【作者】 张占荣；

【导师】 盛谦；

【作者基本信息】 中国科学院研究生院（武汉岩土力学研究所） ， 岩土工程， 2010， 博士

【摘要】 本文在岩石力学试验和工程地质勘察成果的基础上,应用数理统计方法,研究了裂隙岩体变形参数的确定方法,同时从开挖损伤效应、结构效应和尺寸效应出发,结合概率统计、智能反演分析和回归分析等方法,系统探讨了裂隙岩体变形特性的影响因素,主要研究工作有：(1)广泛搜集国内大型水利水电工程现场试验数据,结合地质勘察资料及室内试验数据,参照国标《工程岩体分级标准》得到了岩体变形模量-岩体基本质量指标(Em-BQ)的对应关系；结合数理统计原理,通过相关性分析,建立了基于BQ指标的岩体变形模量估算的经验方法,误差评价表明该方法具有一定的应用和参考价值；进一步结合实测数据,引入岩体状态因子M,建立了通过BQ指标、岩块弹性模量估算岩体变形模量经验方法,应用于三峡工程岩体力学参数的估计,所得结果与现场岩体力学试验数据比较吻合。这些研究为通过BQ指标估算裂隙岩体变形模量提供了条合理的量化途径。(2)结合正交试验设计、支持向量机模型、粒子群优化算法等原理,建立了考虑开挖损伤效应的LSSVM-PSO位移反分析模型；以大岗山、溪洛渡水电站地下厂房为例,结合开挖损伤区测试成果,针对现场变形监测成果进行了位移反演分析,得到了损伤区范围与弱化程度；进一步围绕岩体开挖损伤效应,展开了多个工程对比分析,揭示了地质结构对岩体开挖损伤效应的影响；对比分析表明：①岩体完整性越好,则开挖损伤区范围越小,岩体力学参数弱化程度较大；②有断层岩脉切割地下洞室时,岩体损伤区的范围会相应变大；③损伤区范围与弱化程度大致呈反比关系,随着厂房向下开挖,损伤区内岩体变形参数弱化程度不断增加。(3)根据规则节理岩体变形模量的理论计算公式,在考虑岩体结构面参数概率分布特征基础上,结合变形等效原则,推导了随机节理岩体变形模量的理论计算公式；结合半球形展示图,可对节理岩体变形模量进行三维空间全方位展示；在工程地质调查和岩石力学试验基础上,应用此方法,研究了三峡工程裂隙岩体变形参数的结构效应,与已有研究成果对比表明,此方法计算精度较好、计算过程简便,还可以全方位展示裂隙岩体变形参数的各向异性特征。(4)在大量现场试验数据基础上,通过随机分布函数检验确定了岩体变形模量的概率分布类型,建立不同尺度的随机数值分析模型,研究大岗山水电站岩体变形模量的REV及其随尺度的变化特征；进一步通过多尺度方法,包括室内试验、现场变形试验、声波测试、数值分析、反演分析等方法,确定了不同尺度裂隙岩体的变形参数,通过拟合回归分析,得到大岗山水电站岩体宏观力学参数及其REV尺度；最后,对裂隙岩体变形模量尺寸效应的产生机理进行了探讨,表明结构面的切割是造成裂隙岩体变形参数尺寸效应的内在原因,随着岩体试验尺度的增大,试验结果会受到更大尺度结构面的影响,从而导致了裂隙岩体等效变形模量随着岩体尺度增加而出现有规律的降低现象。更多还原

【Abstract】 Deep research on the determination methods and influential factors for deformation parameters of the fractured rock mass has been carried out based on the rock mechanics tests, engineering geological investigation, probability statistics, intelligent inversion model and multiple linear regression method, etc. The major achievements are listed as follows:(1) The Em-BQ relationship is obtained by combining massive on-site test data, geological investigation records and indoor test logs of a number of large hydropower stations in China with Chinese Standard for engineering classification of rock masses. Then a BQ-based empirical method is put forward to estimate the deformation modulus of rock mass based on correlation analysis. By introducing a rock mass status factor M, a new empirical approach is further proposed to estimate the deformation modulus of rock mass by BQ index and rock elastic modulus. The new approach is used to calculate the rock mass deformation parameters of the Three Gorges Project, and the estimated results match well to the tested data. These methods provide a reliable way to estimate the deformation modulus of rock mass by BQ index.(2) The excavation-damage-considered LSSVM-PSO displacement back-analysis model is established by the orthogonal test, support vector machine and particle swarm optimization, etc. The underground cavern displacements of Dagangshan and Xiluodu hydropower stations are back analyzed, according to the on-site measurement on excavation damage zone and displacement, and then the scope and weaken degree of excavation damage zone are obtained. The excavation damage inversions for other projects are conducted, revealing the effect of geological structure on the excavation damage of rock mass. The results show that the more intact the rock mass is, the smaller the excavation damage zone, with larger weaken degree of rock mass mechanical parameters; the excavation damage zone will extend if there are faults and dikes across the underground caverns; the scope of the excavation damage zone is in inverse proportion to its weaken degree; the deformation parameters of rock mass within the excavation damage zone decreases as the caverns are excavated deeply.(3) Based on a theoretical formula for calculating the deformation modulus of rock mass containing a set of structural joint planes, a novel formula is derived to calculate the deformation modulus of fractured rock mass considering the effects of random joint planes. The hemispherical figures are applied to demonstrate fully the deformation modulus of random jointed rock mass in the 3D space. Furthermore, the proposed formula for deformation modulus of random jointed rock mass is used to study the structure effect of the deformation parameters of fractured rock mass of the Three Gorges Project. Comparing with the existing results, it shows that the new approach can not only be easily operated with satisfied accuracy, but also fully demonstrate the anisotropy property of rock mass.(4) The size effect of rock mass deformation parameters is studied by probability and multi-scale methods. A large number of on-site test data are used to determine the probability distribution type of rock mass deformation parameters by the random distribution function test. A multi-scale random analysis model is employed to estimate the REV of rock mass deformation modulus of the Dagangshan hydropower station and the REV-scale relationship. Then the deformation parameters of different scales are determined by multi-scale methods, including indoor and on-site tests, numerical simulation and back-analysis, etc. The macro-mechanical parameters and REV scale of Dagangshan hydropower station are obtained. Finally, the mechanism of size effect of rock mass deformation modulus is discussed and it shows the structural planes within rock mass are the intrinsic cause for this mechanism. The equivalent deformation modulus will decrease regularly with model size.更多还原