(745a) Accelerated Quantitative DEM Simulation of Large Screw Feeder and Tablet Press Feeder Using Adaptive Coarse Graining Technique

Coarse-graining of particles is unavoidable for practical DEM simulations. For pharmaceutical industry, the typical dimensions of the production scale equipment are ~ 1 m whereas the typical powder contains particles around 0.1 mm. This size ratio requires the coarse-graining ratio to be very high which introduces large inaccuracies in the simulation. To tackle this situation, recently an adaptive coarse-graining strategy has been proposed [1]. This newly proposed technique has been demonstrated using a simple hopper discharge. However, refinement of the algorithm and application for a variety of systems is pending. In this work we will apply the adaptive coarse-graining technique for two problems of practical importance: tablet die feeder and hopper screw feeder.

It will be shown that the adaptive coarse-grained simulation of tablet die filling reveals additional features that were not otherwise seen in the fixed coarse-grained simulations. This is because the model particles from fixed coarse-graining are much larger than actual particles that introduces artifacts near the die mouth (like exaggerated weight RSD and porosity).Jamming in the feeder section is also observed.As shown in figure 1, resolved particles can be taken near the mouth of the die for accurate estimation of the filling process, when adaptive coarse-graining technique is used. Slightly coarse-grained particles will be taken near the paddle wheel for correct estimation of the flow, and larger coarse-grained particles will be taken to fill up the feed chute. With this strategy, the system can be simulated accurately with reasonable computation.

A different type of challenge exists in implementing adaptive coarsegraining for the hopper screw feeder system. Because of the large recirculation near the feed end of the screw, adaptive coarse-graining induces repeated resolution and consolidation of same set of particles which eats up computational resources and renders adaptive coarse-graining less attractive for such systems. It will be shown that an innovative zoning strategy (snapshot given in figure 1) can resolve this issue and the adaptive coarse-graining can improve the efficiency of such systems substantially.

References:

1. A Multi-Level Coarse Graining Technique for Discrete Element Method (DEM) Simulation. Jayanta Chakraborty, Tarun De, Jitendra Kumar, Salvador García-Muñoz, Maitraye Sen. AIChE Annual Meeting, 2020