Identifying Rock Blocks Formed By Complex Fracture Networks

Fractures,also known as discontinuities,that are contained in rock masses may intersect with excavation surfaces and cut the rock masses into various discrete rock blocks.The failure of these blocks may destabilize the rock masses and thus may affect the safety of the entire rock-engineering project.Therefore,it is important to identify rock blocks formed by complex fracture networks during the designing and constructing process of each project.The identified block systems can also be used as the inputs of other advanced discrete modeling methods.In this study,a rock block identification method is proposed based on the element-block assembling approach.This method can handle complex fractures networks with arbitrary arrangements and the blocks formed by both the planar and curved fractures can be modeled.There is no limit to the shape of the resulting blocks and the numerical issues brought by complex fracture networks are analyzed.A systematic scheme to handle the computational error during the block identification process is established.In this study,a hierarchical block system representation model is constructed using the boundary representation approach.The position and shape of each block can be described accurately and effectively using this model.Based on the characteristic of the geological structures encountered in most rock engineering projects,the triangle meshes and polygonal models are used to represent the curved and planar fractures,respectively,which leave enough flexibility for the choice of random fracture models.The proposed representation scheme of block system possesses high flexibility and require less data.This model can also be used in other discrete numerical modeling methods.A block identification process that can handle curved fractures are proposed.A preprocess method is used to remesh the curved triangle meshes under the constraints of the intersection segments.The subdomains that compose the modeling domain are restricted to tetrahedrons in this method,and a constraint Delaunay tetrahedralization algorithm is used to generate each subdomain.The combination of adjacent subdomains that are separated by curved fractures are skipped during the block identification process,which makes it possible to retain the shapes of curved fractures on the identified rock blocks.This method provide a more realistic description of the complex rock block systems.The computational errors arise during the block identification process are studied.An exact computation scheme is proposed and the traditional inexact floating-point arithmetic is replaced by the rational number arithmetic,which is implemented based on a multi-precision integer package to avoid possible overflow during the computation.The filter and lazy evaluation strategy are implemented to avoid unnecessary rational number computations.These strategies bring a considerable improvement to the time efficiency of the exact computation scheme.A regulation process is proposed to deal with finite precision inputs and outputs.The proposed exact arithmetic method can be used to handle arbitrary complex fracture networks,and can provide accurate and realistic inputs for other advanced discrete modeling methods.The established block identification method is applied to a rock block stability analysis process.The proposed method considers the possible cracking of rock bridges,which can be determined based on the block identification process introduced in this paper.The shear resistance on each rock bridges is integrated into the total resistance force of each block when calculating the safety factor.The proposed method provides more accurate results that can be utilized by the designing process of rock engineering projects.