Effect of Geological Uncertainty on the Shaft Resistance Prediction and Reliability of Piles Driven in Multi-Layered Geomaterials

Abstract

This study proposes a method to quantify the effect of geological uncertainty on the predicted shaft resistance and the reliability of piles driven into multi-layered geomaterials. Several geological predicted surfaces with different levels of geological uncertainties are simulated through multiple sets of boreholes using the multinomial categorical prediction approach for the spatial Markov chain. As the number of strategically placed boreholes increases, the geological uncertainties decrease. Respective ground predicted surfaces are also simulated from these geological predicted surfaces using universal kriging and conditional simulation. The shaft resistances of the piles driven at the site are estimated for the multiple geomaterials layers using geotechnical prediction equations specified for the geomaterial type. The design reliability for the shaft resistance obtained from the various geological, ground, and geotechnical predictions is thereafter accessed through its probability of failure evaluated using the Monte Carlo simulation. This methodology is applied to the Lemhi River bridge project site in Idaho. The results illustrate the influence of geological uncertainty on the design of piles driven into multi-layered geomaterials and further help in determining an optimal site investigation plan that will improve pile design.