【作者】 盛金昌； 朱万成； 刘继山；

【Author】 Sheng Jinchang Zhu Wancheng J Liu ( College of Water Conservancy and Hydropower Engineering, Hohai University, Nanjing 210098 China; Department of Mining Engineering, Northeast University, Shenyang 110004 China; School of Oil & Gas Engineering, The University of West Australia, Perth Australia)

【机构】 河海大学水电学院； 东北大学采矿工程系； 西澳大利亚大学石油与天然气工程系；

【摘要】 人们多引用统计学随机抽样方法来描述岩石的非均质性。本文引入图像数字化技术来客观地反映岩石的结构分布,并将其转化成物理参数(渗透系数、弹性模量等)的分布图,然后将它和数值方法相结合, 应用于裂隙化岩石流固热多场耦合分析中。首先需要获得高清晰度的岩体数字图像,岩体中不同的结构对应不同的特征值I值,这样通过I值可以建立起岩体中不同结构材料物理系数的真实分布。将物理参数分布输入到我们开发的基于FEMLAB的岩体多场耦合过程分析程序中,就可以得到真实地反映裂隙岩石的多场耦合作用过程。数值分析结果表明该方法是可行且有效的,比统计抽样方法有明显的优点,其结果更能反映真实的物理现象。更多还原

【Abstract】 The heterogeneity of fractured rock masses has been simulated via statistical tools or random distribution packages in most numerical simulations. In this study, a digital image based (DIB) numerical method is proposed for the two-dimensional coupled multiphysics analysis of fractured rock masses by actually taking into account the spatial distribution of rock structures, i.e., main fractures, minor fractures and intact rock. The digital image processing technique is developed to identify and classify these structures in the fractured rock masses and to translate spatial distribution of rock structures to spatial distributions of material parameters (permeability, Young’s modulus, heat conductivity, etc.). Then the characterized material parameters of rock masses are directly imported into our newly developed three-dimensional (3D) simulator, a MATLAB- and FEMLAB-based Simulator of Coupled Multiphysics in Rock, to evaluate the effect of heterogeneity on the coupled multiphysics process in fractured rock masses. The digital image based numerical results of coupled multiphysics process show that digital image-based technique is capable in characterizing the heterogeneity and fracture structures of fractured rock masses. The digital image based technique is advantageous over the statistical method in characterizing the rock heterogeneity and provides a new approach to predict more realistic numerical results.更多还原