A Rock Strength Prediction Model Utilizing Real-Time Data from Percussion-Rotary Drilling Measurements

The interaction between drilling machinery and rock during the drilling process generates drilling parameters that encapsulate substantial data closely correlated with rock mass strength. The real-time acquisition and analysis of these data for predicting the quality of the rock being drilled is of considerable practical value. However, challenges exist in the real-time, comprehensive collection of drilling data, as well as in data mining and quantitative interpretation. To address these challenges, we undertook equipment development, laboratory tests, and theoretical research. Initially, we established a large-scale experimental platform that integrates percussion-rotary horizontal drilling with real-time monitoring capabilities. This platform facilitates the real-time acquisition of nine drilling parameters, including drilling stroke, rate, pressure, and frequency, and exhibits high controllability, stability, and monitoring precision. Subsequently, we conducted a series of drilling and real-time measurement tests on both homogeneous and layered samples. We identified sensitive drilling parameters, such as drilling stroke, drilling rate, and propulsive force, that indicate rock hardness and layered interfaces. The propulsive force exhibited distinct periodicity, small-scale fluctuations, and amplitude variations, further highlighting its sensitivity to rock strength and stratigraphic boundaries. These characteristics, along with the mechanisms underlying parameter variations during drilling, were elucidated, providing experimental evidence for the real-time identification of engineering rock masses. However, single sensitive parameters were found insufficient for quantitatively predicting rock strength, emphasizing the need for multi-parameter fusion to achieve reliable strength predictions. Finally, we developed a drilling energy model that integrates multiple drilling parameters, based on the energy relationship in drilling processes. Building on this model, we established a rock strength prediction model, thus providing a theoretical foundation for the real-time prediction of engineering rock mass strength.