Composites have been increasingly used in different applications in the last decade, especially in aerospace due to their high strength and lightweight characteristics. Indeed, the latest models of Airbus (A350) and Boeing (B787) have employed more than 50 wt% of composites, mainly Carbon Fibre Reinforced Polymers (CFRP). Yet, the increased use of CFRP has raised the environmental concerns about their end-of-life related to waste disposal, consumption of non-renewable resources for manufacturing and the need to recycle CFRP wastes. In this study, a generic model is developed in order to propose an optimal management of aerospace CFRP wastes taking into account economic and environmental objectives. Firstly, a life-cycle systemic approach is used to model the environmental impacts of CFRP recycling processes focusing on Global Warming Potential (GWP) following the guidelines of Life Cycle Assessment (LCA). The whole supply chain for recycling CFRP pathways is then modelled from aircraft dismantling sites to the reuse of recycled fibres in various applications. A multi-objective optimisation strategy based on mathematical programming, -constraint and lexicographic methods with appropriate decisionmaking techniques (M-TOPSIS, PROMETHEE-GAIA) has been developed to determine CFRP waste supply chain configurations. Various scenarios have been studied in order to take account the potential of existing recycling sites in a mono-period visions as well as the deployment of new sites in a multi-period approach considering the case study of France for illustration purpose. The solutions obtained from optimisation process allow developing optimal strategies for the implementation of CFRP recovery with recycled fibres (of acceptable quality) for the targeted substitution use while minimising cost /maximising profit for an economic criterion and minimising an environmental impact based on GWP.

<p>Composite materials have been successfully applied in various industries, such as aerospace, automobile, and wind turbines, etc. Although the material properties of composites are desirable, the behaviors of composites are complicated. Many efforts have been made to model the constitutive behavior and failure of composites, but a complete and validated methodology has not been completely achieved yet. Recently, machine learning techniques have attracted many researchers from the mechanics field, who are seeking to construct surrogate models with machine learning, such as deep neural networks (DNN), to improve the computational speed or employ machine learning to discover unknown governing laws to improve the accuracy. Currently, the majority of studies mainly focus on improving computational speed. Few works focus on applying machine learning to discover unknown governing laws from experimental data.  In this study, we will demonstrate the implementation of machine learning to discover unknown governing laws of composites. Additionally, we will also present an application of machine learning to accelerate the design optimization of a composite rotor blade.</p> <p><br></p> <p>To enable the machine learning model to discover constitutive laws directly from experimental data, we proposed a framework to couple finite element (FE) with DNN to form a fully coupled mechanics system FE-DNN. The proposed framework enables data communication between FE and DNN, which takes advantage of the powerful learning ability of DNN and the versatile problem-solving ability of FE. To implement the framework to composites, we introduced positive definite deep neural network (PDNN) to the framework to form FE-PDNN, which solves the convergence robustness issue of learning the constitutive law of a severely damaged material. In addition, the lamination theory is introduced to the FE-PDNN mechanics system to enable FE-PDNN to discover the lamina constitutive law based on the structural level responses.</p> <p><br></p> <p>We also developed a framework that combines sparse regression with compressed sensing, which leveraging advances in sparsity techniques and machine learning, to discover the failure criterion of composites from experimental data. One advantage of the proposed approach is that this framework does not need Bigdata to train the model. This feature satisfies the current failure data size constraint. Unlike the traditional curve fitting techniques, which results in a solution with nonzero coefficients in all the candidate functions. This framework can identify the most significant features that govern the dataset. Besides, we have conducted a comparison between sparse regression and DNN to show the superiority of sparse regression under limited dataset. Additionally, we used an optimization approach to enforce a constraint to the discovered criterion so that the predicted data to be more conservative than the experimental data. This modification can yield a conservative failure criterion to satisfy the design needs.</p> <p><br></p> <p>Finally, we demonstrated employing machine learning to accelerate the planform design of a composite rotor blade with strength consideration. The composite rotor blade planform design focuses on optimizing planform parameters to achieve higher performance. However, the strength of the material is rarely considered in the planform design, as the physic-based strength analysis is expensive since millions of load cases can be accumulated during the optimization. Ignoring strength analysis may result in the blade working in an unsafe or low safety factor region since composite materials are anisotropic and susceptible to failure. To reduce the computational cost of the blade cross-section strength analysis, we proposed to construct a surrogate model using the artificial neural network (ANN) for beam level failure criterion to replace the physics-based strength analysis. The surrogate model is constructed based on the Timoshenko beam model, where the mapping is between blade loads and the strength ratios of the cross-section. The results showed that the surrogate model constraint using machine learning can achieve the same accuracy as the physics-based simulation while the computing time is significantly reduced. </p>

A hybrid data-physics driven reduced-order homogenization (dpROH) approach aimed at improving the accuracy of the physics-based reduced order homogenization (pROH), but retain its unique characteristics, such as interpretability and extrapolation, has been developed. The salient feature of the dpROH is that the data generated by a high-fidelity model based on the direct numerical simulations with periodic boundary conditions improve markedly the accuracy of the physic-based model reduction. The dpROH consist of the offline and online stages. In the offline stage, dpROH utilized surrogate-based Bayesian Inference to extract crucial information at the representative volume element (RVE) level. With the inferred data, online predictions are performed using a data-enhanced reduced order homogenization. The proposed method combines the benefits of physics-based reduced order homogenization and data-driven surrogate modeling, striking a balance between accuracy, computational efficiency, and physical interpretability. The dpROH method, as suggested, has the versatility to be utilized across different RVE geometries (including fibrous and woven structures) and various constitutive models, including elasto-plasticity and continuum damage models. Through numerical examples that involve comparisons between different variants of dpROH, pROH, and the reference solution, the method showcases enhanced accuracy and efficiency, validating its effectiveness for a wide range of applications. A novel pseudo-nonlocal eight-node fully integrated linear hexahedral element, PN3D8, has been developed to accelerate the computational efficiency of multiscale modeling for complex material systems. This element is specifically designed to facilitate finite element analysis of computationally demanding material models, enabling faster and more efficient simulations within the scope of multiscale modeling. The salient feature of the PN3D8 is that it employs reduced integration for stress updates but full integration for element matrices (residual and its consistent tangent stiffness). This is accomplished by defining pseudo-nonlocal and local stress measures. Only the pseudo-nonlocal stress is updated for a given value of mean strain or mean deformation measure for large deformation problems. The local stress is then post-processed at full integration points for evaluation of the internal force and consistent tangent stiffness matrices. The resulting tangent stiffness matrix has a symmetric canonical structure with an identical instantaneous constitutive matrix at all quadrature points of an element. For linear elasticity problems, the formulation of the PN3D8 finite element coincides with the classical eight-node fully integrated linear hexahedral element. The procedure is illustrated for small and large deformation two-scale quasistatic problems.

Les recherches actuelles dans le domaine des composites montrent l'utilisation croissante de matrices biodégradables et/ou de renforts fibreux naturels issus de ressources renouvelables. Néanmoins, une étape de compatibilisation fibres/matrice est très souvent nécessaire. Dans cette thèse, nous avons exploré une méthode physique : le traitement corona. Son impact sur les propriétés physicochimiques de différentes fibres a été étudié par XPS, mesure d'angle de contact et MEB. Il ressort qu'il entraîne principalement une oxydation de surface et une augmentation de sa rugosité. Nous avons évalué le comportement mécanique de composites, préparés par extrusion à partir de fibres de chanvre ou miscanthus et de matrices polypropylène (PP) ou acide polylactique (PLA). L'incorporation de renforts accroît la rigidité des matériaux et le transfert de contrainte et leur traitement permet d'obtenir des caractéristiques encore supérieures en raison d'un ancrage mécanique accru. Les valeurs optimales sont obtenues pour un taux massique de fibres de l'ordre de 20%. Les propriétés thermiques et thermomécaniques des composites ont été caractérisées par ATG, DMA et DSC. La stabilité thermique des matériaux est abaissée après incorporation de renforts bruts mais très largement améliorée (+ 15-20°C) après traitement des fibres. Dans certaines conditions, les fibres agissent comme des agents nucléants qui influent sur le processus de cristallisation et le taux de cristallinité. Le traitement des fibres par corona ne permet pas de retarder la dégradation des matériaux au cours d'un vieillissement accéléré en milieu humide et l'évolution des propriétés thermomécaniques est plus prononcée pour les matériaux à base de PLA plus hydrophile / The field of composites materials shows increasing use of biodegradable matrices and / or natural reinforcements from renewable resources. Nevertheless, a compatibilization step between fiber and matrix is necessary. In this PhD study, we have explored a physical method: corona treatment. Its impact on the physicochemical properties of different fibres has been studied by XPS, contact angle measurement and SEM. It appears that it mainly involves surface oxidation and roughness increase. We have also evaluated the mechanical behaviour of composites, prepared by extrusion from hemp or miscanthus fibres and polypropylene (PP) or polylactic acid (PLA) matrices. The incorporation of raw reinforcements increases the stiffness and the stress transfer. Composites based on treated fibres show better mechanical performances, resulting from an enhanced mechanical anchorage. The optimum values are obtained for 20% (wt) fibres content. The thermal and thermomechanical properties of composites have been characterized by TGA, DMA and DSC. The thermal stability of materials is reduced after incorporation of raw reinforcements but very much improved (+ 15 - 20 ° C) after treatment of fibres. Under certain conditions, the fibres act as nucleating agents that affect the crystallization process and crystallinity rate. The corona treatment of fibres does not delay the degradation of materials during an accelerated aging in humid environment and the evolution of the thermomechanical properties is more pronounced for PLA based materials because of its hydrophilic character.

[pt] Este estudo teórico-experimental analisa o comportamento até a ruptura de vigas de concreto armado reforçadas externamente à torção com compósitos de fibras de carbono (CFC). No programa experimental, sete vigas de concreto armado, com seção transversal de 20 cm x 40 cm e 420 cm de comprimento, com mesma armadura de aço longitudinal e transversal e concreto com mesma resistência à compressão, foram ensaiadas até a ruptura. As vigas testadas foram divididas em três séries, sendo uma viga de referência sem reforço, três vigas com reforço transversal externo e três vigas com reforço externo transversal e longitudinal. Para a realização dos ensaios foi montada uma estrutura auxiliar de aço capaz de transferir às vigas a solicitação de torção pura. No estudo teórico foram desenvolvidas duas formulações. A primeira formulação, baseada no modelo da treliça espacial generalizada com abrandamento de tensões, apresenta uma sistemática para traçado da curva momento torçor x ângulo de torção por unidade de comprimento de vigas de concreto armado reforçadas à torção. A segunda formulação, fundamentada no modelo da Analogia da Treliça Espacial de acordo com a filosofia de dimensionamento do Eurocode 2, apresenta uma sistemática para dimensionamento de reforço com CFC . As duas metodologias adotam um modelo para determinação da aderência entre o substrato de concreto e o reforço. A inclusão da aderência nos modelos desenvolvidos é de grande importância porque em geral a ruptura do elemento estrutural ocorre devido ao descolamento do CFC. Os resultados experimentais obtidos nos testes das vigas foram utilizados para validar as duas formulações teóricas desenvolvidas. Os resultados experimentais apresentaram boa aproximação quando comparados com os modelos propostos. Verificou-se que todas as vigas reforçadas apresentaram um acréscimo de resistência à torção em torno de 40% em relação à viga de referência. Verificou-se que, após a fissuração, as vigas reforçadas apresentaram perda de rigidez inferior à da viga de referência. Observou-se que o ângulo da fissura medido experimentalmente, o ângulo de inclinação calculado pelo estado de deformação e o ângulo de inclinação calculado pelo estado de tensão da viga apresentaram valores próximos para cada viga. / [en] A theoretical-experimental research on the torsional behavior up to failure of reinforced concrete beams strengthened with external carbon fiber composites (CFC) was carried out. The experimental study comprises a series of seven reinforced concrete beams with the same compressive strength of concrete loaded to failure and subjected to torsion. The beams dimensions were 20 cm x 40 cm x 420 cm. The test specimens had the same internal steel reinforcement. The beams were divided in three series: the reference beam without strengthening; three beams with the external strengthening applied transversally and three beams with the external strengthening applied transversally and longitudinally. For the accomplishment of the tests an auxiliary steel structure was mounted, capable to transfer to the beams the pure torsion moment. In the theoretical study two analytical procedures were developed. The first formulation, based on the softened space truss model for torsion, presents a systematic to obtain the curve torsion moment x torsion angle per length unit of the reinforced concrete beams with CFC torsion strengthening. The second systematic, based on the Space Truss Model in accordance with the Eurocode 2, presents the design of the CFC strengthening. Both methodologies adopt the Chen and Teng bond model between concrete and CFC. The consideration of the bond in the developed models is very important because the failure of the concrete members often occurs from debonding of the CFC. The experimental results from the beams tests were used to validate the two analytical procedures. Good agreement was obtained with the experimental and analytical results. For all the strengthened beams the average values of torsion strength were increased by 40% when compared to the reference beam. After cracking, the loss of rigidity in the strengthened beams was lower then in the reference beam. The cracking angle experimentally measured and the strut angles evaluated by strain state and stress state presented close values.

[pt] Tubos de material compósito - matriz polimérica reforçada por fibra de vidro - fabricados pela técnica de enrolamento filamentar, foram caracterizados através de microscopia eletrônica digital e processamento de imagens. Três tubos foram fabricados em equipamento próprio seguindo parâmetros de enrolamento similar. Um tubo comercial fabricado por empresa especializada, com parâmetros de enrolamento mais complexos, também foi caracterizado. Para tal, seções circunferenciais foram observadas em um microscópio eletrônico de varredura com captura digital de imagem. Mosaicos de imagens foram gerados, permitindo obter informação com boa resolução local e, simultaneamente, grande abrangência espacial. Assim, foi possível realizar uma caracterização que abrangia desde o tamanho e forma de fibras individuais até a distribuição espacial de milhares de fibras em uma vasta área da amostra. Foram cridas rotinas de processamento e análise de imagens para medir dados como diâmetro, fator de forma, fração volumétrica e ângulo de enrolamento de fibras. Além disso, uma rotina específica foi desenvolvida para a identificação automática das várias camadas de fibras presentes no tubo comercial. / [en] Pipes made from composite material - polymer matrix reinforced with glass fibers - manufactured by filament winding, were characterized by scanning electron microscopy and image analysis. Three pipes were manufactured with equipment owned by the research group, following similar winding parameters. A commercial tube made by a specialized company, with more complex winding conditions, was also characterized. Circumferential sections were observed in a scanning electron microscope with digital image acquisition. Image mosaics were created, providing information with good spatial resolution and, at the same time, wide spatial coverage. Thus, it was possible to characterize size and shape of individual fibers and, simultaneously, obtain the spatial distribution of thousands of fibers within a large sample area. Image processing and analysis routines were created to measure fiber diameter, shape factor, area fraction and winding angle. A specific routine was developed for the automatic identification of the several fiber layers present in the commercial pipe.

Ce travail porte sur porte sur le développement de nouveaux matériaux permettant d’améliorer la visibilité des produits de marquage routier en toutes conditions climatiques. Les études se sont tout d’abord dirigées vers une étude théorique des principes optiques régissant le phénomène de rétroréflexion, puis sur la modélisation de ces phénomènes à l’aide d’un logiciel, pour déterminer les systèmes rétroréflecteurs idéaux par temps sec et par temps humide. De nouveaux verres ayant les propriétés optiques et mécaniques adéquates à ce genre d’applications ont ensuite pu être synthétisés. Des traitements de surface ont enfin été appliqués pour améliorer encore ces propriétés, grâce au phénomène d’adaptation d’indice et à l’étude des interférences dues aux dépôts de couches minces. / This work deals with the development of new materials for improving the visibility of road markings in all weather conditions. The studies moved towards a theoretical study of optical properties responsible for retroreflexion, then towards the establishment of a pattern of these phenomenon with the Zemax software in order to determine ideal retroreflective systems for dry and wet weather. New glasses having the ideal optical and mechanical for these applications were synthesized. Surface treatments have been used to improve these properties, by the study of the interference phenomena induced by thin layers deposited on substrates.

Les développements récents en électronique de puissance visent à augmenter la densité de puissance totale dans les systèmes de conversion d'énergie. Cela contraint alors de plus en plus les matériaux isolants, tels que l'encapsulation dans les modules de puissance. Si les renforcements de champ électrique au sein des polymères isolants atteignent des valeurs critiques, cela peut entraîner une activité de décharges partielles, des arborescences voire la rupture totale de l'isolation. L'objectif de cette thèse est d'étudier l'adaptation des propriétés diélectriques d'un polymère composite isolant afin de réduire les contraintes autour des zones de renforcements de champ électrique. Nous proposons une nouvelle approche de gradation de potentiel pour minimiser les renforcements de champ à travers une structuration locale du matériau composite sous forme d'un gradient de permittivité auto-adaptatif localisé là où les contraintes sont les plus intenses. Cette structuration est réalisée via l'application d'un champ électrique DC lors du procédé d'élaboration du matériau composite, permettant le déplacement par électrophorèse des particules. Le composite à gradient de permittivité est composé d'une matrice époxy chargée en particules à forte permittivité (titanate de strontium SrTiO3 ou titanate de baryum BaTiO3). L'action d'un champ électrique DC sur la résine liquide chargée en particules engendre leur accumulation vers l'électrode de plus fort potentiel, formant ainsi une couche fortement chargée, qui confère à cette région une permittivité plus élevée. Chaque région du composite structuré (zone de la couche accumulée et zone faiblement chargée en particules) a été caractérisée en termes de propriétés diélectriques (permittivité et pertes). Alors que la région des composites faiblement chargée en particules conserve une permittivité voisine de celle des composites homogènes, la couche accumulée présente une augmentation importante liée à l'augmentation de la densité de particules. Les concentrations en particules de chaque région du matériau structuré ont été déterminées précisément, et les valeurs de permittivités associées se corrèlent bien avec les valeurs de permittivité des matériaux composites homogènes de taux de chargement équivalent. Cela montre que la couche accumulée ne s'est pas organisée d'une façon particulière. Concernant la rigidité diélectrique de la couche accumulée, elle présente des valeurs suffisantes pour tenir les contraintes rencontrées et ses valeurs suivent la loi de puissance classique en fonction de l'épaisseur. Des simulations par éléments finis confirment l'intérêt de ces matériaux pour la minimisation des renforcements de champ électrique au niveau du point triple dans les modules de puissance. Ces résultats montrent tout le potentiel applicatif de ces nouveaux matériaux à gradient de champ. Ils pourraient permettre l'amélioration de la fiabilité et de la robustesse des modules de puissance et autres systèmes électriques travaillant sous fort champ. / New developments in power electronics allow increasing the power density of the conversion systems. This means that the insulating materials, such as the encapsulation in power modules, are more are more stressed. If the electric field reinforcements in insulating polymers reach critical values, this can lead to a partial discharge activity, electrical treeing and eventually a complete breakdown of the insulation. The objective of this thesis is to study the appropriate matching of the dielectric properties of insulating polymer composites in order to reduce the electrical stress in the regions of field reinforcement. A new approach to minimize the reinforcements is proposed through a local structuration of the composite material allowing an auto-adaptive permittivity gradient where the largest stresses are present. This structuration is achieved thanks to the application of a DC electric field during the elaboration process of the composite material, leading to the displacement of the particles by electrophoresis. The field grading material is an epoxy matrix filled with high permittivity particles (strontium titanate SrTiO3 or barium titanate BaTiO3). Applying a DC electric field on the liquid resin containing the particles induces their accumulation on the high voltage electrode, building an accumulated layer highly concentrated in particles, conferring to this region a higher permittivity. Each region of the structured composite (accumulated layer and low concentrated region) was characterized in terms of dielectric properties (permittivity and losses). While the low concentrated region of particles keeps a permittivity close to that of homogeneous composites one, the accumulated layer exhibits a significant increase due to the increase in the local particle content. The particle concentration in each region of the structured material were precisely determined, and the related permittivity values are in good agreement with the permittivity values of the homogeneous composite materials of the same filler content. This shows that the accumulated layer was not organized in a particular way. Regarding the dielectric strength of the accumulated layer, its values are large enough for the applications and these values follow the typical power law versus thickness. Finite element methods prove that these materials are appropriate for minimizing the electric field reinforcements at the triple point, between the metal, the ceramic and the encapsulation. These results highlight the interests of these new kind of field grading materials. They could allow improving the reliability and the robustness of power modules or other electrical systems working at high field.

Ce travail de thèse s'est déroulé dans le cadre du programme H2E (Horizon Hydrogène Énergie) piloté par Air Liquide et en partenariat industriel avec Composites Aquitaine. Le stockage d’hydrogène sous forme gazeuse impose que les réservoirs bobinés en composite carbone/époxy utilisés résistent à des pressions internes en service allant de 350 à 700bar.L’état de l’art montre que le procédé d’enroulement filamentaire et son application à des épaisseurs de matériaux composites importantes peuvent générer des variations, dans l’épaisseur, des contraintes résiduelles d’origine thermique, de la teneur en fibres et en porosités et de l’épaisseur des couches. Ces hétérogénéités peuvent alors être à l’origine de modifications des propriétés mécaniques du composite. Notre travail s’est donc appuyé sur la caractérisation physico-chimique à l’échelle des constituants (mesure de températures, nano-indentation, analyse d’images et microtomographie) ainsi que la caractérisation mécanique à l’échelle de la couche de référence mais également à l’échelle de la structure (essais plans spécifiques sur éprouvettes bobinés et essais hors-plan).Les hétérogénéités de la matière ainsi identifiées sur le réservoir ont été intégrées dans un modèle de calcul analytique permettant d'étudier leurs influences sur sa pression d'éclatement. La conception d'un essai de mise sous pression interne d'un anneau a par ailleurs permis de vérifier le comportement mécanique de la structure réelle. / This PhD work was conducted in the H2E (Horizon Hydrogen Energy) program coordinated by Air Liquide and with the industrial partnership of Composites Aquitaine. The hydrogen storage under gaseous form needs the filament-wound carbon/epoxy composite vessels used to resist to service pressures between 350 to 700bar.The influences of the process and of the thickness of the structure on its mechanical behaviour were determined by precise bibliographic work and were supposed to generate thermal residual stresses and also to be responsible of variations in the thickness and fibre content of the plies. These heterogeneities could modify the mechanical properties of the composite material. Our work focuses, first of all, on the physical chemistry characterization at the constituents scale (temperature measurement, nano-indentation, image analysis and microtomography) and then on the mechanical characterization of the reference ply and also of the structure (dedicated tests on filament-wound samples and out-of-plane tests).The material heterogeneities identified on the vessel were integrated in an analytical calculation model to study their influences on its burst pressure. The development of an internal pressure test allows us to verify the global mechanical behaviour of the real structure.

Les matériaux composites jouent un rôle de plus en plus important dans notre société et dans de très nombreux domaines (aéronautique, naval, génie civil…), grâce à leurs avantages en terme de légèreté, d’inaltérabilité et de rigidité. Cependant, ils présentent des faiblesses qui peuvent poser des problèmes au niveau de leur utilisation pour les ouvrages de génie civil. Ces faiblesses concernent notamment leur durabilité. A cause des phénomènes viscoélastiques, les propriétés mécaniques des structures en composites évoluent dans le temps. Le fluage et/ou la relaxation sont des facteurs importants qui peuvent considérablement affecter l’application des composites aux structures. Dans ce travail de doctorat, on effectue une analyse sur le comportement à court et à long terme des composites unidirectionnels renforcés par des fibres de verre/carbone. Afin d’obtenir des résultats quantitatifs sur le comportement mécanique de ces composites, différents types des sollicitations mécaniques seront considérés (ex. compression, cisaillement, tension, flexion). Les analyses sont basées sur deux modèles micromécaniques développés par l'équipe MSA. Le premier modèle est de type shear-lag viscoélastique et le deuxième utilise le logiciel éléments finis Abaqus. Ces deux modèles prennent en compte les différents micro-mécanismes de rupture comme la rupture des fibres, la décohésion des fibres/matrice et le fluage de la résine. Plusieurs analyses numériques sont faites afin de valider les différentes hypothèses de la théorie shear-lag. A partir des analyses menées, des améliorations sont apportées sur le modèle type shear lag. Une étude comparative avec les éléments finis a permis de bien valider les résultats obtenus par la méthode shear-lag. Ayant calibré nos modèles type shear-lag et éléments finis, des simulations types court et long terme sont faites sur des composites unidirectionnels renforcées par des fibres de verre et de carbone. Les analyses sont réalisées sur plusieurs échantillons pour chaque type de fibre (Simulation de Monte-Carlo) / Fiber Reinforced Plastic materials (FRP) are beginning to find more and more applications in the civil engineering domain. Besides the use of FRPs for the reinforcement of existing structures, these materials are utilized quite often today for the construction of bridges and even for new buildings made entirely of FRPs. At the matter of fact, the light weight of composite materials is a considerable advantage compared to conventional materials such as steel or concrete. Another advantage is that they have outstanding fatigue and durability potential and that they are in general very tolerant to environmental effects such as UV radiations, moisture, chemical attack and extreme temperature variations. However, the lack of a comprehensive, validated, and easily accessible database for the durability of fiber-reinforced polymer composites as related to civil infrastructure applications is a critical barrier to their usage as main load bearing systems. The creep behavior of these materials and their failure under sustained loads remains an open research topic. This study gives a detailed analysis on the mechanical behavior of unidirectional fiber reinforced composites (UD FRP) subjected to different loading patterns (tension, compression, shear, and flexure). We develop two micromechanical models that allow us to analyze the instantaneous and the long term response of UD composites subjected to different load patterns. The first model is based on the shear-lag theory and the Beyerlein et al.[1998] developments while the second one is established using the finite element software Abaqus. A Comparative study between the two models allowed to validate the fundamental assumptions of the shear-lag theory (first model) as well as several numerical issues related to time integration and spatial discretization. The Monte Carlo method is used in order to account for the stochastic fiber strength and its impact on the ultimate tensile strength (short term) and creep (long term). A parametric investigation on the fiber type and load level/type on the short/long term behavior of UD composites is also presented