Post-processing of X-ray tomographic images of an acoustic fibrous material and possible use of numerical methods such as FEM and LBM to determine Biot´s porous technical parameters.
Belisario Nina Huallpa1; José Roberto de F. Arruda2; Pierre Lamary3; Franciso Ilson da silva Junior3
1 Engineering Department, Federal University of Lavras (UFLA), MG, Brazil; 2 Department of Computational Mechanics, Faculty of Mechanical Engineering, State University of Campin; 3 Department of Mechanical Engineering, Federal University of Ceará (UFC)
doi:10.20906/CPS/CON-2016-1197
Resumo
We are involved in a project which aims to develop experimental techniques and methods in order to characterise an acoustic absorbing material made of coconut fibres and saturated by air. The material is highly porous and the characterisation aims to determine the technical parameters involved in the homogenized equilibrium equations of Biot and in the semi-phenomenological micro-macro model of Johnson-Champoux-Allard. Tow approaches have been followed, the first one is indirect and is based on the use of acoustic tubes and inverse characterisation techniques, the second one is direct and is based on the analysis of tomographic images of a material sample. It is clear that the two approaches are complementary because any parameter which can be directly determined will ease the indirect identification. The paper will focus on the direct approach and on the post-processing of the tomographic images which is a pre-requisite to any further investigation. The difficulty is that the tomograph produces, slice by slice, two dimensional images that are given in a grey colour scale. The post-processing consists in the analysis of the grey images in order to identify the geometry and the location of the fibres, as to be able to produce black and white images for this bi- phasic material. Several algorithms will be compared from which the Arruda's algorithm build for this purpose. Once this task is achieved, that is to say that the morphology of the material is known, it opens various possibilities. In particular, the porosity of the material which is an important parameter is directly deduced by counting the black and white pixels. Results obtained in our case will be presented. Some algorithms were also proposed to estimate the tortuosity of the material. Another possibility to take benefit from the known morphology for the determination of others parameters such as the air flow resistivity of the material, is to imagine micro-scale numerical simulations. The paper will conclude on this latter point with a pre-study of
Palavras-chave: Acoustics; Characterisation; Porous; Biot; Tomography; LBM