Contribuição à confiabilidade de estacas cravadas através de um estudo de caso com aplicação da teoria Bayesiana.

• Autor(a) principal: Cabral, Eduardo Vidal
• Data de Publicação: 2008
• Tipo de documento: Dissertação
• Idioma: por
• Título da fonte: Biblioteca Digital de Teses e Dissertações da UERJ
• Texto Completo: http://www.bdtd.uerj.br/handle/1/11558
• topic: Deep Foundations; Reliability; Case study; Fundações profundas; Confiabilidade; Estudo de caso; CNPQ::ENGENHARIAS::ENGENHARIA CIVIL

In several branches of Soil Mechanics, particularly on foundation design, engineers are faced with a series of uncertainties. Some uncertainties are due to local soil variability, loading conditions, time effects, differences in construction processes, errors in soil borings that may directly influence the bearing capacity of the foundation during static loading or after its driving, among others. The present research aimed at the adaptation of a procedure originally conceived for offshore piles for use also in onshore piles. It consists in updating the soil resistance estimated during driving, based on driving records. In this approach the updating is performed by the application of Bayesian analysis, assuming that the parameters of the probabilistic distribution are random variables. The uncertainties of the parameters are modeled by a-priori and a-posteriori distributions. The a-posteriori distribution is calculated by updating the a priori distribution, using a maximum likelihood function which contains the information obtained from available driving records. The procedure is applied to piles from an extensive piling of a construction located in the West Zone of Rio de Janeiro City. The updated estimates are then compared to results of dynamic loading tests. Different practical applications can arise by using this procedure, as a better selection of piles for loading tests among those with the lowest updated bearing capacity or the greatest uncertainty in the updated predictions. The extension of this study to different pile types in soil profiles of distinct nature may lead to the development of more adequate rules for execution control, thus contributing to the optimization of future deep foundations design.