岩石缝网拓扑结构特征及破裂预测研究

岩石破裂是岩体工程中普遍存在的现象，研究岩石破裂演化规律，预测岩石破裂时间及其位置，对于基础工程建设、自然资源高效开发、地球深部断层活化认识与评价至关重要。岩石的破裂伴随着应变场的动态演化，且在破裂之前必然会有非均匀的变形，因此可通过监测应变场来预测破裂发生的时间和位置。本项目基于高能加速器CT试验系统在试验加载过程中实时获取岩石切片，采用数字体积相关技术，揭示位移场的变化规律；提出针对裂缝演化的数字图像方法，突破亚像素的裂缝识别，实现空间缝网的精细重构。结合有限元和位移反分析法进行逆向化数字建模，得到一种基于CT图像的非均匀应变场构建方法。采用拓扑结构参数定量表征岩石破裂过程，建立应力-缝网拓扑结构关系，对岩石破裂进行预测。本项目首次提出根据CT图像进行破裂预测的逆向化方法，揭示了岩石破裂的时空演化机制，建立了岩石破裂预测的新方法，突破了根据真实岩石破裂预测的技术瓶颈。

Rock fracturing is frequently encountered in the rock mechanical engineering. To investigate the rock fracturing evolution paradigm and predict the time and location of rock fracturing are of great importance to the infrastructure construction, efficient exploration of the natural resources and deep fault reactivation. Rock fracturing is accompanied by the dynamic evolution of the strain field. The non-uniform deformation is inevitable before fracturing. Therefore, the monitoring of strain field can be used to predict the fracturing time and location. The program will first use the linear accelerator CT scanner to obtain the rock CT slices. Then the digital volume correlation technique will be used to capture the displacement filed variations. A digital image framework for fracture evolution will be proposed to achieve the fracture detection at the sub pixel level and realize the elaborate reconstruction of the spatial fracture network. The finite element method based reverse digital modeling and displacement back analysis methodology will be used to obtain a non-uniform strain field construction method based on CT images. The topological structure parameters are used to quantitatively characterize the rock failure process, and the stress-fracture network topological structure relationship is established to predict the rock failure. In this program the reversal methodology of predicting fracturing from the CT images is proposed for the first time, which reveals the temporal and spatial evolution mechanism of rock fracture, establishes a new method of rock fracture prediction, and breaks through the bottleneck of how to real.