Loose fractured rock mass parameter acquisition method based on sub-particle scale modeling and simulation
By using a sub-particle scale modeling and simulation method, the error problem in obtaining mechanical parameters of soil-rock mixtures was solved, achieving high-precision parameter acquisition, reducing costs, and providing a reliable reference for underground engineering design.
Patent Information
- Application Number
- CN202511583914.X
- Authority / Receiving Office
- CN · China
- Patent Type
- Applications(China)
- Current Assignee / Owner
- Filing Date
- 2025-10-31
- Publication Date
- 2026-01-23
AI Technical Summary
Existing technologies for studying the mechanical properties of soil-rock mixtures suffer from errors due to differences in size or distribution between indoor tests and on-site conditions. They also lack comparative analysis of various soil-rock mixtures, leading to decreased accuracy in obtaining mechanical parameters of soil-rock mixtures and high costs for in-situ testing.
A sub-particle scale-based modeling and simulation method was adopted. A three-dimensional model was constructed by scanning soil-rock mixture samples, the soil and rock were segmented, and micromechanical parameters were obtained by combining indoor screening tests and strength tests. Multi-scale modeling was then performed to simulate the mechanical properties of soil-rock mixtures.
It improves the accuracy of obtaining mechanical parameters of soil-rock mixtures, reduces testing costs, and provides more accurate design references for underground engineering.
Smart Images
Figure CN121389686A_ABST
Abstract
Description
TECHNICAL FIELD
[0001] The application belongs to the technical field of geotechnical and rock mechanics, and particularly relates to a loose broken rock mass parameter acquisition method based on sub-particle scale modeling and simulation. BACKGROUND
[0002] The statements in this section merely provide background information related to the application and do not necessarily constitute prior art.
[0003] In recent years, with the rapid development of economy and the continuous advancement of infrastructure construction, the transportation network continues to extend, and the construction scale and difficulty of tunnel engineering are also increasing. In particular, in some complex geological conditions in mountainous areas, the burial depth of some tunnels reaches the kilometer level, which brings great challenges to engineering design and construction. In these areas, faults and adverse geological phenomena are particularly common, and the strength and stability of the soil and stone mixture in the fault are difficult to accurately predict. Such geological conditions not only increase the difficulty and cost of construction, but also may cause landslides, water inrush and other geological disasters, which seriously threaten the life safety of construction personnel and the overall quality of the project. The fault fracture zone essentially belongs to a soil and stone mixture, which is a heterogeneous multiphase material, and its mechanical properties often determine the mechanical behavior characteristics of the fault fracture zone during the engineering process, so it is of great engineering significance to study it.
[0004] Because the rock blocks and matrix soil in the fault fracture zone show obvious resistance difference during drilling, it is difficult to obtain complete samples. Therefore, domestic and foreign scholars mainly rely on laboratory tests and numerical simulation methods to study the influencing factors of the mechanical properties of soil and stone mixtures. Some scholars have studied the deformation and fracture mechanism of soil and stone mixtures through large-scale laboratory direct shear tests, triaxial compression tests, etc. Some scholars have further studied the meso-mechanical behavior and failure mechanism of soil and stone mixtures through simulation methods such as finite element, boundary element or finite difference method, and revealed the influence of factors such as stone content, block stone inclination and block stone size on the mechanical properties of soil and stone mixtures.
[0005] Although the test and simulation methods for studying soil and stone mixtures have achieved certain results, there are still some problems, such as the error caused by the difference between the laboratory test and the field size or distribution. That is, in the current test and simulation methods for studying soil and stone mixtures, the influence of the actual field and the test equipment on the test is often ignored, and there is also a lack of comparative analysis of various soil and stone mixtures, resulting in a decrease in the accuracy of the mechanical parameters of soil and stone mixtures. The main problem is that the in-situ test of the strength parameters of soil and stone mixtures in current underground engineering is too high in cost, and there is a certain error between the laboratory test and the field due to the difference in size or distribution. SUMMARY
[0006] In order to solve the above problems, the application provides a loose broken rock mass parameter acquisition method based on sub-particle scale modeling and simulation, and the application improves the specific implementation process in the process of acquiring the mechanical parameters of the soil-rock mixture, tests the actual field samples, improves the accuracy of acquiring the mechanical parameters of the soil-rock mixture through numerical simulation, a low-cost method, and comparative analysis of different rock masses, and ensures high accuracy to measure various strength parameters of the soil-rock mixture, thereby providing a certain reference for underground engineering design.
[0007] According to some embodiments, the application provides a loose broken rock mass parameter acquisition method based on sub-particle scale modeling and simulation, which adopts the following technical scheme: The loose broken rock mass parameter acquisition method based on sub-particle scale modeling and simulation comprises the following steps: Acquire soil-rock mixture samples and complete rock mass samples in a target area, and acquire modeling parameters based on the soil-rock mixture samples; Scan the soil-rock mixture samples to construct a mixture three-dimensional model, and segment the soil and stone based on the mixture three-dimensional model to acquire simulation analysis parameters; Perform indoor screening tests on the soil-rock mixture samples to obtain stone samples and soil samples; Analyze the composition of the stone samples and the complete rock mass samples in the soil-rock mixture samples, and determine whether the composition of the complete rock mass samples and the stone samples is the same; Perform strength tests on the complete rock mass samples and the soil samples respectively to obtain micromechanics parameters; According to the modeling parameters, perform accurate modeling from macro scale and sub-particle scale respectively to obtain a rock coupling model; Take the micromechanics parameters as input, perform strength test simulation on the rock coupling model to obtain soil-rock mixture mechanical parameters.
[0008] Further, the scanning of the soil-rock mixture samples to construct a mixture three-dimensional model and the segmentation of the soil and stone based on the mixture three-dimensional model to acquire simulation analysis parameters are specifically as follows: Acquire three-dimensional point cloud data of the soil-rock mixture samples and perform preprocessing to obtain preprocessed point cloud data, and construct a mixture three-dimensional model based on the preprocessed point cloud data; Based on the preprocessed point cloud data, utilize a pre-trained image segmentation deep learning model to segment the stone and soil in the mixture three-dimensional model to obtain a stone three-dimensional model and a soil three-dimensional model; Based on a pre-trained general neural network model, extract simulation analysis parameters of the stone three-dimensional model.
[0009] Further, the indoor screening tests on the soil-rock mixture samples to obtain stone samples and soil samples are specifically as follows: The soil-rock mixture sample is dry screened or wet screened according to the selected standard screen group; The separated particles are divided into two categories: The stone sample: the block stone retained on the 20 mm screen; The soil sample: the crushed stone and fine-grained soil passing through the 20 mm screen.
[0010] Further, the composition of the stone sample and the intact rock sample in the soil-rock mixture sample is analyzed to determine whether the composition of the intact rock sample and the stone sample is the same, specifically: The intact rock sample and the stone sample are cleaned to obtain the cleaned intact rock sample and the cleaned stone sample; Based on the dimensions of mineral composition, element content, microstructure, and physical properties, consistency analysis is performed on the composition of the cleaned intact rock sample and the cleaned stone sample. When all four dimensions meet the set threshold, it is determined that the composition of the cleaned intact rock sample and the cleaned stone sample is consistent, otherwise, the composition is inconsistent.
[0011] Further, when all four dimensions meet the set threshold, it is determined that the composition of the cleaned intact rock sample and the cleaned stone sample is consistent, specifically: If the difference in the main mineral species and content between the intact rock sample and the soil-rock mixture sample is less than or equal to 5%, the mineral composition of the two is consistent; If the oxide mass element content deviation between the intact rock sample and the soil-rock mixture sample is less than or equal to 3%, the element content of the two is consistent; If the structural similarity between the intact rock sample and the soil-rock mixture sample is greater than the set threshold, the microstructure of the two is consistent; If the density difference between the intact rock sample and the soil-rock mixture sample is less than or equal to 2% and the porosity difference is less than or equal to 5%, the physical properties of the two are consistent.
[0012] Further, the intact rock sample and the soil sample are respectively subjected to strength testing to obtain mesoscopic mechanical parameters, specifically: The intact rock sample is processed into a standard cylindrical test piece for uniaxial compression testing to determine the compressive strength, elastic modulus, and Poisson's ratio; The soil sample is prepared into saturated or unsaturated soil samples, different normal stresses are applied, and the cohesion and internal friction angle are measured to complete the soil strength testing; The compressive strength, elastic modulus, Poisson's ratio, cohesion, and internal friction angle are used as mesoscopic mechanical parameters.
[0013] Further, according to the modeling parameters, accurate modeling is performed from the macro scale and the sub-particle scale to obtain a rock coupling model, specifically: From the macro scale, the skeleton particle model is constructed based on the obtained rock skeleton particles; From the sub-particle scale, different modeling is carried out on the cemented medium according to the cementing type, and a cementing model is obtained. The skeleton particle model and the cementing model are coupled to obtain a rock coupling model.
[0014] Further, the skeleton particle model is constructed based on the obtained rock skeleton particles from the macro scale, and specifically, The obtained rock skeleton particles are decomposed into a sub-particle assembly, the microcracks, pores and mineral grain structure inside the rock are simulated, the sub-particle arrangement is optimized through the energy minimization principle, and the optimized sub-particle arrangement is obtained. The optimized sub-particle arrangement is embedded into a grain boundary model, a weakened contact surface is arranged between the sub-particles, the intergranular fracture behavior of the actual rock is simulated, and the skeleton particle model is obtained.
[0015] Further, the cementing model is obtained by different modeling of the cemented medium according to the cementing type from the sub-particle scale, and specifically, For physical cementing, a deformable cementing layer is added between soil particles to generate a physical cementing model. For chemical cementing, a rigid cementing shell is generated on the surface of the skeleton particle to generate a chemical cementing model.
[0016] Further, the soil-rock mixture mechanical parameters are obtained by strength test simulation of the rock coupling model with the mesoscopic mechanical parameters as input, and specifically, Triaxial test loading is simulated, a servo control mechanism is adopted, a constant confining pressure is applied to the model by a rigid wall, different burial depth conditions are simulated, an axial displacement is applied to the rock coupling model in a strain control mode, and the axial stress, volume strain and particle displacement field are recorded. If cementing fracture occurs, the system automatically switches to sliding contact, the soil particle grading curve is generated according to the sieve test data, and the particle size distribution is controlled. In the confining pressure balance stage, the servo control confining pressure is balanced to the target value; in the axial loading stage, the displacement is applied to the rock coupling model until the sample is damaged; in the post-failure stage, the displacement is maintained, and the residual strength, stress-strain curve and failure mode are recorded. According to the stress-strain curve, the peak strength, residual strength, elastic modulus and Poisson's ratio are extracted.
[0017] Compared with the prior art, the beneficial effects of the present application are: The present application improves the accuracy of obtaining the mechanical parameters of the soil-rock mixture by improving the specific implementation process in the process of obtaining the mechanical parameters of the soil-rock mixture, testing the actual field samples, and comparing and analyzing different rock bodies. BRIEF DESCRIPTION OF DRAWINGS
[0018] The accompanying drawings, which form a part of this specification, are included to provide a further understanding of the application and are incorporated in and constitute a part of this specification. The embodiments of the application, together with its
[0019] Figure 1 A flow chart of a calculation process of a method for obtaining mechanical parameters of a soil-rock mixture based on experiments and simulations in an embodiment of the application; Figure 2 A schematic diagram of an analysis principle of a three-dimensional convolutional neural network in an embodiment of the application; Figure 3 A schematic diagram of sub-particle scale three-dimensional modeling in an embodiment of the application; Figure 4 A schematic diagram of a discrete element particle flow numerical simulation software simulating a triaxial test in an embodiment of the application; Figure 5 A schematic diagram of a stress-strain curve of a triaxial test in an embodiment of the application. DETAILED DESCRIPTION
[0020] The application will be further described below with reference to the drawings and embodiments.
[0021] It should be noted that the following detailed description is illustrative only, and is intended to provide further description in order to provide a further understanding of the application. Unless otherwise defined, all technical and scientific terms used herein have the same meaning as commonly understood by one of ordinary skill in the art to which this application belongs.
[0022] It should be noted that the terms used herein are only intended to describe specific embodiments and are not intended to limit exemplary embodiments according to the present application. As used herein, the singular form is intended to include the plural form unless the context clearly indicates otherwise, and it should also be understood that the terms "comprise" and / or "include" as used herein indicate the presence of a feature, step, operation, device, component and / or a combination thereof.
[0023] The embodiments in the present application and the features in the embodiments can be combined with each other without conflict.
[0024] Embodiment One The present embodiment provides a method for obtaining parameters of loose broken rock mass based on sub-particle scale modeling and simulation. In the present embodiment, the method comprises the following steps: Obtain soil-rock mixture samples and intact rock mass samples in a target area, and obtain modeling parameters based on the soil-rock mixture samples; Scan the soil-rock mixture samples to construct a three-dimensional model of the mixture, and segment the soil and rock based on the three-dimensional model of the mixture to obtain simulation analysis parameters; The soil-rock mixture sample is subjected to indoor screening test to obtain the stone sample and the soil sample; The composition of the stone sample in the soil-rock mixture sample and the intact rock sample is analyzed to determine whether the composition of the intact rock sample is the same as that of the stone sample; The intact rock sample and the soil sample are respectively subjected to strength test to obtain the meso-mechanical parameters; According to the modeling parameters, accurate modeling is respectively performed from the macro scale and the sub-particle scale to obtain the rock coupling model; The rock coupling model is subjected to strength test simulation with the meso-mechanical parameters as input to obtain the soil-rock mixture mechanical parameters.
[0025] As shown in Figure 1 The method of the embodiment specifically includes: Step 1, field sampling; obtaining the soil-rock mixture sample and the intact rock sample of the target area.
[0026] The requirement of sampling is to ensure that the soil-rock mixture sample contains typical stone blocks and soil distribution. The sampling depth, position and geological conditions are recorded, and the sample is packaged to maintain the original water content and structure.
[0027] The purpose of recording the depth, position and geological conditions is to provide the corresponding simulation environment in the subsequent discrete element particle flow numerical simulation software. The depth affects the in-situ ground stress of the sample, and the geological conditions affect the lateral stress of the sample. The subsequent results under the corresponding conditions of numerical simulation are compared, and the state of the soil-rock mixture at the fault can be roughly obtained to assist in judging how much the fault negatively affects the safety of the site.
[0028] Step 2, sample scanning and data extraction; the soil-rock mixture sample is scanned to construct a mixture three-dimensional model, and the soil and stone are segmented based on the mixture three-dimensional model to obtain simulation analysis parameters.
[0029] The three-dimensional point cloud data of the soil-rock mixture sample is obtained, and the stone and soil of the soil-rock mixture sample are separately modeled to obtain a stone three-dimensional model and a soil three-dimensional model.
[0030] Step 2.1, 3D scanning and intelligent processing.
[0031] The three-dimensional point cloud data of the soil-rock mixture sample is obtained and preprocessed to obtain preprocessed point cloud data, and a mixture three-dimensional model is constructed based on the preprocessed point cloud data; In the pre-processing, the occluded areas (such as stone block gaps) in the three-dimensional point cloud data of the soil-rock mixture sample are scanned from multiple angles to complete the gaps. Based on the pre-processed point cloud data, the missing parts of the pre-processed point cloud data are repaired using but not limited to a generative adversarial network (GAN) (such as Pix2Pix) to generate a complete mixed three-dimensional model.
[0032] Step 2.2, automatic segmentation of stone and soil.
[0033] Based on the pre-processed point cloud data, the stone and soil in the mixed three-dimensional model are segmented using a pre-trained image segmentation deep learning model to obtain a stone three-dimensional model and a soil three-dimensional model. The image segmentation deep learning model can be implemented using but not limited to PointNet++ or RandLA-Net deep learning model for automatic segmentation: Input: pre-processed point cloud data (including coordinates, RGB color, reflectance intensity).
[0034] Output: classification label (stone, soil, pore).
[0035] The boundary fuzzy area (such as the soil-rock contact surface) of the segmented result is post-processed using conditional random field (CRF) to reduce misclassification, and a stone three-dimensional model and a soil three-dimensional model are obtained.
[0036] Step 2.3, intelligent extraction of geometric parameters.
[0037] Based on a pre-trained general neural network model, the simulation analysis parameters in the stone three-dimensional model are extracted; The general neural network model uses but not limited to a three-dimensional convolutional neural network (3D-CNN) to analyze the stone block morphology in the stone three-dimensional model and extract model analysis parameters such as Figure 2 as shown in the figure; The simulation analysis parameters include key parameters such as stone block volume, aspect ratio, surface roughness, and spatial orientation angle, as well as stone block volume proportion (VBP), average spacing (ASD), and coordination number (number of contact points).
[0038] The output of the general neural network model is structured data (JSON format) of the simulation analysis parameters, which is used as input for the particle flow analysis software to perform simulation analysis and obtain the mechanical parameters of the soil-rock mixture sample.
[0039] Step 3, separate the stone and soil in the soil-rock mixture sample. Perform indoor sieve test on the soil-rock mixture sample to obtain stone sample and soil sample.
[0040] According to the "Soil Test Procedures", select a standard sieve set (such as aperture 2mm, 5mm, 20mm, etc.) to perform dry or wet sieving on the soil-rock mixture sample. The separated particles are divided into two categories: Rock sample components: blocks retained on 20 mm sieve, used for comparative analysis of composition.
[0041] Soil sample components: crushed stone and fine-grained soil (clay, sand, etc.) passing through a 20 mm sieve. Prepare samples according to natural moisture content or remolded state for direct shear testing.
[0042] Weigh and record the mass percentage of each component, take photos for archiving.
[0043] Step 4, analyze the composition of rock samples and intact rock samples in soil-rock mixture samples. Test whether the composition of intact rock samples and rock samples is the same.
[0044] Step 4.1, sample cleaning and numbering.
[0045] Clean the intact rock samples and rock samples to obtain cleaned intact rock samples and rock samples.
[0046] Specifically, use an ultrasonic cleaner to remove soil attached to the surface of the samples, rinse with distilled water and dry (105°C constant temperature oven, 24 hours). Label the sample source (intact rock / mixture), collection location, size and mass.
[0047] Step 4.2, composition testing and comparison.
[0048] Based on the dimensions of mineral composition, element content, microstructure and physical properties, the composition of cleaned intact rock samples and rock samples is analyzed for consistency. When all four dimensions meet the set threshold, the composition of cleaned intact rock samples and rock samples is consistent, otherwise, the composition is inconsistent.
[0049] Verify rock composition consistency through multi-dimensional testing. As shown in Table 1 below.
[0050] Table 1 Rock sample composition testing
[0051] Step 5, determination of geotechnical micromechanics parameters. Test the strength of intact rock samples and soil samples separately to obtain micromechanics parameters.
[0052] Step 5.1, rock strength testing.
[0053] Process intact rock samples into standard cylindrical specimens (diameter 50 mm, height 100 mm) for uniaxial compression testing to determine compressive strength (f σ c ), elastic modulus (E) and Poisson's ratio (v), complete rock strength testing.
[0054] Step 5.2, soil strength testing.
[0055] Direct shear test: the soil sample is prepared as saturated or unsaturated soil sample, different normal stress is applied, and cohesion (c) and internal friction angle (φ) are measured to complete the soil strength test. c φ
[0056] Based on the compressive strength (fc), elastic modulus (E), Poisson's ratio (v), cohesion (c) and internal friction angle (φ), as the meso-mechanical parameters. σ c ν c φ
[0057] According to the modeling parameters obtained by the 3D scanner, the sample is accurately modeled in the discrete element particle flow numerical simulation software; Step 6, discrete element particle flow numerical simulation software numerical simulation modeling (macro-scale and sub-particle scale coupled model).
[0058] Step 6.1, skeleton particle modeling.
[0059] The rock skeleton particle (Clump) obtained by 3D scanning is further decomposed into sub-particle assembly to simulate the internal micro-crack, pore and mineral grain structure of the rock. Voronoi partitioning algorithm is used to generate sub-particles (diameter 0.1-1mm), and the arrangement of sub-particles is optimized by energy minimization principle to obtain the optimized sub-particle arrangement.
[0060] The optimized sub-particle arrangement is embedded in the grain boundary model, and a weakened contact surface is set between the sub-particles to simulate the intergranular fracture behavior of the actual rock, and the skeleton particle model is obtained. Based on the rock meso-test data (such as nanoindentation test, micro-CT scanning), the stiffness (E) and friction coefficient (μ) of the sub-particles are calibrated. k n , k s μ .
[0061] Step 6.2, cemented medium modeling.
[0062] According to the cementing type, different modeling is carried out for the cemented medium.
[0063] For physical cementation (such as clay wrapping), a deformable cement layer (thickness 0.01-0.1mm) is added between the soil particles (Ball), and Hertz-Mindlin with Cohesion contact model is used to obtain the physical cementation model; For chemical cementation (such as carbonate cementation), a rigid cement shell is generated on the surface of the skeleton particle, and ParallelBond model is used to obtain the chemical cementation model.
[0064] Based on the pore structure in the scanning data, the natural deposition process of the cemented medium in the pores is simulated using a diffusion-limited aggregation (DLA) algorithm. The cementation strength is predicted by deep learning: a convolutional neural network (CNN) is trained to output the bonding strength (Gc) a bond ) and stiffness (G) k bond ) of the cemented layer according to the gray value distribution of the CT scan image.
[0065] Step 6.3, multi-scale coupled modeling, as shown in Figure 3 .
[0066] The macro-scale (Clump / Ball assembly) and sub-particle scale (grain / cement) are coupled through a sub-model nesting technique, and the cross-scale force transmission boundary conditions are set to obtain the rock coupled model. A dynamic link library (DLL) is written using the Fish language built-in discrete element particle flow numerical simulation software to realize multi-thread parallel computing.
[0067] The meso-mechanical parameters of the rock and soil obtained by the test are used as input to simulate the strength test of the model in the discrete element particle flow numerical simulation software, and the mechanical parameters of the soil-rock mixture are obtained.
[0068] Step 7, simulation analysis and parameter extraction.
[0069] Step 7.1, set the model loading conditions.
[0070] Simulate triaxial test loading, adopt servo control mechanism, apply constant confining pressure (such as 0.1 MPa, 1 MPa, 5 MPa) to the model through rigid wall, simulate different burial depth conditions. Apply axial displacement (rate 0.05 mm / s) to the rock coupled model in the strain control mode, record axial stress (σ σ n ), volume strain and particle displacement field. The process of simulating triaxial test by discrete element particle flow numerical simulation software is shown in Figure 4 .
[0071] Step 7.2, multi-scale model parameter response adjustment.
[0072] When the local tensile stress > bonding strength (Gc) a bond ) or shear stress > shear strength (Gs) a bond + μ σ n , the local cementation is fractured.
[0073] The contact between the surface formed after the cementation fracture in the rock coupling model and other parts of the model is automatically switched to sliding contact (friction coefficient μ =0.5) and the soil particle size distribution curve (such as the CU curve) generated according to the sieve test data controls the particle size distribution.
[0074] Step 7.3, simulation test process and data acquisition.
[0075] In the confining pressure balancing stage, the confining pressure is servo-controlled to the target value, and the force fluctuation is allowed to be less than 2%, and recorded once every 10 steps; in the axial loading stage, the displacement is applied to the rock coupling model until the sample is destroyed (the axial strain reaches 15%), and recorded once every 100 steps; in the post-failure stage, the displacement is maintained, and the residual strength is recorded, and recorded once every 500 steps. The stress-strain curve and the failure mode are recorded, and the stress-strain curve of the triaxial test is shown in Figure 5 .
[0076] Step 7.4, strength and deformation parameter extraction According to the stress-strain curve, the peak strength ( σ max ), residual strength is extracted; the elastic modulus is the tangent modulus at 50% peak stress, and the Poisson's ratio is the ratio of lateral strain to axial strain in the elastic stage; That is, the final obtained soil-rock mixture mechanical parameters are peak strength, residual strength, elastic modulus and Poisson's ratio.
[0077] A person of ordinary skill in the art can understand that all or part of the processes in the above-mentioned embodiment methods can be completed by a computer program instructing related hardware, and the program can be stored in a computer readable storage medium. When the program is executed, it can include the processes of the above-mentioned embodiment methods. The storage medium can be a magnetic disc, an optical disc, a read-only memory (ROM) or a random access memory (RAM), etc.
[0078] The above describes the specific embodiments of the present application in combination with the accompanying drawings, but is not a limitation on the protection scope of the present application. Those skilled in the art should understand that various modifications or variations made by those skilled in the art on the basis of the technical solutions of the present application without creative labor are still within the protection scope of the present application.
Claims
1. A method for obtaining parameters of loose and fractured rock masses based on sub-grain-scale modeling and simulation, characterized in that, include: Obtain soil-rock mixture samples and intact rock mass samples from the target area, and obtain modeling parameters based on the soil-rock mixture samples; A three-dimensional model of the soil-rock mixture was constructed by scanning the soil-rock mixture sample. Based on the three-dimensional model of the mixture, the soil and rock were segmented to obtain simulation analysis parameters. Indoor sieve tests were conducted on soil-rock mixture samples to obtain rock samples and soil samples. Analyze the composition of rock samples and intact rock samples in soil-rock mixture samples to determine whether the composition of intact rock samples and rock samples is the same; Strength tests were performed on intact rock samples and soil samples to obtain micromechanical parameters. Based on the modeling parameters, precise modeling is performed at both the macroscopic and sub-grain scales to obtain the rock coupling model. Using micromechanical parameters as input, a strength test simulation of a rock coupling model was performed to obtain the mechanical parameters of the soil-rock mixture.
2. The method for obtaining parameters of loose and fractured rock mass based on sub-grain scale modeling and simulation as described in claim 1, characterized in that, The process involves scanning the soil-rock mixture sample to construct a three-dimensional model of the mixture, and then segmenting the soil and rock masses based on this model to obtain simulation analysis parameters. Specifically: The three-dimensional point cloud data of the soil-rock mixture sample is acquired and preprocessed to obtain preprocessed point cloud data. A three-dimensional model of the mixture is then constructed based on the preprocessed point cloud data. Based on preprocessed point cloud data, the rock and soil in the hybrid 3D model are segmented using a pre-trained image segmentation deep learning model to obtain the 3D rock model and the 3D soil model. Based on a pre-trained general neural network model, simulation analysis parameters are extracted from the three-dimensional model of the rock body.
3. The method for obtaining parameters of loose and fractured rock mass based on sub-grain scale modeling and simulation as described in claim 1, characterized in that, The indoor sieve test was conducted on the soil-rock mixture sample to obtain rock samples and soil samples, specifically as follows: Select a standard sieve set and perform dry or wet sieve on the soil-rock mixture sample; The separated particles were divided into two categories: Rock samples: boulders retained on a sieve of 20mm or larger; Soil sample: gravel and fine soil passing through a 20mm sieve.
4. The method for obtaining parameters of loose and fractured rock mass based on sub-grain scale modeling and simulation as described in claim 1, characterized in that, The analysis of the composition of rock samples and intact rock samples in the soil-rock mixture sample, and the determination of whether the composition of the intact rock sample and the rock sample are the same, specifically involves: The intact rock sample and the stone body sample were cleaned to obtain the cleaned intact rock sample and the stone body sample. Based on the dimensions of mineral composition, elemental content, microstructure, and physical properties, a consistency analysis is performed on the composition of the cleaned intact rock sample and the stone body sample. When all four dimensions meet the set thresholds, the composition of the cleaned intact rock sample and the stone body sample is considered to be consistent; otherwise, the composition is inconsistent.
5. The method for obtaining parameters of loose and fractured rock mass based on sub-grain scale modeling and simulation as described in claim 4, characterized in that, When all four dimensions meet the set thresholds, the composition of the cleaned intact rock sample and the stone body sample is determined to be consistent, specifically: If the difference in the types and contents of the main minerals in the intact rock sample and the soil-rock mixture sample is less than or equal to 5%, then the mineral composition of the two samples is consistent. If the difference in oxide mass element content between a complete rock sample and a soil-rock mixture sample is less than or equal to 3%, the element content of the two samples is considered to be the same. If the structural similarity between a complete rock sample and a soil-rock mixture sample is greater than a set threshold, then their microstructures are identical. If the density difference between a complete rock sample and a soil-rock mixture sample is less than or equal to 2% and the porosity difference is less than or equal to 5%, then the two samples have the same physical properties.
6. The method for obtaining parameters of loose and fractured rock mass based on sub-grain scale modeling and simulation as described in claim 1, characterized in that, Strength tests were performed on intact rock and soil samples to obtain micromechanical parameters, specifically: Complete rock samples were processed into standard cylindrical specimens for uniaxial compression tests to determine compressive strength, elastic modulus, and Poisson's ratio. Prepare saturated or unsaturated soil samples, apply different normal stresses, and measure cohesion and internal friction angle to complete the soil strength test. Compressive strength, elastic modulus, Poisson's ratio, cohesion, and internal friction angle are used as micromechanical parameters.
7. The method for obtaining parameters of loose and fractured rock mass based on sub-grain scale modeling and simulation as described in claim 1, characterized in that, Based on the modeling parameters, precise modeling is performed at both the macroscopic and sub-grained scales to obtain the coupled rock model, as follows: From a macroscopic scale, a skeletal particle model is constructed based on the obtained rock skeletal particles; At the sub-particle scale, different models are used to model the cementing medium according to the cementing type, resulting in cementing models; The skeleton particle model and the cementation model are coupled to obtain the rock coupling model.
8. The method for obtaining parameters of loose and fractured rock mass based on sub-grain scale modeling and simulation as described in claim 7, characterized in that, The construction of a skeletal particle model based on the acquired rock skeletal particles at a macroscopic scale specifically involves: The obtained rock skeleton particles are decomposed into sub-particle aggregates to simulate the micro-fractures, pores and mineral grain structures inside the rock. The sub-particle arrangement is optimized by minimizing energy to obtain the optimized sub-particle arrangement. The optimized subgrain arrangement is embedded in the grain boundary model, and a weakened contact surface is set between the subgrains to simulate the intergranular fracture behavior of actual rocks, thus obtaining the skeleton grain model.
9. The method for obtaining parameters of loose and fractured rock mass based on sub-grain scale modeling and simulation as described in claim 7, characterized in that, The cementing model is obtained by modeling the cementing medium differently according to the cementing type at the sub-particle scale, as follows: For physical cementation, a deformable cementation layer is added between soil particles to generate a physical cementation model; For chemical bonding, a rigid cemented shell is generated on the surface of the skeletal particles, thus creating a chemical bonding model.
10. The method for obtaining parameters of loose and fractured rock mass based on sub-grain scale modeling and simulation as described in claim 1, characterized in that, The method of using micromechanical parameters as input to perform strength test simulations on a coupled rock model to obtain the mechanical parameters of the soil-rock mixture is as follows: Triaxial loading was simulated using a servo control mechanism. A constant confining pressure was applied to the model through a rigid wall to simulate different burial depths. Axial displacement was applied to the rock coupling model using a strain control method, and axial stress, volumetric strain, and particle displacement field were recorded. If cementation fracture occurs, it automatically switches to sliding contact and controls the particle size distribution based on the soil particle size distribution curve generated from the sieve test data. During the confining pressure equilibrium stage, the confining pressure is servo-controlled to the target value; during the axial loading stage, displacement is applied to the rock coupling model until the specimen fails; during the post-failure stage, the displacement is maintained, and the residual strength, stress-strain curves, and failure modes are recorded. Based on the stress-strain curve, peak strength, residual strength, elastic modulus, and Poisson's ratio are extracted.