Search

CN-121997632-A - Rock slope discrete element modeling method based on random field theory

CN121997632ACN 121997632 ACN121997632 ACN 121997632ACN-121997632-A

Abstract

The invention discloses a rock slope discrete element modeling method based on random field theory, which comprises the steps of generating an initial homogeneous particle sample according to physical parameters and geometric boundaries, applying target confining pressure through servo control to enable the sample to reach a static equilibrium state, resetting the state of the sample, endowing adhesive characteristics to particle contact, constructing a homogeneous rock model, extracting unique identifiers and space coordinates of each contact, constructing a random field for representing space variability of cohesive force and tensile strength based on the unique identifiers and the space coordinates, distributing corresponding strength parameters to each contact, applying a gravity field amplified in proportion to the model to simulate prototype stress, executing grading excavation simulation, and evaluating slope stability through monitoring system displacement data. The invention realizes the high-precision mapping from random field parameters to discrete element models, and ensures the strict consistency of contact parameters in the whole simulation process under complex working conditions.

Inventors

  • ZHANG JUNHUI
  • XIONG GUANGWEI
  • YANG HAO
  • ZENG CHENG

Assignees

  • 长沙理工大学

Dates

Publication Date
20260508
Application Date
20251223

Claims (10)

  1. 1. A rock slope discrete element modeling method based on random field theory is characterized by comprising the following steps: s1, generating an initial homogeneous particle sample according to a set physical parameter and a geometric boundary; S2, applying target confining pressure to the initial homogeneous particle sample, and enabling the sample to reach a static equilibrium state under the target confining pressure through servo control; s3, resetting the state of the sample after stress initialization, and giving adhesive characteristics to contact among particles to construct a homogeneous rock model; s4, generating and deriving identification information comprising unique identification and space position coordinates for each contact in the homogeneous rock model; S5, constructing a random field representing the space variability of the cohesive force and the tensile strength parameters based on the identification information, and endowing corresponding cohesive force parameters and tensile strength parameters for each contact; And S6, applying a gravitational field amplified in proportion to the model endowed with the corresponding parameters to simulate prototype stress, and executing grading excavation simulation until the model reaches a new equilibrium state after excavation, so as to obtain displacement monitoring data for stability analysis.
  2. 2. The method for modeling a rock slope discrete element based on random field theory according to claim 1, wherein the step S1 comprises: Generating a random distribution particle set in the boundary according to the set physical parameters and geometric boundary, designating a contact model and parameters for the inter-particle contact in the particle set, achieving static equilibrium through cyclic iteration and kinetic energy reduction, and deleting particles outside the boundary to form an initial homogeneous particle sample.
  3. 3. The method for modeling a rock slope discrete element based on random field theory according to claim 1, wherein the step S2 comprises: According to the difference value between the target confining pressure and the monitoring variable, the control variable is adjusted by combining with a servo gain parameter, so that the monitoring variable approaches to the target confining pressure; The servo gain parameters are determined according to the total normal contact stiffness of the model in the pressing direction, the geometric dimension of the boundary wall body and the calculation time step.
  4. 4. The method for modeling a discrete element of a rock slope based on random field theory as claimed in claim 3, wherein the criteria of the static equilibrium state comprises that the relative error between the monitored variable and the target confining pressure is smaller than a preset threshold value, and the unbalanced force ratio of the particle system in the model is smaller than the preset threshold value or the controlled variable approaches zero.
  5. 5. The method for modeling a rock slope discrete element based on random field theory according to claim 1, wherein the step S3 comprises zeroing the displacement and rotation angle of all particles in the sample, and defining a combined contact model comprising a linear portion and a parallel bonding portion for all contacts and parameters thereof.
  6. 6. The method for discrete element modeling of a rock slope based on random field theory according to claim 1, wherein said step S4 comprises traversing all contacts in the model, assigning a unique identifier to each contact and recording its spatial coordinates, and writing said identifiers and spatial coordinates into a data file.
  7. 7. The method for modeling a rock slope discrete element based on random field theory according to claim 1, wherein the step S5 comprises: based on the unique identification and the space coordinates, respectively constructing an auto-covariance core and a cross-covariance core which characterize the spatial correlation of the cohesive force and the tensile strength of the rock mass so as to construct a covariance matrix; based on the monitoring data, the parameter mean vector and the covariance matrix are updated by constructing a conditional random field, and if no monitoring data exists, the unconditional mean and covariance matrix is directly adopted.
  8. 8. The method of modeling a discrete element of a rock slope based on random field theory according to claim 7, wherein in the step S5, the covariance matrix is decomposed by eigenvalues, the number of truncated modes is determined based on the cumulative contribution of eigenvalues, and KL expansion is used to generate a random field sample vector containing cohesive force and tensile strength values of each contact node.
  9. 9. The method for modeling a rock slope discrete element based on random field theory according to claim 8, wherein in the step S5, the random field sample vector is split into a cohesive force parameter field and a tensile strength parameter field, and is organized by a contact unique identifier and exported as a parameter file; and reading the parameter file, searching corresponding parameter values according to the unique identifier of the contact, and respectively endowing the discrete meta-model with the parameter attribute of the bonding strength of the corresponding contact.
  10. 10. The method for modeling a rock slope discrete element based on random field theory according to claim 1, wherein the step S6 comprises: Setting increased gravity acceleration according to geometric scaling coefficients of the model and the prototype size, so that the model reaches initial static balance under a corresponding gravity field; Virtual monitoring points are distributed at key positions of the side slope so as to record displacement and depth data of particles at each monitoring point; and (3) defining the boundary of the excavation area, deleting particles in the boundary, performing iterative calculation, and taking the unbalance force ratio of the model system as a balance convergence criterion when the unbalance force ratio is smaller than a preset threshold value or the displacement rate of the monitoring point is close to zero, thereby obtaining the new balance state and displacement monitoring data.

Description

Rock slope discrete element modeling method based on random field theory Technical Field The invention belongs to the technical field of slope engineering numerical simulation, and particularly relates to a rock slope discrete element modeling method based on random field theory. Background The rock slope engineering is a key component in the field of geotechnical engineering and widely exists in important engineering such as traffic, water conservancy, mine, infrastructure construction and the like. With the development of engineering construction to larger scale and higher complexity, the problem of rock slope stability has become one of key factors affecting engineering safety and economic benefits. In order to deepen understanding of a rock slope destabilizing mechanism and improve prediction accuracy, a numerical simulation technology is widely applied in the field. The Discrete Element Method (DEM) is widely applied to stability analysis of rock slope engineering because the discrete element method can effectively simulate discontinuous deformation processes such as rock mass fracture expansion, block separation and the like. However, in the current modeling process of the rock slope DEM, a homogeneous or simplified heterogeneous parameter assignment mode is still generally adopted, for example, a partition assignment method or a layered parameter assignment method based on geological investigation results is adopted. The modeling mode is difficult to truly reflect continuous variability and correlation of rock mass parameters (such as cohesive force c, internal friction angle phi and the like) in space, so that a simulation result has larger deviation from actual engineering performance in aspects of structural stress transmission, destructive path evolution and the like. In recent years, random field theory provides a strict theoretical basis and effective tool for describing the spatial variability of geologic medium parameters. In particular to a space random field modeling method based on a Karhunen-Loeve (K-L) expansion method and the like, which can efficiently construct a physical parameter distribution field with a predetermined space correlation structure on the premise of known parameter statistical characteristics. However, the application research of the method in the DEM simulation field is relatively few at present, the prior attempt usually directly takes the unit space position coordinate as an index of parameter matching, the method has the risk of inaccurate parameter mapping caused by calculation errors of the particle (or unit) coordinate, and meanwhile, a mechanism for tracking and matching parameters of the particle after the displacement of the particle under a complex working condition is lacking, so that strict consistency of the parameters and a model in the whole simulation process is difficult to ensure. Therefore, development of a modeling method integrating random field theory and Discrete Element Method (DEM) is needed, high-precision assignment of parameters is achieved through a contact parameter mapping mechanism based on unique identification, and strict consistency of contact parameters in the whole simulation process under complex working conditions is ensured. Disclosure of Invention The invention aims to provide a rock slope discrete element modeling method based on random field theory, which solves the problems that the space variability of rock parameters cannot be truly reflected, mapping is inaccurate due to coordinate matching and a parameter tracking mechanism after displacement is lacked in the prior art, realizes high-precision mapping of random field parameters to a discrete element model, and ensures strict consistency of contact parameters in the whole simulation process under complex working conditions. The technical scheme adopted by the invention is that the rock slope discrete element modeling method based on random field theory comprises the following steps: s1, generating an initial homogeneous particle sample according to a set physical parameter and a geometric boundary; S2, applying target confining pressure to the initial homogeneous particle sample, and enabling the sample to reach a static equilibrium state under the target confining pressure through servo control; s3, resetting the state of the sample after stress initialization, and giving adhesive characteristics to contact among particles to construct a homogeneous rock model; s4, generating and deriving identification information comprising unique identification and space position coordinates for each contact in the homogeneous rock model; S5, constructing a random field representing the space variability of the cohesive force and the tensile strength parameters based on the identification information, and endowing corresponding cohesive force parameters and tensile strength parameters for each contact; And S6, applying a gravitational field amplified in proportion to the model endowe