Collaborative Research: Lagrangian Statistics and Acceleration in Turbulent Shear Flows: Simulation and Modeling

Cornell University, Ithaca NY

Investigators

Abstract

ABSTRACT PROPOSAL NO.: CTS-0328329, CTS-0328314 PROPOSAL TYPE: INVESTIGATOR INITIATED (COLLABORATIVE) PRINCIPAL INVESTIGATORS: STEPHEN B. POPE, PUI-KUEN YEUNG INSTITUTION: CORNELL UNIVERSITY, GEORGIA INST. TECH. LAGRANGIAN STATISTICS AND ACCELERATION IN TURBULENT SHEAR FLOWS: SIMULATION AND MODELING This research is focused on the systematic use of direct numerical simulations (DNS) to study the statistical properties of Lagrangian fluid particle motion in turbulent shear flows, and the development of new stochastic models based on the fluid particle acceleration. Turbulent fluid flows are prevalent in several engineering disciplines as well as in oceanography and atmospheric sciences. A Lagrangian viewpoint is essential in describing several phenomena, such as turbulent dispersion, and also in turbulence modeling using the probability density function method, which has proven particularly successful for turbulent combustion. Despite recent progress in the field there is still a lack of knowledge concerning important effects of anisotropy and inhomogeneity, which are generally present in applications. New and comprehensive Lagrangian statistics covering a substantial Reynolds-number range are to be extracted from DNS of canonical flows. Success of the new simulations will contribute to the resolution of many scaling issues and hence enable the rigorous development of a new class of stochastic models for the acceleration, with tensor coefficients appropriate to anisotropic and inhomogeneous flows. This work is expected to lead to significant advances in turbulence modeling applicable in diverse contexts such as: droplet coalescence in atmospheric clouds; modeling of turbulent flame behavior in the design of improved combustion devices; and fluid-mechanical models of (small) marine organism behavior which has implications for the oceanic food chain and fishery management. At least two jointly supervised PhD students will work on this project using very large-scale scientific computing platforms. This training, together with the knowledge they accumulate in the research process, will prepare them well for future careers in academia or industry. *co-funded jointly by the Fluid Dynamics & Hydraulics, and Combustion & Plasma Systems programs