No description available.

Rockfill columns have been used to stabilize the clay riverbanks in the Winnipeg area for over two decades. The construction methods used in Manitoba are uniquely adapted to the local soil conditions. The performance of rockfill columns in Manitoba has generally been satisfactory, except that in some cases significant deformations have occurred during and after construction (Yarechewski and Tallin 2003). This thesis will discuss the full-scale test loading of a riverbank that was stabilized with rockfill columns. The purpose of the test was to measure the load-deformation characteristics of a reinforced slope in order to further the understanding of rockfill column behaviour. Rockfill column technology has evolved from granular shear key methods for stabilizing slopes. The relatively weak and soft lacustrine clay is reinforced with compacted columns of stronger and stiffer limestone rockfill. The research test site is located along the natural banks of the Red River in The City of Winnipeg. The project involved an extensive site investigation, and soils characterization program in preparation for the field test. Eleven columns, 2.1 m in diameter were tested by surcharging the slope with 1920 tonnes of fill. The deformations were measured with standard and in-place inclinometers while the porewater pressure response of the in-situ soils was continuously monitored with vibrating wire piezometers. The research further investigated the mobilization of resistance along the length of the rockfill columns with finite element models. The analysis of the model results illustrated the mobilization of shear resistance within the rockfill and the development of stresses across the column cross section. The important engineering characteristics of rockfill are discussed in the context of rockfill column design and the importance of effective compaction is highlighted. The results of this research are used to develop recommendations for rockfill column design, analysis and construction.

Rockfill columns have been used to stabilize the clay riverbanks in the Winnipeg area for over two decades. The construction methods used in Manitoba are uniquely adapted to the local soil conditions. The performance of rockfill columns in Manitoba has generally been satisfactory, except that in some cases significant deformations have occurred during and after construction (Yarechewski and Tallin 2003). This thesis will discuss the full-scale test loading of a riverbank that was stabilized with rockfill columns. The purpose of the test was to measure the load-deformation characteristics of a reinforced slope in order to further the understanding of rockfill column behaviour. Rockfill column technology has evolved from granular shear key methods for stabilizing slopes. The relatively weak and soft lacustrine clay is reinforced with compacted columns of stronger and stiffer limestone rockfill. The research test site is located along the natural banks of the Red River in The City of Winnipeg. The project involved an extensive site investigation, and soils characterization program in preparation for the field test. Eleven columns, 2.1 m in diameter were tested by surcharging the slope with 1920 tonnes of fill. The deformations were measured with standard and in-place inclinometers while the porewater pressure response of the in-situ soils was continuously monitored with vibrating wire piezometers. The research further investigated the mobilization of resistance along the length of the rockfill columns with finite element models. The analysis of the model results illustrated the mobilization of shear resistance within the rockfill and the development of stresses across the column cross section. The important engineering characteristics of rockfill are discussed in the context of rockfill column design and the importance of effective compaction is highlighted. The results of this research are used to develop recommendations for rockfill column design, analysis and construction.

Le barrage poids en enrochement dont l’étanchéité est assurée par un masque en béton sur sa face amont est de plus en plus utilisé aujourd’hui. Son appellation courante est « CFRD1 ». La conception de ce type d’ouvrage est effectuée en suivant des « règles » de conception empiriques. Dans la dernière décennie, plusieurs barrages de type CFRD de grande hauteur ont subi la rupture de leur masque en béton lors de la première mise en eau, montrant ainsi la défaillance de l’approche empirique.Il s’avère alors nécessaire de comprendre les mécanismes physiques mis en jeu, notamment dans le comportement des enrochements sous fortes contraintes. Ce travail de thèse commence par une recherche bibliographique montrant que les fortes contraintes ont un effet important dans la rupture des particules. Cette rupture est influencée par les caractéristiques des particules, l’assemblage et les conditions mécaniques imposées. Une fois les facteurs influençant la rupture de particules identifiés, on s’intéresse aux modèles de comportement tenant compte de ce phénomène. La plupart des modèles étudiés englobent les différents facteurs d’influence dans un ou plusieurs paramètres difficilement identifiables par les essais mécaniques courants. Un modèle proposé dans le cadre de la thermodynamique prenant en compte la distribution de taille des particules a été retenu pour une analyse plus approfondie. Cette analyse aboutit au besoin d’identifier deux phénomènes importants : i) la distribution de tailles de particules suite à la rupture, et ii) la relation entre la rupture de particules et la dissipation d’énergie par frottement. Concernant le premier phénomène, un modèle probabiliste est proposé. Ce modèle tient compte de l’effet de la taille des particules dans la probabilité de rupture. La comparaison des simulations aux résultats expérimentaux pour un sable montre une bonne approximation de la variation des courbes granulométriques. Concernant la relation entre la rupture de particules et la dissipation d’énergie par frottement, des essais triaxiaux ont été analysés thermodynamiquement sur un plan « énergie reçue/énergie rendue » par le système (l’échantillon de sol). Sur ce plan, on étudie le comportement dissipatif sans rupture qui se traduit par une relation linéaire. Les matériaux présentant une rupture des particules montrent une relation différente. Donc, ce plan permet de mettre en évidence la différence entre l’énergie dissipée par le matériau sans rupture et celui avec rupture de grains. Cette différence est associée au travail mécanique dû à la rupture des particules. Différentes conclusions et perspectives sont proposées à ce point pour le développement de modèles de comportement. Une conséquence directe de la dépendance de la probabilité de rupture avec la taille des particules est l’existence d’un effet d’échelle. Une théorie d’effet d’échelle récemment proposée pour l’enveloppe de rupture a été validée dans le cadre des relations contraintes-déformations. Ceci permet de modéliser la réponse contrainte-déformation d’un matériau contenant des particules de taille significative à partir d’un matériau de granulométrie réduite.Finalement, quelques recommandations de modification aux pratiques existantes dans la conception de barrages à masque amont en béton ont été proposées suite à des analyses de modèles en éléments finis. / The concrete face rockfill dam (CFRD) is a type of dam widely constructed nowadays. The design “rules“ for this type of dam have remained totally empirical. During the last decade, several high CFRDs have experienced the cracking of the concrete face during first reservoir impounding. This shows the limitations of the current empirical state of practice. It is important to understand the different physical mechanisms leading to these problems, especially those concerning the mechanical behavior of rockfill materials. Looking for answers to this problem, this thesis starts by a bibliographic research which shows that particles breakage is an important issue on the mechanical behavior of rockfills. The particles breakage phenomenon is affected by the characteristics of individual particles, the packing conditions of the assembly and the imposed mechanical solicitations. Once the factors affecting particles breakage have been identified, we study the constitutive models that have introduced the particles breakage phenomena. Most of them reduce the influence of the different factors to a couple of parameters, which are not easily identifiable through current laboratory tests. One constitutive model proposed on a thermodynamic framework has been retained for a more detailed analysis. This analysis leads to two main topics of interest: i) the description of the grain-size distribution due to particles breakage, and ii) the relationship between dissipation of energy and particles breakage. Concerning the first topic, a probabilistic model is proposed to describe the evolution of particles breakage and the variation of the grain-size distribution curve. This model takes into account the dependency of breakage probability with particles size. The comparison between model response and laboratory experiences shows good agreement for the variation of the grain-size distribution curves.Concerning the relationship between dissipation of energy by friction and particles breakage, several drained triaxial tests have been studied thermodynamically on a plot called “input/output power” of the system (soil sample). On this plot, we study the dissipative behavior without particles breakage, which is characterized by a straight line. Materials experiencing particles breakage show a different form. Therefore this plot allows identifying the difference between the dissipated energy of materials with and without particles breakage. This difference is associated to the particles breakage phenomena. Several conclusions and future works are proposed at this point related to the description of the mechanical behavior of rockfills.A direct consequence of the dependency of particles breakage on the particles size is the existence of a size-scale effect. A theory recently proposed for the shear-strength envelope has been validated for the stress-strain relationships. This allows estimating the stress-strain response of a material with very coarse particles from a material with reduced grain-size distribution.Finally, some recommendations to the current practice of CFRDs design are proposed, based on analyses with finite element models.

China’s government invests in hydropower and explicitly in larger dams. An increasing share of such large dams is globally represented by the type Concrete Face Rockfill Dam (CFRD) due to its safe behaviour at a relatively low construction cost. China is however building them higher than has earlier been done and the lack of worldwide experience requires thorough studies on their behaviour. This report presents a study of the Cihaxia dam that is to be built in the river basin of Huang He, regarding the optimisation of an increased rockfill compaction and its final effects on the settlements of the concrete face. The settlements of a typical cross section of the Cihaxia dam are modelled with a Finite Element Method (FEM) program with varying values for a number of structural properties. The effect of each variation indicates what properties that will have the largest effect on the behaviour of the finished structure and points out where focus shall be laid when optimising the design of a high CFRD. The outcome of the FEM analysis shows the static comportment of the rockfill body and its concrete face at two stages, at the end of construction and after reservoir impoundment. The difference between the two provides data on what effects the water load would cause, which constitute the main displacements of the concrete slabs. It is evident from the obtained results that the most important of the studied parameters alongside with the level of compaction is the geometrical extension downwards and inwards of the zone subject to increased compaction. A higher degree of compaction will always produce a beneficial decrease of settlements but might not always be economically legitimate as costs rapidly increase with higher levels of compaction. An adequate level of a CFRD’s compaction increase can with advantage be determined by comparing different designs regarding the volumetric percentage to compact against the calculated decrease in slab deflection.

Non-overflow flow-through rockfill structures are river engineering elements used to attenuate and delay inflow hydrographs. They represent expedient places to deposit rather enormous quantities of waste rock at mountainous mine sites. Their application has become so common that matters of safety regarding their design have been laid out in Section 8.5 of the Canadian Dam Safety Guidelines (CDA 2007). The research described herein was directed at investigating the different aspects of the hydraulics of these flow-through rockfill structures. In order to assess the potential for an unraveling failure of flow-through rockfill dams, a systematic study of the hydraulic design of these structures was conducted and the non-linear nature of flow through these structures was dealt with using a p-LaPlacian-like partial differential equation. Subsequently, factors of safety against this type of failure are presented for a range of downstream slopes, thus showing the unsafe combinations of embankment slope and particle diameter. Three different index gradients within the toe of such structures were investigated. In this regard, the gradient most suitable for independently computing the height of the point of first flow emergence on the downstream face is examined and a method for independently computing the variation in hydraulic head within that vertical (which allows for the toe of the structure to be isolated) is presented. An additional gradient that allows for the independent estimation of the default tailwater depth is proposed. In order to provide better tools to assess the behavior of these embankments at the toe, laboratory and analytical studies were undertaken. In this regard, the hydraulics associated with the zone of the downstream toe were studied. The depth variation of the seepage-face was computationally modeled, and two approaches for solving the spatially varied flow (SVF) condition problem within the toe region undertaken. The results show that a dual linear variation in depth can be used to good accuracy, without inducing any unrealistic exit gradients in the zone of primary concern with respect to unraveling. It is hoped that these techniques and computational tools provided herein will aid in facilitating the design and assessment of these flow-through rockfill structures.

O crescimento do número de Barragens de Enrocamento com Face de Concreto (BEFC) no Brasil e no mundo e os problemas relacionados com a movimentação da face motivaram o desenvolvimento deste estudo sobre o comportamento mecânico do enrocamento. Geralmente, as previsões dos movimentos diferenciais do maciço são feitas com base nas características de deformabilidade do material. Este trabalho apresenta um estudo de tensão-deformação de enrocamento baseado em análise numérica de uma seção de barragem de grande altura, com o objetivo de definir o estado e as trajetórias de tensões que influenciam os valores dos parâmetros de deformabilidade. Baseado nas condições de carregamento previstas para o maciço de enrocamento, foi desenvolvida a programação e as metodologias dos ensaios laboratoriais. A partir dos resultados destes ensaios foi possível estimar os parâmetros de deformabilidade do enrocamento a serem utilizados em estudos tensão-deformação. / The important growth in the number of Concrete Face Rockfill Dams (CFRD\'s) in Brazil and in the World, and the reported problems related to movements of the concrete slab, motivated the present study of mechanical behavior of the rockfill. In general, predictions of the differential movements within the rockfill mass are based on the deformation properties of the rockfill materials. This work presents a stress-strain study of the rockfill mass, based on a numerical analysis of a high dam cross section, aiming to define the stress state and stress paths that influence the deformability parameters. Based on the loading condition estimated for the rockfill mass, the program and the methodology for the laboratorial tests was set. From these test results it was possible to estimate the deformability parameters of the rockfill to be used in stress-strain analysis.

An impermeable weir constructed across a stream prevents the longitudinal movement of aquatic life and transportation of physical and chemical substances in water, eventually having a negative impact on river environment. However, a rubble mound weir is considered environmentally friendly, since its permeability allows the streamwise migration of aquatic life. This thesis investigates the performance of this type of weir as a water use facility. The particular objective of the investigation is to study the flow mechanism in terms of water surface profile and discharge through the weir. In the study, flow through the rubble mound weir is considered non-Darcian, steady, and one-dimensional. In the analysis, gradually varied open channel flow algorithm is applied to porous medium flow through the rubble mound weir in which laminar and turbulent components of flow are taken into consideration. Unlike previous studies where Stephenson and Wilkins relations were used, in this thesis Forchheimer equation is used. To verify the validity of numerical solution of governing equation based on Forchheimer relation, an experimental investigation is conducted in the laboratory. The experimentally obtained water surface profiles are compared with the numerical results. It is observed that there is a satisfactory agreement between numerical and experimental results. The water surface profiles obtained by numerical solution are further compared with those based on Stephenson and Wilkins relations. It is concluded that the proposed numerical solution technique for the Forchheimer based governing equation may be used in the analysis of flow through, and design of rockfill weirs.

[pt] A mistura asfáltica tem sido utilizada como núcleo impermeável em barragens de enrocamento desde 1930. A partir de então, estudos foram desenvolvidos, buscando uma melhoria das principais características da mistura asfáltica, como a granulometria, o teor de ligante asfáltico e o volume de vazios; permitindo obter um material asfáltico de maior densidade e com permeabilidade inferior a 10-9 cm s. As técnicas e procedimentos construtivos destas barragens apresentam vantagens quando comparadas às barragens já conhecidas. Os materiais do núcleo, as transições e o enrocamento podem ser representados por modelos constitutivos adequados e o comportamento de barragens de enrocamento com um núcleo de concreto asfáltico (BENCA) pode ser reproduzido por ferramentas computacionais. O presente trabalho apresenta um estudo sobre barragens do tipo BENCA, fazendo uso do programa computacional PLAXIS, de elementos finitos. O programa foi calibrado e validado a partir de resultados de medições de campo do Reservatório Megget, construído na Escócia (1981). Um estudo sobre a influência da inclinação do núcleo na resistência e deformabilidade de barragens de enrocamento é apresentado e discutido. Finalmente, apresenta-se a previsão numérica do comportamento da Barragem Foz de Chapecó, que é a primeira barragem construída com esta técnica no Brasil (2010). Os resultados mostraram-se coerentes com o comportamento observado em campo, indicando que o programa e os modelos constitutivos adotados permitem a previsão do comportamento de barragens de enrocamento com núcleo em concreto asfáltico. / [en] Asphalt mixture has been used as impermeable core rockfill dams since 1930. Since then, studies have been developed, searching an improvement of the main characteristics of the asphalt mixture, such as grading, bitumen content and volume ratio, allowing to obtain an asphalt material of higher density and less permeability than 10-9 cm s. Techniques and constructive procedures of these dams have advantages when are compared to known dams. The core materials, rockfill and transitions can be represented by appropriate constitutive models and the behavior of rockfill dams with an asphaltic concrete core (BENCA) can be reproduced by computational tools. The present work presents a study about the type dams (BENCA), using the finite elements computer program PLAXIS. The program was calibrated and validated from results of field measurements of Megget Reservoir, built in Scotland (1981). A study on the influence of core inclination in the strength and deformability of rockfill dams is presented and discussed. Finally, it is presented the numerical prediction of the behavior of the Foz de Chapecó dam that is the first dam built with this technique in Brazil (2010). The results were consistent with the behavior observed in field, indicating that the program and the adopted constitutive models allow the prediction of the behavior of rockfill dams with asphalt concrete core.

[pt] Este trabalho apresenta um estudo experimental do comportamento de maciços de enrocamento, considerando as variações de deformabilidade e resistência provocadas pela alteração da rocha e pelas variações das características intrínsecas do enrocamento, tais como, litologia, granulometria, efeito de escala, umidade e as características das partículas. Foram considerados o enrocamento de basalto da Barragem de Marimbondo, com cerca de 25 anos de alteração natural no campo, o basalto da Pedreira Rio Grande, semelhante ao utilizado na construção de Marimbondo, e o enrocamento de granito da Barragem de Serra da Mesa. Foi elaborada uma metodologia para avaliar a alterabilidade dos materiais de enrocamento, envolvendo ensaios nos materiais intacto e alterados no campo e no laboratório. A alteração no basalto da Pedreira em laboratório simulou, de forma acelerada, a alteração natural do basalto de Marimbondo no campo, sendo realizada através de ensaios de lixiviação contínua e de ciclos de umidade. A lixiviação em corpos de prova de enrocamento foi feita em um equipamento soxhlet de grandes dimensões desenvolvido no Laboratório da PUC-Rio. O comportamento geomecânico dos enrocamentos foi avaliado a partir dos parâmetros mais relevantes das partículas, da rocha e do maciço de enrocamento. A caracterização do enrocamento como meio granular é obtida através de ensaios de compressão unidimensional, de compressão triaxial e de cisalhamento direto, em equipamentos de grandes dimensões. Os resultados mostram que as características intrínsecas do enrocamento têm um papel importante no comportamento do material. Além disso, a alteração do basalto no campo ou no laboratório pode provocar variações significativas dos parâmetros dos materiais de enrocamento, dentre as quais destacam-se a redução da rigidez e resistência do enrocamento, bem como o aumento do desgaste das partículas. Este estudo também indica que a lixiviação no equipamento soxhlet de grandes dimensões mostrou-se adequada para a simulação da alteração natural do basalto em condições aceleradas no laboratório, possibilitando a previsão, a longo prazo, do comportamento de enrocamentos. Considerando um período de 75 anos de alteração natural no campo, o trabalho apresenta a previsão de comportamento do enrocamento de Marimbondo em termos de resistência ao fraturamento, desgaste das partículas, módulo de compressibilidade M, módulo de deformabilidade E , resistência à compressão uniaxial que envoltória de resistência. / [en] This work presents an experimental study of the behavior of rockfills, focusing on the changes caused by rock alteration on the physical and mechanical characteristics, such as, lithology, gradation, deformability and strength. The basalt rockfill used in the construction of Marimbondo dam has been considered, with about 25 years of in situ alteration, together with intact basalts samples from Rio Grande quarry, which is located about 500m downstream from the dam site. The experimental results were also compared with data from tests on granite rockfill samples from Serra da Mesa dam. A methodology is proposed for evaluating the alterability of rockfills, based on laboratory tests on intact and alterated samples. Procedures for fast rock alteration in the laboratory were also developed, attempting to simulate the naturally slow alteration which occurs in the field. The laboratory alteration made use of both continuous leaching and wetting-drying cycles on intact basalt samples. The leaching was imposed in a large scale soxhlet equipment developed in the geotechnical laboratory at PUC-Rio. For the characterization of the geomechanical behavior of the rockfill masses, large scale equipments have been used, such as an confined compression chamber, triaxial compression machine and direct shear box. The results indicate that the intrinsic characteristics play an important role on the rockfill behavior. Furthermore, basalt alteration occurring in the field or in the laboratory may cause a significant decrease on the rockfill s stiffness and strength parameters, as well as an increase on the abrasion characteristics of rockfil s particles. This study also indicates that leaching in large scale equipment is a proper way to simulate the basalt s natural alteration in accelerated laboratory conditions, making possible the long time prevision of the behavior of rockfills. Considering a time period of 75 years of natural alteration in the field, this work presents a quantitative prediction of Marimbondo s rockfill behavior in terms of particle breakage, abrasion resistance, compressibility modulus M, deformability modulus E, unconfined strength qu and strength envelope parameters.