Analysis of Near Boundary Layer Features in Hypersonic Flow

This paper develops a systematic image processing workflow for the identification and categorization of near boundary layer features arising from instabilities within the hypersonic boundary layer of a $5^\circ$ cone. To overcome the challenges of low signal-to-noise ratios inherent in schlieren image capture, the methodology employs steerable filtering with Gaussian derivatives for visual enhancement and noise suppression. A time reconstruction technique revealed a frequency shift from approximately 1000 kHz for features found upstream to 1170 kHz for wave packets found downstream likely due to turbulence capture or mode switching. A Gabor filter bank is utilized to extract orientation-specific spatial frequency information. The analysis successfully tracked the propagation of upstream and downstream-leaning features. Notably, features detected with this methodology exhibited propagation speeds similar to the near-wall boundary layer instabilities suggesting these features may be related. The Gabor filtering technique is also shown to detect precursor activity to the onset of turbulence in one case where traditional Welch's method analysis could not, suggesting that this methodology may be more sensitive in some cases thereby offering a useful alternative tool for linking experimental observations with theoretical predictions.