Modeling I : a brief history
It is often said that porous media theory started with Biot's series of papers in 1956. Actually, earlier works also brought fundamental insights into the physics of porous material research. In addition, some models, might there be analytical or numerical, may have been unfairly mistaken or misunderstood. This first talk will attempt to draw the global history of the research in the field of acoustical porous media. As an introduction of the seminar, the talk will cover theoretical models, characterization techniques, numerical methods. The aim of this talk is to pave a common background for attendees which are coming with various tuitions and experiences.
Lecture given by L. Jaouen
Modeling II : scaling approaches
Porous media research inherently involves various scales of observation. From the microscale related to the pore size and skeleton morphology, one gains the information about the dissipation mechanisms implied in the material behaviour. From the arrangement of the pores and the distribution of the sizes, global performance of the material may be retrieved. Other intermediate scales may also be considered to account for heterogeneities or porosity networks. This talk will aim at presenting the different techniques which exist to go from one scale to the other. Acoustics, mechanics and flow in porous media will be considered. Keywords for this talk comprise homogenisation, micro-macro approaches and morphological modelling.
Lecture given by C. Boutin
Modeling III : time domain models and outdoor sound propagation
Due to the dissipative nature of porous media, it is convenient to adopt frequency domain models. However, this type of representation poses difficulties when directly translating to a time-domain representation. Time domain modelling has become in recent years a mature numerical technique in acoustics, and is most useful to simulate moving sound sources, in noise control applications involving a broad spectrum like road traffic noise, to include nonlinear effects, etc. The time-domain calculation schemes most often used nowadays in outdoor sound propagation (FDTD, PSTD, TLM) are briefly reviewed. The main focus will be on the finite-difference time-domain technique, and detailed information will be provided concerning numerical accuracy, stability and computational cost. The characteristics largely change when the sound propagation medium is moving, which is often an importance influence in outdoor sound propagation applications due to the presence of wind. The different ways of including porous materials into these models will be discussed, with their pro’s and contra’s. Hybrid modelling between time-domain and frequency-domain models, with the purpose of reducing computational cost, will also be discussed.
Lecture given by T. Van Renterghem
Characterization techniques I : direct approaches
Characterization techniques II : inverse approaches
The relevance and accuracy of the input information related to porous materials is critical. A wide range of techniques have been designed to access the parameters which populate the predicting models. A first category includes direct approaches where a single or a set of parameters are directly measured. Measurement techniques involve techniques among ultrasonics, acoustics, mechanics and fluid dynamics. The second category comprises inverse
techniques which consist in assuming a behaviour model of the porous material and determining the set of parameters which best fit measured data. These two categories are complementary and will be presented and discussed alongside.
Lectures given by R. Panneton
Experience feedback I : material production
Experience feedback II : material assembling
From theory to practice may summarize these two talks. How these tools could serve the material manufacturing while coping with mounting constraints, package assembling, production limitations ? These lectures will present how predicting models and characterization techniques could be used in an industrial context. Each of these talks will be given by the research group leaders of two worldwide leading material manufacturers. Examples will address applications sound packages comprising fibrous materials, felts, foams, granular materials, a wide range in the context of noise control in Buildings, Automotive, Railway, Aircraft and Aerospace, Domestic appliances, Equipment.
Lectures given jointly by S. Berger & A. Duval
Numerics I – FEM : theories
Numerics I – FEM : implementations
With ever increasing computational capabilities, finite element models are now matured enough to be directly involved in the design process of porous materials and sound packages. This talk aims at presenting the basis of the finite element method dedicated to porous media modelling. Classical models together with alternate representations will be discussed. Some basic implementations and practical examples will be proposed to the attendees. Special
attention will be paid to convergence and stability criteria, and numerical efficiency of the implemented models.
Lectures given by P. Göransson
Numerics II – Other approaches
Recently, a number of approaches have been proposed as robust alternatives to purely finite element models. Among these, one could quote the Transfer Matrix Method (TMM), the Wave Based Method (WBM), the Discontinuous Galerkin Methods (DGM) and the Statistical Energy Analysis (SEA), as the most renown. In addition, hybrid models combining finite element models with analytical approaches have proved to be accurate and numerically efficient. This talk will present the principles of several alternates approaches together with practical examples of successful applications of this type of models.
Lectures given by O. Dazel