Abstract

This paper provides a framework for studying the combined effect of model uncertainties and uncertainties due to spatial variability on the side friction of drilled shafts in weak carbonate rocks. High quality site investigation data for a site in Dubai-UAE is used to characterize the spatial variability of weak carbonate rocks. Reliability assessments using Monte Carlo simulations indicated that uncertainty due to spatial variability has a notable effect on the reliability index with the effect being related to the pile length. In addition, the calculated reliability indices were found to be strongly affected by whether the median or the 10 th percentile UCS value was used in the deterministic design. The side friction of drilled shafts in weak carbonate rocks is generally computed using data from unconfined compressive strength (UCS) and rock quality designation (RQD). The method proposed by O'Neill and Reese (1999) is one of the commonly used methods for predicting side friction in these rocks. Recent analyses of instrumented pile load tests (Charif et al. 2010) for shafts in weak carbonate rocks indicated that predictions of side friction exhibited significant scatter when compared to actual back calculated side friction values. Based on the results of ten instrumented load tests conducted on shafts with diameters ranging from 0.6m to 1.5m and lengths ranging from 14m to 32m at seven sites across Dubai-UAE, Charif et al. (2010) evaluated the model uncertainty associated with the O'Neill and Reese (1999) model. The model uncertainty was reflected by the average (bias) and coefficient of variation (COV) of the ratio (λ) of predicted to measured skin friction, which were found to be equal to 1.17 and 0.49, respectively. The calculated COV could be considered to be relatively large reflecting uncertainty in the predictions of skin friction in soft rock. In addition to the model uncertainty, the design of sockets in weak carbonate rocks is affected by vertical and lateral spatial variability in the UCS and RQD values. Spatial variability is generally modeled by the mean, standard deviation, and the spatial correlation structure of the relevant rock property and constitutes an additional source of uncertainty in predicting the side friction. There is very limited

Full Text
Published version (Free)

Talk to us

Join us for a 30 min session where you can share your feedback and ask us any queries you have

Schedule a call