A study of velocity gradient fields at intermediate and small scales of turbulent shear flows via dual -plane stereo particle image velocimetry.

The present study has developed the first highly-resolved dual-plane stereo particle image velocimetry (DSPIV) technique to measure velocity gradient fields ▿u(x,t) on the quasi-universal intermediate and small scales of turbulent shear flows, and has applied this new diagnostic capability to provide the first direct, nonintrusive, highly-resolved, simultaneous measurements of the structure, statistics, and scaling for the complete nine-component velocity gradient tensor field in turbulent flows. Extensive quantitative assessments are presented that clearly demonstrate the accuracy of the velocity gradient fields obtained with this new measurement technique. The technique has been applied to obtain highly-resolved velocity gradient measurements in the fully-developed self-similar far-field of a turbulent shear flow. The results thereby reflect the effects of the large-scale structure, inhomogeneities and anisotropies inherent in such a flow, corresponding to conditions far exceeding those that are currently addressable by DNS of turbulent shear flows. Velocity gradient results and associated turbulence fields are presented at three different combinations of the local outer-scale Reynolds number Redelta and the local mean strain rate S to determine the extent to which the similarity and scaling implied by classical turbulence theory describe various gradient fields. Conditions correspond to Redelta = 6,000 and 30,000, with Taylor-scale Reynolds numbers Relambda = 44 and 113, and to strain rates S = 0 and 1.2. Results presented with normalization on both local inner- and outer-scale variables show good agreement with the Redelta-scalings from classical turbulence theory for the two S = 0 cases, and for S ≠ 0 show departures consistent with a preferred alignment of the vorticity with the local mean strain rate field. Collectively, the results from this study provide new insights relevant to modeling of the quasi-universal intermediate and small scales of turbulence, which can be used for the development of improved physically-based subgrid-scale models for large-eddy simulation of turbulent flows.PhDAerospace engineeringApplied SciencesUniversity of Michigan, Horace H. Rackham School of Graduate Studieshttp://deepblue.lib.umich.edu/bitstream/2027.42/124141/2/3122005.pd