复杂裂隙网络下岩石块体识别方法研究

发布时间：2018-06-30 21:15

本文选题：岩石块体识别 + 曲面裂隙　；参考：《中国地质大学(北京)》2017年博士论文

【摘要】：岩体稳定性分析是水利工程建设过程中非常重要的问题。在大型水电工程中,岩体结构是控制各关键区域稳定性的主要因素之一。块体理论是处理该问题的一个有效方法,岩体中包含的裂隙面与施工开挖面会将岩石切割成形状各异的块体。这些岩石块体的的形态会影响岩体中流体的行为,并且各块体的失稳会破坏岩体的整体稳定性,从而影响岩体工程的安全。因此,准确识别由复杂裂隙网络所形成的岩石块体是岩体工程设计与施工中需要解决的关键问题之一,也是各离散数值模型的重要前处理步骤。本研究在单元块体组合法的基础上,构建了一个考虑复杂裂隙网络的岩石块体识别方法。其可同时处理平面和曲面裂隙,并可构建任意形状的岩石块体。该方法还分析了复杂裂隙网络可能带来的数值计算问题,并建立了系统的计算误差控制方法。本研究基于边界表示法构建了一套逐层递进的块体系统表示方法。该方法可准确有效地描述各离散块体的位置和形状。针对岩体工程地质构造特点,本研究中使用三角网表示曲面裂隙,并可灵活选择随机裂隙模型。该表示方法有灵活性高、数据量小等特点,其同时适用于其它离散岩体数值模型。提出了考虑曲面裂隙的块体识别方法。对曲面三角网进行了预处理,在各三角网相交处进行了三角形的重新划分。使用限定四面体剖分算法将研究区域剖分为有限个四面体子区,并在合并相邻子区的过程中跳过被曲面裂隙分隔的子区,从而将曲面裂隙的形状保留在所识别的块体上。本方法可更加精确地描述工程区域内复杂的岩石块体系统。基于有理数运算模型,研究了块体识别过程中的计算误差控制方法。实现了精确运算方法,使用有理数运算代替传统的浮点数运算,避免了截断误差的产生,并应用大数运算解决了数据溢出问题。构建了过滤器和惰性计算机制,避免了不必要的有理数运算,提升了精确运算方法的效率。结合有限精度模型,提出了输入输出数据的规范化方法。该精确运算方法可处理任意形状的裂隙模型,并可保证块体识别结果的正确性。其计算结果可为进一步的非连续岩体分析提供可靠的输入数据。该方法还可被用来处理复杂裂隙网络模拟过程中所遇到的计算误差问题,并且可以辅助建立离散岩体的力学与水力学模型。应用本研究所实现的块体识别程序,构建了新的岩石块体稳定性分析方法。该方法量化了岩桥对块体稳定性的影响。利用块体识别程序寻找可能发生破坏的岩桥,并在计算块体稳定系数时将岩桥上的抗剪力纳入到块体的抗滑力中。为岩体工程的施工与支护设计提供了更精确的数据支持。
[Abstract]:Rock mass stability analysis is a very important problem in the process of water conservancy construction. In large hydropower projects, rock mass structure is one of the main factors to control the stability of key regions. Block theory is an effective method to deal with this problem. The fissure surface and construction excavation surface in rock mass will cut the rock into blocks with different shapes. The shape of these rock blocks will affect the behavior of fluid in rock mass, and the instability of each block will destroy the overall stability of rock mass, thus affecting the safety of rock mass engineering. Therefore, the accurate identification of rock blocks formed by complex fracture networks is one of the key problems to be solved in the design and construction of rock mass engineering, as well as an important pre-processing step for each discrete numerical model. Based on the element block combination method, a rock block identification method considering complex fracture networks is proposed. It can deal with both plane and surface cracks, and can construct rock blocks of any shape. This method also analyzes the possible numerical problems caused by complex fracture networks, and establishes a systematic calculation error control method. In this paper, a hierarchical progressive block system representation method based on boundary representation is proposed. This method can accurately and effectively describe the position and shape of each discrete block. According to the characteristics of engineering geological structure of rock mass, the triangular network is used to represent the surface fracture in this study, and the random fracture model can be flexibly selected. The method has the advantages of high flexibility and small amount of data, and it is suitable for other discrete rock mass numerical models at the same time. In this paper, a method of block identification considering curved surface cracks is proposed. The surface triangulation is pretreated and the triangles are redivided at the intersection of each triangulation. The study area is divided into finite tetrahedron subregions by the finite tetrahedron algorithm, and the subregions separated by the surface fissure are skipped in the process of merging the adjacent subregions, so that the shape of the surface fissure is retained on the identified block. This method can more accurately describe the complex rock block system in the engineering area. Based on rational number operation model, the calculation error control method in block recognition is studied. The accurate operation method is realized and the rational number operation is used to replace the traditional floating point operation. The truncation error is avoided and the problem of data overflow is solved by using the large number operation. Filter and inert computer system are constructed to avoid unnecessary rational number operation and improve the efficiency of precise operation method. Based on the finite precision model, a normalization method of input and output data is proposed. The accurate calculation method can deal with fracture models with arbitrary shape and ensure the correctness of block identification results. The calculated results can provide reliable input data for further discontinuous rock mass analysis. The method can also be used to deal with the computational errors encountered in the simulation of complex fracture networks and to assist in the establishment of mechanics and hydraulics models of discrete rock masses. A new method of rock block stability analysis is constructed by using the block recognition program. This method quantifies the influence of rock bridge on block stability. The block identification program is used to find the rock bridge which may be destroyed, and the shear force on the rock bridge is incorporated into the block anti-slide force when calculating the block stability coefficient. It provides more accurate data support for the construction and support design of rock mass engineering.
【学位授予单位】：中国地质大学(北京)
【学位级别】：博士
【学位授予年份】：2017
【分类号】：TV223

【相似文献】