Laminar-turbulent transition is important in the design of hypersonic flight vehicles, as a turbulent boundary layer causes local increases in heat transfer and skin friction. The location of transition is influenced by many factors and is difficult to predict. 1 Transition is caused by disturbances within the freestream or on the body, which enter the boundary layer via receptivity mechanisms, grow through various instability modes, and cause turbulence. 2 Among the many factors that can influence hypersonic transition is surface roughness, which can cause transition to occur prior to the natural transition location. 3 From a design standpoint, it may be necessary to determine if a typical, as-built surface roughness will cause transition during reentry, or to determine the surface roughness magnitudes and distributions that will begin to affect the natural transition location. Isolated roughness elements can also be used to intentionally cause transition within a laminar boundary layer, for instance to prevent scramjet engines from unstarting. 4, 5 The physics of roughness-induced transition are poorly understood, and experimental measurements of roughness-induced instabilities at supersonic and hypersonic speeds are limited. An isolated roughness often generates a wake with streamwise vortices originating from within the separation region upstream of the roughness. Streaks from vortices were observed in oil-flow visualization by Whitehead in 1969. 6 Flow structures oriented in the streamwise direction have also been observed in recent high-speed experiments. 5, 7, 8 For large roughnesses (k/δ≈ 1.0 or greater), the dominant instability type is likely to be an “absolute” instability. Absolute instabilities grow in time from an initial disturbance and are oscillatory in nature. 9 Examples of absolute instabilities include periodic vortex shedding, as observed in low-speed roughness experiments, 10 or some other instability caused by the large disturbance of the flow by the roughness. For smaller roughnesses (as the roughness height tends towards zero), it is likely that “convective” instabilities within the shear layers generated by the roughness become the dominant mode of transition. Convective instabilities are characterized by spatial growth as they are swept downstream and are subject to the effects of external disturbances. 11 Other possible transition mechanisms include that of transient growth, 12 interaction of streamwise vorticity from the roughness with the Görtler instability or stationary crossflow, 3 or modification of the receptivity of the boundary layer. 13 Correlations for transition due to roughness are used out of necessity with some success for many flight vehicles. These correlations are based on parameters such as Rek, the Reynolds number at the roughness height k based on conditions in the undisturbed laminar boundary layer at the height k. 14 If not used carefully, correlations can lead to larger-than-desired uncertainty in the predicted transition location. 15 Measurements taken in conventional ground facilities may be subject to tunnel noise effects, which can significantly alter the roughness-induced transition location. 16 Recent NASA flight experiments measured transition due to a controlled roughness on the heat shield of the Space Shuttle. 17, 18 However, controlled measurements of transition in flight are often not available due to the high cost and difficulty inherent in acquiring them. The fact that empirical correlations are used despite sometimes large uncertainty highlights the need for better roughness-induced transition-prediction methods. Ideally, such prediction methods would be based on the growth of disturbances via instabilities within the …