Computation and control of flow-induced noise behind a circular cylinder using an acoustic analogy approach

The computational aeroacoustics (CAA) research, which focuses on predicting acoustics by means of advanced numerical techniques, has recently gained a great deal of progress. In most applications, the prediction of both the sound source and its far-field propagation is necessary as required by regulations. Recently, powerful computers and reliable algorithms have allowed the prediction of far-field noise through the use of Computational Fluid Dynamics (CFD) data as near-field sound sources. One of the most useful analytical methods, used for the computation of noise, is Lighthill's acoustic analogy. The latter will be used in the present study.; Lighthill's acoustic analogy, combined with the two-dimensional incompressible Navier-Stokes flow computation at low Mach Number (M $\ll$ 1), is used to predict the noise generated by laminar vortex shedding from a circular cylinder at the Reynolds number values Re = 100 and Re = 160. The computed velocity and pressure in the flow field are used as input data for noise source functions. The noise prediction is determined by using Curle's solution of Lighthill's acoustic analogy. Due to the fact that the magnitude of the quadrupole noise source ($O$(M3)) for this type of flow is much smaller than that of the dipole source ($O$(M2)) at low Mach Number, this study concentrates on investigating only the effect of the dipole source on the flow field.; The noise amplitude and frequency obtained by using Curle's solution agree well with published data. For both values of Reynolds numbers Re = 100 and Re = 160, the “lift” dipole source function, caused by the lift force acting on a circular cylinder, is the dominant source term that affects the total acoustic density fluctuation. The objective of this research is to study the suppression of flow-induced noise behind a circular cylinder using a flow control method. The selected method is the electro-magnetic feedback control method developed by Chen and Aubry (2000). The results show that at Re = 100 and Re = 160 the nondimensional acoustic density fluctuation is decreased by five orders of magnitude.