Reducing Uncertainty of Prediction from Empirical Correlations

Reliability based design of driven pile groups using combination of pile driving equations and high strain dynamic pile monitoring

“…It is further noted that Eqs. (9)-(11) for q pm = 0, and when expressed in variances rather than CV's, are formally identical to results of a Bayesian updating approach by Zhang et al [13] between global and regional data. Finally, for unmonitored piles, i.e., when R m is not available, CV m /CV p may be set to a very large value leading to the trivial results of w p = 1, w m = 0 and CV pm = CV p .…”

“…McVay et al [9], Paikowsky [1] and Yoon et al [10] investigate reliability based performance of a series of different static and dynamic methods, while Zhang [5], Zhang et al [11] and Liang and Yang [12] propose the use of dynamic measurements and static load tests for reliability based quality control at the end of the construction phase. In an effort to reduce design uncertainty based on large and possibly heterogeneous load test databases, Zhang et al [13] apply Bayesian updating to incorporate less variable regional and site-specific information. Alternative approaches combining uncertainties of relevant design parameters by propagation through design equations rather than comparing predictions against load test results are taken by Foye et al [14] and Kim et al [15].…”

An Approach to Assess LRFD-Φ from Load Test and Borehole Data In and Outside the Footprint of a Drilled Shaft

“…The simplicity in this is very appealing, however, unless extensive load testing is performed at a single site of interest (Yoon & O'Neill 1997), the method does not offer an explicit possibility to account for site specific data (including the levels of load testing and spatial variability within a site) and the characteristics of a site/foundation have to be matched with a sufficient number of corresponding observations from the past. To overcome some of these problems Zhang et al (2004) propose a Bayesian approach for combining empirical data at different scales (global, regional, site specific). As an alternative, other general approaches have been proposed that evaluate the contributing sources of uncertainty separately and then combine them according to appropriate physical and statistical laws (e.g., Phoon & Kulhawy 1999a& b, Kim et al 2011.…”

Bayesian approach for probabilistic characterization of sand friction angles

“…Bayesian approach has been applied to geotechnical engineering, such as characterization of model uncertainty (e.g., Gilbert and Tang, 1995;Zhang et al, 2004, Zhang et al, 2009aNajjar and Gilbert, 2009), calibration of estimated failure probability in liquefaction evaluation (e.g., Juang et al, 1999) and slope reliability assessment (e.g., Cheung and Tang, 2005), and updating of geotechnical parameters (e.g., Miranda et al, 2009). However, research is rare that incorporates the spatial variability in the Bayesian framework.…”