Extrusion lines are used for the continuous processing of plastically deformable substances, for example polymers. The best-known extrusion systems include single and twin-screw extruders. Planetary roller extruders (PWE) are a special design and are classified as multi-screw extruders. The design of the PWE is similar to a planetary gear with a centrally arranged central spindle, rotating planetary spindles and a stationary roller cylinder (see Figure 1). The PWE is suitable for a variety of applications due to the various adjustable spindle configurations, precise temperature control and high mixing quality [Kohlgrüber 2019]. For example, the Virtualization of Process Engineering Processes working group at Ruhr University Bochum has succeeded in producing polymer foams with the PWE using compressed carbon dioxide [Winck 2020]. 

Figure 1: Cross-section of a planetary roller extruder [Radwan 2023]

When operating the PWE, the material data and the operating parameters of the system are of fundamental importance for achieving optimum product properties. Of the materials, viscosity and density are particularly relevant. In terms of the system parameters, the speed, the external temperature control of the central spindle and roller cylinder, the mass flow and the spindle configuration are essential. How these parameters should be set to achieve optimum processing conditions (temperature, pressure, mechanical energy input, dwell time) is usually determined by experimental preliminary tests before the start of production. To date, these have been run on the system in the desired size with the desired target values. As the PWEs are available up to a throughput of several tons per hour [Kohlgrüber 2019], these preliminary tests can be associated with a high consumption of resources (material, energy, time). Consumption increases with the size of the system.

Using the newly developed laboratory PWE from ENTEX Rust & Mitschke GmbH, the optimum production settings of the PWE can be determined in a much more resource-efficient manner. This size is characterized by the fact that the system only has a small process volume compared to previous sizes, but its general design corresponds to that of large systems. This means that similar tests can be carried out across all sizes, whereby the resources required are significantly reduced compared to large systems.

However, so-called scaling algorithms are required in order to be able to transfer the test parameters and results between different sizes. These methods can be used, for example, to convert the throughput for different sizes. This step is necessary because, for example, the throughput on a large system can be more than eight tons per hour, while the maximum throughput for laboratory PWE is around ten kilograms per hour [Kohlgrüber 2019].

Scaling algorithms have already been extensively researched for single and twin-screw extruder designs. However, in the case of the PWE, there have been no approaches to date to make results comparable across different sizes. In this work, initial approaches to scaling the PWE were developed and tested in extensive test series for processing polyethylene and polypropylene on both large and small scales. Further details on the methodology and initial results can be found in the publication [Radwan 2023]. The further objectives of this research project include the experimental and simulative evaluation and optimization of these scaling approaches for the PWE as well as the prediction of process conditions using machine learning.