Roof bolts are the dominant method of ground support in mining and tunneling applications, and the concept of using drilling parameters from the bolter for ground characterization has been studied for a few decades. This refers to the use of drilling data to identify geological features in the ground including joints and voids, as well as rock classification. Rock mass properties, including distribution of joints/voids and strengths of rock layers, are critical factors for proper design of ground support to avoid instability. The goal of this research was to improve the capability and sensitivity of joint detection programs based on the updated pattern recognition algorithms in sensing joints with smaller than 3.175 mm (0.125 in.) aperture while reducing the number of false alarms, and discriminating rock layers with different strengths. A set of concrete blocks with different strengths were used to simulate various rock layers, where the gap between the blocks would represent the joints in laboratory tests. Data obtained from drilling through these blocks were analyzed to improve the reliability and precision of joint detection systems. While drilling parameters can be used to detect the gaps, due to low accuracy of the results, new composite indices have been introduced and used in the analysis to improve the detection rates. This paper briefly discusses ongoing research on joint detection by using drilling parameters collected from a roof bolter in a controlled environment. The performances of the new algorithms for joint detection are also examined by comparing their ability to identify existing joints and reducing false alarms.
All Science Journal Classification (ASJC) codes
- Civil and Structural Engineering
- Geotechnical Engineering and Engineering Geology