Efficient DEM modeling of solid flavor particle mixing in a rotary drum

Rutger van Sleeuwen,Stefan Pantaleev,Mohammadreza Ebrahimi,Leo Feng

DOI: https://doi.org/10.1016/j.powtec.2024.119559

IF: 5.64

2024-02-24

Powder Technology

Abstract:An efficient Discrete Element Method (DEM) modeling methodology for mixing of solid flavor particles in a rotating drum is presented. Machine learning and optimization algorithms are combined with geometric scaling and particle size distribution trimming to produce several orders of magnitude reduction in computational expense relative to the traditional approach of modeling at the physical system scale and performing purely simulation-based model calibration. The validity of the methodology is evaluated by comparing DEM model results for the evolution of the mixture uniformity with a laboratory scale drum and good agreement is observed. The limitations of the approach with respect to the amount of data required and the degree of geometric scaling and particle size trimming that is possible are also discussed. The approach developed could serve as an example for future simulation studies on fine particles in various processes that are relevant to the food, ingredient and pharmaceutical industry.

engineering, chemical