Efficient reliability updating of slope stability by reweighting failure samples generated by Monte Carlo simulation. (September 2015)

• Record Type: Journal Article
• Authors: Li, Dian-Qing; Zhang, Fu-Ping; Cao, Zi-Jun; Zhou, Wei; Phoon, Kok-Kwang; Zhou, Chuang-Bing
• Is Part Of: Computers and geotechnics. Volume 69(2015)
• Page Start: 588
• Page End: 600
• Publication Date: 2015-09
• Subjects: Slopes -- Uncertainty -- Probability of failure -- Reliability updating -- Monte Carlo simulation; Engineering geology -- Data processing -- Periodicals; Soil mechanics -- Data processing -- Periodicals; Rock mechanics -- Data processing -- Periodicals; 624.1510285
• DOI: 10.1016/j.compgeo.2015.06.017
• Languages: English
• ISSNs: 0266-352X

Abstract

Abstract: Direct Monte Carlo simulation (MCS) could be prohibitively expensive for slope reliability updating problem with relatively low-probability level. This paper aims to propose an efficient slope reliability updating method based on the direct MCS. First, the direct MCS is briefly introduced. Thereafter, a weighted approach based on the direct MCS is proposed for slope reliability updating. Furthermore, the implementation procedure and the flow chart of the proposed method are presented. Two illustrative examples are investigated to demonstrate the capacity and validity of the proposed method. The results indicate that the proposed method can properly utilize the information contained in failure samples generated by performing direct MCS based on the baseline distribution, which is much more efficient than repeatedly performing direct MCS as the probability distributions of soil properties are updated. The proposed method can solve the slope reliability updating problems involving multiple uncertain parameters and implicit performance functions. The accuracy of the proposed method depends on the failure samples generated from the baseline distribution. When the failure samples cannot completely cover the failure region underlying the reliability updating problem, the proposed method should be used with caution. The coefficient of variation (COV) of probability of failure from the proposed approach is derived and used as a measure of the accuracy in updated probability of failure. In geotechnical engineering practices, the updated COV of an uncertain parameter is often smaller than its baseline COV. In this case, the proposed method can reasonably update the probability of failure.