The results of probabilistic analysis of simple and layered slopes with linearly increasing (mean) undrained shear strength with depth, and spatial variability using the 2D non-circular Random Limit Equilibrium Method (RLEM) are presented. For the case of simple slopes, the results of the circular RLEM approach and the Random Finite Element Method (RFEM) are also presented and are compared to the results of the non-circular RLEM approach. For the case of simple slopes, it is shown that the non-circular RLEM approach gives higher values of probability of failure compared to circular RLEM and RFEM. For the cases with mean value of factor of safety greater than one, considering spatial variability reduces probability of failure.
The combined availability of topographical, geological, structural, hydrogeological and monitoring data is rapidly increasing. Technology and software advances allow the real time incorporation of this data across various software platforms. This paper describes the back-analysis of a 70 m high, pit slope failure of an open pit gold mine in the Dominican Republic, using data from aerial photogrammetry, ground-based synthetic aperture radar and 3D limit equilibrium and finite element modelling. This back-analysis process is considered leading practice with the latest technology. The (northern) side of the Cumba pit slumped along a non-daylighting plane that was identified after the failure event. Remedial investigations included review of geological data, major structures, rock mass constitutive models and groundwater conditions. Topographical and structural data acquired from aerial photogrammetry, pre- and post-failure event, was input into 3D models to replicate observed ground movement. 3D models of pit progression were compared with displacements recorded by ground-based synthetic aperture radar to calibrate model inputs and increase reliability of forward predictions. Such a technical review was completed in less than one week, and the review process implemented for the Cumba pit slope failure now forms the baseline approach for all future geotechnical analysis at the operating mine.
The disturbance factor (D) is a parameter in the Generalized Hoek-Brown failure criterion for rock slopes in slope stability. It represents the subsurface damage to the rock material properties resulting from blasting and stress relaxation during excavations. Within the region of assumed damage, a number between zero (undisturbed) and unity (very disturbed) is prescribed as the value of the disturbance factor. Most commonly a uniform value of D is assumed within the entire region of damage, but little research has been done to study the impact of the variation in the D parameter on stability. Through use of an example, this paper examines the effect of various distribution functions of D through the damaged region, such namely, as constant, linearly varying, and exponentially varying. The failure surfaces and factors of safety for the slope as determined via limit equilibrium are also compared with finite element analyses. Varying the distribution of the damage function was found to significantly affect the failure surface and factor of safety. It is recommended that practitioners adopt care to select an appropriate distribution for slope stability analysis.
1. BACKGROUND
The Generalized Hoek-Brown method (Hoek and Brown 2018) is widely used for determining rock mass strength in rock mechanics. One application of this method is in the design of open pits in rock masses which requires the evaluation of factor of safety against overall sliding. During excavation, subsurface damage can occur to the surrounding rock masses and cause fracturing. The damage can be caused by two sources: (a) blasting and (b) stress relaxation (Hoek et al. 2002), and can be quantified in the form of a disturbance factor, D, which ranges from a zero (undisturbed) to unity (disturbed) within the rock masses. For slope stability, the shear strength of the rock activated along the slip surface can be obtained by solving Eq. (1) to Eq. (4) (Hoek et al. 2018).
scite is a Brooklyn-based organization that helps researchers better discover and understand research articles through Smart Citations–citations that display the context of the citation and describe whether the article provides supporting or contrasting evidence. scite is used by students and researchers from around the world and is funded in part by the National Science Foundation and the National Institute on Drug Abuse of the National Institutes of Health.