An experimental verification of Glicksman's reduced scaling

A dimensional analysis of the momentum balance in circulating fluidized beds reveals that the fluid dynamics of these units is governed by five dimensionless groups. However, it is not possible to match the hydrodynamics of a range of industrial circulating fluidized beds using a single laboratory facility operated at ambient temperature and pressure. To address this difficulty, Glicksman, et al. (1993) proposed a "reduced set" involving only four dimensionless parameters. They showed that the "reduced set" is complete when the drag on solid particles conforms either to the viscous or to the inertial limit. They assumed that it would also succeed in the intermediate regime, where many industrial beds operate. We verify this assumption by comparing two series of experimental data obtained in the same facility with different combinations of gases and solids that preserve the "reduced set" but exhibit a mismatch of the complete five-parameter set of dimensionless numbers. We find that the dimensionless profiles of pressure along the flow and solid volume fraction in the radial direction are nearly identical in the two series of experiments. These observations validate the reduced scaling proposed by Glicksman et al. for the intermediate regime, at least for the simulated pressurized conditions of our experiments. INTRODUCTION The hydrodynamic scaling of circulating fluidized beds (CFB) remains an industrial challenge. In an effort to increase process efficiency, a recent trend in coal combustion technology has been to pressurize CFB units. Under these conditions, studies of hydrodynamic scale-up are limited. One approach is to construct a smaller pilot facility of similar geometrical aspect ratios fluidized with inert gas mixtures of adjustable density at ambient pressure and temperature. The five dimensionless groups governing the flow in the pilot are then matched to those in the industrial unit to achieve hydrodynamic similarity (Louge, et al., 1999). A difficulty with this method is that the matching imposes onerous constraints on the size of the pilot and properties of the gases and solids. To address this difficulty, Glicksman, et al. (1993) proposed a simpler approach that relaxes these constraints. They noticed that, in the viscous and inertial limits of the drag on individual solid particles, the momentum balances of the two phases reduce to a form giving rise to only four dimensionless groups. They then postulated that this “reduced set” of parameters would also govern the fluid dynamics of suspensions operating