| The dispersion and temperature distribution of inertial particles are important in many turbulent multiphase flow problems. In order to understand these behaviors better, direct numerical simulations (DNS) of inertial particles suspended in both statistically stationary and decaying isotropic turbulence are performed using an Eulerian-Lagrangian approach. Two major sets of information are reported in this dissertation---the behavior of inertial particles in isotropic turbulence and the subgrid-scale information of local fluid as seen by the particles. While the first information set is useful for fundamental knowledge, the behavior of subgrid scales is important in developing models for inertial particle behavior in Large Eddy Simulations (LES) of turbulent multiphase flows. It is found that, for long times, the dispersion of inertial particles is the greatest when the Stokes number, St eta, = taup/taueta, is of order one, where taup and taueta are, respectively, the particle response time and the flow Kolmogorov time scale. A similar result is found for the mean-square particle temperature fluctuations, &angl0;T'2p&angr0;. To understand the DNS results, an equation for &angl0;T'2p&angr0;, along with the short and long time limits, is derived analytically from the thermal energy equation for inertial particles. For subgrid fluid information, this study shows that subgrid fluid quantities can depend significantly on particle Stokes number and the width of the filter employed. Whereas the dependence of the subgrid fluid velocity and its integral time scale on particle inertia show their extreme values for particles with St eta ∼ O (1), their dependence on the filtering width is monotonic. The monotonic dependence on both particle inertia and filtering width is found for the subgrid fluid temperature fluctuations. |
