TS-98
Flow Induced Acoustic Resonances in Exhaust Gas Economizer Tube Banks
Satoru Okamoto*
ABSTRACT
Staggered arrays of closely packed rigid tubes have been tested, in a wind tunnel of varying the flow velocity_', to investigate the conditions under which acoustical resonances do or do not occur. Among the aspects which have been investigated are: the Strouhal numbers at which flow periodicities occur; the relation between these Strouhal numbers and those at which acoustical resonances occur; the effect of the Reynolds numbers and the longitudinal tube spacing on the occurrence of acoustical resonance. This investigation has shown that: (1) The acoustical resonance can be produced at a frequency well removed from that of vortex shedding. The results also show evidence of both phenomena existing, simultaneously at different frequencies. (2) The acoustical resonances behavior is consistent with that of a self excited system. (3) An alternative model of this phenomenon provides a better procedure for avoiding these resonances in closely packed tube banks.
Key Words: Flow Induced Vibration, Flow Measurements, Vortex, Aerodynamic Acoustics, Noise Suppression, Acoustical Resonance, Tube Arrays
1. INTRODUCTION
Acoustical resonances of ducts containing tube banks are often produced by gas flow across the tubes. The resonances occur when the frequency of a flow periodicity inside the tube bank coincides with that of an acoustical mode. At resonance, an intense pure tone noise, which may reach 150 dB, is produced. This level is sufficiently high to cause vibration and noise problems. In some case, however, acoustical resonances do not occur, although the condition of frequency coincidence is satisfied.
The acoustical resonance, encountered in staggered heat, exchangers, is established within the tube bank in a, direction perpendicular to the flow and the tube axis. The noise produced by this phenomenon is practically a pure tone, with sound pressure levels which could produce potentially damaging to the heat exchanger structure.
Externally applied sound can synchronize vortex shedding and increase the strength and correlation of the shed vortices. Okamoto, et al. [1] examined the effect of an acoustical wave propagating parallel to the cylinder axis on vortex shedding. They found some effect at sound levels above about 120 dB. Externally applied sound of more than 140 dB can entrain vortex shedding from cylinders, shift its frequency up by 8 % and increase spanwise correlation. Ffowcs Williams and Zhao [2] showed that, with feedback from the near wake, sound could either enhance or suppress vortex shedding from a cylinder.
It is interesting to note that it is not the sound pressure that is responsible for modification of vortex shedding, but rather the velocity induced by the sound. Thus, for example, it is easier to entrain shedding from a tube or plate located in the center of a duct, where acoustical velocities are maximum, than at the edge of a duct, where acoustical pressures are maximum but acoustical velocities are minimum. The sound field behaves much like an oscillating velocity field as far as an individual tube is concerned, and oscillating flows are well known to entrain vortex shedding.
It has long been recognized that resonances could be excited in heat exchanger tube bundles. This has become more evident as design trends have changed towards more compact units (smaller tube spacings) with higher flow velocities. Despite this there is still a basic shortage of reliable data and procedures for predicting the occurrence of this phenomenon a situation which appears to owe much to the assumption that vortex induced tube vibration and acoustical resonances are produced by the same mechanism in closely packed banks.
