【作者】 倪春中；

【导师】 张世涛；

【作者基本信息】 昆明理工大学 ， 矿产普查与勘探， 2013， 博士

【摘要】 本文结合云南省三年找矿行动项目滇东南薄竹山地区钨、铅、锌、银多金属矿整装勘查及国家自然科学基金项目“地学中裂隙的多尺度空间分布模拟”(项目编号：40902058)的科研选题。裂隙是地壳上部岩石中广泛发育的一种基本构造形式。在矿产资源形成过程中,裂隙的存在为矿液渗透、分散及聚集等提供了有利条件。如何利用现存的构造地质和岩石样品试验等资料对裂隙的空间分布、延伸情况作出科学评价与预测,是地质找矿研究面临的难题。本文在总结前人工作的基础上,以个旧锡矿高松矿田为例,通过系统的野外调查观察,针对如何正确预测裂隙空间分布的难题,借助不同尺度下裂隙建模,建立其相互间联系规律。本文主要取得以下研究成果与创新性认识：(1)个旧锡矿高松矿田裂隙特征的标定研究。裂隙属性包括方向、长度、宽度、位置、充填等。通过分析,裂隙位置服从符合泊松分布,裂隙的长度、位移、大小、密度、宽度属性服从幂指数分布、对数正态分布、指数分布、Gamma分布,方向服从Fisher分布及Binghanm分布。在厘定本文大、中、小三个不同尺度的基础上,揭示裂隙属性的尺度不变性规律,并探讨了用盒子计数法和两点相关函数计算的裂隙多尺度联系的方法。(2)中尺度裂隙三维网络分布特征模拟理论方法的构建。该方法由普通克里格(OK:Ordinary Kriging)、序贯高斯模拟(SGS:Sequential Gaussian Simulation)、主成分分析(PCA:Principal Component Analysis)方法构成。模拟内容包括收集裂隙资料、分解裂隙属性、进行裂隙密度的估计、裂隙空间位置的生成、裂隙方向(走向和倾向)的主成份分析、裂隙方向分布的地质统计学估值、裂隙属性的综合以及基本裂隙面的连接。(3)个旧锡矿区高松矿田裂隙多尺度分布模拟研究。利用遥感、地表、坑道、CT等手段,搜集了不同尺度、不同维数的地质资料和裂隙样本数据。裂隙大尺度模拟主要利用线迹追踪法(Segment Tracing Alogrithm)从云南个旧矿区遥感影像中抽取二维裂隙(线性构造等)的分布,结合地形数据、数字高程模型(DEM)等推定三维裂隙(断裂面等)的分布。裂隙中尺度模拟以矢量数字化技术为主,辅以人工判读,从个旧矿区的坑道、钻孔等编录中提取一维裂隙(断裂、裂隙等)数据,并从岩墙、风化面、工程剖面等提取二维裂隙(裂隙痕迹)数据,在此基础上,模拟了裂隙空间位置分布、对应方向的空间分布,连接了裂隙网络的空间分布。裂隙小尺度模拟利用线迹追踪方法,根据X射线扫描仪(CT)显微图像读取二维裂隙(裂纹等)的分布,显微图像中低维裂隙数据是三维实体与一维和二维截面相交后留下的痕迹,蕴含着三维裂隙的信息,通过体视学技术,从观察到的低维资料反演推断出三维裂隙的分布,实现了小尺度裂隙长度、密度的跨维数转换及属性(方向、长度、宽度)之间关系的建立。(4)根据多尺度联系的规律及其中用到的方法,依据上述三个尺度下裂隙网络模拟结果,编制了相应的Matlab计算机程序,最后建立了裂隙密度、长度以及隙宽的多尺度联系。(5)通过分析个旧锡矿东区375km2遥感影像和高松矿田地质图,统计断裂体系的空间分布特征,研究裂隙网络与矿体分布关系。根据大尺度和中尺度裂隙对比结果,矿体分布与大尺度第一组裂隙,与中尺度第二组、第三组裂隙分布关系密切,两者具有较好的一致性,结果表明在大、中两个尺度范围内矿田规模及矿体位置与裂隙分布具有尺度不变性。更多还原

【Abstract】 In the geological research field, fracture is a basic structure form which is widely developed in the upper crust rock. Fractures provide favorable conditions for ore-forming fluid to rise, disperse, penetrate and aggregate during the formation of mineral resources. In order to make scientific predictions of the spatial distribution of fracture extension, how to make use of existing structural geology and rock mechanics test data is a difficult problem in geological prospecting. In this paper the fracture networks is modeled under different scales on the basis of the previous work, and finally their mutual relation is established. This paper conducts research in the following aspects.(1) Fracture characteristic’s measurement of Gaosong field in Gejiu tin deposit.A variety of fractures exists widely in nature, and contains a number of attributes (such as length, width, direction, position, filling). Fracture position can be considered to meet the Poisson distribution, but the fracture formed in the multiple phrases and by overprint of several geological processes, fracture distribution exhibit complicated characteristics as we see in the field; fracture size (size, diameter length, chord length) has a close relationship with stress intensity and rock properties; unit normal vector of crack surfaces can be used to represent the fracture direction, which is closely related to the stress direction which impacts the rock (body) in the research process; the actual fracture density can use total number fractures per unit to measure, and it is showen a strong nonuniformity in general; fracture width influences the permeability of rock significantly in general, and closely relates with the fracture length; fracture displacement reflects the migration direction and distance in the course of of fracture formation, and it has significance for simulation of fracture formation. Fracture properties, including length, displacement, size, density, width and other major obey power-law distribution, logarithmic normal distribution, exponential distribution, gamma distribution, while direction obeys Fisher distribution and Binghanm distribution. On the basis of previous study, the internal rules of multi-scale fracture attributes through the case study are analysised, and then the method of calculating across-scales law by box counting method and two-point correlation function is discussed, and then sample size, exponential expansion, estimation method, influence factors, the estimation accuracy in the process of estimating the scaling-exponent are detailed analysised. In view of the relationship between the observability of fractre length and other properties, the scale-index of fracture length estimation is especially discussed in detail. According to multi-scale contact law and the used method, the corresponding computer programs are compiled by Matlab.(2) In this paper GEOFRAC method are Constructed, which is concluding Ordinary Kriging (OK), Sequential Gauss Simulation (SGS), Principal Component Analysis (PCA) method. Specific content include:decomposition properties, fracture density estimation, fracture position generation, the principal component analysis and inversion of fracture direction (strike and dip), geostatistics estimation of fracture orientation distribution, fracture property valuation comprehension, basic fracture surface connection. Fracture multiple attribute, the same position and scale-invariant features can be considered through making full use of fractured samples data of different dimension and scale.(3) Fracture3D distribution in Gaosong ore field.By using acoustic emission, CT, surface, tunnel, remote sensing, the geological data and fracture sample data of different dimensions and different scales are collected. Taking this as the foundation, a3D strata model of Gaosong field by using geological data are constructed, and the relations between small-scale crack length, the density cross-dimension conversion and properties(direction, length, width) are established by using CT data, and then fracture spatial situation distributions, corresponding to the direction of the spatial distribution of fracture network in Gaosong field are simulated by using tunnel, surface data, and fracture elements in the spatial distribution is connected, and then linear structure in Gejiu deposit are extracted by using remote sensing DEM data, multi-scale contact of fracture density and length according to the established fracture network distribution analysis of different scales, the results of these studies are verified in the Gaosong ore field.(4) Through stereological analysis, fracture length distributions meet some laws in given certain assumptions. The relationships between the fracture density, one-dimensional, two-dimensional density, three-dimensional density, density, surface density, linear density, bulk density are concluded, then fracture length and chord length are statisticalized, and then fracture direction, the observation surface direction, the trace ray angle are built, and then the relationship between fracture density, length and direction across dimension conversion are established, and the then technique route of the fracture properties of cross dimension conversion are designed, and finally the idea of computer algorithms are conceived, and all of this implementation program was written by Matlab.(5) Through analyzing the remote sensing images of a375km2in Gejiu tin deposit and Gaosong ore field geological map, the spatial distribution characteristics of fracture system, and the fracture network distribution relationship with orebody are researched. According to the large scale fracture modeling results, orebody is closely associated with the first set of fracture, and with second, third groups of fracture distribution in mesoscale. The results show they have a good consistency.更多还原