Non-intrusive reliability analysis of multi-layered slopes using strength reduction FEM

Te Xiao, Dianqing Li*, Chuangbing Zhou, Kok Kwang Phooh

*Corresponding author for this work

Research output: Contribution to journalJournal Articlepeer-review

16 Citations (Scopus)

Abstract

A non-intrusive reliability analysis method has been highly valued by engineers due to its practicability with utilizing commercial finite element software. The method based on strength reduction finite element method(FEM)is proposed to evaluate the reliability of multi-layered slopes, which considers spatial variability of soil properties, employs commercial finite element software to calculate the safety factor, and uses stochastic response surface method(SRSM)to perform efficient slope reliability analysis. The results indicate that the proposed method can not only take into account the influence of spatial varying soil properties on the slope reliability, but also automatically search for the non-circular critical slip surface, which provides an effective way for solving reliability problems of multi-layered slopes with spatially variable soil properties. Secondly, the modified probabilistic collocation method can improve the calculation accuracy of odd order SRSMs. Furthermore, it is of importance for slope stability analysis with no assumption relating to the geometry of critical slip surface a prior, since the circular slip surface may overestimate the slope reliability. Additionally, Sobol indices can effectively rank sensitivities of input random variables, and provide a reference for the optimal design of slopes.

Original languageEnglish
Pages (from-to)718-732
Number of pages15
JournalYingyong Jichu yu Gongcheng Kexue Xuebao/Journal of Basic Science and Engineering
Volume22
Issue number4
DOIs
Publication statusPublished - 1 Aug 2014
Externally publishedYes

Keywords

  • Non-intrusive analysis
  • Random field
  • Slope reliability
  • Spatial variability
  • Strength reduction FEM

Fingerprint

Dive into the research topics of 'Non-intrusive reliability analysis of multi-layered slopes using strength reduction FEM'. Together they form a unique fingerprint.

Cite this