18 min), for which variations in 12. Mori T., Gomuseizouhou, Kagakukougyokai,
vulcanisate properties of TB 237-294 (mean 258), EB p.7l (1943).
505-537 (mean 517), M300 119-139 (mean 133), 13. Macbeth C., “Kautschu kmaschinen”,
and HS 61-68 (mean 65) were reported. In “Handbuch der gesamten
It is clear from the reports that in mixing on open Kautschuktechnologie”. (Union Deutsche
rolls the rubber properties are greatly influenced by Verlagsge­sellschaft), Berlin. p.155 (1935).
the behaviour of the roll bank, specifically the bank 14. Tokita N., White J.L., J. Appl. Polym. Sci., 10.
temperature, and the roll nip as it relates to bank size. 1011 (1966).
15. Tokita N., White J.L., Rubber Chem. Technol.,
40, 1126 (1967).
CONCLUSIONs
16. White J. L., Rubber Chem. Technol., 50, 163
On open rolls the rubber makes contact with compounding (1977).
agents in just one place - at the bank over the nip, and 17. Mizumoto K., In “Gomu Gijyutsu no Kiso (6th
the rubber is milled as it passes through the nip; the imp.)”, Nippon Gomu Kyoukai Ed., Tokyo,
behaviour of the bank formed above the nip is hence p.l46 (1987).
paramount. Moreover, milling behaviour on open rolls
18. Aoe I., Nippon Gomu Kyoukaishi, 41, 314
is directly related to the temperature of the rubber, not
(1968).
the temperature of the rolls; the important factors are
thus the temperature of the bank over the nip and the
decrease in rubber temperature directly beyond the nip.

T/6 International Polymer Science and Technology, Vol. 42, No. 12, 2015