Compressible Turbulence

In order to push flight technology beyond the supersonic regime, fundamental research on viscous hypersonic flow will be essential. It is widely recognized that this research must have both a strong experimental component and a major computational effort, and that the experimental and computational work needs to be highly interactive. Carefully chosen, fundamental experiments are needed to provide insight into the physics of transition and turbulence at high Mach number. These experiments may then be used to develop physical models to improve our understanding, help to develop turbulence models for computations, and then validate the computations so that they may be used with confidence as design tools.

Experimental studies of transition at high Mach numbers have focused almost exclusively on the behavior of free shear layers. For wall-bounded flows detailed experiments simply do not exist. The existing studies have been confined largely to establishing correlations for the transition Reynolds number, and virtually no detailed information is available on the effect of Mach number, Reynolds number or heat transfer on the transition process. For turbulent boundary layers in supersonic and hypersonic flow, there exist only three experiments which give turbulence data at a Mach number greater than 4 (Fernholz and Finley 1977, 1981): the hot-wire measurements by Owen et al. (1975) at Mach 7.2, the hot-wire data obtained by Berg (1977) at Mach 6, and the electron-beam data by Bartlett (1981) at Mach 9.

Measurements in a supersonic (M = 3) turbulent boundary layer have shown differences in length scales, the intermittency function, and large-scale structure when compared with subsonic turbulence measurements (Spina & Smits, 1987). Based on Morkovin's (1962) hypothesis, the effects of compressibility do not become important until the fluctuating Mach number, M', becomes large (i.e., M' = 0.3). In boundary layer turbulence, this does not occur until the freestream Mach number reaches approximately 4 (and the formation of shocklets is not seen in boundary layer turbulence until the freestream Mach number of the flow reaches approximately 7) (Spina, Smits & Robinson, 1994).

So, at what freestream Mach number do the effects of compressibility become important? Since the differences between subsonic and supersonic turbulence grow in proportion to the square of the Mach number, a study of a turbulent boundary layer at Mach 8 was deemed to be sufficient to show any characteristic differences between high and low Mach number turbulence. The scarcity of data at high Reynolds numbers in high Mach number flows has prompted us to attempt to obtain data in this realm. We have proposed a series of experiments intended to expand the database. We plan to investigate some basic questions regarding viscous flows at high speed: (1) the effect of Mach number, Reynolds number, and heat transfer on transition, (2) the structure of a fully turbulent hypersonic boundary layer, with and without heat transfer, and (3) the structure of shock wave boundary layer interactions at high Mach number. Another reason for selecting a Mach number of 8 is that at this Mach number the flow is well into the hypersonic regime, yet real gas effects are small and the air still behaves almost as a perfect gas. Some computational difficulties are therefore removed and the turbulence and transition models may be evaluated without the complications introduced by real gas effects.


Flat Plate Photo Click to see larger photo (63k)

Figure 1. Schlieren Photo of Flat Plate Model taken in the Hypersonic Boundary Layer Facility