A deterministic-stochastic identification and modelling method of discrete fracture networks using laser scanning: Development and case study.

• A Deterministic-Stochastic Identification and Modelling (DSIM) method of discrete fracture networks is proposed. • For joints, orientations are extracted using region growing algorithm. • For bedding planes, orientations and sizes are extracted from 3D point cloud and real image data in a semi-automatic manner. • The proposed DSIM method is validated with real site data and numerical verification. The intelligent identification of and information extraction from discontinuities in fractured rock masses is crucial for the construction of a 3D fracture network model. A Deterministic-Stochastic Identification and Modelling (DSIM) method for use with discrete fracture networks is proposed using laser scanning technology. The identification of and information extraction from discontinuities is carried out using point cloud data processing by analyzing the outcrop characteristics of the joints and bedding planes visible on the surfaces of the rock mass. For joints, their orientations are extracted using a region growing algorithm, while their sizes are fitted using the Gauss function. For bedding planes, their orientations and sizes are extracted from the 3D point cloud and real image data in a semi-automatic manner. Afterwards, a constrained circle algorithm is applied to restrict the spatial size and shape of the polygonal fractures. In this way, the DSIM method is developed, which conforms to the outcrop characteristics of the bedding planes and joints. From the perspectives of the geometry and geostatistics of discrete fractures, complex discontinuities in 3D space are grouped and spatial density distribution functions are formulated. A corresponding program is compiled to identify and model discrete fracture networks. Finally, a case study was carried out in the Hejing Limestone Mine, Guangxi, China. The proposed DSIM method was validated with real site data and numerical verifications. In addition, the case study also demonstrates that the DISM method can effectively reduce the uncertainty in the identification and modelling of discrete fracture networks (DFN). [ABSTRACT FROM AUTHOR]