Abstract
The uncertainties involved in geotechnical design are considered. Using a formulated approach that accounts for model uncertainty, variable correlation and spatial variability, failure probabilities associated with alternative designs of retaining walls are calculated. Application of this approach is also demonstrated through a case study that is based on a real construction project. The study provides evidence that failure probabilities are significantly influenced by considerations regarding the spatial variability of parameters and the model uncertainty, as well as the adopted statistical modelling of the variables. Furthermore, the study facilitates observations regarding the calibration of codes of practice for the design of retaining walls.
Acknowledgements
The work presented herein formed part of the doctoral research of the first author and was carried out at Heriot-Watt University, Edinburgh. The authors would like to thank Dr John Oliphant, previously of Heriot-Watt University, for his contribution to this study.
Notes
First published online 22 August 2005