Cavity resonance suppression using high frequency excitation: The mystery of the cylinder-in-crossflow revisited

Abstract

Previous experiments that have attempted to achieve cavity resonance suppression using high frequency excitation have shown extremely encouraging results. The challenge, however, is to understand the physical mechanism governing this suppression in order to achieve more robust control over a wide operating range. To this end, the primary objective of this paper is to perform spatial linear stability analyses on experimentally measured mean velocity proflles in order to understand the changes to the stability characteristics of the baseline cavity (i.e., the cavity with no control technique) by means of a control device using high frequency excitation. In this paper, high frequency excitation was introduced into the baseline cavity ∞ow by using an appropriately scaled cylindrical rod-in-cross∞ow. The frequency of the excitation was determined by the coherent spanwise shedding from the rod. Additionally, the cylinder-in-cross∞ow conflguration lead to the formation of a wake-type mean velocity proflle downstream of the rod. The stability characteristics of this ‘shedding+wake’ conflguration were compared to those of a conflguration that produced a similar wake-type mean velocity proflle without the introduction of high frequency excitation. This latter conflguration was achieved by modifying the cylinder in a way that it did not shed spanwise coherent vortices, and was referred to as the ‘nonshedding+wake’ conflguration. Unsteady measurements in the shear layer as well as spatial linear stability results revealed that all conflgurations show the presence of large scale structures within the envelope of amplifled instabilities at locations close to the upstream edge of the cavity; the high frequency excitation, however, corresponds to a spatially decaying mode. In the absence of high frequency excitation, the growth rate of all the large scale structures in the baseline and nonshedding+wake are in good agreement with those predicted by linear theory. On the other hand, the shedding+wake conflguration causes a spatial decay of all structures, eventually leading to a featureless spectrum devoid of any dominant scales. Thus, our experimental and analytical results indicate that efiective mechanism for resonance suppression is the rapid spatial decay of all structures; such an efiect is only brought about due to high frequency excitation in the cylinder-in-cross∞ow conflguration. In the absence of such a mechanism, instability modes can be amplifled in the shear layer, and their amplitudes can be reasonably well predicted using spatial linear stability theory.