Many studies with Thermoelastic Stress Analysis (TSA) and Infrared Thermography, in Fiber Reinforced Polymeric materials (FRPs), are concerned with surface detection of "hot spots" in order to locate and infer damage. Such experimental analyses usually yield qualitative relations where correlations between stress state and damage severity cannot be obtained. This study introduces quantitative experimental methodologies for TSA and Digital Image Correlation to expand the use of remote sensing technologies for static behavior, static damage initiation detection, and fatigue damage in FRPs. Three major experimental studies are conducted and coupled with nonlinear anisotropic material modeling: static and TSA of hybrid bio-composite material systems, a new stochastic model for fatigue damage of FRPs, and fracture analysis for FRP single-lap joints. Experimental calibration techniques are developed to validate the proposed macromechanical and micromechanical nonlinear anisotropic modeling frameworks under multi-axial states of stress. The High Fidelity Generalized Method of Cells (HFGMC) is a sophisticated micromechanical model developed for analysis of multi-phase composites with nonlinear elastic and elastoplastic constituents is employed in this study to analyze hybrid bio-composites. Macro-mechanical nonlinear anisotropic models and a linear orthotropic model for fracture behavior using the Extended Finite Element method (XFEM) are also considered and compared with the HFGMC method. While micromechanical and FE results provide helpful results for correlating with quasi-static behavior, analyzing damage progression after damage initiation is not straightforward and involves severe energy dissipation, especially with increasing damage progression. This is especially true for fatigue damage evolution, such as that of composite joints as it is associated with uncertainty and randomness. Towards that goal, stochastic Markov Chain fatigue damage models are used to predict cumulative damage with the new damage indices proposed using full-field TSA image analysis algorithms developed for continuously acquired measurements during fatigue loading of S2-Glass/E733FR unidirectional single-lap joints. Static damage initiation is also investigated experimentally with TSA in single-lap joints with thick adherends providing for new design limitations. The computational modeling, stochastic and experimental methods developed in this study have a wide range of applications for static, fracture and fatigue damage of different FRP material and structural systems.

- Allgemeine Problematik bei der FEM-Simulation von Lagern/Gelenken - Modellierungstechniken in MECHANICA wie z.B. Kontakt-Analysen, Balkenspinnen, gewichtete Verbindungen, orthotropes Material, etc. - Vergleich dieser Techniken und Diskussion der Vor- und Nachteile anhand eines Gleitlagers - Empfehlungen und Regeln

The aim of this study is to develop an explicit analytical formulation based on the anisotropic elasticity theory that determines the behavior of fiber reinforced composite vessel under hygrothermal loading. The loading is studied for three cases separately, which are plane strain case, free ends and pressure vessel cases. For free-end and pressure vessel cases, the vessel is free to expand, on the other hand for plane strain case, the vessel is prevented to expand. Throughout the study, constant, linear and parabolic temperature distributions are investigated and for each distribution, separate equations are developed. Then, a suitable failure theory is applied to investigate the behavior of fiber reinforced composite vessels under the thermal and moisture effects. Throughout the study, two computer programs are developed which makes possible to investigate the behavior of both symmetrically and antisymmetrically oriented layers. The first program is developed for plane strain case, where the second one is for pressure vessel and free-end cases. Finally, several thermal loading conditions have been carried out by changing the moisture concentration and temperature distributions and the results are tabulated for comparison purposes.

In this Thesis, the problems of the Free Flexural (or Bending) Vibrations of Composite, Orthotropic Plates and/or Panels with Various Bonded Joints are formulated and investigated in detail. The composite bonded plate system is composed of Plate Adherends adhesively bonded by relatively very thin adhesive layers. The general problem is considered in terms of the three Main PROBLEMS, namely Main PROBLEM I, Main PROBLEM II and Main PROBLEM III. The theoretical formulation of the Main PROBLEMS is based on Mindlin Plate Theory which is a First Order Shear Deformation Plate Theory (FSDPT). Thus, the transverse shear deformations, the transverse and the rotary moments of inertia of the plates are included in the formulation. Very thin, elastic deformable adhesive layers are considered as continua with transverse normal and shear stresses. The damping effects in the plates and the adhesive layers are neglected. The entire composite bonded joint assembly is assumed to be simple supported along the two opposite edges, so that the Classical Levy&amp / #8217 / s Solutions can be applied in this direction. The dynamic equations of the Bonded Joint System which combines together the Mindlin Plate dynamic equations with the adhesive layer equations are reduced to a system of First Order Ordinary Differential Equations in the state vector form. This special form of the Governing System of the First Order Ordinary Differential Equations are numerically integrated by means of the Modified Transfer Matrix Method which is a combination of the Classical Levy&amp / #8217 / s Method, the Transfer Matrix Method and the Integrating Matrix Method (with Interpolation Polynomials and/or Chebyshev Polynomials). The Main PROBLEMS are investigated and presented in terms of the mode shapes and the corresponding natural frequencies for various sets of boundary conditions. The significant effects of the hard and the soft adhesive layer elastic constants on the mode shapes and on the natural frequencies are demonstrated. Some important parametric studies such as the influences of the Joint Length Ratio, the Joint Position Ratio, the Bending Stiffness Ratio, etc. on the natural frequencies are computed and plotted for the hard and soft adhesive cases for several support conditions.

Composites are materials which cater to the present and future needs of many demanding industries, such as aerospace, as they are weight-sensitive for a given requirement of strength and stiff ness, corrosion resistant, potentially multi-functional and can be tailored according to the application. However, they are in particular difficult to join as they cannot be easily machined, without introducing damages which can eventually grow. Any structure is as strong as its weakest joint. Most of the joints belong to the category of mechanically-fastened joints and they pose enormous challenges in modeling due to contact phenomena, nonlinearity and stress concentration factors. It is therefore a necessity to construct an efficient model that would include all the relevant contact phenomena in the joints, as it has been pointed out in literature that damage typically initiates near the joint holes. The focus of this work is to describe the construction of an asymptotically-correct model using the Variational Asymptotic Method (VAM). Amongst its many potential applications, VAM is a well-established analytical tool for obtaining the stress and strain fields for beams and shells. The methodology takes advantage of the small parameter that is inherent in the problem, such as the ratio of certain characteristic dimensions of the structure. In shells and beams, VAM takes advantage of the dimension-based small parameter(s), thereby splitting the problem into 2-D + 1-D (for beams) and 1-D + 2-D (for shells), in turn offering very high computational efficiency with very little loss of accuracy compared to dimensionally unreduced 3-D models. In this work, the applicability of VAM is extended to two-dimensional (2-D) and three-dimensional (3-D) frictionless contact problems. Since a generalised VAM model for contact has not been pursued before, the `phantom0 step is adopted for both 2-D and 3-D models. The development of the present work starts with the construction of a 2-D model involving a large rectangular plate being pressed against a rigid frictionless pin. The differential equations governing the problem and the associated boundary conditions are obtained by minimizing the reduced strain energy, augmented with the appropriate gap function, by using a penalty method. The model is developed for both isotropic and orthotropic cases. The boundary value problem is solved numerically and the displacement field obtained is compared with the one obtained using commercial software (ABAQUSr) for validation at critical regions such as the contact surfaces. Banking on the validation of the 2-D model, a 3-D model with a pin and a finite annular cylinder was constructed. The strain energy for the finite cylinder was derived using geometrically exact 3-D kinematics and VAM was applied leading to the reduction in the strain energy for isotropic and orthotropic materials in rectangular and cylindrical co-ordinates. As in the 2-D case, the reduced strain energy, subject to the inequality constraint of the gap function, is minimized with respect to the displacement field and the corresponding boundary value problem is solved numerically. The displacements of the contact surface and the top surface of the annular cylinder are compared with those from ABAQUS and thus validated. The displacement fields obtained using the current 2-D and 3-D models show very good agreement with those from commercial finite element software packages. The model could be re ned further by using the gap function derived in this work and applying it to a plate model based on VAM, which could be explored in the future.

Métodos convencionais, tais como o método da chapa ortotrópica, têm sido aplicados por muitos anos no estudo de painéis reforçados pela sua simplicidade e facilidade de aplicação na determinação de tensões agentes nas fases iniciais da espiral projeto. Não estão disponíveis na literatura, porém, análises comparativas do método da chapa ortotrópica com procedimentos numéricos utilizando elementos finitos (MEF) que permitam a determinação da acurácia ou da ordem de grandeza dos desvios inerentes à aplicação desta metodologia. O presente trabalho apresenta análises comparativas entre estas duas metodologias na solução de painéis reforçados submetidos a carga lateral uniforme, tipicamente aplicados a estruturas navais (chapa em apenas um dos lados com reforçadores em T). Com este objetivo foram construídos modelos de painéis simplesmente apoiados e engastados (modelagem com elementos de viga e casca) com diferentes espaçamentos e diferentes inércias de reforçadores, configurando uma ampla matriz de análise paramétrica. Os resultados de deflexões e tensões nas vigas e chapas obtidos dos modelos MEF foram parametrizados em função das variáveis da chapa ortotrópica (razão de aspecto virtual), (coeficiente de torção) e K (parâmetro adimensional de tensões e de deflexão). Esta parametrização permite gerar curvas numéricas de tensão e deflexão dos modelos em estudo. As curvas numéricas assim geradas são comparadas com as curvas propostas pelo método da chapa ortotrópica para painéis reforçados simplesmente apoiados, de tal maneira que sua comparação permita, além de determinar a sensibilidade dos resultados numéricos em função das mudanças de inércia e espaçamento entre reforçadores, aferir o nível de desvio oriundo do uso da metodologia da chapa ortotrópica em relação ao método dos elementos finitos. Resultados mostram que as curvas derivadas da metodologia da chapa ortotrópica fornecem bons resultados para as deflexões e tensões transversais nas vigas no centro do painel reforçado. Para as tensões longitudinais nas vigas, uma curva corrigida de tensões longitudinais máximas é fornecida. No caso das curvas de tensões longitudinais e transversais na chapa, as curvas da chapa ortotrópica fornecem valores conservadores de tensão no centro do painel em relação aos valores obtidos dos modelos MEF. Adicionalmente, uma vez que o método da chapa ortotrópica só fornece curvas para chapa sem reforçadores no caso de condição de engaste, curvas numéricas das diferentes variáveis são fornecidas para esta condição. Analogias são feitas com a solução fornecida pelo método da chapa ortotrópica para painéis reforçados com razão de aspecto =, borda longitudinal engastada e borda transversal apoiada. Adicionalmente, resultados analíticos baseados na teoria de grelhas são comparados com os valores fornecidos pelas curvas numéricas para painel engastado obtendo-se resultados consistentes. Com esta análise foi possível determinar a aplicabilidade e limitações do método da chapa ortotrópica no estudo de painéis reforçados simplesmente apoiados. O estudo também fornece novas curvas numéricas para painéis reforçados engastados. / Conventional methods, such as the orthotropic plate, have been applied for many years in the study of stiffened plates to obtain the stresses acting on the structure in the early stages of the structural design, because of its simplicity and easy application. However, comparative analyses of the orthotropic plate method with numerical methods using finite element analyses (FEM) to determine its accuracy or inherent errors are not available in the literature. This study presents comparative analyses between the solutions of the two methodologies for reinforced panels subjected to lateral uniform load, typically applied to marine structures (plate only on one side with T beams). Models of reinforced panels were implemented for a simply supported and clamped boundary conditions with different spacing between stiffeners and different stiffeners`s inertia, setting up a broad array of parametric analysis. The deflections and stresses in beams and plate derived from the MEF analyses were parameterized as function of the orthotropic plate parameters: (virtual aspect ratio), (torsion coefficient) and K (dimensionless parameter of stress and deflection). This enables the generation of parametric numerical curves of stresses and deflections for the models under study. The numerical curves generated in this way were compared with the analytic curves proposed by the orthotropic plate theory for reinforced panels with simply supported boundary conditions. The comparisons allow, in addition to a sensitivity analysis of the numerical curves as a function of inertia and spacing between stiffeners, the assessment of inherent deviation for the orthotropic plate theory when compared with the finite element analyses. From the comparative analyses, it is possible to conclude that the curves proposed for the orthotropic theory for deflection and stresses of the transverse beam at the center of the reinforced panels have a good correlation with the numeric curves and provide accurate results. For the stresses on longitudinal beam, a revised curved for maximum stresses is provided. For the curves of plate stresses in the longitudinal and transverse directions at the center of the panels, the orthotropic plate theory provides conservative values when compared with the values of FEM models. The orthotropic plate method only provides curves for unstiffened plate under clamped boundary condition. Numerical curves for reinforced panels with clamped boundary condition are provided. Analogies are made between the solution provided by the orthotropic theory for a reinforced panel with an infinite virtual aspect ratio = , longitudinal edges clamped and transverse edges simply supported. Additionally, analytical results based on grillage theory were compared with the values provided by the numerical curves for clamped reinforced panels, obtaining a consistent results and a good correlation. This analysis provides a critic overview of the applicability and limitations of the orthotropic plate method for the analyses of reinforced panels with simply supported boundary condition. The study also provides new numerical solutions for reinforced panels with clamped boundary condition.

Made available in DSpace on 2016-12-23T14:37:42Z (GMT). No. of bitstreams: 1 Abraao Carlos Verdin Filho.pdf: 1208785 bytes, checksum: 69e6d8f8181c12a985a8ff178be50374 (MD5) Previous issue date: 2011-03-29 / Proper management through the use of new techniques has been paramount to the success of good productivity in the coffee culture conilon. Aiming to improve the information management now in place for the culture, related to the spacing of plants in the area and number of stems per plant per hectare, was introduced on an experimental research farm of C. canephora variety clonal Emcapa 8111, early maturity, at the Experimental Farm of Marilândia (FEM), Incaper Marilândia localized in the northern region of Espirito Santo. It was evaluated the yield and grading of grains through sieves in post-harvest. The statistical scheme used for both evaluations was a DBC with 12 treatments with four replications. The experimental plot was composed of eleven plants, all considered useful, where the data portion corresponded to the average evaluation of eleven plants. In the productivity study was carried out a joint analysis for the years 2008 and 2010, and individual analysis for the year 2009. Regarding the study of classification of grains through sieves in post-harvest, analyzed only in the year 2010, we used a split-plot analysis, and the plots had 12 treatments and for subplots by 7 ratings sieves (billy 12, and 11 and 10; boring 17, 15 and 13, and fundagem). The treatments were set taking into account the planting space and number of stems per plant specifications as follows: treatment 1 - 2.0 x 1.0 m spacing, with two stems per plant, producing 10,000 stems / ha; treatment 2 - 2.0 x 1.0 m spacing, with three stem per plant, producing 15,000 stems / ha; treatment 3 - 2.0 x 1.0 m spacing, with four stems per plant, producing 20,000 stems / ha; treatment 4 - 2.5 x 1.0 m spacing, with two stems per plant, resulting in 8000 stems / ha; treatment 5 - 2.5 x 1.0 m spacing, with three stem per plant, resulting in 12,000 stems / ha; treatment 6 - 2.5 x 1.0 m spacing, with four stems per plant, producing 16,000 stems / ha; treatment 7 - 3.0 x 1.0 m spacing, with three stem per plant, producing 10,000 stems / ha; treatment 8 - 3.0 x 1.0 m spacing with five stems per plant, producing 16,667 stems / ha; treatment 9 - 3.0 x 1.0 m spacing, with six stems per plant, originated 20,000 stems / ha; treatment 10 - 3.0 x 1.5 m spacing, with three stems per plant, resulting in 6667 stems / ha; treatment 11 - 3.0 x 1.5 m spacing with five stems per plant, producing 11,111 stems / ha; and treatment 12 - 3.0 x 1.5 m spacing, with six stems per plant, producing 13,333 stems per hectare. In the analysis of variance and comparisons between means studied, there are significant differences in most cases. Reducing the spacing associated with an increased number of stems per plant has a positive effect on productivity of coffee conilon cultivar Emcapa 8111, until the density of 20,000 stems per hectare, under the conditions studied. Under the conditions evaluated, in general, there is an increase in productivity, when it increases the number of stems within a plant density. The best treatment for the production, evaluated in the years 2008 and 2010 is spacing 2.0 x 1.0 conducted with 20,000 stems per hectare. There are differences in the classification screens at the different densities of plants and stems per hectare. In general, treatments with low plant density (3.0 m spacing between rows and 1.0 to 1.5 m between plants) show higher percentages of fundagem between the classification of sieves studied / O manejo adequado através do uso de novas técnicas tem sido de suma importância para o sucesso de boas produtividades na cultura do café conilon. Com o objetivo de aprimorar as indicações de manejo ora existentes para a cultura, relacionadas ao espaçamento das plantas na área e ao número de hastes por planta por hectare, foi introduzido um experimento de pesquisa em lavoura de C. canephora, variedade clonal Emcapa 8111, de maturação precoce, na Fazenda Experimental de Marilândia (FEM), Incaper localizada no município de Marilândia, região Norte do Estado do Espírito Santo. Foram avaliadas as produtividades e classificação de grãos por peneiras na pós-colheita. O esquema estatístico utilizado para ambas as avaliações foi num DBC, com 12 tratamentos com 4 repetições. A parcela experimental foi composta por onze plantas, todas consideradas úteis, onde o dado da parcela correspondia à média da avaliação das onze plantas. No estudo da produtividade foi efetuada uma análise conjunta para os anos 2008 e 2010, e análise individual para o ano de 2009. Quanto ao estudo da classificação de grãos por peneiras na pós-colheita, analisado somente no ano de 2010, foi utilizada a análise de parcelas subdivididas, sendo que as parcelas foram compostas pelos 12 tratamentos e as subparcelas pelas 7 classificações de peneiras (mocas 12, 11 e 10; chatos 17, 15 e 13; e fundagem). Os tratamentos foram montados levando em consideração o espaçamento de plantio e o número de hastes por plantas, conforme especificações a seguir: tratamento 1 espaçamento 2,0 x 1,0 m, com duas hastes por planta, originando 10.000 hastes/ha; tratamento 2 espaçamento 2,0 x 1,0 m, com três hastes por planta, originando 15.000 hastes/ha; tratamento 3 espaçamento 2,0 x 1,0 m, com quatro hastes por planta, originando 20.000 hastes /ha; tratamento 4 espaçamento 2,5 x 1,0 m, com duas hastes por planta, originando 8.000 hastes /ha; tratamento 5 espaçamento 2,5 x 1,0 m, com três hastes por plantas, originando 12.000 hastes/ha; tratamento 6 espaçamento 2,5 x 1,0 m, com quatro hastes por planta, originando 16.000 hastes/ha; tratamento 7 espaçamento 3,0 x 1,0 m, com três hastes por planta, originando 10.000 hastes/ha; tratamento 8 espaçamento 3,0 x 1,0 m, com cinco hastes por planta, originando 16.667 hastes/ha;tratamento 9 espaçamento 3,0 x 1,0 m, com seis hastes por planta, originado 20.000 hastes/ha; tratamento 10 espaçamento 3,0 x 1,5 m, com três hastes por planta, originando 6.667 hastes/ha; tratamento 11 espaçamento 3,0 x 1,5 m, com cinco hastes por planta, originando 11.111 hastes/ha; e tratamento 12 espaçamento 3,0 x 1,5 m, com seis hastes por planta, originando 13.333 hastes por hectare. Nas análises de variância e comparação entre médias estudadas, verificaram-se diferenças significativas na maioria dos casos. A redução do espaçamento associado ao aumento do número de hastes por planta tem efeito positivo na produtividade do café conilon, cultivar Emcapa 8111, até a densidade de 20.000 hastes por hectare, nas condições estudadas. Nas condições avaliadas, em geral, há aumento na produtividade, quando se aumenta o número de hastes dentro de uma mesma densidade de plantas. O tratamento mais adequado à produção, avaliado nos anos de 2008 e 2010 e o espaçamento 2,0 x 1,0 conduzido com 20.000 hastes por hectare. Há diferenças quanto à classificação por peneiras nas diferentes densidades de plantas e hastes por hectare. Em geral, os tratamentos com menor densidade de plantas (espaçamentos 3,0 x 1,0 m e 3,0 x 1,5 m) apresentam maior porcentagem de fundagem entre as classificações de peneiras estudadas

O Brasil, um país de dimensões continentais e inúmeros rios em seu território, tem uma enorme carência de pontes para pequenos e médios vãos. Atualmente, a elevada produção de madeira de reflorestamento traz à engenharia uma opção de material renovável e ecologicamente correto para se utilizar em prol do ser humano nas obras de infraestrutura, mais especificamente na construção de pontes, uma vez que a madeira possui ótimo desempenho quando sujeita a cargas de curta duração. A utilização de Madeira Laminada Colada (MLC) permite a fabricação de peças de grandes dimensões, com excelente estabilidade, e com elevado controle de qualidade, proporcionando segurança e confiabilidade à estrutura. Este trabalho tem como objetivo o estudo teórico de pontes em vigas de MLC e tabuleiros em painéis de MLC utilizando procedimentos desenvolvidos por pesquisadores internacionalmente renomados e adequando estes procedimentos para as normas e condições brasileiras. Apresentam-se neste trabalho tabelas para pré-dimensionamento deste tipo de ponte para vãos de 8m a 20m para TB-30 e TB-45. Além disso, analisam-se numérica e experimentalmente os ganhos de rigidez e redução de deslocamentos relativos proporcionados pela utilização de vigas de enrijecimento, bem como a influência da seção destas vigas e a quantidade de parafusos utilizados para interligar ao tabuleiro. Por fim, apresentam-se algumas disposições construtivas não estruturais fundamentais para o bom funcionamento deste tipo de estrutura. / Brazil, a continental size country and with many rivers in its territory, has a huge necessity for short and medium spans bridges. Nowadays, the high production of reforestation wood allows to engineering an option of renewable material and ecologically correct to be used in favor to humans in infrastructure works, most specifically in bridges construction, once timber has a great performance in quick loads. The use of Glued-Laminated Timber (Glulam) enable the manufacture of large sizes members, with excellent stability, and a high control of quality in its production, providing a safety and reliability structure. This work aim the theoretical study of glulam stringers bridge and glulam panels deck using procedures developed by international researchers and adapting those procedures for Brazilian conditions and normalizations. Pre-design tables are presented in this work for 8m to 20m spans for TB-30 and TB-45. In addition to, analyses numerical and experimentally the gain of stiffness and the reduction of relative displacements between panels provided by the use of stiffener beams, besides the influence of those beams section size and number of bolts used for deck interconnection. Finally, it is presented some not structural constructive arrangements essential for good performance of this kind of structure.

O tabuleiro laminado protendido é uma composição de peças interligadas por uma tensão de compressão aplicada perpendicularmente às lâminas de madeira. Para aplicar tal tensão, são usados elementos de aço tensionados e a cordoalha engraxada aparece no cenário atual como um material econômico para tabuleiros laminados protendidos, em substituição às barras dywidag, que são comumente utilizadas neste tipo de estrutura em outros países. O funcionamento mecânico do tabuleiro é caracterizado como uma placa ortotrópica, e este trabalho visou determinar as relações entre as constantes elásticas do tabuleiro com a utilização da cordoalha engraxada. Foram analisadas também as perdas de protensão ocorridas ao longo do tempo, causadas pela fluência e relaxação dos materiais e, a partir dos resultados, foi proposta uma equação para determinação teórica da perda de protensão nos tabuleiros. O estudo experimental foi realizado em um modelo em escala reduzida construído no LaMEM (Laboratório de Madeiras e Estruturas de Madeira) para determinar as constantes elásticas e as perdas de protensão. Ao fim do trabalho são apresentados os resultados e comparações com estudos anteriores; também foi confeccionado um Manual de Boa Execução para Tabuleiros Laminados Protendidos de Madeira com Cordoalha engraxada. / The stress-laminated bridge deck is a composition of parts, interconnected by compression stress, perpendicularly applied to the wood laminate. In order to apply such stress, stressed steel elements are used, as well as the greased wire rope, which comes along in the current scenario as an economic material for stress-laminated bridge decks, replacing the dywidag bars, commonly used in this sort of structure in other countries. This bridge deck mechanic function is characterized as an orthotropic plate, and this paper aimed at determining the relation among the decks elastic constants with the greased wire rope. Another analyzed point is the stress loss occurred through time, caused by material creep and relaxation, and, through the results, an equation for the decks stress loss theoretical determination was proposed. The experimental study was developed in a reduced scale built in LaMEM (Laboratório de Madeiras e Estruturas de Madeira) to determine the elastic constants and the stress loss. The results and comparisons with previous studies are presented by the end of this paper; In addition, a handbook of Proper Execution for Stress-laminated Wood Bridge Decks with Greased Wire Rope was developed.

Les tabliers métalliques à dalle orthotrope sont sensibles au phénomène de fatigue produit par les charges des poids lourds du trafic. Ce comportement n'est pas précisément prédit avec les méthodes de l'Eurocode 3, compte tenu de la complexité des effets locaux et de la connaissance insuffisante du rôle mécanique du revêtement (diffusion des charges et participation à la flexion locale). De plus l'augmentation du trafic des camions et éventuellement celle des charges admissibles par essieu en Europe tend à rendre ce problème bien plus critique. Le renforcement de ces tabliers est donc souhaitable de façon à prolonger la durée de vie des ponts existants, et aussi augmenter la durabilité des nouveaux ponts. Le béton fibré à ultra hautes performances (BFUP) a été envisagé comme nouvelle solution de revêtement, étant donné ses propriétés mécaniques, ses possibilités de mise en œuvre et sa durabilité. L'objectif de cette thèse, réalisée dans le cadre du projet ANR Orthoplus, est de quantifier expérimentalement l'apport des revêtements couramment utilisés dans les structures à dalle orthotrope et de valider la solution innovante en BFUP. Des essais statiques et dynamiques sur corps d'épreuve à grande échelle (2,40x4,00) m2 ont été réalisés sur la plate-forme d'essai des structures de l'IFSTTAR. Quatre corps d'épreuve ont été testés : tôle de platelage de 14 mm non revêtue et revêtue de 80 mm de béton bitumineux, tôle de 10 mm revêtue de 35 mm de BFUP et tôle de 12 mm revêtue de 35 mm de BFUP. L'influence des différents types de chargement positionnés au centre des corps d'épreuve a été analysée : plaques métalliques type Eurocode 1 et vraies roues de camion. L'étude a porté sur le détail de fatigue: liaison auget-tôle de platelage entre pièces de pont. La contrainte géométrique de fatigue (extrapolation au point chaud) a été évaluée expérimentalement en utilisant deux schémas d'extrapolation linéaire des déformations à proximité du cordon de soudure du détail étudié, le schéma du rapport CECA et celui proposé par l'Institut International de Soudure, à partir des mesures réalisées au-dessous de la tôle de platelage (σT) et sur l'âme de l'auget (σA).La cohérence entre estimation quasi-statique des déformations et comportement sous cycles de fatigue a été vérifiée, ainsi que la rigidification importante apportée par le BFUP, bien que ce dernier ne participe pas avec une connexion totale. Les résultats expérimentaux ont été confrontés à des modèles de différents niveaux de complexité qu'il reste nécessaire de calibrer empiriquement pour prévoir les contraintes géométriques. A partir des contraintes de fatigue obtenues expérimentalement, nous avons calculé la durée de vie des dalles orthotropes testés à l'aide de la règle du cumul linéaire de l'endommagement. Enfin nous avons mené une étude par analyse de cycle de vie d'un pont à dalle orthotrope pour vérifier la pertinence environnementale des différentes solutions de revêtement. Les nombreuses données expérimentales acquises dans ce travail sont de nature à permettre une amélioration significative du dimensionnement rationnel des tabliers à dalle orthotrope et de leur revêtement pour une meilleure prise en compte de leur gestion durable / Orthotropic steel bridge decks are sensitive to fatigue damage induced by live heavy traffic loads. This behaviour is not precisely predicted by Eurocode 3, because of the complexity of local effects. The pavement overlay is not taken into account for calculating the fatigue resistance because of the lack of knowledge concerning its mechanical behaviour (loads diffusion and participation in the local deflection) and the behaviour of the composite structure. Moreover, the increase in heavy traffic and potential regulations evolution in Europe – towards an increase of acceptable loads of truck axles - tend to render the orthotropic decks fatigue behaviour an even more critical issue. The reinforcement of these steel decks is therefore crucial to extend the service life of existing bridges, and also increase the durability of new bridges. Ultra-high performance fibre-reinforced concrete (UHPFRC) has been chosen as a possible alternative topping layer considering its remarkable durability, flowability and mechanical properties. The purpose of this thesis, carried out within the framework of a joint R&D project called Orthoplus, is to quantify experimentally the mechanical contribution of topping layers currently used in orthotropic steel bridge decks and validate an alternative concept using UHPFRC coating. Static and dynamic tests of large scale panels (2,40x4,00) m2 were carried out at the IFSTTAR Structures Laboratory. Four prototypes have been tested: a 14 mm thick deck plate without surfacing, the same deck plate associated with 80 mm of bituminous concrete surfacing, a 10 mm thick deck plate topped with 35 mm of UHPFRC and a panel with the same UHPFRC topping layer and a 12 mm thick deck plate. The influence of different centered load types and configurations has been analyzed: rectangular steel plates according to Eurocode 1 and real truck wheels. The experimental programme has been focused on the rib-to-deck welded joints at mid-span between two transverse crossbeams. The fatigue geometrical stresses in the deck and the trough, respectively denoted as σD and σT, have been derived from two linear extrapolations of measured strains next to the toe of the welded joint: the extrapolation schemes from the ECSC report and from the IIW document. Consistency between quasi static strains and deflections estimate and behaviour under fatigue cycles has been verified, as well as the significant additional stiffness provided by the UHPFRC overlay, although its contribution does not correspond to a perfectly connected composite section. The experimental results have been compared to simple and more complex models which still need empirical calibration for predicting the geometrical stresses. Using the experimentally obtained fatigue geometrical stresses the service life of the tested prototypes were calculated using Miner's rule. Finally a life cycle assessment study of an orthotropic steel bridge deck was carried out to verify the environmental relevance of the alternative topping layer solutions. The numerous experimental data obtained from this work shall make it possible to significantly improve the rational design method of orthotropic slabs and their associated deck overlay, in view of a better accounting of their long term and sustainable structural management