Detection and Tracking of Coherent Features in a Mach 4 Turbulent Boundary Layer

The detection and tracking of coherent vortical structures in a turbulent compressible boundary layer is considered. Several vortex identification methods are evaluated. It is found that for this flow, the discriminant of the characteristic polynomial of the velocity gradient tensor is the most useful. It tends to highlight vortical structures that are located farther from the wall, and the feature detection is less sensitive to the level than the other methods evaluated. The flow fields are visualized using advanced volume rendering methods that allow large datasets to be analyzed efficiently. This methods have also been adapted to volume fill specific structures so that their evolution can be tracked in time. These approaches promise to make the analysis of large DNS datasets more efficient and quantitative.