The similarity of fracture patterns inside rockmass among adjacent regions is characterized by high correlation coefficient of them on stereographic projection. This is confirmed from experimental results at four different survey sites in South Korea.
The calculation of fracture correlation coefficient are carried out on two stereonet windows of two successive segments by moving continuously stereonet window couples along the depth of each borehole. The correlation values obtained from each borehole can be used to delineate structural domain boundary and then these results are also compared with the nearby ones to determine the spatial distribution of fracture patterns inside rockmass. The experimental results from four pairs of boreholes with 1880 fracture measurements and total depth of 560m have confirmed the significance of this method in identifying structural boundary.
A structural domain shows a volume of rock mass, characterized by a distinct pattern distribution of fractures by the intensity, orientation, spacing, size and shape. However, only the number of fractures (intensity) and their orientation distributions are considered for determining structural domain boundary in this study. The identification of structural domain is very important for rock engineering investigation because of its close relationship to hydrologic properties and potential failure of rock blocks. The studies of structural domain have received a lot of attention from the past decades by the authors: Miller, 1983; Kulatilake et al., 1990; Michael et al., 2004. Miller (1983) used Chi Square method to compare Schmidt plots in pairs and evaluate the homogeneity of structural populations from a contingency table analysis based on the frequencies of fracture poles that occur in corresponding cells of the Schmidt plots. However, this method has limitation in terms of the number of fracture poles in each window and the expected frequency of fracture poles. Kulatilake et al. (1990) used the methods of Miller (1983) and Mahtab and Yegulalp (1984) combined with a visual comparison of stereonets to identify homogeneous regions. Michael et al. (2004) calculated correlation according to fracture frequency between two stereonet windows to quantify the degree of their similarities for determining structural boundary. The structural boundary is established wherever the fracture correlation coefficient between two stereonet windows is low. Recently, Nguyen et al. (2012) has also used correlation coefficient method to analyze fracture frequency along a tunnel for determining structural domain boundaries.