Velocimetry and response surface regression analysis of a swirling bed tested with mesh-coupled axial blade distributor

Abstract

Fluidized bed technology is being used in several industrial processes. The well-controlled fluidization of a bed material is a prerequisite for successful implementation of this technology. Several variants of distributors are being tested for fluidized beds; however, scant information is available on their working regimes. In this study, an annular distributor was constructed by fixing blades between two concentric rings and covering the assembly with a wire mesh. This new distributor design was tested for fluidization of batches of a spherical bed material at different blade angles and superficial air velocities (SAVs). High-speed imaging and MATLAB supported velocimetry techniques were used to generate velocity vector fields of fluidized beds. These fields were analyzed further for bed velocity measurements at different bed zones. A response surface regression model was applied on the data to predict the response of the bed parameters. Statistically, the bed velocity was predicted to be about 0.497 m/s with the confidence interval in the range of 0.4948-0.4999. The experimental bed velocity remained between 0.531 m/s and 0.538 m/s under an optimized blade angle of 60°, SAV of 2.3 m/s, and bed weight of 500 g.