Modeling and Simulating Load Transfer of Drilled Shafts in Shales

When conducting probabilistic analyses using service limit state design for deep foundations, modeling and simulating load transfer data are essential. Before the research presented herein, no quantified uncertainty of load transfer for drilled shafts in shales (cohesive intermediate geomaterials) was available nor were there methodologies for incorporating and simulating the variability and uncertainties of a load transfer model for probabilistic analyses. This paper features a study where load transfer data were analyzed to find the best fit model that reflects data and their variability and uncertainty. The load transfer data are side and tip resistance of individual drilled shafts founded in shales. Intensive nonlinear regression analyses were performed using constant standard deviation and constant coefficient of variation assumptions with four different functional models and two approaches to carry out the analyses. Accordingly, two methods to simulate the load transfer data using Monte Carlo simulation are proposed, which incorporate the uncertainty of the models in probabilistic analyses. Recommendation was made to use non-weighted least square regression and collective approach of analysis to analyze the load transfer data. Quantified variability of the load transfer data, as well as the method to simulate the load transfer data, is now available for probabilistic analysis for service limit state of drilled shafts in shales.