Scholarlink

<p>The complex geological conditions of rock masses, particularly in tunnel and underground construction, often involve numerous unstable isolated blocks, leading to frequent collapse-related disasters during excavation. Accurate identification of hazardous blocks is therefore essential to ensure construction safety. This study proposes a novel methodology for constructing three-dimensional (3D) tunnel geological models and identifying hazardous blocks. A parametric calculation approach is developed to reconstruct structural planes from Tunnel Seismic Prediction (TSP) data. By integrating both deterministic and stochastic structural planes, a regionally refined 3D rock mass model is established. For block identification, voxel representation is adopted, and a fast octree-based voxelization algorithm is employed, leveraging voxel encoding features. Furthermore, an octree-guided block localization method is proposed by combining the scanning line technique with a seed-filling algorithm in two-dimensional (2D) space, which is then extended to 3D for the precise identification of complex closed blocks. This approach enables efficient and accurate localization and quantitative prediction of potential collapse hazards. The method was successfully applied to the Letong Tunnel along the BinLai Expressway in Shandong Province, China, where it demonstrated strong performance in detecting hazardous blocks and mitigating the risk of collapse.</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

Three-Dimensional Modeling of Surrounding Rock in Tunnels and Identification of Potentially Unstable Blocks

  • 期刊论文

  • 发表日期：

    2026年1月10日

查看全文
  • Hongliang Liu,
  • Zhongxu Zhou,
  • Xin Gao,
  • Lanyu Cui,
  • Binggang Yan,
  • Shicheng Wang,
  • Hui Cai,
  • Changyuan Chen

摘要

The complex geological conditions of rock masses, particularly in tunnel and underground construction, often involve numerous unstable isolated blocks, leading to frequent collapse-related disasters during excavation. Accurate identification of hazardous blocks is therefore essential to ensure construction safety. This study proposes a novel methodology for constructing three-dimensional (3D) tunnel geological models and identifying hazardous blocks. A parametric calculation approach is developed to reconstruct structural planes from Tunnel Seismic Prediction (TSP) data. By integrating both deterministic and stochastic structural planes, a regionally refined 3D rock mass model is established. For block identification, voxel representation is adopted, and a fast octree-based voxelization algorithm is employed, leveraging voxel encoding features. Furthermore, an octree-guided block localization method is proposed by combining the scanning line technique with a seed-filling algorithm in two-dimensional (2D) space, which is then extended to 3D for the precise identification of complex closed blocks. This approach enables efficient and accurate localization and quantitative prediction of potential collapse hazards. The method was successfully applied to the Letong Tunnel along the BinLai Expressway in Shandong Province, China, where it demonstrated strong performance in detecting hazardous blocks and mitigating the risk of collapse.