Experimental Investigation of Hypervelocity Shock Wave–Boundary-Layer Interactions on a Deflected Control Surface

Author: Stramenga, Michael John

Year: 2026

Degree: Dissertation (Ph.D.)

Advisor: Austin, Joanna M.

Committee Members: Colonius, Tim; Shepherd, Joseph E.; Hornung, Hans G.; Austin, Joanna M.

Option: Space Engineering

Abstract

Three-dimensional shock wave–boundary-layer interactions generated by a cone-slice-ramp geometry were experimentally investigated in the T5 Free-Piston Reflected Shock Tunnel. Four air test conditions were examined, spanning Re_edge = 2 × 10⁶ to 6 × 10⁶ m⁻¹ at nominal h₀ = 8 MJ kg⁻¹, with one higher enthalpy condition at h₀ = 14 MJ kg⁻¹ and Re_edge = 3 × 10⁶ m⁻¹. Surface heat flux measurements, high-speed schlieren imaging, focused laser differential interferometry, and surface dynamic pressure measurements were used to characterize the mean interaction, surface heating, and separated shear layer unsteadiness.

The test conditions produced both laminar and transitional incoming boundary layers, with separated shear layers that either remained laminar or transitioned within the separation region. Increasing unit Reynolds number at h₀ = 8 MJ kg⁻¹ promoted earlier transition within the separated shear layer, reducing the separation extent and altering both the streamwise and spanwise heating distributions on the ramp. For all test cases, the reattached flow transitioned to turbulence on the ramp surface, and peak streamwise heating was measured in this turbulent region downstream of reattachment. Increasing stagnation enthalpy at constant unit Reynolds number delayed transition onset, causing the separated shear layer to remain laminar and increasing the separation extent.

A separated shear layer instability was observed whose dominant frequency decreased and whose amplitude increased downstream. These disturbances appeared above the separated shear layer as Mach radiation and inside the separation bubble as coherent wave-like density gradient structures. The instability frequencies were not collapsed by either Rossiter-type scaling based on the separated shear layer length or Kelvin-Helmholtz scaling based on the separated shear layer thickness. In contrast, wall-normal acoustic scaling based on the shear layer height above the wall collapsed the dominant frequencies upstream of the ramp, suggesting acoustic communication between the separated shear layer and the model surface. Downstream of the ramp leading edge, this scaling breaks down and the phase relationship between structures inside and above the separation bubble changes, indicating that a different, presently unresolved mechanism influences the downstream shear layer instability.