Influence of Spatially Variable Side Friction on Single Drilled Shaft Resistance and LRFD Resistance Factors

Abstract

Load and resistance factor design (LRFD) is a method that aims at meeting specified target reliabilities (probabilities of failure) of engineered systems. The present work focuses on ultimate side friction resistance for axial loads on single cylindrical drilled shaft foundations in the presence of spatially variable rock/soil strength. Core sample data are assumed to provide reliable information about local strength in terms of mean, coefficient of variation and spatial correlation structure (variogram) at a site. The geostatistical principle of support up-scaling is applied to quantify the reduction in variability between local strength and the average ultimate shaft side friction resistance without having to recur to lengthy stochastic finite difference/element simulations. Site and shaft specific LRFD resistance factors (Φ values) are given based on the assumption of lognormal load and resistance distributions and existing formulas recommended by the Federal Highway Administration. Results are efficiently represented in dimensionless charts for a wide range of target reliabilities, shaft dimensions, and geostatistical parameters including nested variograms of different types with geometric and/or zonal anisotropies. Field data of local rock strength is used to demonstrate the method and to evaluate the sensitivity of obtained resistance factors to potentially uncertain variogram parameters.