Method for predicting in-situ leaching mining effect of sandstone type uranium ore after blasting

Abstract

The invention discloses a post-blasting sandstone type uranium mine in-situ leaching mining effect prediction method, and relates to the field of sandstone type uranium mine in-situ leaching mining effect numerical simulation, and the method comprises the steps: building an HJC rock blasting constitutive model based on a finite element analysis platform, and carrying out the blasting numerical simulation; calibrating the blasting crack damage nephogram through a Kriging interpolation algorithm; a COMSOL multi-physical field model is constructed; constructing a coupling interface of the COMSOL multi-physical field model and the PHREEQC based on MATLAB, and optimizing the cycle time and the time step length; performing geochemical reaction calculation and storage, running a COMSOL file at the same time, and inputting a geochemical reaction calculation result into PHREEQC to obtain dynamic loop iteration simulation; and through integration of the LSTM neural network, dynamically correcting the migration coefficient of the uranyl complex, carrying out circulation termination, and outputting a uranium migration multi-scale dynamic simulation result. According to the method, multi-scale dynamic simulation of uranium migration in the leaching agent injection process is realized, and the uranium migration prediction precision is improved.

Description

Technical Field

[0001] The invention relates to the field of numerical simulation of in-situ leaching mining effect of sandstone type uranium mines, in particular to a method for predicting in-situ leaching mining effect of sandstone type uranium mines after blasting. Background Art

[0002] Sandstone-type uranium deposits are an important global uranium resource, and their green and efficient development is a key component of the nuclear fuel cycle. CO2+O2 in-situ leaching technology, which injects leaching solutions to dissolve uranium ore bodies and recover uranium-containing solutions, offers significant advantages, including high resource utilization and minimal environmental disturbance. However, traditional acid in-situ leaching suffers from issues such as the high corrosiveness of leaching agents and high environmental risks, hindering its sustainable development. Neutral in-situ leaching technology based on CO2+O2 has become a research hotspot in the field of uranium mining due to its wide leaching range and environmental friendliness. However, the multi-field coupling and solute transfer mechanisms of CO2+O2 in-situ leaching of sandstone-type uranium deposits after blasting are complex, and systematic research is urgently needed.

[0003] As a pretreatment process for uranium mining, blasting releases energy from explosives to create a network of fractures in the ore body, changing the permeability and mechanical properties of the rock mass. However, the rock fragmentation, stress redistribution, and temperature field changes caused by the blasting load couple with the seepage field and chemical field after the injection of the leaching agent, making it difficult for traditional single physical field simulation methods to accurately reveal the laws of uranium migration. In addition, the CO2+O2 system involves complex multiphase chemical reactions (such as carbonate dissolution, oxidation-dissolution of uranium oxides, etc.) during the leaching process. The dynamic chemical equilibrium and the spatiotemporal evolution characteristics of solute migration have a decisive influence on the leaching efficiency and leaching agent optimization.

[0004] While researchers at home and abroad have conducted extensive research on the multi-field coupling of in-situ leaching, the mechanisms by which blasting damage influences the multiscale pore structure of rocks remain unclear, and a model for the relationship between blasting parameters and energy evolution is lacking. Methods for obtaining multiphase chemical reaction kinetic parameters in CO₂+O₂ leaching systems also have limitations, making it difficult to accurately describe the dynamic equilibrium of mineral dissolution and precipitation, and to characterize the coevolution of the macroscopic seepage field and microscopic pore structure. Summary of the Invention

[0005] The present invention provides a method for predicting the effect of in-situ leaching mining of sandstone-type uranium deposits after blasting, which solves the problem in the current technology that it is difficult to couple the commonly used software COMSOL in the industry with the chemical software PHREEQC to simulate the multi-field coupling and solute transfer of CO2+O2 in-situ leaching mining of sandstone-type uranium deposits. It realizes the multi-scale dynamic simulation of uranium migration and the simulation of multiple situations such as multi-physical field coupling and geochemical reactions under the solute migration of CO2+O2 in-situ leaching, thereby improving the accuracy of uranium migration prediction.

[0006] The present invention provides a method for predicting the in-situ leaching mining effect of a sandstone-type uranium ore after blasting, comprising the following steps:

[0007] S1. Based on the finite element analysis platform, a HJC (Holmquist-Johnson-Cook, a constitutive model used to describe the mechanical behavior of rock under impact and high pressure) rock blasting constitutive model was constructed. Then, numerical simulation of blasting was performed to obtain blasting crack damage cloud maps.

[0008] S2. Calibrate the blasting crack damage cloud map using the Kriging interpolation algorithm based on the ground stress field data measured by the distributed fiber optic acoustic sensor to obtain a calibrated blasting crack damage cloud map;

[0009] S3. Based on the calibrated blasting crack damage cloud map and COMSOL Multiphysics (a multi-physics coupling simulation platform based on the finite element method that supports modeling, simulation, and optimization analysis of complex multidisciplinary systems in engineering and science), an original COMSOL multi-physics model is constructed. The redox potential and pH of the leaching solution monitored by the electrochemical sensor are loaded as boundary conditions into the original COMSOL multi-physics model to obtain the COMSOL multi-physics model.

[0010] S4. Based on MATLAB, a coupling interface was constructed between the COMSOL (a multi-physics coupling simulation platform based on the finite element method that supports modeling, simulation, and optimization analysis of complex multidisciplinary systems in engineering and science) multi-physics model and PHREEQC (a professional hydrogeochemical simulation software used to study complex geochemical reactions, mineral dissolution and precipitation, and quantitative calculation and analysis of multi-component ion balance in aqueous systems). The optimal time step during the seepage-dissolution coupling calculation phase within the cycle time was determined, and the optimal time step was input into the COMSOL multi-physics model to obtain the COMSOL initial concentration.

[0011] S5. Map the COMSOL initial concentration to PHREEQC, and use MATLAB to call PHREEQC to perform geochemical reaction calculations based on the equilibrium constraints of the CO2+O2-water-rock reaction to obtain the geochemical reaction calculation results.

[0012] S6. Save the geochemical reaction calculation results, run the COMSOL file at the same time, and re-input the geochemical reaction calculation results into PHREEQC as the initial value of the next time step to obtain a dynamic cyclic iterative simulation;

[0013] S7, by integrating the parameter adaptive matching function driven by the LSTM neural network, dynamically correcting the uranyl complex migration coefficient during the dynamic cyclic iterative simulation process to obtain the corrected uranyl complex migration coefficient;

[0014] S8. Repeat S4-S7 until the cycle time is reached and terminate the calculation, and output the multi-scale dynamic simulation results of uranium migration.

[0015] Preferably, in S1, based on the finite element analysis platform, an HJC rock blasting constitutive model is constructed, and then a blasting numerical simulation is performed to obtain the specific contents of the blasting crack damage cloud map including:

[0016] In the finite element analysis platform, the HJC rock blasting constitutive model is constructed by defining the unit material type and model size; the unit material types include: rock, air and explosives;

[0017] Define ground stress boundary conditions, dynamic compressive strength of rock, JWL equation of state for explosives, and air material properties;

[0018] In-situ stress boundary conditions are applied to the HJC rock blasting constitutive model. Based on the dynamic compressive strength of rock, the JWL equation of state of explosives, and the material properties of air, a blasting numerical simulation is performed on the HJC rock blasting constitutive model to obtain a blasting crack damage cloud map. The in-situ stress boundary conditions include vertical constraint boundaries and horizontal non-reflecting boundaries.

[0019] Preferably, in S2, the blasting crack damage cloud map is calibrated using a Kriging interpolation algorithm based on the ground stress field data measured by distributed optical fiber acoustic wave sensing, and the specific content of the calibrated blasting crack damage cloud map includes:

[0020] Based on the ground stress field data measured by distributed fiber optic acoustic sensing, the spatial distribution characteristics of the stress field are obtained, and the dynamic damage evolution input conditions are established;

[0021] According to the spatial distribution characteristics of the stress field and the dynamic damage evolution input conditions, the Kriging interpolation algorithm is used to calibrate the anisotropic deformation and crack propagation path of the damaged area in the blasting crack damage cloud map, and a two-dimensional corrected regional damage model is obtained, and a calibrated blasting crack damage cloud map is obtained.

[0022] Preferably, the COMSOL multi-physics model in S3 is a coupled numerical model of the porous media dilute species transport equation, the Darcy seepage equation, the mineral dissolution domain ordinary differential equations, and the uranyl complex reaction-diffusion equation.

[0023] Preferably, in S4, a coupling interface between the COMSOL multi-physics model and PHREEQC is constructed based on MATLAB to determine the optimal time step during the seepage-dissolution coupling calculation phase within the cycle time, and the optimal time step is input into the COMSOL multi-physics model to obtain the COMSOL initial concentration, specifically including:

[0024] Construct a coupling interface between COMSOL multi-physics model and PHREEQC based on MATLAB;

[0025] The cycle time is optimized by leaching efficiency to obtain the optimized cycle time;

[0026] Based on the evolution characteristics of the uranium concentration gradient, a time step adaptive control algorithm is set to obtain the optimal time step for the seepage-dissolution coupling calculation stage within the optimization cycle time.

[0027] The optimal time step is input into the COMSOL multiphysics model to obtain the COMSOL initial concentration.

[0028] Preferably, the geochemical reaction calculation results in S5 include: liquid phase element concentrations in pores and secondary mineral precipitation amount data sets.

[0029] Preferably, S7 dynamically corrects the uranyl complex migration coefficient during the dynamic cyclic iterative simulation process by integrating a parameter adaptive matching function driven by an LSTM neural network to obtain a corrected uranyl complex migration coefficient, specifically including:

[0030] The redox potential and pH of the leaching solution monitored by the electrochemical sensor and the initial uranyl complex migration coefficient calculated by the COMSOL multi-physics model were input into the LSTM neural network to predict the dynamic correction factor and obtain the dynamic correction factor.

[0031] According to the dynamic correction factor, the uranyl complex migration coefficient is automatically adjusted by integrating the parameter adaptive matching function driven by the LSTM neural network to obtain the corrected uranyl complex migration coefficient.

[0032] Preferably, the multi-scale dynamic simulation results of uranium migration in S8 include: uranium concentration distribution data and uranium migration path prediction.

[0033] A prediction system for in-situ leaching mining effects of sandstone-type uranium deposits after blasting, comprising:

[0034] The first model building module is used to build the HJC rock blasting constitutive model based on the finite element analysis platform, and then perform blasting numerical simulation to obtain the blasting crack damage cloud map;

[0035] A calibration module is used to calibrate the blasting crack damage cloud map using the Kriging interpolation algorithm based on the ground stress field data measured by the distributed optical fiber acoustic sensor to obtain a calibrated blasting crack damage cloud map;

[0036] The second model building module is used to build an original COMSOL multiphysics model based on the calibrated blasting crack damage cloud map and COMSOL Multiphysics software, and load the redox potential and pH of the leaching solution monitored by the electrochemical sensor as boundary conditions into the original COMSOL multiphysics model to obtain the COMSOL multiphysics model;

[0037] A determination module is used to build a coupling interface between the COMSOL multiphysics model and PHREEQC based on MATLAB, determine the optimal time step during the seepage-dissolution coupling calculation phase within the cycle time, and input the optimal time step into the COMSOL multiphysics model to obtain the COMSOL initial concentration;

[0038] The calculation module is used to map the COMSOL initial concentration to PHREEQC and use MATLAB to call PHREEQC to perform geochemical reaction calculations based on the equilibrium constraints of the CO2+O2-water-rock reaction to obtain the geochemical reaction calculation results;

[0039] Iterative simulation module is used to save the geochemical reaction calculation results, run the COMSOL file at the same time, and re-input the geochemical reaction calculation results into PHREEQC as the initial value of the next time step to obtain a dynamic cyclic iterative simulation;

[0040] A correction module is used to dynamically correct the uranyl complex migration coefficient during the dynamic cyclic iterative simulation process by integrating the parameter adaptive matching function driven by the LSTM neural network to obtain the corrected uranyl complex migration coefficient;

[0041] The output module is used to repeat the determination module, calculation module, iterative simulation module and prediction module until the cycle time is reached and the calculation is terminated, and the multi-scale dynamic simulation results of uranium migration are output.

[0042] An electronic device, characterized in that it includes a memory and a processor, wherein a computer program is stored in the memory, and when the processor calls the computer program in the memory, the content of the above-mentioned method for predicting the in-situ leaching mining effect of sandstone-type uranium deposits after blasting is implemented.

[0043] In summary, the method for predicting the in-situ leaching effect of sandstone-type uranium deposits after blasting of the present invention has the following advantages over conventional technologies:

[0044] (1) The present invention solves the problem that it is difficult to couple the commonly used software COMSOL in the industry with the chemical software PHREEQC to simulate the multi-field coupling-solute transfer of CO2+O2 in situ leaching of sandstone-type uranium deposits.

[0045] (2) The present invention establishes a seepage-stress-chemistry coupling model of blasting-damaged rock mass through the COMSOL multi-physics field model, combines PHREEQC to perform geochemical reaction calculations, realizes multi-scale dynamic simulation of uranium migration during the injection of solvent, and improves the prediction accuracy of uranium migration.

[0046] (3) The present invention realizes the simulation of multiple physical field coupling, geochemical reactions and other situations under the in situ CO2+O2 leaching solute migration based on the method of MATLAB coupling COMSOL and PHREEQC.

[0047] (4) The present invention has the advantages of small error and strong scalability, which greatly improves the applicability of the reactive solute migration model and increases the possibility of secondary development of COMSOL-PHREEQC.

[0048] The technical method of the present invention is further described in detail below through the accompanying drawings and examples. BRIEF DESCRIPTION OF THE DRAWINGS

[0049] Figure 1 A schematic flow chart of a method for predicting the effect of in-situ leaching mining of sandstone-type uranium deposits after blasting provided by an embodiment of the present invention;

[0050] Figure 2 A schematic diagram of numerical simulation of blasting permeability enhancement under initial ground stress provided by an embodiment of the present invention;

[0051] Figure 3 A schematic diagram of monitoring points for a solute transport model provided in an embodiment of the present invention;

[0052] Figure 4 A schematic diagram of CaCO3 precipitation changes at a monitoring point provided by an embodiment of the present invention;

[0053] Figure 5 A 900-day uranium concentration in-situ leaching cloud map using COMSOL-PHREEQC coupling provided in an embodiment of the present invention;

[0054] Figure 6 This is the uranium concentration variation curve under COMSOL-PHREEQC coupling provided by an embodiment of the present invention. DETAILED DESCRIPTION

[0055] The technical method of the present invention is further described below through the accompanying drawings and embodiments. It should be noted that unless otherwise specifically stated, the relative arrangement of components and steps, numerical expressions and values ​​described in these embodiments do not limit the scope of this application.

[0056] The following description of at least one exemplary embodiment is merely illustrative in nature and is in no way intended to limit the present disclosure, its application, or uses.

[0057] Technologies, systems, and devices known to those of ordinary skill in the relevant art may not be discussed in detail, but where appropriate, they should be considered part of the specification.

[0058] In all examples shown and discussed herein, any specific values ​​should be interpreted as merely exemplary and not limiting. Therefore, other examples of the exemplary embodiments may have different values.

[0059] Unless otherwise defined, technical or scientific terms used in the present invention shall have the same meaning as commonly understood by one of ordinary skill in the art to which the present invention belongs.

[0060] The present invention provides a method for predicting the in-situ leaching mining effect of a sandstone-type uranium ore after blasting, comprising the following steps:

[0061] S1. Based on the finite element analysis platform, the HJC rock blasting constitutive model was constructed, and then the blasting numerical simulation was performed to obtain the blasting crack damage cloud map;

[0062] It can be understood that the finite element analysis platform can be an LS-DYNA finite element analysis platform to construct the HJC rock blasting constitutive model, assign the isotropic boundary condition of the initial ground stress σx=σy=5MPa, define the dynamic compressive strength of the rock, the JWL state equation of the explosive and the air material properties by writing the K file, perform the numerical simulation of blasting under ground stress conditions, and obtain the blasting crack damage cloud map.

[0063] Furthermore, in S1, based on the finite element analysis platform, the HJC rock blasting constitutive model was constructed, and then the blasting numerical simulation was performed to obtain the blasting crack damage cloud map. The specific contents include:

[0064] In the finite element analysis platform, the HJC rock blasting constitutive model is constructed by defining the unit material type and model size; the unit material types include: rock, air and explosives;

[0065] Define ground stress boundary conditions, dynamic compressive strength of rock, JWL equation of state for explosives, and air material properties;

[0066] In-situ stress boundary conditions are applied to the HJC rock blasting constitutive model. Based on the dynamic compressive strength of rock, the JWL equation of state of explosives, and the material properties of air, a blasting numerical simulation is performed on the HJC rock blasting constitutive model to obtain a blasting crack damage cloud map. The in-situ stress boundary conditions include vertical constraint boundaries and horizontal non-reflecting boundaries.

[0067] S2. Calibrate the blasting crack damage cloud map using the Kriging interpolation algorithm based on the ground stress field data measured by the distributed fiber optic acoustic sensor to obtain a calibrated blasting crack damage cloud map;

[0068] Furthermore, in S2, the blasting crack damage cloud map is calibrated using the Kriging interpolation algorithm based on the ground stress field data measured by distributed fiber optic acoustic sensing. The specific contents of the calibrated blasting crack damage cloud map include:

[0069] Based on the ground stress field data measured by distributed fiber optic acoustic sensing, the spatial distribution characteristics of the stress field are obtained, and the dynamic damage evolution input conditions are established;

[0070] According to the spatial distribution characteristics of the stress field and the dynamic damage evolution input conditions, the Kriging interpolation algorithm is used to calibrate the anisotropic deformation and crack propagation path of the damaged area in the blasting crack damage cloud map, and a two-dimensional corrected regional damage model is obtained, and a calibrated blasting crack damage cloud map is obtained.

[0071] S3. Based on the calibrated blasting crack damage cloud map and COMSOL Multiphysics software, an original COMSOL multiphysics model is constructed, and the redox potential and pH of the leaching solution monitored by the electrochemical sensor are loaded into the original COMSOL multiphysics model as boundary conditions to obtain the COMSOL multiphysics model;

[0072] It is understandable that the blasting damage cloud map output by LS-DYNA is imported into COMSOL Multiphysics using the R2V format conversion plug-in, a two-dimensional geological model of 105m long and 70m wide is established, the Eh / pH dynamic boundary conditions with real-time feedback from downhole electrochemical sensors are embedded, the input settings of material parameters and multi-physics field models are completed, and the transient solver is configured.

[0073] Furthermore, the COMSOL multiphysics model in S3 is a coupled numerical model of the porous media dilute species transport equation, Darcy's seepage equation, the mineral dissolution domain ordinary differential equations, and the uranyl complex reaction-diffusion equation.

[0074] S4. Build a coupling interface between the COMSOL multiphysics model and PHREEQC based on MATLAB to determine the optimal time step during the seepage-dissolution coupling calculation phase within the cycle time, and input the optimal time step into the COMSOL multiphysics model to obtain the COMSOL initial concentration.

[0075] Furthermore, S4 built a coupling interface between the COMSOL multiphysics model and PHREEQC based on MATLAB to determine the optimal time step during the seepage-dissolution coupling calculation phase within the cycle time. The optimal time step was then input into the COMSOL multiphysics model to obtain the COMSOL initial concentration, specifically including:

[0076] Construct a coupling interface between COMSOL multi-physics model and PHREEQC based on MATLAB;

[0077] The cycle time is optimized by leaching efficiency to obtain the optimized cycle time;

[0078] Based on the evolution characteristics of the uranium concentration gradient, a time step adaptive control algorithm is set to obtain the optimal time step for the seepage-dissolution coupling calculation stage within the optimization cycle time.

[0079] The optimal time step is input into the COMSOL multiphysics model to obtain the COMSOL initial concentration.

[0080] S5. Map the COMSOL initial concentration to PHREEQC, and use MATLAB to call PHREEQC to perform geochemical reaction calculations based on the equilibrium constraints of the CO2+O2-water-rock reaction to obtain the geochemical reaction calculation results.

[0081] Furthermore, the geochemical reaction calculation results in S5 include: liquid phase element concentrations in pores and secondary mineral precipitation data sets.

[0082] S6. Save the geochemical reaction calculation results, run the COMSOL file at the same time, and re-input the geochemical reaction calculation results into PHREEQC as the initial value of the next time step to obtain a dynamic cyclic iterative simulation;

[0083] S7, by integrating the parameter adaptive matching function driven by the LSTM neural network, dynamically correcting the uranyl complex migration coefficient during the dynamic cyclic iterative simulation process to obtain the corrected uranyl complex migration coefficient;

[0084] Furthermore, S7 integrates the parameter adaptive matching function driven by the LSTM neural network to dynamically correct the uranyl complex migration coefficient during the dynamic cyclic iterative simulation process to obtain the corrected uranyl complex migration coefficient, specifically including:

[0085] The redox potential and pH of the leaching solution monitored by the electrochemical sensor and the initial uranyl complex migration coefficient calculated by the COMSOL multi-physics model were input into the LSTM neural network to predict the dynamic correction factor and obtain the dynamic correction factor.

[0086] According to the dynamic correction factor, the uranyl complex migration coefficient is automatically adjusted by integrating the parameter adaptive matching function driven by the LSTM neural network to obtain the corrected uranyl complex migration coefficient.

[0087] S8. Repeat S4-S7 until the cycle time is reached and terminate the calculation, and output the multi-scale dynamic simulation results of uranium migration.

[0088] Furthermore, the results of the multi-scale dynamic simulation of uranium migration in S8 include: uranium concentration distribution data and uranium migration path prediction.

[0089] Example

[0090] An embodiment of the present invention provides a method for predicting the effect of CO2+O2 in-situ leaching mining of sandstone-type uranium deposits after blasting. The method generates a blasting crack damage cloud map based on ANSYS / LS-DYNA, integrates DAS monitoring data with the Kriging inversion model to calibrate the fracture network parameters, develops a COMSOL-PHREEQC multi-physics field-chemical field coupling interface based on MATLAB, and combines field working condition information to construct a blasting crack network-seepage-chemical solute migration multi-scale coupling model. The method analyzes the regulation mechanism of CO2+O2-water-rock reaction on uranium ore dissolution and uranyl ion migration path under blasting damage conditions, and reveals the evolution law of solute transfer under the dynamic coupling of stress field-seepage field-chemical field, such as Figure 1 As shown, the details are as follows:

[0091] Step S1: Based on ANSYS software or LS-DYNA software, a HJC rock blasting constitutive model (HJC) is constructed, i.e., a two-dimensional blasting model, and meshing is performed. Then, the vertical constraint boundary σ is defined. y and the transverse non-reflective boundary σ x , and impose σ on the model x =σ y =5MPa ground stress boundary condition, and use Visual Studio Code to modify the parameters in the k file to obtain the blasting crack damage cloud map under ground stress.

[0092] Step S1 is to define the element material type in LS-DYNA software in "Preprocessor" → "Element Type / MaterialProps" → "Material Models" and select "3D Solid 164"; then define the size of the HJC rock blasting constitutive model in the "Modeling" module, which is 105m long and 70m wide. The air diameter is R 气 =1m, the blasthole radius is R 炮 = 0.1m; mesh the rock, air, and explosives separately through "Meshing" → "MeshTool"; set the detonation time "Solution" → "Time Controls" → "Solution Time" → "25000" (detonation time: 25000μs); add the ground stress "Keyword" → "DEFIND" → "CURVE" (σ x=σ y =5MPa), and save the files as "dy1.k" and "dy2.k", such as Figure 2 As shown in the figure; delete the explosive and air material parameters in the "dy1.k" file, and then run the "dy1.k" file in "LS-Run"; dynamically compile the "dy1.k" file to generate a dynain initialization file with stress inheritance characteristics, and import it into the explosive and air model to put the rock mass in a state of ground stress; run the "dy2.k" file in "LS-Run" to obtain the blasting damage cloud map under ground stress.

[0093] Step S2: Integrating the measured ground stress field data from distributed acoustic sensing (DAS), the real shape of the blasting damage area is calibrated by the Kriging interpolation algorithm to obtain a calibrated blasting crack damage cloud map.

[0094] Step S2 specifically captures the dynamic strain signal of the blasting stress wave in real time through distributed fiber optic acoustic sensing (sampling frequency ≥ 10 kHz), combines denoising and Fourier transform to extract the frequency domain energy attenuation characteristics, and obtains the spatial distribution characteristics of the stress field; uses the generalized Hooke's law to convert the stress field into a dynamic stress tensor, establishes an anisotropic variation function model based on the Kriging interpolation algorithm, integrates the numerical simulation of the blasting crack damage cloud map and the stress gradient extreme value data, and uses the Levenberg-Marquardt inversion algorithm to dynamically correct the HJC rock blasting constitutive model parameters and the fracture network geometric parameters (connectivity error ≤ 8%), and cross-validates the spatial matching degree of the calibration model through microseismic monitoring data (F1-score ≥ 0.85), finally realizing the dynamic update of the permeability tensor matrix and the seepage-chemistry coupling boundary conditions to obtain the calibrated blasting crack damage cloud map.

[0095] In step S3, the calibrated blasting crack damage cloud map is imported into COMSOL Multiphysics software using the R2V format conversion plug-in. A two-dimensional geological model with a length of 105 m and a width of 70 m is established, i.e., the original COMSOL multiphysics model. The dynamic boundary conditions of the redox potential (Eh) and pH of the leaching solution fed back in real time by the downhole electrochemical sensor are embedded to obtain the COMSOL multiphysics model. The material parameters and the input settings of the multiphysics model are completed, and the transient solver is configured.

[0096] Step S3 specifically opens the "R2V.exe" program and selects the calibration blasting crack damage cloud map under ground stress; exports the calibration blasting crack damage cloud map as a .dsf file through "Image", "Conversion" and "24-bit RGB, Grayscale", and selects "Auto Vectorization" at the vector; opens the COMSOL software, imports the .dxf geometric crack and sets the well group model in the geometry component part, and establishes a coupled numerical model including the porous medium dilute species transport equation, Darcy seepage equation, mineral dissolution domain ordinary differential equation group and uranyl complex reaction diffusion equation, that is, the original COMSOL multi-physics field model; extracts the dynamic changes of Eh and pH monitored in real time by the electrochemical sensors arranged in the underground tunnels of the uranium mine (sampling interval ≤ 5 seconds) to capture the chemically active signals of the solute migration process; the system loads the Eh and pH values ​​measured by the sensors as boundary conditions into the original COMSOL multi-physics field model to obtain the COMSOL multi-physics field model.

[0097] When pH < 5.5, hexavalent uranium (U(VI)) is converted to uranyl ion UO2 2+ It exists as the main form and is related to H + The interaction promotes the formation of complex ions, and the solubility of U(VI) increases sharply; the configuration convergence tolerance is ≤1e -5 The transient solver.

[0098] Step S4: Build the COMSOL-PHREEQC coupling interface based on MATLAB, set the cycle duration to 900 days, and select the optimal time step of 100 days for the seepage-dissolution coupling calculation phase between COMSOL and PHREEQC. The specific code includes:

[0099] start_day = 0;

[0100] end_day = 900;

[0101] total_days=end_day-start_day;

[0102] dt_days = 100;

[0103] num_steps=total_days / dt_days;

[0104] phreeqc_exe='phreeqc.exe';

[0105] comsol_model = '3336.mph';

[0106] import com.comsol.model.

[0107] import com.comsol.model.util.

[0108] model=mphload(comsol_model).

[0109] Step S5: Map the COMSOL initial concentration to the PHREEQC.pqi file, generate equilibrium constraints based on the CO2+O2-water-rock reaction, and use MATLAB to call PHREEQC to perform geochemical reaction calculations to obtain geochemical reaction calculation results.

[0110] Step S5 specifically moves the obtained COMSOL file to a path recognizable by MATLAB, and calls MATLAB to put the required concentration parameters into the PHREEQC1.pqi file. The specific code includes:

[0111] x_coord = 52.9;

[0112] y_coord=61.6;

[0113] O2_values=mphinterp(model,{'comp1.cO2'},'coord',[x_coord;

[0114] y_coord]);

[0115] O2_kgw=O2_values / 1000;

[0116] save('C:\Users\qifei\.comsol\v63\llmatlab\comsol_O2.mat','O2_kgw');

[0117] load('C:\Users\qifei\.comsol\v63\llmatlab\comsol_O2.mat',O2_kgw');

[0118] template=fileread('C:\Users\qifei\.comsol\v63\llmatlab\Phrqc1.pqi');

[0119] filled_template=strrep(template,'{O2}',num2str(O2_kgw(1),'%.4f'));

[0120] output_pqi=sprintf('C:\\Users\\qifei\\.comsol\\v63\\llmatlab\\Phrqc2.pqi');

[0121] fid=fopen(output_pqi,'w');

[0122] fprintf(fid,'%s',filled_template);

[0123] fclose(fid);

[0124] disp('PHREEQC input file has been generated');

[0125] Based on the equilibrium constraint of CO2+O2-water-rock reaction, the aqueous solution with an initial temperature of 25℃ is set to simulate the dynamic dissolution-precipitation process of the leachate and the mineral; "Monitoring Point 1" and "Monitoring Point 2" are set at the matrix and crack of the COMSOL multi-physics model respectively, as shown in Figure 2. Figure 3 Then call MATLAB to run the PHREEQC.pqi input file for iterative calculation, and finally extract the mineral saturation index (SI) and ion concentration (Ca 2+ 、SO4 2- wait).

[0126] In step S6, the obtained mineral and ion contents are saved to a file, the COMSOL file is run at the same time, and the obtained concentrations are re-entered into the PHREEQC file for data replacement to prepare for the simulation of the next time step and realize dynamic cyclic iterative simulation.

[0127] Step S6 is to extract Ca 2+ Mg 2+ 、CO3 2- and CaCO3 plasma and precipitation concentrations are saved to the file "Ca 2+ _tx", "Mg 2+ _txt", "CO3 2- _txt" and "CaCO3_txt", and draw the curves of ions and precipitation over time, such as Figure 4 shown. Figure 4 The concentration of CaCO3 precipitation has an initial period, an acceleration period and a stable period. The initial period is mainly an acidic environment. The pH reversal of the acceleration period intensifies the reaction. As the reaction proceeds, the ion concentration gradually decreases, causing the CaCO3 precipitation concentration to gradually stabilize in the later period.

[0128] When running the COMSOL file, assuming that the pores are fully saturated, the relevant mathematical models and equations can be derived. The seepage of the leachate in the pores of the mineral layer can refer to Darcy's law:

[0129]

[0130] Where t is time, S is the water storage coefficient of the rock mass, P is the water pressure, u is the seepage velocity of the fluid in the matrix, e is the volume strain of the rock mass, Q is the source and sink term of the seepage, k is the permeability of the porous medium, η is the dynamic viscosity of the fluid, and ρ is the density of the fluid. is the water pressure gradient, g is the acceleration due to gravity, is the height gradient.

[0131] A CO2+O2 in-situ uranium leaching model was constructed based on the convection-diffusion equation, which takes into account the adsorption and diffusion of solutes. The solute transport equation in COMSOL can be expressed as:

[0132]

[0133] Where θ is the liquid volume fraction, ρ b is the volume density, k b is the adsorption isotherm, c i is the concentration of species i in the liquid, u1 is the node velocity obtained from the flow model, D is the diffusion coefficient, S i is the source item.

[0134] Ca generated based on the current time step 2+ Mg 2+ and HCO3 - Ion concentration, use scripts to batch modify the water chemical parameters of PHREEQC input files (*.pqi), automatically replace the initial conditions through the code interface, save the updated results as the initial values ​​of the next time step, and realize dynamic cyclic iterative simulation of multi-stage reactions.

[0135] Step S7: integrating the parameter adaptive matching function driven by the LSTM neural network to dynamically correct the uranyl complex migration coefficient to obtain the corrected uranyl complex migration coefficient.

[0136] Step S7 is specifically based on the historical simulation and measured data to construct a model containing pH, Eh, UO2 2+ Ion concentration, CO3 2- Ion concentration, Ca 2+A time series dataset with characteristics such as ion concentration and temperature was obtained. An LSTM neural network was then used for modeling, with the current water chemical parameters input and the predicted value of the uranyl complex migration coefficient (D_pred) output. In each iterative step of the PHREEQC simulation, the LSTM model was called in real time, and the water chemical data obtained online were input into the model to predict D_pred. After verification by sensitivity analysis, the migration parameters calculated using the traditional empirical formula were replaced. The numerical simulation conditions were dynamically updated through the PHREEQC interface, and the multi-physics field coupling and solute transfer in the uranium migration process were optimized to obtain the corrected uranyl complex migration coefficient.

[0137] Step S8, looping through the COMSOL-PREEQC coupled calculation process, monitoring the uranium ion concentration in the leachate in real time, calculating the instantaneous uranium concentration at the current time node, and finally obtaining the uranium concentration cloud map on the 900th day, as shown in Figure 5 The uranium concentration data obtained in each cycle will be automatically stored in time series, and the data visualization module will eventually generate a curve of uranium concentration change over time, as shown in Figure 2. Figure 6 shown.

[0138] The method provided by the present invention for exploring the multi-field coupling-solute transfer mechanism and law of CO2+O2 in-situ leaching mining of sandstone-type uranium deposits after blasting, constructs a numerical model of blasting under ground stress based on ANSYS / LS-DYNA software, and uses MATLAB to construct a coupling interface between COMSOL and PHREEQC. This method is helpful for analyzing the regulation mechanism of CO2+O2-water-rock reaction on uranium ore dissolution and uranyl ion migration path under blasting damage conditions, and the evolution law of solute transfer under the dynamic coupling of stress field, seepage field and chemical field, and has practical guiding significance.

[0139] The present invention provides a system for predicting the effect of in-situ leaching mining of sandstone-type uranium deposits after blasting, comprising:

[0140] The first model building module is used to construct the HJC rock blasting constitutive model based on the finite element analysis platform, and then perform blasting numerical simulation to obtain the blasting crack damage cloud map.

[0141] The calibration module is used to calibrate the blasting crack damage cloud map based on the ground stress field data measured by distributed optical fiber acoustic wave sensing through the Kriging interpolation algorithm to obtain a calibrated blasting crack damage cloud map.

[0142] The second model building module is used to construct an original COMSOL multi-physics field model based on the calibrated blasting crack damage cloud map and COMSOL Multiphysics software, and load the redox potential and pH of the leaching solution monitored by the electrochemical sensor as boundary conditions into the original COMSOL multi-physics field model to obtain the COMSOL multi-physics field model.

[0143] The determination module is used to build a coupling interface between the COMSOL multi-physics model and PHREEQC based on MATLAB, determine the optimal time step during the seepage-dissolution coupling calculation stage within the cycle time, and input the optimal time step into the COMSOL multi-physics model to obtain the COMSOL initial concentration.

[0144] The calculation module is used to map the COMSOL initial concentration to PHREEQC, and according to the equilibrium constraints of the CO2+O2-water-rock reaction, use MATLAB to call PHREEQC to perform geochemical reaction calculations and obtain the geochemical reaction calculation results.

[0145] The iterative simulation module is used to save the geochemical reaction calculation results, run the COMSOL file at the same time, and re-input the geochemical reaction calculation results into PHREEQC as the initial value of the next time step to obtain a dynamic cyclic iterative simulation.

[0146] The correction module is used to dynamically correct the uranyl complex migration coefficient in the dynamic cyclic iterative simulation process by integrating the parameter adaptive matching function driven by the LSTM neural network to obtain the corrected uranyl complex migration coefficient.

[0147] The output module is used to repeat the determination module, calculation module, iterative simulation module and prediction module until the cycle time is reached and the calculation is terminated, and the multi-scale dynamic simulation results of uranium migration are output.

[0148] The present invention also provides an electronic device comprising a memory and a processor, wherein the memory stores a computer program, and when the processor calls the computer program in the memory, the content of the above-mentioned method for predicting the in-situ leaching mining effect of sandstone-type uranium deposits after blasting is implemented.

[0149] Finally, it should be noted that the above embodiments are only used to illustrate the technical method of the present invention rather than to limit it. Although the present invention has been described in detail with reference to the preferred embodiments, those skilled in the art should understand that they can still modify or replace the technical method of the present invention with equivalents, and these modifications or equivalent replacements cannot cause the modified technical method to deviate from the spirit and scope of the technical method of the present invention.

Claims

1. A method for predicting the effect of in-situ leaching mining of sandstone-type uranium deposits after blasting, characterized in that: The following steps are involved: S1. Based on the finite element analysis platform, the HJC rock blasting constitutive model was constructed, and then the blasting numerical simulation was performed to obtain the blasting crack damage cloud map; S2. Calibrate the blasting crack damage cloud map using the Kriging interpolation algorithm based on the ground stress field data measured by the distributed fiber optic acoustic sensor to obtain a calibrated blasting crack damage cloud map; S3. Based on the calibrated blasting crack damage cloud map and COMSOL Multiphysics software, an original COMSOL multiphysics model is constructed, and the redox potential and pH of the leaching solution monitored by the electrochemical sensor are loaded into the original COMSOL multiphysics model as boundary conditions to obtain the COMSOL multiphysics model; S4. Build a coupling interface between the COMSOL multiphysics model and PHREEQC based on MATLAB to determine the optimal time step during the seepage-dissolution coupling calculation phase within the cycle time, and input the optimal time step into the COMSOL multiphysics model to obtain the COMSOL initial concentration. S5. Map the COMSOL initial concentration to PHREEQC, and use MATLAB to call PHREEQC to perform geochemical reaction calculations based on the equilibrium constraints of the CO2+O2-water-rock reaction to obtain the geochemical reaction calculation results. S6. Save the geochemical reaction calculation results, run the COMSOL file at the same time, and re-input the geochemical reaction calculation results into PHREEQC as the initial value of the next time step to obtain a dynamic cyclic iterative simulation; S7, by integrating the parameter adaptive matching function driven by the LSTM neural network, dynamically correcting the uranyl complex migration coefficient during the dynamic cyclic iterative simulation process to obtain the corrected uranyl complex migration coefficient; S8. Repeat S4-S7 until the cycle time is reached and terminate the calculation, and output the multi-scale dynamic simulation results of uranium migration.

2. The method for predicting the effect of in-situ leaching mining of sandstone-type uranium deposits after blasting according to claim 1, characterized in that: In S1, based on the finite element analysis platform, the HJC rock blasting constitutive model was constructed, and then a blasting numerical simulation was performed to obtain the blasting crack damage cloud map. The specific contents include: In the finite element analysis platform, the HJC rock blasting constitutive model is constructed by defining the unit material type and model size; the unit material types include: rock, air and explosive; Define ground stress boundary conditions, dynamic compressive strength of rock, JWL equation of state for explosives, and air material properties; In-situ stress boundary conditions are applied to the HJC rock blasting constitutive model. Based on the dynamic compressive strength of rock, the JWL equation of state of explosives, and the material properties of air, a blasting numerical simulation is performed on the HJC rock blasting constitutive model to obtain a blasting crack damage cloud map. The in-situ stress boundary conditions include vertical constraint boundaries and horizontal non-reflecting boundaries.

3. The method for predicting the effect of in-situ leaching mining of sandstone-type uranium deposits after blasting according to claim 1, characterized in that: In S2, the blasting crack damage cloud map is calibrated using the Kriging interpolation algorithm based on the ground stress field data measured by distributed fiber optic acoustic sensors. The specific contents of the calibrated blasting crack damage cloud map include: Based on the ground stress field data measured by distributed fiber optic acoustic sensing, the spatial distribution characteristics of the stress field are obtained, and the input conditions for dynamic damage evolution are established; According to the spatial distribution characteristics of the stress field and the dynamic damage evolution input conditions, the Kriging interpolation algorithm is used to calibrate the anisotropic deformation and crack propagation path of the damaged area in the blasting crack damage cloud map, and a two-dimensional corrected regional damage model is obtained, and a calibrated blasting crack damage cloud map is obtained.

4. The method for predicting the in-situ leaching effect of sandstone-type uranium deposits after blasting according to claim 1, characterized in that: The COMSOL multi-physics model in S3 is a coupled numerical model of the porous media dilute species transport equation, Darcy's seepage equation, the mineral dissolution domain ordinary differential equations, and the uranyl complex reaction-diffusion equation.

5. The method for predicting the effect of in-situ leaching mining of sandstone-type uranium deposits after blasting according to claim 1, characterized in that: In S4, a coupling interface between the COMSOL multiphysics model and PHREEQC was constructed based on MATLAB. The optimal time step during the seepage-dissolution coupling calculation phase within the cycle time was determined. The optimal time step was then input into the COMSOL multiphysics model to obtain the COMSOL initial concentration. Specifically, the following steps were performed: Construct a coupling interface between COMSOL multi-physics model and PHREEQC based on MATLAB; The cycle time is optimized by leaching efficiency to obtain the optimized cycle time; Based on the evolution characteristics of the uranium concentration gradient, a time step adaptive control algorithm is set to obtain the optimal time step for the seepage-dissolution coupling calculation stage within the optimization cycle time. The optimal time step is input into the COMSOL multiphysics model to obtain the COMSOL initial concentration.

6. The method for predicting the effect of in-situ leaching mining of sandstone-type uranium deposits after blasting according to claim 1, characterized in that: The geochemical reaction calculation results in S5 include: liquid element concentrations in pores and secondary mineral precipitation data sets.

7. The method for predicting the in-situ leaching effect of sandstone-type uranium deposits after blasting according to claim 1, characterized in that: In S7, the parameter adaptive matching function driven by the integrated LSTM neural network is used to dynamically correct the uranyl complex migration coefficient during the dynamic cyclic iterative simulation process to obtain the corrected uranyl complex migration coefficient. Specifically, the following are the steps: The redox potential and pH of the leaching solution monitored by the electrochemical sensor and the initial uranyl complex migration coefficient calculated by the COMSOL multi-physics model were input into the LSTM neural network to predict the dynamic correction factor and obtain the dynamic correction factor. According to the dynamic correction factor, the uranyl migration coefficient is automatically adjusted by integrating the parameter adaptive matching function driven by the LSTM neural network to obtain the corrected uranyl complex migration coefficient.

8. The method for predicting the effect of in-situ leaching mining of sandstone-type uranium deposits after blasting according to claim 1, characterized in that: The results of the multi-scale dynamic simulation of uranium migration in S8 include: uranium concentration distribution data and uranium migration path prediction.

9. A prediction system for in-situ leaching mining effect of sandstone-type uranium deposits after blasting, characterized in that: include: The first model building module is used to build the HJC rock blasting constitutive model based on the finite element analysis platform, and then perform blasting numerical simulation to obtain the blasting crack damage cloud map; A calibration module is used to calibrate the blasting crack damage cloud map using the Kriging interpolation algorithm based on the ground stress field data measured by the distributed optical fiber acoustic sensor to obtain a calibrated blasting crack damage cloud map; The second model building module is used to build an original COMSOL multiphysics model based on the calibrated blasting crack damage cloud map and COMSOL Multiphysics software, and load the redox potential and pH of the leaching solution monitored by the electrochemical sensor as boundary conditions into the original COMSOL multiphysics model to obtain the COMSOL multiphysics model; A determination module is used to build a coupling interface between the COMSOL multiphysics model and PHREEQC based on MATLAB, determine the optimal time step during the seepage-dissolution coupling calculation phase within the cycle time, and input the optimal time step into the COMSOL multiphysics model to obtain the COMSOL initial concentration; The calculation module is used to map the COMSOL initial concentration to PHREEQC and use MATLAB to call PHREEQC to perform geochemical reaction calculations based on the equilibrium constraints of the CO2+O2-water-rock reaction to obtain the geochemical reaction calculation results; Iterative simulation module is used to save the geochemical reaction calculation results, run the COMSOL file at the same time, and re-input the geochemical reaction calculation results into PHREEQC as the initial value of the next time step to obtain a dynamic cyclic iterative simulation; A correction module is used to dynamically correct the uranyl complex migration coefficient during the dynamic cyclic iterative simulation process by integrating the parameter adaptive matching function driven by the LSTM neural network to obtain the corrected uranyl complex migration coefficient; The output module is used to repeat the determination module, calculation module, iterative simulation module and prediction module until the cycle time is reached and the calculation is terminated, and the multi-scale dynamic simulation results of uranium migration are output.

10. An electronic device, characterized in that: The method comprises a memory and a processor, wherein a computer program is stored in the memory, and when the processor calls the computer program in the memory, the method realizes the content of the method for predicting the in-situ leaching mining effect of sandstone-type uranium deposit after blasting as claimed in any one of claims 1 to 8.

Images