Bubble size and frequency in corrugated-wall bubbling fluidized beds – Image analysis and neural network correlations


A. N. Khan Wardag, F. Larachi*, B.P.A. Grandjean

Department of Chemical Engineering, Laval University, Québec, Québec, G1V 0A6, Canada

T 418-656-3566; F 418-656-5993; Email Faical.Larachi@gch.ulaval.ca


Abstract Digital image analysis was implemented to monitor bubbling dynamics in corrugated-wall bubbling fluidized beds (CWBFB) loaded with Geldart D particles. Various geometrical configurations were investigated in terms of corrugation angle, inter-wall clearance, and rest bed height and gas superficial velocity. Implementation of wall corrugation led to improved gas-solid fluidization quality with respect to flat-wall bubbling fluidized beds (FWBFB) as measured in terms of retreat of the onset of bubbling as a function of gas flow rate, of reduction of bubble sizes and rise velocities, and of increase of bubble frequency. Two artificial neural network correlations valid both for FWBFB and CWBFB were recommended for estimation of bubble frequency and size using a common set of independent variables, i.e., gas superficial-minimum bubbling velocity ratio, bed rest height, corrugation angle, average clearance, and vertical location. The bubble frequency explicit correlation accounted additionally for inter-wall minimum clearance and distance between side wall and either neck or hip of front plate, at a given elevation, whereas bubble size correlation needed bubble frequency as a supplementary input variable.



Submitted to Ind. Eng. Chem. Res, July 2012



You can get the  simanndb-fb.zip  file that contains an Excel worksheet simulator to compute bubble diameter and frequency.

Other links:


- Heat and Mass Transfer in Co-Current Gas-Liquid Packed Beds  Analysis, recommendations and new correlations
-  Predictions of mass transfer for randomly packed towers operation

-  Excel worksheets simulators for  Trickle-bed or Flooded Bed reactors or for predictions of mass transfer in randomly packed towers operations

The neural correlation was developped with the software NNFit