各向异性多孔介质数字岩心模型重构通用算法

General algorithm for reconstructing digital core model of anisotropic porous media

多孔介质材料内部的微观孔隙结构直接决定着其宏观的各类物理特性,岩石等天然多孔介质材料内部分布有复杂的孔隙结构,且呈现出较明显的各向异性特征,准确的描述和捕捉该特征对于岩石高精度数字岩心模型的构建以及后续的孔隙尺度模拟至关重要.基于先前所提出的重构加速框架,选用单点概率函数、两点概率函数和线性路径函数等统计函数来对孔隙结构特征进行定量化描述,通过对孔隙结构在不同方向上的分布特征进行独立的提取表征,提出了一种能够适用于岩石等各类天然各向异性多孔介质数字岩心模型构建的改进模拟退火算法.以具有显著各向异性特征的人工合成二维切片图像及真实碳酸盐岩和砂岩三维CT扫描图像作为测试参考样本,分别采用传统方法和本研究所提出方法构建了其相应的数字岩心模型,并对两者的连通性及不同方向上孔隙结构的分布情况和渗透特性进行了全面的综合对比分析,结果表明:相比于传统方法,本研究所提出方法在确保重构岩心模型具有良好连通性的前提下,能够对不同方向上孔隙结构分布的差异性进行更加准确的捕获和表征,且重构岩心模型各个方向上的渗透性也与参考真实岩心表现出了较高程度的一致性,实现了高精度还原重建参考图像中孔隙结构各向异性特征的目标,从而验证了本研究所提出的各向异性多孔介质数字岩心模型重构方法的准确性和有效性.

The micro pore structure in porous media materials directly determines its macro physical properties.There are complex pore structures in natural porous media materials such as rocks,which show obvious anisotropic characteristics.Accurate description and capture of this characteristic is very important for the construction of high-precision digital core model of rocks and subsequent pore scale simulation.Based on the reconstruction acceleration framework proposed previously,the statistical functions such as single-point probability function,two-point probability function and linear path function are selected to describe the characteristics of pore structure.Through the independent extraction and characterization of the distribution characteristics of pore structure in different directions,an improved simulated annealing method is proposed,which is suitable for the reconstruction of digital core models of various natural anisotropic porous media such as rocks.Taking the synthetic two-dimensional slice image and the three-dimensional CT scanning image of real sandstone and carbonate rock with significant anisotropy as the test reference sample,the corresponding digital core model is constructed by using the traditional method and the method proposed in this study,and the connectivity,distribution and permeability of pore structure in different directions are comprehensively compared and analyzed.The results show that compared with the digital core model constructed by the traditional method,the method proposed in this study can more accurately capture and characterize the differences of pore structure distribution in different directions on the premise of ensuring the good connectivity of the reconstructed core model,and the permeability of the reconstructed core model in all directions also shows a high degree of consistency with that of the reference real core,achieving the goal of high-precision restoration and reconstruction of the anisotropic characteristics of pore structure in the reference image,which verifies the accuracy and effectiveness of the reconstruction method of anisotropic porous media digital core model proposed in this study.