Lance R. Collins1, Juan P. L. C. Salazar1, Jeremy de Jong2, Scott H. Woodward2, and Hui Meng2. (1) Sibley School of Mechanical & Aerospace Engineering, Cornell University, 105 Upson Hall, Ithaca, NY 14853, (2) Department of Mechanical & Aerospace Engineering, State University of New York at Buffalo, 324 Jarvis Hall, Buffalo, NY 14260
Until recently, the phenomenon of inertial particle clustering in turbulence has been investigated exclusively using direct numerical simulations (DNS). Numerous DNS studies have quantified the level of clustering in terms of the radial distribution function (RDF), which increases at small separations with increasing Stokes number, for small values of that parameter. The RDF is an important measure of clustering due to its connection with the enhancement in the inter-particle collision rate. A recent advancement in holographic imaging has enabled us to image a population of particles in a turbulent flow chamber in three dimensions to obtain a quantitative measure of the RDF experimentally. We will present comparisons of the experimental RDF with DNS. Additionally, with two rapidly pulsed images, we have measurements of the relative velocity statistics for particle pairs. We will present comparisons of the relative velocity statistics (mean inward velocity and probability density function) obtained from experiments with equivalent statistics taken from DNS as a function of separation distance. We show there is quantitative agreement between the two, with exception of the tails, where noise in the experiment has produced spuriously large relative velocities. We discuss the origins of those errors and some strategies for reducing them in the future.