从岩石露头裂隙迹线估算裂隙三维空间方向

摘要

Two-dimensional information of fissure traces is gathered from outcrops through digital imaging to study the occurrence of fracture surfaces within rocks and simulate the spatial distribution of fissure networks inside rocks.These studies are of great significance to the prediction of subsurface resources (especially fluid resources) and the evaluation of stability subsurface space.Fissures within rocks can be divided into several groups according to their clustering patterns.It is assumed that the fissures in each group spread around a central value.The relations between occurrences of fissure surface and wall surface and dip of fissure trace are defined with the included angle between lens direction and normal line of the wall and are used to estimate the occurrence of a given fissure surface.Random function is used to define fissure distribution based on the Fisher distribution.Dips of fissure traces are used to delineate the three-dimensional distribution of fissure surfaces.The final step is to develop a consistent method for solving the inverse problem and controls the calculation error under 5 % by using simulated annealing method.The method had been applied to the investigation of fissures on outcrop in Gaosong ore field in GeJiu tin deposit.The fissures were divided into two sets with average dips of 49.0° (SD of 0.38) and 128.2° (SD of 0.32) respectively and their occurrences were estimated.The first set has a mean dip direction of 171.l°and a mean dip angle of 75.1° and the second has a mean dip direction of 338.3° and a mean dip angle of 3.9°.They fit well with field data.Simulation of fissure network was carried out based on the concept and yielded very positive results.Overall, the method shows great potential for direct estimation of 3D distribution of fissures according to fissure trace direction on outcrop.%通过提取岩石露头数字影像中裂隙痕迹分布的二维信息,研究岩石内部裂隙面产状特征,进而模拟岩石裂隙网络空间分布,在地下资源(尤其是流体资源)的预测和地下空间稳定性评价等领域具有重要的意义.根据岩石中裂隙分布多表现出丛聚规律的特征,将裂隙分为若干个组；假设每个组内部的裂隙围绕一个中心值分布,考虑镜头方向与岩墙面法线夹角,从裂隙面、岩墙面产状以及裂隙痕迹倾角三者关系入手,估算得到已知裂隙面产状.根据Fisher分布,利用随机函数产生裂隙分布,由裂隙痕迹倾角求得裂隙面三维空间分布,并利用模拟退火法对原问题求逆,误差控制在5％.根据个旧锡矿高松矿田岩石露头裂隙的特征,将岩体裂隙分为两组,平均倾角分别为49.0°和128.2°,倾角的标准方差为0.38和0.32,最后估算出Fisher分布常数K=200时裂隙面产状分别为171.1°/75.1°和338.3°＜3.9°,与现场调查数据吻合.据此对裂隙网络进行了模拟,取得了很好的预测效果,其结果对于根据岩石露头迹线方向直接估算裂隙的三维空间分布具有重要意义.