Acoustique

Chaine de l'institut Le Mans Acoustique

http://lemans-acoustique.fr/fr/innovation.html 

 

Sound absorption of a poroelastic cuboid network , by Nicolas Dauchez (Université de Technologie de Compiègne)

Jan. 24, 2017
Duration: 00:15:32
Number of views 179
Number of favorites 0

DENORMS Action’s Workshop “Modelling of high performance acoustic structures Porous media, metamaterials and sonic crystals”, Rome, 24-25th January 2017

Website of DENORMS Action

Programme of the Workshop

Session on New trends in traditional noise treatments – porous materials and microperforated plates

Speaker: Nicolas Dauchez (Université de Technologie de Compiègne)

Abstract:

In order to reduce noise in automotive or aircraft cabins, sound absorbing materials such as polymer foams or fibrous materials are widely used. To be efficient, their thickness must be in relation with the acoustical wavelength, making them not suitable in the low frequency range. To overcome this limitation while reducing weight, multilayer architectures and inclusions involving air gap or impervious structures, may be used. Up to now, most of the proposed models dealing with inclusions are based on rigid frame modeling. The aim of this paper is to investigate both the effect of air inclusions on the sound absorption of a poroelastic material and of oblique incidence. The studied configuration is based on a cuboid foam sample surrounded by air gaps of constant thickness. Measurements, performed in an impedance tube, are compared with finite element predictions (Biot-Allard model). The influence of the incident angle on the absorption coefficient is addressed assuming a periodic architecture. Both experiments and models show that air gaps may increase the absorption in the low frequency range. When the frame is considered as elastic, it is shown that additional absorption peaks are observed: they are linked to bending, shear and longitudinal motion of the porous sample. Moreover, the location and the amplitude of the maximal absorption coefficient may not be accurately predicted with the rigid frame model.

 Infos