Development of compounding processes for FKM using twin-screw compounders.

Batch mixing continues to be the mainstay of the rubber industry. Fluoroelastomer compounds are mixed in a variety of methods, including internal mixers and open face milling techniques. In these batch processes, the fluoroelastomer is led at ambient temperature with no prior heating. The curatives and other ingredients are incorporated separately, and the elastomer is allowed to band. The production time for each batch varies greatly, depending on batch size and formulation. Batch-to-batch variability continues to be a major drawback in this process, in addition to high labor costs and environmental issues such as emissions. A suitable alternative to overcome these limitations is continuous processing through the use of extrusion technology. Twin-screw compounders are especially useful in processes where product consistency is paramount, with the added benefit of potentially higher throughput rates and improved control over emissions.

Application of thermoplastic twin-screw extrusion technology for processing of fluoroelastomers has not been widely investigated or reported. Managing the large amounts of heat generated during shearing of rubber compounds is critical to avoid scorching and subsequent poor physical properties. The use of thermoplastic twin-screw compounders for fluoroelastomer compounding is challenging because of the limited cooling capacity available. This is especially true at high throughput rates with residence time of a few seconds. Screw elements in co-rotating twin-screw extruders, especially kneading blocks, are designed to impart maximum shear in order to heat and melt the thermoplastic resin in a short time. However, the rheological state of the rubber is different from that of a thermoplastic resin, which is above its glass transition temperature at room temperature. Therefore, use of the standard elements would result in excessive shear heating and degradation, requiring greater control over screw and barrel design. The objective of this study was to evaluate the feasibility of developing a continuous extrusion process to manufacture accurate and precise blends of fluoroelastomer compounds.

Experimental

Laboratory trials and optimization

A commercially available fluoroelastomer material with a fluorine content of 66% was compounded with acid acceptors and fillers. Initial experiments were carried out on a Werner & Pfleiderer ZSK-30, parallel, co-rotating, twin-screw extruder. This extruder provides excellent mixing and de-volatilization capability. A unique characteristic of this equipment is the relatively low free volume when compared to more traditional extrusion systems. This characteristic enables processing of materials sensitive to shear and temperature.

Acid acceptors and fillers were pre-blended in a V-blender. The fluoroelastomer (FKM), pre-ground to 3 mm particles and partitioned with talc, and the additives blend were fed separately into the same feed barrel on the ZSK-30. A successful process was developed at a screw speed of 250 rpm and 9 kg/hr., with an exit melt temperature of 125[degrees]C. The product manufactured under these conditions met all the quality criteria listed in table 1. No die was used in these experiments. The material exiting from the last open barrel section was fed directly to a single-screw extruder for producing strips that were used in subsequent processing. This process was optimized using optimization principles to 105 lb./hr, by varying process L/D, screw configuration, barrel design and total SME (specific mechanical energy). At the optimal condition, feeding was identified as the rate-limiting factor, i.e., the conveying capacity of the extruder was exceeded at the optimum screw speed and torque combination. The optimized process from a ZSK-30 was successfully scaled up to a ZSK-58 commercial size twin-screw compounder. The details of this process are discussed in greater detail.

ZSK-58 process optimization

A unit operations methodology was employed to scale up the process. The major focal points of this development effort were complete homogenization of all raw material constituents while eliminating the potential for premature curing of the fluoroelastomer compound. The unit operations methodology utilized allowed for stage process development. The two major aspects of unit operations methodology are selection and sequencing. While selection is limited to the five basic operations and hybrids of them, sequencing can present some unforeseen challenges in optimization. The five unit operations are listed below:

* Feed--introduction of raw material components:

* melt--mastication of thermoplastic matrix;

* mix--both dispersive and distributive mixing;

* vent--de-volatilization or entrained vapor release, and

* pump--compression and pressure generation.

Although there are potential hybrids or combinations of these, the basic five are the underlying premise of all systems. Techniques utilized to accomplish these operations are infinite. The breakthrough in this methodology is the sequencing and interactions of the unit operations in an effort to maximize product performance and economic impact.

The fluoroelastomer is very temperature-sensitive in terms of its melt viscosity. The shear-thinning behavior of the material causes a decrease in mixing efficiency and product homogeneity. The small free volume of the parallel co-rotating twin-screw extruder was used to ensure proper material temperature for optimum mixing, through the use of cooling and screw geometry selections.

The mixing of the fluoroelastomer compound has a simple unit operations diagram. Figure 1 displays the unit operations selected and their sequenced relationships. The diagram has five unit operations in all, with three sequential mixing operations. These mixing operations are independent of one another due to the nature or style of mixing that they are inducing on the fluoroelastomer compound.

The first unit operation is obviously a feeding operation. This was accomplished using fast-pitched, under-cut screw elements. Due to the desire in subsequent unit operations for a slow screw speed, these elements were necessary to ensure the screw channel would approach the needed material level for shear induction.

The second is a dispersive mix/melt hybrid operation accomplished using a high shear kneading block. During this operation, significant deformation is induced to the fluoroelastomer matrix. Due to the large clumps of raw material, this unit operation is tasked with agglomeration reduction. This type of operation traditionally also induces material mastication and significant increases in material temperature. A slow screw speed ensured a high degree of fill within the channel for optimum efficiency.

The third unit operation is another dispersive mixing operation, separate from the initial dispersive mixing. This dispersive mixing operation is accomplished over 8.2 L/D. This operation completes the breakdown of the raw material components.

The fourth operation is a distributive mixing operation to ensure material homogeneity. This section utilized two 25 mm Burbank kneading blocks (BKB). These elements are designed to provide superior mixing, with minimal shear or temperature generation.

The fifth and final unit operation is pumping. This operation is dependent on the actual material application. The pumping operation could be used in conjunction with strip or pellet production. The operation is often times performed with an auxiliary piece of equipment that is more suited for efficient pumping. Typically, a single-screw extruder capable of delivering uniform pressure to either a pelletizing head or strip die is used.

This technique enables separate optimization of the mixing and forming unit operations. In an effort to speed the development process, this unit operation was not optimized. Samples were collected directly from the mixing equipment prior to pelletization or strip production for quality control testing.

Figure 2 depicts the process schematic utilized during the final production operation. In order to ensure accuracy of blend components, especially in the small free volume and short residence time of the process, gravimetric feeders were utilized. A relatively short processing length of 25:1 L/D ratio was used. Another detail of the process is the absence of a de-volatilization port. Due to the low temperatures that were targeted for the material, a de-volatilization port was not necessary to remove off-gassing vapors.

[FIGURE 2 OMITTED]

Results and discussion

A variety of quality control tests was utilized to evaluate the product that was produced. Wavelength dispersive x-ray fluorescence (WDXRF) was used to quantify the blend ratios. A very high degree of consistency was found as shown in figure 3. A scanning electron microscope (SEM), coupled with electron-dispersive spectroscopy, was employed to check the final particle sizes of the blend components. Figure 4 shows that the target cumulative particle size distribution achieved in a single pass of twin-screw extruder (labeled ARC) is better than three consecutive batch operations. Moving die rheometer (MDR) data w ere used to measure the rheological properties of the final fluoroelastomer blend, with data shown in table 1.

[FIGURE 3 OMITTED]

The final fluoroelastomer blend product produced was found to be very accurate and precise in terms of chemical composition (figure 3). This can be attributed to the accuracy of the gravimetric feeders. The proper sequencing of the unit operations in the screw configuration, aimed at minimizing heat build-up in the material, ensured that all theological properties were met. This also indicates that the material was not subjected to excessive shear heat and was not scorched. The particle size analysis indicates the superior dispersion achieved by the proper configuration of a twin screw extrusion process.

Conclusion

It has been demonstrated that co-rotating twin-screw extruders designed for thermoplastics can be successfully configured to process fluoroelastomers. The use of twin-screw extruders resulted in more consistent product with improved dispersion, lower heat history and reduced batch to batch variations. Further, labor costs, downtime and energy costs are lower. Optimization of screw design and barrel configuration helped achieve output rates of 105 kg/hr, on a laboratory scale extruder. These kinds of throughput rates enable the use of small-scale extruders for short production runs with minimal downtime. Unit operations methodology was successfully employed to scale up the process.

Table 1--quality control specifications for extruded
fluoroelastomer blends

Property                             Value                     Unit

Exit melt temperature     [less than or equal to] 130       [degree]C
t'50                            3.1 [+ or -] 0.1               min.
t'90                            4.8 [+ or -] 0.1               min.
MH                           2.6 (23 [+ or -] 0.5)         Nm (in.-lb.)
Particle size            [greater than or equal to] 90%        [mu]
                               less than 10 [mu]

Figure 1--unit operation sequence in
Twin-screw extruder

    Feed       Mix/Melt         Mix            Mix          Pump

Extruder        High          Cooling       Low shear      Cooling
Loading and     shear      and dispersive   distribute       and
Material      dispersive       mixing        mixing        pressure
Conveying       mixing                                    generation