Combined wave and ray based room acoustic simulations of small rooms : [challenges and limitations on the way to realistic simulation results]

  • Kombinierte wellen- und strahlenbasierte Raumakustik-Simulationen kleiner Räume : Herausforderungen und Grenzen auf dem Weg zu realistischen Simulationsergebnissen

Aretz, Marc; Vorländer, Michael (Thesis advisor)

Berlin : Logos-Verl. (2012)
Dissertation / PhD Thesis

In: Aachener Beiträge zur technischen Akustik 12
Page(s)/Article-Nr.: VIII, 211 S. : Ill., graph. Darst.

Zugl.: Aachen, Techn. Hochsch., Diss., 2012


Classical room acoustic simulation methods based on the principles of geometrical acoustics (GA) have nowadays become an accepted and highly developed tool for acoustic practitioners and researchers in predicting the acoustic characteristics of large rooms like concert halls, theatres or open-space offices. However, when it comes to small rooms, even the most advanced geometrically based methods appear to be flawed due to the inherent negligence of important low frequency wave effects, such as standing waves, diffraction and interference. In order to overcome this limitation the present thesis investigates the potential benefits of the application of the Finite Element Method (FEM) to the modally dominated part of the sound field. However, despite the fact that the FEM is a well-established tool in engineering sciences, which fully captures all relevant wave effects, its application to room acoustics introduces far-reaching and yet unresolved questions regarding the realistic source, boundary and receiver representation. The present thesis therefore establishes a complete framework for the combination of FE and GA-based room acoustic simulation results and discusses the inherent challenges and limitations including all aspects of sound generation, sound reflection and sound reception. Moreover, the thesis establishes detailed guidelines for the best-possible determination of all necessary input data for both simulation domains. The investigations conducted in the course of this thesis aim at two different goals. On the one hand the thesis investigates the influence of selected isolated aspects regarding their influence on the simulation accuracy. In particular, these topics include a study on the potential of the image source method to predict the modal characteristics of the low frequency Room Transfer Function (RTF), a study on the efficient modelling of porous absorbers in the FE domain and finally a study on the possible low frequency coupling of the excitation velocity of a loudspeaker to the sound field at the loudspeaker membrane. On the other hand the thesis investigates the overall potential of the presented combined approach by conducting extensive objective and subjective comparisons of measurement and simulation results for three types of acoustically relevant small spaces (a scale-model reverberation room, a recording studio and two different car passenger compartments). For each room considerable efforts have been made to obtain a best-possible a-priori assessment of all necessary material and source data for the simulations. However, especially with regard to the determination of the acoustic surface impedances at the room boundaries certain inevitable inaccuracies have to be accepted. While the presented results reveal an overall good agreement regarding the energy distribution in time and frequency domain for all considered rooms, the results clearly show that as expected the simulation accuracy considerably degrades with increasing complexity of the room geometry and boundary conditions. Moreover, it is important to mention that even with the FEM a precise prediction of the fine structure of the RTF appears impossible in the frequency range far above the Schroeder frequency. It can thus be concluded, that possible fields of application of the FE extension in room acoustic simulations lie in the prediction of the modally dominated low frequency part of the RTF of well defined rooms and in the prediction of sound fields that are strongly affected by near-field or diffraction effects as in the car passenger compartment. Thus, despite the general potential of the low frequency FE extension to realistically predict the modal structure of the low frequency part of the RTF, the Achilles’ heel of room acoustic FE simulations appears to be the determination of realistic impedance conditions on the room boundaries. Consequently the application of the finite element method to room acoustic applications calls for improved measurement techniques for the acoustic surface impedance.