(181o) The Study of Heat Transfer In Gas-Solid Flow Using Particle-Resolved DNS of Steady Flow Past Fixed Particle Assemblies

Gas-solid heat transfer is important in fluidized beds that are widely used in chemical processing, energy generation using chemical looping combustion, and CO₂ capture using dry sorbents. Computational Fluid Dynamics (CFD) simulations based on averaged equations are increasingly used for design and scale-up of fluidized beds. Among many other modeling and numerical issues, the accuracy of such CFD simulation strongly depends on the closures used for modeling interphase heat transfer. The existing Nusselt number correlations in gas-solid systems such as fixed beds are based on semi-analytical models and experimental data that differ by several orders of magnitude (Wakao et al. 1982). A Nusselt number correlation developed by Gunn (1978) based on experimental data obtained from many sources is widely used in many CFD cods such as MFIX. Gas-solid heat transfer in steady flow through a homogeneous fixed bed of spherical particles is numerically investigated using Direct Numerical Simulations (DNS) based on Particle-resolved Uncontaminated-fluid Reconcilable Immersed Boundary Method (PUReIBM). Numerical results over a wide range of mean slip Reynolds numbers and volume fractions at Prandtl number of 0.7 for random configurations of fixed assemblies of mono-disperse spherical particles are compared with Gunn's correlation. The DNS data agree reasonably well with Gunn's correlation, with an average difference of 20%. A budget of terms in the mean temperature equation reveals that axial conduction in the fluid phase is 20% of interphase heat flux for low Reynolds number, and its neglect in early models is not justified. Using the DNS data we also develop a model for transport of non-turbulent heat flux which is often neglected in current two-fluid formulates. These improved models for unclosed terms in the averaged temperature equation should result in more accurate CFD prediction of gas-solid flow.