Abstract
The complete segregation model of Gibilaro et al. [Chem. Eng. Sci. 41 (1986) 379] has been widely used to predict the layer-inversion phenomenon in a binary solid–liquid fluidized bed. Important features of the model have been analyzed in this paper. Simplifications are proposed to reduce the original two-dimensional optimization problem to mainly a single implicit algebraic equation. A hybrid scheme using a combination of both property-averaging scheme as well as the serial model for evaluating the bed void fraction apparently improves the model predictions. Finally, it has been shown that a straightforward extension of the model from the binary solid to the ternary solid case is generally not tenable.