Modelagem numérica de fundações por estacas Mega em solos colapsíveis

• Ano de defesa: 2025
• Autor(a) principal: Dias, Fabrício Jerônimo Gonzalez
• Orientador(a): Souza, Tiago de Jesus
• Tipo de documento: Dissertação
• Tipo de acesso: Acesso aberto
• Idioma: por
• Instituição de defesa: Universidade Federal de São Carlos; Câmpus São Carlos
• Programa de Pós-Graduação: Programa de Pós-Graduação em Engenharia Civil - PPGECiv
• Palavras-chave em Português: Reforço de fundações; Estacas Mega; Modelagem numérica; Elementos finitos; Solo colapsível
• Área do conhecimento CNPq: ENGENHARIAS::ENGENHARIA CIVIL::GEOTECNICA; ENGENHARIAS::ENGENHARIA CIVIL::GEOTECNICA::FUNDACOES E ESCAVACOES
• Link de acesso: https://hdl.handle.net/20.500.14289/22876

Mega pile foundations are widely used in underpinning works for structures affected by settlement problems. When installed in collapsible soils—whose behavior is strongly influenced by soil moisture—these piles may experience significant losses in load-bearing capacity under saturation conditions. The assessment of load capacity under different moisture states, traditionally performed through static load tests, is technically advisable but entails high costs and long timeframes, which may restrict the understanding of foundation behavior and result in either overly conservative or oversized solutions. In this context, numerical modeling emerges as a promising tool for the early prediction of Mega pile performance in collapsible soils. This study aimed to analyze the behavior of this type of foundation in collapsible soil through two-dimensional numerical modeling carried out using PLAXIS 2D. Experimental data obtained from static load tests on piles under natural moisture and flooded soil conditions were employed. Numerical simulations were conducted with two distinct constitutive models: Mohr-Coulomb and Hardening Soil. The results indicated that, for triaxial laboratory tests under saturated conditions, the Hardening Soil model achieved better convergence with the experimental results, whereas for pile load tests, the Mohr-Coulomb model showed greater agreement. As a complement to the numerical analyses, the semi-empirical methods proposed by Aoki (1989) and Randolph & Wroth (1978) were evaluated. The investigation identified matric suction as the main variable responsible for the differences in strength between natural and saturated states. The methodology adopted for predicting the performance of Mega piles in collapsible soils, although subject to certain limitations, proves to be a relevant tool for the enhancement of geotechnical design in highly complex contexts.