Abstract The sizes of microstructural features determine the mechanical behaviors of rocks, but the complex nature of their structures frequently hinders the precise quantification of the feature sizes. This study presents a digital-image-based quantification for evaluating the representative sizes of the microstructures in rocks. This method, using a madogram function, allows obtaining a single length value that uniquely represents the internal scales of the constituents in a heterogeneous material based on 3D X-ray computed tomographic images. The madogram function has the advantage of being applicable to grayscale images without any post-processing, which overcomes the limited applicability of previous correlation methods. The method was first validated in patterned and random heterogeneous domains that were synthetically prepared. Then, we further validated the method using original 16-bit μ-CT images of 14 rock samples with the consideration of domain size and image resolution. The measurements of the geophysical properties were performed to investigate their correlation with the representative sizes of the microstructures in rocks.

中文翻译：

---

使用 3D X 射线计算机断层扫描图像参数化岩石中显微结构特征的代表性尺寸

摘要 微观结构特征的大小决定了岩石的力学行为，但其结构的复杂性往往阻碍了对特征尺寸的精确量化。本研究提出了一种基于数字图像的量化方法，用于评估岩石中微结构的代表性尺寸。这种使用 madogram 函数的方法允许基于 3D X 射线计算机断层扫描图像获得唯一代表异质材料中成分的内部尺度的单个长度值。madogram 函数的优点是适用于灰度图像，无需任何后处理，克服了以往相关方法适用性有限的问题。该方法首先在合成制备的图案化和随机异质域中得到验证。然后，考虑到域大小和图像分辨率，我们使用 14 个岩石样本的原始 16 位 μ-CT 图像进一步验证了该方法。进行地球物理特性的测量以研究它们与岩石中微结构的代表性尺寸的相关性。