Twin-screw machines explore solid-state extrusion.

Corotating, intermeshing twin-screw extruders have long proved their worth in a wide variety of compounding tasks. More recently, however, a whole new approach to twin-screw extrusion has started to emerge--one in which the main processing occurs while the thermoplastic material remains below its melting point. This processing strategy can enable a twin-screw to serve in several sorts of formerly unthinkable applications. So far, these have included mechanical dewatering of polymers, sintering a resin and additives, and converting pellets into powder.

At first, this "solid-state extrusion" seems to fly in the face of common extrusion theory, in which complete melting of the polymer takes place before the main mixing or reaction process starts. But the solid-state process works nonetheless because the highest shear forces in any extruder occur in the solid bed during its compaction shortly before the melting step. If the resulting frictional heat is removed efficiently, the high shear stress in the product initiates a grinding rather than melting process.

Under the same conditions, a matrix of polymer and low-viscosity dispersed droplets--such as water, oil, or waxes--can be separated by means of a shear-induced "squeezing action." The high shear levels force the low-viscosity components to travel upstream in a counter-current fashion before draining through the rear of the feed section. In a mixing role, the solid-state process can also be useful for simply blending materials that must stay below their melting points.

The three case histories here, two from commercial operations and one from tests in Berstorff's labs, demonstrate just how the corotating, intermeshing twin-screw machine can be modified successfully for solid-state extrusion and then applied to some unconventional tasks.

DEWATERING WET POLYMERS

It's no secret that high levels of moisture accompany the production of certain polymers, such as ABS. Moisture also troubles recycling processes after the washing step--even a 50% water content in reclaim is not uncommon. Twin-screw machines can "vent off" some of this moisture during conventional compounding steps. But more typically, expensive dryers are needed to reduce these water levels after compounding to a point here standard single-screw extruders can accept the product and produce a bubble-free product.

Recent tests conducted on a 40-mm-diameter twin-screw extruder (Berstorff Series ZE40A) reveal an alternative method for dewatering polymers in the extruder itself through solid-state extrusion's propensity to force low-viscosity components in the upstream direction. Though the method still employs a classic melting section followed by a vacuum degassing and pressurizing section, it features only minimal reliance on degassing (see Fig.1).

Two different polymer powders were selected for the trials--a polypropylene with a 2.7 MFI and an LLDPE with a 50 MFI. To gauge the effects of varying moisture levels, water was added in 10% increments up to 50%. Up to this point, the water stays in the powder and does not drain by itself in the feed hopper prior to extrusion. As for the extruder setup, the pilot-scale machine was configured with a side-feed unit for adding the water. The actual draining took place upstream of the side-feed unit through a standard feed-zone barrel section that had been installed upside down.

Unlike some other dewatering systems, where water is drained in high-pressure areas of the extruder through a slotted barrel, our drain and its position proved less likely to become blocked by solids or partly melted polymer. Indeed, only negligible amounts of solids were lost to the dewatering step. When 40% water was added to the feedstream, 34.2% of this amount was collected at the rear drain, and it contained a mere 0.1% solids. At a 50% water level, the amount of solids increased to only 0.3% (Fig.2).

The test results show how mechanical dewatering offers some compelling advantages to the processor. Energy savings from the mechanical dewatering step vs. a drying or degassing process are especially significant. At higher water contents, the energy input can be three times higher when water has to be heated and steam removed through the vent (Fig.3). The favorable results apply to both PP and LLDPE. For PP, especially at the low water content of 10%, the process has a high efficiency of 92% (Fig 4). Aside from modifying the twin-screw extruder for water separation, the principle can be applied to other separation processes where other low-viscosity components, such as waxes or oil, must be reclaimed from the polymer matrix.

SINTERING FLUOROPOLYMER POWDERS

Solid-state extrusion also provided the answer to a customer's mixing problem by sintering a fluoropolymer with an additive, as opposed to conventional mixing during the resin's melt phase.

For a wire-coating product used in the aerospace industry, a liquid chemical blowing agent had to be dispersed in fluoropolymer powder and converted to pellets. These pellets would eventually go into a crosshead extruder for the wire-coating process. The resulting foam offers good electrical and thermal-insulation properties, both of which persist even under high temperatures or flames.

In this case, a straightforward set of processing objectives emerged. The polymer and blowing agent had to be mixed while building up enough pressure for water-ring pelletizing. At the same time, the process had to sidestep the very real potential for premature foaming due to activation of the blowing agent during the pelletizing process. This final objective presented the greatest challenge because the activation temperature of the blowing agent was 36[degrees] F below the polymer's melting point.

To solve the problem, the mixing sections of the twin-screw were operated without pressurization, which created a tumbling action. This action in turn diffused the liquid blowing agent into the polymer powder. The low coefficient of friction of the polymer particles and the additional lubricating effect of the blowing agent allowed compaction of the two components while minimizing frictional heat build-up. It was, however, critical to reduce the die pressure by about 500 psi in order to stay below the activation temperature. Meanwhile, any process upsets could be detected immediately when pellets started to float in water, indicating that they had foamed.

In the end, the sintered pellets had enough strength to survive both treatment in a centrifugal dryer and shipment to the wire and cable plant.

CONVERTING MELT TO POWDER

Traditionally, twin-screw extruders have performed an important role in converting polymer powders into melt and then pellets. By contrast, solid-state extrusion is a way to reach the opposite target of reducing a fused polymer mass to a powder. In one commercial application, these tiny particles ultimately see use in rotational molding.

So far, the best results and most consistent powders have been achieved with LDPE--in this case one with a 2.0 MFI. The process begins with a normal melting of the pellets. In the second part of the extruder, intense barrel cooling is applied to reduce the polymer temperature below its solidification point. To give an indication of just how intensive the cooling had to be, the specific energy added to this process was 0.3-0.5 kwh/kg--or about twice the amount normally needed to melt LDPE. All this "extra" energy had to be dissipated through the barrel with the help of internally cooled screws.

In the absence of normal levels of frictional heat, the high shear forces present in any extruder solids bed then begin to crush the polymer. This grinding process continues until the particle size diminishes and the finished powder exits the extruder via an open discharge. So far, the production rates have been limited by the cooling capacity; they are approximately one-tenth of the rate the twin-screw extruder is capable of when melting and compounding.

The process itself, which is the subject of a 1986 joint patent between Berstorff and the Soviet Academy of Science (U.S. Patent #4607797), is still shrouded in speculation about the best way to achieve uniform powder of a given size. For now, it seems that LDPE tends to produce more uniform particle shape and size than do other resins.