Microscopic cracks identification and nonlinear creep model of shale under creep-disturbance loading

• The software called Fracture Geometry Feature Extraction Tool is developed herein. • The relationship between macroscopic creep characteristics and microscopic damage of shale was clarified using SEM experiments and digital image processing. • The variation law of geometrical characteristics of cracks with disturbance amplitude and disturbance frequency during rock creep is described quantitatively. • A rock disturbance creep model is developed considering the variation of geometrical characteristics of cracks In this study, the action of creep damage evolution law of deep shale under the influence of disturbance was considered as the research object, and using a self-developed rock disturbance creep test rig, disturbance characteristics of rock creep disturbance were examined using different experiments. The creep of rock burst fracture was examined using scanning electron microscopy. The mesoscopic damage crack of rock creep disturbance syndrome was obtained. To further study the law of crack propagation during shale creep fracture, fracture evolution characteristics during rock disturbance creep were analyzed using a self-developed Fracture Geometry Feature Extraction Tool. Furthermore, the macro- and meso-creep damage evolution law of deep shale was determined. The results showed that the disturbance conditions accelerated the damage and fracture process of shale creep, and compared to the disturbance frequency, the disturbance amplitude was more sensitive to the damage and fracture process of shale creep. Under the increase of disturbance amplitude of 2.5∼5.5MPa, the crack width and weighted average area at the fracture site decrease, while under the increase of disturbance frequency of 1∼3Hz, the crack width and weighted average area at the fracture site increase. A nonlinear creep model of rock disturbance based on the geometrical characteristics of cracks was established, and the overall performance of the model's fitting degree was excellent, and the fitting degree was approximately 93%. The findings of this study provide significant insights for ensuring the long-term stability control of roadway surrounding rocks.