U3Si2-cu dispersion type nuclear fuel with stable interface at high temperature and simulation construction method

Abstract

The present application belongs to the technical field of nuclear fuel simulation design, and aims at the problem that the interface reaction between U3Si2 fuel particles and Al matrix of the existing U3Si2-Al fuel is prone to occur at high temperature, and provides a U3Si2-Cu dispersion type nuclear fuel with interface stability at high temperature and a simulation construction method, a U-Si-Xe-M quaternary potential function is developed by a gradient optimization algorithm of multi-objective simulated annealing combined with a MEAM model, wherein M is any one of Al, Cu, Mg and W, and atomic scale information of irradiation defects and fission gas at the interface in the U3Si2-M fuel is obtained by means of calculation simulation, and finally the U3Si2-Cu dispersion type nuclear fuel is obtained. The U3Si2-Cu dispersion type nuclear fuel provided by the present application has high interface stability at high temperature.

Description

Technical Field

[0001] This invention belongs to the field of nuclear fuel simulation and design technology, specifically relating to a U3Si2-Cu dispersion nuclear fuel with interface stability at high temperature and a simulation construction method. Background Technology

[0002] Following nuclear accidents, the demands for nuclear fuel safety have reached new heights, leading to the concept of accident-resistant fuel (ATF) and sparking widespread research interest. ATF aims to provide a longer safe response time during accidents such as coolant failure, while maintaining robust operational characteristics. U-Si systems (especially U3Si2 fuel) are considered highly promising accident-resistant fuel candidates due to their high-temperature thermal conductivity, high uranium density, and excellent resistance to high-temperature water vapor corrosion.

[0003] Currently, composite dispersed fuel elements formed from U3Si2 and Al alloys have been widely used in experimental reactors. However, this fuel can only operate stably at low temperatures. When the temperature exceeds 250°C, interfacial reactions easily occur between the U3Si2 fuel particles and the Al matrix. The resulting reaction layer, combined with the fission gases (mainly Xe) generated during service, leads to a decrease in the mechanical properties at the interface and damage to the fuel element. Therefore, finding a matrix element with better stability to suppress interfacial reactions and fission gas aggregation between fuel particles and the matrix is ​​an urgent need to promote the application of U3Si2 dispersed fuels in high-temperature reactors.

[0004] Due to the high cost and long experimental cycle of nuclear fuel characterization experiments, molecular dynamics simulations can avoid these difficulties and delve deeper into the interaction between fuel particles and the matrix. However, the reliability of molecular dynamics methods heavily depends on the accuracy of the potential function used, and such potential functions are currently scarce. Only one potential function for the U-Si-Xe ternary system is known in the literature, developed based on the MEAM theoretical model. Verification results have shown that this potential function has significant deviations in describing the mechanical properties and point defect properties of U3Si2. Furthermore, no potential function has yet been developed to describe the interfacial properties between U3Si2 and the matrix elements. Therefore, the primary research goal is to develop a potential function that can accurately calculate the mechanical properties and point defect properties of U3Si2 and reflect the interfacial properties between U3Si2 and the matrix elements. Currently, the most suitable potential function model for describing U alloy systems is the MEAM model; however, due to the large number of fitting parameters and the coupling between these parameters, previous analytical methods have not yielded ideal results for this multi-parameter fitting. Summary of the Invention

[0005] The purpose of this invention is to overcome the shortcomings of the prior art and provide a U3Si2-Cu dispersed nuclear fuel with a stable interface at high temperature and a method for its simulation construction.

[0006] The first objective of this invention is to provide a U3Si2-Cu dispersed nuclear fuel that is stable at high temperatures, comprising a Cu matrix and U3Si2 core particles dispersed on the Cu matrix;

[0007] When the temperature is above 250℃, U3Si2-Cu dispersed nuclear fuel, together with the fission gas Xe produced during nuclear fission, forms a U-Si-Xe-Cu quaternary system during service.

[0008] The second objective of this invention is to provide a method for simulating the construction of the above-mentioned high-temperature interface-stable U3Si2-Cu dispersed nuclear fuel, comprising the following steps:

[0009] Construct the U-Si-Xe-M quaternary potential function;

[0010] Where M is any one of the metallic elements Al, Cu, Mg, and W;

[0011] Molecular dynamics simulations were performed based on the constructed U-Si-Xe-M quaternary potential function to simulate the evolution of interface atoms and fission gas Xe at the interface when M is the matrix, and the evolution results of the interface were obtained.

[0012] Based on the evolution of the interface, M was determined to be the metal element Cu. The metal element Cu was combined with U3Si2 core particles to finally obtain U3Si2-Cu dispersed nuclear fuel.

[0013] Preferably, the U-Si-Xe-M quaternary potential function is constructed using the following method:

[0014] A semi-empirical potential function model is selected as the analytical expression for constructing the U-Si-Xe-M quaternary potential function. Boundary constraints on the fitting parameters are defined based on the characteristics of the semi-empirical potential function model, and the initial distribution of the fitting parameters is defined to obtain the initial potential function.

[0015] Establish a database of fitting target quantities based on the physicochemical properties of the fitting target;

[0016] The initial potential function is used to calculate the fitting target quantity, and the calculated value of the fitting target quantity is obtained. The initial potential function is then trained using the calculated value of the fitting target quantity. The parameters of the initial potential function are updated by continuously optimizing the loss function, and the optimal solution set of the loss function is obtained.

[0017] The effectiveness of each solution in the optimal solution set of the loss function is evaluated based on the database of fitted target quantities. The solutions that meet the requirements are used to generate potential files, resulting in the U-Si-Xe-M quaternary potential function.

[0018] Preferably, the physical and chemical properties of the fitting target include the physical and chemical properties of pure elements and the physical and chemical properties of alloys;

[0019] The physical and chemical properties of pure elements include the lattice constant, cohesive energy, elastic constant, bulk modulus, vacancy formation energy, and interstitial formation energy of elemental crystals;

[0020] The physical and chemical properties of alloys include the substitutional formation energy of solute atoms in the substrate, the interstitial site formation energy, and the binding energy of defect clusters.

[0021] Preferably, the calculated value of the fitted target quantity is determined by simulation calculation using lattice dynamics relaxation method, conjugate gradient algorithm, and large-scale parallel molecular dynamics simulation software.

[0022] Preferably, the specific method for evaluating the effectiveness of each solution in the optimal solution set of the loss function based on the database of fitted target quantities, generating potential files from the solutions that meet the requirements, and obtaining the U-Si-Xe-M quaternary potential function is as follows:

[0023] Each solution in the optimal solution set of the loss function corresponds to a set of feasible potential parameters. The potential function generated by this set of potential parameters is used for molecular dynamics simulation, and the calculated value of the fitting target quantity corresponding to this potential function is calculated.

[0024] The calculated values ​​of all fitted target quantities are compared with the corresponding fitted target quantity values ​​in the fitted target quantity database.

[0025] When the maximum error between the calculated value of the fitted target quantity and the fitted target quantity value does not exceed 30%, the solution that meets the requirements will generate a potential file, and the U-Si-Xe-M quaternary potential function will be obtained.

[0026] When the maximum error between the calculated value of the fitted target quantity and the value of the fitted target quantity exceeds 30%, that is, when there is no solution in the optimal solution set of the loss function that meets the requirements, the current optimal solution set of the loss function is used as the initial parameter. The steps of training the multi-objective optimization model and evaluating the effect of each solution in the optimal solution set of the loss function are repeated until the calculated value of the fitted target quantity meets the requirements. The solution that meets the requirements is then used to generate a potential file, and the U-Si-Xe-M quaternary potential function is obtained.

[0027] Preferably, the interface atoms are U, Si, and M; the simulation of evolution behavior includes the formation simulation of the interface reaction layer and the diffusion and accumulation simulation of the fission gas Xe at the interface.

[0028] Preferably, the formation simulation of the interface reaction layer is achieved through the following method:

[0029] Constructing the U3Si2-M interface;

[0030] Based on the U3Si2-M interface, the diffusion behavior of interface atoms under high temperature irradiation was simulated. After 4 ns of simulation evolution, the mean square displacement of each interface atom was counted, the diffusion coefficient was calculated, and the thickness and position of the interface reaction layer were counted to simulate the formation of the interface reaction layer.

[0031] The high-temperature irradiation conditions are: temperature 800~1000K, irradiation energy 1~3keV.

[0032] Preferably, the U3Si2-M interface is constructed using the following method:

[0033] A complete U3Si2 supercell was constructed and expanded to more than 100 angstroms. U and Si atoms in the upper half of the Z-direction 15 lattices were removed to form the surface of a U3Si2 fuel.

[0034] At a distance of 3 Å from the surface, M crystal atoms are created. The total thickness of the M crystal atoms is 15 M lattices. After the entire system is relaxed to equilibrium, it is relaxed again for 5 ns in a canonical ensemble at a temperature of 550 K to construct the U3Si2-M interface.

[0035] Preferably, the diffusion and accumulation of fission gas Xe at the interface are simulated using the following methods:

[0036] Based on the constructed U3Si2-M interface, 0.5% Xe atoms were randomly added to the interstitial positions of the U3Si2 bulk near the U3Si2-M interface to obtain the U3Si2-M-Xe system.

[0037] The U3Si2-M-Xe system was relaxed in a canonical ensemble, and the simulation temperatures were set to 800~1000K. After 4ns of simulation evolution, the mean square displacement of Xe atoms was calculated, and the size of the final Xe clusters was statistically analyzed, thus realizing the simulation of the diffusion and aggregation of fission gas Xe at the interface.

[0038] Compared with the prior art, the advantages of this invention are as follows:

[0039] (1) Compared with existing experimental studies, the method provided by this invention can greatly reduce research costs and time through computer simulation technology, and can obtain microscopic atomic-scale evolution information that cannot be observed in experiments;

[0040] (2) The present invention constructs the U-Si-Xe-M quaternary potential function based on the multi-objective optimization potential function parameter fitting method. Compared with the traditional method, the fitting efficiency is higher and the parameter fitting effect is better. Through the comparison of potential function accuracy, it can be seen that the potential function provided by the present invention is closer to the experimental value or DFT result than the existing potential function in both qualitative and quantitative aspects. This further verifies that the potential function provided by the present invention has high fitting accuracy.

[0041] (3) In the process of fitting potential parameters, the present invention minimizes human intervention, improves the automation of potential function optimization, and greatly reduces the consumption of human resources.

[0042] (4) The U3Si2-Cu dispersed nuclear fuel provided by the present invention has high stability at high temperature. Simulation results show that at high temperature, the Cu matrix of the U3Si2-Cu dispersed nuclear fuel hardly reacts with the U3Si2 core particles, thereby avoiding the mechanical properties at the interface and the damage to the fuel elements caused by the interaction with the fission gas (mainly Xe) generated during service, thus ensuring the normal application of the U3Si2-Cu dispersed nuclear fuel at high temperature. Attached Figure Description

[0043] Figure 1 A flowchart of the method for constructing the U-Si-Xe-M quaternary potential function;

[0044] Figure 2 The U3Si2-Al interface structure after evolving for 5 ns at 550 K;

[0045] Figure 3 The U3Si2-Cu interface structure after evolving for 5 ns at 550 K;

[0046] Figure 4 The model diagram used to simulate the migration and aggregation behavior of Xe at the U3Si2-Al / Cu interface;

[0047] Among them, purple atoms are Al or Cu, yellow atoms are Xe, and red and blue atoms are U and Si, respectively;

[0048] Figure 5 Figures showing the diffusion and aggregation behavior of Xe atoms at the U3Si2-Al interface at different temperatures;

[0049] Among them, the purple and yellow atoms are Al and Xe, respectively, and the red and blue atoms are U and Si, respectively.

[0050] Figure 6 To analyze and statistically determine the final size and number of Xe clusters using Ovito software;

[0051] Among them, large clusters with more than 30 Xe atoms are circled;

[0052] Figure 7 denoted as the diffusion coefficient of Xe at the U3Si2-Al interface, within the vacancy mechanism, and in amorphous U3Si2 at different temperatures;

[0053] Figure 8 The distribution of Xe atoms at the U3Si2-Cu interface after 4 ns of evolution at different temperatures;

[0054] Figure 9 Model diagram used to simulate the diffusion of atoms and the formation of reaction layer at the interface of U3Si2-Al and U3Si2-Cu under different temperature and irradiation conditions;

[0055] Figure 10 The diffusion coefficients of U, Si, and Al atoms at the U3Si2-Al interface under different temperatures and irradiation conditions;

[0056] Figure 11 The mean square displacement and diffusion coefficient at the U3Si2-Cu interface under high-temperature irradiation conditions;

[0057] Figure 12 The figure shows a comparison of the evolution of the reaction layer at the interface between U3Si2-Al and U3Si2-Cu under different irradiation intensities at a temperature of 850K.

[0058] Figure 13 The figure shows a comparison of the evolution of the reaction layer at the interface between U3Si2-Al and U3Si2-Cu under different irradiation intensities at a temperature of 900K.

[0059] Figure 14 The figure shows a comparison of the evolution of the reaction layer at the interface between U3Si2-Al and U3Si2-Cu under different irradiation intensities at a temperature of 950K. Detailed Implementation

[0060] The following will be described in conjunction with embodiments of the present invention. Figures 1 to 14 The technical solutions in the embodiments of the present invention are clearly and completely described herein. Obviously, the described embodiments are only a part of the embodiments of the present invention, and not all of the embodiments. Based on the embodiments of the present invention, all other embodiments obtained by those of ordinary skill in the art without creative effort are within the scope of protection of the present invention.

[0061] Example 1: This embodiment of the invention provides a U3Si2-Cu dispersed nuclear fuel that is stable at high temperature, comprising a Cu matrix and U3Si2 core particles dispersed on the Cu matrix;

[0062] When the temperature is above 250℃, U3Si2-Cu dispersed nuclear fuel, together with the fission gas Xe produced during nuclear fission, forms a U-Si-Xe-Cu quaternary system during service.

[0063] Example 2: This embodiment of the invention provides a method for simulating the construction of U3Si2-Cu dispersed nuclear fuel with a stable interface at high temperature, as described in Example 1. The method specifically includes the following steps:

[0064] S1. Construct the U-Si-Xe-M quaternary potential function;

[0065] Where M is any one of the metallic elements Al, Cu, Mg, and W;

[0066] Specifically, such as Figure 1 As shown, the U-Si-Xe-M quaternary potential function is constructed using the following method:

[0067] S11. Select a semi-empirical potential function model as the analytical expression for constructing the U-Si-Xe-M quaternary potential function. Determine the boundary constraints of the fitting parameters based on the characteristics of the semi-empirical potential function model, define the initial distribution of the fitting parameters, and obtain the initial potential function.

[0068] In this embodiment of the invention, the semi-empirical potential function model is preferably the Modified Embedded Atom Method (MEAM) model. Currently, the development of potential functions for U and U alloys is typically based on semi-empirical potential function models, which offer high computational efficiency and good accuracy. The MEAM model, due to its correction for the angular contribution of the electron density cloud, is most suitable for describing systems like U3Si2 with directional bonding. Therefore, the MEAM model is used in the development of the U-Si-Xe-M quaternary potential function in this invention.

[0069] When defining the boundary constraints for the fitting parameters, a reasonable sampling range needs to be given based on the effect of the parameters. For the MEAM model, solving for the pair potential energy between atoms requires providing a reference structure of the target element, and using the ROSE equation to give the energy of each atom in the reference structure. Theoretically, the selection of the reference structure is arbitrary, but for the convenience of calculating its basic properties, a simpler ground-state structure is usually used.

[0070] In this embodiment of the invention, the reference structures for U and W are body-centered cubic (bcc) structures, Si is a diamond structure, Mg is a hexagonal close-packed (hcp) structure, and Xe, Al, and Cu are face-centered cubic (fcc) structures. The main principle for choosing the reference structure of the alloy is to ensure that all first nearest neighbors of each atom are of another atom, while all second nearest neighbors are of the same atom. This is primarily to facilitate the calculation of the potential interactions between the two types of atoms. Commonly used alloy reference structures include the B1 structure (sodium chloride type) and the B2 structure (cesium chloride type). In this invention, the reference structure for the alloy potential function is entirely the B1 structure.

[0071] When calculating the single-atom energy of a reference structure using the ROSE equation, the equilibrium nearest neighbor distance *re*, bulk modulus *B*, equilibrium volume *Ω*, and cohesive energy *Ec* of the reference structure are required. These values ​​are typically obtained experimentally or through density functional theory (DFT) calculations and can be used in fitting iterations if necessary. The remaining initial parameters need to be given empirically; for example, the value of *a* is usually between 0 and 1, while the values ​​of *β* (0-3) are usually between 0 and 10, *t* (1-3) are between -10 and 10, and *C*... max Between 1.5 and 3.8, C min Between 0 and 2.0.

[0072] S12. Establish a database of fitting target quantities based on the physicochemical properties of the fitting target;

[0073] Specifically, the physicochemical properties of the fitting target are the physicochemical properties of the U-Si-Xe-M system, including the physicochemical properties of pure elements and the physicochemical properties of alloys;

[0074] The physical and chemical properties of pure elements include the lattice constant, cohesive energy, elastic constant, bulk modulus, vacancy formation energy, and interstitial formation energy of elemental crystals;

[0075] The physical and chemical properties of alloys include the substitutional formation energy of solute atoms in the substrate, the interstitial site formation energy, and the binding energy of defect clusters.

[0076] The aforementioned physical and chemical property data can be derived from experiments or first-principles calculations. The database of the target quantity should contain a sufficiently large amount of data to enhance the effectiveness of model training. The data in the database of the target quantity are defined as the target quantity values.

[0077] S13. Calculate the fitting target quantity using the obtained initial potential function, obtain the calculated value of the fitting target quantity, and use the calculated value of the fitting target quantity to train the multi-objective optimization model of the initial potential function. Update the parameters of the initial potential function by continuously optimizing the loss function, and obtain the optimal solution set of the loss function.

[0078] Specifically, the calculated value of the fitting target quantity is determined by simulation calculation using lattice dynamics relaxation method, conjugate gradient algorithm, and large-scale parallel molecular dynamics simulation software.

[0079] The process of S13 will be explained in detail below:

[0080] For MEAM models, the complexity of their formulas and the large number of fitting parameters make determining the analytical expression of the model difficult. Fitting these undetermined parameters is often a tedious, lengthy, and arduous task. Traditional methods determine potential parameters by analytically solving certain physical properties, but this approach is not suitable for potential function models with a large number of fitting parameters. Multi-objective optimization algorithms have a natural advantage in handling large amounts of data fitting. Therefore, this invention uses a multi-objective optimization algorithm instead of traditional fitting methods to solve the problem of difficult model parameter fitting.

[0081] During the development process, the calculation of the target quantities was mainly performed using the lattice dynamics relaxation method. For the lattice constant, cohesive energy, elastic constant, bulk modulus, vacancy formation energy, interstitial formation energy, substitutional and interstitial site formation energies of solute atoms in the substrate, and binding energy of defect clusters, the lattice dynamics relaxation method and conjugate gradient algorithm were combined with the large-scale atomic / molecular massively parallel simulation software (LAMMPS) to simulate and calculate the values.

[0082] The calculated values ​​of the fitted target quantity are used to train a multi-objective optimization model for the initial potential function. Specifically, during the fitting process of the initial potential function parameters, Monte Carlo sampling and iterative methods are combined, and a multi-objective simulated annealing algorithm is used to solve for the global optimum of the loss function, thus obtaining the optimal solution set of the loss function. The loss function used is as follows:

[0083] ;

[0084] in, O f Represents the loss function. w i Represents the target quantity for fitting i The weight it accounts for λ Represents potential parameters, f i ( λ ) represents the potential parameter λ The fitted target quantity calculated by LAMMPS i The value, F i Represents the target quantity for fitting iReference values.

[0085] In this way, the present invention transforms the process of potential parameter optimization into a process of finding the minimum value of the loss function. This approach also has another advantage. Due to the inherent limitations of the potential function model, it is difficult to simultaneously guarantee the accuracy of all fitted properties using a finite number of potential parameters. Therefore, weights can be manually set according to the simulation needs to ensure that the potential function has a more ideal description of a certain type of property. Simultaneously, multiple judgment conditions are set when calculating the target quantity of the potential function. When the error value between the prediction effect of a set of potential function parameters for certain important properties, such as lattice constant and defect energy, and the reference target value exceeds a certain threshold, subsequent calculations are stopped, and this set of potential function parameters is skipped to the next iteration. This ensures both high iteration efficiency of the potential parameters and good predictive ability of the potential function.

[0086] S14. Based on the database of fitted target quantities, evaluate the effectiveness of each solution in the optimal solution set of the loss function, generate potential files from the solutions that meet the requirements, and obtain the U-Si-Xe-M quaternary potential function; the specific method is as follows:

[0087] Each solution in the optimal solution set of the loss function corresponds to a set of feasible potential parameters. The potential function generated using these parameters can be used for molecular dynamics simulations, and the calculated value of the fitted target quantity corresponding to this potential function can be obtained. The fitted target quantity values ​​calculated through first-principles calculations (DFT method) or experiments correspond to data in the fitted target quantity database. By comparing the calculated values ​​of all fitted target quantities with the fitted target quantity values, the goodness or badness of the potential parameter can be determined. The supercell size used in the calculation is 10. a 0x10 b 0x10 c 0 ( a 0、 b 0、 c 0 represents the equilibrium lattice constant), and then defect structures can be obtained by adding or removing atoms as needed.

[0088] When the maximum error between the calculated value of the fitted target quantity and the fitted target quantity value does not exceed 30%, the solution that meets the requirements will generate a potential file, and other properties will be predicted and verified. These predicted properties are those that were not included in the fitting, and the U-Si-Xe-M quaternary potential function will be obtained.

[0089] When the maximum error between the calculated value of the fitted target quantity and the value of the fitted target quantity exceeds 30%, that is, when there is no solution in the optimal solution set of the loss function that meets the requirements, the current optimal solution set of the loss function is used as the initial parameter. The steps of training the multi-objective optimization model and evaluating the effect of each solution in the optimal solution set of the loss function are repeated until the calculated value of the fitted target quantity meets the requirements. The solution that meets the requirements is then used to generate a potential file, and the U-Si-Xe-M quaternary potential function is obtained.

[0090] The accuracy of the U-Si-Xe-M quaternary potential function constructed in the embodiments of the present invention is verified below:

[0091] Due to the lack of relevant DFT or experimental data for reference and verification, the determination of the potential function parameters of U-Si-Xe-Mg / W was based on empirical methods (Kim HK, Jung WS, Lee B J. Modified embedded-atom methodinteratomic potentials for the Fe–Ti–C and Fe–Ti–N ternary systems[J]. Actamaterialia, 2009, 57(11): 3140-3147.). Subsequently, the quaternary potential function of the developed U-Si-Xe-Al / Cu was verified in detail, and the verification results are shown in Tables 1-5 below.

[0092] Table 1. Comparison of bulk properties (lattice constant and elastic properties) of U3Si2 calculated using the U-Si-Xe-M quaternary potential function with existing potential function (MEAM_Beeler) and DFT calculation results:

[0093] ;

[0094] Table 2. Comparison of defect properties of U3Si2 calculated by the U-Si-Xe-M quaternary potential function with existing potential function (MEAM_Beeler) and DFT calculation results:

[0095] ;

[0096] Table 3. Comparison of the defect properties of Xe in U3Si2 calculated by the U-Si-Xe-M quaternary potential function with the results calculated by the existing potential function (MEAM_Beeler) and DFT:

[0097] ;

[0098] Table 4. Comparison of the defect properties of Al in U3Si2 calculated by the U-Si-Xe-M quaternary potential function with the results calculated by DFT:

[0099] ;

[0100] Table 5. Comparison of defect properties of Cu in U3Si2 calculated by the U-Si-Xe-M quaternary potential function with the DFT calculation results:

[0101] ;

[0102] The results in Tables 1-5 show that, compared with the DFT results, the U-Si-Xe-M quaternary potential function constructed by the method of this invention is significantly superior to the calculation results of the potential function proposed in the literature (Beeler B, Andersson D, Cooper MWD, et al. A molecular dynamics study of the behavior of Xe in U3Si2[J]. Journal of Nuclear Materials, 2019, 523: 413-420.). This is sufficient to prove the accuracy of the U-Si-Xe-Al / Cu quaternary potential function constructed in this invention.

[0103] S2. Molecular dynamics simulations were performed based on the constructed U-Si-Xe-M quaternary potential function to simulate the evolution of interface atoms and fission gas Xe at the interface when M is the matrix, and the evolution results of the interface were obtained.

[0104] In this embodiment of the invention, the interface atoms are U, Si, and M; the simulation of evolution behavior includes the simulation of the formation of the interface reaction layer and the simulation of the diffusion and accumulation of fission gas Xe at the interface.

[0105] Specifically, the formation simulation of the interface reaction layer is achieved through the following methods:

[0106] (1) Construct the U3Si2-M interface. The specific construction method is as follows:

[0107] A complete U3Si2 supercell is constructed. The supercell size is taken as a common multiple of the lattice constants of U3Si2 and M in the X and Y directions to ensure periodic boundary conditions. Then the cell is expanded to more than 100 angstroms. The supercell size in the Z direction is 35 U3Si2 lattices. The U and Si atoms in the upper half of the Z direction 15 lattices are deleted to form the surface of a U3Si2 fuel.

[0108] At a distance of 3 Å from the surface, M crystal atoms were created, with a total thickness of 15 M lattices. The entire system was then relaxed to equilibrium using lattice dynamics relaxation and conjugate gradient algorithm, and then relaxed again for 5 ns in a canonical ensemble (NVT) at 550 K to construct the U3Si2-M interface.

[0109] In the specific construction process:

[0110] When M is Al, a 13×13×35 U3Si2 supercell is constructed;

[0111] When M is Cu, a U3Si2 supercell of 11×11×35 is obtained;

[0112] When M is Mg, a U3Si2 supercell of 11×16×35 is obtained;

[0113] When M is W, a 16×16×35 U3Si2 supercell is obtained.

[0114] The choice of supercell size varies for different systems mainly because it is necessary to simultaneously ensure the periodic boundary conditions of U3Si2 and the matrix in the X and Y directions.

[0115] (2) Based on the constructed U3Si2-M interface, the diffusion behavior of interface atoms under high temperature irradiation was simulated. The high temperature irradiation conditions were: temperature 800~1000K, irradiation energy 1~3keV. After 4ns of simulation evolution, the mean square displacement of each interface atom was counted, the diffusion coefficient was calculated, and the thickness and position of the interface reaction layer were counted to realize the simulation of the formation of the interface reaction layer.

[0116] Specifically, the diffusion and accumulation of fission gas Xe at the interface were simulated using the following methods:

[0117] Based on the constructed U3Si2-Al / Cu interface, 0.5% Xe atoms were randomly added to the interstitial sites of the U3Si2 bulk near the U3Si2-Al / Cu interface to obtain the U3Si2-Xe-Al / Cu system, as shown below. Figure 4 As shown;

[0118] The U3Si2-Xe-Al / Cu system was relaxed in a canonical ensemble, and the simulation temperatures were set to 850K, 900K and 950K, respectively. After 4 ns of simulation evolution, atomic coordinate information was output for subsequent Xe atom mean square displacement (MSD) calculation. The size of the final Xe clusters was statistically analyzed using Ovito software to simulate the diffusion and aggregation of fission gas Xe at the interface.

[0119] S3. Based on the evolution of the interface, the optimal choice for M is determined to be the metal element Cu. The metal element Cu and U3Si2 core particles are then combined to obtain U3Si2-Cu dispersed nuclear fuel.

[0120] The interface structure distributions of the evolved U3Si2-Al and U3Si2-Cu in this invention are as follows: Figure 2 and Figure 3 As shown, through Figure 2 We can see that the Al atoms at the interface begin to become disordered. This amorphization, after a longer evolution period or under irradiation conditions, greatly promotes the formation of the reaction layer. Conversely, Figure 3 No amorphization of Cu was observed, indicating that the U3Si2-Cu interface is thermodynamically stable. For the Mg matrix, the constructed U3Si2-Mg interface is extremely unstable due to the significant difference in crystal structure between Mg and U3Si2. The reaction layer formed by W is similar to that of Al; therefore, these two matrix elements will not be included in further research. Future studies will primarily focus on the interfacial behavior of U3Si2-Al and U3Si2-Cu.

[0121] The aggregation behavior of fission gas Xe at U3Si2-Al is as follows: Figure 5 and Figure 6 As shown. (Through) Figure 5 and Figure 6 The results show that at lower temperatures, Xe atoms exist in the form of multiple small clusters, while as the temperature increases, the number of small Xe clusters decreases and they merge into a small number of large clusters. The diffusion coefficients of Xe at the U3Si2-Al interface at different temperatures are fitted using MSD data. Figure 7 As shown, the diffusion coefficients of Xe in the vacancy mechanism (bulk U3Si2) and in amorphous U3Si2 are provided for comparison. Figure 7 The results show that the diffusion rate of fission gas Xe at the interface is much faster than that in bulk U3Si2 and amorphous U3Si2. This indicates that Xe atoms at the U3Si2-Al interface tend to migrate and aggregate to form larger Xe bubbles, which significantly affects the thermal conductivity and mechanical properties at the interface.

[0122] This invention also analyzed the migration and aggregation behavior of Xe atoms at the U3Si2-Cu interface. Surprisingly, regardless of temperature changes, after 4 ns of evolution, the distribution of Xe atoms remained almost unchanged from their initial positions. Figure 8 As shown in the figure. This indicates that Xe atoms at the U3Si2-Cu interface hardly undergo migration and aggregation, so it can be predicted that the nucleation and growth of Xe bubbles at the U3Si2-Cu interface is thermodynamically unfavorable.

[0123] This invention also employs the developed U-Si-Xe-Al / Cu quaternary potential function to study the atomic diffusion at the U3Si2-Al and U3Si2-Cu interfaces under different temperatures and irradiation conditions. The simulation model is as follows: Figure 9 As shown, three different temperature conditions and three different irradiation energies were considered. The specific simulation temperatures were 850K, 900K, and 950K; the PKA energies were 1, 2, and 3 keV; the selected cascade region was located near the interface of the U3Si2 bulk. The yellow area around the model used a 4-layer atomic temperature control (NVT ensemble), while the blue area in the middle used an NVE ensemble. The simulation box size was 13×13×35. After 4 ns of simulation evolution, the MSD of each atom was calculated, and then the average of 10 independent simulation results was taken to determine the diffusion coefficient.

[0124] The diffusion results at the U3Si2-Al interface are as follows: Figure 10 As shown, through Figure 10 It can be seen that under the three different temperature conditions, the effect of PKA energy on the diffusion behavior of U, Si, and Al atoms is almost negligible. However, as the temperature increases, the diffusion coefficients of the three atoms also increase accordingly. The diffusion coefficients under different temperature conditions are listed in Table 6 below.

[0125] Table 6. Diffusion coefficients of U, Si, and Al atoms at the U3Si2-Al interface at different temperatures when irradiated with an energy of 3 keV:

[0126] ;

[0127] Through Table 6 and Figure 10 The results lead to the conclusion that the diffusion behavior of atoms at the U3Si2-Al interface is primarily temperature-sensitive and independent of irradiation intensity. For the U3Si2-Cu system, this invention found that even at 950 K and 3 keV irradiation energy, U and Si atoms, except for short-range diffusion during the cascade stage, hardly move during subsequent evolution. Only Cu interstitial atoms generated during the cascade process undergo continuous diffusion within the Cu bulk. Therefore, this invention only calculated the diffusion coefficient of Cu, and the results are as follows: Figure 11 As shown. (Through) Figure 11 The results show that the atoms at the U3Si2-Cu interface are very stable under both high temperature and irradiation conditions, and large-scale diffusion does not occur.

[0128] This invention utilizes a developed U-Si-Xe-Al / Cu potential function to investigate the formation of the reaction layer at the U3Si2-Al and U3Si2-Cu interfaces under different temperature irradiation conditions. The simulation model used is as follows: Figure 9As shown, three different temperature conditions and three different irradiation energies were considered. The specific simulation temperatures were 850K, 900K, and 950K; the PKA energies were 1, 2, and 3 keV; the selected cascade region was located near the interface of the U3Si2 bulk. The yellow area around the model used a 4-layer atomic temperature control (NVT ensemble), while the blue area in the middle used an NVE ensemble. The simulation box size was 13×13×35. The simulation evolution lasted for 4 ns, and the distribution ratio of U3Si2 fuel and matrix atoms at various points along the vertical interface (Z-axis) was statistically analyzed.

[0129] The simulation results at temperatures of 850K, 900K, and 950K are as follows: Figure 12 , Figure 13 and Figure 14 As shown, the blue dashed line represents the position of the interface before the simulation starts, the black dashed line represents the new U3Si2 interface after 4ns evolution, and the red and yellow dashed lines represent the interface positions of Al and Cu after evolution, respectively. Therefore, the area between the black and red dashed lines is the area occupied by the U3Si2-Al reaction layer, and the area between the black and yellow dashed lines is the area occupied by the U3Si2-Cu reaction layer.

[0130] pass Figure 12 , Figure 13 and Figure 14 The results showed that under all simulation conditions, the reaction layer thickness at the U3Si2-Cu interface was smaller than that at the U3Si2-Al interface. Furthermore, the reaction layer thickness of U3Si2-Al increased with increasing temperature and irradiation energy, while the reaction layer thickness of U3Si2-Cu was almost unaffected by temperature, only by irradiation conditions. This is because with increasing irradiation energy, U or Si atoms are driven deeper into the Cu matrix by PKA atoms, resulting in an increase in the statistical reaction layer thickness. Therefore, it can be concluded that the reaction layer thickness at the U3Si2-Al interface is larger and more sensitive to temperature and irradiation intensity, indicating significant interdiffusion of atoms at the interface.

[0131] For the U3Si2-Cu system, since the statistically significant reaction layer thickness is very small and shows almost no significant change with temperature, it can be considered as an atomic mixing layer generated by irradiation rather than a reaction layer generated by interdiffusion. Therefore, the U3Si2-Cu interface can effectively suppress the formation of a reaction layer even under high-temperature irradiation conditions. Based on the above results, the optimal choice for M is determined to be the metallic element Cu. Cu and U3Si2 core particles are then combined using existing synthesis methods to finally obtain U3Si2-Cu dispersed nuclear fuel.

[0132] Although embodiments of the invention have been shown and described, those skilled in the art will understand that various changes, modifications, substitutions and alterations can be made to these embodiments without departing from the principles and spirit of the invention, the scope of which is defined by the claims and their equivalents.

Claims

1. A U3Si2-Cu dispersion-type nuclear fuel that is interface-stable at high temperatures, characterized in that, Includes a Cu matrix and U3Si2 core particles dispersed on the Cu matrix; When the temperature is above 250℃, U3Si2-Cu dispersed nuclear fuel, together with the fission gas Xe produced during nuclear fission, forms a U-Si-Xe-Cu quaternary system during service.

2. A method for simulating the construction of U3Si2-Cu dispersed nuclear fuel with a stable interface at high temperature, characterized in that, Includes the following steps: Construct the U-Si-Xe-M quaternary potential function; Where M is any one of the metallic elements Al, Cu, Mg, and W; Molecular dynamics simulations were performed based on the constructed U-Si-Xe-M quaternary potential function to simulate the evolution of interface atoms and fission gas Xe at the interface when M is the matrix, and the evolution results of the interface were obtained. The interface atoms are U, Si, and M; Based on the evolution of the interface, M was determined to be the metal element Cu. The metal element Cu was combined with U3Si2 core particles to finally obtain U3Si2-Cu dispersed nuclear fuel.

3. The method according to claim 2, characterized in that, The U-Si-Xe-M quaternary potential function is constructed using the following method: A semi-empirical potential function model is selected as the analytical expression for constructing the U-Si-Xe-M quaternary potential function. Boundary constraints on the fitting parameters are defined based on the characteristics of the semi-empirical potential function model, and the initial distribution of the fitting parameters is defined to obtain the initial potential function. Establish a database of fitting target quantities based on the physicochemical properties of the fitting target; The initial potential function is used to calculate the fitting target quantity, and the calculated value of the fitting target quantity is obtained. The initial potential function is then trained using the calculated value of the fitting target quantity. The parameters of the initial potential function are updated by continuously optimizing the loss function, and the optimal solution set of the loss function is obtained. The effectiveness of each solution in the optimal solution set of the loss function is evaluated based on the database of fitted target quantities. The solutions that meet the requirements are used to generate potential files, resulting in the U-Si-Xe-M quaternary potential function.

4. The method according to claim 3, characterized in that, The physical and chemical properties of the fitting target include the physical and chemical properties of pure elements and the physical and chemical properties of alloys; The physical and chemical properties of pure elements include the lattice constant, cohesive energy, elastic constant, bulk modulus, vacancy formation energy, and interstitial formation energy of elemental crystals; The physical and chemical properties of alloys include the substitutional formation energy of solute atoms in the substrate, the interstitial site formation energy, and the binding energy of defect clusters.

5. The method according to claim 3, characterized in that, The calculated values ​​of the fitted target quantity were determined by simulation calculation using lattice dynamics relaxation method, conjugate gradient algorithm, and large-scale parallel molecular dynamics simulation software.

6. The method according to claim 3, characterized in that, The method for evaluating the effectiveness of each solution in the optimal solution set of the loss function based on the database of fitted target quantities, generating potential files from the solutions that meet the requirements, and obtaining the U-Si-Xe-M quaternary potential function is as follows: Each solution in the optimal solution set of the loss function corresponds to a set of feasible potential parameters. The potential function generated by this set of potential parameters is used for molecular dynamics simulation, and the calculated value of the fitting target quantity corresponding to this potential function is calculated. The calculated values ​​of all fitted target quantities are compared with the corresponding fitted target quantity values ​​in the fitted target quantity database. When the maximum error between the calculated value of the fitted target quantity and the fitted target quantity value does not exceed 30%, the solution that meets the requirements will generate a potential file, and the U-Si-Xe-M quaternary potential function will be obtained. When the maximum error between the calculated value of the fitted target quantity and the value of the fitted target quantity exceeds 30%, that is, when there is no solution in the optimal solution set of the loss function that meets the requirements, the current optimal solution set of the loss function is used as the initial parameter. The steps of training the multi-objective optimization model and evaluating the effect of each solution in the optimal solution set of the loss function are repeated until the calculated value of the fitted target quantity meets the requirements. The solution that meets the requirements is then used to generate a potential file, and the U-Si-Xe-M quaternary potential function is obtained.

7. The method according to claim 2, characterized in that, The simulation of evolutionary behavior includes the formation of the interfacial reaction layer and the diffusion and accumulation of fission gas Xe at the interface.

8. The method according to claim 7, characterized in that, The formation of the interface reaction layer was simulated using the following method: Constructing the U3Si2-M interface; Based on the U3Si2-M interface, the diffusion behavior of interface atoms under high temperature irradiation was simulated. After 4 ns of simulation evolution, the mean square displacement of each interface atom was counted, the diffusion coefficient was calculated, and the thickness and position of the interface reaction layer were counted to simulate the formation of the interface reaction layer. The high-temperature irradiation conditions are: temperature 800~1000K, irradiation energy 1~3keV.

9. The method according to claim 8, characterized in that, The U3Si2-M interface is constructed using the following method: A complete U3Si2 supercell was constructed and expanded to more than 100 angstroms. U and Si atoms in the upper half of the Z-direction 15 lattices were removed to form the surface of a U3Si2 fuel. At a distance of 3 Å from the surface, M crystal atoms are created. The total thickness of the M crystal atoms is 15 M lattices. After the entire system is relaxed to equilibrium, it is relaxed again for 5 ns in a canonical ensemble at a temperature of 550 K to construct the U3Si2-M interface.

10. The method according to claim 8, characterized in that, The diffusion and accumulation of fission gas Xe at the interface were simulated using the following method: Based on the constructed U3Si2-M interface, 0.5% Xe atoms were randomly added to the interstitial positions of the U3Si2 bulk near the U3Si2-M interface to obtain the U3Si2-M-Xe system. The U3Si2-M-Xe system was relaxed in a canonical ensemble, the simulation temperature was set to 800~1000K, and after 4ns of simulation evolution, the mean square displacement of Xe atoms was calculated, and the size of the final Xe clusters was statistically analyzed, so as to realize the simulation of the diffusion and aggregation of fission gas Xe at the interface.

Images