Le dihydrogène (H2) se présente comme le futur vecteur énergétique pour une économie basée sur des ressources propres et respectueuses de l'environnement. Il est le combustible idéal de la pile à combustible régénérative constituée de deux entités : un électrolyseur pour sa production, et une pile à combustible pour sa conversion directe en énergie électrique. Ce système présente l'avantage d'être compact et autonome. Cependant, l'amélioration de l'activité catalytique des matériaux, leur stabilité et l'élimination de métaux nobles dans leur composition sont nécessaires. Des catalyseurs bifonctionnels à base de métaux de transition associés au graphène ont alors été synthétisés. L'interaction oxyde-graphène a été étudiée sur un catalyseur Co3O4/NRGO. À faible teneur en cobalt, l'interaction entre les atomes de cobalt de l'oxyde et les atomes d'azote greffés sur les plans de graphène a été observée par voltammétrie cyclique. Cette interaction est responsable d'une diminution de la taille des nanoparticules de cobaltite et de l'activité de celles-ci vis-à-vis de la réaction de réduction du dioxygène (RRO). La substitution du cobalt par le nickel dans des structures de type spinelle (NiCo2O4/RGO) obtenu par voie solvothermale, a permis d'améliorer les performances électrocatalytiques vis-à-vis de la RRO et de la RDO. Ce matériau et un autre de type Fe-N-C préparé en collaboration avec un laboratoire de l'Université Technique de Berlin ont servi de cathode dans des études préliminaires réalisées en configuration pile à combustible alcaline à membrane échangeuse d'anion (SAFC). / Hydrogen, as an environmentally friendly future energy vector, is a non-toxic and convenient molecule for regenerative fuel cell, which connects two different technologies: an electrolyzer for H2 production, and a fuel cell for its direct conversion to electric energy. This kind of system possesses many advantages, such as lightness, compactness and more autonomy. However, improvement of activity and durability of electrode materials free from noble metals in their composition is needed. Thereby, bifunctional catalysts composed of transition metals deposited onto graphene-based materials were synthesized. The interaction between the metal atom of the oxide and the graphene doped heteroatom in the Co3O4/NRGO catalyst was investigated physicochemically. With a low cobalt loading, the interaction between cobalt and nitrogen was characterized by cyclic voltammetry, which revealed that it was responsible for decreasing the oxide nanoparticle size, as well as increasing the material activity towards the oxygen reduction reaction (ORR). The substitution of Co by Ni in the spinel structure (NiCo2O4/RGO) obtained by solvothermal synthesis, allowed the enhancement of the electrocatalytic performances towards the ORR and OER. Moreover, this catalyst as well as another material prepared in collaborative program with a lab from Technical University of Berlin were used as cathode in preliminary studies undertaken on solid alkaline fuel cell (SAFC).

Various aspects of myrmecochory were investigated. Effect of different storage methods on atractiveness of seeds with elaiosomes for ants was examined. Viable seed bank of refuse piles (places where ants deposit unused objects from ant nests, e.g. seeds after the elaiosome was eaten) and places outside them were compared. Chemical content of five major chemical groups, specifically amino acids, free fatty acids, organic acids, polyols and sugars, of elaiosomes and seeds of selected species was determined, and compared with respect to their taxonomic relatedness.

Various aspects of myrmecochory were investigated. Effect of different storage methods on atractiveness of seeds with elaiosomes for ants was examined. Viable seed bank of refuse piles (places where ants deposit unused objects from ant nests, e.g. seeds after the elaiosome was eaten) and places outside them were compared. Chemical content of five major chemical groups, specifically amino acids, free fatty acids, organic acids, polyols and sugars, of elaiosomes and seeds of selected species was determined, and compared with respect to their taxonomic relatedness.

Dans les piles primaires au lithium, l'anode de lithium peut être couplée principalement à quatre matériaux de cathode (MnO2, SO2, SOCl2 et des composés fluorés CFx). Les piles Li/CFx présentent de nombreux avantages tels qu’une densité d'énergie élevée (jusqu'à 2200 Wh.kg-1), un potentiel de décharge relativement élevé (environ 2,4 V vs Li+/Li) et une longue durée de vie (plus de 10 ans à la température ambiante). Au cours de ce travail, nous avons intensément étudié les propriétés électrochimiques des carbones fluorés nanostructurés. Pour plus de performances, des précurseurs nanocarbonés de dimensionnalité différentes, et des procédés de synthèse gaz-solide variés ont été employés. Ainsi, des structures fermées comme les nanofibres de carbone (1D, tubulaire), ouvertes comme les nanodisques / nanocônes (2D en majorité, discotiques) et intermédiaires avec les noirs de carbone graphitisés. Chaque mode de synthèse a été optimisé en fonction du précurseur de départ (son facteur de forme, sa dimensionnalité, son degré de graphitisation ...). Pour remédier aux limitations des carbones fluorés commerciaux, leur mécanisme de décharge a été finement étudié afin d'améliorer notre compréhension sur la défluoration électrochimique d'une cathode CFx, et de les comparer aux CFx synthétisés au laboratoire, le but principal étant de dépasser les performances actuelles et de synthétiser des matériaux inédits pour la pile primaire au lithium. Des extracapacités par rapport aux valeurs théoriques dans les batteries Li/CFx ont été obtenues avec de nanocarbones fluorés spécifiques et le phénomène électrochimique à l’origine de cette extracapacité a été clairement expliqué. / For primary lithium batteries, the lithium anode can be coupled mainly four cathode materials (MnO2, SO2, SOCl2 and fluorinated carbons CFx). Li/CFx batteries have many inherent advantages such as high energy density (up to 2200 Wh.kg-1), a moderately high discharge potential (about 2.4 V vs. Li+/Li) and a long life (over 10 years at room temperature). During this work, the electrochemical performances in primary lithium battery of nanostructured fluorinated carbons were investigated. To improve their performances, carbonaceous precursors with different dimensionalities, and various gas-solid fluorination methods were selected. Thus, a closed structure such as carbon nanofibres (1D, tubular), opened such as carbon nanodiscs/nanocones (2D in majority, discotic) and intermediate with graphitized carbon blacks have been chosen. Each synthesis route was optimized based on the starting material and their specificities, i.e. the shape factor, dimensionality, graphitization degree, stacking mode. To address to the limitations of commercial CFx materials, their discharge mechanism has been studied in order to improve the understanding about the electrochemical defluorination of a CFx cathode, and to compare it to original CFx synthesized in our laboratory, the main purpose is to exceed the current performances and to synthesize unrealized materials for primary lithium battery. Extracapacities in Li/CFx batteries were obtained with some specific fluorinated nanocarbon materials and their explanation was discussed.

Ce travail de thèse s’est intéressé aux mécanismes de dégradation de nanoparticules de Pt supportées sur carbone utilisées pour catalyser les réactions électrochimiques dans une pile à combustible à membrane échangeuse de protons (PEMFC) et à leur conséquences d’un point de vue cinétique. Nous avons mis en évidence les différents mécanismes (maturation d’Ostwald 3D, corrosion du support carboné, migration/agrégation des cristallites métalliques) conduisant à une perte de surface active électrochimiquement et avons trouvé des conditions permettent d’isoler chacun de ces mécanismes. En premier lieu, nous avons montré que les nanoparticules de Pt supportées sur carbone ne sont pas immobiles mais agrègent en conditions réactionnelles notamment en présence de molécules réductrices. La vitesse de ce processus varie dans l'ordre CO > CH3OH > H2 et a été reliée à (i) la baisse du travail d’adhésion engendrée par la chimisorption de ces molécules et (ii) la réduction des groupements oxygénés présents sur le support carboné natif.Nous nous sommes également intéressés au mécanisme d’électrooxydation électrochimique du Vulcan XC72, un noir de carbone classiquement utilisé dans les couches catalytiques de PEMFC. Des mesures par spectroscopie Raman ont montré que les domaines désordonnés du Vulcan XC72 (non-graphitiques, hybridation sp3) sont corrodés de façon préférentielle dans des conditions expérimentales proches de celles d’une cathode de PEMFC. Les domaines ordonnés du support carboné (carbone graphitique, hybridation sp2) sont également corrodés, la vitesse de ce processus étant largement inférieure à ce qui est observé sur les domaines désordonnés. En conséquence, les nanoparticules de Pt se détachent ou agglomèrent comme le révèlent des expériences de microscopie électronique en transmission couplées à l’électrochimie. L’ensemble de ces mécanismes de dégradation conduit à un abaissement de la densité du nombre de particules métalliques et augmente la distance entre ces dernières. Dans le chapitre IV, nous montrons que des électrocatalyseurs Pt/Sibunit electrocatalysts possédant (i) un faible chargement massique en Pt, et (ii) de grandes distances inter-particules présentant une faible activité pour la réduction du dioxygène de l’air. Le nombre moyen d’électrons transférés par molécule de dioxygène décroît bien sous la valeur théorique de 4 lorsque l’épaisseur de la couche catalytique ou le chargement massique diminue. Nous avons relié cela à un transport et à une ré-adsorption plus difficiles des intermédiaires réactionnels notamment le péroxyde d’hydrogène. Une diminution du nombre de sites catalytiques peut également engendrer une limitation des cinétiques réactionnelles par l’adsorption de l’oxygène. Au vu de l’ensemble des résultats précédents, nous avons conclu que des cristallites de plus grande taille permettraient d’améliorer la durabilité des matériaux contenus dans les couches catalytiques de PEMFC. Des nano-fils de Pt (NWs) avec une taille moyenne de cristallite de 2,1 ± 0,2 nm ont été synthétisés. Nous avons montré que la morphologie du matériau joue un rôle conséquent à la fois en termes d’activité électrocatalytique et de durabilité : les matériaux Pt NWs/C permettent une réduction de prêt de 170 mV de la surtension d’oxydation d’une monocouche de monoxyde de carbone et possèdent une activité catalytique élevée et stable pour l’électrooxydation du méthanol. Cette dernière a été attribuée à (i) l’augmentation de la masse des cristallites de Pt résultant de l’augmentation en taille (nanoparticules à nano-fils) et (ii) une surface de contact élevée avec le support carboné. Ces matériaux possèdent un potentiel intéressant pour résoudre les problèmes de durabilité rencontrés avec les matériaux 0D utilisés de façon conventionnelle. / State-of-the-art catalytic layers of proton-exchange membrane fuel cells (PEMFCs) utilize ionomer-bonded Pt-based nanocrystallites supported on a high surface area carbon support to accelerate the rate of the hydrogen oxidation reaction at the anode, and of the oxygen reduction reaction at the cathode. Post-mortem analysis of PEMFC catalytic layers reveal that four degradation mechanisms yield Pt surface area losses (and hence decreased PEMFC performance): (i) aggregation and/or detachment of the metal nanoparticles, (ii) corrosion of the carbon support, (iii) Ostwald ripening (dissolution/redeposition of the Pt-based crystallites) yielding the formation of ionic species and (iv) chemical reduction of the Ptz+ species in ion conductors, yielding the formation of electrically disconnected Pt crystallites. A major concern of the Ph.D. was to be able to isolate the migration of the Pt nanocrystallites. For that purpose, commercial Pt/Vulcan XC 72 electrocatalysts were aged in “mild” conditions, for which the corrosion of both the Pt nanoparticles and the carbon support could not be considered dominant. Evidences were provided that the Pt/C nanoparticles are not immobile but prone to agglomerate in the presence of H2, CH3OH, and CO, three molecules of interest for PEMFCs. The migration rate of the Pt crystallites was the largest in CO-containing solution and decreased in the order CO > CH3OH > H2. We postulate that the morphological changes of the Pt/C nanoparticles may be caused by (i) a change of the work of adhesion between the metal phase and the carbon support or (ii) the reduction of the oxygen-bearing surface groups strongly interacting with the Pt nanocrystallites. We also investigated the mechanism of the electrochemical oxidation of Vulcan XC72, a carbon black conventionally used in PEMFCs. Raman spectroscopy measurements evidenced that the disordered domains of the Vulcan XC72 support (non-graphitic, sp3-hybridized) are preferentially oxidized. The ordered domains (graphitic carbon, sp2-hybridized) of the Vulcan XC72 support are also oxidized but at a much smaller rate than that observed on the non-graphitic domains. A major consequence of the oxidation of the high-surface area carbon support is the aggregation and the detachment of the supported Pt nanoparticles. In Chapter V, we used model Pt/Sibunit electrocatalysts to show that aged catalytic layers with (i) low Pt to C weight fraction, (ii) large inter-particle distance, feature bad performance for the oxygen reduction reaction (ORR). The average number of transferred electrons produced during the ORR decreases below 4 with the decrease of the catalyst layer thickness or the Pt loading. This was rationalized by considering the lower probability for H2O2 molecules (the major reaction intermediate) to be re-adsorbed and further reduced into water in the catalytic layer when the inter-particle distance decreases. Finally, Pt nanowires (NWs) with 2.1 ± 0.2 nm crystallite size were synthesized by a soft template method. Evidences were provided that the morphology of the Pt material plays a pivotal role both in terms of electrocatalytic activity/stability: the 1D Pt NWs/C demonstrate a reduction by ca. 170 mV of the CO oxidation overpotential and feature high and stable MOR activity with respect to a conventional Pt/C 20 wt. % catalyst. The enhanced durability of Pt NWs/C was rationalized by considering (i) the increase in weight of the Pt nanomaterials resulting from the increase in size (from nanoparticles to nanowires) and (ii) the enhanced contact surface area between the Pt NWs and the carbon support. The development of 1D Pt nanostructures, such as Pt NWs, hold promises to solve the durability issues faced with the 0D materials currently used in PEMFCs.Keywords: proton exchange membrane fuel cell, durability of PEMFC materials, electrochemical carbon oxidation, crystallite migration, Pt nanowires.

Cette thèse est consacrée à l’étude des mécanismes de dégradation de noirs de carbone de forte surface spécifique (HSAC) utilisés comme supports d’électrocatalyseurs dans une pile à combustible à membrane échangeuse de protons (PEMFC). Nous avons montré que le mécanisme et les cinétiques de la corrosion électrochimique du carbone (COR) sont influencés par la présence d’ionomère Nafion®, la limite supérieure de potentiel électrochimique, la nature et le nombre de caractérisations intermédiaires présentes dans des tests de dégradation accélérés. En utilisant la spectroscopie Raman,il apparaît que la COR est sensible à la structure cristallographique des HSAC et procède plus rapidement sur les domaines désordonnés (carbone amorphe, cristallites de graphite présentant des défauts). Le taux de recouvrement en espèces oxygénées évalué par spectroscopie de photoélectrons X a été comparé à celui trouvé en intégrant l’intensité du pic quinone/hydroquinone (Q/HQ) envol tampérométrie cyclique. Finalement, une comparaison avec des matériaux carbonés ayant fonctionné pendant 12860 heures en PEMFC confirme nos principaux résultats et permet d’élaborer des stratégies pour atténuer les conséquences de la COR. / This thesis investigates the degradation mechanism of high surfacearea carbon (HSAC) supports used in proton-exchange membrane fuel cell (PEMFC) electrodes. The structural and the chemical properties of different HSAC supports were established. The effectof the Nafion® ionomer used as a proton conductor, the gas atmosphere, the upper potential limit and the intermediate electrochemical characterizations used to monitor the changes ofthe electrochemical surface area during accelerated stress tests(ASTs) were investigated. The long-term physical and chemical changes of Pt/HSAC electrocatalysts were investigated insimulated PEMFC operating conditions. Using Raman spectroscopy, we showed that the COR is strongly structure sensitive and proceeds more rapidly on disordered domains of the HSAC (amorphous carbon and defective graphite crystallites) thanon graphitic domains. The coverage with carbon surface oxides was investigated with X-ray photoelectron spectroscopy and bridged tothe intensity of the quinone/hydroquinone (Q/HQ) peak monitored by cyclic voltammetry. Finally, the analyses realized on membrane electrode assemblies operated for 12,860h disclosed a perfect agreement between model and real PEMFC operating conditions, and confirmed the structural dependency of the COR kinetics.

L’optimisation des performances d’une pile à combustible (PEMFC) requiert la compréhension microscopique des mécanismes de transport de l’eau et du proton confinés au sein de la membrane électrolyte polymère. La membrane est un matériau nanostructuré chargé, caractérisé par une dynamique de l’eau et du proton complexe et multi-échelle étroitement corrélée à la morphologie confinante. Nous nous sommes intéressés à la relation structure – transport dans i) L’Aquivion, un ionomère perfluorosulfonique récent présentant de bonnes performances en pile, ii) des systèmes «modèles» auto-assemblés de tensioactifs perfluorés formant des phases lamellaires et hexagonales et iii) une nouvelle membrane hybride préparée par dopage en tensioactif. La nano-structuration des différents systèmes a été étudiée par diffusion de rayonnement (X et neutrons), pour caractériser l’évolution de la structure (géométrie de la matrice hôte, taille de confinement) avec l’hydratation. Ensuite, nous avons sondé la dynamique de l’eau à l’échelle moléculaire (de la picoseconde à la nanoseconde) par diffusion quasi-élastique des neutrons (QENS) et à l’échelle micrométrique par RMN à gradients de champs pulsés. La comparaison membranes commerciales / systèmes modèles permet de discuter l’impact de la connectivité, du confinement et de la géométrie sur le transport ionique. Enfin, des membranes hybrides à fort potentiel ont été obtenues par dopage du Nafion et de l’Aquivion avec des tensioactifs. Ces nouveaux matériaux ouvrent une voie prometteuse pour la préparation de membranes polymères fortement anisotropes avec des chemins de conduction préférentiellement orientés. / The optimization of the Fuel Cell’s performances (PEMFC) requires a microscopic understanding of the water and proton’s transport mechanism, which are confined in a polymer electrolyte membrane. The latter is nanostructured, charged and characterized by a complex and multi-scale water and proton dynamics, closely correlated to the confining morphology. We studied the structure-transport interplay in i) the Aquivion, a recent perfluorosulfonic ionomer exhibiting good performances in fuel cell, ii) “model” systems of perfluorosulfonic surfactants, which self-assemble in lamellar and hexagonal phases and iii) a new hybrid membrane doped with surfactant. The nanostructuration of the different systems has been studied by neutron and X-ray scattering, to characterize the structural evolution (host matrix geometry, confinement sizes) with hydration. Then, we probe the water dynamics at the molecular level (from picosecond to nanosecond) with Quasi-Elastic Neutron Scattering (QENS) and at the micrometric scale with Pulsed Field Gradient NMR. The comparison of commercial membranes and model systems bring new insight on the impact of the connectivity, the confinement and the geometry, on the ionic transport. Finally, high potential hybrid membranes have been obtained by doping Nafion and Aquivion with surfactants. Those new materials open a promising way for the preparation of highly anisotropic polymer membrane, with conducting path preferentially oriented.

O monitoramento em campo de fundações em estacas metálicas helicoidais é baseado em correlações empíricas entre a capacidade de carga à tração e o torque medido durante a instalação. Nesta pesquisa é proposto um modelo teórico da relação entre o torque de instalação e a capacidade de carga à tração deste tipo de estaca em areias para ser utilizado como procedimento de controle durante a execução deste tipo de fundação. Para validar esta relação teórica e para avaliar o desempenho desta fundação em solos arenosos, foi realizado um programa experimental dividido em dois grupos: ensaios de campo e ensaios de laboratório. Em campo foram realizadas provas de carga à tração na Escola de Engenharia de São Carlos (EESC) e em Contagem (Minas Gerais). Os ensaios de laboratório foram divididos em três grupos. O primeiro corresponde aos ensaios de modelagem física em centrífuga realizados no Laboratoire Central des Ponts et Chaussées (LCPC) em Nantes, França. O segundo engloba ensaios de cisalhamento direto de interface entre o aço da hélice da estaca e diferentes areias, parte executada no LCPC e parte na EESC. O terceiro grupo refere-se a ensaios de torção de tubos de estacas metálicas helicoidais efetuados na EESC. A partir dos resultados dos ensaios, o modelo teórico proposto nesta tese foi verificado e mostrou-se adequado para estacas metálicas helicoidais profundas com diferentes dimensões e número de hélices instaladas em solos arenosos. Os resultados das provas de carga realizadas em centrífuga e em campo foram examinados por meio dos métodos existentes na literatura para estimativa de capacidade de carga à tração desenvolvidos para este tipo de fundação. Por meio dos resultados obtidos nos ensaios em centrífuga, constatou-se que nas estacas multi-hélices a porcentagem de capacidade de carga relativa à hélice da ponta aumenta de acordo com o diâmetro da hélice e com a densidade da areia. Nos ensaios de torção de fustes de estacas helicoidais observou-se que, para tubos de mesmo diâmetro e espessura, a ruptura se inicia a partir do mesmo valor de momento torçor aplicado, independentemente de conter ou não emenda. / Empirical correlations between installation torque and uplift capacity of helical screw piles are routinely used as a procedure for quality control on site of this type of foundation. In this work, a theoretical model of the relationship between uplift capacity and installation torque of deep helical screw piles in sand was presented to be used as a quality control method for this type of foundation. An experimental program of laboratory and field investigations was carried out in order to validate this theoretical relationship. The field study included uplift tests of full-scale piles installed at the Escola de Engenharia de São Carlos (EESC) and in Contagem (Minas Gerais). The laboratory tests were performed in the Laboratoire Central des Ponts et Chaussées (LCPC/ Nantes) and at the EESC. In the LCPC a program of centrifuge and direct shear interface tests was completed. A set of torsion tests on pile shafts and of direct shear interface tests were carried out at the EESC. The results showed that the theoretical model presented is applicable to different pile dimensions, number of helices, and sand relative density. The methods existing in the literature for estimating uplift capacity of helical screw piles were used to verify the results of uplift capacities measured in field and in centrifuge studies. The results of centrifuge modelling tests demonstrated that the fraction of uplift capacity corresponding to the bottom helix increases with helix diameter and sand density. The torsion tests showed that the failure of the pile shaft begins at the same torque magnitude in experiments on shafts with and without connections.

Analisam-se problemas de interação solo-estrutura através de uma formulação tridimensional obtida da combinação entre o Método dos Elementos de Contorno (MEC) e o Método dos Elementos Finitos (MEF). Os elementos estruturais que interagem com o solo são modelados pelo Método dos Elementos Finitos. E o solo, considerado como um meio semi-infinito, homogêneo, elástico linear e isotrópico, é modelado pelo Método dos Elementos de Contorno, empregando a solução fundamental de Mindlin. A solução fundamental de Mindlin é particularmente adequada para o tipo de problemas em análise, ou seja, problemas envolvendo sólidos tridimensionais semi-infinitos, já que é necessário discretizar apenas a superfície carregada do solo e/ou a linha de carga, e não todo o sólido tridimensional. A discretização da estaca em vários elementos finitos de viga tridimensional, permitindo a consideração de estacas de qualquer tamanho e submetidas a qualquer tipo de carregamento é uma das principais contribuições do trabalho. Outra contribuição diz respeito à consideração da ação horizontal no sistema placa-estaca-solo, diretamente aplicada na placa, e não redistribuída no topo das estacas. Por isso, os elementos estruturais laminares (radiers, sapatas e blocos de fundação) são modelados por elementos finitos de casca plana, possibilitando a consideração dos efeitos de flexão e de membrana. Essa abordagem permite a análise tanto de um grupo de estacas com bloco de capeamento rígido quanto de um radier estaqueado de qualquer rigidez. O acoplamento entre o MEC e o MEF é feito através de uma formulação mista em que a matriz dos coeficientes de influência do solo obtida pelo MEC é adicionada à matriz de rigidez dos elementos estruturais obtida pelo MEF, resultando em uma matriz de rigidez equivalente. Exemplos numéricos de interação estaca-solo, placa-solo e casca plana-estaca-solo são resolvidos para verificar, validar e demonstrar a eficiência das formulações desenvolvidas e implementadas. / Soil-structure interaction problems are analyzed by a tridimensional formulation obtained combining the Boundary Element Method (BEM) and the Finite Element Method (FEM). Structural elements that interact with the soil are modeled by the Finite Element Method. And the soil, considered as a semi-infinite, homogeneous, linear elastic and isotropic medium, is modeled by the Boundary Element Method, using Mindlin\'s fundamental solution. Mindlin\'s fundamental solution is particularly suitable for the type of problems under analysis, viz. problems involving semi-infinite three-dimensional solids, since it is necessary to discretize only the loaded surface of the soil and/or the line-load, and not all the three-dimensional solid. The discretization of the pile in several three-dimensional beam finite elements, allowing the consideration of piles of any size and subjected to any type of loading, is one of the main contributions of this work. Another contribution is about the consideration of the horizontal load in the plate-pile-soil system, directly applied to the plate, and not redistributed on the top of the piles. Therefore, the laminar structural elements (rafts, footings, and foundation blocks) are modeled by flat shell finite elements, making possible the consideration of the effects of flexion and membrane. This approach allows the analysis of both a capped pile group and a piled raft of any stiffness. The coupling between the BEM and the FEM is performed through a mixed formulation in which the matrix of the soil\'s influence coefficients obtained by the BEM is added to the stiffness matrix of the structural elements obtained by the FEM, resulting in an equivalent stiffness matrix. Numerical examples of pile-soil, plate-soil and flat shell-pile-soil interaction are solved to verify, validate and demonstrate the efficiency of the developed and implemented formulations.

O uso de fundações por estacas helicoidais tem aumentado consideravelmente nos últimos anos, devido às suas vantagens comparadas a outros tipos de fundação, como a sua rápida instalação e sua capacidade de resistir simultaneamente a esforços de tração e de compressão. No entanto, os métodos de previsão do comportamento deste tipo de fundação submetida a esforços de tração (denominadas ancoragens helicoidais neste caso), ainda são insatisfatórios, e normalmente são observadas discrepâncias entre os resultados estimados por métodos teóricos e medidos em provas de carga. Entre outras razões, estas diferenças ocorrem principalmente pelo fato dos efeitos da instalação destas ancoragens no solo não serem considerados nos métodos de previsão disponíveis. No momento da instalação, os parâmetros do solo atravessado são modificados, e como é o volume de solo alterado que deve resistir ao carregamento da ancoragem sob tração, é essencial entender e quantificar este efeito de instalação para uma previsão adequada da capacidade de carga deste tipo de ancoragem. Para contribuir com o entendimento deste efeito, no presente trabalho foram realizadas simulações numéricas com o software FLAC3D, para se conhecer as modificações ocorridas no solo devido à instalação de uma ancoragem helicoidal em areia compacta. Para este fim, na simulação do solo penetrado pela hélice, foram reduzidos os valores de ângulo de atrito e do módulo de elasticidade da areia. O modelo foi validado por meio da comparação com a curva carga-deslocamento obtida experimentalmente em ensaios realizados em centrifuga. Após o ajuste do modelo numérico, foi realizada uma análise paramétrica com o objetivo de investigar o efeito da instalação no fator de capacidade de carga (Nq). Por meio da análise paramétrica também foram avaliados o deslocamento da ancoragem sob carga máxima e as tensões horizontais atuantes na superfície de ruptura. Os resultados estimados de Nq foram comparados com estudos anteriores, e mostraram um bom ajuste com resultados de provas de carga in situ e em centrifuga. Além disso, o estudo paramétrico indicou que as tensões horizontais atuantes na zona de ruptura são distribuídas em forma de tronco de cone, e variam de acordo com o diâmetro da hélice. / The use of helical foundations has increased considerably in recent years due to its advantages compared to other types of foundations, such as rapid installation and the capacity of resisting tensile and compressive loads. However, the methods normally used to predict the behavior of this type of foundation under tensile loads (called helical anchors in this case), are still unsatisfactory, and discrepancies are usually observed between the results estimated by theoretical methods and measured in load tests. Among other reasons, these differences occur mainly because the effects of helical anchor installation in the soil are not considered in the available predicting methods. During installation, the parameters of the soil penetrated are modified, and since it is this disturbed soil volume that must support the loading of the anchor under tension, it is essential to understand and quantify this installation effect for an adequate prediction of the uplift capacity of this type anchor. In order to contribute to the understanding of this effect, in the present work numerical simulations with the software FLAC3D were performed to evaluate the modifications occurred in the soil due to the installation of a helical anchor in dense sand. For this purpose, in the simulation of the soil penetrated by the helices, the values of friction angle and modulus of elasticity of the sand were reduced. The model was validated by means of the comparison with a load-displacement curve obtained experimentally in centrifuge model tests. After adjusting the numerical model, a parametric analysis was carried out to investigate the effect of the installation on the breakout factor (Nq). The parametric analyses were also used to investigate the anchor displacement at the peak load and the horizontal stresses acting on the failure surface. The estimated Nq values were compared with previous studies, and showed a good agreement with experimental results obtained by field and centrifuge tests. Additionally, the parametric study indicated that the horizontal stresses acting on the failure zone are distributed in the shape of a truncated cone, and vary with the helix diameter.