ABSTRACT
One of the key elements for practical implementation of RBD in geotechnical engineering is the characterization of calculationmodel uncertainty. The calculationmodels used in the analysis and design of geotechnical structures are in general incomplete and inexact as a result of lack of knowledge or simplification formathematical convenience. The objective of this chapter is to present available methodologies for characterizing calculationmodel uncertainty for geotechnical structures. Although the examples given are limited, the methodologies presented are applicable to a wide range of geotechnical structures. First, the general statistical characterization entailing (a) exploratory data analysis, (b) outliers detection and correction of anomalous values, (c) using the corrected data to compute the sample moments (mean, standard deviation, skewness, and kurtosis), (d) verification of the randomness of the model factor, (e) determining the appropriate probability distribution for M and (f) removal of statistical dependencies is presented. This is followed by a presentation of published model factor statistics for geotechnical structures at ultimate and serviceability limit states. It is concluded that significant data on model uncertainty statistics is now available as prior information to enable reliability calibration of partial factors required for semi-probabilistic design approach.
