An open-pit landslide risk identification method driven by spatiotemporal distribution evolution of rock mass mechanical parameters

Abstract

This invention discloses a method for identifying landslide risks in open-pit mines driven by the spatiotemporal distribution evolution of rock mass mechanical parameters, relating to the fields of rock mechanics and open-pit mine slope stability analysis. The main components include: constructing a spatial distribution model of rock mass using the Geophysical Indicator (GSI) based on borehole core data and geostatistical methods; measuring the physical parameters and conducting uniaxial compression tests on rock samples from the landslide area; establishing the weakening relationship of rock mechanical parameters with immersion time; calculating the degradation equations of rock mass cohesion and friction angle with immersion time and spatial coordinates based on the GSI spatial distribution model and the Hoek-Brown criterion; constructing a three-dimensional spatial distribution model of rock mass mechanical parameters considering immersion time, and performing three-dimensional numerical simulations to analyze rock mass stability and obtain low safety factors and damaged areas. This invention solves the technical problem of the inability to accurately identify catastrophic events such as landslides and collapses caused by the degradation of rock mass mechanical parameters due to the weakening of soft rock strata caused by immersion in water in existing open-pit mines.

Description

Technical Field

[0001] This invention belongs to the field of rock mechanics and open-pit mine slope stability, and specifically discloses a method for identifying open-pit mine landslide risks driven by the spatiotemporal distribution evolution of rock mass mechanics parameters. Background Technology

[0002] Due to the anisotropy of rocks and the complex mechanical conditions resulting from mutual support and compression within open-pit mine environments, analyzing the evolution mechanism of soft rock slope stability under water-bearing conditions by considering the spatial variability of its non-homogeneous mechanical parameters is crucial for mitigating engineering risks. Methods for characterizing the spatial variability of rock mass mechanical parameters include image mapping, statistical methods, rock mass quality classification, and spatial correlation, combined with deep learning algorithms such as octree theory for refined 3D modeling. However, a method for integrating field data with slope model simulations is lacking. Summary of the Invention

[0003] To address the technical problem of inaccurate identification of landslides and collapses caused by the degradation of rock mass mechanical parameters due to water immersion in soft rock strata in existing open-pit mines, this invention provides a landslide risk identification method driven by the spatiotemporal distribution evolution of rock mass mechanical parameters. This invention targets the challenge of water-sensitive catastrophic changes in open-pit mine slopes. First, it systematically collects rock samples and conducts indoor immersion tests to reveal the quantitative decay law of mechanical parameters such as uniaxial compressive strength and elastic modulus with immersion time. Based on this, it combines geostatistical methods with field borehole data to construct a three-dimensional non-uniform distribution model of rock mass mass, accurately characterizing the GSI parameter features at different spatial coordinate points. Subsequently, it establishes a quantitative relationship between rock mass mechanical parameters and immersion time and spatial GSI values, embedding this relationship into COMSOL to obtain a three-dimensional spatiotemporal distribution model of rock mass mechanical parameters. Finally, by comparing the probability distribution of high-risk areas output by the model with the actual landslide areas, it guides the adjustment of open-pit mine boundaries and optimization of drainage measures, achieving the prediction and prevention of water-sensitive catastrophic changes in slopes and improving the mine's safety production assurance capabilities.

[0004] The technical means employed in this invention are as follows:

[0005] A method for identifying landslide risks in open-pit mines driven by the spatiotemporal distribution evolution of rock mass mechanical parameters includes the following steps: S1. Collect core data from boreholes at different locations in the open-pit mine and rock samples from the landslide area; S2. Based on the borehole core data and geostatistical methods, a spatial distribution model of rock mass GSI was constructed; and physical parameters and uniaxial compression tests were performed on the rock samples from the landslide area. S3. Based on the results of uniaxial compression tests, the weakening relationship of rock mechanical parameters with immersion time is established. According to the GSI spatial distribution model, combined with the Hoek-Brown criterion, the degradation equations of the cohesion and friction angle of the rock mass with immersion time and spatial coordinates are calculated. S4. The quantitative degradation equation is embedded in COMSOL software to construct a three-dimensional spatial distribution model of rock mass mechanical parameters that takes into account the immersion time. Based on the three-dimensional spatial distribution model of rock mass mechanical parameters, a three-dimensional numerical simulation is performed to analyze the rock mass stability and obtain the low safety factor and damage area.

[0006] Furthermore, the method also includes: S5. Compare the low safety factor and damaged areas with the actual landslide areas on site to verify the accuracy of the three-dimensional spatial distribution model of the rock mass mechanics parameters. Use the verified three-dimensional spatial distribution model of the rock mass mechanics parameters to predict and infer high-risk unstable areas in the open-pit mine.

[0007] Furthermore, based on the borehole core data and geostatistical methods, a spatial distribution model of rock mass quality (GSI) is constructed, including: S21. Based on the Mask R-CNN deep learning network, identify single-row cores from core images, and then identify core segments greater than or equal to 10cm from single-row cores; S22. Based on the identified core segments, calculate the RQD value for each image using the following formula:

[0008] In the formula, RQD is the rock quality index, m is the number of core segments with a length ≥10cm in the core box, li is the length of each core segment with a length ≥10cm, and L is the drilling footage. S23. Based on the RQD value, a borehole database containing a positioning table, an inclinometer table, and an RQD table is compiled in 3DMine software. The positioning table and the inclinometer table are used to jointly determine the three-dimensional trajectory of the borehole, and the RQD table is used to store the RQD value obtained by automatic analysis based on core photos. S24. Define the research scope of the borehole database, combine the high-precision surface morphology obtained by UAV oblique photography, and use the distance power inverse ratio interpolation method to obtain the block model of RQD; S25. Based on engineering geological surveys and mine data, determine other evaluation factors in the RMR grading system for rock blocks at different locations, including uniaxial compressive strength score R1, joint spacing score R3, joint condition R4, groundwater condition R5, and joint attitude R6, and sum them up as R0. S26. Obtain the RQD value and cumulative value R0 at each location in the RQD block model to construct the RMR three-dimensional block spatial distribution model, and obtain the spatial distribution model of GSI based on the relationship between GSI and RMR.

[0009] Furthermore, the RQD values ​​and cumulative values ​​R0 at various locations in the RQD block model are obtained to construct a three-dimensional spatial distribution model of the RMR block, including calculating the geological strength coefficient GSI according to the following formula:

[0010]

[0011] Among them, R1 is the uniaxial compressive strength score of the rock block, R2 is the RQD value score, R3 is the joint spacing score, R4 is the joint condition, R5 is the groundwater condition, and R6 is the joint attitude.

[0012] Furthermore, based on the results of uniaxial compression tests, the weakening relationship of rock mechanical parameters with immersion time was established, including: S31. Uniaxial compression tests were conducted on rock samples with different immersion times, and their basic mechanical parameters and compressive strength were recorded. Elastic modulus E, Poisson's ratio μ; S32. Calculate the weakening relationship of mechanical parameters with immersion time according to the following equations:

[0013] In the formula, P(t) represents the mechanical parameters after immersion time t, including compressive strength. Elastic modulus E, Poisson's ratio μ , , All parameter values ​​were obtained by fitting experimental results; S33. Derive the cohesion and friction angle of rock mass using formulas from the Hoek-Brown criterion.

[0014]

[0015]

[0016]

[0017]

[0018] In the formula, φ is the friction angle, c is the cohesion, and D is the coefficient of rock mass disturbance caused by blasting vibration and excavation stress release. The Hoek-Brown material constants for intact rock. This is used to demonstrate the effect of confining pressure conditions on the MC strength parameter.

[0019] S34. The mapping relationship between the output rock mass mechanical parameters, friction angle φ, cohesion c, and GSI and immersion time t, is as follows:

[0020]

[0021] In the formula, a, b, and c are all determined by fitting the surface of the experimental results.

[0022] Furthermore, a three-dimensional spatial distribution model of rock mass mechanical parameters considering immersion time is constructed, including: S41. Using the solid mechanics module of the three-dimensional space dimension in COMSOL software, write the code of the elastic softening constitutive model based on damage mechanics, compile it into a dynamic link library (DLL) file, call the external material interface in COMSOL Multiphysics, load the DLL file, and realize the dynamic update of damage variables during the iteration process; S42. Based on the geological data of the open-pit mine, a three-dimensional geological model of the open-pit mine was constructed and imported into the COMSOL software. Corresponding rock mechanics parameter values ​​were assigned to different lithological regions. For the soft rock region, the calculated rock mechanics friction angle φ and cohesion c were substituted into the formula to obtain the results. S43. In COMSOL software, set boundary conditions for the mesh of the three-dimensional geological model of the open-pit mine, fix the bottom of the model to constrain vertical displacement, constrain the lateral displacement boundaries in the X and Y directions to limit horizontal movement, set the slope as a free boundary to allow displacement by gravity and external disturbances, and load the initial ground stress through the gravity stress field. S44. For the three-dimensional geological model of the open-pit mine, steady-state study is conducted using COMSOL. The strength reduction method is adopted to proportionally reduce the rock mass mechanical parameter values ​​and simulate the damage accumulation process. The plastic zone and critical instability state of the slope are monitored. The final output results are the three-dimensional distribution of safety factor and damage value, as well as the spatial distribution of rock mass mechanical parameters friction angle φ and cohesion c.

[0023] Furthermore, the accuracy of the three-dimensional spatial distribution model of the rock mass mechanical parameters is verified. The verified three-dimensional spatial distribution model of the rock mass mechanical parameters is then used to predict and infer high-risk unstable areas in open-pit mines, including: S51. Obtain the location of the landslide area delineated by the UAV model at the site, and compare this location with the area with the smaller safety factor value in the simulation results to verify the accuracy of the model; S52. Using the validated three-dimensional spatial distribution model of rock mass mechanics parameters for prediction, and based on the prediction results, for areas where the safety factor value is less than the preset threshold, take monitoring or early warning measures in advance at the open-pit mine site.

[0024] Compared with the prior art, the present invention has the following advantages: The technical solution provided by this invention collects core data from open-pit mine boreholes and rock samples from landslide areas, combines geostatistical methods to construct a spatial distribution model of rock mass GSI, and conducts physical parameter measurements and uniaxial compression tests. It establishes the weakening law of rock mechanical parameters with immersion time and derives the degradation equation of rock mechanical parameters. It embeds this model into COMSOL software to construct a three-dimensional model and conducts numerical simulation. After verifying the accuracy of the model by comparison with actual landslide areas in the field, it predicts high-risk unstable areas, forming a complete landslide risk identification method. This provides scientific and technological support for solving the problems of rock mechanical parameter degradation and landslide disasters caused by soft rock immersion, and effectively improves the safety and stability of open-pit mining. Attached Figure Description

[0025] To more clearly illustrate the technical solutions in the embodiments of the present invention or the prior art, the drawings used in the description of the embodiments or the prior art will be briefly introduced below. Obviously, the drawings described below are some embodiments of the present invention. For those skilled in the art, other drawings can be obtained based on these drawings without creative effort.

[0026] Figure 1 This is a schematic diagram of landslide risk identification based on a three-dimensional simulation of the spatiotemporal distribution of water-weakening and temporal distribution of open-pit mine rock mass mechanical parameters provided in an embodiment of the present invention. Figure 2 This is the spatial distribution model of RQD provided in the embodiments of the present invention; Figure 3 This is the spatial distribution model of RMR provided in the embodiments of the present invention; Figure 4 This refers to the GSI spatial distribution model provided in this embodiment of the invention. Figure 5 This is the relationship between rock compressive strength and immersion time provided in the embodiments of the present invention; Figure 6 This is the correspondence between the rock elastic modulus and immersion time provided in the embodiments of the present invention; Figure 7 This is the correspondence between Poisson's ratio of rock and immersion time provided in the embodiments of the present invention; Figure 8 This is the correspondence between rock mass cohesion, immersion time, and GSI provided in the embodiments of the present invention; Figure 9This is the correspondence between the rock mass friction angle, immersion time, and GSI provided in the embodiments of the present invention; Figure 10 This is a spatiotemporal evolution model of the cohesion of rock mass on open-pit mine slopes provided in this embodiment of the invention; Figure 11 This is a spatiotemporal evolution model of the friction angle of open-pit mine slope rock mass provided in this embodiment of the invention; Figure 12 This is a spatiotemporal evolution model of the simulated safety factor for open-pit mine slopes provided in this embodiment of the invention; Figure 13 This is a spatiotemporal evolution model of simulated damage values ​​for open-pit mine slopes provided in this embodiment of the invention; Figure 14 This is a comparison diagram between the simulated safety factor of the open-pit mine slope and the actual landslide area provided in this embodiment of the invention. Detailed Implementation

[0027] To enable those skilled in the art to better understand the present invention, the technical solutions of the present invention will be clearly and completely described below with reference to the accompanying drawings of the embodiments of the present invention. Obviously, the described embodiments are only some embodiments of the present invention, and not all embodiments. Based on the embodiments of the present invention, all other embodiments obtained by those skilled in the art without creative effort should fall within the scope of protection of the present invention.

[0028] like Figure 1 As shown, this embodiment of the invention provides a method for identifying open-pit mine landslide risks driven by the spatiotemporal distribution evolution of rock mass mechanical parameters, including the following steps: S1. Collect borehole core data and rock samples from landslide areas at different locations in the open-pit mine.

[0029] Preferably, the core sample point data in step S1 of this example includes borehole location coordinates, elevation, borehole depth, lithology, core recovery rate, rock quality index RQD, joint and fracture development characteristics, weathering degree, and groundwater development status; wherein, the joint and fracture development characteristics include the number of joint groups, occurrence, spacing, continuity, aperture, roughness, and filling status, etc. The above core sample point data is used to calculate the geological strength coefficient GSI, which is obtained by formula (1). (1) In the formula, GSI is the geological strength coefficient, R1 is the uniaxial compressive strength score of the rock block, R2 is the RQD value score, R3 is the joint spacing score, R4 is the joint condition, R5 is the groundwater condition, and R6 is the joint attitude.

[0030] Preferably, before collecting rock samples from the landslide area in step S1 of this example, an in-situ test needs to be conducted at this location to determine the water content of the rocks. The size of the sampled rocks is generally not less than 15cm × 15cm × 15cm to ensure the subsequent processing of standard rock samples.

[0031] S2. Using the core data collected in step S1, a spatial distribution model of rock mass GSI is constructed using geostatistical methods; and physical parameters and uniaxial compression tests are performed on the samples collected in step S1. The physical parameters mainly include density, water content, porosity, and P-wave velocity.

[0032] Preferably, in this embodiment, the construction of the rock mass quality GSI spatial distribution model includes: S21. Based on the Mask R-CNN deep learning network, identify single-row cores from core images, and then identify core segments greater than or equal to 10cm from the single-row cores.

[0033] S22. Based on the core segments identified in step S21, the RQD value of each image can be calculated statistically. The RQD value can be calculated using formula (2). (2) In the formula, RQD is the rock quality index, m is the number of core segments with a length ≥10cm in the core box, li is the length of each core segment with a length ≥10cm, and L is the drilling footage.

[0034] S23. Based on the RQD values ​​from step S22, a borehole database containing a location table, an inclinometer table, and an RQD table is compiled in 3DMine software. The location table and inclinometer table jointly determine the three-dimensional borehole trajectory, while the RQD table stores the RQD values ​​obtained through automatic analysis of core images.

[0035] S24. Define the research scope for the borehole database, and use the inverse power distance interpolation method to obtain the block model of RQD by combining the high-precision surface morphology obtained by UAV oblique photography. It should be noted that this application focuses on the open-pit mine slope and potential landslide impact area to be evaluated. The research scope is determined based on the effective control range of the borehole data, combined with the high-precision surface morphology obtained by UAV oblique photography, the slope top and toe boundaries, and the main engineering geological interfaces. Specifically, on the plane, the target slope range and potential instability impact area are the core, covering the borehole distribution control area; vertically, the upper slope top boundary, lower slope toe boundary, and potential slip depth are used as constraints to form a three-dimensional research scope for spatial interpolation and block modeling. Through this method, the RQD interpolation results and the subsequent GSI spatial distribution model can be guaranteed to have clear data support and engineering geological significance. S25. Based on engineering geological surveys and mine data, determine other evaluation factors in the RMR grading system for rock blocks at different locations, including uniaxial compressive strength score R1, joint spacing score R3, joint condition R4, groundwater condition R5, and joint attitude R6, and sum them up as R0.

[0036] S26. The RQD values ​​R2 at each position in the RQD block model established in step S24 and the cumulative value R0 in S25 are used to construct the RMR three-dimensional block spatial distribution model according to formula (1), and the spatial distribution model of GSI is obtained according to the relationship between GSI and RMR.

[0037] Preferably, the determination of the physical and mechanical parameters of the rock sample in this example includes: S27. Perform wave velocity tests on the processed standard rock samples with a diameter of 50 mm and a height of 100 mm to screen out samples with similar wave velocities, thereby reducing errors caused by internal differences in the samples themselves.

[0038] S28. Perform a quality determination on the samples screened in step S27, and record it as... m And use formula (3) to calculate its density in its natural state. ρ After drying in a drying oven for 48 hours, the quality was measured and recorded as follows. m d Its dry density is obtained using the same calculation method as described above. ρ d , (3) S29. Based on the dried rock samples from step S28, group them and prepare rock samples for different immersion times, and record their final mass. m wi ,in i This represents the immersion time, and the density of the rock sample under different immersion times is calculated using the same formula as in S27 above. ρ wi ; S31. Based on the rock samples with different immersion times in step S29, conduct uniaxial compression tests and record their basic mechanical parameters and compressive strength. σ c elastic modulus E Poisson's ratio μ ; S32. Preferably, the weakening law of each mechanical parameter in this example with immersion time can be calculated according to equation (4): (4) In the formula, P(t) Mechanical parameters (compressive strength) after immersion time t σ c elastic modulus E Poisson's ratio μ A0, B0, and c0 are parameter values ​​obtained by fitting experimental results; S33. The cohesion and friction angle of the rock mass can be derived using formulas (5)-(9) from the Hoek-Brown criterion, where (8) is the friction angle. φ The calculation formula (9) is the cohesive force. c Calculation formula.

[0039] (5) (6) (7) (8) (9) In the formula, D is the coefficient of rock mass disturbance caused by blasting vibration and excavation stress release. It is taken as 1.0 for conventional blasting, 0.7 for mechanical excavation, and 0.8~0.9 for controlled blasting. This is used to demonstrate the effect of confining pressure conditions on the MC strength parameter. The Hoek-Brown material constant for intact rock is used to characterize the failure characteristics of intact rock materials. In this embodiment, it can be selected according to the rock type by referring to existing empirical Hoek-Brown value tables.

[0040] S34. The process of obtaining rock mass mechanical parameters using the HB strength criterion in S32 above can be performed in batches using a program developed with MATLAB software. The output rock mass mechanical parameters include the friction angle. φ The mapping relationship between cohesion c and GSI and immersion time t is given by the following equations (10) and (11).

[0041] (10) (11) In the formula, α1, β1, γ1, α2, β2, and γ2 are all determined by fitting the surfaces from the experimental results. Specifically, these parameters are obtained through nonlinear surface fitting driven by experimental data. The specific method is as follows: 1. Based on uniaxial compression tests of rock samples under different immersion times, the variation laws of uniaxial compressive strength, elastic modulus, and Poisson's ratio of intact rock with immersion time are obtained. 2. Multiple representative GSI values ​​are selected and combined with different immersion time nodes. The rock mass friction angle φ and cohesion c under each group of conditions are calculated in batches using the Hoek-Brown criterion, forming discrete data point sets (tGSI,φ) and (t,GSI,c). 3. Target surface functions are established with t and GSI as independent variables and φ and c as dependent variables, and the fitting parameter α1 is solved using the nonlinear least squares fitting method in MATLAB. 、 β1 、 γ1 、 α2 、 β2 、 γ2 。 4. Determine the final parameter values ​​with the goal of minimizing the sum of squared residuals, and use the coefficient of determination R², adjusted coefficient of determination, and root mean square error RMSE to evaluate the fitting accuracy; when the fitting accuracy meets the requirements, the obtained parameters are used as the final fitting parameters.

[0042] S4. The quantitative degradation equations of friction angle and cohesion are embedded in COMSOL software to construct a three-dimensional spatial distribution model of rock mass mechanical parameters that takes into account the immersion time. Based on the three-dimensional spatial distribution model of rock mass mechanical parameters, a three-dimensional numerical simulation is performed to analyze the stability of the rock mass and obtain the low safety factor and the damage area.

[0043] Preferably, constructing a three-dimensional spatial distribution model of rock mass mechanical parameters in COMSOL software that considers immersion time includes: S41. Using the three-dimensional solid mechanics module in COMSOL software, an elastic softening constitutive model based on damage mechanics is written in C language, compiled into a dynamic link library (DLL), and loaded into COMSOL Multiphysics via an external material interface (EMI) to achieve dynamic updates of damage variables during iteration. It should be noted that this elastic softening constitutive model based on damage mechanics is not a pre-built constitutive model in COMSOL software, but a custom constitutive model constructed based on continuous damage mechanics theory and combined with the tensile-compressive failure mechanism of rock. The EMI program is written in C language, compiled into a DLL file, and then implemented in COMSOL Multiphysics. Specifically, this model is based on the isotropic linear elastic relationship of undamaged rock mass, uses tensile equivalent strain and compressive equivalent strain as driving variables for damage evolution, uses the maximum tensile stress criterion and the Mohr-Coulomb type compression-shear criterion to determine damage initiation, and updates the elastic modulus through damage variables to construct the elastic softening constitutive relationship. In terms of numerical implementation, the external material interface program reads the current strain state and material parameters in real time, and updates the damage variables, stress tensor and tangent stiffness matrix to achieve dynamic solution of material damage evolution during finite element iteration.

[0044] S42. Based on the open-pit mine geological data, construct a three-dimensional geological model of the open-pit mine and import it into COMSOL software. Focus on different lithological regions and assign corresponding rock mass mechanical parameter values ​​to different lithologies. For soft rock regions, substitute the rock mass mechanical friction angle from S33. φ Formula for cohesive force c; S43. In COMSOL software, boundary conditions need to be set for the three-dimensional model mesh in step S42 above. Fix the bottom of the model to constrain vertical displacement, constrain the lateral displacement boundaries in the X and Y directions to limit horizontal movement, set the slope as a free boundary to allow displacement caused by gravity and external disturbances, and load the initial ground stress through the gravity stress field.

[0045] S44. For this model, steady-state studies are conducted using COMSOL. The strength reduction method is employed to proportionally reduce the rock mass mechanical parameter values ​​and simulate the damage accumulation process. The plastic zone and critical instability state of the slope are monitored. The final output includes the safety factor, the three-dimensional distribution of damage values, and the friction angle of the rock mass mechanical parameter. φ Spatial distribution diagram of cohesion c.

[0046] Preferably, in this embodiment, the open-pit mine landslide risk identification method driven by the spatiotemporal distribution evolution of rock mass mechanical parameters includes step S5, which verifies the accuracy of the proposed model and predicts and infers high-risk unstable areas in open-pit mines, as follows: S51. The location of the landslide area on site is delineated using the UAV model, and this location is compared with areas in the simulation results that have lower safety factor values ​​(higher damage values) to verify the accuracy of the model. In this application, the 'comparison' mainly refers to the consistency of the spatial location of the on-site landslide area and the simulated hazardous area in the open-pit mine slope model. Specifically, after delineating the on-site landslide area based on the UAV model, it is compared with areas with low safety factor and / or concentrated damage areas in the simulation results. The main focus is on whether their relative positions, slope segment ranges, and elevation distributions in the slope are consistent, thereby verifying the accuracy of the model. The low safety factor area is defined as an area where the safety factor is lower than a preset threshold. This preset threshold can be determined based on the slope safety level, load combination, and the "Technical Specification for Slope Engineering of Non-Coal Open-Pit Mines (GB 51016–2014)". Based on the analysis in this case study and the paper, for a Class I overall slope under load combination I (self-weight + groundwater), areas with FS < 1.20 can be defined as low-safety-factor areas, where FS < 1.00 is an instability area, and areas with 1.00 ≤ FS < 1.20 are high-risk areas that do not meet design requirements. Damage concentration areas are those where damage variables reach a preset damage threshold and are spatially continuously clustered. Furthermore, in this embodiment, the spatial consistency between the on-site landslide area and the low-safety-factor areas in the simulation results is compared to verify the accuracy of the model; in some implementations, damage concentration areas can also be used to further verify the model results.

[0047] S52. Based on the model accuracy in step S51, the simulation results of the model can be used to take monitoring or other early warning measures in advance at the open-pit mine site for some areas with small safety factor values ​​(not observed on site) to prevent collapse in the area.

[0048] The following specific application examples will further illustrate the solution and effects of the present invention.

[0049] This embodiment collected and organized data from 72 boreholes, obtaining 5010 core images. Each core box represents one image, with a total length of 26627.12m. Each core image recorded the core recovery rate. The core box was 1m long, covering the entire open-pit mine (3581 m × 2101 m × 521 m). The average block size was 15m. Using 3DMine software, combined with high-precision surface morphology obtained from UAV oblique photography, the RQD block model obtained through inverse power law interpolation was as follows: Figure 2 As shown.

[0050] S25. Based on engineering geological surveys and mine data, determine other evaluation factors in the RMR grading system for rock blocks at different locations, including uniaxial compressive strength score R1, joint spacing score R3, joint condition R4, groundwater condition R5, and joint attitude R6, and sum them up as R0.

[0051] Uniaxial compressive strength, Brazilian tensile strength, elastic modulus, and Poisson's ratio tests were conducted on samples of various lithologies to provide a basis for determining the RMR index value. A table of RMR grading system evaluation factor values ​​(excluding RQD values) was compiled, and the total score R0 was calculated. The RQD values ​​of each region in step S24 were then converted into the corresponding values ​​R2 in the RMR evaluation system.

[0052] S26. Summing the RQD value R2 with the cumulative value R0 in S25, and constructing the RMR three-dimensional block spatial distribution model according to formula (1), the RMR block spatial distribution model is as follows. Figure 3 Based on the relationship between GSI and RMR in formula (1), the spatial distribution model of GSI is as follows: Figure 4 As shown.

[0053] Preferably, this embodiment uses mudstone samples, and its physical parameters (mass before and after immersion in water) are measured, and its density is calculated using the mass, so as to provide data support for the subsequent application of the HB criterion.

[0054] S31. Based on the rock samples with different immersion times in step S29, conduct uniaxial compression tests and record their basic mechanical parameters and compressive strength. σ c elastic modulus E Poisson's ratio μ ; S32. Preferably, the weakening curves of each mechanical parameter in this example with immersion time are referenced. Figure 5-7 The goodness of fit of the curves is above 0.98, and they can be calculated according to the following equations (4)-(6). (4) (5) (6) The aforementioned mechanical parameters are important indicators of a rock's resistance to instability. The gradual weakening of rock mechanical parameters plays a very important role in open-pit mining and slope safety.

[0055] S32. Substitute the aforementioned parameter data into the generalized Hawke-Brown criterion, referring to formulas (7)-(11), and calculate the fitted surface of the cohesion and friction angle of the mudstone mass as a function of immersion time and GSI, as follows: Figure 8-9 As shown. The quantitative equations for the fitted surface can be calculated according to equations (12)-(13). The calculation method for the mechanical parameters of sandstone is the same as that for mudstone.

[0056] (7) (8) (9) (10) (11) (12) (13) S33. Based on step S26, each coordinate point of the three-dimensional model in this embodiment has a corresponding GSI value. Therefore, by substituting into formulas (12)-(13), the rock mechanics parameters of the model can be spatially assigned.

[0057] S41. Based on step S33, the rock mechanics parameters of the three-dimensional model are a function related to the immersion time. By inputting variables with different immersion times (0, 7, 14, 30, 60, and 90 days in this embodiment), the spatial distribution model of the cohesion and friction angle that varies with the immersion time is calculated.

[0058] Preferably, this embodiment provides as follows: Figure 10-11 The spatiotemporal evolution model of rock mass mechanical parameters under water immersion conditions is shown.

[0059] Figure 10 The evolution of cohesion shows a continuous decreasing trend with increasing immersion time, and the deterioration of mudstone is significantly stronger than that of sandstone. Under dry conditions, the average cohesion of sandstone and mudstone areas are 4.5 MPa and 4.0 MPa, respectively. After immersion for 7 days, the cohesion of sandstone drops to 4.0 MPa, a decrease of 11.1%; while that of mudstone drops sharply to 1.7 MPa, a decrease of 56.3%. As the immersion time is extended to 90 days, the maximum cohesion of the entire model decreases from 5.0 MPa to 3.0 MPa, a decrease of 40%, and the minimum cohesion decreases from 1.0 MPa to 0.3 MPa, a decrease of 70%.

[0060] Figure 11For the evolution of the internal friction angle, under dry conditions, the internal friction angle ranges from 45° to 51°. After 7 days of immersion in water, the internal friction angle in the sandstone area decreased from 50.5° to 48.0°, a decrease of 3.9%; in the mudstone area, it decreased from 46.5° to 43.0°, a decrease of 7.5%. During the immersion period of 14–90 days, the overall maximum internal friction angle decreased from 54.0° to 40.5°, a decrease of 25.0%; the minimum value decreased from 37.5° to 24.0°, a decrease of 36.0%.

[0061] S42. This embodiment employs a numerical simulation based on an elastic softening constitutive model derived from damage mechanics. The model is compiled into a dynamic link library (DLL) using the C programming language and implemented in COMSOL Multiphysics via an external material interface. This allows for the dynamic transfer and updating of damage variables during iterative calculations, resulting in the calculated... Figure 12 The spatiotemporal evolution model of the safety factor shown and Figure 13 The spatiotemporal evolution model of damage values.

[0062] Preferably, in this embodiment Figure 12 The FS value gradually decreases with increasing immersion time. In the dry state (0 days), the overall FS value is between 1.20 and 1.35, indicating a stable slope. From 14 to 60 days of immersion, the low FS area evolves from discrete points into a continuous zone, with FS values ​​in some areas dropping to 0.80–1.01. At 90 days of immersion, the cumulative structural damage within the mudstone zone reaches a critical level, with FS values ​​in some areas dropping below 0.80. The low FS area extends to adjacent sandstone strata and the upper part of the slope, indicating a significant decrease in overall stability.

[0063] Preferably, in this embodiment Figure 13 Damage accumulates and spreads gradually with immersion time. In the initial dry state, damage primarily stems from excavation disturbance and self-weight stress, resulting in relatively low overall damage. After 7 days of immersion, damage values ​​increase slightly, and microcracks begin to propagate near the surface. From 14 to 60 days of immersion, as water infiltration accelerates porosity and fracture development, high-damage zones expand and begin to merge spatially within shallow mudstone and weak interlayers. After 90 days of immersion, high-damage zones are widespread across the slope, with damage values ​​approaching 1.0 in some areas, indicating severe weakening of the rock mass and impending failure.

[0064] Preferably, in order to verify the decisive influence of "considering the spatial variability of the rock mass" on the accuracy of the calculation results, this embodiment sets up two comparison schemes. Both schemes are set to a 7-day immersion condition, and the calculation results are compared with the landslide data actually monitored on site.

[0065] Option 1: This embodiment adopts a non-uniform model that considers spatial variability. Based on the GSI spatial distribution model constructed in step S26, geostatistical interpolation is used to assign independent lithological parameters to each block after 7 days of water immersion and weakening, so as to realize point-to-point non-uniform assignment of rock mass properties, which can accurately characterize the spatial distribution of weak interlayers.

[0066] Option 2: Adopt a homogeneous model that does not consider spatial variability, ignore the non-homogeneity inside the rock mass, and use the traditional layered homogenization treatment. Treat the entire mudstone layer and sandstone layer as homogeneous materials, and assign them the average values ​​of the rock mass mechanical parameters after 7 days of water immersion.

[0067] Figure 14 Table 1 presents the simulation results of a non-uniform model considering spatial variability under 7 days of immersion in water, and their correspondence with the actual situation on site. Table 1 provides a detailed comparison of the calculated data from the two schemes with the actual site conditions. It can be seen that, compared to Scheme 2 in this embodiment, Scheme 1 more accurately identifies the factors significantly affected by water immersion in the mudstone area, resulting in a low-safety zone with a banded distribution that spatially highly overlaps with the landslide location observed on site. Scheme 2, however, uses parameter averaging, causing the calculated damage to present as a large-area, blurred, arc-shaped destruction. Therefore, the method studied in this patent can eliminate the calculation errors caused by the homogenized model, achieving the goal of accurate early warning of local instability disasters under complex hydrogeological conditions in open-pit mines.

[0068] Table 1 Comparison of Optimized Data for Scheme 1 and Scheme 2

[0069] This embodiment establishes a three-dimensional non-uniform block model of rock mass quality based on the spatial distribution of GSI and the mechanism of water-rock interaction, considering lithological differences (mudstone / sandstone) and the duration of immersion. Combined with numerical simulation technology, mechanical parameter fields, safety factor fields, and damage fields at different time points are constructed. Analysis shows that water-induced damage exhibits strong localization characteristics, with the mudstone layer showing rapid deterioration in the early stages of immersion, making it a key controlling factor for slope instability. This method can effectively identify potential landslide risk zones that expand over time (such as the new potential instability hazards appearing on the northeast slope in this embodiment), providing a basis for engineering design and disaster monitoring in open-pit mines under complex hydrological conditions.

[0070] Finally, it should be noted that the above embodiments are only used to illustrate the technical solutions of the present invention, and not to limit them; although the present invention has been described in detail with reference to the foregoing embodiments, those skilled in the art should understand that modifications can still be made to the technical solutions described in the foregoing embodiments, or equivalent substitutions can be made to some or all of the technical features; and these modifications or substitutions do not cause the essence of the corresponding technical solutions to deviate from the scope of the technical solutions of the embodiments of the present invention.

Claims

1. A method for identifying landslide risk in open-pit mines driven by the spatiotemporal distribution evolution of rock mass mechanical parameters, characterized in that, Includes the following steps: S1. Collect core data from boreholes at different locations in the open-pit mine and rock samples from the landslide area; S2. Based on the borehole core data and geostatistical methods, a spatial distribution model of rock mass GSI was constructed; and physical parameters and uniaxial compression tests were performed on the rock samples from the landslide area. S3. Based on the results of uniaxial compression tests, the weakening relationship of rock mechanical parameters with immersion time is established. According to the GSI spatial distribution model, combined with the Hoek-Brown criterion, the degradation equations of the cohesion and friction angle of the rock mass with immersion time and spatial coordinates are calculated. S4. The quantitative degradation equation is embedded in COMSOL software to construct a three-dimensional spatial distribution model of rock mass mechanical parameters that takes into account the immersion time. Based on the three-dimensional spatial distribution model of rock mass mechanical parameters, a three-dimensional numerical simulation is performed to analyze the rock mass stability and obtain the low safety factor and damage area.

2. The method for identifying open-pit mine landslide risk driven by the spatiotemporal distribution evolution of rock mass mechanical parameters according to claim 1, characterized in that, The method further includes: S5. Compare the low safety factor and damaged areas with the actual landslide areas on site to verify the accuracy of the three-dimensional spatial distribution model of the rock mass mechanics parameters. Use the verified three-dimensional spatial distribution model of the rock mass mechanics parameters to predict and infer high-risk unstable areas in the open-pit mine.

3. The method for identifying open-pit mine landslide risk driven by the spatiotemporal distribution evolution of rock mass mechanical parameters according to claim 1, characterized in that, Based on the borehole core data and geostatistical methods, a spatial distribution model of rock mass quality (GSI) was constructed, including: S21. Based on the Mask R-CNN deep learning network, identify single-row cores from core images, and then identify core segments greater than or equal to 10cm from single-row cores; S22. Based on the identified core segments, calculate the RQD value for each image using the following formula: In the formula, RQD is the rock quality index, m is the number of core segments with a length ≥10cm in the core box, li is the length of each core segment with a length ≥10cm, and L is the drilling footage. S23. Based on the RQD value, a borehole database containing a positioning table, an inclinometer table, and an RQD table is compiled in 3DMine software. The positioning table and the inclinometer table are used to jointly determine the three-dimensional trajectory of the borehole, and the RQD table is used to store the RQD value obtained by automatic analysis based on core photos. S24. Define the research scope of the borehole database, combine the high-precision surface morphology obtained by UAV oblique photography, and use the distance power inverse ratio interpolation method to obtain the block model of RQD; S25. Based on engineering geological surveys and mine data, determine other evaluation factors in the RMR grading system for rock blocks at different locations, including uniaxial compressive strength score R1, joint spacing score R3, joint condition R4, groundwater condition R5, and joint attitude R6, and sum them up as R0. S26. Obtain the RQD value and cumulative value R0 at each location in the RQD block model to construct the RMR three-dimensional block spatial distribution model, and obtain the spatial distribution model of GSI based on the relationship between GSI and RMR.

4. The method for identifying open-pit mine landslide risk driven by the spatiotemporal distribution evolution of rock mass mechanical parameters according to claim 3, characterized in that, To construct a three-dimensional spatial distribution model of the RMR (Real-Mean Ratio), the RQD values ​​and cumulative values ​​R0 at various locations in the RQD block model are obtained, including the calculation of the geological strength coefficient GSI according to the following formula: Among them, R1 is the uniaxial compressive strength score of the rock block, R2 is the RQD value score, R3 is the joint spacing score, R4 is the joint condition, R5 is the groundwater condition, and R6 is the joint attitude.

5. The method for identifying open-pit mine landslide risk driven by the spatiotemporal distribution evolution of rock mass mechanical parameters according to claim 1, characterized in that, Based on the results of uniaxial compression tests, the weakening relationship of rock mechanical parameters with immersion time was established, including: S31. Uniaxial compression tests were conducted on rock samples with different immersion times, and their basic mechanical parameters and compressive strength were recorded. Elastic modulus E, Poisson's ratio μ; S32. Calculate the weakening relationship of mechanical parameters with immersion time according to the following equations: In the formula, P(t) represents the mechanical parameters after immersion time t, including compressive strength. Elastic modulus E, Poisson's ratio μ , , All parameter values ​​were obtained by fitting experimental results; S33. Derive the cohesion and friction angle of rock mass using formulas from the Hoek-Brown criterion. In the formula, φ is the friction angle, c is the cohesion, and D is the coefficient of rock mass disturbance caused by blasting vibration and excavation stress release. The Hoek-Brown material constants for intact rock. This is used to demonstrate the effect of confining pressure conditions on the MC strength parameter. S34. The mapping relationship between the output rock mass mechanical parameters, friction angle φ, cohesion c, and GSI and immersion time t, is as follows: In the formula, a, b, and c are all determined by fitting the surface of the experimental results.

6. The method for identifying open-pit mine landslide risk driven by the spatiotemporal distribution evolution of rock mass mechanical parameters according to claim 1, characterized in that, A three-dimensional spatial distribution model of rock mass mechanical parameters considering immersion time is constructed, including: S41. Using the solid mechanics module of the three-dimensional space dimension in COMSOL software, write the code of the elastic softening constitutive model based on damage mechanics, compile it into a dynamic link library (DLL) file, call the external material interface in COMSOL Multiphysics, load the DLL file, and realize the dynamic update of damage variables during the iteration process; S42. Based on the geological data of the open-pit mine, a three-dimensional geological model of the open-pit mine was constructed and imported into the COMSOL software. Corresponding rock mechanics parameter values ​​were assigned to different lithological regions. For the soft rock region, the calculated rock mechanics friction angle φ and cohesion c were substituted into the formula to obtain the results. S43. In COMSOL software, set boundary conditions for the mesh of the three-dimensional geological model of the open-pit mine, fix the bottom of the model to constrain vertical displacement, constrain the lateral displacement boundaries in the X and Y directions to limit horizontal movement, set the slope as a free boundary to allow displacement by gravity and external disturbances, and load the initial ground stress through the gravity stress field. S44. For the three-dimensional geological model of the open-pit mine, steady-state study is conducted using COMSOL. The strength reduction method is adopted to proportionally reduce the rock mass mechanical parameter values ​​and simulate the damage accumulation process. The plastic zone and critical instability state of the slope are monitored. The final output results are the three-dimensional distribution of safety factor and damage value, as well as the spatial distribution of rock mass mechanical parameters friction angle φ and cohesion c.

7. The method for identifying open-pit mine landslide risk driven by the spatiotemporal distribution evolution of rock mass mechanical parameters according to claim 2, characterized in that, Verify the accuracy of the three-dimensional spatial distribution model of the rock mass mechanics parameters, and use the verified three-dimensional spatial distribution model of the rock mass mechanics parameters to predict and infer high-risk unstable areas in open-pit mines, including: S51. Obtain the location of the landslide area delineated by the UAV model at the site, and compare this location with the area with the smaller safety factor value in the simulation results to verify the accuracy of the model; S52. Using the validated three-dimensional spatial distribution model of rock mass mechanics parameters for prediction, and based on the prediction results, for areas where the safety factor value is less than the preset threshold, take monitoring or early warning measures in advance at the open-pit mine site.

Images