Microstructural damage is an important cause of macrofracture in fissured rocks. To investigate the microscopic fracture mechanism of fissured rocks, a fissured rock model was established using the CZM-FDEM method. The model considers the mechanical behaviour of cohesive elements between particles within the rock and the propagation of cracks through the damage of cohesive elements. In the damage evolution stage of the cohesive element ontological relationship, the damage variable D is introduced to describe the damage degree of the unit stiffness, which in turn characterises the microscopic damage features of the model. The results show that the stress–strain curve morphology and strength characteristics of the numerical model are in high agreement with the indoor test results, and the model exhibits significant advantages in crack extension, which verifies the applicability of the method in rock fracture simulation. The microscopic damage field exhibits X-shaped conjugate fracture characteristics, and the crack inclination leads to differences in fracture paths by changing the stress field distribution at the crack tip. The number of cohesive element fractures on the microscopic scale shows a high correlation with the macroscopic strength, and the cleavage inclination angle directly affects the load carrying capacity of the structure by regulating the fracture element rate.

Economic investment and mechanical characteristics represent primary limitations to the widespread adoption of cemented tailings backfill. This study introduces an approach of using quarry waste to generate rock powder, which is subsequently mixed with tailings to produce cemented tailings-rock powder backfill (CTRPB). This method achieves cost efficiency while facilitating synergistic waste management. Through uniaxial compression testing, the influences of rock powder type and content on the mechanical behavior, failure modes, and energy transformation of CTRPB were examined. A piecewise damage constitutive model was developed to explore damage progression mechanisms. Findings demonstrate that the incorporation of rock powder markedly improves the deformation resistance of CTRPB. Both uniaxial compressive strength (UCS) and elastic modulus exhibit an initial increase followed by a decline as content rises, with optimal levels and enhancement effects differing across various rock powders. The failure process of CTRPB encompassed four distinct phases, where rock powder amplified post-peak energy release and fragmentation behaviors. Energy assessment indicated that rock powder significantly boosted energy storage and dissipation capacities, elevating total energy density, elastic energy density, and dissipated energy density by 252.17%, 213.87%, and 478.99%, respectively. The formulated piecewise damage constitutive model correlated well with experimental data in the pre-peak regime. Damage evolution may be categorized into four stages, and damage values can act as indicators for assessing failure conditions in the resource utilization of mining waste and rock powder-tailings backfill technology. This research offers a theoretical foundation for the resource-oriented use of mine waste and rock powder-tailings backfill technology.

Water inrush and mud burst disasters pose severe challenges to the safe and efficient construction of underground engineering. Water inflow prediction is closely related to drainage design, disaster prevention and control, and the safety of the surrounding ecological environment. Thus, assessing the water inflow accurately is of importance. This study proposes a Bayesian Optimization-Long Short-Term Memory (BOA-LSTM) recurrent neural network for predicting tunnel water inflow. The model is based on four input parameters, namely tunnel depth (H), groundwater level (h), rock quality designation (RQD), and water-richness (W), with water inflow (WI) as the single-output variable. The model first processes and analyzes the data, quantitatively characterizing the correlations between input parameters. The tunnel water inflow is predicted using the long short-term memory (LSTM) recurrent neural network, and the Bayesian optimization algorithm (BOA) is employed to select the hyperparameters of the LSTM, primarily including the number of hidden layer units, initial learning rate, and L2 regularization coefficient. The modeling process incorporates a five-fold cross-validation strategy for dataset partitioning, which effectively mitigates overfitting risks and enhances the model’s generalization capability. After a comprehensive comparison among a series of machine learning models, including a long short-term memory recurrent neural network (LSTM), random forest (RF), back propagation neural network (BP), extreme learning machine (ELM), radial basis function neural network (RBFNN), least squares support vector machine (LIBSVM), and convolutional neural network (CNN), BOA-LSTM performed excellently. The proposed BOA-LSTM model substantially surpasses the standard LSTM and other comparative models in tunnel water inflow prediction, demonstrating superior performance in both accuracy and generalization. Hence, it provides a reference basis for tunnel engineering water inflow prediction.

Corrosion of rock bolts in engineering exhibits random spatial distribution characteristics. To elucidate the influence mechanism of stochastic corrosion on the surface of rock bolts on the propagation behavior of ultrasonic guided waves, this study establishes a finite element model of rock bolts that incorporates stochastic corrosion characteristics. The coupled effects of corrosion depth and area ratio on guided wave propagation characteristics, time-domain response, energy distribution, and wave velocity variation are systematically investigated. Results indicate that corrosion depth and area ratio synergistically deteriorate guided wave morphology, transforming the stress field from symmetric and uniform to asymmetric and spiral. Reflections, scattering, and mode conversion induced by defects lead to a significant increase in the attenuation rate of pulse amplitude, with the two parameters governing the vertical interaction intensity and horizontal interference scope, respectively. Analysis of the Hilbert curve reveals that corrosion characteristics disrupt energy concentration. Under constant corrosion depth, an increase in area ratio disperses energy toward delayed scattered waves, while under constant area ratio, greater corrosion depth reduces the peak amplitude of the envelope curve. Overall, the energy integral exhibits an increasing trend with the degree of corrosion, whereas the peak-to-peak wave velocity shows a declining trend. The established multivariate nonlinear model accurately describes the coupled influence of corrosion parameters on wave velocity. This stochastic corrosion model overcomes the limitations of traditional simplified models and provides critical theoretical support for parameter calibration and engineering application of ultrasonic guided wave technology in the quantitative assessment of rock bolt corrosion.

Rock bolts in complex underground environments are prone to randomly distributed localized corrosion damage, posing a severe threat to the stability of anchorage structures. Ultrasonic guided waves represent a core technology for nondestructive testing of bolt corrosion; however, existing studies predominantly focus on the influence of total corrosion extent on guided wave propagation characteristics, neglecting the mechanism of defect size discretization effects under a fixed total corrosion amount. To this end, a numerical model of a free rock bolt containing random pitting defects with different individual pit areas is established. Under strictly controlled total corrosion area and corrosion depth, with individual defect area as the sole variable, the propagation behavior of the L(0,1) mode ultrasonic guided wave at a center frequency of 50 kHz is systematically analyzed. Through time-domain waveform analysis, Hilbert transform, energy statistics, and peak-to-peak velocity calculation, the mechanisms by which individual defect area influences waveform distortion, energy scattering, and peak-to-peak velocity are revealed. The results indicate that: under a fixed total corrosion amount, increasing the individual pit area progressively exacerbates waveform distortion and scattered clutter; the cumulative energy generally exhibits a declining trend, yet extremely small-sized dense pitting causes weak scattered energy to be temporally dispersed, resulting in abnormally higher values than the control group; the amplitude attenuation rate shows no obvious monotonic pattern; the reduction in peak-to-peak velocity, within a certain range, exhibits a nonlinear characteristic that first recovers rapidly and then stabilizes as the individual pit area increases, which can be described by a logarithmic relationship. The findings can provide a theoretical basis for ultrasonic guided wave detection of random corrosion in rock bolts.

Tailings dam safety is a cornerstone of green and sustainable mine development. The instability of such facilities directly threatens the safety of surrounding ecosystems, human lives, and property, contravening the Sustainable Development Goals regarding clean water and terrestrial ecosystem protection. Rainfall is the primary external factor triggering tailings dam instability, yet the failure mechanisms may differ fundamentally depending on the temporal distribution of rainfall. Against the background of the Poyang Lake Basin, this study integrates observed rainfall data from 2020 to 2025 with a survey of 101 active tailings dams (of which Class IV dams account for 43.6% and upstream-method construction accounts for 88.1%), on the basis of which a representative Class IV upstream-method tailings dam is selected as the prototype. A coupled unsaturated seepage and stability numerical model is constructed, and two sets of rainfall scenarios are designed: one varying the rainfall peak position coefficient t and the other varying the rainfall duration. The results indicate that infiltration induced by short-duration intense rainfall is confined to shallow strata. The earlier the rainfall peak, the more fully the deep infiltration develops and the greater the factor of safety reduction, with the critical variable governing stability degradation being the length of time available for deep redistribution after the rainfall peak. Long-duration sustained rainfall drives an overall rise of the phreatic line, and the seepage response is dominated by the average rainfall intensity. Extending the duration cannot compensate for the attenuation of the infiltration driving force caused by the reduced rainfall intensity, and the core factor determining disaster risk is whether the average rainfall intensity can exceed the drainage threshold of the deep low-permeability layers. Within the Poyang Lake Basin, concentrated heavy continuous rainfall during the Meiyu period lasting 20 to 40 days represents the decisive working condition governing tailings dam stability and should be prioritized in flood-season safety supervision. The rainfall-driven instability mechanisms revealed in this study provide a theoretical basis for establishing differentiated monitoring and early-warning thresholds and for the sustainable operation of tailings dams.