Large downstream movements of transition observed when the leading edge of a hollow cylinder or a flat plate is slightly blunted are explained in terms of the reduction in Reynolds number at the outer edge of the boundary layer due to the detached shock wave. The magnitude of this reduction is computed for cones and wedges for Mach numbers to 20. Concurrent changes in "outer-edge" Mach number and temperature are found to be in the direction that would increase the stability of the laminar boundary layer.
Experiments carried out in the 12-inch supersonic wind tunnel to investigate the effect of three dimensional roughness elements (spheres) on boundary-layer transition on a 10-degree (apex angle) cone without heat transfer are described. The local Mach number for these tests was 2.71. The data show clearly that the minimum (effective) size of trip required to bring transition to its lowest Reynolds number varies power of the distance from the apex of the cone to the trip. Use of available data at other Mach numbers indicates that the Mach number influence for effective tripping is taken into account by a simple expression. Some remarks concerning the roughness variation for transition on a blunt body are made. Finally, a general criterion is introduced which gives insight to the transition phenomenon and anticipates effects of external and internal disturbances, Mach number transfer.
This paper surveys the available material and summarizes what is known to date about boundary-layer transition at supersonic speeds. Variables studied include Mach number, Reynolds number, pressure gradients, heat transfer, surface roughness, and angle of attack. The discussion is limited to bodies of revolution because similar reliable data for wings is lacking.
Boundary-layer transition by the sublimation and impact-pressure techniques and force tests have been performed on three Haack-Adams bodies of revolution of fineness ratios 7, 10, and 13 at zero angle of attack for free-stream Mach numbers of 2.00, 2.75, and 4.63 and a range of Reynolds numbers based on model length of 6 to 15 X 10(to the 6 power) with and without a roughness strip. The grit method of inducing turbulence was found to provide for a nearly complete turbulent flow over the models at the lower Mach numbers and higher Reynolds numbers considered in this study while the amount of trip drag was less than 8 percent of the model drag with transition fixed. A method of interpreting sublimation data was discussed and used and the results compared well with the impact-pressure results.
