The deterioration of rock mass in the Three Gorges reservoir area results from the coupled damage effects of macro-micro cracks and dry-wet cycles, and the coupled damage progression can be characterized by energy release rate. In this study, a series of dry-wet cycle uniaxial compression tests was conducted on fractured sandstone, and a method was developed for calculating macro-micro damage (DR) and energy release rates (YR) of fractured sandstone subjected to dry-wet cycles by considering energy release rate, dry-wet damage and macro-micro damage. Therewith, the damage mechanisms and complex microcrack propagation patterns of rocks were investigated. Research indicates that sandstone degradation after a limited cycle count primarily exhibits exsolution of internal fillers, progressing to grain skeleton alteration and erosion with increased cycles. Compared with conventional methods, the DR and YR methodologies exhibit heightened sensitivity to microcrack closure during compaction and abrupt energy release at the point of failure. Based on DR and YR, the failure process of fractured sandstone can be classified into six stages: stress adjustment (I), microcracks equal closure (II), nonlinear slow closure (III), low-speed extension (IV), rapid extension (V), and macroscopic main fracture emergence (VI). The abrupt change in damage energy release rate during stage V may serve as a reliable precursor for inducing failure. The stage-based classification may enhance traditional methods by tracking damage progression and accurately identifying rock failure precursors. The findings are expected to provide a scientific basis for understanding damage mechanisms and enabling early warning of reservoir-bank slope failure. (c) 2025 Institute of Rock and Soil Mechanics, Chinese Academy of Sciences. Published by Elsevier B.V. This is an open access article under the CC BY-NC-ND license (http://creativecommons.org/licenses/by-nc-nd/ 4.0/).

The natural property of rock material, whether impact occurs, is the key influencing factor of the occurrence of rock burst disaster. To accurately assess rock burst proneness, this study focuses on typical sandstone as the research object. Uniaxial cyclic loading and unloading tests were conducted to measure the elastic strain energy accumulated in sandstone under different stress levels and a relationship between elastic strain energy and stress level was established. The results show that: (1) The peak stress under cyclic loading and unloading conditions is slightly lower than the uniaxial compressive strength. With an increase in the number of cycles, the internal damage of sandstone continues to accumulate, and the mechanical properties such as compressive strength continue to deteriorate; (2) With an increase in stress, the input strain energy, elastic strain energy, and dissipated strain energy also increase; (3) When the stress is low, the increase in elastic strain energy is large and shows a steady growth; with an increase in stress, the increase of elastic strain energy decreases; (4) The square of stress at any time has a good linear relationship with elastic strain energy. According to the relationship obtained from the test, the elastic strain energy at the peak stress time can be obtained; (5) A new criterion for assessing rock burst proneness is proposed: residual energy release rate index W-T, which characterizes the energy release per unit time when the rock burst occurs. The intervals for evaluating the rock burst proneness of the residual energy release rate index W-T are as follows: W-T 2, indicating strong rock burst proneness; and (6) The rationality of the proposed residual energy release rate index W-T is verified by the multi-index method and the multi-sample method, and the proposed residual energy release rate index is used to determine the rock burst proneness of 10 kinds of rock samples. The evaluation accuracy is shown to be high, and it can reflect the actual rock burst proneness (c) 2024 Institute of Rock and Soil Mechanics, Chinese Academy of Sciences. Published by Elsevier B.V. This is an open access article under the CC BY-NC-ND license (http://creativecommons.org/licenses/by-nc-nd/ 4.0/).

Micron-scale crack propagation in red-bed soft rocks under hydraulic action is a common cause of engineering disasters due to damage to the hard rock-soft rock-water interface. Previous studies have not provided a theoretical analysis of the length, inclination angle, and propagation angle of micron-scale cracks, nor have they established appropriate criteria to describe the crack propagation process. The propagation mechanism of micron-scale cracks in red-bed soft rocks under hydraulic action is not yet fully understood, which makes it challenging to prevent engineering disasters in these types of rocks. To address this issue, we have used the existing generalized maximum tangential stress (GMTS) and generalized maximum energy release rate (GMERR) criteria as the basis and introduced parameters related to micron-scale crack propagation and water action. The GMTS and GMERR criteria for micron-scale crack propagation in red-bed soft rocks under hydraulic action (abbreviated as the Wmic-GMTS and Wmic-GMERR criteria, respectively) were established to evaluate micron-scale crack propagation in red-bed soft rocks under hydraulic action. The influence of the parameters was also described. The process of micron-scale crack propagation under hydraulic action was monitored using uniaxial compression tests (UCTs) based on digital image correlation (DIC) technology. The study analyzed the length, propagation and inclination angles, and mechanical parameters of micron-scale crack propagation to confirm the reliability of the established criteria. The findings suggest that the Wmic-GMTS and Wmic-GMERR criteria are effective in describing the micron-scale crack propagation in red-bed soft rocks under hydraulic action. This study discusses the mechanism of micron-scale crack propagation and its effect on engineering disasters under hydraulic action. It covers topics such as the internal-external weakening of nano-scale particles, lateral propagation of micron-scale cracks, weakening of the mechanical properties of millimeter-scale soft rocks, and resulting interface damage at the engineering scale. The study provides a theoretical basis for the mechanism of disasters in red-bed soft-rock engineering under hydraulic action. (c) 2024 Institute of Rock and Soil Mechanics, Chinese Academy of Sciences. Production and hosting by Elsevier B.V.

Accurate prediction of rockburst proneness is one of challenges for assessing the rockburst risk and selecting effective control measures. This study aims to assess rockburst proneness by considering the energy characteristics and qualitative information during rock failure. Several representative rock types in cylindrical and cuboidal sample shapes were tested under uniaxial compression conditions and the failure progress was detected by a high-speed camera. The far-field ejection mass ratio (FEMR) was determined considering the qualitative failure information of the rock samples. The peak-strength energy impact index and the residual elastic energy index were used to quantitatively evaluate the rockburst proneness of both cylindrical and cuboidal samples. Further, the performance of these two indices was analyzed by comparing their estimates with the FEMR. The results show that the accuracy of the residual elastic energy index is significantly higher than that of the peak-strength energy impact index. The residual elastic energy index and the FEMR are in good agreement for both cylindrical and cuboidal rock materials. This is because these two indices can essentially reflect the common energy release mechanism characterized by the mass, ejection velocity, and ejection distance of rock fragments. It suggests that both the FEMR and the residual elastic energy index can be used to accurately measure the rockburst proneness of cylindrical and cuboidal samples based on uniaxial compression test. (c) 2024 Institute of Rock and Soil Mechanics, Chinese Academy of Sciences. Production and hosting by Elsevier B.V. This is an open access article under the CC BY-NC-ND license (http://creativecommons.org/ licenses/by-nc-nd/4.0/).