[Noise and Vibration Research Group] Aeroacoustics Research Group

The research of the Aeroacoustics Research Group focuses on the numerical prediction and experimental validation of aerodynamic noise generation and propagation mechanisms for confined subsonic flow applications. These low Mach-number (M<0.3) applications (mufflers, expansion chambers, confined fans, bends, contractions,...) are commonly encountered in automotive exhauts systems, HVAC systems, turbomachinery,... In order to gain more insight in the aerodynamic noise generating and propagation mechanisms of the devices and to increase the accuracy of the existing numerical prediction techniques the research aims at the continuous development of:

Experimental aeroacoustics of confined subsonic flows
An multipurpose aeroacoustic test set-up is build to study both the sound propagation in duct systems carying a non-uniform mean flow and the aerodynamic noise generation in confined, subsonic flows. Current research is focusing on the experimental identification of noise generating mechanisms in expansion chambers, commonly installed in exhaust ducts.

LES for the determination of aerodynamic noise sources
Large Eddy Simulations (LES) provide a good compromise between accuracy and computational effort. When solved in a compressible way, LES allows to numerically obtain more detailed insight in aerodynamic noise-generating mechanisms for engineering applications.

Quadrature free discontinuous Galerkin for noise propagation in non-uniform flow field
Sound propagation and attenuation in 3D duct systems is simulated in time-domain using a Quadrature-Free Discontinuous Galerkin Method (DGM) for the Linearized Euler Equations (LEE), which allow to model all convective effects introduced by the presence of a non-uniform mean flow.

Time-domain impedance formulation
To model acoustic impedance in time-domain, an efficient recursive formulation is develloped. This formulation can also be applied to broadband simulations and is validated on the NASA Langley Flow Impedance Tube benchmark for characterization of lining material in a square duct, without and with grazing mean flow at several frequencies.

Active two-port representation of confined, subsonic flows
Current research is focusing on increasing the accuracy for the numerical and experimental determination of both the passive two-port characteristics (four-pole parametres), using LEE, and the aerodynamically generated active part, using LES, for confined subsonic flow applications.

Validation of hybrid coupling techniques
Due to the large disparity between acoustic and flow scales, Direct Noise Computation (DNC), where the Navier-Stokes equations are accurately solved in the whole domain of interest is nowadays prohibitively expensive except for a few academic applications. Looking towards a methodology that can be applied to real industrial problems, hybrid methods, where the flow is solved and the acoustic waves are propagated separately, have become very important in the last few years. The source region and the acoustic propagation region can be coupled through equivalent aeroacoustic sources (acoustic analogies) or an acoustic continuation of the source region simulations (acoustic B.C.). Current research is validating and improving the accuracy of the different hybrid couppling strategies and special focus is being paid towards the influence of the source domain results on the accuracy of the final acoustic far-field radiation.

Aerodynamic/acoustic splitting techniques
The results of the source domain simulation contain pressure, velocity and density fluctuations of an acoustic nature and and non-acoustic fluctuations (pseudo-sound). The accuracy of direct and hybrid CAA techniques can be improved if the source domain results are split into an aerodynamic and an acoustic fluctuating part. For this reason, a Low-Mach number aerodynamic/acoustic splitting technique is develloped and validated. This filtering procedure is based on the fact that all vortical fluid movement can be assumed to be of aerodynamic nature while the irrotational acoustic field can be assumed, at low Mach number, to be responsible for all compressibility effects.