A Experimental Investigation of Unsteadiness in Swept Shock Wave/turbulent Boundary Layer Interactions.

The unsteadiness of various swept shock wave/turbulent boundary layer interactions has been examined by measuring wall pressure fluctuations using multiple, high frequency response, pressure transducers mounted flush with the test surface. The interactions were generated by a variety of test models to study the effect of shock strength and model geometry on the unsteadiness. Tests with nominally 2-D interactions generated by compression ramps were also made for comparison with the results of swept interactions. All tests were made with an incoming boundary layer thickness of 17 mm, at a nominal Mach number of 3, and at a unit Reynolds per meter of 6.7 times 10^{7}.. The unsteadiness is characterized by an intermittent region of relatively low frequency and large amplitude pressure fluctuations, accompanied by a local rms pressure peak at the beginning of the interactions. A large part of the mean and rms pressure distribution is governed by the overall inviscid shock strength rather than by the model geometry. In particular, as in 2-D ramp interactions, the rms peak increases with increasing shock strength. However, the value of the rms peak for swept interactions is only half that of the corresponding 2-D ramp interactions with a similar shock strength. Results further indicate that the level of the rms peak, for either swept or 2-D interactions, can be understood in terms of the motion of the initial pressure gradient and the spatial extent of the intermittent region. These results lead to the observation, for the first time, that local conical scaling law can also be applied to statistical quantities. Results of the two-point correlation and VITA analyses show that the swept and 2-D ramp interactions are inherently different. Conventional correlation and conditional sampling techniques both point to existence of some large scale turbulent structures in the incoming boundary layer. These turbulent structures travel at a significant fraction of freestream velocity and are coherent up to four boundary layer thicknesses in streamwise direction. Results further show these large scale structures have a small effect on unsteadiness, suggesting that other mechanisms are responsible for unsteadiness in interactions.