Skip to Main content Skip to Navigation
Theses

Développement d’un modèle numérique de propagation acoustique dans un local délimité par des parois à relief géométrique

Abstract : The study of sound propagation in an enclosed space requires a precise knowledge of the acoustic and geometrical characteristics of its boundaries. Indeed, the geometric relief on the walls of a room causes complex acoustic phenomena that can significantly impact the sound propagation: scattering, diffraction or even resonance if the relief has a particular geometry. The objective of this study is to develop a numerical model of sound propagation in an enclosed space bounded by walls with geometric relief. First, the Adaptive Rectangular Decomposition (ARD) method is used to simulate the propagation in a room. Then, this method is coupled with the Finite Difference in Time Domaine (FDTD) method and the use of digital impedance filters (DIF) to include boundaries with geometric relief and frequency-dependent impedance. The integration of the FDTD method into the ARD method is made possible by the use of perfectly matched layers (PML). The numerical model is validated by comparison with the Kobayashi Potential (KP) and image source methods as well as experimental results. Finally, the model is used to study the sound scattering caused by several experimental relief walls. An apparent sound absorption coefficient is thus estimated to characterize each of these different walls
Complete list of metadatas

Cited literature [108 references]  Display  Hide  Download

https://tel.archives-ouvertes.fr/tel-01760121
Contributor : Abes Star :  Contact
Submitted on : Friday, April 6, 2018 - 9:25:08 AM
Last modification on : Saturday, April 7, 2018 - 1:24:46 AM

File

DDOC_T_2017_0140_RABISSE.pdf
Version validated by the jury (STAR)

Identifiers

  • HAL Id : tel-01760121, version 1

Collections

Citation

Kevin Rabisse. Développement d’un modèle numérique de propagation acoustique dans un local délimité par des parois à relief géométrique. Acoustique [physics.class-ph]. Université de Lorraine, 2017. Français. ⟨NNT : 2017LORR0140⟩. ⟨tel-01760121⟩

Share

Metrics

Record views

110

Files downloads

100