Cette thèse porte sur le confort acoustique et, en particulier, sur les matériaux absorbants destinés au traitement des nuisances sonores dans le domaine de l’habitat. L’objectif est double : il s’agit de proposer un matériau bio-sourcé et absorbant dans la gamme des basses fréquences.Notre choix s’est porté sur un matériau poreux bio-sourcé basé sur la fabrication de céramiques poreuses. Il est constitué de cellulose homogène et isotrope. Son procédé de fabrication repose sur la construction d’une empreinte en polyméthacrylate de méthyle permettant un contrôle fin de sa structure interne. Ensuite, nous avons procédé à la caractérisation expérimentale de ses propriétés acoustiques et des paramètres intrinsèques du matériau. L’analyse de ces données, associée à la modélisation numérique, a permis de définir les caractéristiques physiques influant sur ses performances acoustiques et d’identifier les leviers d’amélioration de celles-ci.Enfin, une mise en œuvre de concepts empiriques est présentée, à savoir : (i) inclusion d’aérogel dans la matrice poreuse, (ii) exploitation de la double porosité dans deux configurations différentes, soit dans le cas d’un composite cellulose/aérogel, soit celui d’un matériau constitué d’un double réseau mésoporeux/microporeux et (iii) utilisation de matériaux à gradient de propriétés (interconnexions ou porosité). Pour chacune de ces trois approches, le procédé de fabrication et la caractérisation de ces nouveaux matériaux sont détaillés et leurs performances acoustiques sont discutées. / This thesis is about acoustic comfort and deals particularly with absorptive materials used to treat noise pollution in the housing environment. Our purpose is a double-sided one as the idea is to propose a bio based and absorptive material within the range of low frequencies.We have focused on porous bio based material on the same manufacturing process of porous ceramic. It is composed of homogeneous and isotropic cellulose. The making of a polymethyl methcrylate print enabling a thorough control of its internal structure. Then, we have dealt with the experimental characterization of its acoustic properties and of the intrinsic parameters of the material. The analysis of these data together with the numerical modelling permitted to define the physical characteristics having an influence on acoustic performances and to identify their means of improvement.Finally, the implementation of empirical means will be presented, namely: (i) the inclusion of aerogel in the porous mould, (ii) the exploitation of double porosity in two different configurations, either in the case of cellulose/aerogel composite, or the one of a material composed of a double mesoporous/microporous structure and (iii) the use of materials containing gradients of properties (interconnection or porosity). For each of these three approaches, the manufacturing process and the characterization of these new materials are detailed and their acoustic performances are discussed.

In this thesis, the parallel barrier analysis feature in the Federal Highway Administration Traffic Noise Model (FHWA TNM), which is based on RAYVERB was used to explore the effects of multiple reflections due to single and parallel barriers and the use of absorptive treatment. Database was developed from the data collected from previous research efforts was used to generate a best fit equation model that can be used as a predetermining tool to determine the magnitude of parallel barrier insertion loss. The best fit equation model was then used to test against measured/model result and TNM prediction results for its validity. Absorptive materials were also studied such that 3 top of them were selected and recommended for Florida highway barrier use. It was found that the top three absorptive treatments for use on Florida highway barriers have been determined to be cementitous material, metal wool and glass fiber. These materials can be used to reduce the sound reflections for single and parallel barriers. The developed best fit equation model from this research is Deg = -2.17NRC - CW[superscript 0.42] + 1.97eln(BH) + RH[superscript 0.29] + DBB[superscript 0.27]; the prediction results give moderately high R[superscript 2] value of 0.55 if compared to the results from database. Prediction results from best fit equation model was also found to be consistent with the results from the measure/modeled results, providing further proof of the validity of the model. However, if compared results from equation model, TNM and measured/model (measured and model compared results using ANSI method), TNM was shown to provide higher insertion loss degradation. It was found that the most effective placement of absorptive material was the pattern which covers the barrier from the bottom up; it was also found that only about 60% from the bottom of the barrier area requires covering with high NRC absorptive treatment (NRC greater than or equals to 0.8) without sacrificing insertion loss. Also, if the barrier area near the top includes an easily obtainable NRC value of 0.4, only 40% to 50% of the bottom barrier needs absorptive treatment with a higher, more expensive NRC rating. These findings can substantially reduce the cost of conventional absorptive barrier which have full coverage of high NRC absorptive treatment. This research has begun important improvements in noise barrier design, additional work can be continued to further verify all the findings in this thesis such that easier and better equation model can be developed to calculate insertion loss degradation and cheaper absorptive barrier with less absorptive material usage can be built.