Mathematical modelling of particle-laden turbulent flows

点击翻译按钮获取中文摘要

This research is concerned with the development of mathematical models to predict the motion of particulates in a gaseous turbulent flow. This problem is relevant to many engineering applications including the combustion of liquid fuel droplets, pollution dispersion in the atmosphere, and the design of pharmaceutical aerosol drug delivery systems. Computational fluid dynamics (CFD) is used increasingly to aid in the analysis of these flows and is useful for development of improved engineering designs. The complex geometries and typically high Reynolds numbers encountered in such flows necessitate the use of Reynolds Averaged Navier-Stokes (RANS) modelling since the high computational effort associated with large eddy simulation (LES) or direct numerical simulation (DNS) precludes the use of those methods for most engineering applications. It is thus necessary to develop mathematical models that capture the essential physics of the particulate/turbulence interactions, and yet are simple enough to be incorporated into standard CFD codes that**use RANS methodologies. The problem is challenging since numerous physical phenomena such as**turbulence modulation, preferential concentration, turbulence anisotropy and inhomogeneity must be accounted for in the mathematical models. ****The proposed research will focus on using particle-laden DNS for RANS model development. These simulations will provide information on correlation time and length scales (critical for improvement of Lagrangian tracking models) and will also provide insight into the physics of the interaction of the particulates with the coherent structures of the turbulence. LES will also be used to allow for higher Reynolds number flows to be considered. Care will be taken to assess the effect of the subgrid scale modelling on the particle motion. The effects of the particles on the fluid turbulence, through both attenuation and enhancement of turbulent kinetic energy, will also be studied and careful validations performed. ******************