Reliability analysis of unsaturated soil slope stability using spatial random field-based Bayesian method
- PDF / 2,151,431 Bytes
- 13 Pages / 595.276 x 790.866 pts Page_size
- 75 Downloads / 208 Views
M. L. Huang I D. A. Sun I C. H. Wang I Y. Keleta
Reliability analysis of unsaturated soil slope stability using spatial random field-based Bayesian method
Abstract Rainfall and water level change are two of the main factors causing failure of reservoir slopes. Thus, soil-fluid mechanics is applied with the fluctuation of groundwater table. However, many of the geotechnical parameters needed for the analysis are highly varied. Spatial random field-based Bayesian method is proposed, which can systematically assimilate prior knowledge, borehole testing data, and long-term monitoring data to obtain posterior distributions. There are three major components of the method. They include unsaturated soil-fluid coupling mechanics, spatial multivariate discretization, and subset Monte Carlo simulation of reliability analysis. The approach is applied to the Shiliushubao slope of the Three Gorges Reservoir area, which is located 9.0 km downstream of the Badong County in Hubei Province, China. Results prove that the updated key geotechnical parameters will quantitatively predict the geohazards of landslide, especially for unsaturated soil slope conditions that suffer an unprecedented heavy rainfall subject to low reservoir water level in the upcoming summer. Keywords Geotechnical parameters . Spatial random fields . Bayesian method . Unsaturated soil slope . Reliability analysis Introduction Heavy rainfall and water level fluctuation are two of the main factors contributing to reservoir landslides (De Vita et al. 1998; Tsai 2008; Zhu et al. 2015; Huang et al. 2016). For example, Vajont landslide was caused by a rising reservoir level and torrential rainfall in Italy (Kilburn and Petley 2003). Many landslides were reactivated by the first impoundment of the Three Gorges Reservoir, China (Wang et al. 2008; Jian et al. 2014). Early warning system of long-term monitoring data has been used to prevent and mitigate geohazards. However, it is difficult to make quantitative predictions on the reactivation of reservoir landslides. Many researchers have been studying the soil-fluid coupling mechanics of unsaturated soil (Huang and Griffiths 2015; Oh and Lu 2015; Vahedifard et al. 2016; Karantgis et al. 2017; Yang et al. 2017a; Yang et al. 2017b; Yang et al. 2018; Wang et al. 2019). At the same time, spatial variability of geotechnical parameters has been widely recognized. A stochastic soil-fluid coupling mechanics is more reasonable since it enables engineers to identify the uncertainty propagation of soil mechanical behavior (Vanmarcke 1977; Phoon and Kulhawy 1999; Baecher and Christian, 2003). Griffiths and Fenton (1993) innovatively proposed a random finite element method. Srivastava et al. (2010) assumed the hydraulic conductivity as a random variable of logarithmic normal distribution into seepage analysis. Cho (2014) discussed the heterogeneous effects of hydraulic conductivity on the failure patterns of unsaturated slopes due to rainfall infiltration. Cai et al. (2017) analyzed the spatial variability of hydraulic conductivity an
Data Loading...
