Discrete Element Refined Modeling and Calibration Method for In-situ Fractured Rock Mass

Abstract

This invention relates to a method for refined discrete element modeling and calibration of undisturbed fractured rock masses. The method involves acquiring three-dimensional microstructural data of rock samples via CT scanning to generate a three-dimensional digital model of the rock sample; conducting sieving tests on the broken particles of the same rock sample; establishing a database of realistic rock block morphology and generating particle cluster templates for each representative rock block; and establishing a refined discrete element numerical model based on the identification results of joints and fractures within the rock sample, the full gradation distribution curve from the sieving test, the database of realistic rock block morphology, and the particle cluster templates. Shear stress-shear displacement curves are obtained through numerical experiments and compared point-by-point with the measured curves from field direct shear tests. Model parameters are adjusted until the curves show a basic match, completing parameter calibration. This invention, based on CT scanning and three-dimensional reconstruction technology, realistically restores the irregular morphology of rock blocks, the directional distribution of fractures, and the multi-component structure in fractured rock masses, avoiding distortion caused by random particle modeling and significantly improving the model's accuracy in representing the microstructural characteristics of undisturbed rock masses.

Description

Technical Field

[0001] This invention relates to the field of engineering exploration technology, specifically to a method for fine-grained discrete element modeling and calibration of undisturbed fractured rock masses. Background Technology

[0002] Fractured rock masses are typical tectonic rocks, widely distributed in orogenic belts, fault fracture zones, and strongly weathered crusts. Under intense tectonic stress, the original rock is broken into porphyritic and granular structures, often accompanied by secondary mud inclusions, carbonatization, or ferruginous dissemination, exhibiting massive, banded, or network structures. Especially under conditions of high mud inclusions, fractured rock masses are generally soft, poorly diagenetic, and have extremely low cohesion, even approaching zero. The internal friction angle is controlled by the interlocking and alignment of clastic particles, exhibiting significant discontinuities, heterogeneity, and anisotropy. The mechanical behavior of these rock masses is mainly controlled by the morphology, size, spatial arrangement, contact relationships, and muddy or calcareous filling state of the internal clastic particles. In tunnel engineering, slope engineering, and dam foundation engineering, fractured rock masses are often key geological units controlling overall stability. Their deformation and failure mechanisms are complex, requiring appropriate rock mechanics testing methods to reveal their inherent laws, reflect the true microstructure, and provide accurate geological information for engineering design to ensure the safety of construction and operation.

[0003] Currently, the analysis of the mechanical properties of fractured rock masses mainly relies on two methods: indoor remodeling tests and in-situ tests. Indoor remodeling tests involve remodeling rock fragments collected in the field according to a certain gradation before conducting triaxial or direct shear tests. This method is simple to operate and highly repeatable; however, the remodeling process destroys the original internal structure of the rock mass, failing to preserve the in-situ spatial distribution and contact state of fragments, grains, and inclusions, especially losing the interlocking and interlocking characteristics between large rock blocks. This often leads to an underestimation of peak strength and an overestimation of shear dilatation effects. In-situ tests (such as in-situ large direct shear tests and bearing plate tests) can preserve the original structure of the rock mass, and the results are relatively reliable. However, they are costly, time-consuming, heavily limited by site conditions, and cannot perform parameter sensitivity analysis or reproduce the failure process, making it difficult to meet the multi-condition simulation requirements in engineering design.

[0004] In numerical simulation, discrete element methods (such as PFC and 3DEC) are widely used for analyzing the mechanical behavior of discontinuous media because they can simulate the slippage, opening, and rotation between particles. However, existing discrete element modeling methods have significant shortcomings when dealing with fractured rock masses: most studies use random particle stacking to generate samples, simplifying the rock block morphology to disks, spheres, or simple polyhedra, ignoring the irregular morphology of real clastic particles and their contribution to interlocking friction; joints and fractures are treated as randomly distributed contact surfaces or pre-set cracks, failing to reflect the actual fracture orientation, density, and topological relationship with the rock block; the strength difference between cement and rock blocks is rarely distinguished, often using uniform contact model parameters, resulting in simulation results that cannot accurately reproduce the stress-strain characteristics of fractured rock masses, which are characterized by "interlocking followed by shearing". Furthermore, due to the highly nonlinear relationship between the model's microscopic parameters (such as particle stiffness, friction coefficient, and bond strength) and the macroscopic mechanical response, the parameter calibration in existing methods often relies on empirical trial and error or simple comparison with uniaxial tests. There is a lack of a systematic calibration process that compares the results of large-scale direct shear tests in the field point by point, resulting in low reliability of the calculated parameters and a large deviation between the simulation results and the actual situation.

[0005] Therefore, it is necessary to propose new measures to overcome the above-mentioned shortcomings. Summary of the Invention

[0006] The purpose of this invention is to provide a method for fine-grained discrete element modeling and calibration of undisturbed fractured rock masses, in order to solve the problem that existing analysis methods cannot accurately reproduce the mechanical characteristics of fractured rock masses.

[0007] To achieve the above objectives, the technical solution adopted by the present invention is as follows:

[0008] A method for refined discrete element modeling and calibration of undisturbed fractured rock masses is provided, the method comprising:

[0009] Three-dimensional microstructural data of rock samples are obtained by CT scanning, and a three-dimensional digital model of the rock samples is generated. Joints and fissures inside the rock samples are identified based on slice images from different directions.

[0010] Sieve analysis was conducted on the crushed slag particles of the same rock sample, and the full gradation distribution curve was obtained through the sieve analysis data;

[0011] Representative rock blocks were selected from the three-dimensional digital model of the rock samples and their surface meshes were extracted to establish a database of real rock block morphology and generate a grain cluster template for each representative rock block.

[0012] Based on the identification results of joints and fissures inside rock samples, the full gradation distribution curve of sieve tests, the real rock block morphology database, and the particle cluster template, a discrete element refined numerical model consistent with the size of the direct shear sample in the field was established from four aspects: rock block size distribution, rock block morphology distribution, rock block spatial distribution, joint and fissure morphology and contact state.

[0013] Numerical experiments were conducted based on the refined discrete element numerical model to obtain the shear stress-shear displacement curve. The curves were compared point by point with the measured curves from the field direct shear test. The model parameters were adjusted until the two curves basically matched in terms of peak strength, residual strength, and softened section morphology, thus completing the parameter calibration.

[0014] Furthermore, a three-dimensional digital model of the rock sample is generated, including:

[0015] Based on the three-dimensional microstructure data of the rock sample, color enhancement, binarization segmentation, noise reduction filtering, and three-dimensional surface reconstruction are performed sequentially to generate a three-dimensional digital model of the rock sample.

[0016] Furthermore, in the sieve analysis, the rock mass components in the rock sample were divided into three categories according to the particle size distribution: large rock blocks with a diameter of 3 to 20 cm, small rock blocks with a diameter of 0.5 to 3 cm, and cement with a diameter of no more than 0.5 cm.

[0017] Furthermore, a grain cluster template is generated for each representative rock block, including:

[0018] For each representative rock block, a corresponding rigid cluster template is generated using the particle cluster method. The outer envelope of the rigid cluster template is consistent with the surface mesh geometry of the real rock block.

[0019] The rigid cluster template was converted into a deformable flexible cluster template to simulate the deformation and local fracturing of rock blocks under stress.

[0020] Furthermore, a refined discrete element numerical model is established based on the rock block size distribution, including:

[0021] The total gradation distribution curve obtained by sieve analysis is used to quantitatively control the quantity or mass percentage of large rock blocks, small rock blocks, and cementing material in the model.

[0022] Furthermore, a refined discrete element numerical model is established based on the morphological distribution of rock blocks. This involves using a particle cluster template to generate irregularly shaped rock blocks, where the particle cluster template is a deformable flexible cluster template.

[0023] Furthermore, a refined discrete element numerical model is established based on the spatial distribution of rock blocks, including:

[0024] Based on the actual spatial location and attitude angle of the large rock blocks in the CT images, position mapping or orientation constraint methods are used to make the long axis of the large rock blocks oriented in a directional correlation along the dominant direction of the on-site geostress, simulating the interlocking and mosaic characteristics between the rock blocks.

[0025] Furthermore, a refined discrete element numerical model is established based on the joint and fracture morphology and contact state, including:

[0026] For the contact between large rock blocks, a smooth joint model is used to simulate macroscopic cracks. The number of cracks is quantitatively characterized by the number of smooth joint contacts, and the crack size is quantitatively characterized by the characteristic size of the large rock block.

[0027] For the relationships between large and small rock blocks, between small rock blocks, between rock blocks and cement, and between cement particles, a parallel bonding model is used, and the bonding performance is quantitatively controlled by adjusting the bonding strength parameters.

[0028] Furthermore, the refined discrete element numerical model includes upper and lower shear boxes, a normal loading plate, and a shear seam.

[0029] Furthermore, numerical experiments were conducted based on the refined discrete element numerical model to obtain shear stress-shear displacement curves, including:

[0030] In the numerical experiment, the normal load was kept constant by a servo control mechanism, and the horizontal shear displacement was applied at a constant rate. The calculated shear stress-shear displacement curve was recorded in real time.

[0031] Compared with the prior art, the beneficial effects of the present invention are as follows:

[0032] This invention provides a refined discrete element modeling method for undisturbed fractured rock masses based on CT scan data. Based on CT scanning and 3D reconstruction technology, it realistically restores the irregular morphology of rock blocks, the directional distribution of fractures, and the multi-component structure in fractured rock masses. It avoids the distortion problem of traditional random particle modeling and significantly improves the accuracy of the discrete element model in representing the microscopic features of undisturbed rock masses. It is suitable for analyzing the mechanical behavior of fractured rock masses with almost no cohesion under tectonic compression. It can realistically reflect their discontinuous and non-uniform characteristics and significantly improve the accuracy and reliability of discrete element simulation. Attached Figure Description

[0033] 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 only some embodiments of the present invention. For those skilled in the art, other embodiments can be obtained from these drawings without creative effort.

[0034] Figure 1 This is a flowchart of the method provided in an embodiment of the present invention.

[0035] Figure 2 This is a schematic diagram of rock sample CT scanning and three-dimensional reconstruction provided in an embodiment of the present invention.

[0036] Figure 3 This is a schematic diagram of the rock sample slag particle screening test provided in an embodiment of the present invention.

[0037] Figure 4 This is a schematic diagram of the rock sample discrete element fine modeling process provided in the embodiment of the present invention.

[0038] Figure 5 This is a numerical model and calibration diagram of the direct shear test for rock samples provided in an embodiment of the present invention. Detailed Implementation

[0039] To facilitate understanding of the present invention, a more complete description will be given below with reference to the accompanying drawings. Preferred embodiments of the invention are shown in the drawings. However, the invention can be implemented in many different forms and is not limited to the embodiments described herein. Rather, these embodiments are provided to provide a thorough and complete understanding of the disclosure of the invention.

[0040] It should be noted that although the order of steps is mentioned in the method description, in some cases, they may be performed in a different order than that described here, and this should not be interpreted as a restriction on the order of steps.

[0041] Existing methods for analyzing the mechanical properties of rock masses are mostly focused on intact rocks or simple fractured rock masses. For multi-component fractured rock masses composed of rock blocks, infill materials, and cement, accurate analytical methods are lacking. CT scanning and 3D reconstruction techniques can non-destructively obtain the internal microstructure of rock samples and are expected to have advantages in the analysis of multi-component fractured rock masses. Therefore, this invention proposes a novel discrete element method for refined modeling and calibration of undisturbed fractured rock masses. Based on CT scanning and 3D reconstruction technology, it realistically reconstructs the irregular morphology of rock blocks, the directional distribution of fractures, and the multi-component structure within the fractured rock mass. As a complete technical solution from CT data to discrete element model and then to parameter calibration, it can effectively solve the problems of modeling distortion and parameter calibration difficulties in existing analytical methods.

[0042] The method of this invention first uses CT scanning and Avizo software to perform three-dimensional reconstruction of rock samples, identify internal fractures, and combine with sieve analysis to determine that the rock mass is composed of a rock block skeleton, filling material, and cement. Representative rock blocks are selected to establish a three-dimensional morphology library and particle cluster template. Then, a refined discrete element numerical model is established by optimizing the rock block distribution, particle morphology, and fracture distribution. A discrete element sample with the same size as the field direct shear sample is established. A servo-held normal load is applied to apply horizontal shear displacement, and the shear stress-shear displacement curve is recorded. Finally, by comparing the curves point by point with the field test curves, the microscopic parameters are repeatedly adjusted using a trial-and-error method until the peak strength, residual strength, and softening section show good agreement.

[0043] Specifically, such as Figure 1 The method includes the following steps:

[0044] S1: CT scan and 3D reconstruction of rock samples:

[0045] Original fractured rock mass refers to a fractured rock mass subjected to strong tectonic compression, with almost no cohesion (cohesion approaches 0 kPa), and exhibiting discontinuous, heterogeneous, and anisotropic characteristics. It includes fault fracture zones, mylonitized zones, and strongly weathered fractured rocks.

[0046] Uncircumferential fractured rock samples are collected on-site, referred to as rock samples. Three-dimensional microstructural data of the rock samples are obtained through CT scanning. Three-dimensional image visualization and analysis software (such as Avizo software) is used to reconstruct the three-dimensional microstructural data of the rock samples in three dimensions. Figure 2 The sample undergoes color enhancement, binarization segmentation, noise reduction filtering, and three-dimensional surface reconstruction to generate a three-dimensional digital model of the rock sample. Based on slice images from different directions, the internal joints and fissures of the rock sample are identified, including their morphology and spatial distribution.

[0047] S2: Slag particle sieving test and component classification:

[0048] Sieve analysis was conducted on crushed slag particles from the same rock sample, and the full gradation distribution curve was obtained from the sieve analysis data. This serves as the quantitative basis for particle size control in the subsequent discrete element model.

[0049] In the sieve analysis, the rock mass components in the rock sample were divided into three categories based on particle size distribution: large rock blocks, small rock blocks, and cement. The diameter d ranges were as follows:

[0050] Large rock blocks: 3 cm < d ≤ 20 cm;

[0051] Small rock fragments: 0.5cm < d ≤ 3cm;

[0052] Cementing material: d≥0.5cm.

[0053] The full gradation distribution curve is used to quantitatively control the number or mass percentage of particles in different particle size ranges (large rock block range, small rock block range, cementation range) in the discrete element refined numerical model, so that the model gradation is consistent with the actual gradation of the undisturbed rock sample.

[0054] S3: Establishment of rock block morphology database and grain cluster template:

[0055] Representative rock blocks are selected from the three-dimensional digital model of rock samples, and surface meshes are extracted to establish a database of real rock block morphology. Particle cluster templates for each representative rock block are generated to characterize the irregular morphology of real rock blocks in discrete element software.

[0056] For each representative rock block, a corresponding rigid cluster template is generated using the particle cluster method in discrete element software (such as Particle Flow Code, PFC). The rigid cluster template is composed of multiple particles connected rigidly, and the outer envelope of the rigid cluster template is consistent with the surface mesh geometry of the real rock block.

[0057] Then, the rigid cluster template is converted into a deformable flexible cluster template to simulate the deformation and local fragmentation of the rock block during the stress process, while maintaining its macroscopic morphological characteristics.

[0058] Figure 3 This diagram illustrates the rock sample slag particle sieving test, including sieving and grading, three-dimensional morphological extraction of rock blocks in different particle size ranges, establishment of particle cluster templates, and statistical analysis of the probability distribution of microscopic parameters. Through sieving and morphological analysis, the gradation characteristics of the rock mass, consisting of a large rock block skeleton, small rock block fillers, and cementing materials, are determined, providing a statistical basis for the random generation of particle clusters.

[0059] S4: Establishment of a refined discrete-element numerical model:

[0060] Based on the identification results of joints and fissures inside the rock samples, the full gradation distribution curve of the sieve test, the real rock block morphology database, and the particle cluster template, a discrete element refined numerical model consistent with the size of the direct shear sample in the field was established from four aspects: rock block size distribution, rock block morphology distribution, rock block spatial distribution, joint and fissure morphology and contact state.

[0061] (1) Regarding the distribution of rock block size:

[0062] The total gradation distribution curve obtained by sieve analysis is used to quantitatively control the quantity or mass percentage of large rock blocks, small rock blocks, and cementing material in the model.

[0063] (2) Distribution of rock block morphology:

[0064] Irregularly shaped rock blocks are generated using particle cluster templates, avoiding the distortion problems associated with traditional disc or spherical particles. The particle cluster template is a deformable, flexible cluster template.

[0065] (3) Spatial distribution of rock blocks:

[0066] Based on the actual spatial location and orientation angle of the large rock blocks in the CT images, position mapping or orientation constraint methods are used to make the long axis of the large rock blocks oriented and correlated along the dominant direction of the field stress, simulating the interlocking and mosaic characteristics between rock blocks, and ensuring that the shape and position of the large rock blocks have the same non-uniformity and anisotropy in spatial distribution as the original rock sample.

[0067] (4) Regarding the morphology and contact state of joints and fractures:

[0068] For the contact between large rock blocks, a smooth joint model is used to simulate macroscopic fractures. This model allows for slippage and opening at the contact surface without tensile strength, and is used to characterize macroscopic joint fractures in the rock mass. The number of fractures is quantitatively characterized by the number of smooth joint contacts, and the fracture size is quantitatively characterized by the characteristic dimensions of the large rock blocks.

[0069] For large rock blocks and small rock blocks, small rock blocks and small rock blocks, and rock blocks and cementing materials, a parallel bonding model is used. By adjusting parameters such as normal stiffness, tangential stiffness, normal strength and tangential strength of the parallel bonding, the bonding performance between rock blocks and cementing materials is quantitatively controlled to reflect the actual bonding state of the fractured rock mass.

[0070] For the contact between cementing particles, a parallel bonding model is also used.

[0071] Figure 4 This is a schematic diagram of the refined discrete element modeling process for rock samples according to the present invention. It shows the process of establishing a rigid cluster model from a particle cluster skeleton, transforming it into a deformable flexible cluster, and assigning a parallel bond or smooth joint (fracture) contact model based on the composition. This process achieves an integrated conversion from the real microstructure to the computational model, ensuring the model's ability to represent discontinuous and non-uniform mechanical behavior.

[0072] S5: Field Direct Shear Test and Parameter Calibration:

[0073] The geometric dimensions of the refined discrete element numerical model are completely consistent with the dimensions of the field direct shear specimen. The refined discrete element numerical model includes the upper and lower shear boxes, the normal loading plate, and the shear seam.

[0074] Numerical experiments were conducted based on a refined discrete element numerical model. A normal load was applied and kept constant through a servo control mechanism, while a horizontal shear displacement was applied at a constant rate. The calculated shear stress-shear displacement curves were recorded in real time. The shear stress-shear displacement curves were compared point by point with the measured curves from the field direct shear test. The model parameters (contact modulus, stiffness ratio, friction coefficient, parallel bond strength, smooth joint normal stiffness, tangential stiffness, friction coefficient, etc.) were repeatedly adjusted using a trial-and-error method until the two curves basically matched in terms of peak strength, residual strength, and softened section morphology, thus completing the parameter calibration.

[0075] The criteria for determining a basic match are:

[0076] The relative error between the calculated curve and the measured curve at the peak strength does not exceed 5%, the relative error at the residual strength does not exceed 8%, and the Pearson correlation coefficient of the softening section (from the peak strength to the residual strength section) is not less than 0.95;

[0077] If the above standards are not met, the following micro-parameters will be modified in the preset adjustment order: contact modulus, stiffness ratio, friction coefficient, parallel bond strength, smooth joint normal stiffness, tangential stiffness, and friction coefficient, until the convergence criterion is met.

[0078] Figure 5This paper presents a numerical model and calibration diagram for the direct shear test of rock samples in this invention. A large-scale discrete element model with the same dimensions as in-situ is constructed to illustrate the application of normal and tangential loads, particle displacement during shearing, and the evolution of the failure zone, outputting shear stress-shear displacement curves. By comparing numerical results with laboratory curves, a trial-and-error method is used to calibrate the micro-parameters, forming a parameter calibration method based on the large shear test.

[0079] The method of the present invention has the following technical advantages:

[0080] 1. The method of the present invention is based on CT scanning and three-dimensional reconstruction technology, which can realistically restore the irregular shape of rock blocks, the directional distribution of fractures and the multi-component structure in fractured rock masses, avoid the distortion problem of traditional random particle modeling, and significantly improve the characterization accuracy of discrete element model for the microscopic features of undisturbed rock masses.

[0081] 2. The method of the present invention systematically optimizes the modeling from four aspects: rock block size distribution, morphological distribution, spatial distribution, and joint and fracture contact state. It establishes a discrete element model of the rock block skeleton, filling material and cementing material working together, which can accurately simulate the discontinuous, non-uniform and anisotropic mechanical behavior of fractured rock mass, especially reproduce the typical stress-strain characteristics of "first interlocking and then shearing".

[0082] 3. The method of this invention combines a large-scale direct shear test numerical model with a point-by-point comparison trial-and-error calibration process, achieving effective matching between calculated parameters and actual physical and mechanical responses, providing reliable constitutive inputs and simulation basis for deformation control in fractured rock mass engineering. This method has clear parameter convergence criteria, avoiding the subjectivity of traditional empirical trial-and-error methods.

[0083] 4. The method of the present invention is universal and can be applied not only to fractured rock masses under tectonic compression, but also to discrete element modeling and parameter calibration of similar discontinuous media such as fault fracture zones, mylonitized zones, strongly weathered fractured rocks and rockfill dam materials.

[0084] The above examples illustrate the present invention only to aid in understanding it and are not intended to limit the scope of the invention. Those skilled in the art can make various simple deductions, modifications, or substitutions based on the principles of this invention.

Claims

1. A method for refined discrete element modeling and calibration of undisturbed fractured rock mass, characterized by: The method includes: Three-dimensional microstructural data of rock samples are obtained by CT scanning, and a three-dimensional digital model of the rock samples is generated. Joints and fissures inside the rock samples are identified based on slice images from different directions. Sieve analysis was conducted on the crushed slag particles of the same rock sample, and the full gradation distribution curve was obtained through the sieve analysis data; Representative rock blocks were selected from the three-dimensional digital model of the rock samples and their surface meshes were extracted to establish a database of real rock block morphology and generate a grain cluster template for each representative rock block. Based on the identification results of joints and fissures inside rock samples, the full gradation distribution curve of sieve tests, the real rock block morphology database, and the particle cluster template, a discrete element refined numerical model consistent with the size of the direct shear sample in the field was established from four aspects: rock block size distribution, rock block morphology distribution, rock block spatial distribution, joint and fissure morphology and contact state. Numerical experiments were conducted based on the refined discrete element numerical model to obtain the shear stress-shear displacement curve. The curves were compared point by point with the measured curves from the field direct shear test. The model parameters were adjusted until the two curves basically matched in terms of peak strength, residual strength, and softened section morphology, thus completing the parameter calibration.

2. The method for refined discrete element modeling and calibration of undisturbed fractured rock mass according to claim 1, characterized in that: Generate a three-dimensional digital model of the rock sample, including: Based on the three-dimensional microstructure data of the rock sample, color enhancement, binarization segmentation, noise reduction filtering, and three-dimensional surface reconstruction are performed sequentially to generate a three-dimensional digital model of the rock sample.

3. The method for refined discrete element modeling and calibration of undisturbed fractured rock mass according to claim 2, characterized in that: In the sieve analysis, the rock mass components in the rock sample were divided into three categories according to the particle size distribution: large rock blocks with a diameter of 3 to 20 cm, small rock blocks with a diameter of 0.5 to 3 cm, and cement with a diameter of no more than 0.5 cm.

4. The method for refined discrete element modeling and calibration of undisturbed fractured rock mass according to claim 3, characterized in that: Generate a grain cluster template for each representative rock block, including: For each representative rock block, a corresponding rigid cluster template is generated using the particle cluster method. The outer envelope of the rigid cluster template is consistent with the surface mesh geometry of the real rock block. The rigid cluster template was converted into a deformable flexible cluster template to simulate the deformation and local fracturing of rock blocks under stress.

5. The method for refined discrete element modeling and calibration of undisturbed fractured rock mass according to claim 4, characterized in that: A refined discrete element numerical model was established based on the rock block size distribution, including: The total gradation distribution curve obtained by sieve analysis is used to quantitatively control the quantity or mass percentage of large rock blocks, small rock blocks, and cementing material in the model.

6. The method for refined discrete element modeling and calibration of undisturbed fractured rock mass according to claim 5, characterized in that: Establishing a refined discrete element numerical model from the distribution of rock block morphology refers to generating irregularly shaped rock blocks using a particle cluster template, which is a deformable flexible cluster template.

7. The method for refined discrete element modeling and calibration of undisturbed fractured rock mass according to claim 6, characterized in that: Establishing a refined discrete element numerical model based on the spatial distribution of rock blocks includes: Based on the actual spatial location and attitude angle of the large rock blocks in the CT images, position mapping or orientation constraint methods are used to make the long axis of the large rock blocks oriented in a directional correlation along the dominant direction of the on-site geostress, simulating the interlocking and mosaic characteristics between the rock blocks.

8. The method for refined discrete element modeling and calibration of undisturbed fractured rock mass according to claim 7, characterized in that: A refined discrete element numerical model is established based on the joint and fracture morphology and contact state, including: For the contact between large rock blocks, a smooth joint model is used to simulate macroscopic cracks. The number of cracks is quantitatively characterized by the number of smooth joint contacts, and the crack size is quantitatively characterized by the characteristic size of the large rock block. For the relationships between large and small rock blocks, between small rock blocks, between rock blocks and cement, and between cement particles, a parallel bonding model is used, and the bonding performance is quantitatively controlled by adjusting the bonding strength parameters.

9. The method for refined discrete element modeling and calibration of undisturbed fractured rock mass according to claim 8, characterized in that: The refined numerical model of discrete element method includes upper and lower shear boxes, a normal loading plate, and a shear seam.

10. The method for refined discrete element modeling and calibration of undisturbed fractured rock mass according to claim 9, characterized in that: Numerical experiments were conducted based on the refined discrete element numerical model to obtain shear stress-shear displacement curves, including: In the numerical experiment, the normal load was kept constant by a servo control mechanism, and the horizontal shear displacement was applied at a constant rate. The calculated shear stress-shear displacement curve was recorded in real time.

Images