Reliability-Based Design for the Serviceability Limit State of Axially-Loaded Driven Piles in Ontario Soils

ABSTRACT: For steel driven piles subjected to axial loads, this paper presents a simplified reliability-based design (RBD) approach for the allowable load at the serviceability limit state (SLS). First, load-displacement responses were obtained from 40 piles that were located in Ontario, Canada and tested to plunging failure. Second, the measured resistance was interpreted from the test results with the Davisson failure criterion and was used to normalize the load-displacement curves. Conventional design methods determined the capacity with measurements from standard penetration tests (SPT), and their uncertainty was modeled with the capacity bias, a ratio of the measured to predicted capacity. Third, a simple hyperbolic curve was adopted to represent the pile load-displacement responses, and the hyperbolic model parameters were determined through regression. Afterwards, copula theory was adopted to represent the correlation structure of the hyperbolic model, and Monte Carlo simulations computed the reliability of the driven piles at the SLS. High lumped-load capacity factors were obtained for the SLS since the SPT design method predicted the capacity with a significant amount of variability. Since a variety of soil conditions were encountered, the analysis aims to provide greater insight on the reliability of SLS designs.