Reliability-based design of rock slopes - A new perspective on design robustness

Lei Wang, Jin Hung Hwang, C. Hsein Juang, Sez Atamturktur

Research output: Contribution to journalArticlepeer-review

66 Scopus citations

Abstract

Traditional reliability-based rock slope designs, in which the lowest-cost design is selected from all designs meeting target reliability requirements, are often sensitive to variations in noise factors such as rock shear properties. Consequently, a design that was initially judged acceptable may not satisfy reliability requirements if the variation of rock properties has been greatly underestimated. The authors present a Robust Geotechnical Design (RGD) approach for purposes of addressing this dilemma, by considering the robustness explicitly in the design process. In the context of rock slope design, this proposed RGD approach aims to make the response (i.e., failure probability) of a rock slope system insensitive to, or robust against, the variation of rock shear properties by adjusting design parameters (i.e., such as slope angle and height). Compared to traditional reliability-based design, the RGD approach adds design robustness as one of its design objectives. Thus, multi-objective optimization, considering both cost and robustness, is needed to select optimal designs in the acceptable design space where safety is guaranteed by a constraint on the reliability. In this paper, the concept of the Pareto Front, a collection of optimal designs that reflect the trade-off between cost and robustness, is implemented in the RGD approach. The proposed RGD approach is demonstrated with an example of a rock slope design.

Original languageEnglish (US)
Pages (from-to)56-63
Number of pages8
JournalEngineering Geology
Volume154
DOIs
StatePublished - Feb 28 2013

All Science Journal Classification (ASJC) codes

  • Geotechnical Engineering and Engineering Geology
  • Geology

Fingerprint Dive into the research topics of 'Reliability-based design of rock slopes - A new perspective on design robustness'. Together they form a unique fingerprint.

Cite this