DEM-based tunnel slope rockfall process simulation method
By combining the DEM method with the DFN joint model and the local particle refinement algorithm, the problems of slope deformation and roughness in the simulation of rockfall in tunnel slopes were solved, achieving more accurate simulation of rockfall motion and improving the geometric realism and computational stability of the model.
Patent Information
- Authority / Receiving Office
- CN · China
- Patent Type
- Applications(China)
- Current Assignee / Owner
- THE 5TH ENG OF CHINA RAILWAY 22TH BUREAU GROUP
- Filing Date
- 2026-01-23
- Publication Date
- 2026-05-01
AI Technical Summary
Existing discrete element methods cannot accurately account for slope deformation and roughness when simulating rockfalls on tunnel slopes, resulting in inaccurate rockfall trajectories. This is especially true on slopes with complex shapes, where computational interference and unreasonable stress concentrations are likely to occur.
Using a DEM-based approach, a parallel bond model and a DFN joint model are constructed, combined with a local particle densification algorithm, to accurately represent the complex slope morphology and spatial location of unstable rock masses. The DFN and joint models are used to cover the slope surface to simulate the movement process of falling rocks, ensuring the accuracy of contact quantity and parameters.
It improves the geometric realism and reproducibility of tunnel slope rockfall simulation, reduces computational interference, and enhances the stability and accuracy of kinematic response, while also taking into account computational efficiency, enabling more accurate simulation of the interaction between rockfall and slope.
Smart Images
Figure CN121960083A_ABST
Abstract
Description
A simulation method for rockfall process on tunnel slopes based on DEM Technical Field
[0001] This invention relates to the field of numerical calculation technology in tunnel engineering, and more specifically to a method for simulating rockfall processes on tunnel slopes based on DEM. Background Technology
[0002] The southwestern mountainous region has complex terrain and geological conditions, resulting in numerous geological disasters. Frequent heavy rains and earthquakes in the southwest, combined with the impact of human-induced construction blasting, further develop mountain fissures, making the soil and rock mass more prone to loosening and deformation, and thus more susceptible to slope geological disasters. Improper human engineering activities can cause slope deformation and instability, leading to slope geological disasters. Slope geological disasters are mainly characterized by collapses, landslides, and debris flows. A unstable rock mass or collapse refers to a complete rock mass on a steep rock slope that has been fragmented into multiple isolated, unstable rock blocks and combinations by multiple structural planes. When a unstable rock mass is detached from the slope surface due to some reason (earthquake, rainfall, or artificial blasting), it undergoes various dynamic evolution processes such as collision, rebound, jumping, rolling, or sliding down the slope, eventually stopping in a flat area or due to obstacles. This process is called a rockfall event or collapse event. Generally, larger-scale collapses are called landslides, while smaller-scale collapses are called rockfalls. Discrete element method (DEM) simulation of slope rockfalls helps to deepen the understanding of rockfall hazards and enable faster and more comprehensive construction of protective facilities. In simulating landslides and rockfalls, the Discrete Element Method (DEM) can consider various contact patterns between rocks and slopes. In DEM, walls can be viewed as planar foundations, and parallel-bonded particles can be arranged to form spherical rockfall impact foundations to simulate the rockfall collision and fragmentation process. Alternatively, particles can be used to replace walls to simulate mountain slopes, providing a more comprehensive consideration of the interaction between slopes and rocks.
[0003] However, the existing discrete element method for simulating rockfall on slopes has the following shortcomings: (1) When using walls to simulate mountain slopes, the mountain slope is usually regarded as a rigid body that does not deform. Only the influence of rigid slopes on rockfall movement is considered. However, actual slopes are often covered with a weak and rough weathered layer, and the underlying layer is rigid rock mass. Rockfall may also cause some disturbance to the slope rock mass. Such disturbance will also affect the trajectory of rockfall movement. (2) The slope surface simulated by walls is flat and smooth, while the slope surface simulated by particles is uneven. The particles on the discrete element slope surface are disc-shaped or spherical and randomly generated, with initial roughness. (3) When particles replace walls to simulate mountain slopes, the more complex the simulated slope shape, the more easily the rockfall movement will be hindered or even stalled on the particle slope compared to using walls as slopes, resulting in unreasonable stress concentration in some areas.
[0004] Therefore, how to accurately simulate rockfall on tunnel slopes is a problem that urgently needs to be solved by those skilled in the art. Summary of the Invention
[0005] In view of this, the present invention provides a simulation method for rockfall process on tunnel slopes based on DEM, which realizes accurate simulation of rockfall on tunnel slopes and is used to study the contact form between rockfall and steep slopes and the movement characteristics of rockfall.
[0006] To achieve the above objectives, the present invention adopts the following technical solution: a method for simulating rockfall processes on tunnel slopes based on DEM, characterized by comprising: constructing a rock slope composed of parallel bonded models; setting a DFN and assigning a joint model to the surface of the rock slope; calibrating the microscopic parameters of the joint model using slope experiments; determining the shape and size parameters of the mountain and the falling rocks; constructing a discrete element model of complex-shaped mountain slopes and falling rocks; using a local densification algorithm to densify the slope and falling rock particles to ensure the contact quantity between the slope and the unstable rock mass; covering the surface of the complex-shaped mountain slope with the DFN and the joint model to weaken the joint parameters at the unstable rock mass; simulating the process of the unstable rock mass peeling off from the slope surface and the movement process of the falling rocks rolling down a steep slope; and recording the movement characteristics of the falling rocks.
[0007] Preferably, the micro-parameter calibration of the joint model using slope experiments includes: constructing a rock slope composed of parallel bonded models, setting DFN on the slope surface and assigning it a joint model; generating cubic-shaped rocks composed of parallel bonded particles, and freely releasing them from the top of the slope to simulate the sliding motion of rocks on a simple slope; generating circular rocks composed of parallel bonded particles, and freely releasing them from the top of the slope to simulate the rolling motion of rocks on a simple slope; generating circular rocks composed of parallel bonded particles, releasing them from a fixed height and causing them to collide with the simple slope; calibrating the micro-parameters by simulating single motions on different slope surfaces and comparing the experimental results.
[0008] Preferably, by using the calculation formulas for the sliding friction coefficient, rolling friction coefficient, and coefficient of restitution, the experimental results and simulation results are compared through the calculation formulas to inversely derive the microscopic parameters of different slopes. The calculation formulas are shown below: ; ; ; ;in, It is the acceleration due to gravity. The coefficient of sliding friction is Let be the velocity at time t. Let t be the coordinates at time t; A constant related to the mass and shape of the falling rock. Where L is the coefficient of rolling friction and L is the distance traveled. For the final speed; The normal restitution coefficient, : represents the tangential restitution coefficient. The normal velocity perpendicular to the slope before the collision is given. Let be the normal velocity perpendicular to the slope after the collision. The tangential velocity parallel to the slope before the collision. The velocity is the tangential velocity parallel to the slope after the collision.
[0009] Preferably, the micro parameters include the elastic modulus, normal stiffness, stiffness ratio, cohesion, and friction coefficient among the parallel bonding parameters, and the joint parameters include joint normal stiffness and joint tangential stiffness, normal damping coefficient, tangential damping coefficient, joint cohesion, and joint friction coefficient.
[0010] Preferably, the construction of a discrete element model of a complex-shaped mountain slope and rockfall includes: determining the size and shape of the mountain and the location, size and shape of the unstable rock mass before it transforms into a rockfall; and using discrete element software and CAD software to jointly divide the mountain and unstable rock mass and construct a discrete element model.
[0011] Preferably, the step of using discrete element method (DEM) software and CAD software to jointly divide the mountain and unstable rock mass includes: determining a suitable simulation box based on the CAD drawing of the geological profile, determining the average radius of the particles, generating a particle-filled simulation box, and grouping the particles according to the geological profile, further subdividing the mountain rock, and also grouping the locations of the unstable rock mass to achieve particle balance within the simulation box; using a local particle densification algorithm to densify the particles on the slope surface and within the unstable rock mass range to reduce the impact of surface unevenness on motion calculation; deleting particles outside the mountain, and assigning different parallel bonding parameters to different geological groups according to the geological zoning of the profile.
[0012] Preferably, the step of using a local particle densification algorithm to densify particles on the slope surface and within the unstable rock mass area includes: determining the densification factor k and the particle size ratio for each area. After obtaining the coordinate information (x, y) and radius information R of a particle in this region, the tangential restitution coefficient is... The encryption multiplier k and radius information R are used to calculate the radii (R1, R2) of two new particles. The sum of the volumes of the two new particles is the same as the volume of the original particle. The particle radius data is updated using the radius calculation formula. Next, the position coordinates of the two newly generated particles are calculated from the original particle. The two particles are in contact but do not overlap. The coordinates of the contact point between the new particles are located at the center of the original coordinates, and the positions of the two particles are randomly distributed around the contact point. The original particle is deleted, and two new particles are generated at the original particle's location based on the position and coordinate information. If the value of k is greater than 1 at this point, the newly generated particles will be further split into four smaller particles until k is less than 1 and the calculation stops. After encryption, the number of particles increases, the volume of each individual particle decreases, and finally the number of particles in the region becomes the original. The encryption process is repeated to bring the new model to equilibrium. ; .
[0013] Preferably, the method of using DFN and joint model to cover complex-shaped mountain slope surfaces includes: obtaining slope surface node information from geological profile CAD drawings, then using MATLAB to rewrite the node data into PFC-readable line segment data, using the line segments to generate complex-shaped DFNs to cover the slope surface, and assigning different joint parameters to the DFNs of different slope segments.
[0014] As can be seen from the above technical solutions, compared with the prior art, the present invention discloses a method for simulating the rockfall process of tunnel slopes based on DEM. The beneficial effects include: proposing an integrated modeling process of "CAD geometry - DEM particles - DFN joint coverage", which can accurately express complex slope morphology, spatial location of dangerous rock masses and their transformation into rockfall within the discrete element framework, significantly improving the geometric realism and reproducibility of the numerical model of rockfall on steep tunnel slopes.
[0015] A local particle refinement algorithm is proposed, which adaptively refines only the key contact control areas such as the slope surface and unstable rock mass. Under the premise of ensuring volume conservation and non-overlapping particles, it significantly increases the number of contact pairs and reduces the interference of the geometric "step effect" caused by the dispersion of particles on the slope surface on the calculation of slip, rolling and collision. This improves the stability and accuracy of kinematic response calculation while taking into account computational efficiency.
[0016] By using DFN and joint contact models to achieve segmented coverage and differentiated assignment of complex slopes, it is ensured that newly generated contact pairs automatically adopt the joint contact model. This allows slope roughness, structural surface weakening, and anisotropic friction / recovery characteristics to be directly reflected at the particle contact level, enhancing the ability to characterize the dynamic differences of real slopes (soil, turf, rock slopes, etc.). Attached Figure Description
[0017] 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 embodiments of the present invention. For those skilled in the art, other drawings can be obtained based on the provided drawings without creative effort.
[0018] Figure 1 is a flowchart of the method in the embodiment provided by the present invention; Figure 2 is a discrete element calibration model provided by the present invention; Figure 3 is a simple motion test diagram provided by the present invention; Figure 4 is a comparison diagram of discrete element calibration results provided by the present invention; wherein, (a) is sliding, (b) is rolling, (c) is normal collision, and (d) is tangential collision; Figure 5 is a geological profile of dangerous rock mass on a steep slope provided by the present invention; Figure 6 is a discrete element modeling diagram of dangerous rock mass on a steep slope provided by the present invention; Figure 7 is a flowchart of the local particle densification algorithm provided by the present invention; Figure 8 is a comparison diagram of the local particle densification effect provided by the present invention; Figure 9 is a segmented diagram of a steep slope provided by the present invention; Figure 10(a) is a diagram of the trajectory and velocity of falling rocks provided by the present invention; Figure 10(b) is a diagram of the rotation angle and rotation speed of falling rocks provided by the present invention. Detailed Implementation
[0019] The technical solutions of the embodiments of the present invention will be clearly and completely described below with reference to the accompanying drawings. 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 are within the scope of protection of the present invention.
[0020] This invention discloses a method for simulating rockfall processes on tunnel slopes based on a Discrete Element Model (DEM). The method includes: constructing a rock slope composed of parallel bonded models; setting a Discrete Element Model (DFN) on the surface of the rock slope and assigning a joint model; calibrating the microscopic parameters of the joint model using slope experiments; determining the shape and size parameters of the mountain and the falling rocks; constructing a discrete element model of the complex-shaped mountain slope and the falling rocks; using a local densification algorithm to densify the slope and falling rock particles to ensure the contact quantity between the slope and the unstable rock mass; covering the surface of the complex-shaped mountain slope with the DFN and the joint model to reduce the joint parameters at the unstable rock mass; simulating the process of the unstable rock mass peeling off from the slope surface and the movement of the falling rocks as they roll down a steep slope; and recording the characteristics of the falling rock movement.
[0021] On a simple slope (a straight slope with a flat surface), the basic motion process of falling rocks was simulated using a DFN (Discrete Element Model) and a joint model. The results were compared with experimental results to calibrate the microscopic parameters of the joint model. The shape and size parameters of the mountain and the falling rocks were determined, and a discrete element model of a complex-shaped mountain slope (an irregularly shaped slope) and falling rocks was constructed. A local densification algorithm was used to densify the slope and falling rock particles. The surface of the complex slope was covered with a DFN and a joint model, and falling rocks were initiated to simulate the complex motion process of falling rocks rolling down a steep slope, combining various motion forms. The contact form between the falling rocks and the steep slope and the motion characteristics of the falling rocks were studied.
[0022] A two-particle collision model was constructed to verify the characteristics of the joint model. A discrete element simple slope model was constructed, and the single motion forms of falling rocks, including sliding, rolling and collision, were simulated by combining the DFN method and the joint model.
[0023] The construction of the steep slope and rockfall motion model includes: determining the size and shape of the mountain and the location, size, and shape of the unstable rock mass before it transforms into a rockfall; using discrete element method (DEM) software and CAD software to jointly divide the mountain and unstable rock mass, and constructing a DEM mountain and rockfall model; using a local densification algorithm to densify the particles on the mountain slope surface and the unstable rock mass to ensure the contact quantity between the slope and the unstable rock mass; importing the slope shape using CAD software to create a DFN and joint model to cover the slope surface; reducing the joint parameters at the unstable rock mass to simulate the process of the unstable rock mass peeling off from the slope surface under rainfall or earthquake action; and simulating the motion characteristics of the unstable rock mass after it transforms into a rockfall and falls from the steep slope.
[0024] As shown in Figure 1, the specific process is as follows: S1. Construct a rock slope composed of parallel bonded models, set DFN on the slope surface and assign it a joint model to ensure that the newly generated contact pairs are joint contact models; generate cubic-shaped rocks composed of parallel bonded particles and release them freely from the top of the slope to simulate the sliding motion of rocks on a simple slope; generate circular rocks composed of parallel bonded particles and release them freely from the top of the slope to simulate the rolling motion of rocks on a simple slope; generate circular rocks composed of parallel bonded particles and release them from a fixed height to collide with the slope, as shown in Figure 2; S4. Simulate the sliding, rolling, and collision motions on soil, turf, and granite slopes and compare them with the experimental results to calibrate microscopic parameters, as shown in Figure 3. Using the calculation formulas for the sliding friction coefficient, rolling friction coefficient, and coefficient of restitution, compare the experimental and simulation results through calculation formulas to inversely determine the microscopic parameters of different slope surfaces. The calculation formulas are as follows: ; ; ; ;in, It is the acceleration due to gravity. The coefficient of sliding friction is Let be the velocity at time t. Let t be the coordinates at time t; A constant related to the mass and shape of the falling rock. Where L is the coefficient of rolling friction and L is the distance traveled. For the final speed; The normal restitution coefficient, The tangential restitution coefficient, The normal velocity perpendicular to the slope before the collision is given. Let be the normal velocity perpendicular to the slope after the collision. The tangential velocity parallel to the slope before the collision. The velocity is the tangential velocity parallel to the slope after the collision.
[0025] The calibrated microscopic parameters include the elastic modulus, normal stiffness, stiffness ratio, cohesion, and friction coefficient in the parallel bond parameters, and joint parameters including joint normal stiffness and joint tangential stiffness, normal damping coefficient, tangential damping coefficient, joint cohesion, and joint friction coefficient. The calibrated parameters are shown in Table 1 below (the values within the parameter definition domain represent values for soil, turf, and granite slopes, respectively). The calibration results are shown in Figure 4.
[0026] Table 1 Calibration Parameter Table
[0027] S5. Clearly define parameters such as mountain size, slope shape, and shape and size of unstable rock mass in the geological profile map; based on the ground survey and preliminary and detailed drilling data of this section, determine the calculation parameters of rock and strata after considering the degree of fracture penetration, fracture filling degree and fracture bonding. It is used as a road filling material and has high strength parameters. It is composed of relatively soft clay, with relatively low parameters. Lower. Granite The mechanical properties vary depending on the number and degree of fissures, as well as the degree of weathering. Specific rock mass calculation parameters are shown in Table 2, and the specific geological profile of the mountain is shown in Figure 5.
[0028] Table 2 Rock Mass Calculation Parameters
[0029] S6. Using particles to simulate mountains and boulders, and joints to simulate the contact between boulders and the mountain, based on the given geological profile of the dangerous rock section, determine a suitable simulation box, determine the average particle radius, generate particle-filled simulation boxes, and group the particles according to the geological profile. Granite is further subdivided into weakly weathered granite, moderately weathered granite, and so on. , This process involves bringing the particles within the simulation chamber to equilibrium; deleting particles outside the mountain; and assigning different parallel bonding parameters to different geological groups based on the geological zoning of the cross-section. The walls are then deleted, and gravity and boundary constraints are set to bring the discrete element model back to equilibrium, as shown in Figure 6.
[0030] S7. A local particle density algorithm is used to densify the particles on the slope surface and within the unstable rock mass area to reduce the impact of surface unevenness on motion calculations. The flowchart of the local particle density algorithm is shown in Figure 7. The density factor k and particle size ratio are determined for each region. After obtaining the coordinates (x, y) and radius (R) of a particle in the region, the radii (R1, R2) of two new particles are calculated based on Rt, k, and R. The sum of the volumes of the two new particles is the same as the volume of the original particle. The radius calculation formula is as follows. Next, the position coordinates of the two newly generated particles are calculated from the original particle. The two particles are in contact with each other but do not overlap. The contact point between the new particles is located at the center of the original coordinates, and the positions of the two particles are randomly distributed around the contact point. The original particle is deleted, and two new particles are generated at the original particle's position based on the position and coordinate information. If the value of k is greater than 1 at this time, the newly generated particles will be further split into four smaller particles until k is less than 1 and the calculation stops. The comparison before and after particle encryption is shown in Figure 8. The number of particles increases, the volume of a single particle decreases, and finally the number of particles in the region becomes the original. The encryption process is repeated to bring the new model to equilibrium.
[0031] ; S8. Obtain the slope surface node information from the CAD profile drawing, then use MATLAB to rewrite the node data into line segment data readable by discrete element software, and use the line segments to generate complex-shaped DFNs to cover the slope surface. As shown in Figure 9, different slope joint parameters are assigned to the DFNs of different slope segments. The different slope joint parameters are shown in Table 3.
[0032] Table 3 Calculation parameters for joint model
[0033] S9. Set an appropriate time step, release the unstable rock mass, and record the trajectory, speed, rotation angle, rotation speed and other motion characteristics of the falling rock as shown in Figure 10.
[0034] The various embodiments in this specification are described in a progressive manner, with each embodiment focusing on its differences from other embodiments. Similar or identical parts between embodiments can be referred to interchangeably. The methods disclosed in the embodiments are described simply because they correspond to the methods disclosed in the embodiments; relevant parts can be found in the method section.
[0035] The above description of the disclosed embodiments enables those skilled in the art to make or use the invention. Various modifications to these embodiments will be readily apparent to those skilled in the art, and the general principles defined herein may be implemented in other embodiments without departing from the spirit or scope of the invention. Therefore, the invention is not to be limited to the embodiments shown herein, but is to be accorded the widest scope consistent with the principles and novel features disclosed herein.
Claims
1. A method for simulating rockfall processes on tunnel slopes based on DEM, characterized in that, include: A rock slope composed of parallel bonded models was constructed. A discrete element model (DFN) was set on the surface of the rock slope and a joint model was applied. The microscopic parameters of the joint model were calibrated using slope experiments. The shape and size parameters of the mountain and falling rocks were determined. A discrete element model of the complex-shaped mountain slope and falling rocks was constructed. A local densification algorithm was used to densify the slope and falling rock particles to ensure the amount of contact between the slope and the unstable rock mass. The surface of the complex-shaped mountain slope was covered with the DFN and joint model to reduce the joint parameters at the unstable rock mass. The process of the unstable rock mass peeling off from the slope surface and the movement of falling rocks rolling down a steep slope were simulated, and the movement characteristics of the falling rocks were recorded.
2. The method for simulating rockfall processes on tunnel slopes based on DEM according to claim 1, characterized in that, The method of calibrating the micro-parallel bonded rock slope model using slope experiments includes: constructing a rock slope composed of parallel bonded particles, setting DFN on the slope surface and assigning it a jointed model; generating cubic-shaped rocks composed of parallel bonded particles and releasing them freely from the top of the slope to simulate the sliding motion of rocks on a simple slope; generating circular rocks composed of parallel bonded particles and releasing them freely from the top of the slope to simulate the rolling motion of rocks on a simple slope; generating circular rocks composed of parallel bonded particles and releasing them from a fixed height to collide with the simple slope; calibrating the micro-parallel parameters by simulating single motions on different slope surfaces and comparing the experimental results.
3. The method for simulating rockfall processes on tunnel slopes based on DEM according to claim 2, characterized in that, By using the calculation formulas for the sliding friction coefficient, rolling friction coefficient, and coefficient of restitution, the experimental results and simulation results are compared through the calculation formulas to inversely derive the microscopic parameters of different slopes. The calculation formulas are shown below: ； ； ； ;in, It is the acceleration due to gravity. The coefficient of sliding friction is Let be the velocity at time t. Let be the coordinates at time t; A constant related to the mass and shape of the falling rock. Where L is the coefficient of rolling friction and L is the distance traveled. For the final speed; The normal restitution coefficient, : represents the tangential restitution coefficient. The normal velocity perpendicular to the slope before the collision is given. Let be the normal velocity perpendicular to the slope after the collision. The tangential velocity parallel to the slope before the collision. The velocity is the tangential velocity parallel to the slope after the collision.
4. The method for simulating rockfall processes on tunnel slopes based on DEM according to claim 2, characterized in that, The microscopic parameters include the elastic modulus, normal stiffness, stiffness ratio, cohesion, and friction coefficient in the parallel bonding parameters, and the joint parameters include joint normal stiffness and joint tangential stiffness, normal damping coefficient, tangential damping coefficient, joint cohesion, and joint friction coefficient.
5. The method for simulating rockfall processes on tunnel slopes based on DEM according to claim 1, characterized in that, The construction of a discrete element model for complex mountain slopes and rockfalls includes: determining the size and shape of the mountain and the location, size, and shape of the unstable rock mass before it transforms into a rockfall; and using discrete element software and CAD software to jointly divide the mountain and unstable rock mass and construct a discrete element model.
6. The method for simulating rockfall processes on tunnel slopes based on DEM according to claim 5, characterized in that, The method of using discrete element method (DEM) software and CAD software to divide mountains and unstable rock masses includes: determining a suitable simulation box based on the CAD drawings of the geological profile, determining the average radius of the particles, generating a particle-filled simulation box, and grouping the particles according to the geological profile. The rocks of the mountain are further subdivided, and the locations of unstable rock masses are also grouped to achieve particle balance within the simulation box; a local particle densification algorithm is used to densify the particles on the slope surface and within the range of unstable rock masses to reduce the impact of surface unevenness on motion calculation; particles outside the mountain are deleted, and different parallel bonding parameters are assigned to different geological groups according to the geological zoning of the profile.
7. The method for simulating rockfall processes on tunnel slopes based on DEM according to claim 6, characterized in that, The method of using a local particle densification algorithm to densify particles on the slope surface and within the unstable rock mass area includes: determining the densification factor k and the particle size ratio for each area. After obtaining the coordinates (x, y) and radius R of a particle in the region, the tangential recovery coefficient is used... The encryption multiplier k and radius information R are used to calculate the radii (R1, R2) of two new particles. The sum of the volumes of the two new particles is the same as the volume of the original particle. The particle radius data is updated using the radius calculation formula. Next, the position coordinates of the two newly generated particles are calculated from the original particle. The two particles are in contact but do not overlap. The coordinates of the contact point between the new particles are located at the center of the original coordinates, and the positions of the two particles are randomly distributed around the contact point. The original particle is deleted, and two new particles are generated at the original particle's location based on the position and coordinate information. If the value of k is greater than 1 at this point, the newly generated particles will be further split into four smaller particles until k is less than 1 and the calculation stops. After encryption, the number of particles increases, the volume of each individual particle decreases, and finally the number of particles in the region becomes the original. The encryption process is repeated to bring the new model to equilibrium. ； 。 8. The method for simulating rockfall processes on tunnel slopes based on DEM according to claim 1, characterized in that, The method of using DFN and joint model to cover complex-shaped mountain slope surfaces includes: obtaining slope surface node information from geological profile CAD drawings, then using MATLAB to rewrite the node data into PFC-readable line segment data, using the line segments to generate complex-shaped DFNs to cover the slope surface, and assigning different joint parameters to the DFNs of different slope segments.