Pehrson, John C (2023-05). Characterization of Femtosecond Laser Electronic Excitation Tagging Measurement Uncertainty for Hypersonic Boundary Layer Profiling. Master's Thesis. Thesis uri icon

abstract

  • Reduced-order computational design tools used to model hypersonic shock wave / boundary layer interactions have been known to provide unreliable estimates of surface heat flux and total pressure, which are key parameters for designing hypersonic engine intakes. Ground test experiments are desired to validate reduced order models. Femtosecond laser electronic excitation tagging (FLEET) was used as a non-intrusive diagnostic tool to measure spatially resolved mean and fluctuating velocity in the boundary layer above a canonical inlet model. Experiments were conducted in the Texas A&M University Actively Controlled Expansion (ACE) tunnel at Mach 6 and a Reynolds number of 6 x10^6 /m. The ACE tunnel operated with a test gas of air at a pressure of 3 Torr, which resulted in FLEET emissions 10x weaker than in pure nitrogen tunnels and rapid signal quenching. Two test campaigns were conducted to mature FLEET boundary layer profiling capabilities. A third test campaign was conducted in the wall-normal orientation, focusing the focused femtosecond laser beam into a 2 mm beam port in the test article to capture velocity 10 mm to just 10's of um above the surface. The rapid emission quenching in the tunnel caused a systematic under-prediction in the mean velocity on the order of 25 m/s (3%), while the low signal in the tunnel caused an over-prediction of velocity fluctuations on the order of 50 m/s (50%). Calibration curves were generated using a model replicating decaying fluorescence and measurement error. The calibrated mean velocity had a relative uncertainty of 2% which was dominated by uncertainty in the image scale and decay time. The calibrated fluctuating velocity had a relative uncertainty of 50% which was dominated by imprecise identification of the displaced emissions. A local maximum in the fluctuating velocity in the boundary layer that was consistent with Direct Numerical Simulations. Mean velocity profiles were found to be repeatable and accurate when compared against RANS simulations in a turbulent boundary layer above a height of y+ ? 5. Below this height, superimposed emissions at low velocity dramatically increased measurement error. Reflected emissions from the beam port also obscured the signal near the surface.

publication date

  • May 2023