流形元法(NMM)在岩体稳定性分析中的应用研究
详细信息    本馆镜像全文|  推荐本文 |  |   获取CNKI官网全文
摘要
土木建筑、水利水电、铁路公路交通、矿山开采、文物保护、地质灾害治理、能源开发及国防建设等众多的工程实践中,常常会遇到大量的岩土工程问题,尤其是工程岩体稳定性、岩体强度及变形的问题。作为主要工程对象的岩体是在漫长的地质历史发展过程中经受各种地质营力作用而形成的成因和构造复杂,岩性多样且包含了连续介质,裂隙介质和散体介质特征的,兼有弹性、塑性、粘性、流变性的各向异性的非均质地质体。岩体在不同的工程实践中往往表现为加载和卸载两种不同的力学状态,并且由于其固有的裂隙介质特征,又不可避免地赋存于一定形态的液态介质一各种地表水、潜水、承压水等中。由于实际工程岩体其固有的非线性和复杂性,使得求解对应的诸如其稳定性、变形等问题面临较大的困难,而有关的工程结构的设计必须保证该结构在外来因素的作用下具有足够的安全度、经济性和合理性,这必然要求对岩体及其工程结构的受力与变形有一套量化评判体系和优化技术,抓住实际工程问题的本质特征并建立可行的符合几何仿真、本构仿真、受力仿真、过程仿真四原则的求解模型,通过该模型的数值模拟成果来指导岩体工程的设计、施工及运营、管理;而岩体结构面的存在使得基于传统连续介质力学理论的理论分析和数值模拟面临巨大的挑战,物理模拟的试验周期和成本也大大增加,而岩土体工程问题则成为典型的数据有限、了解程度有限类问题,这类问题的解决需要综合应用理论分析、经验判断、物理模拟和数值模拟等方法,数值模拟可以完成目前许多技术手段无法完成的实验,如参数控制,复杂条件下的边界条件的处理,同时数值模拟具有高可重复性,且数值模拟的成本和人力开销等远低于物理模拟,因此研究岩土体工程问题的流形元数值模拟方法是一项具有理论和实际工程应用价值的重要课题。
     流形元法是十年在有限覆盖系统基础上发展起来的新的岩体工程数值模拟方法,流形的基本结构是有限覆盖系统和覆盖间的接触函数。与通常的网格不同,覆盖可重叠或折叠产生一个组合空间,因此流形方法是追求一切空间上的所有解法。在流形一般化定义下,流形方法具有两种独立的网格——数学网格和物理网格。数学网格由折叠的覆盖组成,物理网格则由实际的材料边界、裂缝、块体、不同材料的交接面甚至不同的材料相的界面等组成。数学网格所含的数学覆盖定义了近似解的精度,由用户自由选择。常规的有限元分析网格等都可以转换成数学覆盖;而物理网格则定义了积分区域,不能人为选择。流形元法的最大优点可以统一处理连续变形与非连续变形问题, 同时可以统一解决有限元法、非连续变形分析及解析法的计算问题,具有广泛的应用前景。
     本文以流形元法的基本理论为基础,主要研究对应的流形元有限覆盖系统的生成算法及流形元法的接触判断、接触转换、流形元法总体控制方程的建立与稀疏存储、动态连通性分析及分片直接求解,并基于泛型程序设计技术及面向对象的系统分析实现了对应的C++流形元法分析工具NMMG,并用具体的计算模型检验了流形元法分析的结果。
    
     具体的研究内容主要为:
     1.流形元有限覆盖系统生成法
     在计算儿何的实际直线求交法、剪切算法及空间元素拓扑关系确定法等方法和非连续变
    形分析的裂隙切割块体生成算法的基础上,结合有向图遍历理论,本文构造并实现一个适用
    于以三角形常应变单元和四节点等参元为数学网格的流形元有限覆盖系统生成算法,并进行
    了相关算例的检验,结果表明该算法能满足实际问题的容错性及鲁棒性要求;并能较好地扩
    充到以其它有限元单元类型为数学网格的流形元有限覆盖系统的生成,开发了对应的C++流
    形元有限覆盖系统的生成类(elass ME_Builder)。
    2.流形元法中的接触分析
     针对二维流形元法实际的角对角、角对边、边对边三种接触类型,在石根华提出的相关
    接触准则的基础上借鉴离散单元法中对接触的处理方法,利用接触判断的距离准则、角度准
    则及嵌入准则对流形元法中的接触进行处理,同时研究了接触状态在不同时间步之间的转换
    方法,开发了对应的e++接触处理类(elass ContaetS)。
    3.流形元法总体控制方程的建立
     在流形元有限覆盖基础上推导了以三角形常应变单元和四节点等参元为数学网格的对
    应的各种应力边界条件、位移边界条件及各种荷载作用下的流形元单元刚度子矩阵,各种接
    触类型对应的刚度子矩阵、最终控制方程的集成和无嵌入、无张拉约束条件的处理方法以及
    对应的基于单纯形积分理论的数值积分法,开发了对应的C++类(class KMatrioes)。
    4.流形元法总体控制方程求解理论的初步研究
     主要研究了基于有向图理论的流形元系统节点编码的优化、对应稀疏矩阵的压缩存储与
    流形元总体控制方程的动态连通性分析、分片直接求解法及流形元接触的开一合迭代方法,
    并开发了对应的e++类(elass Solver)。
    5.基于范型程序、模式设计技术的流形元分析系统的实现
     结合最新的范型程序设计技术及模式设计技术,用C++实现了上述算法及必要的辅助类,
    并最终集成一个实用的二维流形元分析工具一确蒯G(而merieal Manifold Met
There are many geotechnical engineering problems involved in the practice of civil construction, water conservancy and hydroelectric structures, railway and highway project, mine exploitation, cultural relic protection, geological hazard prevention, energy exploiture and national defence works etc, especially the problems for the stability, strength and deformation of the rock mass. And the rock mass as the main object of most engineering practice is developed by all kinds of geotechnical action and force in the long geology history to the complex characteristic of continuous medium, crannied medium and granular medium, to the anisotropic and heterogeneous entity with elasticity, plasticity, viscidity, creepocity and rheology. At the meantime, the rock mass may alternate between loading and unloading and it exists in certain kind of liquid, such as surface water, unconfined water, confined water etc for its intrinsic crannies. The practical rock mass concerned project must solve the key stability pr
    oblem after the understanding of the complicated mechanical characteristic and the deformation trends to guide the following project design and construction for the demand of security, economy, feasibility and validity. However, the intrinsic nonlinearness and complexity of the engineering rock mass become the main difficulty to predict the stability and deformation, the corresponding structure design must ensure enough safety with all the determinate or random force combination, so a model without the geometry distortion and constitutive equation warp is necessary to be built for the quantificational analysis of practical structure's stress, for the simulation of the real process and for the determinate evaluation system and optimization. It is also indispensable to direct the design, construction, operation, maintenance and management of specific rock mass concerned project. Discontinuities of rock mass are the primary challenge for theoretically analysis and numerical model based on the traditional contin
    uous medium mechanics theory, and sharply increase the cost, extend the period of the physical modeling. And the geotechnical engineering concerned problems are typical data limited and understanding limited issues. All available theoretically analysis methods, simple or complex, qualitative assessment and prediction by experience, physical modeling and numerical modeling can not solve these problems solely. However, numerical modeling can make use of the result of physical modeling and theoretically analysis and its result can be compared with the practical monitoring data to verify the validity of theoretically analysis and it may decrease the cost of physical modeling. The most advantage of numerical modeling is its wide solution scope and its repetition. So the study on the numerical manifold method for geotechnical engineering problem is an important subject of academic and applied value for practical projects.
    Numerical Manifold Method (NMM) is a thorough new numerical modeling method for geotechnical engineering in past decade. Manifold, as the main theme of differential geometry and analysissitus, has a basic structure composed of finite cover system and contact functions between
    
    
    covers, which could be overlapped or even folded. There are two type meshes, mathematic mesh and physical mesh for the general definition of manifold. Mathematic mesh is composed of folded covers and could be random set by final user, and physical mesh is composed of real material boundary, joints or discontinuities, phase interfaces etc and couldn't be artifical. Traditional finite element method's mesh can be used as the mathematic mesh for the NMM. And NMM is a promising numerical method to solve both the continuous and discontinuous deformation problem of rock mass and to unify the finite element method(FEM),discontinuous deformation analysis(DDA) and analytic method.
    This paper will focus on the generation of finite cover system for the numerical manifold method, distance and angle criteria for contact jud
引文
1.曹文贵,速宝玉.流形元有限覆盖系统的自动生成方法之研究.岩土工程学报,2001,23(2)
    2.严登俊.有限元网格生成技术.微特电机,1999(1)
    3.庞作全.在节理网络图上全自动生成有限元网格.岩石力学与工程学报,1999,18(2)
    4.刘强.论两条直线的求交.计算机学报,1997,20(12)
    5.吴兵.具有拓扑关系的任意多边形剪切算法.小型微型计算机系统,2000,21(11)
    6.石根华.数值流形方法与非连续变形分析.北京:清华大学出版社
    7.裘衷燕.C++程序设计语言 第三版.北京:机械工业出版社
    8.王泳嘉.离散单元法及在岩石力学中的应用.沈阳:东北工学院出版社
    9.李学文.三维摩擦问题的非光滑方程组方法研究.大连理工大博士学位论文
    10.王水林等.流形元方法在模拟裂纹扩展中的应用.岩石力学与工程学报,1997,16(5)
    11.王芝银等.岩石大变形分析的流形元方法.岩石力学与工程学报,1997,16(5)
    12.张国新等.基于流形方法的结构破坏分析.岩石力学与工程学报,2001,20(3)
    13.王书法,朱维申,李术才等.考虑侧向影响的数值流形方法及其工程应用.岩石力学与工程学报,2001,20(3)
    14.洪帆.离散数学基础.武汉:华中理工大学出版社
    15.吴文虎.王建德.实用算法的分析与程序设计.北京:电子工业出版社
    16.唐辉明等.岩土工程数值模拟新方法.地质科技情报,1998(2)
    17.唐辉明等.工程地质数值模拟的理论与方法.武汉:中国地质大学出版社
    18.刘锦华.块体理论在工程岩体稳定性分析中的应用.北京:水利电力出版社
    19.刘佑荣等.岩体力学.武汉:中国地质大学出版社
    20.潘别桐,黄润秋.工程地质数值法.北京:地质出版社
    21.黄润秋,张倬元.高边坡稳定性的系统工程地质研究.成都:成都科技大学出版社
    22.冯夏庭.智能岩石力学导论.北京:科学出版社
    23.王泳嘉,冯夏庭.关于计算岩石力学发展的几点思考.岩土工程学报,1996,8(4)
    24.李景勇.有限元法.北京:北京邮电大学出版社
    25.陈崇希.地下水流动问题数值方法.武汉:中国地质大学出版社
    26.卓家寿.不连学介质力学问题的界面元法,北京:科学出版社
    27.徐芝纶.弹性力学.北京:高等教育出版社
    28.卓家寿.弹塑性力学中的广义变分原理.北京:水利水电出版社
    29.周维垣.高等岩石力学.北京:水利电力出版社
    30.孙广忠.岩体结构力学.北京:科学出版社
    31.伍法权.统计岩体力学原理.武汉:中国地质大学出版社
    32.魏群.散体单元法的基本原理.北京:科学出版社
    33.王卫东.弹性力学问题的无网格方法.山东工业大学学报,2002,32(1)
    
    
    34.刘欣等.基于流形覆盖思想的无网格方法的研究.计算力学学报,2001,18(1)
    35.田春松等.平面区域渐变三角形网格的自动生成.数值计算与计算机应用,1993(4)
    36.王建华.平面任意区域自适应网格逼近.数值计算与计算机应用,1996(3)
    37.O.C辛克维奇.土壤与多孔介质静力和动力特性的基本公式.应用数学和力学,1998,3(4)
    38.王彦武.地下采矿工程岩体质量的可拓模糊评价方法.岩石力学与工程学报,2002,21(1)
    39.郑宏.界面问题的混合有限元法.岩石力学与工程学报.2002.21(1)
    40.黎勇.非连续变形计算模型基本原理及其线性规划解.大连理工大学学报,2000,40(3)
    41.栾茂田,黎勇,杨庆.非连续变形计算模型在岩体边坡稳定性分析中的应用.岩石力学与工程学报,2000,19(3)
    42.胡小荣.岩土力学参数随机场的空间变异性分析及单元体力学参数赋值研究.岩石力学与工程学报,2000,19(1)
    43.韦忠瑄.计算力学中现代数学方法的若干问题.解放军理工大学学报,2001,2(4)
    44.杨海天.有限体积法在固体力学中的应用.大连理工大学学报.1995,35(3)
    45. Tang Huiming. A Study on Slope Stability by Method of Damage Mechanics,8th Int Conf Enginnering Geology, A.A.Balkma, 1998
    46. Yan Tongzhen,Tan Huiming. Global Environmental Changes And Engineering Geology,1997 Wuhan: China University of Geosciences
    47. Shi G.H.. Manifold Method. Proceedings of 1st International Forum on Discontinuous Deformation Analysis and Simulations of Discontinuous Media, 1996, Berkeley. California
    48. Lin J.S. Continuous and Discontinuous Analysis Using the Manifold Method. Working Forum on the Manifold Method of Material Analysis, 1995,California
    49. Shi G.H.. Modeling Rock Joints and Blocks by Manifold Method. Proceedings of 32nd U.S. Symposium on Rock Mechanics, 1992,Sana Fe New Mexico
    50. Goodman R.E. Block Theory and its Application.Geotechnique,1985
    51. Cundall P.A A Computer Model for Simulating Progress Large-scale Movement in Blocky Rock System. Proceedings for the Symposium of International Society of Rock Mechanics. 1971.Nancy, France
    52. Feng kang. Construction of Canonical Difference Schemes for Hamilton Formalism via Generating Function.J.Comp.Math, 1989,7
    53. E. Hoek. Rock Slope Engineering, The Institution of Mining and Metallurgy, London
    54. Bjarne Stroustrup. The C++ Programming Language. Addison-Wesley Inc.
    55. Stanley B.Lippman. C++ Primer. Addison-Wesley Inc.
    56. Nicolai M.Josuttis. The C++ Standard Library. Addison-Wesley Inc.