Análise de confiabilidade estrutural de pilares mistos de aço e concreto com seção circular preenchida

• Ano de defesa: 2020
• Autor(a) principal: Menezes, Alberto Brigeel Noronha de
• Orientador(a): Nardin, Silvana de
• 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: Confiabilidade estrutura; Erro de modelo; Pilar misto preenchido; Compressão axial
• Palavras-chave em Inglês: Structural reliability; Model error; CFST columns; Axial compression
• Área do conhecimento CNPq: ENGENHARIAS::ENGENHARIA CIVIL::ESTRUTURAS
• Link de acesso: https://repositorio.ufscar.br/handle/20.500.14289/13885

The developed study deals with the analysis of structural reliability of circular concrete filled steel tube columns under concentric axial load. This type of composite column presents high load capacity when compared to columns with other geometries and materials, mainly due to the confinement effect promoted by the steel tube in the concrete core. A structural reliability analysis seeks to evaluate the probability of failure to meet the limit states and was applied in the formulations applicable to the standards ABNT NBR 8800 (2008), EN 1994-1-1 (2004) and AS / NZS 2327 (2017) for axially compressed columns. The objective was to obtain the reliability indexes and sensitivity factors associated with variables that influence the analysis such as steel and concrete strength, dead and live loads and model error for each standard. For this, the starting point was the construction of a large database with ultimate load values. Computational code was elaborated considering the FORM method; this code was used in a parametric study in which several values of resistance and geometric dimensions were used. It was found that ABNT NBR 8800 (2008) was the only standard to reach the target index, as it is a more conservative standard among those available. EN 1994-1-1 (2004) has the lowest dispersion of model error results, but has not reached the target index. In AS / NZS 2327 (2017), however, it presents the lowest average model error, but with a high dispersion of results influenced the reduction of the reliability index. The factor that most influenced the result was the live load, due to its high coefficient of variation. The concrete strength and the evaluation of the model error were also of great influence. The other variables had little effect on the results.