MODEL EXTENSION OF RANDOM ATMOSPHERIC INHOMOGENEITIES DURING SOUND PROPAGATION FOR ENGINE NOISE AURALIZATION

Abstract

Modelling the impact of atmospheric turbulence is not often encountered in aircraft noise auralization, despite its physical relevance to human-ear perception. This paper presents an extension of a method first developed by Rietdijk et al., for modelling the amplitude and phase fluctuations of spherical sound waves travelling through randomly inhomogeneous media. The main advantage of Rietdijk’s method over other approaches is that ground reflections do not have to be considered simultaneously. That makes it possible to split both effects in auralization tools and investigate them independently. The innovations of the method presented in this paper are as follows: More realistic von Kármán spectra are implemented that also consider the inhomogeneity of atmospheric turbulence with increasing height which makes it possible to model slanted or vertical sound propagation; an overlap-add-method is introduced, which is needed to model a moving sound source when height-dependent changes of turbulence characteristics are considered; abstract physical input parameters describing the meteorological conditions are substituted with input parameters that can be easily obtained from simple measurements. The method is integrated into a global framework dedicated to engine noise simulation, propagation and auralization, developed and validated by the German Aerospace Center (DLR). A good match with measured functions of phase and amplitude fluctuations are observed, as well as realistic reproduction of spectrograms, time signals and the psychoacoustic fluctuation strength of the auralizations with respect to real flyovers.