【作者】 吴月秀；

【导师】 刘泉声；

【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】 众所周知,节理是控制岩体水力学行为的重要因素,研究节理的变形行为及渗透行为是对裂隙岩体工程安全评估的重要组成部分。随着地下开采、隧道、边坡及核废料地下处置库等岩体工程的兴建,对裂隙岩体水力耦合研究的需求越来越迫切。然而目前有关节理变形行为及渗透行为的描述和数值模拟,均局限于光滑、平行板模型,节理产状参数相互独立且节理处于常法向应力边界等假设；在使用裂隙岩体水力学模型时,常常忽略了其前提条件。针对目前研究的不足,结合国际合作项目DECOVALEX,以核废料地下处置库安全评估为最终目的,在国内外研究的基础上,采用离散裂隙网络模型方法,通过离散元数值计算(DEM)分析了节理粗糙性、非贯通节理的弱化作用、产状参数相关性以及应力状态等多因素对裂隙岩体水力耦合特性的影响,为分析实际裂隙岩体的水力学特性提供研究方法,其具体内容如下：1)提出了一种新的粗糙节理网络模拟方法—SAW法,将粗糙节理视为不断行走的SAW链,通过控制SAW链的一些基本参数(如键长、凸起高度等)来控制粗糙节理形态。本文详细讲述了SAW法的主要思想、实现程序、基本参数的取值以及其使用条件,最终给出了较为系统的粗糙节理网络模拟方法。且应用SAW,研究了不同取样长度下,描述节理粗糙性的各统计参数与JRC之间的相关关系。2)考虑节理模型的广泛适用性以及其参数取值的方便性,引入适用于任何边界条件下的粗糙节理本构模型—SA模型,将其嵌入UDEC自定义节理本构模型,并通过数值单轴压缩试验和直剪实验验证了该自定义节理本构模型的正确性,为水、力学计算提供基础。3)基于SAW,生成单条变隙宽、粗糙节理,对其进行数值渗透实验,分析了其水流规律,得出：节理等效水力隙宽与力学隙宽均值呈指数关系,与节理粗糙度系数JRC呈负指数关系。拟合三者的关系,将其与BB模型修正立方定律进行对比分析,得到了较好的相似性,从数值模拟的角度验证了BB模型的修正立方定律,并将修正的立方定律与SA节理模型联合使用,最终给出了粗糙节理水力学本构模型,为研究裂隙岩体水力学特性打下基础。4)为考虑非贯通节理对裂隙岩体力学强度的弱化作用,引入一损伤变量D,将其定义为块体内所有非贯通节理的实际长度总和与延长至块体边界的长度总和之比。并给出了各块体内损伤变量D的实现程序,然后将含有非贯通节理的岩块弹性模量进行折减,分析非贯通节理对裂隙岩体水力学特性的影响规律,得出：非贯通节理的弱化作用对裂隙岩体力学REV的影响较小,对岩体等效弹性模量有较大的消减作用；其弱化作用对裂隙岩体水力耦合REV的影响较小,对裂隙岩体渗透特性有消弱作用；尤其对岩体渗透率的最大值影响较大,对渗透率的最小值和渗透主方向的影响较小。在实际应用中,应考虑非贯通节理弱化作用的影响。5)研究不同分布规律的节理迹长与隙宽的相关性,重点讲述了迹长与隙宽均符合幂律分布时的相关性,在此基础上,对比分析了常隙宽情况下、变隙宽且不相关情况下以及变隙宽且相关情况下裂隙岩体的水力耦合特性,得出：考虑迹长与隙宽相关性对裂隙岩体的渗透REV、渗透率张量均有很大影响；只考虑隙宽的变化对裂隙岩体渗透REV的影响较小,但对其渗透率张量的影响较大。因此,在实际工程中,应根据地质调查结果对其分别进行研究。6)基于上述研究的基础,结合国际合作项目DECOVALEXⅢBMT 2,计算了Sellafield处核废料处置库各个区域内裂隙岩体的等效渗透系数张量,并研究了不同地应力状态下,其渗透系数张量的变化规律,得出：A)区域1裂隙岩体的等效渗透系数张量主渗透系数K1和K2分别为3.89×10-6m/s和2.27×10-6m/s,其主渗透方向为-15.32°；B)区域2裂隙岩体的等效渗透系数张量主渗透系数K1和K2分别为5.76×10-6m/s和1.06×10-6m/s,其主渗透方向为-32.03°；C)断层带裂隙岩体的等效渗透系数张量主渗透系数K1和K2分别为5.75×10-6m/s和1.97×10-6m/s,其主渗透方向为-31.33°；D)区域1裂隙岩体等效渗透系数张量随深度的变化规律为：Kxx=8.209×10-6·d-0.536, Kxy=3.081×10-6·d-0905,Kyy=2.205×10-6·d-0.635；E)区域2裂隙岩体等效渗透系数张量随深度的变化规律为：Kxx=4.275×10-6·d-0.255,Kxy=5.299×10-6·d-0.325,Kyy=5.971×10-6·d-0.115；F)断层带裂隙岩体等效渗透系数张量随深度的变化规律为：Kxx=4.04×10-6·d-0.141,Kxy=3.048×10-6·d-0.246,Kyy=5.99×10-6·d-0.114；G)与Blum等人的研究结果对比得,本文的计算结果偏小,这是因为本文考虑了节理的粗糙性及非贯通节理的弱化作用的影响。更多还原

【Abstract】 It is known that fracture is the important factor which controls the Hydro-Mechanical behavior of rock mass. Research on deformation and seepage behavior of fracture is a key component of performance assessment for rock mass engineering. With the construction of rock mass engineering such as underground mining, slopes, excavations, nuclear waste disposal and so on, the requirement of research on Hydro-Mechanical behavior of fractured rock mass is more and more strong. But, the present studies on deformation and seepage behavior of fracture are limited to the smooth and parallel fracture and the joint papameters (such as trace, aperture, dip) are assumed to be independent with each other. And the joint is assumed to be under the constant normal load condition. On the application of Hydro-Mechanical model of fractured rock mass, the related prerequisite is ignored. Aiming at those shortage of present studies, combining the international coorperated project DECOVALEX, with a view of performance assessment for nuclear waste disposal, the discrete fracture network (DFN) model and the discrete element method are adopted to analyze the effect of joint roughness, non-persistent joint, correlations between joint parameters and stress boundary on Hydro-Mechanical characteristics of fractured rock mass. A methodology for Hydro-Mechanical analysis of fractured rock mass in nature is presented. The content is list as following:1) A new method for generation of rough fracture network is proposed which is named SAW method. The rough joint is regard as a self-avoiding work. The morphology of fracture is controlled by changing the basic parameters of SAW, such as bone length, the pile-up height and so on. The main idea, implementation program and the determination of basic parameters are presented in this paper. At last, a systematic method for generation of rough joint network is presented. Considering the effect of sampling length on morphology of rough joint, SAW method is applied to analyze the relationship between JRC and other statistic parameters on description of joint roughness.2) Considering the broad applicability of joint constitutive model and the convenience of determination on the related material parameters, the SA joint model is adopted for any kind of boundary condition and is implemented into UDEC as user defined joint model. The numerical uniaxial compression test and direct shear test are used to verify the user defined joint model. This is a foundation for Hydro-Mechanical calculation of fractured rock mass.3) Based on SAW method, a rough joint with variable aperture is generated and the numerical seepage test is done to analyze the flow rule. It is concluded that the relationship between equivalent hydraulic aperture and mean mechanical aperture is an exponential function and the relationship between equivalent hydraulic aperture and JRC is a negative exponential function. After fitting these three parameters, the result of flow rule is similar with that of Banton-Bandis model (BB model). The modified cubic law of BB model is verified from a numerical simulation. SA joint model is used, combining with this modified cubic law. And then the Hydro-Mechanical joint constitutive model is presented for rough joint which is the base of study on Hydro-Mechanical characteristics of fractured rock mass.4) To consider the weakened effect of non-persistent joint on mechanical strength of fractured rock mass, a damage variable D is defined as the ratio of total trace length of non-persistent joints in a block to the total trace length of fiction joints which are the extention of non-persistent joints to the block boundaries. The implementation program of damage variable for each block is presented and the Young’s modulus of intact rock is deduced with the damage variable. Then, the effect of non-persistent joints on Hydro-Mechanical characteristics of fractured rock mass is analyzed. It is concluded that the non-persistent joints have a little effect on mechanical REV and a large weakened effect on the equivalent elasitic modulus of fractured rock mass. It is shown that the non-persistent joints have a little effect on Hydro-Mechanical REV and a certain weakened effect on the equivalent elasitic modulus of fractured rock mass. Especially, the maximum equivalent permeability reduces very significantly, while the minimum equivalent permeability and the main permeability direction change very little. In practical application, the weakened effect of non-persistent joint should be considered.5) The correlation between trace length and aperture with different distribution is studied. Especially, the correlation between trace length and aperture with power-law distribution is analyzed. And the effect of correlation on Hydro-Mechanical behavior of fractured rock mass is studied. Three cases are analyzed. The first one is the fracture with constant aperture. The second one is the fracture with variable aperture following a power-law distribution which is not correlated with trace length. The last one is the fracture with variable aperture which is correlated with trace length both following a power-law distribution. It is concluded that the correlation has a significant effect on hydraulic REV and permeability of fractured rock mass. The variable aperture has a little effect on hydraulic REV, but a significant effect on permeability. Thus, the correlation between trace and aperture should be analyzed based on the geological investigations and then the research should be done for different cases in practical engineering.6) based on the above research and combining with the international cooperated project DECOVALEXⅢBMT 2, the permeability of fractured rock mass for each zone in nuclear waste disposal in Sellafield is calculated. And the permeability under different stress boundaries is analyzed. It is concluded that the permeability of fractured rock mass for zone 1, zone 2 and the fault zone respectively. The maximum permeability for zone 1, zone 2 and the fault zone are 3.89×10-6 m/s,5.76×10-6 m/s and 5.75×10-6 m/s respectively. The minimum permeability for zone 1, zone 2 and the fault zone are 2.27×10-6 m/s, 1.06×10-6 m/s and 1.97×10-6 m/s respectively. The permeability direction for zone 1, zone 2 and the fault zone are -15.32°,-32.03°and -31.33°. The permeability compenents in zone 1 change with deep and the relationships are Kxx=8.209×10-6·d-0.536,Kxy=3.081×10-6·d-0.905 and Kyy=2.205×10-6·d-0.635 respectively.The permeability compenents in zone 2 change with deep and the relationships are Kxx=4.275×10-6·d-0.255,Kxy=5.299×10-6·d-0.325 and Kyy=5.971×10-6·d-0.115 respectively.The permeability compenents in fault zone change with deep and the relationships are Kxx=4.04×10-6·d-0.141, Kxy=3.048×10-6·d-0.246 and Kyy=5.99×10-6·d-0.114 respectively.Compared with the result of Blum,the calculation result in this paper is smaller.The reason for that difference is the joint roughness and the weakened effect of non-persistent joint which is considered in this paper.更多还原