OasisLMS

Abstract: This paper presents a method for calculating drilled shaft resistance using both random variable (RV) and random field (RF) approaches to account for soil variability. The study focuses on the Circunvalacion Norte viaduct project in Costa Rica, which includes 58 piles and 426 drilled shafts. Soil characterization is based on a comprehensive dataset from 138 pressuremeter (PMT) tests. Reliability analysis is performed using Monte Carlo Simulation (MCS), and the spatial variability of the soil is modeled through the RF approach with Karhunen– Loè ve expansion. A comparison of deterministic and probabilistic analyses reveals significant discrepancies between the two approaches. While deterministic calculations, commonly used in conventional practice, suggest that the proposed drilled shaft lengths may be adequate, probabilistic analyses using RV and RF models show a higher risk, with an unacceptable probability of failure under variable terrain conditions. The paper further examines the sensitivity of RF models to changes in correlation lengths and demonstrates how this parameter affects predicted shaft resistance. These findings underscore the importance of considering soil variability and the inherent uncertainties in geotechnical design. The study emphasizes the need for incorporating probabilistic methods to more accurately account for uncertainties in soil properties. By applying reliability analysis, the study provides a more robust framework for evaluating drilled shaft performance, leading to safer and more resilient geotechnical designs.

Ing. BERNARDO JIMENEZ; Engr. Dr. Adrian Torrico Siacara

Annual Conference Proceedings