(125a) Mixing of Freeflowing Spherical Particles in Model Cylindrical Silos with Inserts

Authors: 
Davies, C. E., Massey University
Manaf, F., Massey University
Fullard, L., Massey University
Storage of feedstocks, products and intermediaries in silos is an integral part of many processing operations and inevitably some mixing takes place while a material transits a storage vessel. Flow patterns within a silo are influenced by many factors including silo geometry, material and silo friction properties, mechanical strength of the material, and particle size. As part of a comparative analysis of mixing in model silos we have carried out flow-through drainage experiments with spherical particles ~2 mm in diameter.

We have used 80 mm diameter cylindrical vessels with conical hoppers, 3D printed in ABS plastic, to obtain reference data for half angles spanning the range 24 degrees to 90 degrees (flat bottom) for three different wall finishes, viz as-made, smoothed by acetone vapor, and roughened by a monolayer of test material glued in place. In each test run, the hopper was loaded with an equal mass (250 g) of particles of two optically distinct colours, so that there was a step change in particle color well above the hopper transition; apart from color all particle properties were the same.

We followed mixing patterns by collecting the hopper discharge in a segmented rotary sampler. The relative proportions of particles of each color in each sample were determined by a photographic technique. These trials provided indications, consistent with expected trends, of a progressive increase in the extent of stagnant regions as hopper half angle and wall roughness increased.

An additional focus here is the effect of inserts on the development of stagnant regions; the inserts used were: a sphere, small diamondback, and large diamondback. The experimental technique was similar to that outlined above, but with a single hopper with a half angle of 60 degrees and an as-made wall finish. The discharge opening was 14 mm. Experiments were carried out initially with no insert, then repeated three times with a different insert each time. With no insert, there was evidence of a significant stagnant zone; the small diamondback and spherical inserts somewhat improved the flow, and with the large diamondback insert in place, there was no evidence of a stagnant zone.