Passive control of pressure oscillations in hypersonic cavity flow

The primary objective of this experimental investigation, which was carried out in a Mach 5 wind tunnel, was to determine if pressure oscillations in hypersonic cavity flow could be suppressed by appropriate changes in the cavity geometry. A major part of the study was to understand better the empty cavity flow structure and the mechanism responsible for the pressure oscillations. In this first phase, fluctuating surface pressures were measured for several baseline cavities for different ratios of the cavity length (L) to depth (H). The ratio of L/H varied from 3 to 6 such that all cavity flows were of the open type. The results showed that: (i) from surface flow visualization all cavity flows were 3-D and symmetric about the longitudinal center line. (ii) three vortices were recognized, namely, the front vortex, the large trailing-edge vortex, and the small rear corner vortex. (iii) Heller and Bliss's model was somewhat better than Rossiter's model for predicting the oscillation frequencies, although neither predicted the values to better than about 10%. (iv) L was the most important factor in determining the oscillation frequencies. (v) the shock induced by the flow impingement on the rear wall (RW) only penetrated a short distance into the cavity as demonstrated by conditional sampling analysis and through calculation of the growth of the instability of the compressible shear layer. (vi) although the evidence is inconclusive, the results suggest that the acoustic wave instead of the shedding vortex was the cause of the flow impingement event on the RW. (vii) the number of consecutive acoustic waves existing simultaneously inside the cavity determined the number of the modes and the oscillation frequencies.; In the second phase, several passive control methods were tested. Changes in geometry were made to the front wall (FW), the RW, and the cavity floor. A 2-D vented wall, a 2-D slotted wall, three 2-D slanted walls, and two 3-D walls were tested to examine their effects on attenuating the strength of cavity flow oscillations. One of the slanted walls and the wall base were also attached to the cavity floor in order to alter the trailing-edge vortex's shape. Three Wheeler doublet vortex generators (VGs) and a full-span wedge were placed upstream of the cavity to act as BL spoilers to change the incoming shear-layer characteristics. The results showed that the vented and slotted walls were ineffective and both kinds of spoilers excited stronger oscillations. One of the 3-D walls, referred to as the "beak" RW, was found to be most effective. This geometry attenuated the strongest oscillations by factors of about 3.5 and 6.8 compared to the baseline cavity with L/H = 3 and 4 respectively. With the wall base or the slanted wall (#3) attached to the cavity floor and with a spoiler upstream of the cavity, the changes in internal flow structure did not alter the mode frequencies. The wall base obstacle effectively attenuated the strongest mode. This attenuation suggested that not only the acoustic wave but also the trailing-edge vortex could affect the pressure oscillations inside the cavity.