Abstract Mineralogical composition has a great influence on the mechanical behavior and the micro-cracking process of crystalline rocks for CO2 and natural gas storage. This study numerically investigates the influence of mineralogical composition (i.e., quartz content) of a dominantly felsic phaneritic igneous rock with respect to rock strength and the associated micro-cracking behavior using a grain-based modeling approach in two-dimensional Particle Flow Code (PFC2D). First, numerical specimen models with different mineralogical compositions are generated. The generated numerical models have the same geometry of the assembled grain structure to minimize the effect of grain scale heterogeneity on the simulation results. Micro-parameters previously calibrated to match the macro-properties of the Bukit Timah granite are then assigned to the numerical models. In the numerical simulation of uniaxial compression tests, the strength and Young's modulus are found to increase with the increase of quartz content in the numerical model, while the Poisson's ratio and the maximum volumetric strain gradually decrease. The simulated strength behavior is in good agreement with the laboratory test results obtained from previous studies. However, the crack damage stress seems not to be affected by the quartz content. The total number of generated micro-cracks is also found to gradually increase as the quartz content in the numerical model increases. The rock strength shows a good correlation with the total number of generated micro-cracks. Two mechanisms are identified to initiate the nearly vertical macroscopic fractures which are generated in tension. At last, the influence of spatial distribution of mineral grains on the simulated strength property and micro-cracking behavior is discussed.

中文翻译：

---

矿物成分对结晶岩强度及微裂纹行为影响的数值研究

摘要 矿物成分对用于储存二氧化碳和天然气的结晶岩的力学行为和微裂化过程有很大影响。本研究使用二维粒子流代码 (PFC2D) 中基于颗粒的建模方法，以数值方式研究了主要长英质显晶火成岩的矿物组成（即石英含量）对岩石强度和相关微裂纹行为的影响）。首先，生成具有不同矿物成分的数值样本模型。生成的数值模型具有相同的组装晶粒结构几何形状，以最大限度地减少晶粒尺度异质性对模拟结果的影响。然后将先前校准以匹配武吉知马花岗岩的宏观特性的微观参数分配给数值模型。在单轴压缩试验的数值模拟中，发现强度和杨氏模量随着数值模型中石英含量的增加而增加，而泊松比和最大体积应变逐渐减小。模拟的强度行为与先前研究中获得的实验室测试结果非常一致。然而，裂纹损伤应力似乎不受石英含量的影响。随着数值模型中石英含量的增加，产生的微裂纹总数也逐渐增加。岩石强度与产生的微裂纹总数具有良好的相关性。确定了两种机制来引发在张力下产生的几乎垂直的宏观裂缝。最后讨论了矿物颗粒的空间分布对模拟强度特性和微裂纹行为的影响。