Separation length in high-enthalpy shock/boundary-layer interaction

Abstract

Experiments were performed in the T5 Hypervelocity Shock Tunnel to investigate nonequilibrium real-gas effects on separation length using a double-wedge geometry and nitrogen test gas. Local external flow conditions were estimated by computing the inviscid nonequilibrium flow field. A new scaling parameter was developed to approximately account for wall temperature effects on separation length for a laminar nonreacting boundary layer and arbitrary viscosity law. A classification was introduced to divide mechanisms for real-gas effects into those acting internal and external to viscous regions of the flow. Internal mechanisms were further subdivided into those arising upstream and downstream of separation. Analysis based on the ideal dissociating gas model and a scaling law for separation length of a nonreacting boundary layer showed that external mechanisms due to dissociation may decrease separation length at low incidence but depend on the free-stream dissociation at high incidence. A limited numerical study of reacting boundary layers showed that internal mechanisms due to recombination occurring in the boundary layer upstream of separation cause a slight decrease in separation length relative to a nonreacting boundary layer with the same external conditions. Correlations were obtained of experimentally measured separation length using local external flow parameters computed for reacting flow, which scales out external mechanisms but not internal mechanisms. These showed the importance of the new scaling parameter in high-enthalpy flows, a linear relationship between separation length and reattachment pressure ratio, and a Reynolds-number effect for transitional interactions. A significant increase in scaled separation length was observed in the experimental data at high enthalpy. The increase was attributed to an internal mechanism arising from recombination in the free-shear layer downstream of separation, perhaps altering its velocity profile. This real-gas effect depends on the combined presence of free-stream dissociation and a cold wall.