GPU-based discrete element simulation on a tote blender for performance improvement

Abstract

The mixing and flow of granular materials in a conical tote blender are investigated using a GPU-based DEM software to explore new approaches to enhance mixing. The structure and dimensions of the blender and other simulation conditions are set according to experimental data from literature. A parametric study on fill level and rotation rate is carried out from which optimum values are found with respect to mixing rate, productivity and energy consumption. It is also found that the standard horizontal installation of the blender results in poor axial mixing, while inclining the blender at a certain angle can enhance mixing effectively. This may be ascribed to the larger mean particles velocity and better velocity distribution under such conditions, which is confirmed to be a consistent character for relatively large-scale systems. Furthermore, the effect of operating conditions for the inclined blender is also examined. It is close to those for the standard blender but has less effect on mixing rate. Inclining the blender at a certain angle is found to be an effective and simple approach to speed up mixing especially if the inclining angle varies from 30 to 60°. This may be ascribed to the larger mean particle velocity and better velocity distribution under such conditions which are confirmed to be consistent character for relatively large-scale systems. ► We explore new approaches to enhance mixing for a conical tote blender. ► The optimal fill level and rotation rate are found with respect to industrial demand. ► Inclined the blender at a certain angle can enhance mixing effectively. ► We optimize the inclined angle. ► We examine the effect of the inclined angle on axial and radial particle velocities.