Please login first
Investigation of mixing non-spherical particles in a double paddle blender via experiments and GPU based DEM modeling
1 , 2 , * 2 , 2
1  Toronto Metropolitan University
2  Ryerson University
Academic Editor: Blaž Likozar

Abstract:

Solid mixing is an essential unit operation in various industries including food, pharmaceutical, and cosmetics. Thus, a better understanding of the underlying mechanisms affecting the performance of solid mixers can help industries determine the best design and operating conditions for the mixing processes. Both experimental and numerical techniques have been used to analyze the operations of the solid particle blenders. The limitations associated with experimental techniques such as disturbance of the granular flow path, cost, and cumbersome implementation made the Discrete Element Method (DEM) a complementary tool to obtain comprehensive particle-level information about mixing systems. However, the DEM technique suffers from high computational time and requires enormous computing power. To address these challenges, some DEM studies have used spherical particles models even though the experimental and numerical studies have demonstrated the pronounced effect of the particle shape on the mixing quality. However, the use of graphics processing units (GPUs) has enabled us to run DEM simulations of mixing systems containing non-spherical particles with less computation time. In this study, mixing kinetics and flow patterns of non-spherical particles in a horizontal double paddle blender are investigated using both experiments and DEM. Firstly, a set of experimental data is obtained using image analysis from a rotary drum containing cubical and cylindrical particles. Additionally, the EDEM v2021 commercial software is utilized as the GPU-based DEM solver. Then, the DEM model was calibrated using the experimental data. Using the calibrated DEM model, the effects of operating parameters such as vessel fill level, particle loading arrangement, and impeller rotational speed on the mixing performance are examined. The relative standard deviation (RSD) is calculated to assess the mixing performance. Analysis of the DEM results enables us to find the optimum design parameters and processing conditions for the mixing of non-spherical particles in a double paddle blender.

Keywords: Discrete element method (DEM); Solid mixing; Non-spherical particles; Double paddle blender; Mixing kinetics and mechanism; Batch mixing
Top