Search

Fractures have significant impact on the stability of underground projects such as coal mining. The identification of geological features based on drilling parameters is a promising option for intelligent detection of rock formations. A series of studies conducted to identify fractures under unstable drilling conditions are presented in this paper. First, the forces acting around the drill bit are analyzed to further improve the force model of the two-wing PDC drill bit. Then, based on the self-developed borehole drilling device, the drilling parameters are collected in real time while drilling in concrete of different strengths and different fracture widths. Finally, the response characteristics of drilling parameters are analyzed, and the multi-parameter voting method is proposed for fracture identification. The experimental results show that when fracture width increases, the crushing area, which forms around the borehole when the drill bit encounters the fracture, tends to increase. The rate of penetration (ROP) increases suddenly, the revolution per minute (RPM) decreases, and the torque increases at fracture encounter. The sudden changes are recovered after passing the fracture. The multi-parameter voting method has a high recognition rate for fractures of width greater than 2 mm, but a relatively low recognition rate for fractures of 1 mm width. In addition to detecting fractures, the method performs well in predicting fracture location and fracture width. Highlights: The response characteristics of multiple drilling parameters at the fracture were analyzed. As the thrust (torque) increases, the amount of change in penetration rate caused by the unit change in thrust (torque) decreases. Results show that the proposed multi-parameter voting method has a high recognition rate for fractures with width greater than 2 mm. [ABSTRACT FROM AUTHOR]