DYNAMICS AND DISPERSION IN 3D UNSTEADY EULERIAN-EULERIAN SIMULATIONS OF TWO-PHASE FLOWS

Particle-laden ows are of great interest since they occur in a variety of industrial applications, such as chemical reactors or internal combustion engines in which either solid or liquid particles are injected in a gas ow. These two-phase ows are characterized by a high level of dynamic coupling, heat and mass transfer between the phases. Other complex phenomena of these ows are linked to the size of the particles, depending on the formation process, heterogeneous chemical reactions, coalescence and break-up. Commonly used Lagrangian particle tracking techniques are able to handle most of these complex physical processes. However they are also known to be numerically expensive, as they require a high particle number density to reach a minimum level of accuracy. For unsteady simulations of practical applications of turbulent industrial ows in complex geometries, Lagrangian methods can not be eciently used today. As an alternative a three dimensional unsteady Eulerian-Eulerian approach is proposed to simulate dispersed two-phase ows. In this approach the dispersed phase is treated as the continuous gas phase, through an averaging operator that leads to a system of con-servation equations very similar to the ones for the gas. The main advantage is therefore that the gas phase algorithm may be used on the dispersed phase with the same numer-ical accuracy and computational eciency. However this averaging operation introduces unclosed terms [15] in the equations that need to be modeled. These terms are the non-linear term in the transport operator and other terms related to the unresolved part of the particle velocities. The Eulerian-Eulerian approach has already been successfully applied to Reynolds-Averaged Simulation (RANS) of turbulent ows. The objective of the present work is to extend this approach to unsteady three dimensional simulations of complex ows. Long term objective is to extend this method to an approach comparable to Large Eddy Simulation (LES) as being used in aerodynamics or recently in reactive ows. LES has become