(458a) Direct Numerical Simulation of Inertial Particles in Isotropic Turbulence At High Reynolds Numbers

Collins, L. R., Cornell University
Ireland, P. J., Cornell University

We use direct numerical simulations (DNS) to study the motion of inertial particles in isotropic turbulence. The primary motivation is to advance our understanding of the role turbulence plays in the evolution of atmospheric clouds.  It is known that clouds develop faster than microphysical models based on gravitational settling alone can predict, and it is our hypothesis that turbulence-driven collisions are responsible for that discrepancy.  An open question is how to extrapolate DNS at Reynolds numbers (based on the Taylor microscale) in the low hundreds to atmospheric conditions, with Reynolds numbers of order 10,000.  While the computational demands of DNS make the atmospheric condition unachievable on today's computers, a recent advance in our simulation code has enabled us to perform DNS on a 2048 cubed lattice and achieve a Reynolds number of about 600, the highest value to date for a multiphase flow simulation.  We will present results from those simulations, with a focus on the statistics that impact the collision kernel, namely the radial distribution function, which quantifies the effect of particle clustering, and the relative velocity probability density function, which controls the number of intersecting trajectories per unit time.  The results reveal interesting trends with Reynolds number that are informing a theory under development in our group, and that will enhance our ability to extrapolate to atmospheric conditions.