Molecular dynamics model construction method and system for non-spherical rigid particles
By combining a multi-sphere combination model with LAMMPS, the problem of high computational complexity in the simulation of non-spherical particles was solved, and efficient and accurate particle packing simulation and performance evaluation were achieved.
Patent Information
- Authority / Receiving Office
- CN · China
- Patent Type
- Applications(China)
- Current Assignee / Owner
- Filing Date
- 2025-12-20
- Publication Date
- 2026-04-17
AI Technical Summary
Existing particle simulation techniques, such as the discrete element method and the Monte Carlo method, have high computational complexity and are time-consuming when simulating non-spherical particles. Furthermore, commercial or open-source simulation platforms have limited support for particles with complex shapes, resulting in inaccurate simulation results and high computational costs.
A molecular dynamics model construction method for non-spherical rigid particles is adopted. The ellipsoidal particles are geometrically approximated by a multi-sphere combination model, dynamic simulation is performed by combining LAMMPS, and performance indicators such as intersection rate are calculated by analytical module to generate three-dimensional visualization results.
It enables high-precision, automated modeling of non-spherical particles, significantly improving simulation efficiency and accuracy, reducing computational resource consumption, and providing a tool for quickly evaluating particle packing performance.
Smart Images
Figure CN121884969A_ABST
Abstract
Description
Technical Field
[0001] This invention relates to the field of model building technology, and in particular to a method and system for building molecular dynamics models of non-spherical rigid particles. Background Technology
[0002] In many fields, such as electronic component manufacturing, accurate modeling of non-spherical particle (especially ellipsoidal particle) packed systems is crucial for predicting the mechanical properties, flow characteristics, and packing density of materials. However, existing general particle simulation techniques, such as the Discrete Element Method (DEM) and the Monte Carlo Method, are typically limited to perfectly spherical particles in their standard implementations.
[0003] The Discrete Element Method (DEM) simulates system evolution by accurately calculating the interaction forces and trajectories between each particle. However, its computational accuracy is greatly affected by boundary effects and initial conditions, especially when simulating small containers or particles with complex shapes, where the simulation results are prone to distortion. The Monte Carlo method uses a large number of random samples to statistically analyze system characteristics, but its computational complexity increases dramatically when simulating high-density, large-scale particle systems, leading to high computational costs and long processing times.
[0004] While these methods, in theory, can be applied to non-spherical particles such as ellipsoids with complex modifications, existing commercial or open-source simulation platforms (such as LAMMPS) have very limited native support for particles with complex shapes. This forces researchers to use idealized spherical models to approximate irregular particles in industrial practice, resulting in models that cannot accurately capture the detailed features of real particle systems, thus affecting the accuracy and reliability of simulation results.
[0005] Insufficient simulation capability for complex-shaped particles: Mainstream open-source simulation software such as Lammps typically only supports rigid particle models for perfect spheres, failing to accurately represent more common non-spherical particles such as ellipsoids, resulting in fundamental discrepancies between the model and actual materials.
[0006] The accuracy of simulation results is limited by boundary and initial conditions: when simulating small-sized containers, the boundary effect is significant, which will affect the final arrangement of particles; at the same time, the simulation results are highly sensitive to the initial arrangement of particles, and unreasonable initialization will amplify the error of the results.
[0007] High computational resource costs: Whether it is the long relaxation process required by the discrete element method or the large number of samples required by the Monte Carlo method to achieve statistical significance, both face the problems of high computational complexity and excessive time consumption when simulating large-scale, high-density particle systems. Summary of the Invention
[0008] The purpose of this invention is to address the shortcomings of existing technologies, such as high computational resource costs: whether it is the long relaxation process required by the discrete element method or the large number of samples required by the Monte Carlo method to achieve statistical significance, both face the problems of high computational complexity and excessive time consumption when simulating large-scale, high-density particle systems. Therefore, this invention proposes a method and system for constructing molecular dynamics models for non-spherical rigid particles.
[0009] To achieve the above objectives, the present invention adopts the following technical solution: A molecular dynamics modeling system for non-spherical rigid particles includes an input module for receiving particle gradation parameters input by a user and generating particle gradation; an ensemble modeling module for geometrically approximating ellipsoidal particles using a multi-sphere combination model; a simulation module for generating and executing LAMMPS input files for dynamic simulation; an analysis module for parsing simulation results and calculating and outputting performance indicators such as intersection rate; and a visualization module for generating three-dimensional visualization results of particle packing.
[0010] In a preferred embodiment, the input module includes a graphical user interface for interactively receiving particle parameters input by the user and automatically generating particle size distribution and LAMMPS input files.
[0011] In one preferred embodiment, the analysis module processes the simulation results using a vectorized calculation method to accurately calculate the intersection rate of each particle and outputs the results to a log file. At the same time, the performance indicators such as the intersection rate are displayed in the graphical user interface as a reference for the particle packing density.
[0012] In a preferred embodiment, the input module receives particle size distribution parameters input by the user and generates a particle size distribution. Then, the ensemble modeling module performs geometric approximation modeling of ellipsoidal particles using a multi-sphere combination model. Subsequently, the simulation module performs dynamic simulation using LAMMPS. The analysis module analyzes the simulation model of the simulation module and outputs performance indicators such as the intersection rate. The visualization module generates a three-dimensional visualization result of particle packing.
[0013] A method for constructing molecular dynamics models for non-spherical rigid particles, specifically including the following steps: S1. Receive multiple powder parameters input by the user through a graphical interface, including mean particle size, standard deviation of particle size, mean aspect ratio, standard deviation of aspect ratio, density, and mass fraction, and discretize them based on a log-normal distribution to generate a particle size distribution with specific statistical characteristics. S2. For each target ellipsoidal particle, a central sphere and two symmetrical side spheres are rigidly combined to calculate and determine the diameter and position of each sphere in order to approximate the shape, volume and inertial tensor of the ellipsoidal particle. S3. Automatically generate and load the input script through LAMMPS to execute the kinetic relaxation process during the heating, cooling, and maintenance phases; S4. Based on the simulation results, automatically parse the coordinates and diameters of all spheres in the output file, and efficiently calculate the overlap volume between all spheres. Calculate the intersection rate based on the overlap volume as a performance indicator of particle packing density.
[0014] In a preferred embodiment, in step S1: S1.1. Convert the user-input mean particle size and mean aspect ratio into continuous log-normal distribution parameters, calculate the quantiles, and discretize into multiple particle size intervals and aspect ratio intervals. S1.2 Calculate the representative particle size and aspect ratio for each particle type to generate a particle gradation with specific statistical characteristics.
[0015] In a preferred embodiment, in step S2: S2.1. The target ellipsoid is approximated by a central sphere and two symmetrical lateral spheres. The diameter of the central sphere is the diameter of the minor axis of the ellipsoid, and the diameters of the two lateral spheres are twice the difference between the major axis and the minor axis of the ellipsoid. S2.2. Based on the major axis, minor axis, mass, and rotational inertia tensor of the target ellipsoid, accurately calculate the volume and geometry of the combined sphere.
[0016] In a preferred embodiment, in step S3: S3.1 Calculate the size of the simulated region based on the particle size distribution, and generate the corresponding molecular, bond and angle definitions; S3.2 Generate a LAMMPS input file containing atomic creation commands, simulation box definitions, etc., so that it can be run in the LAMMPS simulator.
[0017] In a preferred embodiment, in step S4: S4.1 The simulation process includes a heating stage for increasing the temperature of the particles to a set value through LAMMPS simulation, a cooling stage for gradually reducing the temperature of the particles to ensure that the particles reach a stable state, and a maintenance stage for maintaining the particles at a stable temperature to perform multi-step relaxation and finally obtain the particle packing result in an equilibrium state. S4.2. Using vectorized programming, calculate the vector distance between all pairs of spheres and make corrections based on periodic boundary conditions; S4.3 Calculate the cross volume between all pairs of overlapping spheres using the formula based on the sum of the distance and radius between the spheres. S4.4. The intersection rate is obtained by dividing the total cross volume by the total volume of the particle system, and is used as a quantitative basis for particle packing density. S4.5 The program automatically extracts the coordinates and diameter data of the sphere from the LAMMPS output file and parses it using an efficient algorithm; S4.6 Use the open-source molecular model viewing software Ovito to visualize the simulation results so that users can intuitively observe the results of particle accumulation.
[0018] As a preferred implementation, the program also provides a visualization interface to display each stage of the particle packing simulation process. Users can intuitively view the particle changes during the heating, cooling, and maintenance processes and generate corresponding three-dimensional visualization images.
[0019] The beneficial effects of this invention are as follows: This invention achieves high-precision, automated modeling of non-spherical particles, from abstract parameters to concrete simulation. It automatically transforms abstract statistical distribution parameters into a physically realistic particle model library that can be directly used for simulation, overcoming the shortcomings of inefficient manual modeling and the lack of native support on general platforms in existing technologies. Through a three-sphere combination model and its volume correction algorithm, the accuracy of geometric simulation is significantly improved while maintaining computational efficiency. Full-process automation greatly improves simulation efficiency and ease of use, and reduces human error: the complex process that originally required deep expertise and involved script writing, file preparation, command execution, and result processing is integrated into a simple "input parameters - one-click run - get results" mode. This reduces the preparation time for a single simulation from several hours to minutes and completely eliminates errors caused by manual operation; By automating and standardizing the calculation of the key indicator "intersection rate" and linking it with parameterized input, engineers can quickly evaluate the packing performance of different gradation schemes, thus providing an efficient, reliable, and practical tool for computer-aided design and optimization of material formulations. Attached Figure Description
[0020] Figure 1 This is a schematic diagram of a molecular dynamics model construction system for non-spherical rigid particles according to the present invention.
[0021] Figure 2 This is a schematic diagram illustrating the use of multiple rigid spheres combined to construct an approximate ellipsoid in an embodiment of the present invention.
[0022] Figure 3This is a schematic diagram of the result of the approximately ellipsoidal particle close packing model constructed using the present invention in an embodiment of the present invention. Detailed Implementation
[0023] The technical solutions of the present invention will be clearly and completely described below with reference to the accompanying drawings of the embodiments of the present invention. Obviously, the described embodiments are only some embodiments of the present invention, and not all embodiments.
[0024] like Figure 1 As shown, a molecular dynamics modeling system for non-spherical rigid particles includes an input module for receiving particle gradation parameters input by the user and generating particle gradation; an ensemble modeling module for geometrically approximating ellipsoidal particles using a multi-sphere combination model; a simulation module for generating and executing LAMMPS input files for dynamic simulation; an analysis module for parsing simulation results and calculating and outputting performance indicators such as intersection rate; and a visualization module for generating three-dimensional visualization results of particle packing.
[0025] The input module includes a graphical user interface for interactively receiving particle parameters input by the user and automatically generating particle size distribution and LAMMPS input files.
[0026] The analysis module processes the simulation results using a vectorized calculation method to accurately calculate the intersection rate of each particle and outputs the results to a log file. At the same time, the performance indicators such as the intersection rate are displayed in the graphical user interface as a reference for the particle packing density.
[0027] The input module receives the particle size distribution parameters input by the user and generates the particle size distribution. Then, the ensemble modeling module performs geometric approximation modeling of the ellipsoidal particles using a multi-sphere combination model. The simulation module then performs dynamic simulation using LAMMPS. The analysis module analyzes the simulation model of the simulation module and outputs performance indicators such as the intersection rate. The visualization module generates a three-dimensional visualization result of the particle packing.
[0028] In the above embodiments: 1) A geometric modeling method that uses a central sphere and two symmetrical side spheres to form a rigid spatial combination to approximate ellipsoidal particles and calculate their accurate volume and inertia tensor; 2) A parameterization method that discretizes the user-input log-normal distribution parameters into multiple sets of specific particle sizes and aspect ratios, and automatically generates a corresponding number of LAMMPS molecular template files containing bond and angle constraints; 3) An integrated automated simulation system that can automatically calculate the simulation domain and particle number based on user input, synthesize a complete LAMMPS input script, automatically call the calculation core, and finally automatically parse the result data and calculate the intersection rate.
[0029] When calculating the intersection rate, an efficient algorithm based on vectorized operations and strictly handling the inter-sphere cross volume under periodic boundary conditions is used.
[0030] Through the above embodiments, a method is established to approximate the shape of ellipsoidal particles by rigidly combining multiple spheres in space. By adjusting the relative positions and sizes of the constituent spheres, ellipsoids with different aspect ratios can be flexibly approximated, thereby constructing a physical model that is closer to the actual particle shape. We will build an integrated visualization program based on the Windows system, providing users with a graphical interface to input powder gradation information. It can automatically generate the input files required by LAMMPS based on user input, automatically call the LAMMPS core program for calculation, and automatically analyze and visualize the results after the calculation is completed, which greatly reduces the threshold for use and improves work efficiency. By combining the aforementioned multi-sphere combination model with automated processes, a stable method for simulating non-spherical particles that can be executed on a general simulation platform is provided to users while ensuring geometric accuracy and physical realism. This effectively controls the consumption of computing resources and promotes the practical application of this technology in industry.
[0031] A method for constructing molecular dynamics models for non-spherical rigid particles, specifically including the following steps: S1. Receive multiple powder parameters input by the user through a graphical interface, including mean particle size, standard deviation of particle size, mean aspect ratio, standard deviation of aspect ratio, density, and mass fraction, and discretize them based on a log-normal distribution to generate a particle size distribution with specific statistical characteristics. S2. For each target ellipsoidal particle, a central sphere and two symmetrical side spheres are rigidly combined to calculate and determine the diameter and position of each sphere in order to approximate the shape, volume and inertial tensor of the ellipsoidal particle. S3. Automatically generate and load the input script through LAMMPS to execute the kinetic relaxation process during the heating, cooling, and maintenance phases; S4. Based on the simulation results, automatically parse the coordinates and diameters of all spheres in the output file, and efficiently calculate the overlap volume between all spheres. Calculate the intersection rate based on the overlap volume as a performance indicator of particle packing density.
[0032] In step S1: S1.1. Convert the user-input mean particle size and mean aspect ratio into continuous log-normal distribution parameters, calculate the quantiles, and discretize into multiple particle size intervals and aspect ratio intervals. S1.2 Calculate the representative particle size and aspect ratio for each particle type to generate a particle gradation with specific statistical characteristics.
[0033] In step S2: S2.1. The target ellipsoid is approximated by a central sphere and two symmetrical lateral spheres. The diameter of the central sphere is the diameter of the minor axis of the ellipsoid, and the diameters of the two lateral spheres are twice the difference between the major axis and the minor axis of the ellipsoid. S2.2. Based on the major axis, minor axis, mass, and rotational inertia tensor of the target ellipsoid, accurately calculate the volume and geometry of the combined sphere.
[0034] In step S3: S3.1 Calculate the size of the simulated region based on the particle size distribution, and generate the corresponding molecular, bond and angle definitions; S3.2 Generate a LAMMPS input file containing atomic creation commands, simulation box definitions, etc., so that it can be run in the LAMMPS simulator.
[0035] In step S4: S4.1 The simulation process includes a heating stage for increasing the temperature of the particles to a set value through LAMMPS simulation, a cooling stage for gradually reducing the temperature of the particles to ensure that the particles reach a stable state, and a maintenance stage for maintaining the particles at a stable temperature to perform multi-step relaxation and finally obtain the particle packing result in an equilibrium state. S4.2. Using vectorized programming, calculate the vector distance between all pairs of spheres and make corrections based on periodic boundary conditions; S4.3 Calculate the cross volume between all pairs of overlapping spheres using the formula based on the sum of the distance and radius between the spheres. S4.4. The intersection rate is obtained by dividing the total cross volume by the total volume of the particle system, and is used as a quantitative basis for particle packing density. S4.5 The program automatically extracts the coordinates and diameter data of the sphere from the LAMMPS output file and parses it using an efficient algorithm; S4.6 Use the open-source molecular model viewing software Ovito to visualize the simulation results so that users can intuitively observe the results of particle accumulation.
[0036] The program also provides a visualization interface to display each stage of the particle packing simulation process. Users can intuitively view the particle changes during the heating, cooling, and maintenance processes and generate corresponding three-dimensional visualization images.
[0037] Through the above embodiments, high-precision and automated modeling of non-spherical particles, from abstract parameters to concrete simulation, is achieved. This invention automatically transforms abstract statistical distribution parameters into a physically realistic particle model library that can be directly used for simulation, overcoming the shortcomings of inefficient manual modeling and the lack of native support on general platforms in existing technologies. By using a three-sphere combination model and its volume correction algorithm, the accuracy of geometric simulation is significantly improved while ensuring computational efficiency. Full-process automation greatly improves simulation efficiency and ease of use, and reduces human error: the complex process that originally required deep expertise and involved script writing, file preparation, command execution, and result processing is integrated into a simple "input parameters - one-click run - get results" mode. This reduces the preparation time for a single simulation from several hours to minutes and completely eliminates errors caused by manual operation; By automating and standardizing the calculation of the key indicator of intersection rate and linking it with parametric input, engineers can quickly evaluate the packing performance of different gradation schemes, thus providing an efficient, reliable and practical tool for computer-aided design and optimization of material formulations.
[0038] Three-sphere model construction: such as Figure 2 As shown, for a target ellipsoid (major axis D, minor axis D / A) R (Using three spheres for approximation: Central sphere: Centered at the center of the ellipsoid, with a diameter set to D / A) R This is used to simulate the minor axis diameter of an ellipsoid. Side spheres: The centers of the two side spheres are located at (D / 2A) respectively. R , 0, 0) and (-D / 2A R (, 0, 0), diameter set to DD / A R This is used to extend and simulate the major axis of the ellipsoid. Physical property calculation: Based on the ellipsoid's geometry, its volume, mass, and rotational inertia tensor are accurately calculated. Combined volume correction: To accurately calculate the volume of the combined model, the system considers the cross-volume between the spheres and corrects it using the formula: Total Volume = Central Sphere Volume + 2 * Side Sphere Volume - 2 * Cross-volume, resulting in a high-precision combined volume.
[0039] Working Principle: In this invention, the user inputs relevant parameters of the powder through a graphical interface, including mean particle size, standard deviation of particle size, mean aspect ratio, standard deviation of aspect ratio, density, and mass fraction. The system discretizes these parameters using a log-normal distribution, generating multiple particle size and aspect ratio intervals, ultimately generating a particle size distribution. For ellipsoidal particles in each distribution, the system uses a three-sphere model for geometric approximation. This model approximates the geometry of the target ellipsoid by calculating the size and position of the central and lateral spheres. Based on the generated particle size distribution information, the system dynamically calculates the size of the simulation area and the quantity of each type of particle. The simulation first enters a heating stage, where the temperature of the particle system gradually increases, and the interactions between particles gradually activate. Next, it enters a cooling stage, where the system temperature slowly decreases, and the particles reach an equilibrium state. Finally, it enters a maintenance stage, where the temperature remains constant, and the particle system relaxes for a long time, eventually reaching an equilibrium state. After the simulation ends, the system automatically parses the output relaxed.data file, which contains the coordinates and diameter information of all particles. The program calculates the intersection volume between all particle pairs using a vectorized programming method. During this process, the program strictly considers periodic boundary conditions, determines which spheres overlap by calculating the vector distance between all spheres, and calculates their intersection volume. After the simulation is completed, the system uses the open-source molecular model viewing software Ovito to read the coordinate data output by LAMMPS and generate a three-dimensional visualization image of the particle packing, specifically as follows: Figure 3 As shown.
[0040] The above description is only a preferred embodiment of the present invention, but the scope of protection of the present invention is not limited thereto. Any equivalent substitutions or modifications made by those skilled in the art within the scope of the technology disclosed in the present invention, based on the technical solution and inventive concept of the present invention, should be covered within the scope of protection of the present invention.
Claims
1. A molecular dynamics modeling system for non-spherical rigid particles, characterized in that, The system includes an input module for receiving particle size distribution parameters from the user and generating particle size distributions; an ensemble modeling module for geometrically approximating ellipsoidal particles using a multi-sphere combination model; a simulation module for generating and executing LAMMPS input files for dynamic simulation; an analysis module for parsing simulation results and calculating and outputting performance indicators such as intersection rate; and a visualization module for generating three-dimensional visualization results of particle packing.
2. The system for constructing a molecular dynamics model of non-spherical rigid body particles according to claim 1, wherein, The input module includes a graphical user interface for interactively receiving particle parameters input by the user and automatically generating particle size distribution and LAMMPS input files.
3. The molecular dynamics model construction system for non-spherical rigid particles according to claim 1, characterized in that, The analysis module processes the simulation results using a vectorized calculation method to accurately calculate the intersection rate of each particle and outputs the results to a log file. At the same time, the performance indicators such as the intersection rate are displayed in the graphical user interface as a reference for the particle packing density.
4. The molecular dynamics model construction system for non-spherical rigid particles according to claim 1, characterized in that, The input module receives the particle size distribution parameters input by the user and generates the particle size distribution. Then, the ensemble modeling module performs geometric approximation modeling of the ellipsoidal particles using a multi-sphere combination model. The simulation module then performs dynamic simulation using LAMMPS. The analysis module analyzes the simulation model of the simulation module and outputs performance indicators such as the intersection rate. The visualization module generates a three-dimensional visualization result of the particle packing.
5. A method for constructing molecular dynamics models for non-spherical rigid particles, characterized in that, The molecular dynamics modeling system for non-spherical rigid particles, as described in any one of claims 1-4, specifically includes the following steps: S1. Receive multiple powder parameters input by the user through a graphical interface, including mean particle size, standard deviation of particle size, mean aspect ratio, standard deviation of aspect ratio, density, and mass fraction, and discretize them based on a log-normal distribution to generate a particle size distribution with specific statistical characteristics. S2. For each target ellipsoidal particle, a central sphere and two symmetrical side spheres are rigidly combined to calculate and determine the diameter and position of each sphere in order to approximate the shape, volume and inertial tensor of the ellipsoidal particle. S3. Automatically generate and load the input script through LAMMPS to execute the kinetic relaxation process during the heating, cooling, and maintenance phases; S4. Based on the simulation results, automatically parse the coordinates and diameters of all spheres in the output file, and efficiently calculate the overlap volume between all spheres. Calculate the intersection rate based on the overlap volume as a performance indicator of particle packing density.
6. The method for constructing a molecular dynamics model for non-spherical rigid particles according to claim 5, characterized in that, In step S1: S1.
1. Convert the user-input mean particle size and mean aspect ratio into continuous log-normal distribution parameters, calculate the quantiles, and discretize into multiple particle size intervals and aspect ratio intervals. S1.2 Calculate the representative particle size and aspect ratio for each particle type to generate a particle gradation with specific statistical characteristics.
7. The method for constructing a molecular dynamics model for non-spherical rigid particles according to claim 5, characterized in that, In step S2: S2.
1. The target ellipsoid is approximated by a central sphere and two symmetrical lateral spheres. The diameter of the central sphere is the diameter of the minor axis of the ellipsoid, and the diameters of the two lateral spheres are twice the difference between the major axis and the minor axis of the ellipsoid. S2.
2. Based on the major axis, minor axis, mass, and rotational inertia tensor of the target ellipsoid, accurately calculate the volume and geometry of the combined sphere.
8. The method for constructing a molecular dynamics model for non-spherical rigid particles according to claim 5, characterized in that, In step S3: S3.1 Calculate the size of the simulated region based on the particle size distribution, and generate the corresponding molecular, bond and angle definitions; S3.2 Generate a LAMMPS input file containing atomic creation commands, simulation box definitions, etc., so that it can be run in the LAMMPS simulator.
9. A method for constructing a molecular dynamics model for non-spherical rigid particles according to claim 5, characterized in that, In step S4: S4.1 The simulation process includes a heating stage for increasing the temperature of the particles to a set value through LAMMPS simulation, a cooling stage for gradually reducing the temperature of the particles to ensure that the particles reach a stable state, and a maintenance stage for maintaining the particles at a stable temperature to perform multi-step relaxation and finally obtain the particle packing result in an equilibrium state. S4.
2. Using vectorized programming, calculate the vector distance between all pairs of spheres and make corrections based on periodic boundary conditions; S4.3 Calculate the cross volume between all pairs of overlapping spheres using the formula based on the sum of the distance and radius between the spheres. S4.
4. The intersection rate is obtained by dividing the total cross volume by the total volume of the particle system, and is used as a quantitative basis for particle packing density. S4.5 The program automatically extracts the coordinates and diameter data of the sphere from the LAMMPS output file and parses it using an efficient algorithm; S4.6 Use the open-source molecular model viewing software Ovito to visualize the simulation results so that users can intuitively observe the results of particle accumulation.
10. A method for constructing a molecular dynamics model for non-spherical rigid particles according to claim 5, characterized in that, The program also provides a visualization interface to display each stage of the particle packing simulation process. Users can intuitively view the particle changes during the heating, cooling, and maintenance processes and generate corresponding three-dimensional visualization images.