Incorporating stratigraphic boundary uncertainty into reliability analysis of slopes in spatially variable soils using one-dimensional conditional Markov chain model

• Published in: Computers and geotechnics Vol. 118; p. 103321
• Main Authors: Liu, Lei-Lei, Cheng, Yung-Ming, Pan, Qiu-Jing, Dias, Daniel
• Format: Journal Article
• Language: English
• Published: New York Elsevier Ltd 01.02.2020; Elsevier BV
• Subjects: Boreholes; Boundary conditions; Computer simulation; Conditional Markov chain; Dimensional stability; Markov analysis; Markov chains; Mathematical models; Monte Carlo simulation; Probability theory; Reliability; Reliability analysis; Reliability aspects; Slope; Slope stability; Soil conditions; Soil investigations; Soil layers; Soil properties; Soil stability; Soils; Spatial variations; Spatially variable soils; Stability analysis; Statistical analysis; Statistical methods; Stratigraphic boundary uncertainty; Stratigraphy; Uncertainty; Uncertainty analysis; Variability; Slope stability; Reliability analysis; Spatially variable soils; Conditional Markov chain; Stratigraphic boundary uncertainty
• DOI: 10.1016/j.compgeo.2019.103321
• ISSN: 0266-352X

Slope stability analysis based on different stratigraphic boundary conditions may have significant differences in terms of factor of safety (FS) and the location of the critical slip surface. Therefore, traditional reliability analysis of layered slopes that assumes the stratigraphic boundary between two soil layers being a deterministic line or surface can be misleading. This study aims to investigate the influence of the stratigraphic boundary uncertainty (SBU) arising from limited site investigation data on slope stability analysis with the consideration of soil spatial variability. A modified one-dimensional conditional Markov chain model is proposed to model the SBU. Cholesky decomposition technique is subsequently utilized to simulate the inherent spatial variability of soil properties based on the simulated stratigraphic boundary within the framework of Monte Carlo simulation. A sample example on a two-layered soil slope with limited number of boreholes shows that the proposed method can well simulate the stratigraphic boundary based on limited borehole information. The statistics of FS and probability of failure are found not to increase or decrease monotonically with the number of boreholes, but can converge to the correct results if the number of boreholes increases. Moreover, the conventional reliability analysis with an implicit assumption of deterministic stratigraphic boundary condition can significantly overestimate the reliability of the slope, but this overestimation decreases with the increase in the scale of fluctuation.