Analysis of Complex Flows From Experimental Data and Numerical Experiments

Fluid mechanics is considered to be a privileged field in physics because phenomena can be made visible. This is unfortunately not the case in turbulence where diffusion and mixing of passive tracers are enhanced by turbulent transport. Consequently, the analysis of the rich flow physics provided by direct numerical simulations (DNS) and by modern optical diagnostic techniques require advanced post-processing tools to extract fine flow details. In this context, this paper reviews most recent techniques used to reveal coherent structures and their dynamics in turbulent flows. In particular, results obtained with standard Eulerian techniques are compared to those obtained from a more recent Lagrangian technique. Even if this latter technique can provide finer details, it is found that the two methods are complementary. This is illustrated with DNS results and with experimental data including planar measurements as well as time-resolved measurements converted to quasi-instantaneous volumetric data by using the Taylor hypothesis.