L'objectif de cette thèse est l'étude du bruit propre d'origine hydrodynamique au niveau de l'antenne d'étrave d'un sous-marin pour des fréquences comprises entre 10 Hz et quelques kHz. Dans cette optique, trois problématiques particulières doivent être étudiées : la modélisation du comportement vibro-acoustique des structures en fluide lourd, l'intégration de l'excitation couche limite turbulente (CLT) dans les modèles numériques déterministes et la mesure de la réponse d'une structure sous écoulement turbulent. Le premier chapitre est consacré à la synthèse bibliographique des différentes méthodes existantes pour étudier de tels problèmes. Dans un second chapitre, nous présentons la méthode PTF (Patch Transfer Functions), qui est une approche par sous-structuration de domaines permettant l'étude des problèmes vibro-acoustiques en basses et moyennes fréquences. Cette méthode n'ayant été appliquée qu'en fluide léger, l'un des enjeux de ce chapitre est d'étudier son applicabilité en fluide lourd et de mettre en lumière d'éventuels points de blocage. Nous montrons ainsi que des problèmes de convergence de la méthode apparaissent en fluide lourd. Pour pallier ces difficultés, nous présentons, dans le troisième chapitre, deux stratégies de calcul permettant d'accélérer la convergence globale de la méthode. Le quatrième chapitre propose une application de la méthode PTF pour caractériser la réponse vibro-acoustique en fluide lourd d'une structure soumise à une CLT et couplée à un milieu fluide extérieur et à une cavité acoustique, encombrée par des structures déformables. Pour traiter ce problème, nous complétons, tout d'abord, le modèle PTF par l'introduction de l'excitation CLT dans les modèles déterministes, à partir d'une décomposition de l'interspectre des fluctuations de pression pariétale sur une base d'ondes planes décorrélées. Nous présentons, ensuite, le calcul de l'impédance de rayonnement par pavé pour des surfaces de couplage de géométrie quelconque, ainsi qu'un modèle d'encombrement permettant, à partir de considérations physiques, de tenir compte simplement de la complexité structurale de structures encombrant une cavité acoustique. Cette application nous permet de montrer que l'influence de l'encombrement des cavités sur la réponse vibro-acoustique du système ne peut pas être négligée, lorsque l'on s'intéresse à l'étude du bruit hydrodynamique au niveau d'une antenne d'étrave de sous-marin. Dans le cinquième chapitre, nous développons une technique expérimentale permettant de caractériser le comportement des structures sous écoulement turbulent. Cette technique est basée sur la décomposition en ondes planes décorrélées présentée au chapitre 4 et nécessite la génération de champs de pression pariétale correspondant à ceux que l'on obtiendrait à partir d'ondes propagatives et évanescentes pour tenir compte des caractéristiques spectrales de la CLT. Pour cela, nous utilisons une antenne de monopôles acoustiques, dont nous déterminons numériquement les principaux paramètres (taille, nombre de monopôles et distance antenne paroi). Le nombre de monopôles augmentant avec la fréquence, la mise en oeuvre de cette méthode peut s'avérer délicate. Pour contourner cette difficulté, nous présentons le principe de l'antenne synthétique, dont la validation expérimentale est étudiée dans le dernier chapitre.

A vibroacustografia (VA) é uma técnica de imagem de alta resolução lateral (<0,7 mm), livre de speckles, não invasiva e que utiliza radiação não-ionizante, baseada no fenômeno de combinação não-linear de ondas de ultrassom (MHz) que gera uma região focal estreita de baixa frequência (kHz). Neste trabalho, propõe-se uma avaliação tridimensional baseada em imagens de VA da área descoberta de implantes, após o procedimento cirúrgico de artroplastia total de quadril (ATQ). A ATQ é uma técnica terapêutica de substituição de articulações do quadril por implantes, em pacientes com osteoartrose avançada. Atualmente, o acompanhamento pós-operatório é auxiliado pela análise de radiografia plana por raios-X. Contudo, esta técnica apresenta limitações na avaliação da estabilidade dos implantes, tais como: avaliação imprecisa da área, uma vez que, a radiografia plana trata o problema tridimensional como bidimensional; impossibilidade de visualização de algumas regiões do implante, devido ao sobreposicionamento das regiões numa imagem planar; restrição do uso de radiação ionizante na região da pelve em pacientes jovens. Devido à alta atenuação da onda de ultrassom incidindo sobre objetos densos, a VA pode ser ajustada para adquirir imagens relativas, principalmente, à superfície destes. Uma correção da topologia baseada na geometria do implante é apresentada, a fim de se avaliar com precisão a área superficial na representação tridimensional da imagem de VA. A correção da topologia baseada em aquisições em modo-B também é apresentada, sendo útil na avaliação da forma de ossos ou nos casos em que a forma do objeto estudado é desconhecida. Além disso, é apresentado um estudo numérico da utilização de transdutores matriciais reconfiguráveis (RCA, do inglês reconfigurable array) para a formação de feixe em VA. Um estudo paramétrico da seleção de abertura, número de canais, número de elementos, distância focal e parâmetros de varredura eletrônica do feixe é apresentado para demonstrar a viabilidade e avaliar o desempenho da imagem de VA utilizando o RCA. A função de espalhamento de ponto do sistema foi calculada com base em métodos de espectro angular usando a aproximação de Fresnel para fontes retangulares. Foram discutidas as vantagens de transdutores bidimensionais e RCA, e aspectos relacionados à importância clínica da implementação do RCA na VA, tais como resolução espacial, taxa de aquisição da imagem e implementação do RCA em máquinas comerciais. Conclui-se que: a aquisição de imagens por VA foi menos sensível ao ângulo de incidência das ondas de ultrassom do que por modo-B e apresentou alta resolução e alto contraste entre o osso e o implante, o que permitiu uma avaliação precisa da área exposta do implante; os transdutores RCA podem aprimorar a aquisição de imagens por VA, devido à sua resolução espacial similar à de transdutores confocais e à possibilidade de deflexão eletrônica do feixe (em inglês steering) nos planos de elevação e azimutal. / Vibro-acoustography (VA) is a high lateral resolution (<0.7 mm), speckle-free, non- ionizing and non-invasive image based on the nonlinear combination phenomenon of ultrasound waves (MHz) that generates a narrow focus region of low frequency (kHz). In this work, we propose a VA-based tridimensional evaluation of the exposed area of implants, after total hip arthroplasty (THA). THA is a therapeutic surgery procedure in which the hip joints are replaced by metal prostheses in patients with advanced osteoarthritis. Currently, the postoperative assessment of THA is performed by analyzing planar X-ray radiographies. However, this technique presents limitations in the implant stability evaluation such as: inaccurate evaluation of the area, since the planar radiography reduces the tridimensional problem into a two-dimensional one; problems in the visualization of some regions of the implants due to the superimposition of the regions; limitations of using ionizing radiation in the pelvis region of young patients. Due to the high attenuation of the ultrasound wave into objects of high density, the VA image can be set to mainly acquire the surface information. Topological correction based on the geometry of the implant is presented in order to accurately evaluate the surface area in tridimensional representation of the VA image. The topological correction based on B-mode slices is also presented and is useful in the evaluation of bone shape or in cases which the shape of the object is unknown. Furthermore, it is presented a numerical study of the use of reconfigurable arrays (RCA) for VA beam formation. A parametric study of the aperture selection, number of channels, number of elements, focal distance and steering parameters is presented in order to show the feasibility and evaluate the performance of VA imaging based on RCA. The point-spread function of the system is calculated based on angular spectrum methods using the Fresnel approximation for rectangular sources. Advantages of two-dimensional and RCA arrays and aspects related to clinical importance of the RCA implementation in VA such as spatial resolution, image frame-rate, and commercial machine implementation are discussed. It is concluded that: the VA image is less sensitive to the wave incident angle than B-mode; the VA image presents high-resolution and high-contrast between bone and implant, which allowed an accurate evaluation of the uncovered area; RCA transducers improve the VA acquisition by producing spatial resolution similar to confocal transducers, and due to the possibility of steering in elevation and azimuthal planes.

Orientador: Jose Roberto de França Arruda / Dissertação (mestrado) - Universidade Estadual de Campinas, Faculdade de Engenharia Mecanica / Made available in DSpace on 2018-08-06T11:13:15Z (GMT). No. of bitstreams: 1 Padilha_PauloEduardoFranca_M.pdf: 721382 bytes, checksum: 8fae387ba335cdb5b3e7b462935121c1 (MD5) Previous issue date: 2006 / Resumo: Neste trabalho é feita uma comparação de técnicas analíticas e experimentais de Análise de Caminhos de Transferência Vibroacústicos. A Análise de Caminhos de Transferência, ou TPA como é conhecida pela abreviatura em inglês, é uma ferramenta para se avaliar a contribuição de diferentes caminhos de propagação de energia vibroacústica entre uma fonte e um receptor ligados entre si por várias conexões. Uma aplicação típica de TPA é quantificar e determinar a importância relativa desses caminhos numa dada banda de freqüência, verificando-se a contribuição mais significativa para o receptor. Nessa análise, basicamente duas quantidades devem ser obtidas: as forças operacionais em cada caminho de transferência e a Função de Resposta em Freqüência desses caminhos. Essas funções podem ser obtidas tanto experimentalmente quanto analiticamente, considerando-se ou não a impedância mecânica da fonte vibroacústica. As forças operacionais podem ser diretamente obtidas de medições, usando um transdutor de forças, ou indiretamente estimadas a partir de medições auxiliares. Neste trabalho são apresentados dois métodos de obtenção indireta das forças operacionais ¿ o Método da Rigidez Complexa e o Método da Inversão de Matrizes ¿ associados com as duas possíveis configurações para determinação das Funções de Resposta em Freqüência ¿ incluindo e excluindo a impedância da fonte. Essas técnicas são aplicadas a dois modelos: um analítico e um experimental. avaliadas com esse modelo analítico, comentando-se suas vantagens e desvantagens. São avaliados ainda os efeitos do acoplamento entre fonte e receptor e entre os caminhos de transferência. Os experimentos foram realizados numa estrutura concebida como um modelo simplificado de um automóvel. As forças foram obtidas através dessas duas técnicas indiretas e comparadas com as medidas com transdutores de força. Como no modelo analítico, as forças e FRF foram estimadas na presença e na ausência da fonte, e a contribuição de cada caminho calculada através dessas técnicas e configurações. A resposta do receptor, calculada como uma combinação dos diferentes caminhos de transferência, foi comparada à medida nas condições operacionais, de modo a se verificar a exatidão de cada método utilizado. As vantagens e limitações desses métodos e configurações experimentais são então comentadas para uso em aplicações práticas / Abstract: In the present work, analytical and experimental techniques of Vibro-Acoustic Transfer Path Analysis are compared and discussed. Transfer Path Analysis (TPA) is a tool to evaluate the contribution of different vibro-acoustic energy propagation paths between a source and a receiver linked to each other by a number of connections. TPA is typically used to quantify and rank the relative importance of these paths in a given frequency band, determining the most significant one to the receiver. When applying TPA two basic quantities have to be determined: the operational forces at each transfer path and the Frequency Response Functions of these paths. These functions can be obtained either experimentally or analytically, considering or not the mechanical impedance of the source. Operational forces can be directly measured by force transducers or indirectly estimated from auxiliary response measurements. Two methods to obtain the operational forces indirectly ¿ the Complex Stiffness Method and the Matrix Inversion Method ¿ associated with two possible configurations to determine the Frequency Response Functions ¿ including and excluding the source impedance ¿ are presented and discussed in this work. These estimation techniques are applied to two models: an analytical and an experimental model. Analytically, these techniques are applied to a lumped parameter model. The advantages and drawbacks of them are commented and the effect of coupling between source and receiver and among transfer paths is analyzed with this theoretical approach. The experiments are done in a test structure conceived as a simplified model of an automotive vehicle. The forces are obtained via these two techniques and compared to those measured by force transducers. Like in the analytical model, forces and Frequency Response Functions are estimated in the presence and in the absence of the source, and the contribution of each path is calculated based on the combinations of these different estimation methods and setups. The contributions of the different paths at the receiver location are calculated using each estimation method and are compared to those obtained in the operational condition in order to verify the accuracy of these methods. The advantages and limitations of the methods and test setups are then discussed for use in practical applications / Mestrado / Mecanica Computacional / Mestre em Engenharia Mecânica

Dans le domaine de l’acoustique du bâtiment, la structuredes constructions en bois est légère et ses capacités d’isolementacoustique sont faibles en particulier pour les basses fréquencesen comparaison à d’autres techniques de constructioncomme le béton. Les travaux de cette thèse se donc sont centréssur l’élaboration d’un outil prédictif visant à déterminer lesperformances acoustiques des planchers solivés en bois afin decompléter les outils prédictifs actuellement sur le marché. Laméthode des éléments finis est adaptée pour résoudre ce problèmevibro-acoustique car le comportement basses fréquencesest de type modal. Pour gagner en temps de calcul, une résolutionmixte éléments finis – formulation intégrale d’Huygens etdifférentes optimisations ont été mises en place. Grâce à l’utilisationde la méthode des éléments finis, il a été possible demettre en évidence des résonances de ce problème couplé quifont chuter localement la valeur de l’affaiblissement ; celles-cine peuvent être prises en compte par la méthode des matricesde transfert. Néanmoins, ces deux méthodes fournissent des alluresmoyennes de courbes d’affaiblissement équivalentes, enparticulier lors d’une étude en tiers d’octave. Pour aborder leproblème des transmissions latérales, nous avons montré lafaisabilité de la détermination expérimentale des propriétés mécaniqueséquivalentes d’un élément de jonction courant entreles planchers et les murs : le sabot métallique. Un modèle mécaniquesimplifié par éléments finis permet ainsi un couplagemécanique entre les solives et les poutres en intégrant les effetslocaux de dissipation présents pour ce type de jonction. / In building acoustics, the structure of wooden construction islightweight and sound insulation at low frequency are weakcompared to other construction techniques like concrete. Thework of this thesis is focus on the development of predictivetools for determining the acoustic performance of wooden joistfloors to complete the existing predictive tools available on themarket. The finite element method (FEM) is adapted to solvethis vibro-acoustic problem because behavior at low frequenciesis modal. To save computation time, a mixed resolutionfinite element with Huygens integral formulations and severaloptimizations were implemented. We could note the interest ofthe FEM : we observed that resonances of the coupled systemdrive locally down the value of the TL, which can not be takeninto account by the method of transfer matrices. However, wenote that these two methods provide equivalent average curveshapes of TL’s spectrum. To deal with the problem of flankingtransmissions, penalizing the lightweight construction, we experimentalydetermine the mechanical properties of a equivalentjunction element between floors and walls : the wood joisthanger. Through a simplified mechanical model, this numericalfinite element model allows a mechanical coupling between thejoists and beams by incorporating the local effects of dissipationpresent for this type of junction.

A cold extrusion process has been developed to tailor a porous structure from polymeric waste. The use of an extruder to manufacture acoustic materials from recycled waste is a novel idea and the author is not aware of any similar attempts. The extruder conveys and mixes the particulates with a reacting binder. The end result is the continuous production of bound particulates through which a controlled amount of carbon dioxide gas that is evolved during the reaction is used to give the desired acoustic properties. The cold extrusion process is a low energy consuming process that reprocesses the post manufacturing waste into new vibro-acoustic products that can be used to meet the growing public expectations for a quieter environment. The acoustical properties of the developed products are modelled using Pade approximation and Johnson-Champoux-Allard models. Applications for the developed products are widespread and include acoustic underlay, insulation and panels in buildings, noise barriers for motorways and railway tracks, acoustic insulation in commercial appliances and transport vehicles.

La thèse s’inscrit dans le cadre industriel de la détection de la réaction entre l’eau et le sodium dans un générateur de vapeur d’un réacteur à neutrons rapides à caloporteur sodium. Plus précisément, il s’agit de détecter une micro-fuite d’eau dans le sodium (débit < 1 g/s) en moins de 10 secondes à partir des signaux vibratoires mesurés sur la virole externe du générateur de vapeur. Le bruit de fond très important de l’installation en fonctionnement risque de mettre en défaut la détection basée sur un dépassement de seuil de l’énergie vibratoire. Nous nous intéressons dans cette thèse à la technique de la formation de voies qui permet d’augmenter le rapport signal à bruit et de détecter et de localiser une source acoustique à l’intérieur du système à partir de signaux vibratoires mesurés par une antenne de capteurs accéléromètres placés sur la coque cylindrique (i.e. virole du générateur de vapeur). Une étude numérique a d’abord été réalisée. Des modèles numériques sont développés afin de simuler les signaux mesurés (virtuellement) par les accéléromètres de l’antenne linéaire considérée. Les performances de la formation de voies sont alors étudiées suivant différents paramètres (i.e. position source, amortissement, fréquence, type de bruit de fond, etc.). Le premier modèle consiste en une plaque plane infinie en contact avec un fluide lourd et excitée par un monopole acoustique situé dans ce fluide. L’analyse des déplacements transverses de la plaque nous permet d’établir un critère sur l’échantillonnage spatial du champ vibratoire par l’antenne. Un second modèle, plus représentatif de la virole du générateur de vapeur, est ensuite proposé. Une coque cylindrique élastique infinie, remplie par un fluide lourd est considérée. Le caractère « fini » de ce système dans les directions radiales et circonférentielles fait apparaitre un comportement modal. Son impact sur la formation de voies est étudié. La méthode est finalement testée expérimentalement. Des mesures sont réalisées sur une maquette composée d’un cylindre en acier rempli d’eau, placé dans un circuit hydraulique dont on maitrise le débit. La source acoustique est générée à partir d’un hydrophone. On étudie alors le gain d’antenne pour différents débits et niveaux d’excitation de la source. / This thesis deals with a new method to detect a sodium-water reaction in a steam generator of a fast sodium-cooled nuclear reactor. More precisely, the objective is to detect a micro-leak of water (flow < 1 g/s) in less than 10 seconds by measuring the external shell vibrations of the component. The strong background noise in operation makes impossible the use of a detection system based on a threshold overrun. A beamforming method applied to vibrations measured by a linear array of accelerometers is developed in this thesis to increase the signal-to-noise ratio and to detect and locate the leak in the steam generator. A numerical study is first realized. Two models are developed in order to simulate the signals measured by the accelerometers of the array. The performances of the beamforming are then studied in function of several parameters, such as the source location and frequency, the damping factor, the background noise considered. The first model consists in an infinite plate in contact with a heavy fluid, excited by an acoustic monopole located in this fluid. Analyzing the transverse displacements in the wavenumber domain is useful to establish a criterion to sample correctly the vibration field of the plate. A second model, more representative of the system is also proposed. In this model, an elastic infinite cylindrical shell, filled with a heavy fluid is considered. The finite dimensions in the radial and circumferential directions lead to a modal behavior of the system which impacts the beamforming. Finally, the method is tested on an experimental mock-up which consists in a cylindrical pipe made in stainless steel and filled with water connected to hydraulic circuit. The water flow speed can be controlled by varying the speed of the pump. The acoustic source is generated by a hydrophone. The performances of the beamforming are studied for different water flow speeds and different amplitude and frequencies of the source.

This thesis is part of a major research project to analyse vibro-acoustic characteristics from variable speed inverter driven induction motors (VSIDIM). The overall projects??? aimed at providing a better understanding of the mechanisms of sound generation from electromagnetic origins and developing a numerical model to predict the sound power emitted from a VSIDIM. The scope of this thesis is to assess experimentally the effect of various controller strategies on the radiated sound power and to develop a finite element method for calculating the electromagnetic force distribution over the stator. Various sources of noise in induction motors and their behaviour with speed and load have been reviewed. Models of the electromagnetic field and vibro-acoustic character have been discussed. An outline of various techniques of reducing noise in induction motors through design of inverters and modifications to the motor structure has been given. Experiments were conducted to assess the effect of controller strategies on the radiated sound power. Three different supplies were tested: a dynamotor which produces an almost sinusoidal supply with very low harmonic content, an inverter with a low switching frequency (less than 1kHz) and an inverter with a high switching frequency (8kHz) and various levels of random modulation. Results indicate that the sound power level of the MSC drive is a lot higher than that of the VSC 2000 drive and the dynamotor drive. The sound power level of the VSC 2000 drive and the dynamotor drive increases almost linearly with motor speed, that for the MSC drive is almost independent of speed. The sound power level of the MSC drive is almost 28dB higher than that of the dynamotor drive at 450rpm and the difference is reduced to 14dB at 1500rpm where the aerodynamic noise becomes more dominant. It has been found that at the rated speed (1500rpm), the sound power level varies by less than 3dB from no load to full load for all three sources. Although increasing the switching frequency increases the cost of the inverters and switching losses, results from the MSC and VSC 2000 drives clearly show that it reduces the radiated sound power by shifting the harmonics into higher and inaudible frequency range. The tonal nature around the switching frequency has been reduced by increasing the levels of random modulation to spread the energy over a wider range of frequencies, although the sound power level has not varied by more than 0.2dB. A finite element model has been developed to calculate the electromagnetic force distribution. The quasi-static solution method has been implemented by stepping the rotor through the time domain using a fine regular mesh in the air gap. The stator currents were experimentally obtained while the rotor currents were obtained using a 4 parameter state space model of the motor. Results of the simulation indicate the influence of stator and rotor slots, saturation and time harmonics in the current. The calculated electromagnetic force distribution has been used in a FEM/BEM acoustic model and SEA acoustic model to predict the radiated sound power which agrees reasonably well with the measured sound, thus validating indirectly the electromagnetic force simulations.

With the advent of power electronics, inverter-driven induction motor are finding increased use in industries because of applications that demand variable speed operations and because of the potential savings in energy usage. However, these drives sometimes produce unacceptably high levels in vibration and acoustic noise. A literature survey has revealed that while there has been intensive research on the design of inverters to minimize acoustic noise radiation from these drives, the vibro-acoustic behaviour of an induction motor structure has received relatively little attention. The primary objective of this research project, therefore, is to develop a general strategy/algorithm for estimating the acoustic noise radiated from inverter-driven induction motors. By using a three-phase, 2.2 kW induction motor, the vibration modes due to various structural components (such as the rotor, the stator/casing, the endshields and the base plate) of the motor structure were analysed by experimental modal testing. Results indicate that the vibration modes due to the rotor are only important at low frequencies. It has been found that the power injection method gives more accurate measurement of the damping of a motor structure than the modal testing and the time decay methods. If a point force excitation is used, then it is more accurate to measure the sound radiation efficiency than the power conversion efficiency for motor structures. The effect of three different inverter designs (an ideal ???almost sinusoidal??? controller and two commercially available PWM inverters) on the radiated acoustic power were assessed for both no-load and load conditions using sound intensity measurements conducted in an anechoic room. The results indicate that although the sound power level due to aerodynamic and mechanical noise increases at a rate of 12 dB per doubling of the motor speed, the electromagnetic noise dominates at low motor speeds and is still a significant noise source even at high motor speeds. For inverters with low switching frequencies, the radiated sound power level is almost 15 dB higher than the ideal case at low speeds and is relatively insensitive to the motor speed. For inverters that implement the random modulation technique, the change in the total sound power level with the level of the random modulation is very small but the tonal nature of the noise is greatly reduced. The vibration behaviour of a motor structure was modeled using the finite element method (FEM) and validated using the experimental modal testing results. It has been found that it is essential to model the laminated stator as an orthotropic structure. While the details of other structural components (such as the endshields, the teeth in the stator and the windings) are not so important, it is essential that they are incorporated into the structural model as simplified structures to account for their mass, stiffness and boundary conditions imposed on the motor structure. Based on this structural model, the radiated acoustic power for various operating conditions has been predicated using the boundary element (BEM) and the electromagnetic force calculated from an electromagnetic finite element model. The predicted results agree reasonably well with experimental measurements. Despite the success of the FEM/BEM approaches, they can be prohibitively expensive (in terms of computer resources required) to apply to large motors and high frequencies. Thus the feasibility of using a statistical method, namely, the statistical energy analysis (SEA), to estimate the radiated acoustic sound power from an inverter-driven induction motor has been examined. In order to carry out this analysis, analytical expressions for calculating the natural frequencies and radiation efficiency of finite length circular cylindrical shells (which are simplified models of the stator and casing of a motor structure) were firstly derived. The internal loss factors and coupling loss factors of the motor structure were determined experimentally using the power injection method. Then by introducing an equivalent surface mobility of circular cylindrical shells for the electromagnetic force, the vibration response and the acoustic noise radiated from each part of the motor structure were estimated. Results indicate that SEA method is potentially an efficient and effective tool in estimating the noise radiated from inverter-driven induction motors.

Ce travail présente deux méthodes expérimentales de caractérisation des propriétés élastiques et d'amortissement de matériaux poreux acoustiques en régime dynamique.<br />Les modules d'Young ou de cisaillement ainsi que les coefficients<br />d'amortissement de mousses polymères ou matériaux fibreux sont estimés dans leurs conditions usuelles d'utilisation, i.e. en flexion ou cisaillement et dans des gammes de température et de fréquence habituellement rencontrées dans les industries du bâtiment ou des transports. La théorie de Biot-Johnson-Champoux-Allard est utilisée pour décrire le comportement de ces matériaux poro-visco-élastiques modélisés comme des systèmes diphasiques constitués d'une phase solide et d'une phase fluide, l'air, couplées dans le temps et l'espace.<br /><br />La première méthode est dérivée de celle de la poutre d'Oberst : un<br />déplacement transverse est imposé au centre d'une poutre en conditions limites libre-libre. Un calcul par éléments finis hiérarchiques et un algorithme non-linéaire d'inversion sont utilisés afin d'estimer les paramètres inconnus des matériaux et de déterminer leurs évolutions en fonction de la fréquence et de la température.<br /><br />La seconde méthode est basée sur l'étude des vibrations d'une plaque multicouche en flexion. Un code numérique hiérarchique<br />simplifié est utilisé conjointement au précédent algorithme d'inversion dans le même but de caractérisation des matériaux poro-visco-élastiques.<br /><br />Des applications à quelques matériaux, visco-élastiques légers ou<br />mousses aux propriétés très différentes, ont permis de vérifier la pertinence de ces méthodes face à celles déjà existantes et d'en fixer les limitations.

Ce mémoire de thèse présente le développement de la méthode "statistical modal energy distribution analysis (SmEdA)" pour des systèmes vibro-acoustiques couplés. Cette méthode de calcul est basée sur le bilan énergétique dans des sous-systèmes fermés couplés, comme une structure ou une cavité. L'interaction entre de tels systèmes est décrite par des couplages entre les modes. La version initiale de SmEdA prend en compte seulement les modes qui ont une fréquence propre dans le bande d'excitation. Le travail présenté ici étudie l'effet des modes non résonants sur la réponse et identifie les cas dans lesquels un tel effet devient important. L'introduction des modes non résonants permet d'utiliser la méthode SmEdA dans des cas d'applications plus larges. En outre, une nouvelle méthode de post-traitement a été développée pour calculer des distributions d'énergie dans les sous-systèmes. Finalement, une nouvelle méthode d'approximation pour la prise en compte des modes de systèmes de grandes dimensions ou mal définis a été formulée. Toutes ces méthodes ont été comparées avec d'autres méthodes de calcul via des exemples académiques et industriels. Ainsi, la nouvelle version de SmEdA incluant le post-traitement pour obtenir des distributions d'énergie a été validé et les avantages et possibilités d'applications sont montrés.