COMPARISON OF THEORY AND EXPERIMENT ON AEROACOUSTIC LOADS AND DEFLECTIONS

Abstract

The correlation of acoustic pressure loads induced by a turbulent wake on a nearby structural panel is considered: this problem is relevant to the acoustic fatigue of aircraft, rocket and satellite structures. Both the correlation of acoustic pressure loads and the panel deflections, were measured in an 8-m diameter transonic wind tunnel. Using the measured correlation of acoustic pressures, as an input to a finite-element aeroelastic code, the panel response was reproduced. The latter was also satisfactorily reproduced, using again the aeroelastic code, with input given by a theoretical formula for the correlation of acoustic pressures; the derivation of this formula, and the semi-empirical parameters which appear in it, are included in this paper. The comparison of acoustic responses in aeroacoustic wind tunnels (AWT) and progressive wave tubes (PWT) shows that much work needs to be done to bridge that gap; this is important since the PWT is the standard test means, whereas the AWT is more representative of real flight conditions but also more demanding in resources. Since this may be the first instance of successful modelling of acoustic fatigue, it may be appropriate to list briefly the essential “positive” features and associated physical phenomena: (i) a standard aeroelastic structural code can predict acoustic fatigue, provided that the correlation of pressure loads be adequately specified; (ii) the correlation of pressure loads is determined by the interference of acoustic waves, which depends on the exact evaluation of multiple scattering integrals, involving the statistics of random phase shifts; (iii) for the relatively low frequencies (one to a few hundred Hz) of aeroacoustic fatigue, the main cause of random phase effects is scattering by irregular wakes, which are thin on wavelength scale, and appear as partially reflecting rough interfaces. It may also be appropriate to mention some of the “negative” features, to which may be attached illusory importance; (iv) deterministic flow features, even conspicuous or of large scale, such as convection, are not relevant to aeroacoustic fatigue, because they do not produce random phase shifts; (v) local turbulence, of scale much smaller than the wavelength of sound, cannot produce significant random phase shifts, and is also of little consequence to aeroacoustic fatigue; (vi) the precise location of sound sources can become of little consequence, after multiple scattering gives rise to a diffuse sound field; and (vii) there is not much ground for distinction between unsteady flow and sound waves, since at transonic speeds they are both associated with pressures fluctuating in time and space.