Two-Fluid Simulation of Gas Fluidized Beds

Our previous studies have clearly shown that two-fluid models are able to capture much of the complex system behavior featured by gas fluidized beds. For example, Kuipers et al. (1991) and Nieuwland et al. (1996a) studied bubble forma- 

Measuring Technique Quantity Measured Point/Whole Field Technique Instantaneous/ Time-Averaged Intrusive/ Nonintrusive Type of Flow Application 

HWA Velocity Point Instantaneous Intrusive Single phase G + L 

LDA Velocity Point Instantaneous Nonintrusive Single phase G-l-L Multiphase GL -1- dilute GS 

---

Two-fluid simulations have also been performed to predict void profiles (Kuipers et al, 1992b) and local wall-to-bed heat transfer coefficients in gas fluidized beds (Kuipers et al., 1992c). In Fig. 18 a comparison is shown between experimental (a) and theoretical (b) time-averaged porosity distributions obtained for a 2D air fluidized bed with a central jet (air injection velocity through the orifice 10.0 m/s which corresponds to 40u ). The experimental porosity distributions were obtained with the aid of a nonintrusive light transmission technique where the principles of liquid-solid fluidization and vibrofluidization were employed to perform the necessary calibration. The principal differences between theory and experiment can be attributed to the simplified solids rheology assumed in the hydrodynamic model and to asymmetries present in the experiment. 

Consider a fluidized bed operated at an elevated temperature, e.g. 800°C, and under atmospheric pressure with ah. The scale model is to be operated with air at ambient temperature and pressure. The fluid density and viscosity will be significantly different for these two conditions, e.g. the gas density of the cold bed is 3.5 times the density of the hot bed. In order to maintain a constant ratio of particle-to-fluid density, the density of the solid particles in the cold bed must be 3.5 times that in the hot bed. As long as the solid density is set, the Archimedes number and the Froude number are used to determine the particle diameter and the superficial velocity of the model, respectively. It is important to note at this point that the rale of similarity requires the two beds to be geometrically similar in construction with identical normalized size distributions and sphericity. It is easy to prove that the length scales (Z, D) of the ambient temperature model are much lower than those in the hot bed. Thus, an ambient bed of modest size can simulate a rather large hot bed under atmospheric pressure. 

---