(69c) Discrete Element Method (DEM) Investigation of Particle Properties on Bulk Powder Flowability:  Toward a Particle Calibration Library for Pharmaceutical Materials

Authors: 
Ketterhagen, W. R., Pfizer Worldwide Research and Development
Shoemaker, B., Pfizer Inc.
Cole, S., DEM Solutions Ltd.
Curry, D., DEM Solutions Ltd.
Application of the discrete element method (DEM) to industrial processes can provide insight into the performance of powder processes. For improved confidence in simulation results, the simulation input parameters, namely the particle contact parameters, must be known with some certainty. For some processes, e.g. tablet coating operations, the particle size is such that the contact parameters can be directly measured and the number of particles is small enough to permit reasonable simulation times. In most industrial processes, however, the particle size may be too small to make meaningful direct measurement of the contact parameters and the number of particles far exceeds what is computationally tractable in an industrially-relevant time scale. One common approach to address this shortcoming is to make simplifications to the particle shapes, size, and size distribution such that a smaller number of particles are simulated. In doing so, the link between contact parameters for the real and model particles is dubious if not non-existent. A particle calibration approach where contact parameters for the model particles are selected such that the simulated bulk powder behavior matches that of the real material may be a useful approach to simulate large-scale processes with improved confidence.

In this work, the shear flow of particles in a periodic section of an annular ring shear tester is simulated to examine the effects of various particle properties on the bulk flow behavior. The impact of particle shape along with coefficients of restitution, sliding and rolling friction, and cohesion are examined through thousands of cloud-based DEM simulations. The simulated bulk flow behavior spans the range of experimental results determined for a wide array of typical pharmaceutical materials. Thus, this extensive simulation dataset provides a prerequisite to a potential DEM particle calibration utilizing a lookup table for contact parameter values. Simulated and experimental shear tester results will be presented with a focus on the link between the DEM input parameters and macro-scale bulk flow measurements. Further, an outlook for extension to calibration of DEM particle models will be discussed.