Optimization of site investigation program for reliability assessment of undrained slope using Spearman rank correlation coefficient

Site investigation programs (e.g., boreholes) are crucial in characterizing soil properties and stratigraphic con-figurations. However, the traditional borehole patterns are generally of equally spaced distribution for the slope design, and the locations and total number of boreholes are considerably determined depending on engineers' experience, which may lead to cost-inefficient geotechnical design, especially considering the soil spatial vari-ability. To address this dilemma, this paper presents a Spearman rank correlation coefficient-based scheme to optimize site investigation in slope design, where both locations and total number of boreholes are optimized. Conditional random field simulations are performed to consider the effect of the borehole data on the estimation of the soil property distribution. The superiority of the proposed method to the traditional method is illustrated by a comparison study in an undrained slope example. In this example, the accuracy of the characteristics of the slope (i.e., the factor of safety, location of slip surface, and sliding volume), robustness of the estimated char-acteristics of the slope, and risk reduction are examined. The comparison results show the effectiveness of the proposed method in accurately estimating the characteristics of the slope without prior knowledge about the slip surface, since the slip surface is unknown for most practical cases prior to the site investigation. The most robust estimate results and risk reduction are obtained using the proposed method. This study can also provide useful references to build an adaptive unequally spaced borehole pattern in practice.

Automated summary

This paper presents a scheme based on Spearman rank correlation coefficient to optimize site investigation in slope design, optimizing the locations and total number of boreholes. Conditional random field simulations are performed to consider the effect of borehole data on the estimation of soil property distribution. The proposed method is effective in accurately estimating slope characteristics and improving the ability to withstand risks.