Scaling and Transition Effects on Hollow-Cylinder/Flare SBLIs inWind Tunnel Environments

A comprehensive investigation into the flow over a Hollow-Cylinder/Flare has been conducted at Mach 5 with Re_L ≈ 11×105 and a flare deflection θ = 15◦. Experiments of two similar models have been conducted in LT5 at the University of Arizona (Tucson, USA) and R2Ch at ONERA (Meudon, France). Despite similar non-dimensional scaling of the models, a considerable difference in reattachment behavior was observed from Infrared Thermography measurements, indicating that the reattachment in LT5 was located approximately twice as far from the flare base as observed in R2Ch. This discrepancy has driven the investigation in an attempt to identify the cause of this difference. Simulations have been performed at the University of Arizona, ONERA, and the Technical University of Munich (Germany) in support of this study, targeting a range of potential factors that are relevant to the challenge, to quantify the various influences. Amongst the effects reviewed are: differences in the freestream Mach number (M_∞) modulating boundary layer development and the flare-induced inviscid pressure rise, differences in the wall temperature conditions (T_w/T₀) also affecting boundary layer development, 3D relief effects due to different normalized cylinder diameters (D/L), differences in the bluntnesses of the two nominally sharp configurations (r_nose/L), and the impact of freestream disturbances. The noise environment appears to play a significant role in scaling the Shock Boundary Layer Interaction (SBLI) by affecting the transition behavior along the separated shear layer and causing the bubble to grow/shrink to accommodate. Simulations show that the amplitude can be modulated to control the SBLI size, and produce a close match to the experimental results. However, the distribution of noise in the frequency spectra remains unclear. Experimental investigation of the respective noise environment between the two facilities showed that despite each tunnel exhibiting similar noise magnitudes (expressed as p′/p̄_∞) they differed considerably in the range of frequencies (by a factor of 6.5 when considering freestream Strouhal number), suggesting additional parameters are required when quantifying wind tunnel freestream noise conditions beyond its simple amplitude. This study was conducted as part of an international collaborative effort in support of NATO STO AVT-346 Research Task Group.