Williams, Owen J. H. O.W.Pulimeno, Alessandro2025-08-012025-08-012025-08-012025Pulimeno_washington_0250O_28501.pdfhttps://hdl.handle.net/1773/53354Thesis (Master's)--University of Washington, 2025The sensitivity of the flowfield and the resulting loads/moments are examined on a 4% scale, half-span, high-speed version of the Common Research Model (CRM-HS). A range of Reynolds numbers and angles of attack were examined using a combination of tuft flow-visualization and Particle Image Velocimetry (PIV) in the 8 ft by 12 ft Kirsten Wind Tunnel (KWT) at the University of Washington. The tufts showed a leading edge–tip stall at every tested condition, with separation delayed to higher angles of attack at higher Reynolds numbers (Re). Approximate Reynolds number independence was observed for Rec > 10^6, based on the mean aerodynamic chord. The separation point drifts towards the root for larger angles of attack at the same Rec condition. PIV data reveal a thickening of the separated shear layer and the formation of reversed flow with consequent vortical structures in the downstream region at a higher angle of attack. We evaluate and compare background subtraction and denoising methods to mitigate low signal-to-noise ratio (SNR) typical of large scale wind tunnels. A POD-based background removal approach proved more robust than average subtraction or high-pass filters, especially in removing model-induced reflections. Conversely, an entropy-based criterion designed to isolate higher order POD modes representing noise, was found to also attenuate turbulent flow features and so it was unsuitable for this dataset. Proper Orthogonal Decomposition (POD) of the top-down PIV data identified a dominant mode pair (modes 1 and 2) exhibiting large-scale spanwise periodicity. These modes strongly resemble the spanwise alternation seen in stall cells and are phase-shifted to represent a traveling structure in the spanwise direction. The coherent footprint of these modes across the field of view suggests that a low-dimensional mechanism may underlie the separation dynamics. In summary, a highly dynamic reversed flow region and spanwise periodicity are observed which are poorly represented by the mean flow.application/pdfen-USnoneFluid mechanicsAerospace engineeringAeronautics and astronauticsThesis