Simple and efficient geotechnical reliability-based design based on transformed limit state functions by target reliability-oriented sampling

Abstract

This paper develops a novel reliability-based design (RBD) approach from a new perspective. The approach is relatively simple as it does not need direct reliability calculations for different designs and gives the critical design satisfying the target reliability requirement without involving a trial-and-error procedure. It specifies the critical design, which is unknown, as the one that achieves the limit state at its corresponding design point given a prescribed target reliability requirement. This is inverse to geotechnical reliability calculations based on the design point where the design point is determined for calculating the reliability, which is unknown, given a design. From this inverse perspective, geotechnical RBD is transformed into determination of a design point given a target reliability level and its corresponding design. In this study, this is accomplished by a target reliability-oriented sampling algorithm and transforming the original limit state function (LSF) to represent the design parameter as a function of uncertain parameters. The proposed approach is illustrated by three design examples. Results show that the proposed approach provides accurate design outcomes in a cost-effective way. It provides a useful complementary tool to existing RBD methods due to its simplicity and efficiency.