Computer control of internal mixer for more consistent EPDM compounds.

Ethylene propylene elastomers are now ranked third among synthetic rubbers, with a worldwide production capacity installed of more than 600 k tons. This success is attributable to their wide processing latitude and excellent resistance to ozone, weathering and heat. The EPDM rubber is used extensively in production of extruded articles, in particular for the automotive industry. Examples of application are door sealing; sponge and dense. glass run channels; weatherstrips for static windows and coolant hoses. Analysis of rubber parts in recent vehicles shows that more than 60% of the rubber, excluding tires, is EPDM-based. This illustrates the unique position of the EPDM in the automotive industry for body sealing and coolant hoses.

Being major components in the car, EPDM rubber goods are now a focal point of the car makers wanting to be more cost effective and to improve their overall quality Some are targeting a reduction of component cost as much as 5% per year. This value can be achieved only by working out the production efficiencies along the supply chain.

By compounding and mixing, the rubber producer adds several independent parameters related to the different fillers, plasticizers and polymers used. This generally enlarges the final characteristics range of different batches mixed from the same formulation, like the plasticity or the Mooney viscosity. To reduce the production cost of the component, the supplier of the car industry looks for the best consistency from batch to batch, with the narrowest range of plasticity. It allows smooth extrusion characteristics and tight dimensional control of the profile shape, eliminating continuous line-setting adjustments.

This article will show that a modem and computerized mixing room helps the rubber producer to control very accurately the weight of the ingredients per batch, run reproducible mixing cycles and ultimately monitor rotor speeds and thermal exchanges of the internal mixer.

By combining these features with tailored MWD EPDM grades which have already proven their contribution for more consistent processing, viscosity variations inherent to the rubber processing can be leveled out to get the best batch-to-batch reproducibility.

Experimental

Piloting a mixer with computer assistance

To conduct the experiments, we have used an intermeshing mixer Werner & Pfleiderer GK5E, connected to a Siemens computer PKS20. This laboratory mixer has a capacity of 5 liters, i.e., a working volume of 3.3 liters. The intermeshing mixer presents certain advantages in the EPDM process which we can summarize as follows:

* A high energy input is delivered in a short period of time because of the configuration of the rotors and the chamber. It is beneficial to the EPDM, as this polymer develops less interaction with the carbon black than do NR or SBR.

* The EPDM does not have very long molecular weight chains as the NR or the SBR do, thus has more difficulty breaking the carbon black aggregates, so higher energy input is required to get a finer dispersion of the fillers.

* The ratio of batch volume to cooling surface is more favorable for a better temperature control of the mixing operation, and so it is well-suited for an efficient automation. The computer is linked directly to the control of the mixer with a system that follows step by step the different operations of the machine during the cycle. There are five possible conditions to monitor a mixing step:

* step time;

* absorbed energy;

* rotors revolutions per minute (rpm);

* calculated reaction value (viscosity);

* compound temperature.

All these steps can be started with a given condition and/or another condition, which increases the control and monitoring of the mixing.

The development of a technique to improve the batch-to-batch reproducibility takes into account process variations like the EPDM Mooney viscosity. We generated a reference base line so that the computer can record the different parameters representative of average conditions at each step of 45 seconds (table l). Centered settings have been chosen such as: EPDM Mooney viscosity - ML 1 + 8, 125[degrees]C: 81; mixer temperature: 50[degrees]C; gum temperature: 25[degrees]C.

We have developed a conventional mixing cycle with eight different steps, which is representative of the conditions used today in the rubber industry.

This reference mix will be used for all the experiments conducted with the computer assistance. Temperature and energy evolution with the time are the reference points that the computer will use to reproduce the mixing cycle (figure 1).

To conduct this process optimization, we used a tailored MWD EPDM (Vistalon 8600). This unique molecular structure is characterized by an overall narrow molecular weight distribution (MWD) with a tailored amount of long and short molecular chains, highlighted by the high value of the ratio Mz/Mw. It offers the best balance in mixing performance: fast carbon black incorporation and best dispersion rate. As an example, we have selected the EPDM sample at the center of the production range (Mooney viscosity) and compared it with mixes made with polymer at the extremity of the specified production range.

Results and discussion

Optimization of the computer assistance model

Once a model for the time, temperatures and energy has been stored in the computer, several options are opened to pilot the mixer in order to get the best reproducibility. We will study three possibilities: time/temperature, energy and temperature/energy.

Time and temperature control

The first and easiest choice is to duplicate for each batch, step by step, the evolution of the temperature with the time and with the machine going automatically to the next step once the target is achieved. An example of the settings is given in table 3.

The mixer will call (1) for the polymer premastication of 30 seconds at 75 rpm, then (3) calls the other ingredients for a mixing cycle of 180 seconds or a maximum temperature of 150[degrees]C. It is the easiest approach to use computer assistance, but it does not take into account any other parameter such as polymer Mooney variation or bale temperature.

For example, if we use an EPDM with a high Mooney viscosity (85) for a series of batches and one with a low Mooney viscosity (77) in another series, we can observe two different populations of compounds for their Mooney viscosity, with large variation of cure rate (s = 10 over an average of 86 dNm/min) and mixing energy (s = 20 over an average of 365 kJ/kg), as seen in figures 2 and 3.

These variations are too significant to provide a very consistent process without adjusting the settings of the extrusion line, such as screw speed and curing temperature.

Energy control

The specific energy absorbed by a compound is representative of the carbon black dispersion and also of the polymer degradation (its viscosity variation). This mixing energy is the power absorbed by the mixer relative to the volume of mixed ,compound, integrated with the time of mixing:

E= [integral] P(t) m.dt in kJ/kg

In theory, the regulation with the energy should overcome the process inconsistencies like:

* raw material temperature (winter-summer);

* mixer temperature (start-up, cooling water variation);

* raw material weight;

* volume loading;

* polymer viscosity. In this study, we have set variable rotor speed controlled by the energy developed during the mixing operation, with the reference energy curve as a target.

The mixer calls (1) for the EPDM premastication, then the fillers (3), and in (4) the rotor speed is controlled by the energy, with a temperature limit to prevent excessive heating generated by a too-high rotor speed.

The results show that this type of regulation is not effective (figure 4). In the series of eight experiments, 75% of the batches were not able to reproduce the reference energy curve, and therefore the rotor speed was kept high (100 rpm), developing an excess of heat through the shear and finally reaching the temperature limit (175[degrees]C) before the batch was well mixed. This resulted in lower energy than targeted for 65% of the batches, and very short mixing time was achieved in some cases (90 and 110 sec.). Compound viscosity of the batches presented large variations (figure 5) and the extrusion of a flat profile presented a high level of non dispersed particles. Simple energy control of the mixer creates a concern. The mixer cannot recover a lower generation of energy because of change in polymer viscosity or mixer temperature. This results in a poor control of the process.

Rotor speed controlled by the temperature over the energy

With the previous experience, we made a synthesis between the capacity of the computer to control the rotor speeds and the somewhat good control obtained from the simple temperature-time regulation. Therefore, in this series of batches, the rotor speed is controlled by the reference curve of temperature, with the energy curve superimposed as a second controlling variable.

This series of mixes was well under control without excessive change in the rotor speed (between 35 and 55 rpm). We also used EPDM with two different Mooney viscosities (77 and 85), and the computer monitored the mixer to reach the targeted energy of 440 kJ/kg for each batch, allowing more time for mixing when a low viscosity polymer was used.

It resulted in a very consistent series of compound Mooney viscosity (s = I for an average of 58.5) in figure 6.

The different compounds look well mixed with a maximum cure rate (100 dNm/min with a s = 3) in figure 7, higher and more consistent than those obtained by controlling only the time and temperatures of the mixing operation

Mixer temperature

In this study we have also conducted a series of experiments designed to assess the relative importance of other mixing parameters, such as the temperature of the mixer and the temperature of the EPDM bale at the introduction in the internal mixer.

The results (figure 8) show that an increase in temperature of 20[degrees]C (from 40 to 60[degrees]C) of the mixer chamber will change the energy developed by the mixing operation. It cannot be controlled and monitored by the computer. It results in a large spread of Mooney viscosity of the different compounds.

This result can be explained by the change of rheological behavior of the mixed compound against the wall of the chamber. If the metal is too hot, there is formation of a thick layer of polymer with low viscosity, which generates a low shear. Conversely, a cold metal would create a situation of slippage, not favorable for a fast black incorporation. The experience shows that there is an optimum for each polymer. For the tailored MWD EPDM used in this study, the optimum temperature of the mixer would be around 40[degrees]C.

Conclusions

In this article, we studied some of the main parameters which influence the consistency of the EPDM mixing operation.

The experimental designs we have conducted have shown that one of the key variables to control is the temperature of the wall chamber. The shear developed during the carbon black incorporation depends on this control, which influences the final quality of the mixing.

There is an optimum for each polymer, depending on its molecular weight and its molecular weight distribution.

Comprehensive use of the computer to assist the batch-to batch reproducibility is defined by close temperature control during the mixing, combined with an energy reference. The experiments have demonstrated the possibility of leveling out the variations of the different compounding parameters, such as the Mooney viscosity of the EPDM.

Using these tools, the rubber transformer can improve his production efficiencies. A reduction of the cost may impact most of the stages of a dense seal production, such as (figure 9):

* mixing with a shorter cycle, single pass and good control of the temperature, which can reduce a dense profile cost up to 3%.

* a higher cure rate, which could allow an increase of the extrusion speed (5% faster could impact by 2% on the component cost).

* extrusion with a better consistency in the rheology of the mixes. It can reduce the scrap generated by too-frequent changes in line setting (2% less scrap may save 2% cost).

* production rejects as more efficient work in the mixer helps to reduce the number of undispersed particles visible on the surface of the profile.

All these efficiencies will contribute to achieving the general objective of component cost reduction and help to meet some very tough targets required today by the marketplace, including the automotive industry.

References

[1.] "Elastomere in VW-Fahrzeugen" Kautschuk+Gummi Kuntstoffe 1, 1991, p. 10. [2.] "Conduite Automatisee d'un Melangeur Interne" P. Hygounenc, PRI conference, Belgium, Feb. 1993. [3.] "La Rheologie, une Aide aux Transformateurs" J.L. Leblanc, Caoutchouc & Plastiques, N719, Dec. 1992. [4.] "Use of a computer for rubber processing" communication, Werner & Pfleiderer. [5.] "Optimization strategy for the mixing room" W. Hader, Rubber World, February 1993. [6.] "Designing EPDM for production efficiency" G. Stella and N.P. Cheremisinoff, IRC Australia, Oct. 1988. [7.] "New generation FPDM polymers for automotive body sealing" S.A. Huser, ACS, May 1990. [8.] "Design technology advancements for FPDM sponge seals" G. Stella and N.P. Cheremisinoff, IRC, Paris, June 1990.