基于控制圆法和迭代反演的岩体多边形裂隙网络模拟方法

Dimensional discrete fracture network(DFN) is a common approach to analyze the structure of rock masses. However, Baecher disk method which is most popular, is entirely based on statistics and the assumption that fractures are disks rather than polygons for the convenience of mathematical derivation. In this study, a polygonal DFN modeling method more really reflecting fracture feature was developed to improve the traditional Baecher disk method. A circle-controlled method was proposed to determine the shape of random polygonal fractures and its validity was proven through mathematical derivation. For determining the size distribution of fractures, an iterative inversion algorithm was presented. To ensure that the algorithm is able to rapidly converge, a set of adjusting and initialization rules of the parameters of the algorithm were designed. A case study, in which comparison between the trace maps of two DFNs generated by the proposed method and Baecher method is carried out, shows that the iterative inversion algorithm can almost achieve the same results as Baecher method when assuming that the fractures are disks. Further, a polygonal discrete fracture network is generated using the presented method. This research provides a new way to analyze the structure of rock masses.