Operational characteristics of the interlocking rotor internal mixer.

Mixing in the rubber industry is dominated by the batch internal mixers. Irrespective of which mixing system is employed, tangential rotors or interlocking rotors, and despite the long period for which these machines have been used in production, reducing reject rates and lowering costs by process optimization has not yet been fully accomplished. The difference between the tangential rotor and interlocking rotor machines is quite an old topic and has been reported (refs. 1-3). Familiarity with the basic differences that distinguish these machines is the first step in determining which mixing approach is preferred and suited for the manufacture of the end product.

Internal mixing is a complex operation, non-steady-state, multi-variable process, where strong interaction between process variables is usually found (ref. 4). There are many different factors in the processability of a rubber compound that if not properly controlled, may cause improper quality in the end product properties. Major factors are:

* Initial conditions:

- polymer starting temperature;

- polymer size;

- polymer grade and rheology;

- internal mixer metal temperature (body & rotors);

* Process variables:

- rotor speed;

- batch size;

- water temperature;

- batch temperature;

- unit work (MJ/[m.sup.3])

- ram pressure;

* Properties:

- rheological characteristics

- cure characteristics

- physical properties

- end product performance.

The internal mixing operation can therefore be influenced via a substantial number of process variables which can be adjusted to optimize performance. However, the initial condition variables are generally ignored or overlooked by many rubber processors, but when adjusted could yield the best consistency in the manufacturing process.

The first interlocking rotor design appeared in a British patent application by R.T. Cooke of Francis Shaw & Co. in 1934. On the basis of the Cooke patent, Francis Shaw commercially introduced an interlocking rotor internal mixer, which they called the "Intermix." Francis Shaw is now part of Farrel Corp.

The rotor shape remained virtually unchanged until the early 1980s, with the introduction of the Mark-4 range machines with. NR2 range of rotors. The basic design has remained the same in Mark-5 machines, which were introduced in the early 1990s, with the addition of a hydraulically driven top ram assembly and NR5 rotors.

This article is directed at the effectiveness of combining the cooling and shear characteristics of the interlocking rotor design mixer and its practical applications.

Importance of cooling

Because of the relationship between viscosity (B) and temperature (T),

[Eta] = A exp (B/T)

where A and B are constants which depend on the nature of material and the importance of maintaining a high viscosity and high level of shear stress, it follows that for effective mixing, temperatures during mixing should be kept low. However, in any practical rubber process the compound is subjected to complex deformation rates and temperatures. Furthermore, because the low thermal conductivity of rubbers and the considerable proportion of the power required for mixing is transformed into wasteful heat, this in general makes the requirement of low temperature difficult to achieve. This seems true despite the considerable emphasis which has been placed upon cooling efficiency in the present generation internal mixers. Figure 1 shows the dependency of compound Mooney viscosity on testing temperature.

[Figure 1 ILLUSTRATION OMITTED]

Cooling efficiency

The mixer is equipped with water cooling which can be electrically heated to all parts of the internal surfaces of the mixer that come in contact with the rubber during the compounding process. To improve the heat transfer efficiency of the machine, the following features were incorporated into the design: drilled chamber bodies; water cooled rotor end wear plates; cooling passages to the rotors; cooling to the plunger; and cooling to the drop door.

For example, typical water flow rates in the above regions for a K4-Mark-5 mixer (91 liters) are summarized in table 1.

Table 1 - water flow rate in different zones of K4 mixer (91L)

Zone   Controlling                 Flow rate, 1/min.

 1     a. Rotors and                  100 (each)
       b. Chamber end wearplates       45 (each)
 2     a. Body and                        67
       b. Plunger                         42
 3     Drop door                          42

In the mid 1990s, a new rotor shape was developed - the NR5 profile. This type of rotor has given good results when tested on a variety of compounds, and practically all new interlocking rotor design mixer units now incorporate NR5 rotors. The NR5 rotor incorporates two very successful design features - enhancing the heat transfer properties and improving the flow pattern of the material being mixed. Basically, the modified design has been arrived at by changes in the cooling passages in the long helical projection (large nog), the small nogs and the main part of the rotor body, as indicated in figure 2. The NR5 design retains the NR2 rotor clearances, nog-chamber wall and nog-root of the adjacent rotor, but the nogs' profiles have been modified. The nogs' profiles have been modified by creating a gradual slope of the leading edges and reducing the area on the top of the large nog of these rotors.

[Figure 2 ILLUSTRATION OMITTED]

Comparison between the NR5 and NR2 rotors' performance in the field with regard to productivity, energy input and Mooney viscosity for two formulations mixed in two K6A (257L) mixers is given in table 2.

Table 2 - comparison between the NR5 and NR2 rotors' performance in the field with regard to productivity, energy input and Mooney viscosity for two formulations mixed in two K6A (257L) mixers

                           Base                 Batch
Mixer type   Compound     polymer   Mooney   weight (kg)

K6A (NR2)    Printing-     HNBR     60-66        204
K6A (NR5)    blanket                63-67        235
K6A (NR2)    Injection-     NBR     18-20        197
K6A (NR5)    molded                 19-21        214

              Cycle     Kg per           KWH
Mixer type   time (s)    hour    KWH    per Kg

K6A (NR2)      260      2,825    14     0.069
K6A (NR5)      286      2,958    19     0.081
K6A (NR2)      386      1,837    14.8   0.075
K6A (NR5)      395      1,950    14.8   0.069

Table 2 indicates that the modified rotor geometry of the NR5 rotor, which produced improvement in the ability to pull in the compound, enables users to increase output by some 10%, while maintaining quality.

Surface-to-volume ratio

Heat transfer in the internal mixer is proportional to the surface area in contact with the batch being mixed, while heat generation is proportional to the volume of material being mixed. The rate of temperature build-up in larger machines will be higher than in smaller machines because the ratio between cooling area and mixer volume is higher in the latter. Figure 3 shows the calculation of the surface area to the net volume for the different sizes of the interlocking rotor design mixer. From the point of view of dispersive mixing, the shear stress must always be above the critical value for the particle concerned, and temperature effect in the larger machine may mean that the shear stress is below the value calculated from machine geometry and would give incorrect predictions when used for scale-up operation. This results in different energy inputs into the material during the mixing cycle. To this end, the initial temperature of the large machine should be lower (more cooling); also, preheating the mixer up to its thermal equilibrium level before starting a production run will lead to uniform energy inputs and less batch-to-batch variation in the final product.

[Figure 3 ILLUSTRATION OMITTED]

Batch size

One of the most significant factors in ensuring the maximum effective mixing in a batch mixer is the batch size. For every compound there is an ultimate batch size which will produce the best results with respect to dispersion and other compound parameters that are significant. In general, the amount of compound which can be mixed in this internal mixer depends on two influential factors:

* Total free volume (liquid volume) of mixing chamber;

* mix viscosity (compound dependent).

Recent developments in rotor design have increased the fill factor of the interlocking rotor design mixer by about 10%. The nominal fill factor now used in the range of machines fitted with the latest design rotors, NR5, is about 0.7. This fill factor could be slightly higher or lower based on the process method and the viscosity of the rubber compound being mixed.

However, generally the mixer is less dependent on having the correct fill factor and has a broader operating window for the batch size, which makes it suited to situations where variable amounts of materials need to be processed. To check this feature of operation in the interlocking rotor design mixer, the effect of the fill factor on the compound properties, such as the tensile strength and power consumption, were tested using a laboratory size mixer KO (1.80L) fitted with NR2 rotors. An SBR tire tread compound was mixed at different fill factors. A conventional mixing cycle was used in processing the compound by adding all ingredients at the start, ram lifted for 10 seconds when completing 90 seconds of mixing and dropped at 150 seconds. The rotor speed used was 70 rpm throughout and circulating water temperature operating at 60 [degrees] C. The end point of the mixing process was measured as 10 seconds after the amps reached a peak after incorporation of the black.

To determine the useful batch volume range, the graph of tensile strength versus fill factor (figure 4), and also the graph of amps versus mixing time at different fill factors (figure 5), were considered. Figure 4 shows that as the batch volume increases, the tensile strength increases, then starts to fall off as the batch volume becomes too large at fill factors of 0.7 to 0.75. This graph has a plateau region of tensile strength, hence giving a useful range of fill factors between 0.46 and 0.65. Figure 5 appeared to confirm this conclusion. For instance, at the lower fill factors of 0.28 and 0.37, there is insufficient material in the chamber for efficient mixing to take place. The current load, therefore is low, and the final peak in amps is not reached, indicating that the end point of the mixing process had not been achieved. This inefficient mixing is mainly due to the lack of interaction between the compound and the machine internal surfaces. At the higher fill factors between 0.46 and 0.65, the current load is higher and the final peak in amps, as well as the end point of mixing are achieved, giving satisfactory mixing results.

[Figures 4-5 ILLUSTRATION OMITTED]

Ram pressure

Until the late 1950s, the standard internal mixer was operated with a ram pressure around 16 psi. At this pressure it was found that a large proportion of the mixing cycle was used to shred and masticate the rubber sufficiently for the matrix to flow and fill the flow channel. The effect of increasing pressure is to increase the contact force between the rubber and the rotor surface. This has the effect of increasing the critical stress so that flow begins at a lower temperature. This means that effective mixing begins much earlier in the cycle, and because the stock temperature is lower at this point, the viscosity is higher and the power peak is higher. The rotor speed can be increased before slip occurs so that the mixing time for a constant total shear can be decreased. Starov et al. (ref. 5) compared the power requirements to achieve the same dispersion as a function of ram pressure and rotor speed. They found the empirical expressions:

P = A + B Pr

P = C + DN

where P is the power in kW and Pr and N are the ram pressure and rotor speed, respectively.

However, these correlations do not include machine geometry or material variables, and are usually approximations to more complex equations (ref. 6). Therefore, the simple operating lines would give a reasonable approximation or response over a limited range of operating conditions for a constant machine geometry and material.

Whitaker (ref. 7) examined the combined effects of increased ram pressure and increased rotor speed on the mixing time and power when working with a K2A mixer (49L) and a standard SBR recipe. He found that mixing time could be greatly reduced by increasing the ram pressure to 60 psi (4 bar) with little further change at higher pressures.

The interaction between the pressure exerted by the ram and the compound has been found to be of paramount importance to achieve satisfactory mixing in the interlocking rotor design mixer. Today, the Mark-5 is equipped with an energy efficient hydraulic actuated ram with controllable pressure up to a maximum of 5 bar.

Mixing action

The mixing action in the interlocking rotor design mixer is designed to increase the effectiveness of distributive and dispersive mixing of viscoelastic materials.

Distributive mixing

The rotors in the interlocking rotor design mixer move in opposite directions, and the nogs are designed so that the portions of the material move from one rotor to the other and between the nogs of each rotor. This random transfer between rotors enhances distributive mixing. Because of the complex geometries of the rotors and the mixing chamber, the mechanics of the rubber flow in the mixing chamber of the mixer are complex and difficult to analyze. Instead, the distributive mixing characteristics were investigated by direct flow visualization studies using the interlocking rotors as used in the mixer fitted with a transparent mixing chamber. Figure 6 shows the distribution of a red marker within the matrix of a clear silicone rubber at a different number of revolutions. The quality of the distributive mixing of the rotors, which is controlled mainly by the number of revolutions, can be judged easily by the absence of the red and white specks in the final compound after nine revolutions.

[Figure 6 ILLUSTRATION OMITTED]

Dispersive mixing

During the mixing operation in this mixer, the deformation of rubber occurs largely by shear forces and also by elongation flow of materials entering high shear regions between the rotor nogs and the chamber wall and between rotors.

In practice over most of the shear rate range of interest the shear stress increases with the shear rate in a manner according to the power law equation:

[Tau] = k [[Gamma].sup.n]

where [Tau] is shear stress, y is shear rate, k is constant, which is material dependent, and n is power law index (for many rubbers, n = 0.2 to 0.3).

The moving rotors in this mixer cause drag flow between the nogs and the chamber wall (region A, figure 7), as well as between the root of the rotor and the wall (region C). Usually, the nog-wall clearance is about 1/13 of the root-wall clearance so that the former is the region of a high shear stress. In addition to the friction ratio of about 1.4:1 caused by the difference between peripheral speeds of the nog diameter of one rotor and that of the root of the opposing rotor, the clearance is small enough to give shear in this region (region B). To cover for the complex geometry of the interlocking rotor design mixer, the above three regions in the mixer were used for the shear rate measurement, using the relative movement of two walls with a certain clearance h. The average shear rate obtained from pure drag flow can be calculated from

[Gamma] = [Omega].r/h

where [Omega] is angular velocity and r is radius of the moving surface.

[Figure 7 ILLUSTRATION OMITTED]

Figure 8 shows the shear rate versus rotor speed, calculated from the machine geometry of a K7 mixer (306L) for the three regions A, B and C. It is clear that region A is characterized by a high shear stress.

[Figure 8 ILLUSTRATION OMITTED]

To demonstrate the importance of region A, the author has conducted a series of experimental programs in conjunction with the Institute of Polymer Technology and Materials Engineering of Loughborough University of Technology in the East Midlands of England, with which the author was associated, to measure the high shear stress and the level of viscosity reduction during the passage of material through this region (ref. 8).

The region selected for this study (ref. 9) was between the rotor and the chamber wall of an internal mixer (figure 9a), and was reasonably represented by a fixed-blade and roll geometry, as shown in figure 9b. The rotor is assumed as fixed, whereas the chamber wall can move with different rotations. Four pressure transducers were fixed at different locations in the mixing volume and the blade-wall clearance. The tips of the pressure transducers come into contact with the flowing compound, thus allowing them to detect the distribution of pressure and evaluate the magnitude of the flow of the compound (ref. 10). A simple compound containing only natural rubber (SMR10) and 30 phr of N330 carbon black was used in this work. The compound was mixed in a K1 (5.5L) mixer and re-milled, allowing the viscosity of the final product to be reduced by about 40%, hence giving an ideal compound for the pressure distribution study.

[Figure 9 ILLUSTRATION OMITTED]

To determine whether there was a change in the flow property of the compound after passing only once through the clearance; flow curves of the compound were measured at different shear rates in a variable speed biconical rotor rheometer named the TMS rheometer, and the results are shown in figure 10. This figure compares the flow behavior of the compound before and after one passage through the clearance, in consequence, the compound became softer as indicated by the lower shear stress values of the flow curve.

[Figure 10 ILLUSTRATION OMITTED]

Figure 11 shows the pressure values as measured by pressure transducers at positions 1, 2, 3 and 4, and obtained at 6 rpm with 2 mm clearance between the blade-wall region. It also shows that the maximum peak pressure appears to fluctuate considerably, especially those measured by transducers 3 and 4. In the case of transducers 1 and 2, it is believed that this observed fluctuation occurs as a result of the compound experiencing a degree of slippage. In the case of transducers 3 and 4, the maximum peak pressure not only fluctuates, but it shows zero readings. From these zero readings, one can conclude that there is no compound which exerts pressure upon transducers 3 and 4. Evidence of this observation was confirmed by melt fracture of the compound in the clearance region. Furthermore, transducers 3 and 4 show that a substantial pressure (up to 6 MPa) can be generated in this clearance because of the flow constriction in this region.

[Figure 11 ILLUSTRATION OMITTED]

To summarize the mixing action, the interlocking rotor design mixer mixes the compound in two areas, but at different times during the cycle. In the early part of the cycle, the elastic behavior of the material predominates, and the elastic forces exceed the frictional forces at the rubber-metal interface, preventing the material from entering the region (A) between the nog and the mixing chamber wall. This mechanism is beneficial in the early stages of mixing, preventing material from entering the small clearance where it could generate excessively high torques and power peaks, and perhaps stall the mixer. Therefore, most of the shear is occurring between the rotors and is akin to the action of a two-roll mill. As the material warms up and its viscosity drops, the viscous behavior of the material predominates, enabling it to flow through the region between the rotor nogs and the mixing chamber wall under the influence of the drag flow and pressure gradients generated by movement of the rotor.

The material is accelerated as it passes in this region, leading to elongational flow and effecting very high dispersive mixing on the material.

Practical examples

A K2A mark-5 interlocking rotor design mixer of chamber net volume capacity 49 liters, which gives reasonably comparable performance to that in considerably larger scale production mixer, was used for mixing two formulations in a single pass technique; an antivibration compound based upon NR/BR (table 3) and a typical EPDM truck hose compound (table 6). Tables 4, 5, 7 and 8 show the K2A parameters used in mixing and some of the material property responses.

Table 3 - NR antivibration

NR          39%
BR          10%
Black       36%
Silica       3%
Oil          1%
Chemicals   11%

Table 4 - mixer parameters and responses (NR-antivibration)

Water temperature, C                                 30
Zone 1 (rotors + wear plates)
Water temperature, ([degrees] C)                     35
Zone 2 (body + ram + door)
Rotor.speed, rpm                             60, 40, 30
Ram pressure, bar                                     5
Batch weight, kg                                   35.5
Discharge batch temperature, ([degrees] C)          125
Actual batch temperature, ([degrees] C)             128
Cycle time, sec.                                    228
Total energy input, kwh                            4.84
Specific energy, kWh/kg                           0.136

Table 5 - compound property response (NR-antivibration)

Mooney, ML (1 +4)121 ([degrees] C)     90
Hardness, IRHD                         80
Specific gravity, g/[cm.sup.3]       1.19
Tensile strength, Mpa                  19
Elongation @ break, %                 350
Table 6 - EPDM truck hose

EPDM        27%
Black       40%
Fiber        5%
Oil         24%
Chemicals    4%

Table 7 - mixer parameters and responses (EPDM truck hose)

Water temperature, ([degrees] C)                     30
Zone 1 (rotors + wear plates)
Water temperature, ([degrees] C)                     35
Zone 2 (body + ram + door)
Rotor speed, rpm                             40, 25, 20
Ram pressure, bar                                     5
Batch weight, kg                                  33.50
Discharge batch temperature, ([degrees] C)          100
Actual batch temperature, ([degrees] C)             104
Cycle time, sec.                                    240
Total energy input, kwh                            2.35
Specific energy, kWh/kg                            0.07

Table 8 - compound property response (EPDM truck hose)

Mooney, ML (1 +4) 121 ([degrees] C)     65
Hardness, IRHD                          88
Specific gravity, g/[cm.sup.3]        1.16
Tensile strength, Mpa                    7
Elongation @ break, %                  240

Figures 12 and 13 show the power, batch temperature and ram position profiles obtained during mixing of the two compounds together with the mixing steps. For instance, in figure 12 the rubbers and peptizer were added to the mixer which was operated at 60 rpm and masticated to a preselected temperature of 100 [degrees] C. Carbon black, small chemicals and oil were added and mixed at 40 rpm. At 105 [degrees] C the cure system was added and mixing continued at 30 rpm until discharge at a pre-selected batch temperature of 125 [degrees] C. Further demonstration on the effect of changing rotor speed on the mixing parameters and how it can be used to control the batch temperature is shown in figure 13. To avoid any additional mixing, both compounds were passed twice through a two roll mill (60 x 22 inch) with samples being taken for quality evaluation (tables 5 and 8) at the end of the second pass.

[Figures 12-13 ILLUSTRATION OMITTED]

Silica mixing

The continued increase in the utilization of silica in passenger tire treads due to improved rolling resistance (ref. 11) and the benefits produced in lower fuel consumption is forcing the tire industry to take advantage of the interlocking mixers in processing silica-filled compounds.

When processing rubber compound containing high levels of silica and the required silane, special attention has to be paid to the thermal reactions of silane with silica and with the rubber. The silane-silica reaction, which must be achieved in the mixer, is highly sensitive to the mixing parameters, time and temperature. The silane-silica reaction time can be shortened by increasing the temperature, but this rise must be limited to 160 [degrees] C to avoid the second reaction of silane with polymer leading to premature vulcanization. The silane-silica reaction is best achieved between 140-160 [degrees] C, whereas with temperatures above 160 [degrees] C, the silane-polymer reaction becomes active (ref. 11).

With silica technology, the rubber compound passes three to four times in the mixer. This multi-pass process technology is needed for two reasons. Firstly to achieve easy processing ranges of Mooney viscosity in the compound, a good degree of silica silanization must be achieved. Secondly, in a large mixer with tangential rotors, the batch cannot be maintained at a high temperature. Instead, in each pass, the batch is mixed at an elevated temperature for a short time up to three minutes and dropped prematurely to avoid the onset of thermal reaction of silane with polymer.

By way of analogy, because of its good temperature control, a steady-state condition can be achieved in this internal mixer; that is when the mixer power requirement and the batch temperature have both reached constant values for certain batch size and rotor speed, as shown in figure 14. Operating the interlocking rotor design mixer under equilibrium conditions enables the silica filled compound to be processed in a smaller number of mixing stages. For instance, at the beginning of the cycle the batch can be mixed efficiently at high speeds to reach a good degree of dispersion, then the speed is reduced and held at a level where the heat dissipation and the heat flow to the metal surfaces of the mixer are in an equilibrium. However, it is interesting to note that the increasing trend towards operating the interlocking rotor mixers in a steady-state condition has also been extended to tangential rotor mixers fitted with high efficiency ST rotors (tip cooled synchronous technology). Results to-date have shown that reacting of silica with different silane coupling agents in a rubber compound could be performed in a F80-ST (70 L chamber volume) operating in a steady-state at 160 [degrees] C for eight minutes.

[Figure 14 ILLUSTRATION OMITTED]

Conclusions

The combination of the increased heat transfer through the intensive cooling of the full surface of the rotor and the specifically modified rotor geometry all contribute to the uniqueness of the NR5 rotor which produced improvement in the circulation of the material being mixed, ability to pull in the compound and enabled users to increase fill factor to typically 0.7.

The interlocking rotor design mixer is undoubtedly the most efficient for mixing technical rubber compounds. However, with silica technology resulting in the benefits of tread compounds with better rolling resistance, wet grip and abrasion resistance, we are now entering a new phase for processing silica-filled compounds. Once the appropriate rotor speed and batch size are selected, a general procedure for operating the mixer in a steady-state can be established. It is therefore foreseeable that the interlocking rotor design mixer will gradually be favored by the tire industry for the silanization process to which operating the mixer in a steady state condition could successfully be applied.

Acknowledgments

"Operational characteristics of the interlocking rotor internal mixer," is based on a paper given at the September, 1999 meeting of the Rubber Division.

"QC of the discontinuous compounding process in a rubber internal mixer by regression and neural networks process models," is based on a paper given at the April, 1999 meeting of the Rubber Division.

"NR elastomer composite: Process, material and performance characteristics," is based on a paper given at the September, 1999 meeting of the Rubber Division.

References

(1.) W.M. Wiedmann and H.M. Schmid, Rubber Chem. Technol., vol. 55, p. 363 (1982).

(2.) P.K. Freakley, Rubber Chem. Technol, vol. 65, p. 706 (1992).

(3.) P. Wood, Tire Technol. International, p. 51 (June 1998).

(4.) J. Batchelor, P.K. Freakley, S.N. Ghafouri and D. W. Southwart, Rubber World, vol. 200, No. 4, p. 18 (1989).

(5.) I.M. Starov, A.A. Suschehenko and S.V. Gelfreikh, Sov. Rubber Technol., vol. 20, p. 17 (1961).

(6.) J.M. Funt, "Mixing of Rubber," Rapra, Shrewsbury, U.K. (1977).

(7.) P. Whitaker, J IRI, vol. 4, p. 153 (1970).

(8.) S.N. Ghafouri, "Indirect flow visualization" Report No. 9, Loughborough University, U.K. (1993).

(9.) S.N. Ghafouri and P.K. Freakley, Polymer Testing, vol. 13 (1994).

(10.) P.K. Freakley and S.A. Patel, Polym. Eng. Sci., vol. 27, No. 18, p. 1,358 (1987).

(11.) S. Wolff, Tire Sci. Technol., vol. 15, p. 276 (1987).