Microfractures in the coal matrix have complex spatial structure because of the joint effects of coalification and tectonism. The accurate modeling and characterization of microfractures directly affect mechanism analysis of coal rock fracturing and mining. In this study, taking the high-rank coal in the north of Qinshui Basin, structural characteristics of microfracture in intact and tectonic coals were analyzed systematically using 2D thin section, 3D micro-CT modeling and mercury intrusion based on digital image analysis and fractal theory. The result indicates that the response sensitivity of fractal parameters obtained from different methods to microfractures from two types of coal structures is different greatly. Single fractal dimension (D_f) in tectonic coal is higher than that of intact coal. Specially, for 2D thin sections, the indicative effect on the D_f of manually corrected images is better than that of automatically extracted and linear ones. D_f has a positive correlation with the pixel-based microfracture porosity, as well as generalized fractal parameters. Moreover, microfracture permeability based on the Cubic-Law increases with the increase of D_f. For 3D micro-CT modeling, the classical Watershed algorithm cannot segment the spatial microfracture network, which, however, can be segmented and characterized by the self-developed FracSC3D program. Spatial microstructural parameters including volume, surface and length at three-axis directions meet bifractal characteristics. Specially, for 2D micro-CT slices, with the increase of porosity, the multifractal spectrum width Δα also increased. Besides, the fractal parameter based on Menger model from mercury intrusion experiment can act as a good indicator for identifying microfractures, flow pores and diffusion pores.

中文翻译：

---

高阶煤裂缝网络的显微结构分析

由于煤化作用和构造作用的共同作用，煤基质中的微裂缝具有复杂的空间结构。微裂缝的准确建模和表征直接影响煤岩压裂和开采的机理分析。本研究以沁水盆地北部高阶煤为研究对象，基于数字图像分析和分形理论，利用二维薄片、3D显微CT建模和压汞等方法，系统分析了完整煤和构造煤微裂缝的结构特征。 . 结果表明，不同方法得到的分形参数对两类煤结构微裂缝的响应灵敏度差异很大。单分形维数 ( D _f) 在构造煤中高于完整煤。特别是对于二维薄片，手动校正图像对D _f的指示效果优于自动提取和线性图像。D _f与基于像素的微裂缝孔隙度以及广义分形参数呈正相关。此外，基于三次定律的微裂缝渗透率随着D _{f 的}增加而增加. 对于3D微CT建模，经典的Watershed算法无法对空间微裂缝网络进行分割，但可以通过自主开发的FracSC3D程序进行分割和表征。三轴方向的体积、表面和长度等空间微观结构参数满足双分形特征。特别地，用于2D显微CT切片，与孔隙率的增加，该多重分形谱宽度Δ α也增加。此外，压汞实验中基于Menger模型的分形参数可以作为识别微裂缝、流动孔和扩散孔的良好指标。