PROCESSING POLYETHYLENE TEREPHTHALATE ON A SINGLE SCREW EXTRUDER WITHOUT PREDRYING USING HOPPER- AND MELT DEGASSING Walter Michaeli Torsten Schmitz Institute of Plastics Processing at RWTH Aachen University Abstract Conventional PET processing requires predrying, an energy- and cost-consuming process limiting production flexibility. The paper presents combined hopper- and melt degassing in a single screw extruder as a real alternative to predrying and investigates the influence of extrusion- and degassing parameters, screw-design and water content on both intrinsic- and melt viscosity.
Introduction The amount of processed polyethylene terephthalate (PET) has been increasing rapidly for several years, thus rising its economic importance. In 2002, the PET market has grown 17 % in Asia, 11 % in Eastern Europe, 8,6 % in North America, 8,4 % in West Europe, 8,3 % in South America, 8,1 % in Middle East and Africa and 4,7 % in Japan. World consumption of PET has thus reached a volume of 31,3 million tons in the year 2002 [1]. When processing PET, the challenge of coping with its various possibilities of degradation has to be met. Mechanical, thermal, oxidative and especially hydrolytic degradation exert negative influence on the resulting properties of the product and have to be kept within certain limits. In short terms, PET has to be processed applying a low shear stress- and temperature-level, exposure to oxygen must be minimized and the PET pellets have to be dried thoroughly in order to achieve a sufficiently low water content. As predrying of PET pellets is an energy- and cost-consuming process limiting production flexibility due to long residence times, alternatives for efficient water removal are highly welcome. A very promising method, combined hopper- and melt degassing in a
294 / ANTEC 2004
single screw extruder, is presented in this paper. The Institute of Plastics Processing at RWTH Aachen University has installed a pilot plant for the production of polyethylene terephthalate film with a single screw venting extruder. The influence of water content, screw design as well as degassing and extrusion parameters on film quality, intrinsic viscosity (iv) and melt viscosity is investigated [2-7] with this equipment. Main objective of these experiments is to avoid hydrolytic degradation or to even build up molecular weight, thus presenting a highly cost- and energy-efficient method for processing PET without predrying.
Theoretical Background of Melt Degassing The process of melt degassing involves the transport of volatiles to a polymer/vapor-interface, evaporation at the interface and their removal through a vacuum system. In addition to simple diffusion, the degassing process is in many cases enhanced by a foaming process: bubbles are formed in the melt, containing the substances to be removed. The bubbles may grow, coalesce and finally rupture at the polymervapor interface, thus releasing their volatile contents to the vapor phase [8]. Diffusion can be described using Fick´s laws. Deriving from these, criteria for an efficient melt degassing can be developed. Increasing melt temperature leads to lower viscosity, thus endorsing bubble growth, and to a higher diffusion constant. However, rising melt temperatures may also lead to thermal degradation. The concentration gradient of the volatile at the polymer/vapor-interface may be increased by reducing the partial pressure of the volatile in the gas phase, which can be achieved by applying vacuum. Polymer residence time in the venting zone should be as high as possible in order to improve degassing. On the other
hand, the risk of thermal degradation rises with residence time. High rotational speed and specific design of the screw in the venting zone lead to a regular renewal of the melt layers, thereby using the highest possible concentration gradient at any time. High rotational screw speed, on the other hand, leads to a significant amount of heat produced through shear, thereby increasing the melt temperature. Further demands on degassing technology are a completely molten product in the venting zone, an abrupt and large pressure drop before the venting zone in order to initiate foaming, a large screw channel volume to supply sufficient room for the foaming polymer, thin melt layers, a narrow residence time distribution provided by steady material transport, large diameter vent openings in order to reduce gas flow velocity and pressure loss, and sufficient heating capacity to compensate the enthalpy of evaporation.
Description of Equipment and Process At IKV Aachen, a 60 mm single screw venting extruder with a length of 38D has been integrated in a flat film extrusion line for processing PET, Fig. 1. The pellets are fed into the extruder by a vacuum hopper system.They are molten in the first 20D of the extruder and subsequently conveyed into the venting zone with a length of 9D and 3 degassing domes. After the following compression- and metering-section, a melt pump has been installed for building up melt pressure. The melt pump can also be used to vary the extruder back pressure. The two-stage vacuum pump system is able to provide a large volume flow rate and a very low pressure in the degassing zone.
Methods of Analyzing the Process Quality The common analysis method to determine the average molecular weight of a PET sample is the intrinsic viscosity (iv). With a suitable solvent, for example dichloroacetic acid, solutions are made from the PET samples. These solutions are afterwards analyzed in a capillary rheometer. This method, however, is very time-consuming, contains several possible sources of error and it does not help the processor to evaluate the process and product quality in time. Therefore, an online criterion is introduced, using conventional melt pressure measurement.
The pressure loss of a known geometry, in this case the flat film die, the melt temperature and the throughput are measured during the process. The pressure loss depends on the geometry of the flow channel, the volumetric throughput and the melt viscosity. The latter depends on shear rate, temperature and molecular weight. As a result, when shear rate and temperature are constant, the melt viscosity contains information on the molecular weight. When investigating degassing efficiency, the well-established process of predrying serves as a reference. Such a comparison between the conventional processing of predried material and combined hopper- and melt degassing of non-predried material can be seen in Fig. 2. Here, intrinsic viscosity and pressure loss of the flat film die are shown in dependence of the melt temperature. While the processing of predried material is almost indifferent to changes of the melt temperature within the shown temperature range, the degassing efficiency increases with higher melt temperatures, as it can be seen from the intrinsic viscosity. The pressure loss indicates the same effect: The pressure loss of the melt decreases at elevated temperatures when processing without degassing. By employing the degassing technology, however, this decrease is compensated with an increase in molecular weight. The figure also shows that the single screw venting extruder used for the experiments is suitable for processing high quality film from non-predried material. The classification figure for the further evaluation of the degassing quality is called viscosity ratio (VR). The method for calculating the VR is shown in Fig. 3. With thoroughly predried material, pressure loss in a certain geometry is measured for a variety of process points. Using the Hagen-Poiseuille equation, the viscosity is then calculated for these process points. By employing temperature shift (WLF) and the Carreau-Equation, a viscosity master curve is developed. From this curve, one can calculate a viscosity corresponding to every single process point measured in the degassing trials, with exactly the same volumetric throughput and temperature. A VR larger than one, accordingly, means a higher molecular weight in the polymer than it can be achieved with predrying under the same process conditions.
ANTEC 2004 / 295
Results of Melt Degassing Fig. 4 shows the influence of the melt degassing pressure on the pressure loss of the die as well as on the viscosity ratio. One can here observe a linear dependence. It can also be seen that the viscosity ratio is a suitable means for describing the degassing quality. The influence of volumetric throughput, represented by the rotational speed of the melt pump, on the VR is shown in Fig. 5. With combined hopperand melt degassing a VR=1.07 has been achieved for 20 rpm of the melt pump, which equals about 60 kg/h. The local maximum can be explained as follows: At low rotational speed of the screw, there is a long residence time of the melt in the degassing section, which leads to a good degassing. With higher rotational speed, the residence time decreases but the renewal rate of the melt layers and the bubble disruption speed increase, which are important for the degassing quality. Accordingly, somewhere between low and high screw speed, a maximum occurs.
Hopper Degassing Tests have shown that pellets with an initial water content of 1650 ppm (parts per million) do not release a significant amount of water at 20 mbar and room temperature, Fig. 6. This can be explained by a low diffusion coefficient in combination with large diffusion distances from pellet core to its surface. Pellets which have been sprayed with water 24 hours before the test and are exposed to the same vacuum conditions show a greater reduction of water content. This can be traced back to a higher concentration gradient and smaller diffusion distances, because the water has not yet reached the pellet core. The third test has been carried out using flakes made from 0.5 mm sheet. Because of their large surface-tovolume ratio, both the initial water content and the water loss during vacuum exposition are significantly higher than observed with pellets. Although an exposure to vacuum does not significantly decrease the moisture content of the pellets, experiments on hopper degassing show a great effect on the molecular weight, indicated by the pressure loss of the die, Fig. 7. This can be attributed to reduced oxidative degradation and moisture
296 / ANTEC 2004
reduction in the feed section, where the diffusion coefficient is higher due to the elevated temperature of the pellets. The results show that employing hopper degassing has a positive effect on the molecular weight of the PET melt and that it is especially recommendable for processing bottle- or film flakes. Further experiments have to be carried out investigating the influence of material and extrusion parameters, especially the temperature of the feed section.
Conclusions and Perspective A pilot plant for processing PET with combined hopper- and melt degassing has been installed. For the evaluation of the degassing efficiency, the viscosity ratio has been introduced as a classification figure comparing the well-established process of predrying with the degassing technology. Experiments have shown that with degassing it is possible to process non-predried PET with the same or higher molecular weight as with predrying. The influence of the process parameters throughput and degassing pressure has been discussed. In addition, experimental results concerning the hopper degassing have been presented. In further investigations, a direct correlation between intrinsic viscosity and melt viscosity will be established using an online rheometer. Further tests will also show the influence of different screw geometries. In a second step, a flat film extrusion line with an 80 mm degassing extruder will be designed, using the results from the 60 mm pilot plant.
Acknowledgements The investigations set out in this report received financial support from the German Federal Ministry of Education and Research (01RC0193), to whom we extend our thanks. We would also like to thank KoSa GmbH & Co. KG and M&G Polimeri Italia for the donation of raw materials.
References
Illustrations
1. N.N.: PET-Rasantes Wachstum bei Verpackungen setzt sich fort. KI - Kunststoff Information 1671, 27.10.2003 2. SCHMITZ, T., MICHAELI, W.: Extrusion von Kondensaten - Kann man sich das Trocknen sparen? Conference Proceedings of Folienextrusion - Rohstoffe, Verarbeitung, Anwendungen. Institut für Kunststoffverarbeitung, Aachen, 2003 3. MICHAELI, W., SCHMITZ, T.: PET für Verpackungen Recycling und Direktverarbeitung. Kunststoffe 93 (2003) 10
Fig. 1: Laboratory Extrusion Line
4. MICHAELI, W., SCHMITZ, T.: Entgasen von PET auf Einschneckenextrudern - Kann man sich das Trocknen sparen? Plastverarbeiter 54 (2003) 9 5. MICHAELI, W., SCHMITZ, T.: Degassing Polyethylene Terephthalate on a Single Screw Extruder without Pre-Drying. Conference Proceedings of PPS 2003, Athens, Greece, 2003 6. MICHAELI, W., SCHMITZ, T.: Ist PETVortrocknung überflüssig? K-Zeitung, 16.10.2003, S. 13-14
Fig. 2: Influence of Melt Temperature
7. MICHAELI, W., SCHMITZ, T.: Entgasen von PET in der Extrusion (2003), Conference Proceedings of Innovative Compoundieraufgaben: Herausforderungen, Perspektiven, Lösungen. Institut für Kunststoffverarbeitung, Aachen, 2003 8. ALBALAK, R. J.: An Introduction to Devotilization. In: Polymer Devotilization, Marcel Dekker, Inc., Cambridge, 1997, S. 1-12
Key Word Index PET, Drying, Melt Degassing, Hopper Degassing, Single Screw Extruder, Intrinsic Viscosity, Melt Viscosity
Fig. 3: Calculating Viscosity Ratio
ANTEC 2004 / 297
Fig. 4: Influence of Degassing Pressure
Fig. 5: Influence of Throughput
Fig. 6: Influence of Vacuum Residence Time
298 / ANTEC 2004
Fig. 7: Influence of Hopper Degassing
