Publication Abstract

Proceedings-DFI/EFFC 11th International Conference on Piling and Deep Foundations, 2014, Stockholm, Sweden, (DFI)

Analysis of the Uncertainty by the Random Finite Elements Method of a Vertical Load Test of a Piled Raft Founded in Tropical Soil
Cristhian C. Mendoza, Pilot University of Colombia, Colombia; Renato da Cunha, University of Brasilia, Brazil; Bernardo Caicedo, University of the Andes, Colombia, Brazil; Max Barbosa, SOLOTRAT Company, Brazil

This paper shows that a raft alone is not sufficiently appropriate to address the uncertainty of underlying soil variability; therefore, a piled raft system is a better option for mitigating such uncertainty. To study the variability of soil, the behavior of the Piled-Raft Foundation was analyzed with a type of self-drilling micropiles: the "Alluvial Anker" (common name for this pile type in Brasilia city). These micropiles have been used in the city and have the advantage that the drill pipe also serves as reinforcement. The studied parameters belong to an elastoplastic constitutive model for strain-stress analysis of soil; variability of these parameters can be intrinsic and/or epistemic. Simulations were performed for a hypothetical Piled-Raft Foundation subjected to vertical loading using the random finite element method. The average parameters of the constitutive model were obtained through laboratory tests. The simulations of the load test were compared with results of a real scale experimental test conducted exclusively for this work in the field. It is analyzed the relative importance of the variability of some geomechanical parameters and their variation with depth is considered. In addition, it is discussed a comparison between the uncertainty of the geomechanical parameters and the uncertainty of the results of the vertical loading test. Finally, this paper shows that a decrease of the safety factor may occur when taking into account the variability of the parameters.

 article #1927; publication #100 (IC-2014)