“…It is well known that algebraic turbulence models yield good predictions for attached turbulent flows (see, for example, Kline, Cantwell & Lilley 1981) at low to moderate subsonic speeds, where an extensive base of reliable experimental data is available to validate the computed results. Such models are often used almost without modification in prediction methods for supersonic turbulent boundary layers (see, for example, Talcott & Kumar 1985), the implicit assumption being that a compressible 'law of the wall' exists which is similar to the incompressible form (see, for example, Viegas, Rubesin & Horstman 1985;Carvin, Debieve & Smits 1988). However, the generalization of the 'law of the wall' to compressible flow has been somewhat controversial, and a number of different functional forms have been proposed (see, for example, Van Driest 1951 ; Rotta 1960;White & Christoph 1972;White 1992).…”