Scholarlink

<p>Accurate determination of the equivalent mechanical parameters of jointed rock masses is essential for tunneling and underground excavation. However, the random distribution of joints hinders a comprehensive understanding of the mechanical behavior of such complex systems. This study proposes a methodology based on the discrete fracture network–discrete element method (DFN–DEM) to determine the equivalent mechanical parameters of rock masses with different joint densities. First, an enhanced Mask R-CNN algorithm was employed to extract joint geometries from tunnel surrounding rocks, and statistical features were used to construct discrete fracture networks. Synthetic rock mass technology was used to generate jointed rock specimens with varying densities and sizes, enabling analysis of anisotropy, size effects, and representative elementary volumes (REV). Numerical simulations were complemented by tests on 3D-printed rock-like specimens, from which elastic modulus, cohesion, and internal friction angle were obtained through triaxial compression experiments. Results for limestone with low, medium, and high joint densities (<i>P</i><sub><i>32</i></sub> = 3.1, 5.8, and 8.2&#xa0;m⁻¹) indicated REV sizes of 8&#xa0;m, 8&#xa0;m, and 10&#xa0;m, respectively. Equivalent parameters under different loading directions varied by less than 70%, and results deviated by under 16% from Hoek–Brown criterion estimates, confirming method reliability. The integration of intelligent joint identification, DFN–DEM modeling, and 3D printing provides a precise parameter determination method. The method assumes statistical representativeness of extracted joint features, while laboratory validation remains limited in scale. The findings of this study provide a theoretical basis for underground tunnel design and construction.</p>

错误：搜索内容不能为空，请输入英文关键词
错误：关键词超出字数限制，请精简
高级检索

A DFN–DEM approach to determine equivalent mechanical parameters of rock mass with different joint densities

  • 期刊论文

  • 发表日期：

    2025年12月19日

查看全文
  • Liang Wang,
  • Shijie Zhou,
  • Juan Huang,
  • Mingfeng Lei,
  • Yonggang Han,
  • Jian Zhang,
  • Chaojun Jia

摘要

Accurate determination of the equivalent mechanical parameters of jointed rock masses is essential for tunneling and underground excavation. However, the random distribution of joints hinders a comprehensive understanding of the mechanical behavior of such complex systems. This study proposes a methodology based on the discrete fracture network–discrete element method (DFN–DEM) to determine the equivalent mechanical parameters of rock masses with different joint densities. First, an enhanced Mask R-CNN algorithm was employed to extract joint geometries from tunnel surrounding rocks, and statistical features were used to construct discrete fracture networks. Synthetic rock mass technology was used to generate jointed rock specimens with varying densities and sizes, enabling analysis of anisotropy, size effects, and representative elementary volumes (REV). Numerical simulations were complemented by tests on 3D-printed rock-like specimens, from which elastic modulus, cohesion, and internal friction angle were obtained through triaxial compression experiments. Results for limestone with low, medium, and high joint densities (P32 = 3.1, 5.8, and 8.2 m⁻¹) indicated REV sizes of 8 m, 8 m, and 10 m, respectively. Equivalent parameters under different loading directions varied by less than 70%, and results deviated by under 16% from Hoek–Brown criterion estimates, confirming method reliability. The integration of intelligent joint identification, DFN–DEM modeling, and 3D printing provides a precise parameter determination method. The method assumes statistical representativeness of extracted joint features, while laboratory validation remains limited in scale. The findings of this study provide a theoretical basis for underground tunnel design and construction.