(81f) Particle-Level Residence Times in a Twin-Screw Feeder

Residence time distribution (RTD) and flow sheet modeling have proven a valuable tool for material tracking in a pharmaceutical continuous manufacturing process. Feeders are often considered ideal first-in-first-out systems in flow sheets, neglecting the intermixing between batches inside the feeder. In addition, experimental methods to measure the RTD of a discharging feeder are complex and material intensive.

In this work, we present a discrete element model (DEM) of a full discharge of a twin-screw feeder. Feeding is an inherently transient process; there is no steady-state condition. The DEM software XPS (extended particle system, [1]) is capable performing a 16-minute-long discharge process within two months on a single consumer-grade GPU (Nvidia GTX 1080Ti). The initial number of particles was 2.5 million, with the particle number decreasing over time due to discharge. Upon discharge, the starting position, particle radius, and residence time of each particle is logged before the particle is deleted from the system. The log files of individual particle residence times using a free flowing material are available as open dataset [2].

With this dataset it is possible to extract information that is relevant at the process level, for example the feed factors over time. In addition, because the residence time and starting position of each particle are known, it is possible to extract RTDs of arbitrary spatial regions in the feeder. The RTD data per layer enables material tracking inside the feeder for different refill levels [3]. Thus, the particle-level residence time data enable the development of a refill strategy considering both the feed rate and the intermixing of batches inside the feeder.

References

[1] T. Forgber et al., „Extended validation and verification of XPS/AVL-FireTM, a computational CFD-DEM software platform“, Pow Tech, 361, 880–893 (2020). doi:10.1016/j.powtec.2019.11.008.

[2] P. Toson und J. G. Khinast, Dataset: „Particle-Level Residence Time Data in a Twin-Screw Feeder“. Mendeley Data (2019). doi:10.17632/D76RZZD8R7.2.

[3] P. Toson und J. G. Khinast, „Particle-level residence time data in a twin-screw feeder“,
Data in Brief, 27, 104672 (2019). doi:10.1016/j.dib.2019.104672.