[pt] No estado do Rio de Janeiro, os movimentos de massa em encostas passaram a representar um dos problemas mais importantes em geotecnia devido ao seu relevo acidentado, forma de ocupação e condições climáticas rigorosas. Visando identificar uma alternativa para avaliar encostas usando métodos não invasivos que possibilitem o trabalho em locais de difícil acesso, com rapidez e aquisição contínua de dados, neste trabalho é exposto uma análise teórica e prática deste processo de avaliação utilizando métodos geofísicos e o monitoramento contínuo da água no solo por meio de sensores. Os levantamentos geofísicos com uso do georadar e da análise multicanal de ondas superficiais para investigação da estratigrafia da encosta estudada, possibilitaram a identificação de três camadas: solo maduro, solo residual jovem e rocha. Para o monitoramento contínuo, são usados sensores de umidade volumétrica, sucção e um pluviômetro durante um período de 6 meses com a coleta de dados a cada 10 minutos. Com a interpretação dos dados coletados, verificou-se uma boa funcionalidade do monitoramento para avaliar as respostas de sucção e umidade volumétrica ao longo de períodos com intervalos irregulares de chuva. Os dados de precipitação e estratigrafia identificados são usados nas análises de infiltração e estabilidade saturada e não saturada, que possibilitaram a identificação de seções críticas com fator de segurança abaixo de 1,5, onde são necessárias intervenções para garantir a segurança da vizinhança. / [en] In the state of Rio de Janeiro, mass movements on slopes have become one of the most important geotechnical problems due to their rugged terrain, occupation and harsh weather conditions. In order to identify an alternative to evaluate slopes using non-invasive methods that enable work in hard to reach places, with rapid and continuous data acquisition, this paper presents a theoretical and practical analysis of this evaluation process using geophysical methods and continuous monitoring ground water through sensors. Geophysical surveys using georadar and multichannel analysis of surface wave to investigate the stratigraphy of the studied slope, allowed the identification of three layers: mature soil, young residual soil and rock. For the continuous monitoring, volumetric humidity and suction sensors and a rain gauge are used for a period of 6 months with data collection every 10 minutes. With the interpretation of the collected data, a good monitoring functionality was verified to evaluate the suction and volumetric humidity responses during periods with irregular rain intervals. The identified precipitation and stratigraphy data are used in the infiltration and saturated stability analyzes, which allowed the identification of critical sections with safety factor below 1.5, where interventions are necessary to ensure the safety of the neighborhood.

Cette thèse concerne la caractérisation expérimentale d’un sable limoneux provenant du barrage de Livet – Gavet (38) dans le cadre du projet ANR TerreDurable avec plusieurs objectifs : 1- Caractériser au travers d’essais de laboratoire le comportement hydromécanique d’un sable fin limoneux (sol A1 dans la classification GTR) en fonction de son état de saturation. Lors de cette étude, un accent particulier est porté sur la caractérisation de ce comportement dans le domaine proche de la saturation. 2- Interpréter le comportement hydrique du matériau sur chemin de drainage – imbibition en relation avec l’analyse de sa microstructure. 3- Fournir d’un point de vue général une base de données et d’analyses exhaustive permettant le développement et la calibration de modèles de comportement des sols fins proches de la saturation, en particulier en considérant des chemins de chargement hydromécanique complexes. Pour l’ensemble de cette étude, le matériau est considéré sous deux états : soit à l’état de pâte (matériau normalement consolidé) préparée à une teneur en eau proche de la limite de liquidité, soit sous forme compactée (matériau sur-consolidé) à différentes énergies de compactage et différentes teneurs en eau initiales. / This thesis concerns the experimental characterization of a silty sand from the Livet - Gavet dam (38) as part of the ANR TerreDurable project, for following objectives: 1- Through laboratory tests, characterize the hydro-mechanical behaviour of a fine silty sand (Type A1 in the GTR classification) according to its saturation state. In this study, particular emphasis is placed on the characterization of this behaviour in the near-saturated domain. 2- Interpret the water behaviour of material on the drainage - imbibition cycles, in relation to the analysis of its microstructure. 3- From a general point of view, provide a comprehensive database and analysis allowing the development and calibration of models of near-saturated fine soil's behaviour, in particular, by considering complex hydro-mechanical loading paths. For all tests in this study, the material is considered in two states: either in the state of paste (normally consolidated material) prepared at water content close to the limit of liquidity, or in compacted state (over consolidated material) at different compaction energies and different initial water contents.

Les routes sont normalement pourvues de systèmes de drainage dimensionnés et exécutés conformément aux règles de l’art en vue d’évacuer, le plus rapidement possible de l’emprise de la route, les eaux des précipitations à considérer sur la base des périodes de retour prises en compte. Toutefois, il subsiste souvent des écoulements d’eau indésirables au niveau des talus et parfois dans les accotements et/ou les surfaces de roulement non imperméabilisées. Une succession de pluies entraîne des infiltrations d’eau qui varient notamment en fonction des conditions climatiques et suivant la texture et la structure du sol. De telles infiltrations ont pour conséquence la réduction des coefficients de sécurité des talus.<p><p>Bien qu’il existe plusieurs publications scientifiques traitant de pluies qui ont conduit à des glissements de terrain (Lim et al. 1996 ;Cho et al. 2001 ;Kim et al. 2004 ;Xue et al. 2007 ;Gavin et al. 2008), les incidences des infiltrations résultant de pluies successives sur le comportement des couches superficielles des sols non saturés ne sont généralement pas prises en compte. Les modèles permettant le calcul de la stabilité de talus des massifs de sols non saturés exigent beaucoup de paramètres parfois difficiles à évaluer et se rapportent ordinairement aux cas d’instabilité provoqués par une remontée du niveau piézométrique des nappes phréatiques.<p><p>Sur la base des essais réalisés en laboratoire, une méthodologie adaptée permettant d’évaluer la variation spatio-temporelle de la teneur en eau du sol en fonction d’une suite de pluies a été développée. Cette méthodologie facilite la prise en compte des effets cumulés des taux d’infiltration associés aux évènements pluvieux et permet d’en déduire le profil de succion ainsi que celui de la cohésion apparente à utiliser en vue de calculer, pour une inclinaison β du talus, l’intervalle de variation du coefficient de sécurité Fs. La méthodologie développée présente un intérêt particulier dans le cas de budgets et infrastructures limités.<p><p>ABSTRACT<p><p>Roads are normally equipped with drainage systems sized and implemented in accordance with the rules of art to evacuate as quickly as possible to the right of way, water precipitation to be considered on the basis of return periods taken into account. However, there are often water flows at the slope side and sometimes in the shoulders and / or running surfaces that are not waterproof. A succession of rain causes a certain amount of water infiltration, which varies according to climatic conditions and depending on the soil texture and structure. Such infiltrations have resulted in reduced safety factor of slopes.<p><p>Although there are several scientific publications on rainfall leading to landslides (Lim et al. 1996; Cho et al. 2001, Kim et al. 2004; Xue et al. 2007; Gavin and al. 2008), impacts resulting from infiltration of successive rains on the behavior of surface layers of unsaturated soils are usually not taken into account. Models for calculating the slope stability of unsaturated soils require many parameters that can be, in certain circumstances, difficult to assess and refer generally to cases of instability caused by a rise in groundwater level.<p><p>Based on laboratory tests, a suitable methodology for assessing the spatial and temporal variation of soil water content induced by a set of rains has been developed. This methodology facilitates the inclusion of the cumulative effects of the infiltration rates associated with rain events and infers from them the profile of suction and that of the apparent cohesion to be used to calculate, for a slope angle β, the range of variation of the safety factor Fs. This methodology is particularly relevant in the case of limited budgets and infrastructures. / Doctorat en Sciences de l'ingénieur / info:eu-repo/semantics/nonPublished