Three-dimensional laser imaging for rock mass characterization

Three-dimensional (3D) laser imaging has recently emerged as a tool for rock mass characterization. An image is composed of millions of 3D points which are collectively termed point cloud data. Each image is a digital representation of the rock face containing information for rock mass characterization. The objective of this research project was to (1) measure joint orientation and (2) surface roughness from 3D data. The third objective was to (3) remove the obstructive wire mesh from 3D data. In an above ground field trial, joint orientation was measured using two methods: a 2.5D method using a triangular irregular network and a 3D pole density contouring method using a fully 3D triangular surface discretization model. In each method, the normal vector of each triangle was used to determine the strike and dip. The fully automated 3D method was very accurate in a validation test against manual measurements. The effect of image resolution, triangle size, and joint face geometry were also assessed. In an underground field trail, surface roughness was measured from point cloud data using principal component analysis (PCA). For geo-referencing, joint orientation was first measured using the normal vector from the PCA best fit plane. A 2D surface profile was generated and the maximum asperity amplitude was measured to estimate the Joint Roughness Coefficient. The methodology was successfully validated against manual measurements and applied to generate a surface roughness map from the entire image. Surface roughness anisotropy was also evaluated. In underground mining, support elements such as wire mesh are imaged in front of the rock face. Two filters were developed to remove the wire mesh. The first was based on the instrument noise, measured in-situ from rockbolt face plates, and an estimate of the maximum asperity amplitude. The second was based on the amplitude distance of each point from the PCA best fit plane. In a third field trial, these filters were applied with success. For joint orientation, the presence of wire mesh had minimal impact. Once removed, surface roughness measurements were successfully made using the same methodology as in the previous field trial. Acknowledgements This thesis would not have been possible without the valuable contributions of several individuals. I am most grateful for my supervisor and mentor, Dr. Claire Samson, Chair of the Department of Earth Sciences at Carleton University, for her academic advice, encouragement, and friendship. All of which were invaluable during this thesis program. At the same time, I express sincere gratitude to my co-supervisor, Dr. Steve McKinnon, of Queen's University, for his guidance, support, and technical expertise. I give thanks to Dr. Denis Thibodeau, of Vale, for sponsoring my thesis program and providing access to Vale mining operations. I would like to thank the staff at the 175 Orebody mine (Sudbury, Ontario) and the T1 nickel mine (Thompson, Manitoba) for their assistance and support during my field trials. This research would not be possible if not for the generous donations from Vale. I would like to thank Neptec Design Group for providing the laser scanner and support equipment. I would also like to acknowledge the staff at Neptec for sharing their expertise in image processing and analysis. I gratefully acknowledge the entire Earth Sciences department at Carleton University for their support. Thanks to the graduate students past and present. Last but not least, I would like to express my deepest thanks to my family and friends for their encouragement and patience throughout my academic studies.