A tailings dam multi-field coupling discrete element catastrophe early warning method based on energy evolution criterion

Abstract

The application relates to the technical field of tailing dam safety monitoring and early warning, in particular to a tailing dam multi-field coupling discrete element catastrophe early warning method based on an energy evolution criterion. Five modules are constructed in the Fangzhizhun software: a discrete element model construction module: a non-continuous medium model of a tailing particle accumulation body is established to represent the mechanical behavior between particles; a multi-field coupling calculation module: dynamic coupling simulation of a seepage field-stress field-chemical field-temperature field is realized; an energy evolution analysis module: a real-time tracking system energy accumulation, transformation and release process is realized, and a critical feature before catastrophe is extracted; an efficient algorithm optimization module: parallel computing and a reduced model are used to improve the calculation efficiency; a dynamic early warning decision module: based on the energy criterion and monitoring data fusion, a hierarchical early warning signal is output. The method breaks through the limitation of traditional single parameter early warning such as displacement and pore water pressure, realizes accurate capture of tailing dam catastrophe precursors from the energy driving perspective, and improves the timeliness and reliability of early warning.

Description

Technical Field

[0001] This invention relates to the field of tailings dam safety monitoring and early warning technology, specifically to a tailings dam multi-field coupled discrete element disaster early warning method based on energy evolution criteria. Background Technology

[0002] Tailings dams are structures used to store tailings generated during mining operations, typically constructed from materials such as cement, stone, and sand. With the continuous development of the mining industry, tailings dams have been widely used globally. However, the safety of tailings dams remains a major challenge in mining engineering. Instability and damage to tailings dams often lead to enormous property losses and casualties. Therefore, effectively monitoring the operational status of tailings dams and predicting potential catastrophic events has become an urgent issue to be addressed.

[0003] Currently, tailings dam safety monitoring and early warning technologies are mainly divided into the following categories: (1) Traditional monitoring methods. Traditional methods mainly rely on physical sensor monitoring, including displacement gauges, piezometers, GPS monitoring, etc., combined with manual inspection to periodically collect parameters such as surface deformation and internal pore water pressure of tailings dams. However, traditional monitoring methods rely on threshold alarms, which can only trigger early warnings after deformation or seepage reaches dangerous values, and cannot capture the energy accumulation process before the disaster. In addition, the sensor deployment density is limited, making it difficult to cover the complex internal structure of tailings dams and easily missing local instability signals. It only focuses on a single physical field (such as displacement field or seepage field) and ignores the influence of multi-field coupling on stability. (2) Numerical simulation methods. Existing numerical simulation technologies mainly rely on the finite element method (FEM) and the finite difference method (FDM) to predict the stress-strain response of tailings dams through continuous medium mechanical models. However, due to the significant heterogeneity and discreteness of tailings dam materials (such as particle accumulation and crack development), continuous medium models are difficult to accurately characterize their mechanical behavior. Existing models often employ unidirectional or weak coupling (e.g., considering only seepage-stress coupling), neglecting the influence of chemical fields (such as tailings particle dissolution and cementation) and temperature fields. Furthermore, high-precision multi-field coupling models involve enormous computational demands, making it difficult to meet real-time early warning requirements.

[0004] Due to the shortcomings of the above methods, the discrete element method (DEM) has been introduced into tailings dam analysis in recent years to simulate the behavior of discontinuous media through interparticle contact mechanics. However, existing DEM models mostly focus on mechanical field simulation and do not integrate multi-physics field coupling mechanisms such as seepage field and chemical field, and the energy evolution mechanism is unclear: there is a lack of quantitative correlation analysis between the spatiotemporal evolution law of system energy (such as elastic energy, dissipated energy, and kinetic energy) and the catastrophic critical characteristics.

[0005] In summary, the following key bottlenecks exist in the field of tailings dam disaster early warning: The multi-field coupling mechanism is imperfect; existing models fail to achieve fully coupled dynamic simulation of the seepage field, stress field, chemical field, and temperature field, leading to a one-sided analysis of disaster causes. Energy evolution criteria are lacking; quantitative early warning indicators based on the system's energy accumulation, transformation, and release processes are insufficient, making it difficult to capture critical disaster characteristics. The discrete medium characterization is inadequate; the continuous medium assumption does not match the actual discrete structure of tailings dams, limiting simulation accuracy. There is a conflict between real-time performance and computational efficiency; high-precision multi-field coupling models are computationally time-consuming and cannot meet the engineering requirements for real-time early warning.

[0006] Therefore, it is necessary to develop a tailings dam disaster early warning method based on the system's energy accumulation, conversion and release processes, with multi-field coupling, good computational efficiency and real-time performance. Summary of the Invention

[0007] To address the shortcomings of existing technologies in traditional tailings dam safety monitoring methods, such as insufficient analysis of multiphysics coupling effects, unclear disaster evolution mechanisms, and limited early warning indicators, this invention proposes a multiphysics coupling discrete-element disaster early warning method for tailings dams based on energy evolution criteria, as detailed below:

[0008] A multi-field coupled discrete-element disaster early warning method for tailings dams based on energy evolution criteria is proposed, and modeled in simulation software, including the following steps:

[0009] Step 1: Constructing the Discrete Element Model: Establish a discontinuous medium model of the tailings particle accumulation, and introduce a cementation model to calculate the interparticle mechanical behavior: interparticle normal force F n and tangential force F t The unit is N;

[0010] Step 2: Multi-field coupling calculation: Based on the discrete element model obtained in Step 1, perform seepage field-stress field coupling, chemical field coupling, and temperature field coupling;

[0011] Step 3: Energy Evolution Analysis: Based on the model obtained in Steps 1 and 2, extract Fn from the model under equilibrium conditions. 、 F t Subsequently, energy classification and energy mutation rate index extraction are performed, while physical parameters such as velocity, force, and displacement are extracted from the set monitoring points.

[0012] Step 3.1: Energy Classification and Calculation:

[0013] Elastic energy E e Energy stored in particle contact deformation

[0014] Where, δ n δ tThese represent the normal and tangential displacements, respectively, in meters (m).

[0015] Dissipated energy E d Energy dissipated by friction and viscosity, E d =∑μF n Δδ t +∑η(Δv) 2 Δt+∑βγ 2 V cell Δt;

[0016] Where, μ: interparticle friction coefficient, μ = 0.3~0.6; Δδ t : Tangential displacement increment, in meters; η: Fluid dynamic viscosity, in Pa·s; Δv: Fluid-particle relative velocity, in meters per second; Δt: Time change, in seconds; β: Viscous dissipation coefficient, β = 2η; γ: Local shear rate, in seconds. -1 V cell Fluid mesh volume, in meters (m). 3 .

[0017] Kinetic energy E k Kinetic energy of particle motion

[0018] Where, m p The mass of the particles is expressed in kg; v p Particle velocity, in m / s;

[0019] Step 3.2: Energy Mutation Feature Extraction:

[0020] Energy mutation rate index: defined as the ratio of the rate of accumulation of elastic energy to the rate of release of dissipated energy. critical value δP ext For external input power, when R total >R crit The time is determined to be energy instability, and R is set to... total >R crit As one of the early warning criteria;

[0021] Where R crit Solve by simultaneously applying the energy conservation equation and the instability criterion:

[0022] Energy conservation equation:

[0023]

[0024] P ext External input power (such as rainfall seepage, seismic disturbance).

[0025] Instability Criterion (Hill Condition):

[0026]

[0027] That is, the system becomes unstable when its second-order energy variation is negative.

[0028] Solving the system of equations simultaneously, we get:

[0029]

[0030] Numerical solutions are required using material constitutive relations (such as the Drucker-Prager model) and boundary conditions.

[0031] Energy entropy increase index: Calculate the energy distribution entropy of the system S = -k B ∑p i lnp i , where k B Let be the Boltzmann constant, taken as 1.38 × 10⁻⁶. -23 J / K, p i The critical threshold for the energy production rate, representing the energy proportion of each sub-region. Where λ is the damping coefficient (determined by shaking table test; typical value for sandy tailings is 1000–1500 W·s / m). 3 ), τ relax The energy relaxation time for discrete element simulation fitting (DEM simulation fitting, typically 10–30 s). For the kinetic energy spatial gradient (extracted from pre-catastrophic DEM data), S th The critical velocity at which the energy distribution within a tailings dam system shifts from order to disorder is used to quantify the rate of increase of entropy. As one of the early warning criteria;

[0032] Step 4: Early Warning Decision: Based on the characteristics of energy mutation, output tiered early warning signals:

[0033] Normal: R total <0.8R crit and

[0034] Level 3 warning: 0.8R crit ≤R total <R crit or

[0035] Level 2 Warning: R total ≥R crit or

[0036] Level 1 Warning: R total ≥1.5R crit or kinetic energy E kThe surge exceeded the historical peak by more than 20% and lasted for more than 10 minutes.

[0037] Among them, the level 1 warning is more dangerous than the level 2 warning, and the level 2 warning is more dangerous than the level 3 warning.

[0038] Furthermore, the discrete element model construction in step 1 specifically includes the following steps:

[0039] Step 1.1 Particle size classification configuration

[0040] The particle size distribution and percentage of particles were determined by indoor sieving experiments and laser particle size analyzer. The tailings gradation curve was plotted, and the particle size range was determined based on the tailings gradation curve.

[0041] Step 1.2 Particle Template Generation

[0042] To determine the particle size range, the hyperellipsoid equation is used. Generate a non-spherical particle template, where n controls the sharpness of the edges and corners to simulate the actual shape distribution of tailings particles;

[0043] Step 1.3 Input particle properties

[0044] Input the tailings particle parameters measured in the laboratory, including density ρ, elastic modulus E, thermal expansion coefficient α and thermal conductivity k, introduce the cementation model, and finally obtain the discontinuous medium model of the tailings particle accumulation.

[0045] Step 1.4 Calculate the interparticle normal force F based on the discontinuous medium model of tailings particle accumulation. n and tangential force F t .

[0046] Furthermore, in step 1.4 of step 1, the Hertz-Mindlin contact theory is used to calculate the interparticle normal force. Calculate the tangential force based on the Mindlin-Deresiewicz model. In the formula: E * Equivalent elastic modulus, in Pa, R * The equivalent radius is in meters (m), μ is the interparticle friction coefficient, and δ is the effective radius. n ,δ t These are the normal and tangential displacements, respectively.

[0047] Furthermore, the bonding model introduced in step 1, wherein the tensile strength σ of the bonding bond is input... max and shear strength τ max The unit is kPa, which simulates the formation and destruction process of cementitious materials between tailings particles.

[0048] Furthermore, in the seepage field-stress field coupling in step 2, based on the discrete element model in step 1, the following steps are performed:

[0049] Step 2.1 Pore fluid simulation

[0050] Fluid motion is solved based on the Navier-Stokes equations, using porosity... Dynamically modified fluid mesh permeability Darcy-Forchheimer equation, where V particle V is the volume of the grain mesh. cell For fluid mesh volume;

[0051] Step 2.2 Fluid-structure interaction

[0052] Calculate the drag force of the fluid on the particles. Where C d ρ is the drag coefficient. f Fluid density, unit: kg / m³ 3 d p Particle diameter, in meters (m) and um. f The velocity vector of the undisturbed fluid at the center of the particle, in m / s, u p The particle velocity vector is expressed in m / s, |u f -u p | represents the relative velocity between the fluid and the particles, in m / s. (u) in the formula... f -u p This ensures that the direction of the force is consistent with the direction of the relative velocity.

[0053] Furthermore, in the chemical field coupling of step 2, based on the discrete element model of step 1, the following steps are performed:

[0054] Step 2.3 Dissolution-precipitation equation

[0055] The dissolution rate of tailings particles is defined based on the Arrhenius formula. Where k0 is the reaction rate constant, R is the gas constant, T is the temperature in K, S is the solution saturation, and E is the gas saturation. a Activation energy, in kJ / mol;

[0056] Step 2.4 Formation of cementitious material

[0057] Based on the solution ion concentration (e.g., Ca) 2+ SO4 2- (etc.) Dynamically generate cementitious materials and update the tensile strength σ of interparticle cement bonds. max and shear strength τ max , unit: kPa.

[0058] Furthermore, in the temperature field coupling of step 2, based on the discrete element model of step 1, the following steps are performed:

[0059] Step 2.5 Heat Conduction Equation

[0060] Calculating interparticle heat conduction using Fourier's law And consider the heat transfer q of fluid convection. conv =h(T) p -T f The impact of )

[0061] Where q is the heat flux density, in W / m³ 2 k is the thermal conductivity, with units of W / (m·K). For temperature gradient, the unit is K / m, q conv h is the convective heat transfer density, in W / m³, and h is the convective heat transfer coefficient, in W / (m³). 2 ·K), T p The particle surface temperature is expressed in Kelvin (K) and temperature (T). f This represents the local temperature of the fluid, in Kelvin (K).

[0062] Step 2.6 Thermal Expansion Effect

[0063] The particle radius changes with temperature Δr=r0αΔT, where r0 is the original particle radius in meters; α is the coefficient of thermal expansion, which affects the contact force and porosity.

[0064] Furthermore, the computational process in the method is optimized by allocating particle contact force calculation, fluid solution, and energy statistics tasks to the GPU core, and using the CUDA architecture to achieve real-time simulation of the billion-particle model; the tailings dam model is divided into several subdomains, each of which is calculated independently and then the boundary data is synchronized via the MPI protocol; the time step Δt is dynamically adjusted according to the energy evolution rate |dE / dt|.

[0065]

[0066] Where Δt new To adjust the step size, Δt old The step size before adjustment is ∈, where ∈ is the allowable energy error threshold.

[0067] Furthermore, in step 4, the early warning decision-making process, the energy mutation characteristics are fused with the monitoring data to output a graded early warning signal, and the monitoring data is monitored by sensors.

[0068] Furthermore, the simulation software used is COMSOL or PFC.

[0069] This invention employs a comprehensive analysis method combining multi-field coupling and energy evolution criteria to comprehensively and accurately assess the stability of tailings dams and issue effective early warnings before disasters occur. Compared to traditional single-physics field analysis methods, this invention provides a more accurate, real-time, and forward-looking tailings dam disaster early warning scheme. Utilizing high-precision simulation using the discrete element method, sensitivity analysis based on energy evolution criteria, and an intelligent automatic response mechanism, this invention effectively improves the accuracy and efficiency of tailings dam disaster early warning, reducing the occurrence and losses of disasters. Specific beneficial effects are as follows:

[0070] (1) Improving the accuracy of tailings dam disaster early warning: The catastrophic behavior of tailings dams is caused by the interaction of multiple physical fields, including solid mechanical fields, seepage fields, and thermal fields. This invention introduces a multi-field coupling model, which can simultaneously consider the interaction of multiple physical fields within the same analytical framework, thereby more accurately simulating the real behavior of tailings dams. Through multi-field coupling simulation, this invention can comprehensively consider multiple factors such as soil stress changes, the hydraulic effects of seepage fluids, and possible temperature changes, and comprehensively assess the stability of tailings dams. For example, water seepage causes soil softening, which may lead to dam settlement, while temperature changes may affect the mechanical properties of materials. Through this comprehensive coupling analysis, a more accurate basis for judging the instability of tailings dams can be provided.

[0071] (2) Early Instability Prediction Based on Energy Evolution Criterion: As an effective physical judgment tool, the energy evolution criterion plays a crucial role in early warning of system instability. Changes in a system's energy often reveal its stability or instability trends in advance. Through energy evolution analysis, changes in tailings dams due to external or internal forces during operation can be accurately captured. This invention, based on the energy evolution criterion, monitors the energy accumulation, consumption, and transformation processes of the tailings dam system, enabling the identification of potential instability risks in the early stages of a disaster. In particular, sudden energy changes or abnormal accumulation are often precursors to tailings dam instability. By combining the energy evolution criterion, the sensitivity and accuracy of tailings dam disaster early warning can be effectively improved. When a part of the tailings dam experiences overload or internal deformation, its energy changes are usually reflected; when the system's energy exceeds a certain threshold, instability may be triggered. Through this energy evolution determination, early warning signals can be issued in the early stages of a disaster, buying valuable time for emergency response.

[0072] (3) High-precision simulation and prediction using the discrete element method: The discrete element method can accurately simulate the interaction between particles and perform detailed analysis of material changes, stress distribution, and deformation inside tailings dams. Through the simulation of the discrete element method, the detailed dynamic behavior of tailings dams under complex working conditions can be captured, including local settlement, crack development, and particle flow behavior. Traditional continuous medium theory cannot fully consider the interaction between particles in tailings dams and their nonlinear characteristics, while the discrete element method can accurately simulate the collision, slip, and distribution changes between particles, revealing the microscopic changes of tailings dams during disasters. Discrete element simulation can realistically reflect the deformation and movement process of materials inside tailings dams, capturing possible instability behaviors before disasters occur, thus providing a scientific and reliable basis for disaster early warning of tailings dams.

[0073] (4) Dynamic and Real-Time Disaster Early Warning Capability. This invention, through the combination of multi-field coupling and energy evolution criteria, enables tailings dam disaster early warning not only to reflect the static stability state but also to dynamically assess the evolution process of the tailings dam under different operating conditions. This dynamic analysis capability allows the tailings dam disaster early warning system to respond in real-time to changes in the stability of the tailings dam and promptly detect potential disaster risks. Tailings dam disasters usually occur when some minor, localized changes begin, which traditional monitoring methods struggle to capture in a timely manner. This invention, through energy evolution and multi-field coupling analysis, can provide danger warnings to relevant personnel when minor changes in energy occur within the tailings dam. Furthermore, the dynamic analysis of real-time monitoring data allows for real-time adjustment of the disaster early warning strategy based on the latest operating status of the tailings dam, improving the accuracy and timeliness of disaster prediction.

[0074] (5) Enhancing the intelligence level of the tailings dam disaster early warning system. This invention also achieves automated and intelligent processing of tailings dam disaster early warning by combining intelligent algorithms based on a multi-field coupling model. The system can automatically assess the stability of the tailings dam based on real-time monitoring data and energy evolution trends, and automatically issue early warning signals according to set thresholds, reducing human intervention and improving response speed and accuracy. The intelligent early warning system can automatically trigger corresponding emergency procedures when an anomaly occurs in the tailings dam, such as adjusting the water level of the tailings dam, reinforcing the dam structure, and evacuating personnel. This intelligent response mechanism not only improves the efficiency of early warning but also ensures that tailings dam disaster events are responded to in a timely and effective manner.

[0075] In summary, this invention not only has significant technological innovation in tailings dam disaster early warning technology, but also provides practical technical support for the safety management and emergency response of tailings dams, with broad application prospects and huge social benefits. Attached Figure Description

[0076] The embodiments of the present invention will be further described below with reference to the accompanying drawings, wherein:

[0077] Figure 1 A flowchart of the method of the present invention is shown. Detailed Implementation

[0078] To make the objectives, technical solutions, and advantages of this invention clearer, the invention will be further described in detail below with reference to the accompanying drawings and specific embodiments. It should be understood that the specific embodiments described herein are merely illustrative and not intended to limit the invention.

[0079] In one embodiment, a tailings dam multi-field coupled discrete-element disaster early warning method based on energy evolution criteria is modeled in the simulation software COMSOL, including the following steps:

[0080] Step 1: Construct the discrete element model:

[0081] Step 1.1 Particle size classification configuration

[0082] The particle size distribution and percentage of particles were determined through indoor sieving experiments and laser particle size analyzer. A tailings gradation curve was plotted, and based on this curve, the particle size range (d) was determined. p =0.05mm~5mm;

[0083] Step 1.2 Particle Template Generation

[0084] To determine the particle size range, the hyperellipsoid equation is used. A non-spherical particle template is generated, where n controls the sharpness of the edges and corners, taking values ​​from 1.2 to 2.5, and the major axis ratio a:b:c = 1:(0.6 to 0.9):(0.4 to 0.7). The actual shape distribution of tailings particles is simulated based on the tailings gradation curve, according to the particle size range d. p =0.05mm~5mm to form particle aggregates;

[0085] Step 1.3 Input particle properties

[0086] Input the tailings particle parameters obtained from indoor experiments, including density ρ = 2650 kg / m³. 3 Based on the elastic modulus E = 50 GPa, the coefficient of thermal expansion α, and the thermal conductivity k, a cementation model is introduced, with the cementation bond tensile strength σ as input. max =12.6 kPa and shear strength τ max =18.5kPa, to simulate the formation and destruction process of cement between tailings particles, and finally obtain the discontinuous medium model of tailings particle accumulation;

[0087] Step 1.4 Calculate the interparticle normal force F based on the discontinuous medium model of tailings particle accumulation.n and tangential force F t

[0088] The Hertz-Mindlin contact theory was used to calculate the normal force between particles. Calculate the tangential force based on the Mindlin-Deresiewicz model. In the formula: E * This is the equivalent elastic modulus, in Pa. v = 0.23 is Poisson's ratio, R * The equivalent radius is in meters (m), μ is the interparticle friction coefficient, μ = 0.6, δ n ,δ t These are the normal and tangential displacements, respectively.

[0089] Step 2: Multi-field coupling calculation:

[0090] Based on the discrete element model obtained in step 1, the CFD-DEM method is used to perform seepage field-stress field coupling:

[0091] Step 2.1 Pore fluid simulation

[0092] Fluid motion is solved based on the Navier-Stokes equations, using porosity... Dynamically modified fluid mesh permeability Darcy-Forchheimer equation, where V particle V is the volume of the grain mesh. cell For fluid mesh volume;

[0093] Step 2.2 Fluid-structure interaction

[0094] Calculate the drag force of the fluid on the particles. Where c d The drag coefficient is 0.44 (when Reynolds number Re > 1000), ρ f For a fluid density of 1000 kg / m³ 3 d p Particle diameter, in meters (m) and um. f The velocity vector of the undisturbed fluid at the center of the particle, in m / s, u p The particle velocity vector is expressed in m / s, |u f -u p | represents the relative velocity between the fluid and the particles, in m / s. (u) in the formula... f -u p This ensures that the direction of the force is consistent with the direction of the relative velocity;

[0095] Perform chemical field coupling:

[0096] Step 2.3 Dissolution-precipitation equation

[0097] The dissolution rate of tailings particles is defined based on the Arrhenius formula. The reaction order n = 1.2, and the solution saturation S is updated in real time via an ion concentration monitoring module. When the Ca in the solution... 2+ When the concentration exceeds 0.1 mol / L, the cementitious material formation algorithm is triggered, enhancing the strength of the interparticle cementation bonds. Where k0 is the reaction rate constant, R is the gas constant, T is the temperature in Kelvin, S is the solution saturation, and E... a =35kJ / mol;

[0098] Step 2.4 Formation of cementitious material

[0099] Based on the dynamic generation of cementitious material according to the solution ion concentration, the tensile strength σ of the interparticle cementation bonds is updated. max and shear strength τ max , unit: kPa.

[0100] Perform temperature field coupling:

[0101] Step 2.5 Heat Conduction Equation

[0102] Calculating interparticle heat conduction using Fourier's law And consider the heat transfer q of fluid convection. conv =h(T) p -T f The effect of ), where q is the heat flux density, in W / m³. 2 k is the thermal conductivity, with units of W / (m·K). For temperature gradient, the unit is K / m, q conv h is the convective heat transfer density, in W / m³, and h is the convective heat transfer coefficient, in W / (m³). 2 ·K), T p The particle surface temperature is expressed in Kelvin (K) and temperature (T). f This represents the local temperature of the fluid, in Kelvin (K).

[0103] Step 2.6 Thermal Expansion Effect

[0104] The particle radius changes with temperature as Δr = r0αΔT, where r0 is the original particle radius in meters (m) and α is the coefficient of thermal expansion (8.56 × 10⁻⁶). -6 K -1 This affects contact force and porosity.

[0105] Step 3: Energy Evolution Analysis:

[0106] Based on the model obtained in steps 1 and 2, extract F from the model under equilibrium conditions. n F t Then, energy classification and energy mutation rate index extraction were performed:

[0107] Step 3.1: Energy Classification and Calculation:

[0108] Elastic energy E e Energy stored in particle contact deformation

[0109] Where, δ n δ t These represent the normal and tangential displacements, respectively, in meters (m).

[0110] Dissipated energy E d Energy dissipated by friction and viscosity, E d =∑μF n Δδ t +∑η(Δv) 2 Δt+∑βγ 2 V cell Δt;

[0111] Where μ: interparticle friction coefficient; Δδ t : Tangential displacement increment, in meters; η: Fluid dynamic viscosity, in Pa·s; Δv: Fluid-particle relative velocity, in meters per second; Δt: Time change, in seconds; β: Viscous dissipation coefficient, β = 2η; γ: Local shear rate, in seconds. -1 V cell Fluid mesh volume, in meters (m). 3 .

[0112] Kinetic energy E k Kinetic energy of particle motion

[0113] Where, m p The mass of the particles is expressed in kg; v p Particle velocity, in m / s;

[0114] Step 3.2: Energy Mutation Feature Extraction:

[0115] Energy mutation rate index: defined as the ratio of the rate of accumulation of elastic energy to the rate of release of dissipated energy. critical value δP ext For external input power, when R total >R crit The time is determined to be energy instability, and R is set to... total >R crit As one of the early warning criteria;

[0116] Energy entropy increase index: Calculate the energy distribution entropy of the system S = -k B ∑p i ln p i , where kB Let be the Boltzmann constant, taken as 1.38 × 10⁻⁶. -23 J / K, p i The critical threshold for the rate of increase of energy entropy, representing the energy proportion of each sub-region. Where λ is the damping coefficient, τ relax The energy relaxation time for the discrete element simulation fitting, in seconds. The kinetic energy gradient is expressed as the rate of increase of entropy. As one of the early warning criteria;

[0117] Step 4: Early Warning Decision: Based on the characteristics of energy mutation, output tiered early warning signals:

[0118] Normal: R total <2.0 and

[0119] Level 3 warning: 2.0≤R total <2.5 or

[0120] Level 2 Warning: R total ≥2.5 or

[0121] Level 1 Warning: R total ≥3.75 or kinetic energy E k The surge exceeded the historical peak by more than 20% and lasted for more than 10 minutes.

[0122] Among them, the level 1 warning is more dangerous than the level 2 warning, and the level 2 warning is more dangerous than the level 3 warning.

[0123] The calculation process in the above method is optimized:

[0124] GPU acceleration strategy: Distribute particle contact force calculation, fluid solution and energy statistics tasks to GPU cores, and use CUDA architecture to realize real-time simulation of the 100 million particle model (1 hour of physical process simulation time ≤ 10 minutes);

[0125] The region decomposition method divides the tailings dam model into several subdomains, and each subdomain is calculated independently and then the boundary data is synchronized through the MPI protocol.

[0126] Adaptive time step dynamic adjustment: The time step Δt is dynamically adjusted according to the energy evolution rate |dE / dt|.

[0127]

[0128] Where Δt new To adjust the step size, Δt old To adjust the time step size, ∈ represents the allowable energy error threshold of 1 kJ, and the time step range is 10. -7s~10 -5 s.

[0129] This embodiment accurately simulates the internal deterioration process of tailings dams through four-field coupling. Combining the dual criteria of energy entropy increase and mutation rate, the early warning response time is 4 to 5 hours earlier than the traditional stress monitoring method, and the false alarm rate is reduced to below 6%.

[0130] The foregoing descriptions have outlined some exemplary embodiments of the present invention. It is understood that these embodiments are merely illustrative and do not constitute a limitation on the scope of protection of the present invention. Features in these embodiments can be rearranged in suitable ways, and the resulting solutions remain within the scope of protection claimed by the present invention. All other embodiments obtained by those skilled in the art based on the foregoing embodiments without inventive effort, i.e., all modifications, equivalent substitutions, and improvements made within the spirit and principles of this application, fall within the scope of protection claimed by the present invention.

Claims

1. A multi-field coupled discrete-element disaster early warning method for tailings dams based on energy evolution criteria, characterized by modeling in simulation software, wherein... Includes the following steps: Step 1: Constructing the Discrete Element Model: Establish a discontinuous medium model of the tailings particle accumulation, and introduce a cementation model to calculate the interparticle mechanical behavior: interparticle normal force F n and tangential force F t The unit is N; Step 2: Multi-field coupling calculation: Based on the discrete element model obtained in Step 1, perform seepage field-stress field coupling, chemical field coupling, and temperature field coupling; Step 3: Energy Evolution Analysis: Based on the model obtained in Steps 1 and 2, extract F from the model under equilibrium conditions. n F t Then, energy classification and energy mutation rate index extraction were performed: Step 3.1: Energy Classification and Calculation: Elastic energy E e Energy stored in particle contact deformation Where, δ n δ t These represent the normal and tangential displacements, respectively, in meters (m). Dissipated energy E d Energy dissipated by friction and viscosity, E d =∑μF n Δδ t +∑η(Δv) 2 Δt+∑βγ 2 V cell Δt; Where μ: interparticle friction coefficient; Δδ t : Tangential displacement increment, in meters; η: Fluid dynamic viscosity, in Pa·s; Δv: Fluid-particle relative velocity, in meters per second; Δt: Time change, in seconds; β: Viscous dissipation coefficient, β = 2η; γ: Local shear rate, in seconds. -1 V cell Fluid mesh volume, in meters (m). 3 ; Kinetic energy E k Kinetic energy of particle motion Where, m p The mass of the particles is expressed in kg; v p Particle velocity, in m / s; Step 3.2: Energy Mutation Feature Extraction: Energy mutation rate index: defined as the ratio of the rate of accumulation of elastic energy to the rate of release of dissipated energy. critical value δP ext For external input power, when R total >R crit The time is determined to be energy instability, and R is set to... total >R crit As one of the early warning criteria; Energy entropy increase index: Calculate the energy distribution entropy of the system S = -k B ∑p i lnp i , where k B Let be the Boltzmann constant, taken as 1.38 × 10⁻⁶. -23 J / K, p i The critical threshold for the rate of increase of energy entropy, representing the energy proportion of each sub-region. Where λ is the damping coefficient, τ relax The energy relaxation time for the discrete element simulation fitting, in seconds. The kinetic energy gradient is expressed as the rate of increase of entropy. As one of the early warning criteria; Step 4: Early Warning Decision: Based on the characteristics of energy mutation, output tiered early warning signals: Normal: R total <0.8R crit and Level 3 warning: 0.8R crit ≤R total <R crit or Level 2 Warning: R total ≥R crit or Level 1 Warning: R total ≥1.5R crit or kinetic energy E k The surge exceeded the historical peak by more than 20% and lasted for more than 10 minutes. Among them, the level 1 warning is more dangerous than the level 2 warning, and the level 2 warning is more dangerous than the level 3 warning.

2. The method for early warning of multi-field coupled discrete-element disasters in tailings dams based on energy evolution criteria according to claim 1, characterized in that, Step 1, the construction of the discrete element model, specifically includes the following steps: Step 1.1 Particle size classification configuration The particle size distribution and percentage of particles were determined by indoor sieving experiments and laser particle size analyzer. The tailings gradation curve was plotted, and the particle size range was determined based on the tailings gradation curve. Step 1.2 Particle Template Generation To determine the particle size range, the hyperellipsoid equation is used. Generate a non-spherical particle template, where n controls the sharpness of the edges and corners to simulate the actual shape distribution of tailings particles; Step 1.3 Input particle properties Input the tailings particle parameters measured in the laboratory, including density ρ, elastic modulus E, thermal expansion coefficient α and thermal conductivity k, introduce the cementation model, and finally obtain the discontinuous medium model of the tailings particle accumulation. Step 1.4 Calculate the interparticle normal force F based on the discontinuous medium model of tailings particle accumulation. n and tangential force F t .

3. The tailings dam multi-field coupled discrete-element disaster early warning method based on energy evolution criterion as described in claim 2, characterized in that, Step 1.4 in step 1 uses the Hertz-Mindlin contact theory to calculate the interparticle normal force. Calculate the tangential force based on the Mindlin-Deresiewicz model. In the formula: E * Equivalent elastic modulus, in Pa, R * The equivalent radius is in meters (m), μ is the interparticle friction coefficient, and δ is the effective radius. n ,δ t These are the normal and tangential displacements, respectively.

4. The tailings dam multi-field coupled discrete-element disaster early warning method based on energy evolution criterion as described in claim 1, characterized in that, The bonding model introduced in step 1, where the tensile strength σ of the bonding bond is input. max and shear strength τ max The unit is kPa, which simulates the formation and destruction process of cementitious materials between tailings particles.

5. The method for early warning of multi-field coupled discrete-element disasters in tailings dams based on energy evolution criteria according to claim 1, characterized in that, In the seepage field-stress field coupling in step 2, based on the discrete element model in step 1, the following steps are performed: Step 2.1 Pore fluid simulation Fluid motion is solved based on the Navier-Stokes equations, using porosity... Dynamically modified fluid mesh permeability Darcy-Forchheimer equation, where V particle V is the volume of the grain mesh. cell For fluid mesh volume; Step 2.2 Fluid-structure interaction Calculate the drag force of the fluid on the particles. Where C d ρ is the drag coefficient. f Fluid density, unit: kg / m³ 3 d p Particle diameter, in meters (m) and um. f The velocity vector of the undisturbed fluid at the center of the particle, in m / s, u p The particle velocity vector is expressed in m / s, |u f -u p | represents the relative velocity between the fluid and the particles, in m / s. (u) in the formula... f -u p This ensures that the direction of the force is consistent with the direction of the relative velocity.

6. The tailings dam multi-field coupled discrete-element disaster early warning method based on energy evolution criterion according to claim 5, characterized in that, In the chemical field coupling of step 2, based on the discrete element model of step 1, the following steps are performed: Step 2.3 Dissolution-precipitation equation The dissolution rate of tailings particles is defined based on the Arrhenius formula. Where k0 is the reaction rate constant, R is the gas constant, T is the temperature in K, S is the solution saturation, and E is the gas saturation. a The activation energy is expressed in kJ / mol, and n is the reaction order. Step 2.4 Formation of cementitious material Based on the dynamic generation of cementitious material according to the solution ion concentration, the tensile strength σ of the interparticle cementation bonds is updated. max and shear strength τ max , unit: kPa.

7. A tailings dam multi-field coupled discrete-element disaster early warning method based on energy evolution criterion as described in claim 6, characterized in that, In the temperature field coupling of step 2, based on the discrete element model of step 1, the following steps are performed: Step 2.5 Heat Conduction Equation Calculating interparticle heat conduction using Fourier's law And consider the heat transfer q of fluid convection. conv =h(T) p -T f The effect of ), where q is the heat flux density, in W / m³. 2 k is the thermal conductivity, with units of W / (m·K). For temperature gradient, the unit is K / m, q conv h is the convective heat transfer density, in W / m³, and h is the convective heat transfer coefficient, in W / (m³). 2 ·K), T p The particle surface temperature is expressed in Kelvin (K) and temperature (T). f The local temperature of the fluid, in Kelvin (K). Step 2.6 Thermal Expansion Effect The particle radius changes with temperature Δr=r0αΔT, where r0 is the original particle radius in meters; α is the coefficient of thermal expansion, which affects the contact force and porosity.

8. The method for early warning of multi-field coupled discrete-element disasters in tailings dams based on energy evolution criteria according to claim 1, characterized in that, The computational process in the method is optimized by allocating particle contact force calculation, fluid solution, and energy statistics tasks to GPU cores, and using the CUDA architecture to achieve real-time simulation of a model with hundreds of millions of particles. The tailings dam model is divided into several subdomains, each calculated independently, and boundary data is synchronized via the MPI protocol. The time step Δt is dynamically adjusted according to the energy evolution rate |dE / dt|. Where Δt new To adjust the step size, Δt old The step size before adjustment is ∈, where ∈ is the allowable energy error threshold.

9. A method for early warning of multi-field coupled discrete-element disasters in tailings dams based on energy evolution criteria, as described in claim 1, is characterized in that... In step 4, the early warning decision is made by fusing energy mutation characteristics with monitoring data to output a graded early warning signal. The monitoring data is monitored by sensors.

10. A method for early warning of multi-field coupled discrete-element disasters in tailings dams based on energy evolution criteria, as described in claim 1, is characterized in that... The simulation software used is COMSOL or PFC.

Images