Particle reinforced metal matrix composite modeling method for molecular dynamics simulation
By constructing a spherical cavity and embedding a spherical particle atomic model in molecular dynamics simulations, and performing energy minimization and structural relaxation treatments, the problem of precise control over enhanced particle distribution was solved, thereby improving the accuracy and reliability of the simulation.
Patent Information
- Authority / Receiving Office
- CN · China
- Patent Type
- Applications(China)
- Current Assignee / Owner
- Filing Date
- 2025-12-08
- Publication Date
- 2026-03-27
AI Technical Summary
In existing molecular dynamics simulations, the arrangement of reinforcing particles is simplified to an ideal or random distribution, making it difficult to precisely control the number, position, size, and volume fraction. This leads to distortion of the initial model, excessive internal stress, and reduced accuracy and reliability of the simulation results.
By combining the crystal structure information of the matrix material, a spherical cavity is constructed and an atomic model of spherical reinforcing particles is embedded. By combining energy minimization and structural relaxation processing, an initial atomic model that precisely controls the microstructure of the reinforcing particles is formed.
It enables precise control over the number, location, size, and volume fraction of reinforcing particles, eliminates atomic overlap and internal stress, improves the accuracy and reliability of molecular dynamics simulations, and supports the design of multiple particles with uniform distribution and different size combinations.
Smart Images
Figure CN121747784A_ABST
Abstract
Description
Technical Field
[0001] This invention belongs to the field of metal matrix composite modeling technology, specifically relating to a method for modeling particle-reinforced metal matrix composites for molecular dynamics simulation. Background Technology
[0002] In the field of materials science and engineering, particularly in aerospace and high-end equipment manufacturing, particle-reinforced metal matrix composites (such as aluminum-based silicon carbide) have attracted considerable attention due to their excellent specific strength, specific stiffness, and heat resistance. Their macroscopic mechanical properties are closely related to the distribution, size, and volume fraction of the reinforcing particles at the microscopic scale. Molecular dynamics simulations are a key computational tool for revealing the microscopic deformation mechanisms of such materials and predicting their macroscopic properties. However, accurate molecular dynamics simulations highly depend on high-quality initial atomic models.
[0003] Currently, methods for constructing atomic models of materials for molecular dynamics analysis have significant limitations: the arrangement of reinforcing particles is often simplified to an ideal or random distribution, making it difficult to precisely control their number, position, size, and volume percentage; atomic overlap or unreasonable initial forces at interfaces easily occur during modeling, leading to model distortion and excessive internal stress. These factors cause the initial model to deviate significantly from the microstructure of the real material, severely reducing the accuracy and reliability of subsequent simulation results, and failing to provide a reliable computational basis for the optimized design and performance prediction of materials. Summary of the Invention
[0004] The purpose of this invention is to provide a modeling method for particle-reinforced metal matrix composites for molecular dynamics simulations, so as to construct an initial atomic model of the particle-reinforced metal matrix composites that can accurately control the microstructural parameters of the reinforcing particles (such as number, position, size and volume fraction) and has no atomic overlap and low internal stress, thereby improving the accuracy and reliability of subsequent molecular dynamics simulations.
[0005] The present invention achieves the above objectives through the following technical solutions: This invention proposes a modeling method for particle-reinforced metal matrix composites for molecular dynamics simulations, the method comprising: An atomic model of the metal matrix is established by combining the crystal structure information of the target matrix material; In the three-dimensional space of the atomic model of the metal matrix, all atoms in the set spherical region are defined and removed to form a spherical cavity in the matrix to accommodate the reinforcing particles. Construct an atomic model of spherical reinforced particles that matches the spherical cavity; The atomic model of the spherical reinforcing particles is embedded into the corresponding spherical cavity of the metal matrix to form the initial atomic model of the composite material; The initial atomic model is subjected to energy minimization and structural relaxation to obtain the atomic model of particle-reinforced metal matrix composite material.
[0006] Furthermore, the step of defining and removing all atoms within a defined spherical region in the three-dimensional space of the atomic model includes: Based on a preset enhanced particle distribution scheme, cavity distribution parameters for N spherical cavities are generated; the cavity distribution parameters include the center coordinates (x, y) of the i-th cavity. i , y i , z i ), radius r i and particle type identifier T i Where i = 1, 2, ..., N; For each cavity distribution parameter, based on the cavity's center coordinates, radius, and particle type identifier, all atoms whose atomic coordinates are located within the spherical geometric space are selected from the metal matrix atomic model, and the selected atoms are removed from the model.
[0007] Furthermore, the construction of the spherical reinforced particle atomic model that matches the spherical cavity includes: Obtain the crystal structure file of the reinforcing particles and convert it into an atomic data file that can be recognized by molecular dynamics simulation software; A cell copying operation is performed on the converted atomic data file to construct a supercell model whose three-dimensional dimensions are all larger than the radius of the spherical cavity to be embedded; Using a preset point in the supercell model as the center of the sphere, and with a target radius R... p Define a spherical selection area with radius R; wherein, the target radius R p Based on the radius R of the corresponding spherical cavity c Determined, and satisfying R p < R c ; All atoms whose atomic coordinates are located within or on the surface of the spherical selection area are extracted from the supercell model to form a preliminary spherical atomic cluster. The atom types of the preliminary spherical atomic groups are identified and statistically analyzed, and the results are output as spherical enhanced particle atomic models that independently contain atomic coordinates and atom type information.
[0008] Furthermore, the embedding of the spherical reinforcing particle atomic model into the corresponding spherical cavity of the metal matrix includes: Read as the center coordinates (x) of the current spherical cavity i , y i , z iBased on the center coordinates, a global translation transformation is performed on the coordinates of all atoms in the spherical reinforced particle atomic model, so that the geometric center of the spherical reinforced particle atomic model is aligned with the coordinates (x, y). i , y i , z i Overlap; By applying a preset numbering remapping rule, the atom type numbers of all atoms in the spherical enhanced particle atom model are remapped based on the maximum atom type number M to generate continuous global atom type number data. The translated atomic coordinate data and mapped atomic type encoding data are concatenated with the atomic data of the metal matrix atomic model, and the box boundary information of the model is updated to generate a unified initial atomic model data file for composite materials.
[0009] Furthermore, the preset number remapping rule is as follows: the atom with the original atom type number j in the spherical enhanced particle atom model is remapped to a new type number M + j, where j is a consecutive positive integer starting from 1.
[0010] Furthermore, the energy minimization and structural relaxation processing of the initial atomic model includes: The initial atomic model is imported into molecular dynamics simulation software, and the conjugate gradient method is used to perform energy minimization iterative calculations on the model until the total potential energy of the system satisfies the first convergence criterion. Based on energy minimization, the model is placed under the NPT ensemble for molecular dynamics relaxation; the NPT ensemble is set with target temperature T and target pressure P, and the simulation is run until the system energy and pressure fluctuations satisfy the second convergence criterion, thus obtaining the atomic model of the particle-reinforced metal matrix composite material that has reached thermodynamic equilibrium.
[0011] Furthermore, the first convergence criterion is that the change in the total potential energy of the system is less than a first threshold, or the number of iterations reaches the upper limit; the second convergence criterion is that the rolling average of the total potential energy and pressure of the system fluctuates less than a second threshold within a set step size.
[0012] Furthermore, the method also includes: outputting the obtained atomic model of the particle-reinforced metal matrix composite material as an atomic data file that conforms to the input requirements of at least one of the following software: LAMMPS (Large-scale Atomic / Molecular Massively Parallel Simulator), GROMACS (GROningen MAchine for Chemical Simulations), NAMD (Nanoscale Molecular Dynamics), or Materials Studio.
[0013] The beneficial effects of this invention are as follows: (1) This invention can strictly control the number, position, size and volume ratio of reinforcing particles in the matrix through precise spherical space removal and particle embedding operations. It supports modeling of multiple particles in uniform distribution and can also realize multi-scale combination design of reinforcing particles of different sizes. At the same time, the matrix model can also be adapted to different spatial scale requirements by adjusting the number of cell replications in the command.
[0014] (2) This method is not limited to specific matrix or reinforcing materials. By replacing the initial model file, it is applicable to a variety of particle-reinforced metal matrix composite systems and can be easily extended to more complex particle arrangement designs. (3) In the modeling process, this invention solves the atomic overlap problem at its root by adopting the modeling steps of "matrix-cavity-particle-embedding" and controlling the size parameters. By introducing key energy minimization (conjugate gradient method) and NPT ensemble relaxation (isothermal and isobaric conditions) for dual optimization, it can effectively eliminate unreasonable atomic forces and internal stresses in the initial configuration of the model, so that the model can reach a thermodynamic equilibrium state. The atomic arrangement, particle distribution and interface bonding are all consistent with the real microstructure of the composite material, providing a high-precision material model for subsequent molecular dynamics simulations such as material forming and cutting. Attached Figure Description
[0015] Figure 1 This is a flowchart of a modeling method for particle-reinforced metal matrix composites for molecular dynamics simulation in one embodiment of the present invention; Figure 2 This is another flowchart of a method for modeling particle-reinforced metal matrix composites for molecular dynamics simulation in one embodiment of the present invention; Figure 3 This is a schematic diagram of a metal substrate model with spatial dimensions of 300*200*200Å established in this invention; Figure 4 This is a schematic diagram of the metal substrate model after the uniform spherical space has been removed in this invention; Figure 5 This is a schematic diagram of the metal matrix model after the removal of the heterogeneous spherical space in this invention; Figure 6 This is a schematic diagram of the supercell and multiphase spherical reinforced particle model in this invention; Figure 7 This is a schematic diagram of the supercell and uniform spherical reinforcing particle model in this invention; Figure 8 This is a schematic diagram of a composite material model in which spherical particles are embedded in the present invention; Figure 9 This is a schematic diagram of a modeling result for the data output during the relaxation stage in this invention. Detailed Implementation
[0016] The present application will now be described in further detail with reference to the accompanying drawings. It should be noted that the following specific embodiments are only used to further illustrate the present application and should not be construed as limiting the scope of protection of the present application. Those skilled in the art can make some non-essential improvements and adjustments to the present application based on the above application content.
[0017] This invention proposes a modeling method for particle-reinforced metal matrix composites for molecular dynamics simulations. The method employs a step-by-step modeling strategy. First, an atomic model of the metal matrix is established. Then, spherical cavities are created at predetermined locations within the matrix. Next, spherical atomic models of the reinforcing particles are constructed and embedded into the matrix cavities. Finally, energy optimization and structural relaxation are performed using molecular dynamics software to obtain a stable atomic model of the particle-reinforced metal matrix composite. This invention supports flexible control over the number, spatial location, size, and volume fraction of particles, exhibiting not only good versatility and scalability but also effectively improving modeling accuracy, providing a reliable foundation for molecular dynamics simulations of material forming and cutting.
[0018] The following describes the method in detail, using a preferred embodiment with aluminum (Al) as the matrix and silicon carbide (SiC) as reinforcing particles, and employing ATOMSK software for preprocessing modeling and LAMMPS software for kinetic optimization. Those skilled in the art should understand that this invention is not limited to this specific material combination or software tool; any material system and software platform capable of achieving the same or similar functions is within the scope of this invention.
[0019] Please see Figure 1 and Figure 2 A preferred embodiment of the present invention proposes a modeling method for particle-reinforced metal matrix composites for molecular dynamics simulations, the method comprising the following steps: S1: Based on the crystal structure information of the target matrix material, establish an atomic model of the metal matrix.
[0020] Combination Figure 3 The diagram illustrates a schematic of a metal substrate model with spatial dimensions of 300*200*200 Å. In this embodiment, the target substrate is aluminum with a face-centered cubic (FCC) structure. Using a crystal modeling tool (such as ATOMSK), a unit cell is created based on the lattice constant of aluminum (lattice constant α = 4.046 Å). The model size is expanded by replicating the unit cell along the X, Y, and Z axes to meet the simulation scale requirements. Here, Å (Ångström) is a unit of length used to measure minute distances at the atomic and molecular level; 1 Å = 10⁻⁶. -10 m.
[0021] Example commands are as follows: atomsk --create fcc 4.046 Al -duplicate 75 50 50 Al.lmp The command creates a matrix model with a size of approximately 300×200×200 Å and contains approximately 750,000 aluminum atoms, and outputs a LAMMPS-readable data file Al.lmp.
[0022] S2: In the three-dimensional space of the atomic model of the metal matrix, define and remove all atoms in the specified spherical region to form a spherical cavity in the matrix to accommodate the reinforcing particles.
[0023] In this step, the arrangement of reinforcing particles needs to be defined according to design requirements. This is achieved by specifying a set of "cavity distribution parameters," each of which includes the center coordinates (x, y, y) of the cavity. i , y i , z i ) and radius r i .
[0024] In one implementation, within the three-dimensional space of the atomic model, all atoms within a defined spherical region are defined and removed, including: Based on a preset enhanced particle distribution scheme, cavity distribution parameters for N spherical cavities are generated; the cavity distribution parameters include the center coordinates (x, y) of the i-th cavity. i , y i , z i ), radius r i and particle type identifier T i Where i = 1, 2, ..., N; for each cavity distribution parameter, from the atomic model of the metal matrix, based on the center coordinates, radius and particle type identifier of the cavity, select all atoms whose atomic coordinates are located in the spherical geometric space, and remove the selected atoms from the model.
[0025] Specifically, in combination Figure 4It shows a schematic diagram of the metal matrix model after the spherical space is removed. Using ATOMSK's selection and deletion functions, all atoms located in the corresponding spherical space are removed from the matrix model based on each cavity parameter.
[0026] An example command (for a cavity centered at (60,60,60) with a radius of 20 Å) is as follows: atomsk Al.lmp -select in sphere 60 60 60 20 -rmatom select By repeatedly performing similar operations, multiple spherical cavities can be precisely formed in the matrix, and the number, location, and size of the cavities are entirely controlled by the input parameters.
[0027] In addition, to specifically demonstrate the ability of the method of the present invention to construct models of heterogeneous, multiphase particle-reinforced composite materials, an implementation example based on the atomic-scale modeling software (Atomsk) is provided as follows: Combination Figure 5 and Figure 6 First, an aluminum (Al) matrix model with dimensions of 120 × 40 × 40 Å was constructed. Then, four spherical cavities were predefined and removed from this matrix, with their centers aligned sequentially along the x-axis and radii of 6 Å, 8 Å, 10 Å, and 12 Å, respectively. Next, initial models of silicon carbide (SiC) and elemental carbon (C) and silicon (Si) with different crystal structures were prepared, and spherical nanoparticles with radii of 5 Å, 7 Å, 9 Å, and 11 Å were extracted through geometric selection. Finally, these reinforcing particles of varying sizes and materials were sequentially embedded into the corresponding cavity positions according to the aforementioned cavity arrangement order; that is, 5 Å SiC particles, 7 Å C particles, 9 Å SiC particles, and 11 Å Si particles were filled sequentially. This successfully constructed an atomic model of a heterogeneous multiphase particle-reinforced aluminum matrix composite material containing different scales and reinforcing phases.
[0028] This implementation example demonstrates that by controlling the distribution parameters (location, size) of the cavity and the geometric and material properties of the reinforcing particles, the method of the present invention can flexibly and controllably generate highly customized simulation models of complex multiphase composite materials.
[0029] S3: Construct an atomic model of spherical reinforced particles that matches the spherical cavity.
[0030] In one implementation, a spherical reinforced particle atomic model matching the spherical cavity is constructed, including: Combination Figure 7This diagram illustrates a supercell and spherical reinforcing particle model. The crystal structure file of the reinforcing particle is obtained and converted into an atomic data file recognizable by molecular dynamics simulation software; for example, silicon carbide (SiC) reinforcing particles. First, the crystal structure file of SiC (e.g., SiC.cif) is obtained and converted into a format recognizable by simulation software (e.g., the molecular dynamics simulation software LAMMPS).
[0031] Example command: This can be done using the ATOMSK software: atomsk SiC.cif .lmp. This command converts the SiC.cif file to the LAMMPS data format file SiC.lmp.
[0032] The `-orthogonal-cell` command converts the unit cell to an orthogonal unit cell (Cartesian coordinate system) to ensure that subsequent operations are performed in the orthogonal coordinate system.
[0033] Example command: atomsk SiC.cif lmp atomsk SiC.lmp -orthogonal-cell -dup 3020 20 sic-large-supercell.lmp.
[0034] Perform a cell replication operation on the converted atomic data file (such as SiC.lmp) to construct a supercell model whose three-dimensional dimensions are all larger than the radius of the spherical cavity to be embedded.
[0035] Example command: atomsk SiC.lmp -duplicate 30 20 20 sic_supercell.lmp.
[0036] This command replicates the SiC unit cell 30×20×20 times, generating a large supercell file sic_supercell.lmp.
[0037] Using a preset point in the supercell model as the center of the sphere, and the target radius R p Define a spherical selection area with a radius (defined by the cavity distribution parameters in step S2); where the target radius R p Based on the radius R of the corresponding spherical cavity c Determined, and satisfying R p < R c For example, R p R c Smaller than 0.2-2 Å, or R p = (0.9-0.98)R c .
[0038] Extract all atoms whose atomic coordinates lie within or on the surface of a spherical selected region from the supercell model (i.e., those satisfying a distance from the center of the sphere ≤ R).p ), forming the initial spherical atomic clusters.
[0039] Example command: (Assuming the center of the sphere is at (22,22,22), and the target radius is R) p (19 Å) atomsk sic_supercell.lmp -select out sphere 22 22 22 19 -rmatomselect In this command, -select out sphere ... defines a spherical selection area, and -rmatom select performs a "reverse selection and deletion" operation: deleting atoms outside the selection area, thus retaining the atoms within the spherical selection area, forming a preliminary spherical atomic cluster, and outputting it as the SiC_sphere.lmp file.
[0040] The preliminary spherical atomic clusters obtained from the above steps are then subjected to atomic type identification and quantity counting. Finally, they are output as a single, self-contained atomic model file. This file fully contains the three-dimensional coordinate information of all atoms constituting this spherical reinforcing particle and their corresponding atomic type information, preparing it for the next step of embedding into the matrix.
[0041] At this point, a cavity that is strictly matched in size to a specific spherical cavity (R) has been created. p < R c The atomic model of the spherical enhanced particles has been completed.
[0042] S4: Embed the atomic model of the spherical reinforcing particle into the corresponding spherical cavity of the metal matrix to form the initial atomic model of the composite material.
[0043] In one implementation, the spherical reinforcing particle atomic model is embedded into the corresponding spherical cavity of the metal matrix, including: Spatial alignment: Read the center coordinates (x, y) of the current spherical cavity. i , y i , z i Based on the central coordinates, a global translation transformation is performed on the coordinates of all atoms in the spherical reinforced particle atomic model, so that the geometric center of the spherical reinforced particle atomic model is aligned with the coordinates (x...). i , y i , z i (overlap)
[0044] Example command: atomsk SiC_sphere.lmp -shift 38 38 38 SiC_sphere_shifted.lmp 1. Atom type number remapping: Combination Figure 8 This illustrates a schematic diagram of a composite material model with embedded spherical particles. This step resolves data conflicts when merging two independent models: the matrix and the particles. Specifically, it includes: First, identify the maximum defined atom type number M in the metal matrix atomic model after step S2 (i.e., the model with cavity atoms removed, such as Al_with_cavity.lmp). For example, if the matrix is only aluminum, its atom type number may be 1, in which case M=1.
[0045] Then, a preset numbering remapping rule is applied to the translated spherical reinforced particle atomic model (SiC_sphere_shifted.lmp). A preferred rule is: let the original atom type number in the particle model be j (j is a consecutive positive integer starting from 1, for example, Si is type 1 and C is type 2 in SiC), then it is remapped to a new atom type number new. t = M + j. This rule ensures that all atom type numbers are globally consecutive and unique after remapping.
[0046] This process typically requires writing scripts or utilizing software features that support atom renumbering (such as the scripting features of ATOMSK, Ovito, or the set type command in LAMMPS). For example, a program can read the particle file, iterate through each atom, calculate the new type M+j based on its original type j, and modify the corresponding field in the data file.
[0047] 2. Model Integration: The translated atomic coordinate data and mapped atomic type encoding data are concatenated with the atomic data of the metal matrix atomic model, and the box boundary information of the model is updated to generate a unified initial atomic model data file for composite materials.
[0048] Specifically: Use the merging function of the modeling software to combine the processed particle model file and the matrix model file into one file. This operation concatenates data such as the atomic coordinate lists and atomic type lists of the two models.
[0049] After merging, the size of the composite model may change. Therefore, the box boundary information in the model data file needs to be recalculated and updated based on the actual spatial distribution of all atoms after merging, to ensure that it correctly surrounds all atoms.
[0050] Finally, a standardized, formatted initial atomic model data file for the composite material (such as Composite_initial.lmp) is generated. This file contains complete atomic information for the matrix and all embedded particles, with globally consistent atom type numbering, and can be directly used as input for molecular dynamics simulation software.
[0051] S5: Perform energy minimization and structural relaxation on the initial atomic model to obtain the atomic model of particle-reinforced metal matrix composite material.
[0052] In one implementation, molecular dynamics simulation software such as LAMMPS is used, which specifically includes two stages: energy minimization and structural relaxation.
[0053] In one implementation, the initial atomic model undergoes energy minimization and structural relaxation, including: Energy minimization: The initial atomic model is imported into the molecular dynamics simulation software, and the conjugate gradient method is used to perform energy minimization iterative calculations on the model until the total potential energy of the system satisfies the first convergence criterion. The first convergence criterion is that the change in the total potential energy of the system is less than the first threshold, or the number of iterations reaches the upper limit.
[0054] Structural relaxation: Based on energy minimization, the model is placed under the NPT ensemble for molecular dynamics relaxation. The NPT ensemble is set with target temperature T and target pressure P (e.g., target temperature 300 K, zero pressure). The simulation is run until the system energy and pressure fluctuations meet the second convergence criterion, and an atomic model of particle-reinforced metal matrix composite material that has reached thermodynamic equilibrium is obtained. The second convergence criterion is that the rolling average of the total potential energy and pressure of the system fluctuates less than a second threshold within a set step size.
[0055] In one implementation, the method further includes: outputting the obtained atomic model of the particle-reinforced metal matrix composite as an atomic data file that conforms to the input requirements of at least one of the software LAMMPS, GROMACS, NAMD, or Materials Studio.
[0056] Example operation: The particle-reinforced metal matrix composite model was imported into the molecular dynamics simulation software LAMMPS, where energy minimization and relaxation stages were performed sequentially. Within LAMMPS, the model was read using the `read_data` command, energy minimization was performed using the `min_style` and `minimize` commands, and the NPT ensemble was set using the command `fix npt_relax all npt temp 300 300 0.1 iso 0 01.0`, followed by a relaxation process at a target temperature of 300 K and a target pressure of zero. Energy minimization and relaxation eliminated unreasonable atomic forces or overlaps in the initial model, achieving thermodynamic equilibrium and obtaining a stable material model suitable for subsequent molecular dynamics simulations such as material forming and cutting.
[0057] Combination Figure 9 The data output during the relaxation phase, and the final modeling result are: (1) Final model atomic number: 761,625 (of which the number of reinforcing particles is 44,000, accounting for about 3.8% of the volume); (2) Model dimensions: approximately 300×200×200 Å; (3) Particle distribution: 16 reinforcing particles are evenly distributed in the metal matrix; (4) The material model is suitable for molecular dynamics simulations such as stretching, compression, and cutting.
[0058] Optionally, to verify the geometric correctness and reliability of the cutting simulation model generated by the above method, the model file can be imported and checked using visualization software such as OVITO (Open Visualization Tool) after the model is generated. Through the visualization rendering and interactive analysis functions of this software, the spatial distribution of reinforcing particles in the matrix can be intuitively examined, confirming whether they are strictly located within the preset particle area and meet the condition of non-interference. Simultaneously, it checks for any abnormal overlap or penetration between particles and the matrix boundary or between particles, thereby directly verifying the effectiveness of the modeling algorithm and ultimately confirming the model quality.
[0059] According to the above embodiments, the specific implementation of the present invention can be summarized as two technical processes: preprocessing modeling and molecular dynamics optimization.
[0060] In the preprocessing modeling stage, based on the input matrix and reinforcing particle material parameters and cavity distribution design parameters, the following steps are executed sequentially: atomic model construction of the metal matrix, accurate generation of spherical cavities, spherical model construction of reinforcing particles, and final model embedding and data integration. The core of this stage is to construct the initial three-dimensional atomic configuration of the composite material through atomic-scale "hollowing out" and "filling" operations based on preset geometric parameters. The output is a structurally controllable initial model data file containing all atomic coordinates and type information.
[0061] In the molecular dynamics optimization stage, the initial model file is imported into the molecular dynamics simulation environment for physical correction. This stage involves the following steps: energy minimization to eliminate unreasonable short-range interatomic forces (such as overlap) that may be introduced during modeling; and structural relaxation under a constant temperature and pressure (NPT) ensemble, allowing the entire atomic system to evolve freely under set temperature and pressure conditions until the fluctuations in macroscopic quantities such as total potential energy and pressure reach a stable state, thus obtaining a stable composite material atomic model that has reached thermodynamic equilibrium. This model has eliminated initial internal stress, and the atomic arrangement conforms to physical laws, making it directly usable for subsequent molecular dynamics simulations such as stretching, compression, and cutting to predict the microscopic mechanical behavior of materials.
[0062] The above embodiments are only used to illustrate the technical solutions of this application, and are not intended to limit them. Although this application has been described in detail with reference to the foregoing embodiments, those skilled in the art should understand that modifications can still be made to the technical solutions described in the foregoing embodiments, or equivalent substitutions can be made to some of the technical features. Such modifications or substitutions do not cause the essence of the corresponding technical solutions to deviate from the spirit and scope of the technical solutions of the embodiments of this application.
Claims
1. A method for modeling particle-reinforced metal matrix composites for molecular dynamics simulations, characterized in that, The method includes: An atomic model of the metal matrix is established by combining the crystal structure information of the target matrix material; In the three-dimensional space of the atomic model of the metal matrix, all atoms in the defined spherical region are defined and removed to form a spherical cavity in the matrix to accommodate the reinforcing particles. Construct an atomic model of spherical reinforced particles that matches the spherical cavity; The atomic model of the spherical reinforcing particles is embedded into the corresponding spherical cavity of the metal matrix to form the initial atomic model of the composite material; The initial atomic model is subjected to energy minimization and structural relaxation to obtain the atomic model of particle-reinforced metal matrix composite material.
2. The method for modeling particle-reinforced metal matrix composites for molecular dynamics simulation according to claim 1, characterized in that, The process of defining and removing all atoms within a defined spherical region in the three-dimensional space of the atomic model includes: Based on a preset enhanced particle distribution scheme, cavity distribution parameters for N spherical cavities are generated; the cavity distribution parameters include the center coordinates (x, y) of the i-th cavity. i , y i , z i ), radius r i and particle type identifier T i Where i = 1, 2, ..., N; For each cavity distribution parameter, based on the cavity's center coordinates, radius, and particle type identifier, all atoms whose atomic coordinates are located within the spherical geometric space are selected from the metal matrix atomic model, and the selected atoms are removed from the model.
3. The method for modeling particle-reinforced metal matrix composites for molecular dynamics simulation according to claim 1, characterized in that, The construction of the spherical reinforced particle atomic model that matches the spherical cavity includes: Obtain the crystal structure file of the reinforcing particles and convert it into an atomic data file that can be recognized by molecular dynamics simulation software; A cell copying operation is performed on the converted atomic data file to construct a supercell model whose three-dimensional dimensions are all larger than the radius of the spherical cavity to be embedded; Using a preset point in the supercell model as the center of the sphere, and with a target radius R... p Define a spherical selection area with radius R; wherein, the target radius R p Based on the radius R of the corresponding spherical cavity c Determined, and satisfying R p < R c ; All atoms whose atomic coordinates are located within or on the surface of the spherical selection area are extracted from the supercell model to form a preliminary spherical atomic cluster. The atom types of the preliminary spherical atomic groups are identified and statistically analyzed, and the results are output as spherical enhanced particle atomic models that independently contain atomic coordinates and atom type information.
4. The method for modeling particle-reinforced metal matrix composites for molecular dynamics simulation according to claim 3, characterized in that, The embedding of the spherical reinforced particle atomic model into the corresponding spherical cavity of the metal matrix includes: Read as the center coordinates (x) of the current spherical cavity i , y i , z i Based on the center coordinates, a global translation transformation is performed on the coordinates of all atoms in the spherical reinforced particle atomic model, so that the geometric center of the spherical reinforced particle atomic model is aligned with the coordinates (x, y). i , y i , z i Overlap; By applying a preset numbering remapping rule, the atom type numbers of all atoms in the spherical enhanced particle atom model are remapped based on the maximum atom type number M to generate continuous global atom type number data. The translated atomic coordinate data and mapped atomic type encoding data are concatenated with the atomic data of the metal matrix atomic model, and the box boundary information of the model is updated to generate a unified initial atomic model data file for composite materials.
5. The method for modeling particle-reinforced metal matrix composites for molecular dynamics simulation according to claim 4, characterized in that, The preset number remapping rule is as follows: the atom with the original atom type number j in the spherical enhanced particle atom model is remapped to a new type number M + j, where j is a consecutive positive integer starting from 1.
6. The method for modeling particle-reinforced metal matrix composites for molecular dynamics simulation according to claim 1, characterized in that, The energy minimization and structural relaxation process performed on the initial atomic model includes: The initial atomic model is imported into molecular dynamics simulation software, and the conjugate gradient method is used to perform energy minimization iterative calculations on the model until the total potential energy of the system satisfies the first convergence criterion. Based on energy minimization, the model is placed under the NPT ensemble for molecular dynamics relaxation; the NPT ensemble is set with target temperature T and target pressure P, and the simulation is run until the system energy and pressure fluctuations satisfy the second convergence criterion, thus obtaining the atomic model of the particle-reinforced metal matrix composite material that has reached thermodynamic equilibrium.
7. The method for modeling particle-reinforced metal matrix composites for molecular dynamics simulation according to claim 6, characterized in that, The first convergence criterion is that the change in the total potential energy of the system is less than a first threshold, or the number of iterations reaches the upper limit. The second convergence criterion is that the rolling average of the total potential energy and pressure of the system fluctuates less than a second threshold within a set step size.
8. The method for modeling particle-reinforced metal matrix composites for molecular dynamics simulation according to claim 6, characterized in that, The method further includes: outputting the obtained atomic model of the particle-reinforced metal matrix composite material as an atomic data file that meets the input requirements of at least one of the following software: LAMMPS, GROMACS, NAMD, or Materials Studio.