The single-rotor continuous mixing system.

Continuous mixing holds out the promise of efficient and consistent rubber processing. This has been recognized for many years, with the last major development activity occurring in the 1970s (ref. 1). Then, the lack of a reliable supply of technologically and economically viable particulate rubber, coupled with a relatively undeveloped mixing technology, caused the movement to founder, despite the best efforts of enthusiasts. Now, continuous mixing is back on the agenda with major materials suppliers offering a range of elastomers in particulate form. There is also a much better understanding of mixing to draw upon for process design (refs. 2-4). It is also recognized that supplying a consistent feedstock to the continuous mixer is a pre-requisite for success.

In this article, a new powder blender and a continuous mixer are described, followed by a presentation of results from a prototype system. A layout of the system, equipped with a roller die output device, is shown in figure 1. Alternatively, a screw extruder can be substituted for the roller die. The objective is a simple, lightweight, versatile and efficient system.

[FIGURE 1 OMITTED]

Design of the carousel blender/feeder

Rubber compounds in particulate blend form present an acute problem for supplying a consistent feedstock to the continuous mixer. The obvious solution, separate metered feeds of each ingredient, is very expensive and unwieldy because of the number and variety of ingredients. This points to conventional batch weighing, followed by some form of pre-blending and metering to the mixer. For this route to be successful, two difficulties have to be overcome:

* The blending action must be gentle, to avoid damage to pelletized materials, particularly fillers such as carbon black and silica;

* segregation during blending and subsequent feeding to the continuous mixer must be avoided.

The latter difficulty is particularly acute. The ingredients of a typical rubber compound have a wide range of particle shapes, sizes and densities. From figure 2, which shows a capability map for currently available powder blenders, it can be seen that the requirements of rubber compound blending fall in a region which is not serviced. In addition, even if a consistent powder blend can be produced, segregation in transport to the continuous mixer must be avoided to gain any benefit from this achievement.

[FIGURE 2 OMITTED]

The carousel blender/feeder (ref. 5) shown in figure 3 has been designed to overcome the difficulties outlined above. Its action is based on ordered subdivision to render it insensitive to segregation during blending, and as shown in figure 1, it can be mounted directly onto the continuous mixer to avoid segregation during transport.

[FIGURE 3 OMITTED]

During blending, the contents of each cell are subdivided into two equal parts as they fall into the cell below. Sequenced rotation of the carousels and opening of the cells gives a consistent composition in each of the final cells, irrespective of the sequence of weighing or initial distribution of ingredients in the top carousel. The final carousel then delivers a pulsed feed as each cell is discharged. So, from a batch weighed input, a continuous feed is delivered to the mixer. Referring back to figure 2, it can be seen that the carousel blender occupies the essential territory of "gentle action combined with insensitivity to segregation." The dependence on gravity for the blending action enables a light construction to be used and gives a low energy requirement. The prototype system has a diameter of 0.4 m, a height of 1.3 m and has a maximum design output of 250 kg/hr.

Design of the single rotor continuous mixer (SRM)

From analysis of existing continuous mixers (refs. 3 and 4), it is clear that some separation of incorporation, distribution and filler dispersion functions is desirable for an efficient design. Similarly, separation of conveying and mixing is necessary for versatility, to bring residence or mixing time under operational control. A schematic cross-section of the prototype, which has a simple fixed die, is shown in figure 4.

[FIGURE 4 OMITTED]

The prototype system shown in figure 4 has a 90 mm diameter feed unit screw and a 200 mm diameter mixing unit rotor. Maximum speeds are 30 rev./min. for the former and 100 rev./min. for the latter. Simple water circulation temperature control is used for the screw, rotor and barrels.

Mixing trials and results

Two dissimilar compounds were selected for initial evaluation of the SRM. The purpose of the SBR compound shown in table 1 is to explore the dispersive mixing behavior of the SRM, while that of the EPDM compound in table 2 is to determine its ability to deal with high filler and oil loadings.

Union Carbide's Elastoflo EPDM is produced directly from polymerization with a spherical granule in the size range 0.5 to 1 mm, while the SBR is granulated bale material with an irregular particle in the size range 4 mm to 7 mm. No difficulties were experienced with the feeding of the SRM system. In the results which follow, rubber compound viscosity is used as a measure of filler dispersion (refs. 6 and 7).

A great deal has been learned from the mixing trials to date. The screw and rotor speeds of eight and 40 rev./min. for the SBR compound and 20 and 40 rev./min, for the EPDM compound are simply the first found to give a visually mixed rubber compound.

The results in tables 3 and 5 confirm that the mixing treatment can be varied over a wide range by independent adjustment of feed unit and mixing unit speeds. In addition, the outlet impedance (die resistance) of the mixing unit has been found to be a powerful and useful variable. A reduction of impedance reduces the effective fill factor in the mixing unit. The second result in table 3, marked by the *, was obtained with a reduced impedance. Unit work and rubber temperature are reduced, but there is an unacceptable deterioration in dispersion, indicated by the high viscosity. In contrast, table 5 shows that reducing the outlet impedance for the EPDM compound gives an overall improvement, reducing unit work and rubber temperature while maintaining a good dispersion. Outlet impedance clearly has a strong influence on the intensity of mixing; and a much lower intensity is needed for dispersive mixing of N550 carbon black than for N330.

The unsatisfactory filler dispersion of both the SBR compound and the EPDM compound at maximum speed is attributed to the high rubber temperature causing a substantial drop in the stress available for dispersion. There is substantial scope for improving cooling. The prototype SRM was designed for simplicity and ease of modification, which precluded the incorporation of efficient cooling. In both tables 3 and 5, the last row gives the best result that can be achieved with the current prototype--maximum output with satisfactory filler dispersion. This will be increased when the cooling efficiency is improved.

In tables 4 and 6, the results for satisfactory carbon black dispersion in a Shaw K1 Intermix (intermeshing rotor internal mixer of 5.5 liters chamber volume) are given, together with the batch viscosity, which provides a measure of this state-of-mix. Comparison with the unit work results in tables 3 and 5 shows the high energy efficiency of the continuous mixer. For the SBR compound, the energy required is approximately a third that of the batch internal mixer and half for the EPDM compound. In addition in the obvious energy saving, this improvement enables the size of the whole mixing system to be reduced. With the simplicity and stiffness of the cylindrical feed and mixing unit barrels, the mass of the continuous mixer can also be much lower than that of an equivalent batch mixer.

Physical properties

Tensile properties have been measured to determine if the state-of-mix results in tables 3 through 6 are carried forward into the vulcanizate. The SRM results in table 7 are derived from feed and mixing unit speeds of 8/40 for the SBR compound and 20/40 for the EPDM.

The similarity of results from the SRM and the Intermix for both compounds confirms that the former gives effective curative distribution and dispersion in addition to good filler dispersion.

Purging characteristics

Minimum wastage of rubber compound and time during a compound changeover are natural objectives of continuous mixer design. Experience with the prototype SRM has shown that the feed unit screw and the dispersive zone are completely self purging, but physical removal of rubber from the feed unit head and the distributive zone is necessary. Design modifications to improve the self-emptying of the distributive zone are being investigated. In addition, a trial to explore the extent to which a direct changeover is possible, without emptying the machine of rubber, has been carried out. The SBR and EPDM compounds, with and without curatives, were used to track the progress of purging. The feed hopper was run empty of the particulate blend with curatives and followed immediately with a blend which omitted the curatives. Samples of mixed compound were then taken at intervals for cure testing. The results, shown in figure 5, were obtained at an output rate of 35 kg/hr. for the SBR and 44 kg/hr. for the EPDM.

[FIGURE 5 OMITTED]

The reduction in the amount of material present with curatives is tracked by the change in crosslink density ([T.sub.max.] - [T.sub.min.]), until it is eliminated, at one minute for the SBR and two minutes for the EPDM. After each of these trials, the mixer was opened and samples of compound removed from the head of the feed unit for further cure tests, to check for residual compound with curatives. These samples showed zero crosslink density and confirmed that the head is tree of regions of re-circulatory flow or stagnation.

Scale-up

Scaling rules have been developed for the SRM, using non-isothermal, non-Newtonian flow analyses to extrapolate from the prototype results. They are conservative in that they do not assume any increase in dispersive mixing efficiency resulting from the planned improvements in heat transfer efficiency and output pressure control. The predictions in tables 8 and 9 are for single-pass mixing with all the curatives included in the blended particulate feedstock. Consequently, the maximum output rate is dictated by robber temperature, and a target outlet value of 100[degrees]C has been used. The specific heat transfer rates have been set at levels expected for drilled cooling channels, except for the prototype characteristics in the top row, which are experimentally determined. For comparison, the experimentally determined heat transfer coefficient of the flood cooled Intermix used in the work reported here is 770 W/[m.sup.2]/C. Net power does not include allowances for drive and friction losses.

Conclusions and further development

Results from a prototype continuous rubber mixing system have been presented which provide proof of principle and show the way forward for improvement and for the design and build of production systems. In particular, the new system:

* is compact, simple and robust;

* has simple and versatile operational control to mix a wide range of compounds;

* feeding is insensitive to rubber particle size and shape within the range investigated;

* is energy efficient;

* can achieve complete mixing in a single pass;

* has good purging behavior for rubber compound changes.

Substantial improvements to the heat transfer efficiency of the mixing unit are possible, to combine high output with low rubber temperature. This will also improve the dispersive mixing capability. The prototype unit is relatively inefficient due to simplicity of construction and design for adaptability.

Table 1 - SBR compound

Material                    Pphr

SBR 1500                     100
Carbon black N330             40
Zinc oxide                     5
Stearic acid                   2
Sulfur                         2
CBS                            1
Table 2 - EPDM compound

Material                    Pphr

EPDM (includes 20 phr CB))   120
Carbon black N550            100
Oil                           75
Zinc oxide                     5
Stearic acid                   2
Sulfur                         2
CBS                            1
Table 3 - single rotor continous mixer results for the SBR compound

Speed               Power        Output       Unit work
(rev./min.)         (kW)         (kg/h.)    (MJ/[m.sup.3])

8/40               1.0/4.8         43            527
8/40 *             1.1/3.0         43            343
30/100             4.9/16         150            620
16/100             2.4/10.8        78            620

Speed              Output       Viscosity
(rev./min.)         temp.        (kPa S)
                ([degrees]C)

8/40                  97            97
8/40 *                88           120
30/100               153           121
16/100               121            98
Table 4 - Intermix results for the SBR compound

    Speed           Power        Mixtime      Unit work
 (rev./min.)        (kW)         (kg/h.)    (MJ/[m.sup.3])

     30          21.5(avg.)         4           1,718

    Speed          Output       Viscosity
 (rev./min.)        temp.        (kPa S)
                ([degrees]C)

     30              135           95
Table 5 - single rotor continous mixer results for the EPDM compound

Speed               Power        Output       Unit work
(rev./min.)         (kW)         (kg/h.)    (MJ/[m.sup.3])

20/40              2.9/8.0          73           537
20/40 *            2.7/3.7          73           315
30/100             4.2/12.6        130           471
25/100             3.5/7.4         105           372

Speed              Output       Viscosity
(rev./min.)         temp.        (kPa S)
                ([degrees]C)

20/40                131           83
20/40 *               96           84
30/100               127           94
25/100               120           85
Table 6 - Intermix results for the EPDM compound

   Speed          Mixing        Output       Unit work
(rev./min.)        time         (kg/h)     (MJ/[m.sup.3])
                  (min.)

     32             19             2            754

   Speed          Output       Viscosity
(rev./min.)        temp.        (kPa S)
               ([degrees]C)

     32             118           84
Table 7. comparison of tensile properties

SBR             M100 (MPs)      EB (%)        TS (MPs)

SRM                 2.8           494          21.4
K1 Intermix         2.6           436          20.0

EPDM
SRM                 4.3           428          13.6
K1 Intermix         5.1           361          14.1
Table 8 - scale-up predictions for the SBR compound

        Screw D         Q      Power
Feed      (mm)    (kg/hr.)   (kW-net)
unit      90          80       1.7
         120         178         4
         150         333         8
         180         560        13
         210         871        21
         230       1,149        28
         260       1,569        41

        Rotor D     Power     Rubber       Heat trans.
Mixing  (mm)      (kW-net)   temp (C)   (W/[m.sup.2]/C)
unit     200          9        148             410
         300         20        102             920
         400         38        102             980
         500         64        102           1,060
         600        100        102           1,150
         700        132        103           1,100
         800        180        103           1,180
Table 9 - scale-up predictions for the EPDM compound

        Screw D      Q         Power
Feed     (mm)     (kg/hr)    (kW-net)
unit      90        106         2.3
         120        236         5.5
         150        443          11
         180        744          19
         210      1,157          29
         230      1,527          39
         260      1,917          50

        Rotor D     Power     Rubber       Heat trans.
Mixing  (mm)      (kW-net)   temp (C)   (W/[m.sup.2]/C)
unit     200          9        100           460
         300         14        100           590
         400         27        101           630
         500         45         99           700
         600         69        101           750
         700         92        100           730
         800        115        100           710

References

(1.) C.W. Evans, Powdered and Particulate Rubber Technology, Applied Science Publishers, London, (1978).

(2.) J.L. White, Rubber Processing, Hanser/Gardner, Cincinnati (1995).

(3.) I. Manas-Zloczower and Z. Tadmor (Eds), Mixing and Compounding of Polymers, Hanser/Gardner Cincinnati (1994).

(4.) H. Ellwood, "A tale of continuous developments, " Eur. Rubb. J., 169 (3), 26-32 (1987).

(5.) J. Clarke and P.K. Freakley, Patent Application No. PCT/GB00/ 02882 (WO 01/07153).

(6.) J. Clarke and P.K. Freakley, "Reduction in viscosity of an SBR compound by mastication and dis-agglomeration during mixing, " Rub. Chem. Technol., 67, 4, 700-715 (1994).

(7.) P.K. Freakley and J. Clarke, "Comparisons of the mixing of rubber with carbon black in an internal mixer and in a biconical rotor rheometer," J. Appl. Polym. Sci., 53, 121-132 (1994).

(8.) J. Clarke and P.K. Freakley, "Modes of dispersive mixing and filler agglomerate size distribution in rubber compounds," Plast. Rub. and Comps., Proc. and Appl., 24, 5, 261-266 (1995).

P.K. Freakley, Loughborough University and J.B. Fletcher, Carter Bros. Ltd.