Chasing eddies and their wall signature in DNS data of turbulent boundary layers

We present an automated and physically rooted method to identify and temporally track hairpin packets and their wall signatures in direct numerical simulation data of turbulent boundary layers. Statistical tools and pattern-recognition algorithms are combined to identify the coherent structures and their signature on the wall, and object segmentation and feature-tracking algorithms are assessed and enhanced to achieve automatic monitoring of the temporal evolution of individual packets and their wall signatures. The visualization algorithms are validated against the statistical analysis. We demonstrate that the average geometric packet is representative of strong statistical ones. Satisfactory results are presented for the canonical case of an isolated hairpin packet convecting in channel flow, and for fully turbulent boundary layers. The method is also suitable for use in combination with experimental particle image velocimetry (PIV) data.