An Efficient Approach to DEM Simulation of Fine Powder Mixing using Machine Learning and Optimization Techniques

Mixing of fine powders is a common unit operation in many industries, most prominently in the pharmaceutical industry. DEM simulations can provide valuable insights into this process, allowing for rapid product development and optimization of processes and equipment. Using machine learning and optimization techniques, this study proposes a novel approach to material model calibration for DEM simulation. Through image analysis, Lacey's mixing index and volume fraction of an ingredient were evaluated at a region of interest. These responses were then captured in the simulation by optimizing material interaction parameters.

Presented by Ambrish Singh, Solutions Engineer- EDEM at Altair as part of the virtual event: DEM Simulation for Pharmaceutical Manufacturing Processes - Latest Advances, Methodologies & Applications in June 2025.