Composite filter-based detection of rock joints from drilling parameters

In geotechnical engineering, rock joints are the key factor in determining the stability of rock. Quickly and accurately identifying the joints in the rock has always been a concern in geotechnical research. A drilling experiment was designed for jointed rock with joint spacings of 20 cm and 30 cm, and a high-precision drilling parameter acquisition device was used to collect the drilling parameters. It investigated the response of passive drilling parameters—specifically rotation pressure and penetration velocity—to the presence of rock joints. The Kalman filter and moving average filter were used to process the raw data. Moreover, an algorithm for detecting the position of a rock joint was proposed. The results show that when the drill bit of the hammer rotary drilling rig passes through the position of the rock joint, the passive drilling parameters change abruptly. This type of abruptly change occurs when the drill bit passes through the rock joint, due to the difference in strength between the joint and the intact rock. Thus generating phenomena such as “velocity peaks” and “pressure valleys”. The average relative errors of the joint detection algorithm based on the penetration velocity and rotation pressure were calculated to be 5.56% and 4.65%, respectively. Finally, the relative error of joint position detection based on the rotation pressure criterion was 0.45%, and the relative error of joint position detection based on the penetration velocity criterion was 0.48%.