(166e) A Fluid-Bed Continuous Classifier of Polydisperse Granular Solids

Solids mixtures produced in different fields of chemical engineering, alimentary factories, cosmetics industries, may be separated in theirs components by means of several techniques: liquid classifying, elutriation, sieving (static or vibrated), screening. Even though these techniques are applied in several cases, they suffer of some disadvantages. Solids classifications carried out in presence of a liquid phase must be carried out at low temperatures to reduce the liquid evaporation rate. Moreover, liquid classifying processes ask for a separation time often too long. Dry classification processes are often characterized by high mechanical stress of particles that may induce solids comminution.

Separation processes based on fluidization technology partially overcome above cited problems. These processes are based on the propensity of dissimilar solids to segregate or to mix evenly in a fluidized bed depending on the operating conditions. Typically, high superficial gas velocity (U) promotes solids mixing while lower values of U promotes solids segregation with stabilization of a lower fluidized bed enriched in jetsam (heavy/large) particles and a surnatant fluidized bed enriched in flotsam (light/small) particles (Nienow and Chiba, in JF Davidson, R Clift, D Harrison. Fluidization 2nd Ed. London. Academic Press., 357-382, 1985). The range of U characterized by uneven distribution of solids in the fluidized beds depends on solids properties and average bed composition. Still, the same variables strongly affect the dynamics of particle segregation in dissimilar solids fluidized beds (Olivieri, Marzocchella and Salatino, AICHE J., 50, 3095-3106, 2004).

Even though classification processes based on fluidization technology could be successful, limited information is available in literature to support design and control of apparatus.

This study reports preliminary results of a research program aiming at the separation of components of solids mixtures in a fluidized bed plant. The attention was focused on solids distribution in the classification apparatus and on its performances.

The apparatus consists of a conical shape column (0.70 m high; ID bottom 40 mm, ID top 185 mm) and a cylindrical column (1.0 m high, 40 mm ID) connected each other at the bottom by an inclined pipe. Columns, made in Plexiglas, are equipped with fluidizing gas distributors at the bottom and along the walls, solids discharge ports and pressure taps. The plant is equipped with pressure transducers, solids delivering device and a data acquisition unit.

Solids mixtures were fed in the conical shape column. The jetsam enriched stream was collected at the top of the cylindrical column, that enriched in the flotsam solids at the top of the cylindrical column. Composition and rate of enriched streams were measured by collecting classified streams at prefixed interval of time. Eventually, solids distributions along the columns were measured directly ?freezing? the bed by suddenly shutting the fluidizing gas flow off and collecting the solids slice by slice.

The plant was operated with a mixture of 120 μm silica sand (density = 2600 kg/m³) and 500 μm glass beads (density = 2540 kg/m³), in equal mass ratio. Mass rate was changed between 20 and 65 g/min. The fluidizing gas was technical air split into two streams: the first was sent to the dehumidifier and the second to the humidifier. Mixing of these two streams at fixed proportion served to the purpose of controlling the relative humidity in the range from dry to saturated.

Classification effectiveness was defined in agreement with the relationships reported by McCabe and Smith (Unit Operation of Chemical Engineering Third Edition, McGraw-Hill International Book Company, 1976). The classification effectiveness with respect jetsam (flotsam) particles was estimated as the ratio of the jetsam (flotsam) mass rate in the jetsam (flotsam) enriched stream and that in the feed stream. The product of the classification effectiveness with respect to jetsam and to that with respect to flotsam gives the overall classification effectiveness of the process.

Each run was characterized in terms of overall classification effectiveness, rate and solids composition of enriched streams and solids distribution along the columns.

Preliminary runs were aimed at finding the operating conditions characterized by steady state regime. Results showed that under the investigated conditions the jetsam fraction in the jetsam enriched stream was 100% and that in the flotsam enriched stream was 10%. Overall classification effectiveness was as high as 85%. The operation was carried out successfully for six hours.