Discussion on necessity and implementation strategy of sub-classification of engineering rock mass

Addressing the common deficiencies in existing engineering rock mass classification methods—such as wide intervals between grades and vague classification indicators—this paper takes the BQ method as a typical representative to systematically demonstrate the necessity of conducting rock mass quality sub-classification. Through parameter sensitivity analysis of Grade IV rock mass, it is found that the selection of upper and lower limit parameters within the same grade can significantly affect the crown settlement of tunnel engineering, the extent of the plastic zone in surrounding rock, as well as the safety factor and deformation magnitude of slope engineering, leading to notable differences in engineering response. An engineering economy case study further indicates that for an open-pit mine slope designed based on the upper and lower limit parameters of the same grade, the maximum safe slope angle can differ by up to 31.56°, resulting in a difference in stripping costs that can reach the order of nearly one million yuan per linear meter. On this basis, this paper systematically explores, from a methodological perspective, the implementation strategies and technical concepts for refined rock mass quality sub-classification: in the data acquisition stage, a conceptual design of an in-situ measurement system is proposed, integrating multi-source measurement-while-drilling data, borehole image information, and computer vision recognition functions; in the indicator screening stage, strategies for selecting sub-classification indicators and analyzing their influence mechanisms are discussed, combining correlation analysis, principal component analysis, and expert experience; in the grade evaluation stage, a multi-dimensional sub-classification technical pathway is conceived, integrating traditional methods such as fuzzy mathematics with intelligent algorithms like machine learning and deep learning, thereby forming a complete methodological framework of "data acquisition – mechanism analysis – intelligent evaluation." The work presented in this paper aims to provide theoretical reference and methodological insights for establishing a scientific and unified standard for engineering rock mass quality sub-classification.