Search

An experimental study of open cavity flow has been made in a high Reynolds number, Mach 5 turbulent boundary layer. The majority of measurements made were of fluctuating wall pressures. The objectives were: (1) examine how effective changes in front and rear wall geometry were at attenuating the pressure oscillations, (2) explore how impingement of a shock wave (variable strength and position) the cavity flow, and (3) how stores (different geometries and positions) affected the cavity flow. In addition, techniques which were judged effective at attenuating pressure oscillations for the empty cavity were used with shock impingement and with stores in order to explore their effectiveness under perturbed flow conditions. The results show that vented and slotted walls, and spoilers are ineffective. A 3-D rear wall (swept in both planes and symmetric about the center line) attenuated the strongest oscillations by factors of up to 7 compared to the baseline rectangular cavity. Regardless of shock impingement position, shock strength, store position, store dimensions, store to cavity volume ratio and asymmetric store arrangement the cavity oscillation frequencies remain essentially unchanged. Based on the mean and rms Pressure distributions (whose magnitude varies substantially but whose basic shape does not change significantly) and surface flow patterns it appears that the essential flow structure also remains largely unchanged. These similarities suggest that control techniques developed for the empty cavity flow should be effective with shock impingement or store release. Tests using two passive control rear walls in perturbed cavity flow support this conclusion.

The purpose of this work was to (I) examine separation shock wave unsteadiness in different turbulent interactions and determine whether a universal model describing the unsteadiness could be developed, and (II) determine whether or not the observed unsteadiness is a feature of turbulent flow in general, or is specific to the wind tunnel environment. To this end, wall and pitot pressure fluctuation measurements were made in interactions generated by unswept and 25 deg swept compression ramp models, and by 8 deg and 30 deg swept blunt-fin models in a high Reynolds number, Mach 5 turbulent boundary layer. It is clear that the high-frequency, jittery motion of the separation shock is the result of the passage through the wave of individual large-scale turbulent structures. Thus, this component of the unsteadiness is an inherent feature of all turbulent flows. The primary outstanding question concerns the cause of the low-frequency expansion/contraction of the separated flow which is characterized by the large-scale, long-duration excursions of the separation shock wave. Preliminary experimental work to address this question has revealed two very interesting, complementary results. First, there is a distinct correlation between large-scale expansion or contraction of the separated flow and long duration (i.e., low-frequency) falls or rises in pitot pressure in the incoming turbulent boundary layer. Second, results from the same experiment show that the ensemble-averaged pitot pressure at a fixed location in the incoming undisturbed boundary layer correlates with separation shock wave position.