Addressing Subgrid Stress Modeling Issues in Large Eddy Simulations Using Conditional Sampling of HPIV Data

Department of Mechanical Engineering
The Johns Hopkins University

Abstract-
The seminar introduces a series of experiments aimed at evaluating and developing sub-grid scale (SGS) stress models for Large Eddy Simulation (LES) of turbulent flows. 2-D PIV measurements of locally isotropic and rapidly strained turbulence have been used for studying the correlation between measured and modeled SGS stresses. Recent three dimensional Holographic Particle Image Velocimetry (HPIV) measurements of a turbulent flow in a square duct are being used for evaluating geometric relationships between the filtered vorticity, strain rate and SGS stress tensors. Similar to previous experiments and DNS data, the vorticity has a preferred alignment with the intermediate strain-rate eigendirection. Contrary to eddy viscosity models, the most compressive strain-rate and most extensive SGS stress have a strongly preferred relative angle of 34 degrees. The orientations of the other stress eigendirections are more complex and less deterministic. Comparison to a similarity model shows strong alignment between modeled and measured extensive stresses, weaker but still clear agreements for the most compressive stresses and less deterministic relation for the intermediate stress. Results of conditional sampling show a variety of trends. The seminar will conclude with samples of recent PIV measurements of turbulence in the bottom boundary layer of the coastal ocean.

GALCIT Home Page

Fluids Seminar Page