[pt] Historicamente é sabido que muitas das rupturas ocorridas em barragens ou taludes naturais podem ser atribuídas ao fenômeno da liquefação de solos arenosos, causada pela ação de carregamentos dinâmicos gerados por explosão ou, mais freqüentemente, por terremotos. Quando liquefação ocorre, um súbito aumento da poropressão faz decrescer a resistência ao cisalhamento do solo e sua capacidade de suportar pontes e edifícios é significativamente reduzida. Solo liquefeito também pode exercer altas pressões sobre estruturas de contenção, causando inclinações da mesma e movimentos do solo que, por sua vez, originam recalques e destruição de estruturas localizadas sobre a superfície do terreno. O termo liquefação tem sido empregado para descrever fenômenos relacionados, que produzem efeitos similares, mas cujos mecanismos de formação são bastante diferentes. Estes fenômenos são modernamente descritos como fluxo por liquefação e mobilidade cíclica. Fluxo por liquefação é o fenômeno no qual o equilíbrio estático é destruído por carregamentos estáticos ou dinâmicos em um depósito de solo com baixa resistência residual. Colapsos causados por fluxo por liquefação são freqüentemente caracterizados por movimentos rápidos e de grande extensão. Mobilidade cíclica, por outro lado, é causada por carregamentos cíclicos em solos sob tensões cisalhantes estáticas inferiores à resistência ao cisalhamento do material, com as deformações desenvolvendo-se gradualmente. A execução de barragens de rejeito usando a técnica de construção à montante pode levar à ocorrência de liquefação estática se a velocidade de construção for suficientemente alta para causar o desenvolvimento de excessos de poropressão. A resposta de liquefação pode ser observada em amostras de solo fofo quando as tensões de cisalhamento atingem um pico seguido por uma fase de amolecimento aparente que, no caso de carregamento não drenado, é associado com a tendência do material em contrair de volume. Para alguns estados iniciais, a parte descendente da resposta do material pode ser seguida por uma fase crescente que se atenua à medida que o estado permanente ou crítico for atingido. Nesta dissertação, a modelagem da resposta de liquefação por carregamento estático, foi feita considerando-se modelos propostos na literatura por Juárez-Badillo (1999b) e Gutierrez e Verdugo (1995). Este último, principalmente após modificação introduzida pela dependência de alguns parâmetros em relação à tensão de confinamento, levou a resultados bastante satisfatórios nas retroanálises consideradas, apesar da relativa simplicidade da formulação. / [en] Historically it is known that many failures in earth dams and natural slopes can be attributed to the phenomenon of sand liquefaction, caused by dynamic loads generated by earthquake shaking or other rapid loading, such as blasts. When liquefaction occurs, the strength of the soil decreases and its ability to support foundations for buildings and bridges is significantly reduced. Liquefied soil can also exerts higher pressure on retaining walls, which can cause them to tilt or slide, yielding settlement of the retained soil with risks of destruction of structures on the ground surface. The term liquefaction has actually been used to describe a number of related phenomena, which produce similar effects but whose mechanisms are quite different in nature. These phenomena can be divided into two main categories: flow liquefaction and cyclic mobility. Flow liquefaction is a phenomenon in which the static equilibrium is destroyed by static or dynamic loads in a soil deposit with low residual strength. Failures caused by flow liquefaction are often characterized by large and rapid movements. Cyclic mobility, on the other hand, is a liquefaction phenomenon triggered by cyclic loading, occurring in soil deposits with static shear stresses lower than the soil strength. Deformations due to cyclic mobility develop incrementally because of static and dynamic stresses that exist during an earthquake. The rising of tailing dams using the upstream construction technique can lead to static liquefaction failure if the rate of construction is sufficiently high to cause excess pore pressure to develop in the tailings. The liquefaction response is observed for loose specimens when the shear stress exhibits a peak followed by a phase of apparent softening that, in undrained loading, is associated with the tendency of the material to contract (densify). For some initial loading states, the descending part of the response is followed by an increasing part, leveling-off eventually when the material reaches the final, critical (steady) state. In this thesis, the modeling of the phenomenon of static liquefaction is carried out considering the constitutive models proposed in the literature by Juárez-Badillo (1999b) and Gutierrez & Verdugo (1995). The latter, mainly after introducing the assumption that some material parameters are stress dependent and not simple constants, as in the original version, produced good matching between experimental and predicted results, in spite the simplicity of the mathematical formulation.

<div>Underground structures must be able to support static overburden loads, as well as to accommodate additional deformations imposed by seismic motions. Progress has been made in the last few years in understanding the soil-structure interaction mechanisms and the stress and displacement transfer from the ground to the structure during a seismic event. It seems well established that, for most tunnels, the most critical demand to the structure is caused by shear waves traveling perpendicular to the tunnel axis. Those waves cause distortions of the cross section (ovaling for a circular tunnel, and racking for a rectangular tunnel) that result in axial forces (thrusts) and bending moments. While all this has been well-studied for structures placed in linear-elastic ground, there is little information regarding the behavior of buried structures placed in nonlinear ground, especially under undrained conditions, i.e., when excess pore pressures generate and accumulate during the earthquake.</div><div><br></div><div><div>Two-dimensional dynamic numerical analyses are conducted to assess the seismic response of deep circular tunnels located far from the seismic source, under drained or undrained loading conditions. It is assumed that the liner remains elastic and that plane strain conditions apply. </div><div> A new cyclic nonlinear elastoplastic constitutive model is developed and verified, to simulate the nonlinear behavior and excess pore pressures accumulation with cycles of loading in the ground. The results of the numerical analyses show negligible effect of input frequencies on the normalized distortions of a tunnel for input frequencies smaller than 5 Hz (the distortions of the tunnel are normalized with respect to those of the free field); that is, for ratios between the wavelength of the seismic input and the tunnel opening larger than about eight to ten. The results also show that undrained conditions, compared with drained conditions, tend to reduce deformations for flexible liners and increase them for stiffer tunnels, when no accumulation of pore pressures with cycles of loading is assumed. However, when pore pressures increase with the number of cycles, the differences in distortions between drained and undrained loading are reduced, i.e., the normalized distortions increase for flexible and decrease for stiff tunnels, compared to those with drained conditions. </div></div><div><br></div><div><div>Undrained loading produces larger thrust in the liner than drained loading for stiff tunnels with flexibility ratio F ≤ 2.0.</div><div>For more flexible tunnels with F > 2.0, the behavior is the opposite, i.e., smaller axial forces are obtained for undrained loading than for drained loading. Including excess pore pressure accumulation does not introduce significant changes in the axial forces of the liner, irrespective of the flexibility of the tunnel, compared to those obtained from undrained loading without pore pressure accumulation.</div><div>The drainage loading condition (drained or undrained) or the magnitude of the free-field excess pore pressures during undrained loading do not affect the normalized bending moments for flexible tunnels, with F ≥ 2. For stiffer tunnels, with F < 2, the normalized bending moments increase from drained to undrained loading, and with the free field excess pore pressures.</div></div><div><br></div><div><div>It is found that the tunnel’s response is determined by the load on the liner, or by the distortions of the cross section, depending on the flexibility ratio. For stiff structures, with F ≤ 2.0, important axial forces and bending moments are produced in the structure, with larger magnitudes for the undrained case; while the distortions of the cross section are very small. When the tunnel becomes more flexible, the loading on the liner decreases, but the distortions of the cross section start to be important. For flexible structures with initial F ≥ 10 (for the cases investigated), the performance is largely determined by the distortions of the cross section, while the axial forces and bending moments are almost negligible. Such distortions are drastically affected by the drainage loading condition and by the magnitude of pore pressures in the free field. </div></div><div><br></div>

[pt] O fenômeno de liquefação continua sendo um dos temas mais importantes, complexos e controversos da engenharia geotécnica, sendo a liquefação dinâmica, causada por terremotos, o maior contribuinte de risco sísmico urbano em vários países andinos. O movimento causa incrementos da poropressão que reduz a tensão efetiva e conseqüentemente a resistência ao cisalhamento de solos arenosos. O presente trabalho de pesquisa apresenta e compara algumas das metodologias existentes para a avaliação do fenômeno de liquefação dinâmica, variando desde o método semi-empírico de Seed-Idriss para estimativas do potencial de liquefação até a execução de algumas análises numéricas, no contexto dinâmico, através do programa computacional FLAC 2D v.5. Um estudo de caso, para uma comparação dos diversos procedimentos, consiste na análise dinâmica de uma pilha de lixiviação de minério de cobre, situada em região de atividade sísmica no Peru que, devido a problemas no sistema de drenagem interna, resultou com um alto nível de saturação que poderia colocar sua estabilidade em risco devido à possibilidade de liquefação sob carregamento sísmico. Dos resultados obtidos nessas análises, pode-se afirmar que o método semi-empírico de Seed-Idriss, para determinação do potencial de liquefação, compara-se bastante bem com modelos mais sofisticados baseados em análises dinâmicas empregando tanto o modelo constitutivo de Finn quanto o modelo constitutivo elastoplástico UBCSAND. / [en] The phenomenon of liquefaction is still one of the most important, complex and controversial subjects of the geotechnical engineering, being the dynamic liquefaction, caused by earthquakes, the major contributor to urban seismic risks in several Andesian countries. The shaking increases the pore water pressure which reduces the soil effective stress and, therefore, the shear strength of sandy soils. This dissertation presents and compares some of the proposed methodologies to evaluate the phenomenon of dynamic liquefaction, ranging from the semi-empirical method of Seed-Idriss to estimate the liquefaction potential to the execution of some numerical analyses, within the dynamic context, through the computational program FLAC 2D v.5. The case study, for the comparison among the several procedures, consists of a copper ore heap leach, situated in a high seismic activity zone in Peru, that experimented high levels of saturation, due to problems in the internal drainage system, that could put in risk the stability of the leach pad to the possibility of liquefaction under seismic loading. From the results of such analyses, one can say that the semi-empirical method proposed by Seed- Idriss for the determination of the liquefaction potential compares quite well with outputs from more sophisticated numerical analyses based on dynamic studies that incorporate either the Finn s or the elasto-plastic UBCSAND constitutive models.