(52b) The Investigation of Cohesive Particle Mixing At Sub-Agglomerate Scale Using Discrete Element Method

Davé, R., New Jersey Institute of Technology
Deng, X., New Jersey Institute of Technology
Scicolone, J. V., New Jersey Institute of Technology

Recent experimental results show the magnetically assisted impaction mixing (MAIM), a high-shear mixing device, is capable of mixing nanoparticles at sub-agglomerate scale. Likewise, other high-shear and/or high-strain devices may be able to mix very cohesive powders due to their ability to break down agglomerates. In this work, MAIM system was used a model high-shear mixing device for cohesive particle mixing at sub-agglomerate scale. In order to better understand the dynamic process of the MAIM system, discrete element method (DEM) based modeling was employed. In a typical DEM mixing simulation, the initial state of the particles is in deagglomerated form. However, in order to truly capture the high cohesion of very fine powders that are micron or smaller in size, it would be desirable to examine them in an agglomerated form as the initial condition. Accordingly, a DEM analysis was performed to investigate the relationship between homogeneity of mixing (HoM) and various parameters such as magnets size, surface energy of non-magnet particles, and magnets-to-sample mass ratio. The cohesive force between non-magnets is based on the JKR model. Agglomerates were formed before mixing, thus better capturing the effect of cohesion on the initial state. Simulation results indicate that the mixing will be faster for decreasing surface energy, increasing mass ratio, and decreasing magnet size at fixed mass ratio. The results related to collision frequency and collision energy between magnets and non-magnets show that homogeneous mixing required longer processing times for higher surface energy since higher collision numbers and collision energies were necessary to deagglomerate the particles. Moreover, when the collision energy between magnets and non-magnets exceeds the cohesive energy, the HoM would achieve the steady value within a shorter processing period. The simulation results are in agreement with previously published experimental results, indicating the system model involving the evolution of agglomerates is applicable to cohesive power mixing.