Correlation of rock quality designation and resistivity using unmanned aerial vehicle and two-dimensional electrical resistivity tomography

• Main Author: Muhammad Junaid, Muhammad Junaid
• Format: Thesis
• Language: English
• Published: 2022
• Subjects: TA Engineering (General). Civil engineering (General)
• Online Access: http://eprints.utm.my/id/eprint/101473/1/MuhammadJunaidPSKA2022.pdf; http://eprints.utm.my/id/eprint/101473/; http://dms.library.utm.my:8080/vital/access/manager/Repository/vital:150611

Rock Quality Designation (RQD) is a widely applied rock mass classification system for quantifying rock mass quality because it is simple and easily obtained compared to other rock mass classification systems. The rock mass quality using RQD can be identified from drill cores and scanline surveys. However, the calculation of RQD from core samples is expensive and directional-dependent. On the other hand, the scanline survey of obtaining RQD, provides point base information, is time-consuming, and is not practicable in large areas. In addition, the information by scanline survey is limited to the rock outcrop only, and subsurface rock mass quality remains unidentified. For subsurface investigation of rock mass conditions, 2D Electrical Resistivity Tomography (2D ERT) has been extensively applied; however, no comprehensive and detailed correlation of RQD and resistivity values exists to date. This study utilised an integrated Unmanned Aerial Vehicle survey (UAV) and 2D ERT survey at two sites with similar geological formations and aims to establish the correlation between resistivity and RQD indexes. The UAV survey enables the reconstruction of 3D point cloud that calculates the RQD on the surface indirectly from 1 m × 1 m block utilizing Volumetric Joint Count (Jv). This was achieved in ShapeMetrix (SMX) software. At the same time, the 2D ERT survey allows extracting the corresponding resistivity values for each RQD indexes from the same block using ZonRes2D software. A series of Linear Regression (LR) analysis and k-Nearest Neighbour (k-NN) algorithm were performed in Python to obtain continuous projections of RQD and rock resistivity and assigned resistivity values to respective RQD indexes. Two hundred twenty-three data points were obtained representing RQD and corresponding resistivity values. These data points successfully provide a continuous projection of RQD with resistivity using LR analyses, and it was confirmed that the resistivity of rock mass increases 30 Om for each unit increase in RQD index. Whereas the k-NN efficiently assigned resistivity values to various RQD indexes, the very poor rock shows a resistivity value of less than 350 Om; for poor rock, it ranges from 350-1150 Om. While for fair rock, the resistivity varies between 1150 to 1850 Om, for good rock, the resistivity ranges from 1850 to 2500 Om, and excellent rock has a resistivity value greater than 2400 Om. The established correlation of RQD obtained via k-NN characterize the surface and subsurface rock mass quality along the slope in RQD mapping. It was found that the subsurface rock mass quality was at higher quality compared to the surface at both sites. It can be concluded that the integrated UAV and 2D ERT have been successfully applied in this study. In addition, the established correlation will help in obtaining the RQD values using expeditious, inexpensive, and environmental non-destructive approach.