The influence of particle size and vessel capacity on segregation and scale-up of granular dynamics in tumbling blenders

Tumbling blenders of increasingly more complex geometry are used as model systems for experimental investigation of segregation tendencies and scale-up criteria associated with variations in both particle size and blender capacities. The rotating cylinder, which represents the simplest possible geometry, is used to measure variations in particle velocities with changes in cylinder diameter and rotation rate. A new non-dimensional scaling parameter is introduced that effectively matches velocity profiles and an independent particle size effect is also determined.; Segregation dynamics of glass beads mixtures containing particles of two sizes are investigated in the rotating cylinder, the double cone blender, and the v-blender. In the rotating cylinder, the relative size of the particles in relation to the cylinder diameter is shown to significantly influence segregation dynamics. Specifically, when the cylinder diameter is less than 40 times larger than the average particle size, axial segregation is completely inhibited, regardless of particle size ratio or vessel rotation rate.; In the double cone blender, a number of segregation patterns were found to form when a single mixture was rotated at different rotation rates from the same initial conditions. Furthermore, simply changing the particle size ratio at a single rotation rate also resulted in varying pattern formation. Specific mechanisms are identified that produce the observed. Changes in particle size ratio are shown to have an impact on both relative particle velocities and the relative mobility of large particles in comparison to smaller particles.; In the most complex blender geometry, the v-blender, segregation pattern formation is also studied with changes in vessel size and fill level. A number of segregation patterns are shown to form in a single vessel with changes in rotation rate and/or fill level and a new segregation mechanism, surface-flow induced trajectory segregation, is introduced. These patterns are found to form in vessels ranging in capacity from 1 to 30 quarts. A ‘transition rotation rate’ between two patterns is determined for each blender size and the non-dimensional parameter presented to scale particle velocities is shown to effectively predict the transition rotation rate at all blender scales tested.