裂隙岩体隧洞危石批量识别与垮塌失稳演化规律研究

One of the main geological disaster faced by tunnel engineering constructed in fractured rock mass zone is dangerous rock collapse. Fundamental reason that the disaster is hard to prevent and control is the serious different between complex surrounding rock mass system and mechanical model caused by the uncertainty of rock mass structure itself. To ascertain the spatial evolution law and distribution characteristics of dangerous rock group, the information obtain of fractured rock mass structure, batch simulation of fractured network model and breadth first search of block group are researched respectively. The self-adaptive extraction algorithm of structural information for fractured rock mass is put forward by using multiple image process and feature recognition operators, and the shear strength characteristics considering 3D morphology features of fractures is researched. The self-adaptive extraction and mechanical properties representation of rock mass structure are realized. The batch fractured network models can be simulated based on parallel computation, and the geometric topological criterion and convergence condition of block with the mechanism of breadth-first search is put forward. The shear failure criterion of block which is suitable for undigraph style is established, and the failure probabilistic prediction of dangerous rock group based on batch simulation models is realized. The spatial evolution law and form distribution characteristics of block group under different condition of fracture combination can be revealed. The research achievements will provide a significant method to forecast, identify and prevent dangerous rock collapse for tunnel engineering located in fractured rock mass zone.

危石垮塌是裂隙岩体隧洞工程修建中面临的主要地质灾害之一，其难以防控的根本原因在于岩体结构本身的不确定性致使数据不完备的围岩体系统与理论严密的数学模型之间严重脱节。本项目以岩体结构信息获取-裂隙网络模型仿真构建-块体广度优先搜索为主线，研究危石垮塌的展布特征与空间形态演化规律。基于多元图像处理与特征识别算子，提出裂隙岩体结构自适应特征提取算法，开展考虑裂隙三维形貌特征的剪切强度试验特性研究，实现岩体结构信息的自适应提取与力学属性表征；基于并行运算实现裂隙网络仿真模型的批量构建，研究广度优先搜索机制下不同批次可动块体的几何拓扑准则与收敛条件，提出适用于有权无向图结构的块体剪切滑移失效准则；建立基于批量仿真模型的隧洞危石识别与失效概率预测模型，揭示不同岩体裂隙组合条件下隧洞危石群空间形态演化规律与响应特征。研究成果对裂隙岩体隧洞危石垮塌灾害预测、临灾判识与防控治理具有重要科学意义与工程实用价值。

裂隙岩体地下工程不同于有着比较明确数学模型的结构工程，作为一种非均质、非连续性介质，其最为凸显的特性即是岩体结构的不确定性，尤其是与开挖历程有密切关联的裂隙岩体隧洞工程，所建数学模型分析结果均与工程实际出入较大。数据不完备的复杂围岩体系统与理论严密的数学模型之间严重脱节已然成为裂隙岩体隧洞危石几何识别、稳健分析与垮塌防控的“瓶颈”所在。为此，充分考虑岩体裂隙的几何形态与展布特征，基于多元图像处理与特征识别算子，提出了裂隙岩体结构自适应特征提取算法；基于并行运算实现了裂隙网络仿真模型的批量构建，研究了广度优先搜索机制下不同批次可动块体的几何拓扑准则与收敛条件，提出了适用于有权无向图结构的块体剪切滑移失效准则；建立了基于批量仿真模型的隧洞危石识别与失效概率预测模型，基于数值优化模型，开展了隧道遭遇断层条件下其影响区内块体垮塌空间展布规律研究，对不同开挖循环内垮塌块体的形状、体积、数量等进行了系统分析与数理统计；同时针对不同连通率、DFN密度与节理间距等敏感地质参数，通过108组仿真模拟，揭示了既定岩体裂隙组合条件下隧洞危石群的空间形态演化规律。研究结果为裂隙岩体区域隧洞及地下工程修建过程中危石垮塌灾害预测提供了科学依据。