Floors have traditionally been designed only for strength and deflection serviceability. As technological advances have been made in medical, scientific and micro-electronics manufacturing, many types of equipment have become sensitive to vibration of the supporting floor. Thus, vibration serviceability has become a routinely evaluated limit state for floors supporting sensitive equipment. Equipment vibration tolerance limits are sometimes expressed as waveform peak acceleration, and are more often expressed as narrowband spectral acceleration, or one-third octave spectral velocity. Current floor vibration prediction methods, such as those found in the American Institute of Steel Construction Design Guide 11, Floor Vibrations Due to Human Activity, the British Steel Construction Institute P354, Design of Floors for Vibration: a New Approach and the British Concrete Centre CCIP-016 A Design Guide for Footfall Induced Vibration of Structures, have limitations. It has been observed that non-structural components such as light-weight partitions could significantly change floor dynamic properties. Current prediction methods do not provide a fundamental frequency manual prediction method nor finite element modeling guidance for floors with non-structural components. Current prediction methods only predict waveform peak acceleration and do not provide predictions for frequency domain response including narrowband spectral acceleration or one-third octave spectral velocity. Also, current methods are not calibrated to provide a specific level of conservatism. This research project provides (1) a fundamental frequency manual prediction method for floors with lightweight partitions; (2) an improved finite element modeling procedure for floors with light-weight partitions; (3) a procedure to predict the vibration response in narrow-band spectrum and one-third octave band spectrum which can be directly compared with vibration tolerance limits; and (4) a simplified experimental procedure to estimate the floor natural frequencies. An experimental program including four steel-framed building floors and a concrete was completed. Modal tests were performed on two of the steel-framed buildings and the concrete building using an electrodynamic shaker. Experimental modal analysis techniques were used to estimate the modal properties: natural frequencies, mode shapes, and damping ratios. Responses to walking excitation were measured several times in each tested bay for individuals walking at different walking speeds. During each test, the walker crossed the middle of the bay using a metronome to help maintain the intended cadence. The proposed method was used to predict the modal properties and responses to walking. The measurements are used to assess the precision of the proposed methods and to calibrate the prediction methods to provide a specific probability that the actual response will exceed the predicted response. Comparison of measurements and predictions shows the proposed methods are sufficiently accurate for design usage.

Given the advantages of parallel manipulators and lightweight manipulators, a 3-PRR planar parallel manipulator with three lightweight intermediate links has been developed to provide an alternative high-speed pick-and-place positioning mechanism to serial architecture manipulators in electronic manufacturing, such as X-Y tables or gantry robots. Lightweight members are more likely to exhibit structural defection and vibrate due to the inertial forces from high speed motion, and external forces from actuators. Structural flexibility effects are much more pronounced at high operational speeds and accelerations. Therefore, this thesis presents the dynamics and vibration control of a 3-PRR parallel manipulator with three flexible links. Firstly, a procedure for the generation of dynamic equations for a 3-PRR parallel manipulator with three flexible intermediate links is presented based on the assumed mode method. The dynamic equations of the parallel manipulator with three flexible intermediate links are developed using pinned-pinned boundary conditions. Experimental modal tests are performed using an impact hammer and an accelerometer to identify the mode shapes, frequencies, and damping ratios of flexible intermediate links. The mode shapes and frequencies, obtained from experimental modal tests, match very well the assumed mode shapes and frequencies obtained based on pinned-pinned boundary conditions, and therefore the dynamic model developed is validated. Secondly, this thesis presents the investigation on dynamic stiffening and buckling of the flexible links of a 3-PRR parallel manipulator by including the effect of longitudinal forces on the modal characteristics. Natural frequencies of bending vibration of the intermediate links are derived as the functions of axial force and rigid-body motion of the manipulator. Dynamic stiffening and buckling of intermediate links is investigated and configuration-dependant frequencies are analyzed. Furthermore, using Lagrange multipliers, the fully coupled equations of motions of the flexible parallel manipulator are developed by incorporating the rigid body motions with elastic motions. The mutual dependence of elastic deformations and rigid body motions are investigated from the analysis of the derived equations of motion. Open-loop simulation without joint motion controls and closed-loop simulation with joint motion controls are performed to illustrate the effect of elastic motion on rigid body motions and the coupling effect amongst flexible links. These analyses and results provide valuable insight into the design and control of the parallel manipulator with flexible intermediate links. Thirdly, an active vibration control strategy is developed for a moving 3-PRR parallel manipulator with flexible links, each of which is equipped with multiple PZT control pairs. The active vibration controllers are designed using the modal strain rate feedback (MSRF). The amplification behavior of high modes is addressed, and the control gain selection strategy for high modes is developed through modifying the IMSC method. The filters are developed for the on-line estimation of modal coordinates and modal velocity. The second compensator is used to cut off the amplified noises and unmodeled dynamics due to the differentiation operation in the developed controller. The modal coupling behavior of intermediate links is examined with the modal analysis of vibrations measured by the PZT sensors. The error estimation of the moving platform is examined using the measurement of PZT sensors. Finally, an active vibration control experimental system is built to implement the active vibration control of a moving 3-PRR parallel manipulator with three flexible links. The smart structures are built through mounting three PZT control pairs to each intermediate flexible link. The active vibration control system is set up using National Instruments LabVIEW Real-Time Module. Active vibration control experiments are conducted for the manipulator moving with high-speed, and experimental results demonstrate that the vibration of each link is significantly reduced.

Given the advantages of parallel manipulators and lightweight manipulators, a 3-PRR planar parallel manipulator with three lightweight intermediate links has been developed to provide an alternative high-speed pick-and-place positioning mechanism to serial architecture manipulators in electronic manufacturing, such as X-Y tables or gantry robots. Lightweight members are more likely to exhibit structural defection and vibrate due to the inertial forces from high speed motion, and external forces from actuators. Structural flexibility effects are much more pronounced at high operational speeds and accelerations. Therefore, this thesis presents the dynamics and vibration control of a 3-PRR parallel manipulator with three flexible links. Firstly, a procedure for the generation of dynamic equations for a 3-PRR parallel manipulator with three flexible intermediate links is presented based on the assumed mode method. The dynamic equations of the parallel manipulator with three flexible intermediate links are developed using pinned-pinned boundary conditions. Experimental modal tests are performed using an impact hammer and an accelerometer to identify the mode shapes, frequencies, and damping ratios of flexible intermediate links. The mode shapes and frequencies, obtained from experimental modal tests, match very well the assumed mode shapes and frequencies obtained based on pinned-pinned boundary conditions, and therefore the dynamic model developed is validated. Secondly, this thesis presents the investigation on dynamic stiffening and buckling of the flexible links of a 3-PRR parallel manipulator by including the effect of longitudinal forces on the modal characteristics. Natural frequencies of bending vibration of the intermediate links are derived as the functions of axial force and rigid-body motion of the manipulator. Dynamic stiffening and buckling of intermediate links is investigated and configuration-dependant frequencies are analyzed. Furthermore, using Lagrange multipliers, the fully coupled equations of motions of the flexible parallel manipulator are developed by incorporating the rigid body motions with elastic motions. The mutual dependence of elastic deformations and rigid body motions are investigated from the analysis of the derived equations of motion. Open-loop simulation without joint motion controls and closed-loop simulation with joint motion controls are performed to illustrate the effect of elastic motion on rigid body motions and the coupling effect amongst flexible links. These analyses and results provide valuable insight into the design and control of the parallel manipulator with flexible intermediate links. Thirdly, an active vibration control strategy is developed for a moving 3-PRR parallel manipulator with flexible links, each of which is equipped with multiple PZT control pairs. The active vibration controllers are designed using the modal strain rate feedback (MSRF). The amplification behavior of high modes is addressed, and the control gain selection strategy for high modes is developed through modifying the IMSC method. The filters are developed for the on-line estimation of modal coordinates and modal velocity. The second compensator is used to cut off the amplified noises and unmodeled dynamics due to the differentiation operation in the developed controller. The modal coupling behavior of intermediate links is examined with the modal analysis of vibrations measured by the PZT sensors. The error estimation of the moving platform is examined using the measurement of PZT sensors. Finally, an active vibration control experimental system is built to implement the active vibration control of a moving 3-PRR parallel manipulator with three flexible links. The smart structures are built through mounting three PZT control pairs to each intermediate flexible link. The active vibration control system is set up using National Instruments LabVIEW Real-Time Module. Active vibration control experiments are conducted for the manipulator moving with high-speed, and experimental results demonstrate that the vibration of each link is significantly reduced.

Este trabalho realiza uma investigação sobre técnicas experimentais para a determinação de Funções de Resposta em Freqüência (FRFs) angulares com aplicações em estruturas do tipo viga. Estas FRFs são definidas considerando-se como variável de saída o movimento angular (deslocamento, velocidade ou aceleração) exibido pela estrutura sob estudo quando a mesma é excitada por uma força linear ou um momento puro. Dada a grande dificuldade em se aplicar um momento puro à estrutura sob estudo, este trabalho utiliza como forma de excitação apenas esforços lineares que podem ser aplicados através de técnicas usuais de excitação em análise modal, tais como o excitador eletrodinâmico e o martelo impulsivo. Portanto as FRFs obtidas descrevem relações de saída e entrada do tipo Angular/Linear. Uma das técnicas utilizadas na determinação das FRFs angulares consiste na excitação de uma estrutura com um excitador eletrodinâmico e utilização de um corpo rígido na forma de um bloco T que é montado sobre a estrutura sob estudo. Dois acelerômetros lineares devem ser montados sobre o bloco T e a partir das duas acelerações lineares medidas buscam-se estimar um sinal proporcional à aceleração angular da estrutura no ponto de conexão. Outra técnica utiliza uma formulação via diferenças finitas, onde dois ou três acelerômetros (de acordo com a formulação de diferenças finitas utilizada) igualmente espaçados são montados diretamente sobre a estrutura sob estudo e são usados para se derivar à aceleração angular. Os resultados obtidos a partir destas técnicas são comparados com resultados obtidos a partir da utilização de um acelerômetro angular piezelétrico recentemente disponível no mercado. Também foram desenvolvidos modelos analíticos e computacionais via método dos elementos finitos a fim de se gerar subsídios adicionais para a análise dos resultados. Foram feitas várias constatações e dentre elas destacam-se resultados onde as FRFs angulares/lineares resultantes podem sofrer alterações significativas dependendo de como os dados experimentais são processados. / The present work aims to perform an investigation on experimental techniques for the determination of angular Frequency Response Functions (FRFs) in Modal Testing. Angular FRFs are those where the output variable is given by angular displacement, velocity or acceleration, whereas the input is given in terms of linear or angular quantities (a pure moment). Since the application of a pure moment as an excitation source still remains as a challenge, this work is focused in studying techniques to estimate angular/linear types of angular FRFs. One of these techniques consists of exciting the structure with a shaker and using a rigid T-block to measure the linear accelerations and then calculating the angular FRFs from these linear accelerations. Another technique employs finite differences formulations to get the angular motions. This technique uses the closely spaced accelerometers mounted directly to the structure under test, where at least two (according to the finite difference formula employed) accelerometers are used. The translational measurements are gathered and finite difference formulas are used to derive the necessary angular quantities. Additional tests are performed with an angular piezoelectric accelerometer recently available in the market in order to provide a comparison basis for the results obtained using the two techniques. The results are also compared whit theoretical models developed using analytical and Finite Element Formulations. Among all results obtained, it was understood that depending on the level of angular vibrations exhibited by the structure, and how the signals are processed, the resulting angular FRFs can suffer some significant changes

Orientador: João Antonio Pereira / Banca: Luiz de Paula do Nascimento / Banca: Mário Francisco Mucheroni / Resumo: Este trabalho apresenta o estudo, a implementação e a aplicação de duas técnicas de identificação de parâmetros modais utilizando apenas as respostas da estrutura, denominadas: Identificação Estocástica de Subespaço (IES) e Decomposição no Domínio da Freqüência (DDF). A IES é baseada na Decomposição em Valores Singulares (DVS) da projeção ortogonal do espaço das linhas das saídas futuras no espaço das linhas das saídas passadas. Uma vez realizada a DVS da projeção ortogonal é possível obter o modelo de espaço de estado da estrutura e os parâmetros modais são estimados diretamente através da decomposição em autovalores e autovetores da matriz dinâmica. A DDF é baseada na DVS da matriz de densidade espectral de potência de saída nas linhas de freqüências correspondentes a região em torno de um modo. O primeiro vetor singular obtido para cada linha de freqüência contém as respectivas informações daquele modo e os correspondentes valores singulares levam a função densidade espectral de um sistema equivalente de um grau de liberdade (1GL), permitindo a obtenção dos parâmetros do respectivo modo. Os métodos são avaliados utilizando dados simulados e experimentais. Os resultados mostram que as técnicas implementadas são capazes de estimar os parâmetros modais de estruturas utilizando apenas as respostas. / Abstract: This work presents the study, implementation and application of the two techniques for the modal parameters identification using only response data: Stochastic Subspace Identification (SSI) and Frequency Domain Decomposition (FDD). The SSI is based on Singular Value Decomposition (SVD) of the orthogonal projection of the future output row space in the past output row space. After the completion of the SVD of the orthogonal projection, is possible to get the state space model of the structure and the modal parameters are estimated directly through the eigenvalues and eigenvectors decomposition of the dynamic matrix. The FDD is based on the SVD of the output power spectral density matrix in the frequencies lines around a mode. The first singular vector obtained for each frequency line contains the respective information about this mode and the corresponding spectral density function leads to an equivalent system of one degree of freedom (1 DOF), allowing the calculation of the parameters of the mode. The methods are evaluated using simulated and experimental data. The results show that the techniques implemented are capable to estimate the modal parameters of structures using only response data. / Mestre

Le crissement de frein est une nuisance sonore récurrente dans l’automobile qui, bien qu’il n’existe pas de solution robuste lors de la conception, se traduit souvent par des pénalités économiques envers les équipementiers. La simulation numérique et les caractérisations expérimentales sont les deux voies classiques pour analyser les phénomènes crissants. La simulation permet une analyse fine des comportements vibratoires et l’évaluation de l’impact de modifications de conception, mais les modèles sont imparfaits et leur domaine de validité doit être établi. La mesure garantit que tous les paramètres sont pris en compte, mais elle est souvent spatialement pauvre et peu adaptée à la prédiction de modifications. Dans ce contexte, les objectifs de la thèse sont de faire un bilan des méthodes d’exploitation des essais, de fournir des outils permettant une interaction facilitée entre les équipes d’essai et de simulation et de mettre en œuvre les développements dans une application métier.Après un rappel sur les hypothèses et les méthodes pour l’identification modale, une analyse critique des résultats d’identification aboutit à la caractérisation des biais et à l’introduction de critères détaillés permettant d’évaluer la qualité du résultat. La corrélation calcul/essai permet de qualifier les modèles et des outils sont intégrés pour faciliter sa mise en œuvre. Plusieurs critères sont définis pour mettre en évidence les sources de mauvaise corrélation venant de la mesure d’une part et du modèle d’autre part. Les sources de mauvaise corrélation sont également identifiées avec l’algorithme d’expansion MDRE, dont les limitations sont comparées à celles des expansions statiques et dynamiques.Le recalage des modèles est ensuite traité par une procédure systématique imbriquant les étapes de recalage de la géométrie, des propriétés matériau et des contacts entre composants. Une paramétrisation du contact est proposée avec une réduction multi-modèle permettant des temps de calculs compatibles avec le temps industriel. Elle permet également des études de sensibilités avec l’introduction de la notion de modes de composant dans un assemblage.Une analyse détaillée d’une campagne de mesure sur un système de frein en conditions crissantes est enfin réalisée. Un parallèle entre les évolutions du cycle limite et le degré de couplage des modes complexes est proposé. Dans le domaine temps-fréquence, la variabilité, la répétabilité, la reproductibilité et la possibilité d’agréger des mesures séquentielles sont évaluées. Le résultat obtenu est finalement étendu sur le modèle éléments finis, ce qui permet des interprétations utiles. / Brake squeal is a nuisance commonly encountered in the car industry which often results in financial penalties towards brake manufacturers, even if no robust solution exists for the conception. Numerical simulation and experimental characterizations are the classical two-track approaches to analyze squeal phenomena. Numerical simulation allows a fine analysis of vibration behaviors and the evaluation of conception modification impacts, but models are not perfect and their validity domain must be defined. Measurement guarantees that every parameter is taken into account, but it is often spatially poor and not really suited for modification prediction. In this context, the thesis objectives are to make an overview of test operating methods, to provide tools allowing an easier interaction between both test and simulation teams and to enforce the developments in a business application.After a review of the hypotheses and methods for modal identification, a critical analysis of the identification results leads to the characterization of biases and the introduction of detailed criteria to evaluate the quality of the result. Calculation/test correlation allows to qualify the models and tools are integrated to facilitate its implementation. Several criteria are defined to highlight the sources of bad correlation coming from the measure on the one hand and the model on the other hand. Sources of poor correlation are also identified with the MDRE expansion algorithm, whose limitations are compared with those of static and dynamic expansions.Model updating is then processed by a systematic procedure imbricating the steps of updating of geometry, material properties and contacts between components. A contact parametrization is proposed with a multi-model reduction allowing calculation times compatible with industrial time. It also allows sensitivity studies with the introduction of the notion of component modes in an assembly.Finally, a detailed analysis of a measurement campaign on a braking system under squeal conditions is carried out. A parallel between the changes of the limit cycle and the degree of coupling of the complex modes is proposed. In the time-frequency domain, variability, repeatability, reproducibility and the ability to aggregate sequential measurements are evaluated. The result is finally extended on the finite element model, which allows useful interpretations.

A análise dinâmica experimental tem sido amplamente pesquisada como uma ferramenta de avaliação de integridade de estruturas de concreto armado. Existem técnicas de identificação de danos baseadas em propriedades modais como frequências de ressonâncias, deformadas modais, curvaturas modais e amortecimento. Há também técnicas baseadas na não linearidade da resposta dinâmica, que apesar do grande potencial na detecção de danos, têm sido pouco exploradas nos últimos anos. Este trabalho tem por objetivo avaliar a integridade estrutural de vigas de concreto armado através do comportamento da resposta dinâmica. Foram realizados ensaios dinâmicos em duas vigas de concreto armado com 3,5 m de comprimento, 25 cm de largura, 35 cm de altura e idênticas taxas de armaduras, mas configuradas com barras de aço de diferentes diâmetros, 2 &#981 16 mm e 8 &#981 8 mm, respectivamente. Tais vigas, inicialmente íntegras, foram submetidas a ciclos de carregamento e descarregamento com intensidades crescentes até atingir a ruptura do elemento. Após cada ciclo, as propriedades dinâmicas foram avaliadas experimentalmente, com o emprego de técnicas de excitação por sinais do tipo aleatório e tipo transiente, respectivamente, visando determinar parâmetros que indiquem a deterioração gradativa do elemento. Nesses ensaios dinâmicos aplicaram-se diferentes amplitudes da força de excitação. Verificou-se que o aumento da amplitude da força dinâmica de excitação provocou reduções nos valores das frequências de ressonância de 1,1% e 2,4%, associadas, respectivamente, às excitações aleatórias e transientes; e um comportamento não linear dos índices de amortecimento, associados às excitações aleatórias, mantendo um crescimento linear com as excitações transientes. Constatou-se, ainda, que os valores das frequências de ressonância decrescem com a redução de rigidez mecânica, diminuída com o aumento do nível de fissuração induzido nos modelos. Já os valores dos índices de amortecimento, após cada ciclo, se comportaram de forma não linear e assumiram diferentes valores, conforme a técnica de excitação empregada. Acredita-se que esta não linearidade está relacionada aos danos provocados no elemento pela solicitação estrutural e, por consequência, ao processo de como a dissipação de energia é empregada no processo de instauração, configuração e propagação das fissuras nos elementos de concreto armado. / The experimental dynamic analysis has been widely investigated as a tool to assess integrity of reinforced concrete structures. State-of-the-art modal-based techniques for structural damage analysis use resonance frequencies, modal deformed, curvature and modal damping analysis. There are also techniques based on nonlinear dynamic response that despite the large potential for damage detection, have not been considered in recent years. This work aims to evaluate the structural integrity of reinforced concrete beams through the dynamic response behavior. Dynamic tests were performed on two reinforced concrete beams with 3.5 meters of length, 250 mm of width, 350 mm of height and identical reinforcement rates, but different steel bars diameters, 2 &#981 16 mm e 8 &#981 8 mm, respectively. Beams were initially intact and were subjected to loading and unloading cycles with increasing intensity until the rupture of the concrete beam. After each cycle, the dynamic properties were evaluated experimentally with random and transient excitation signals to determine parameters that indicate gradual deterioration of the beam. Different amplitudes of the excitation force were applied during dynamic testing of concrete beams. These results indicated that an increase in amplitude of dynamic excitation force caused reductions in frequency resonances by 1.1% and 2.4%, associated with the random and transient excitations, respectively. Furthermore, a non-linear behavior of the damping ratios related to random excitations was observed while a linear relationship with transient excitation was found. Moreover, it was found that the resonance frequency decreased with the cracking-related reduction of beam stiffness, caused by increasing level of loading. Additionally, the values of the damping ratios after each cycle behaved non-linearly assuming different values according to the used excitation technique. It is believed that this non-linearity can be related to the loading-related structural damage of the reinforced concrete beam. Therefore, the energy dissipation related to cracking initiation, configuration and propagation in reinforced concrete elements played an important role in the damping ratios of concrete beams.

Orientador: Samuel da Silva / Resumo: Os métodos de análise modal tradicionalmente são limitados aos sistemas vibrando em regime linear de movimento. Assim, quando as estruturas sofrem altas amplitudes de excitação ou são muito flexíveis, gerando possíveis vibrações não-lineares, estes métodos acabam perdendo a sua validade e as propriedades características. Com base nesta motivação, este trabalho apresenta um estudo detalhado para mostrar quais as limitações de se aproximar por parâmetros modais sistemas vibrando em regime de movimento não linear. Para ilustrar a formulação, assume-se uma viga engastada e livre emulando um oscilador de Duffing com não linearidade concentrada, suave e polinomial (rigidez cúbica). Observa-se que para regimes de excitação baixa, pode-se extrair parâmetros modais do modelo e ajustá-los para níveis de excitação mais altos quando se induz vibração não-linear pelo aumento do nível da amplitude de excitação. Para situações de vibração não-linear opta-se por aproximar os sinais e saídas pelo método de superfície de resposta e identificar a dependência amplitude-frequência para extração de modos normais não-lineares. Os resultados apresentados com a formulação descrita neste trabalho permitem adaptar adequadamente as ferramentas convencionais de análise modal linear para validade e aplicação direta em casos de vibração em regime não linear, quando estes ainda são considerados de fraca influência. / Abstract: Modal analysis methods have traditionally been limited to systems vibrating in linear motion regime. Thus, when the structures undergo high excitation amplitudes or are very flexible, generating possible nonlinear vibrations, these methods end up losing their validity and characteristic properties. Based on this motivation, this work presents a detailed study to show the limitations of approaching by modal parameters systems vibrating in nonlinear regime. To illustrate the formulation, a cantilever beam is assumed to emulate a Duffing oscillator with concentrated, smooth, polynomial nonlinearity (cubic stiffness). It is observed that for low excitation regimes, one can extract modal parameters from the model and adjust them to higher excitation levels when inducing nonlinear vibration by increasing the excitation amplitude level. For nonlinear vibration situations, we choose to approximate the signals and outputs by the response surface method and identify the amplitude-frequency dependence for extraction of nonlinear normal modes. The results presented with the formulation described in this work allow to adapt adequately the conventional tools of linear modal analysis for validity and direct application in cases of vibration in nonlinear regime, when they are still considered of low influence. / Mestre

Les méthodes de modification structurale sont l'ensemble des techniques qui permettent d'estimer l'influence d'une modification quelconque sur le comportement dynamique d'une structure. Dans le cadre général, le modèle de comportement de la structure cible, comme celui de la modification, peut être aussi bien numérique qu'expérimental. Dans le cadre de ce travail, seul le cas particulier du couplage d'un modèle expérimental de la structure et d'un modèle numérique de la modification sera traité.<br />Les concepts fondamentaux utilisés dans cette thèse sont ensuite présentés. Les relations de la dynamique des structures pour les problèmes discrets sont rappelées, ainsi que les principes de la synthèse modale, de la sous-structuration dynamique et de la réduction de modèle, tout comme la notion de modes d'interface. Les formulations classiques des méthodes de modification structurale sont ensuite détaillées pour en illustrer les limitations et les restrictions.<br />Une formulation originale permettant de prendre en compte les incompatibilités entre les mesures et les DDL de l'interface structure/modification et de régulariser la construction d'un modèle de comportement couplé est alors proposée. Cette première contribution de la thèse repose sur l'utilisation des techniques d'expansion de données et de réduction de modèle. Des indicateurs sont également construits pour estimer la cohérence de la prédiction réalisée. Les évolutions sont appliquées au cas d'un démonstrateur numériques et les résultats sont comparés avec les prédictions réalisées par les méthodes classiques. La méthodologie associée à cette nouvelle formulation est alors largement exposée.<br />L'influence des différents facteurs intervenant dans la construction du modèle couplé et la qualité de la prédiction est ensuite analysée en détail. Cette analyse permet de dresser une liste non exhaustive des précautions à prendre lors de la mise en œuvre de la méthode proposée, depuis la réalisation pratique de l'analyse modale expérimentale jusqu'à l'interprétation des premiers résultats.<br />Enfin, plusieurs applications sont présentées. Une première structure académique démontre la faisabilité de la méthode. Une deuxième étude, réalisée sur un cas industriel, illustre les gains de temps potentiels en comparant la prédiction avec les résultats d'une étude basée sur un modèle EF recalé de la structure. La troisième étude illustre l'application de la méthode dans un cas type. L'analyse modale de la structure cible permet de comprendre le problème, une modification est conçue, réalisée et mise en place. La prédiction est ensuite comparée aux résultats de l'analyse modale de la structure modifiée. Enfin, la dernière application montre les limites de la méthodologie. L'étude multi-objectifs sur une large bande de fréquences d'une structure industrielle permet de faire une ouverture vers la suite des travaux et montre la nature des difficultés à surmonter.

In this study, experimental modal analysis was studied. Experimental modal analysis includes modal testing, modal parameter estimation and calibration. For this purpose a 4 span skewed steel frame was constructed in Structural Mechanics Laboratory of Civil Engineering Department of METU. The model was transported to Vibration and Acoustic Laboratory of Mechanical Engineering Department of METU. The tests were conducted by cooperation with Vibration and Acoustics Laboratory. Due to lack of experimental modal analysis software in Structural Mechanics Laboratory, modal parameter estimation and finite element updating softwares were written in Matlab platform. The written softwares were executed on the data obtained from modal testing. 15 reasonable modes are extracted from the FRFs that are obtained from modal testing. 59.23 percent consistency is found for the nominal modal comparison. At the end of calibration process 76.14 percent consistency is achieved between the experimental results and analytical results.