著者

早野 明 板倉 賢一

出版者

一般社団法人 資源・素材学会

雑誌

Journal of MMIJ

巻号頁・発行日

vol.133, no.4, pp.76-86, 2017

---

<p>Fracture mapping conducted during gallery excavation on construction projects of a large-scale underground facility is based on traditional geological techniques such as visible observations and simple measurements of fracture orientation using a clinometer. Therefore, some difficulties persist, such as securing data quality and worker safety. Three-dimensional (3D) laser scanning, which can instantly acquire point clouds representing the 3D shape of an object surface, is can be effective for resolving these difficulties. In order to confirm the applicability of the three-dimensional laser scanning to the fracture mapping on a gallery wall, this study proposes the method for obtaining attribute information from a point cloud, such as trace length and orientation, associated with the spatial distribution of fractures. In the proposed method, the two-dimensional image is generated from the point cloud to obtain the trace map. After the proposed method was applied to an approximately 50-m-long horizontal gallery excavated into the granitic rock, the reproducibility of the attribute information of fractures obtained solely from the point cloud was examined. Results show that the number of fractures extracted from the point cloud is approximately 80% of those extracted by traditional geological techniques. Although the trace length reproducibility ratio of fractures is approximately 70%, fracture orientations calculated from the point cloud compare favorably with those measured by an on-site researcher. Most fractures that were not extracted from the point cloud do not act as water-conducting fractures because they had short lengths and because they were bonded tightly. Even if the fracture data obtained from the point cloud were applied to modelling for hydrogeological analysis, the results would probably not be influenced strongly. Results of this study indicate the possibility of applying 3D laser scanning to fracture mapping.</p>