O objetivo deste trabalho é estudar métodos dinâmicos de detecção de dano em vigas, em especial os métodos baseados na variação da flexibilidade medida dinamicamente. Os métodos revisados formam parte das técnicas de Detecção de Dano Não Destrutivas (DDND). Nas técnicas DDND o dano é localizado por comparação entre o estado sadio e o danificado da estrutura. Neste trabalho, o problema de vibração inverso é apresentado e a matriz de flexibilidade estática da estrutura é determinada a partir de seus parâmetros modais. Com ajuda de um Modelo de Elementos Finitos (MEF) são mostrados os diferentes padrões de variação da matriz de flexibilidade produzidos pela presença do dano. Baseando-se nestes padrões é possível identificar a posição do dano dentro da estrutura, como indicado pelos diversos exemplos apresentados. / The purpose of this work is to study dynamic methods for damage detection in beam structures. The attention is devoted to the methods based on dynamically measured flexibility. The reviewed methods are part of Nondestructive Damage Detection techniques (NDD). In the NDD techniques the damage is determined through the comparison between the undamaged and damaged state of the structure. In this work the inverse vibration problem is presented and the structure\'s flexibility matrix calculated from his modal parameters. The Finite Elements Model (FEM) is employed to show that a clear pattern exist for the changes in the flexibility matrix produced due to the presence of damage. The flexibility matrix changes is used to identify and locate damage as indicated by the several examples presented.

O presente trabalho tem como objetivo desenvolver novas alternativas de identificação modal para estruturas sob excitações em condição de operação baseadas em funções de transmissibilidade. Recentes metodologias formuladas sobre conceitos de transmissibilidade têm surgido como alternativa para a identificação de parâmetros modais de estruturas. A identificação nestas metodologias é independente do espectro da excitação, sendo uma vantagem importante com respeito a metodologias anteriores no domínio da frequência que supõem a excitação como ruído branco. Dessa forma, aproveitando os diferentes trabalhos dirigidos a avaliar parâmetros modais com uso da transmissibilidade, são propostas três novas alternativas. A primeira delas propõe a decomposição de valores singulares sobre matrizes de funções de transmissibilidade escalar com densidade espectral para estimar frequências naturais e modos de vibração. A segunda alternativa propõe o conceito de funções de transmissibilidade multivariável com diferente referência para a identificação modal. E a terceira introduz uma melhora na primeira alternativa incluindo a possibilidade da estimação de taxas de amortecimento. Uma ferramenta computacional para a análise modal é desenvolvida como apoio para as simulações numéricas de verificação das metodologias de identificação modal propostas. Diferentes exemplos numéricos com uma viga submetida a excitações de ruído colorido mostram que os métodos propostos são capazes de identificar parâmetros modais sem a introdução das frequências adicionais devido às excitações de ruído colorida utilizadas. Além disso, os dados de um teste de vibrações sobre uma ponte em operação foram utilizados para verificar os métodos. / This research aims to develop new alternatives of modal identification for structures under excitation in operation condition based on transmissibility functions. Latest methodologies based on transmissibility concepts have been arising as alternatives for modal parameter identification of structures. Modal parameter identification in this type methodology is input spectrum independent being an important advantage with respect previous frequency domain methods that assumes white noise excitation. Different alternatives of modal identification based on transmissibility functions are proposed in this work. The first of them proposes singular value decomposition on scalar transmissibility functions matrices with spectral density to estimate natural frequencies and vibration modes (PSDTM-SVD method). A second alternative proposes the concept of multivariable transmissibility functions with different transferring outputs for modal parameter identification. And the third alternative proposes an enhanced PSDTM-SVD method, which permits to identify modal damping. Computational tool for modal analysis is developed as a support for the numerical simulations of verification of modal identification methodologies proposed. Different numerical examples of a beam model subjected to colored noise excitations show that the proposed methods are capable of identifying modal parameters without the introduction of the additional frequencies due to the excitations used. Furthermore, data from an operational vibration bridge test were used to verify the methods.

This thesis deals with the complex analysis of dynamic characteristics of the exhaust system and the transmission of noise from exhaust system to the vehicle. In the first part is performed modal analysis, which is compared to experimentally measured data. Modal analysis of the complete exhaust system is made with focusing on the influence of temperature of the exhaust system. There is also performed an analysis of noise transmission from exhaust system to the car cockpit.

Cieľom tejto diplomovej práce je predstavenie Experimentálnej modálnej analýzy a štúdium a použitie algoritmov pre výpočet modálnych parametrov z odmeraných vibrácií pri Experimentálnej modálnej analýze. Bodom záujmu sú predovšetkým algoritmy, pracujúce s odmeranými dátami v časovej oblasti. Diplomová práca sa zaoberá programovaním algoritmu ITD a jeho implementáciou pre Experimentálnu modálnu analýzu. Ďaľšou časťou práce je štúdium citlivosti algoritmu na okrajové podmienky meranej sústavy pri výpočte modálnych parametrov a štúdium možnosti využitia algoritmu pri Operačnej modálnej analýze.

Machining chatter is an undesirable phenomenon of material removal processes and hardly to control or avoid. Its occurrence and extent essentially depend onthe kinematic, which alters with the position of the Tool Centre Point, of the machine tool. Research as to chatter was done widely but rarely with respect to changing structural dynamics during manufacturing. This thesis applies intelligent methods to learn the underlying functions of modal parameters – natural frequency, damping ratio, and mode shape – and defines the dynamic properties of a system firstly at this extent. To do so, it embraces three steps: first, the elaboration of the necessary dynamic parameters, second, the acquisition of the data via a simulation,and third, the prediction of the modal parameters with two kinds of Machine Learning techniques: Gradient Boosting Machine and Multilayer Perceptron. In total, it investigates three types of kinematics: cross bed, gantry, and overhead gantry. It becomes apparent that Light Gradient Boosting Machine outperforms Multilayer Perceptron throughout all studies. It achieves a prediction error of at most 1.7 % for natural frequency and damping ratio for all kinematics. However, it cannot really control the prediction of the participation factor yet which might originate in the complexity of the data and the data size. As expected, the error rises with noisy data and less amount of measurement points but at a tenable extent for both natural frequency and damping ratio. / 'Bearbetningsvibrationer är ett oönskat fenomen i materialborttagningsprocesser och är svåra att kontrollera eller undvika. Dess förekomst och omfattning beror i huvudsak på kinematiken, som förändras med positionen för verktygets centrumpunkt på verktygsmaskinen. Det har gjorts mycket forskning om bearbetningsvibrationer, men sällan om förändrad strukturell dynamik under tillverkningen. I denna avhandling tillämpas intelligenta metoder för att lära sig de underliggande funktionerna hos modalparametrar – egenfrekvens, dämpningsgrad och modalform – och definierar systemets dynamiska egenskaper för första gången i denna omfattning. För att göra detta omfattar den tre steg: för det första utarbetandet av de nödvändiga dynamiska parametrarna, för det andra insamling av data via en simulering och för det tredje förutsägelse av modalparametrarna med hjälp av två typer av tekniker för maskininlärning: Gradient Boosting Machine och Multilayer Perceptron. Sammanlagt undersöks tre typer av kinematik: crossbed, gantry och overhead gantry. Det framgår tydligt att Light Gradient Boosting Machine överträffar Multilayer Perceptron i alla studier. Den uppnår ett prediktionsfel på högst 1,7 % för egenfrekvens och dämpningsförhållande för alla kinematiker. Den kan dock ännu inte riktigt kontrollera förutsägelsen av deltagarfaktorn, vilket kan bero på datans komplexitet och datastorlek. Som väntat ökar felet med bullrig data och färre mätpunkter, men i en acceptabel omfattning för både naturfrekvens och dämpningsförhållande.

[pt] Identificação paramétrica de sistemas mecânicos é uma das principais aplicações das técnicas de identificação de sistemas na Engenharia Mecânica, especificamente para a identificação de parâmetros modais de estruturas flexíveis. Um dos principais problemas na identificação é a presença de ruido nas medições. Este trabalho apresenta uma análise na presença de ruído de alguns dos métodos no domínio do tempo aplicáveis na identificação de parâmetros modais de sistemas mecânicos lineares invariantes no tempo com múltiplas entradas e múltiplas saídas (MIMO), usando como base o modelo em espaço de estados tipicamente usado em Dinâmica e Vibrações. Os algoritmos de subespaço envolvidos neste estudo destacam-se pela utilização da decomposição em valores singulares (SVD) dos dados, obtendo subespaços ortogonais dos modos associados ao sistema e dos modos associados ao ruído. Outros complicadores no processo de identificação que serão explorados neste trabalho são: flexibilidde e baixo amortecimento. Comparam-se as técnicas usando o modelo no espaço de estado da estrutura Mini- mast desenvolvida pela NASA Langley Research Center e simulações são feitas variando o nível de ruído nos dados, o amortecimento e a flexibilidade da estrutura. O problema de identificação de parâmetros estruturais (matrizes de massa, rigidez e amortecimento) também é estudado, as características e limitações dos algoritmos utilizados são analisados. Como exemplo de aplicação prática, um experimento foi realizado para identificar os parâmetros modais e estruturais de um rotor flexível e os resultados são discutidos. / [en] Parametric identification of mechanical systems is one of the main applications of the system identification techniques in Mechanical Engineering, specifically for the identification of modal parameters of flexible structures. One of the main problems in the identification is the presence of noise in the measurements. This work presents an analysis in the presence of noise of some of the time domain methods applicable in modal parameters identification of linear time-invariant mechanical systems with multiple inputs and multiple outputs (MIMO), using as base the state-space model typically used in Dynamics and Vibrations. The subspace algorithms involved in this study are distinguished for the use of the singular values decomposition (SVD) of the data, obtaining ortogonal subspaces of the modes associates to the system and of the modes associates to the noise. Other complicators in the identification process that will be explored in this work are: flexibility and low damping. The techniques are compared using the state-space model of the Mini-mast structure developed for NASA Langley Research Center and simulations are made varying the level of noise in the data, the damping and the flexibility of the structure. The problem of identification of structural parameters (mass, stiffness and damping matrices) also is studied, the characteristics and limitations of the used algorithm is analyzed. As example of practical application, an experiment was made to identify the modal parameters of a flexible rotor and the results are discussed.

There has been increasing interest in evaluating the performance of existing reinforced concrete (RC) bridges just after natural disasters or man-made events especially when the defects are invisible, or in quantifying the improvement after rehabilitations. In order to obtain an accurate assessment of the reliability of a RC bridge, it is critical to incorporate information about its current structural properties, which reflects the possible aging and deterioration. This dissertation proposes to develop an adaptive reliability analysis of RC bridges incorporating the damage detection information obtained from nondestructive testing (NDT). In this study, seismic fragility is used to describe the reliability of a structure withstanding future seismic demand. It is defined as the conditional probability that a seismic demand quantity attains or exceeds a specified capacity level for given values of earthquake intensity. The dissertation first develops a probabilistic capacity model for RC columns and the capacity model can be used when the flexural stiffness decays nonuniformly over a column height. Then, a general methodology to construct probabilistic seismic demand models for RC highway bridges with one single-column bent is presented. Next, a combination of global and local NDT methods is proposed to identify in-place structural properties. The global NDT uses the dynamic responses of a structure to assess its global/equivalent structural properties and detect potential damage locations. The local NDT uses local measurements to identify the local characteristics of the structure. Measurement and modeling errors are considered in the application of the NDT methods and the analysis of the NDT data. Then, the information obtained from NDT is used in the probabilistic capacity and demand models to estimate the seismic fragility of the bridge. As an illustration, the proposed probabilistic framework is applied to a reinforced concrete bridge with a one-column bent. The result of the illustration shows that the proposed framework can successfully provide the up-to-date structural properties and accurate fragility estimates.

Les modèles d'éléments finis des automobiles donnent des grandes tailles de problèmes matriciels, ce qui demeure coûteux en ressources numériques pour une procédure d'optimisation. La multiplicité des phénomènes couplés du problème d'interaction de l'air de l'habitacle et de la superstructure rend plus sensible, à des variations mineures des paramètres, une optimisation directe du véhicule. Pour réduire les temps de calculs et l'espace mémoire liés à la simulation numérique en éléments finis, une méthode de double synthèse modale est appliquée sur la structure et le fluide. Ceci permet de diminuer le nombre de degrés de liberté de frontière. Egalement, un algorithme a été développé pour minimiser le nombre d'évaluations de fonction au cours des itérations d'optimisation. L'approche modale permet également de décomposer le problème d'optimisation de la réponse vibro-acoustique par des sous-problèmes couplés d'optimisation de critères modaux. Ces critères modaux explicitent les couplages fréquentiels par des termes d'amplification et les couplages spatiaux par des paramètres effectifs modaux. Ils favorisent ainsi le développement d'une stratégie d'optimisation robuste par le contrôle modal des effets prépondérants sur la qualité vibro-acoustique des véhicules. / Finite element models and the complexity of vehicle passenger compartments make it harder the optimization, mainly because of expensive computing resources and multiple coupled phenomena of fluid-structure problems. Strategies to improve time and memory performance consist in the use of reduction methods, and combined with efficient optimization techniques, vibro-acoustic solutions of better quality can be performed. The complexity of the system is taken into account thanks to a hierarchical optimization process. Both reduction method and gradient-based optimization algorithm are investigated. Based on modal synthesis, special criteria help to determine critical vibration propagation paths. A modified SQP (Sequential Quadratic Programming) algorithm is also developed in order to provide a faster convergence speed. Such process is to be applied on an academic example and hollow parts and panels of a whole passenger compartments. It allows to find relevant and non obvious solutions by minimizing noise and vibration transfer functions in a relatively wide range of frequencies.