Análise da influência do uso de geogrelhas em aterros sobre solos moles utilizando os métodos do equilíbrio limite e elementos finitos

Abstract

The settling of Brazil began from the coast of the territory, a place whose predominant soil is soft and has low support capacity. In consequence, soil improvement techniques that could make these places safe for the population to occupy over time were developed, with the use of geosynthetics to reinforce soft soils being one of the most widespread. The use of software based on the Limit Equilibrium Method (LEM) and the Finite Element Method (FEM), such as Slide and Plaxis, has become increasingly common, as they allow us to computationally simulate field situations, which facilitates analysis before, during and after construction. The Embankment 2, presented by Huat (1995), was studied with the intention of verifying how the variation in the stiffness modulus (J) of the reinforcement, the thickness of the soft clay layer (D) below the embankment and the form of application of the geosynthetic would impact the safety factor (SF) upon embankment failure and also the measured settlement. These simulations were carried out using two different calculation methods, using the softwares Slide and Plaxis 2D, in order to compare the approaches of the two programs. Most of the analyses were carried out based on the method of Rowe and Soderman (1985), which recommends the use of the undrained shear strength (Su) as constant throughout the entire layer of soft clay below the embankment. However, with the aim of comparing different methods, the method of Hinchberger and Rowe (2003) was also used, in which Su can vary throughout the soft soil layer. Finally, in order to validate the model built in the Plaxis 2D software, experimentally measured data by Magnani (2006) and Magnani et al. (2009) were used and compared with the results obtained by modeling the embankment by Magnani (2006) in Plaxis. The results demonstrate that both the LEM analysis and the FEM approach converge to similar answers in most analyses. Furthermore, the methods of Rowe and Soderman (1985) and Hinchberger and Rowe (2003) presented similar results for the analyses carried out, demonstrating how it would be most interesting to use the first, simpler and more widespread in Geotechnics. Validation of the model through the experimental embankment by Magnani (2006) showed that it is representative.