Aiming at optimization of bench height for open-pit mines in high-altitude cold regions, a multi-objective decision-making method considering the influence of freeze-thaw cycles was proposed. Based on parameters including shoveling equipment selection and toe burden, the bench height was preliminarily determined to be within the range of 12-18 m, and seven alternative schemes were proposed. Numerical simulation of slope stability after 25 freeze-thaw cycles was conducted by adopting GeoStudio software, and the safety factors of each scheme ranged from 1.196 to 1.214. Models for ore dilution and loss, drilling and blasting cost, and ore-rock transportation work were then established to calculate the performance values of each scheme in terms of dilution rate, loss rate, drilling cost, blast cost, and transportation work. On this basis, the entropy-weight method was used to determine the weight of each index, and the comprehensive scores of alternative schemes were calculated. The results show that the 15 m bench scheme achieves the highest comprehensive score (0.5166), and is identified as the optimal solution. Incorporating the mechanical effects of freeze-thaw cycles into the multi-objective evaluation framework, this method provides a quantitative decision-making basis for bench height design of open-pit mines in high-altitude cold regions.

As for underground engineering projects in shallowburied soft rock, extensive blast vibration monitoring was carried out along the axial direction on the tunnel ground surface. Based on the monitoring data, probability distribution analysis confirmed that the peak particle velocity (PPV) followed the lognormal distribution. Considering the randomness of PPV, the lognormal distribution theory and confidenclevel analysis method were introduced. A blast vibration safety prediction and control model was established by combining the PPV attenuation model with the probability distribution function, and then applied to the sensitive sections of the project. The results show that in a blasting scheme with the maximum charge per delay at 43.2 kg, the horizontal distance between sensitive point 2^# and the blast source should be greater than 35.044 m to satisfy the safety criterion of a 95% confidence; while in a scheme with the maximum charge per delay at 18.6 kg, that horizontal distance shall exceed 25.660 m to meet the 95% confidence. On the basis of this model, differentiated blasting control schemes are formulated for sensitive tunnel sections. This proposed method provides theoretical support for blasting operation for similar underground softrock projects.

To understand the splitting-tensile behavior of internal cross-cracks in rock, discrete element splitting simulation tests were carried out by using a particle flow code (PFC2D) based on the physical and mechanical parameters of rock obtained from laboratory tests. The results show that the tensile strength of specimens with pre-fabricated cracks is lower than that of intact specimens. When the angle between the main crack and the loading direction ranges from 0° to 60°, the peak strength of specimens decreases with the increase of the secondary crack angle; when the angle ranges from 60° to 90°, the peak strength increases with the increase of the secondary crack angle. In general, the tensile strength of specimens follows the rule that the larger the angle difference between the main and secondary cracks, the smaller the tensile strength. Under splitting loading, pre-fabricated cracks do not change the loading mode of specimens but only reduce their bearing strength. Micro-cracks in cross-crack specimens are dominated by tensile micro-cracks, which initiate near the tips of pre-fabricated cracks, propagate toward the loading ends and coalesce with increasing load, resulting in specimen failure. Irreversible transformation of internal energy occurs during rock failure. When the loading reaches the storage limit of elastic strain energy, the elastic strain energy is rapidly converted into dissipated energy for micro-crack initiation and propagation as well as particle movement. A large number of acoustic emission events occur simultaneously, and the number of micro-crack increases exponentially.

When drilling and blasting are adopted for excavation of high slopes, blasting-induced vibration signals are susceptible to noise interference and their effective features are difficult to extract accurately, a noise reduction method based on complete ensemble empirical mode decomposition with adaptive noise (CEEMDAN) combined with wavelet threshold is proposed. Blasting vibration signals are reprocessed based on mean ratio (MR), sample entropy, and correlation coefficient. The proposed method is verified by simulated and field measured blasting vibration signals. The simulation results show that the signal-to-noise ratio (SNR) of the proposed method reaches 27.80 dB and the root-mean-square error (RMSE) is 0.2091, presenting a superior noise-reduction performance compared to the other four methods. Engineering application results indicate that this proposed method can effectively suppress high-frequency noise and improve the resolution and stability of time-frequency analysis for blasting vibration signals. Further analysis reveals that the dominant frequency of blasting vibration signals is mainly within the range of 10－60 Hz with energy concentrated within 16－125 Hz; the dominant frequency of rockfall impact signals is mostly in the range of 40－720 Hz with more complex spectral structures, and its energy is mainly distributed in 62－1000 Hz.

Aiming at the inconsistency of grading results and difficulty in determining credibility for rock mass quality classification methods such as the RMR method, Q-system method and BQ method, a credibility evaluation method for rock mass quality classification based on analytic hierarchy process (AHP), fuzzy comprehensive evaluation (FCE) and D-S evidence theory is proposed. Firstly, the evaluation vector of rock mass quality grade is obtained via the AHP-FCE model. The evaluation vectors obtained by the three classification methods are taken as independent bodies of evidence. The D-S evidence theory and credibility weighting model are introduced to fuse and revise multi-source evidence, so as to obtain an optimal rock mass quality classification result integrating multi-source information. On this basis, a calculation platform for the credibility of rock mass quality classification is established with MATLAB GUIDE. Verification with 16 groups of measured data from No.2 Ping’an Phosphate Mine shows that the evaluation results obtained by the proposed method are more consistent with engineering practice than those from any single classification method. After revision by the credibility weighting model, the conflict coefficients of 16 groups of original evidence decrease from 0.60-0.82 to less than 0.01. The conflict problem of multi-source evidence is effectively solved, and the credibility of rock mass quality classification results is improved.

With the slopes of an open-pit mine in Inner Mongolia as the research background, the creep characteristics of soft rock in the slopes were investigated by means of field monitoring, laboratory test and theoretical analysis. Real-time monitoring results of the slopes from radar and GNSS found that the creep deformation is relatively larger on the eastern side of the northern slope, which requires key protection. When the axial stress exceeds the long-term strength, the soft rock goes through four deformation stages: instantaneous deformation, decelerating creep, steady-state creep and accelerated creep. Creep deformation and creep rate of soft rock are positively correlated with axial stress. A nonlinear creep model for slope soft rock is established by combining elastic elements, fractional-order element and nonlinear viscoplastic body (NVPB). This model can well describe the whole creep process of slope soft rock, with higher fitting accuracy than that of traditional Burgers model, and also demonstrate favorable applicability.