particles, the gravity and inertial forces are important and the liquid distribution is not very sensitive to the wettability of the packing surface. Closure models used by Yin et al. (2000) to account for the gravity and inertial forces are discussed above. Jiang et al. (2000a, 2000b) also used the Ergun equation to account for these forces. In addition, it is necessary to account for interfacial tension and packing wettability in the physical sub-models to simulate gas-liquid flow through a bed of smaller particles. Based on the experimental data on liquid hold-up distribution with and without particle pre-wetting and an empirical relationship between particle wetting factor and velocity and pressure gradient (of Al-Dahhan and Dudukovik, 1995), Jiang et al. (2000a, 2000b) introduced an additional term representing capillary effect in their two-fluid model. When particles are completely wetted, these additional terms reduce to zero. These additional models are not yet sufficiently validated. It is, however, expected that further research along these lines will allow one to capture the influence of particle wetting on flow distribution within the packed beds.

Apart from these attempts to simulate macroscale flow distribution within the trickle bed, some attempts at understanding microscale flow phenomena using computational flow models have also been carried out. Higler et al. (1999) simulated counter-current flow of gas and liquid through a structured packed bed reactor. They simulated flow of liquid through a sandwich structure. One cross-over of tubes of triangular section was considered in their simulations. The results were useful to understand residence time distribution and the influence of cross-over on this distribution. Simulations of microscale flow phenomena may lead to better understanding of flow regimes of trickle bed reactors (co-current operations) or absorption columns. Casey et al. (1998) reported some of the early results of simulations of film flow over inclined surfaces with a gas flow in the opposite direction. They were able to capture the onset of wavy flow. Much more work is needed in these directions to enhance our understanding of flooding and regime transitions in gas-liquid flows in packed columns. Some of the applications of computational flow modeling to other chemical reactors are briefly reviewed in the following section.

0 0

Post a comment