Scattering of turbulent-jet wavepackets by a swept trailing edge

Resumo

© 2015 by The Authors, Airbus.Installed jet noise is studied by means of a simplified configuration comprising a flat plate in the vicinity of a round jet. The effects of jet Mach number, jet-plate radial distance and trailing-edge sweep angle are explored. Acoustic measurements are performed using a traversable 18-microphone azimuthal array, providing pressure data at 360 points on a cylindrical surface surrounding the jet-plate system. Key obsevations include: a decrease, with increasing Mach number, of the relative level of the scattered field in comparison to the uninstalled jet; an exponential dependence of the scattered sound pressure level on the radial jet-plate separation (consistent with a wavepacket source), provided the plate is not in a region of strong mean flow; and, perhaps most interestingly, considerable sideline noise reductions with increasing sweep angle—an evaluation of the integrated sound power radiated by the system shows that this is more than just a directivity modification, the acoustic efficiency of the system decreasing as the sweep angle is increased. The measurements are compared with results obtained using a kinematic wavepacket source model—previously identified from uninstalled measurements—whose radiation is computed in two ways. The tailored Green’s function for a semi-infinite flat plate is used to provide a low-order approximation of the scattering effect; the computation is fast and straightforward, and good agreement with measurements demonstrates that much of the essential source and scattering mechanisms are correctly captured. Use of a more computationally intensive Boundary Element Method provides additional precision, lobes associated with secondary scattering from the other edges and corners being correctly predicted. That such simplified models can reproduce so much of the aeroacoustic behaviour of this high-Reynolds-number, fully turbulent, installed jet, is not only encouraging from the perspective of low-cost prediction strategies, it is first and foremost a demonstration that the models comprise the essential sound generation mechanisms.