A Semi-Automated DEM Parameter Calibration Technique of Powders Based on Different Bulk Responses Extracted from Auger Dosing Experiments

Discrete Element Method (DEM) proved to be an essential tool to optimize the industrial auger dosing process for pharmaceutical powders. During the DEM parameter calibration process of a certain powder, several parameter combinations might lead to a similar bulk response, which could also vary for other bulk responses. Therefore, a methodology is needed in order to narrow down the number of combinations and be at once close to reality representation. In this study, a vertical auger dosing setup is used as a standard calibration device to extract three different bulk responses, i.e., angle of repose, bulk density, and mass flow rate. Simulations using LIGGGHTS software package are performed based on Design of Experiments (DoE) by varying four input factors, i.e., auger speed, particle-particle and particle-wall static friction coefficients, and particle-particle rolling friction coefficient. The successful application of multivariate regression analysis (MVRA) results in predicting the bulk behavior within the studied ranges of parameters. In this regard, clustering the different predicted behaviors of the three responses together allows to dramatically reduce the admissible parameter combinations. Consequently, an optimized set of calibrated DEM parameters is chosen, where the simulation results accurately match the experimental reference data. This simple dynamic calibration tool proves to strongly verify and predict the flowability of free-flowing bulk materials.

Automated summary

The study optimized the industrial auger dosing process for pharmaceutical powders using the Discrete Element Method (DEM) and proposed a method to reduce the number of parameter combinations while approaching reality. Multivariate regression analysis successfully predicted bulk behavior within the studied parameter ranges.