Reports: AC9

Back to Table of Contents

43393-AC9
On the Description of Flows Involving Polydisperse Solids

Christine M. Hrenya, University of Colorado (Boulder)

During the first year of the grant cycle, a kinetic theory of N discrete species (that vary in size and/or mass) was developed starting with the Enskog equation, and making no assumptions regarding the form of the velocity distribution or partition of energy. Although previous models have been applied to describe the behavior of binary mixtures, relatively little work has been done in applying these theories to continuous size distributions. As a first step in this direction, work this past year focused on gathering experimental data for purposes of validation.

The experimental system under consideration is a low-velocity, gas-fluidized bed composed of a Gaussian size distribution of Geldart Group B particles. The fluidizing gas is air, and sand is used as the particulate medium. A Plexiglas column of either 12 cm or 18.5 cm in diameter is used in conjunction with a stainless steel distributor plate. The superficial velocity is determined based on upstream measurements from an orifice plate meter. Pressure drops across the orifice plate flow meter, across the distributor plate, from above the distributor plate to a point 20 cm above it, and across the entire fluidized bed are measured using differential pressure transmitters.

Axial and radial concentration profiles are obtained by means of bed “freezing”, slicing, and sieving. In particular, following a stabilization period for a given experiment, the air supply was rapidly shut off and the plenum vented to“freeze” the bed. The frozen bed was subsequently sliced in bands 4 cm in height or smaller using a slicing tool. Each of the 16 slicer sections was vacuumed out of the column. The particles were then separated by size via sieving and then weighed to determine the components' mass fractions. Note that this method is restricted to relatively low gas velocities since bed expansion in low in such systems, thereby precluding significant segregation during free-fall.

In addition to the segregation measurements described above, bubbling measurements were also obtained using a fiber-optic probe. The probe consisted of two tips in parallel, each with an individual emitter and receptor. An analysis of individual traces and their correlations can be used to obtain information on the bubble frequency, size and velocity. Furthermore, this probe can be inserted at various axial and radial positions throughout the bed, giving local information on bubbles throughout the bed. Unlike the segregation measurements, these measurements are not limited to low velocities.

Both segregation and bubbling measurements were taken over a range of low gas velocities and particle size distribution widths. The bubbling measurements were also obtained at higher velocities for the reasons described above. Overall, the segregation results indicate the tendency of large particles to reside preferentially near the bottom of the bed (axial segregation), with a slighter tendency toward the walls (radial segregation) at some bed heights. The bubbling data is in the process of being analyzed. Both data sets will be used for purposes of validating the theory developed earlier in the project.

Back to top