Cost-effective framework for simplified geotechnical Reliability-Based Design

Abstract

Reliability-Based Design (RBD) is known to provide a rational basis for incorporating uncertainties in the design environment explicitly into geotechnical design. In principle, RBD should be able to translate site investigation efforts into design savings. However, the constant partial factor approach prohibits such translation. This study shows that the combination of two recent developments made by the authors allows such translation. The first development is the multivariate probability distribution of soil parameters. Such a multivariate distribution provides a solid basis that can translate site investigation efforts into smaller parameter variability, e.g. smaller Coefficient Of Variation (COV). The second development is called the Quantile Value Method (QVM). The QVM adopts constant quantile values as the design values, so the resulting partial factors scale with the parameter COV-when parameter COV is small, partial factors are closer to one. The combination of the above two novel developments is a cost-effective framework that is able to translate site investigation efforts into smaller parameter COVs and in turn translate into partial factors that are closer to one. © 2014 Taylor & Francis Group, London.