Numerical approach for modeling particle transport phenomena in a closed loop of a circulating fluidized bed

Numerical modeling of a large scale circulating fiuidized bed （CFB） imposes many complexities and difficulties. Presence of a dense solid phase, a variety of spatial and time scales as well as complex model geometries requires advanced numerical techniques. Moreover, the appropriate selection of a numerical model capable of solving granular flow, and geometrical model simplification can have a huge impact on the predicted flow field within the CFB boiler. In order to reduce the cost of the numerical simulations, the complex CFB boiler geometry is reduced to that of the combustion chamber. However, a question arises as to bow much one can simplify the geometrical model without losing accuracy of numerical simulations. To accurately predict the gas-solid and solid-solid mixing processes within subsequent sections of the CFB boiler （combustion chamber, solid separator, drain section）, a complete 3D geometrical model should be used. Nevertheless, because of the presence of various spatial and temporal scales within subsequent boiler sections, the complete model of the 3D CFB boiler is practically unrealizable in numerical simulations. To resolve the aforementioned problems, this paper describes a new approach that can be applied for complete boiler modeling. The proposed approach enables complex particle transport and gas flow problems within each of the boiler sections to be accurately resolved, It has been achieved by dividing the CFB boiler geometry into several submodels, where different numerical approaches can be used to resolve gas-solid transport. The interactions between computational domains were taken into account by connecting the inlets/outlets of each section using a set of user-defined functions implemented into the solution procedure. The proposed approach ensures stable and accurate solution within the separated boiler zones.

Lattice Boltzmann method and RANS approach for simulation of turbulent flows and particle transport and deposition

Using the lattice-Boltzmann computational approach in conjunction with the Reynolds averaged Navier-Stokes （RANS） model, several turbulent flows and the transport and deposition of particles in different passages were studied. The new lattice Boltzmann method （LBM） solved the RANS equations coupled with the standard and renormalization group k-E turbulence models. In particular, the LBM formulation was augmented by the addition of two transport equations for the probability distribution function of populations of k and 8. The discrete random walk model was used to generate the instanta- neous turbulence fluctuations. For turbulent channel flows, the analytical fits to the root mean-square velocity fluctuations obtained by the direct numerical simulation of the turbulent flow were used in the analysis. Attention was given to the proper evaluation of the wall normal turbulent velocity fluctuations particularly near the wall. The simulation results were compared with the available numerical simulation and experimental data. The new LBM-RANS model is shown to provide a reasonably accurate description of turbulent flows and particle transport and deposition at modest computational cost.