A post-processing display method to improve the visualization quality of particle simulation results

By constructing a three-dimensional structured background mesh and performing kernel function interpolation and isosurface extraction in the post-processing stage of particle simulation results, the discrete coarseness of free surface display in particle methods and the interface diffusion problem in mesh methods are solved, achieving high-quality three-dimensional visualization effects.

CN123118971APending Publication Date: 2026-11-17JIANGSU OCEAN UNIV
View PDF 0 Cites 0 Cited by

Patent Information

Application Number
CN202611438526.7
Authority / Receiving Office
CN · China
Patent Type
Applications(China)
Current Assignee / Owner
Filing Date
2026-09-16
Publication Date
2026-11-17

AI Technical Summary

Technical Problem

Existing particle numerical methods suffer from poor visualization quality in post-processing, including discrete coarseness in free surface display, lack of systematic mesh interpolation and extraction processes for color mapping, and interface diffusion in mesh methods during large deformation flows.

Method used

A three-dimensional structured background mesh is constructed, and particle physical quantities are mapped to mesh nodes through kernel function interpolation. A contour surface extraction algorithm is used to generate free surface mesh data, which is then smoothed to achieve continuous and smooth three-dimensional free surface display.

Benefits of technology

Without altering the original particle calculation results, a smooth and continuous free surface display effect is achieved, avoiding interface diffusion and quality degradation common in mesh-based methods, and supporting the display of lighting, materials, and physical quantities.

✦ Generated by Eureka AI based on patent content.

Smart Images

  • Figure CN123118971A_ABST
    Figure CN123118971A_ABST
Patent Text Reader

Abstract

This invention provides a post-processing display method to improve the visualization quality of particle simulation results, relating to the field of scientific computing visualization technology. The method includes: reading particle simulation result files to obtain particle spatial coordinates at each time step and physical quantities including at least density; constructing a three-dimensional structured background mesh covering the computational domain based on the particle coordinate range and setting the resolution; mapping particle physical quantities to mesh nodes using kernel function interpolation to obtain a physical quantity field, and calculating a labeling function or volume fraction to determine free surfaces; setting a threshold based on the physical quantity field, and generating a free surface mesh containing vertices, normals, and topological relationships using an isosurface extraction algorithm; smoothing the mesh to eliminate jagged edges and fluctuations; and then importing it into a rendering engine for three-dimensional free surface display. This invention solves the problems of discrete coarseness in direct particle rendering, lack of a systematic mesh interpolation extraction process for color mapping, and poor visualization quality caused by mesh interface diffusion.
Need to check novelty before this filing date? Find Prior Art

Description

Technical Field

[0001] This invention relates to the field of scientific computing visualization technology, and in particular to a post-processing display method for improving the visualization quality of particle simulation results. Background Technology

[0002] Particle-based numerical methods belong to the Lagrangian meshless method category and have natural advantages in simulating large deformation flows, free surface breakage, splashing, and fluid-structure interaction problems. They have been widely used in numerical simulations in fields such as marine engineering, shipbuilding engineering, and hydraulic engineering.

[0003] Currently, post-processing visualization of existing particle numerical methods mainly relies on the following three methods: the first is direct particle rendering, which involves directly rendering and displaying 3D data through particle point clouds or particle spheres, such as... Figure 1 As shown; the second is color mapping based on particle physical quantities, that is, in general visualization software (such as ParaView) or the post-processing module of each solver, color mapping is performed based on particle position or physical quantities (such as velocity, density); the third is the grid-based CFD method, which usually uses VOF (volume fraction method) or Level Set method to track free surfaces, and its post-processing can obtain a smooth and continuous free surface display effect through isosurface extraction algorithms (such as Marching Cubes).

[0004] However, the above-mentioned technical methods all have obvious shortcomings in actual use, specifically: Firstly, in direct particle rendering, since the particles themselves are discrete Lagrange sampling points, the free surface will exhibit discrete, rough, and jagged features, and there are visible gaps between particles. It is impossible to present the continuous and smooth shape of the fluid free surface. Especially in 3D display scenes, the wave surface topology is blurred, making it difficult to clearly observe the true 3D shape and evolution process of the fluid free surface.

[0005] Secondly, while color mapping based on particle physical quantities can provide some visualization information, existing general visualization software or solver modules mainly support scatter plots or particle sphere rendering. They lack a complete post-processing workflow that systematically interpolates particle information to the background mesh and extracts free surfaces. It is difficult to learn from mature isosurface extraction techniques in mesh-based methods (such as Marching Cubes) to obtain smooth and continuous free surface display effects, which limits the depth of visualization analysis.

[0006] Third, although mesh-based CFD methods can obtain smooth and continuous free surface display effects through isosurface extraction, such methods face problems such as interface diffusion and numerical dissipation when dealing with complex problems such as large deformation flow, free surface breakage, and splashing. This results in insufficient capture of small-scale vortices and detailed features. In addition, the computation and storage overhead of dense meshes is large, which limits their application in scenarios with drastic changes in free surface topology.

[0007] Therefore, how to improve the 3D visualization quality of free surfaces in particle simulation results by drawing on the post-processing techniques of mesh methods to extract free surfaces while retaining the computational advantages of particle methods is a technical problem that urgently needs to be solved in this field. Summary of the Invention

[0008] To address the shortcomings of existing technologies, this invention provides a post-processing display method to improve the visualization quality of particle simulation results. It solves the problems of poor 3D visualization quality caused by the discrete and coarse display of free surfaces in the post-processing of existing particle rendering methods, the lack of a systematic mesh interpolation and extraction process in color mapping methods, and the interface diffusion in mesh methods.

[0009] According to an embodiment of the present invention, a post-processing display method for improving the visualization quality of particle simulation results, which only applies to the post-processing display stage of particle simulation results and does not change the original particle calculation results, includes the following steps: Read the result file of the particle numerical simulation output, and obtain the spatial coordinates and physical quantities of each particle at each output time. The physical quantities include at least density. Based on the particle computational domain spatial range determined by the minimum and maximum values ​​of all particle coordinates, a three-dimensional structured background mesh covering the entire computational domain is constructed, and the mesh resolution is set according to the display accuracy requirements. The physical quantities carried by each particle are interpolated to each node of the three-dimensional structured background grid using the kernel function interpolation method to obtain the physical quantity field on the grid node. At the same time, the identification function or volume fraction for identifying the fluid region is calculated on each grid node to facilitate the determination of the subsequent free surface. On the three-dimensional structured background mesh, a free surface determination threshold is set based on the physical quantity field obtained by interpolation, and an isosurface extraction algorithm is used to extract the free surface isosurface, generating free surface mesh data composed of triangular facets. The free surface mesh data includes vertex coordinates, normal vectors, and facet topological connection relationships. The extracted free surface mesh data is smoothed to eliminate jagged edges and fluctuations caused by mesh discretization and particle noise; The processed free surface mesh data is imported into the visualization rendering engine for 3D free surface display.

[0010] Compared with the prior art, the present invention has the following beneficial effects: By constructing a three-dimensional structured background mesh, particle physical quantities are mapped to mesh nodes using kernel function interpolation to form a continuous physical quantity field. Based on this field, a free surface threshold is set, and an isosurface extraction algorithm is used to generate a triangular free surface mesh containing vertices, normals, and topological relationships. After smoothing, a "continuous curved surface topology" is used to replace "discrete particle display," outputting a smooth, continuous free surface mesh that can carry normal and lighting information. This eliminates particle gaps and spherical accumulation, resulting in a high-quality three-dimensional display effect.

[0011] The process involves reading the result file to obtain particle coordinates and physical quantities, constructing a background mesh covering the computational domain and setting the resolution as needed, interpolating the nodal physical quantity field using a kernel function, calculating the identifier function or volume fraction to determine the free surface, and then performing threshold setting, isosurface extraction, smoothing, and rendering. This forms a complete post-processing workflow of "reading → mesh construction → interpolation → identifier → extraction → smoothing → rendering", rather than sporadic color mapping, giving the free surface a clear topology and supporting the display of lighting, materials, and physical quantities.

[0012] By limiting this method to the post-processing and display stage of particle simulation results without altering the original particle calculation results, and by reading the already solved and output result file, and by interpolating only to reconstruct the display field on the background mesh, this method retains the calculation accuracy of the particle method for free surfaces, large deformations, and splashing problems while obtaining continuous, smooth, and physically-property-enabled triangular free surfaces in the display stage. This avoids interface diffusion and quality degradation caused by full-field solving or resampling of meshes, achieving mesh-level display quality without changing the original calculation. Attached Figure Description

[0013] Figure 1 This is a schematic diagram illustrating the display effect of a 3D free surface rendered directly from traditional particles.

[0014] Figure 2 This is a flowchart illustrating the overall process steps of a post-processing display method for improving the visualization quality of particle simulation results according to Embodiment 1 of the present invention.

[0015] Figure 3 This is a schematic diagram illustrating the principle of interpolating particle information to background grid nodes in Embodiment 1 of the present invention (blue circles represent particle results, red represents background grid nodes, and blue lines represent fluid free surfaces).

[0016] Figure 4 This is a schematic diagram of free surface isosurface extraction based on the Marching Cubes algorithm in Embodiment 1 of the present invention.

[0017] Figure 5This is a schematic diagram of the three-dimensional free surface display effect of a post-processing display method for improving the visualization quality of particle simulation results according to Embodiment 2 of the present invention. Detailed Implementation

[0018] The technical solutions of the present invention will be further described below with reference to the accompanying drawings and embodiments.

[0019] Example 1 like Figure 2 As shown in the figure, this embodiment of the invention proposes a post-processing display method to improve the visualization quality of particle simulation results. This method only applies to the post-processing display stage of particle simulation results and does not change the original particle calculation results. The method includes the following steps: S1. Read the result file of the particle numerical simulation output, and obtain the spatial coordinates and physical quantities of each particle at each output time. The physical quantities include at least density. The particle numerical simulation is preferably a smoothed particle hydrodynamic (SPH) simulation, the particles in the result file are SPH particles, and the physical quantities include at least density. Pressure P, velocity vector v, particle mass m, and particle number.

[0020] Specifically, the SPH method is a meshless Lagrangian particle method where the fluid domain is represented by a large number of discrete particles, each carrying spatial position coordinates (x, y, z) and related physical quantities, including but not limited to density. After obtaining the simulation results of the SPH method, the system only reads the already solved and output result file, without participating in or modifying the internal calculation process of the SPH solver, including parameters such as pressure P, velocity vector v, particle mass m, and particle number. In problems such as free surface flow, large deformation, and splashing, the SPH method does not require pre-fixed meshes. Therefore, it can obtain high-quality 3D display effects in the display stage while maintaining the calculation accuracy of complex interfaces.

[0021] S2. Based on the particle computation domain spatial range determined by the minimum and maximum values ​​of all particle coordinates, construct a three-dimensional structured background mesh covering the entire computation domain, and set the mesh resolution according to the display accuracy requirements. Specifically, the minimum and maximum coordinates of all particles in the x, y, and z directions are set as follows: , , )and( , , Then the background grid covers ( , )×( , )×( , The entire bounding box region of the background mesh is a three-dimensional structured mesh, which is a set of cubes / cubic prisms arranged at equal or set intervals along the three coordinate axes to carry the Euler mesh data field obtained by subsequent interpolation.

[0022] It is worth noting that the grid resolution is set according to the display accuracy requirements. For example, the grid node spacing can be 0.5-2 times the initial particle spacing. In areas with intense free surface activity, an adaptive densification strategy can be adopted to improve local display accuracy.

[0023] S3. Using the kernel function interpolation method, the physical quantities carried by each particle are interpolated to each node of the three-dimensional structured background grid to obtain the physical quantity field on the grid node. At the same time, the identification function or volume fraction for identifying the fluid region is calculated on each grid node to facilitate the determination of the subsequent free surface. Specifically, such as Figure 3 As shown, blue circles represent SPH particles and the physical quantities they carry, red squares represent nodes of the 3D structured background mesh, and blue curves represent the fluid free surface. For each background mesh node i, all particles j are searched within the kernel function support domain determined by its smooth length h. The physical quantities carried by the particles are interpolated onto the mesh node by weighted summation through the kernel function W.

[0024] The physical quantity field at the obtained grid node is used to calculate the physical quantity value of the node using a kernel function weighted summation method. The specific calculation formula is as follows:

[0025] in, Let i be the target physical quantity at grid node i; Let j be the mass of particle j; Let be the density of particle j; Let be the physical quantity carried by particle j; W be the kernel function (preferably a cubic spline kernel function, a Gaussian kernel function, or a quintic kernel function); h be the smooth length. , These are the spatial position vectors of grid node i and particle j, respectively.

[0026] This formula originates from the SPH kernel approximation idea: a smooth kernel function W is used to perform a weighted average of the surrounding particles in the support domain, thereby converting the discrete information of Lagrange particles into continuous field values ​​on the grid nodes, thus realizing the conversion from Lagrange particle data to Eulerian grid data field.

[0027] Simultaneously, the fluid identification function or volume fraction F is calculated on each grid node. When calculating the volume fraction F, if there are particles in the node's support domain and the interpolation density is greater than a set threshold, the node's support domain is marked as a fluid region, and F=1. Otherwise, the node's support domain is marked as an empty region, and F=0. Meanwhile, the transition region consists of continuous values ​​between 0 and 1, which are used for subsequent free surface determination.

[0028] S4. On the three-dimensional structured background mesh, a free surface determination threshold is set based on the physical quantity field obtained by interpolation, and the free surface isosurface is extracted using the isosurface extraction algorithm to generate free surface mesh data composed of triangular facets. The free surface mesh data includes vertex coordinates, normal vectors and facet topological connection relationships. The free surface determination threshold is 30%-70% of the fluid reference density, or the volume fraction F=0.5. The isosurface extraction algorithm is one of Marching Cubes algorithm, Marching Tetrahedra algorithm, or asymptote decoder.

[0029] Specifically, the Marching Cubes algorithm is a classic algorithm for extracting isosurfaces from 3D volume data. Specifically, it processes each voxel one by one, finds the corresponding triangulation pattern based on the relative relationship between the scalar values ​​of the eight vertices of the voxel and the threshold, and linearly interpolates the intersection points on the edges, thereby converting the scalar field into a surface mesh composed of triangular facets.

[0030] like Figure 4 As shown, when using the Marching Cubes algorithm, at the eight vertices of a single voxel (cube mesh cell), vertices are classified into internal and external vertices based on the free face determination threshold. Each of the eight vertices of the voxel has an interpolated scalar value (such as density or volume fraction), as shown in the diagram for example, the vertices... The scalar value is 0.8, which is greater than the threshold. The vertex is classified as an internal vertex, while the scalar values ​​of the remaining 7 vertices are less than the threshold. It is classified as an external vertex.

[0031] Then, on the edge where the scalar value crosses the threshold, the intersection point of the isosurface and the edge is obtained by linear interpolation, such as the three intersection points P01, P03, and P04 in the figure. Then, the adjacent intersection points on the same cube are connected to form a triangular facet, which is the free surface isosurface fragment in the voxel. The triangular facets of all voxels are combined to form the complete free surface mesh data.

[0032] S5. The extracted free surface mesh data is smoothed to eliminate jagged edges and fluctuations caused by mesh discretization and particle noise. Specifically, the smoothing process for the extracted free surface mesh data includes: The extracted free surface mesh data is smoothed using one or more of the following methods: Laplacian smoothing, curvature-based smoothing, or Humphrey Classes smoothing. During the smoothing process, volume conservation constraints are set to prevent the overall shape of the free surface from shifting.

[0033] Specifically, Laplacian smoothing is simple to implement and computationally fast by moving each vertex to the geometric center of its adjacent vertices and iterating several times, making it ideal for quick previews or light denoising; curvature-based smoothing adjusts vertex displacements according to local curvature, eliminating high-frequency noise while preserving the large-scale shape of free surfaces; Humphrey Classes smoothing is specifically designed for mesh smoothing by adding a "volume-preserving" correction step after the Laplacian smoothing step, which preserves the original details of the mesh's geometry and volume while avoiding the shrinkage and deformation caused by traditional Laplacian smoothing.

[0034] It is worth noting that during the smoothing process, a volume conservation constraint is set to prevent the smoothing process from causing a shift in the overall shape of the free surface.

[0035] S6. Import the processed free surface mesh data into the visualization rendering engine for 3D free surface display.

[0036] Specifically, importing the processed free-face mesh data into the visualization rendering engine for 3D free-face display includes: The processed free surface mesh data is imported into the visualization rendering engine, and rendering parameters such as lighting model settings, material shading, and transparency adjustment are configured to realize the display of three-dimensional free surfaces. The lighting model is either the Phong lighting model or the PBR physical rendering model.

[0037] Furthermore, when importing the processed free surface mesh data into the visualization rendering engine for three-dimensional free surface display, post-processing elements are also superimposed simultaneously to achieve three-dimensional visualization display with multi-element fusion. The post-processing elements include one or more of particle position, velocity vector, pressure cloud map, or floating body structure.

[0038] In this embodiment, the post-processing display method for improving the visualization quality of particle simulation results of the present invention has the following beneficial effects: (1) Solve the problem of rough discrete free surfaces caused by direct particle rendering: Traditional rendering of particles as points, spheres, or scattered dots results in only discrete particle sets, leading to gaps, jagged edges, and visual clutter. This invention constructs a 3D structured background mesh, interpolates it using kernel functions to obtain a continuous physical field, and then uses an isosurface extraction algorithm (such as Marching Cubes) to generate a free-face mesh of triangular facets containing vertices, normals, and topological relationships. This is further smoothed, thus replacing "discrete particle display" with a "continuous curved surface topology," resulting in smooth, continuous 3D free-facets capable of carrying lighting information.

[0039] (2) Solving the problem of the lack of a systematic mesh interpolation and extraction process in color mapping methods: If only color mapping or simple mapping is performed on particles, only the drawing level processing is completed, lacking a systematic process from particles to fields, and from fields to continuous surface meshes. This invention forms a complete post-processing chain of "reading result files → constructing background meshes → kernel function interpolation → calculating identifier functions / volume fractions → setting thresholds → isosurface extraction → triangular patch mesh generation → smoothing → rendering and displaying", so that free surfaces have clear vertex coordinates, normal vectors and patch topological connections, and can display lighting, material, transparency and physical quantity distribution.

[0040] (3) Solve the problems of interface diffusion and poor visualization quality in grid-type methods: This invention explicitly applies only to the post-processing and display stage of particle simulation results, without altering the original particle calculation results. Furthermore, it reads the result file that has already been solved and output by SPH. Interpolation is only used to reconstruct the field on the background mesh for display, without recalculating the interface during the solution process using the mesh method. Therefore, while preserving the calculation accuracy of SPH for problems such as free surfaces, large deformations, and splashing, it obtains continuous, smooth, and physically-property-enabled triangular free surfaces at the display stage, avoiding interface diffusion and display quality degradation common in mesh-based full-field solutions or intermediate resampling methods.

[0041] (4) Multi-element integration and practical engineering effects: While displaying three-dimensional free surfaces, post-processing elements such as particle positions, velocity vectors, pressure cloud maps, or floating body structures can be superimposed to achieve multi-element fusion three-dimensional visualization, thus enabling the mechanistic display and result interpretation of complex engineering scenarios such as dam breaks, surges, and floating body interactions.

[0042] Example 2 Based on Example 1, the specific implementation process of the present invention is explained using the classic three-dimensional dam-break problem as the application object.

[0043] The SPH method was used to numerically simulate the dam failure process in a water tank. The initial particle spacing dp = 0.01 m, the total number of particles was approximately 500,000, and the output time interval was 0.01 s. Each particle contained spatial coordinates (x, y, z) and related physical quantities, including density. Pressure P, velocity vector v, particle mass m, and particle number, etc.

[0044] S1. Read the particle result data at a certain output time and obtain the coordinates and physical quantities of all particles.

[0045] S2. Construct a three-dimensional structured background mesh based on the particle coordinate range (x: 0-1.0m, y: 0-0.5m, z: 0-0.4m), with a mesh spacing Δx=dp=0.01m and a mesh size of approximately 100×50×40 voxels / nodes, covering the entire water body and the computational domain range of possible motion.

[0046] S3. Using a cubic spline kernel function with a smoothing length h = 1.3dp, physical quantities such as particle density and pressure are interpolated to each grid node. For each background grid node i, all particles j are searched within the support domain of the kernel function determined by its smoothing length h. The physical quantities carried by the particles are interpolated to the grid node by weighted summation through the kernel function W. At the same time, the fluid identification function of each node is calculated based on the interpolation density. Nodes with a density greater than 50% of the reference density are marked as fluid regions (F = 1), otherwise they are empty regions (F = 0). The transition region takes continuous values ​​(F = (0,1)) to provide field data for the determination of free surfaces.

[0047] S4, using density threshold =0.5 ( Using the fluid reference density as the criterion for determining free surfaces, the MarchingCubes algorithm is used to extract isosurfaces and generate a free surface mesh composed of triangular facets.

[0048] S5. Apply Laplacian smoothing to the free surface mesh, iterating 5 times, with a vertex displacement coefficient of 0.5 each time, while applying volume conservation constraints.

[0049] S6. Import the smoothed free surface mesh into the visualization rendering engine, use the Phong lighting model, set the water body to a blue semi-transparent material, and render a high-quality 3D free surface image.

[0050] In this embodiment, as Figure 5 As shown, compared to traditional direct particle rendering (such as displaying particle spheres), Figure 1 Compared to the previous method, the free surface obtained by this method is smooth and continuous, without discrete particle gaps and jagged edges. The dam break wavefront morphology and surge structure are clearly distinguishable in the three-dimensional scene, and the visualization quality is significantly improved.

[0051] Finally, it should be noted that the above embodiments are only used to illustrate the technical solutions of the present invention and are not intended to limit it. Although the present invention has been described in detail with reference to preferred embodiments, those skilled in the art should understand that modifications or equivalent substitutions can be made to the technical solutions of the present invention without departing from the spirit and scope of the technical solutions of the present invention, and all such modifications or substitutions should be covered within the scope of the claims of the present invention.

Claims

1. A post-processing display method for improving the visualization quality of particle simulation results, which only applies to the post-processing display stage of particle simulation results and does not change the original particle calculation results, characterized in that, Includes the following steps: Read the result file of the particle numerical simulation output, and obtain the spatial coordinates and physical quantities of each particle at each output time. The physical quantities include at least density. Based on the particle computational domain spatial range determined by the minimum and maximum values ​​of all particle coordinates, a three-dimensional structured background mesh covering the entire computational domain is constructed, and the mesh resolution is set according to the display accuracy requirements. The physical quantities carried by each particle are interpolated to each node of the three-dimensional structured background grid using the kernel function interpolation method to obtain the physical quantity field on the grid node. At the same time, the identification function or volume fraction used to identify the fluid region is calculated on each grid node to determine the subsequent free surface. On the three-dimensional structured background mesh, a free surface determination threshold is set based on the physical quantity field obtained by interpolation, and an isosurface extraction algorithm is used to extract the free surface isosurface, generating free surface mesh data composed of triangular facets. The free surface mesh data includes vertex coordinates, normal vectors, and facet topological connection relationships. The extracted free surface mesh data is smoothed to eliminate jagged edges and fluctuations caused by mesh discretization and particle noise; The processed free surface mesh data is imported into the visualization rendering engine for 3D free surface display.

2. The post-processing display method for improving the visualization quality of particle simulation results as described in claim 1, characterized in that, The obtained physical quantity field on the grid node specifically includes: The kernel function interpolation method is used to interpolate the physical quantities carried by each particle onto each node of the three-dimensional structured background mesh. For each mesh node i, all particles j within its smooth length h support domain are searched, and the physical quantity value of that node is calculated using the kernel function weighted summation method. The calculation formula for the kernel function weighted summation method is as follows: ; in, Let i be the target physical quantity at grid node i; Let j be the mass of particle j; Let be the density of particle j; Let be the physical quantity carried by particle j; Here, h is the kernel function; h is the smooth length. , These are the spatial position vectors of grid node i and particle j, respectively.

3. The post-processing display method for improving the visualization quality of particle simulation results as described in claim 2, characterized in that, The kernel function is one of a cubic spline kernel function, a Gaussian kernel function, or a quintic kernel function.

4. The post-processing display method for improving the visualization quality of particle simulation results as described in claim 1, characterized in that, The particle numerical simulation is a smoothed particle hydrodynamics (SPH) simulation, and the particles in the result file are SPH particles. The physical quantities include at least density, pressure, velocity vector, particle mass, and particle number.

5. The post-processing display method for improving the visualization quality of particle simulation results as described in claim 1, characterized in that, The free surface determination threshold is 30%-70% of the fluid reference density, or a volume fraction F=0.5; The isosurface extraction algorithm is one of the Marching Cubes algorithm, the Marching Tetrahedra algorithm, or the asymptote decoder.

6. The post-processing display method for improving the visualization quality of particle simulation results as described in claim 5, characterized in that, When using the Marching Cubes algorithm, at the 8 vertices of a single voxel, the vertices are classified into internal vertices and external vertices according to the free face determination threshold, where the voxel is a cube mesh cell; The intersection of the isosurface and the edge is obtained by linear interpolation on the edge where the scalar value crosses the threshold; adjacent intersections on the same cube are connected to form triangular patches, and the triangular patches of all voxels are combined to form complete free surface mesh data.

7. The post-processing display method for improving the visualization quality of particle simulation results as described in claim 1, characterized in that, The smoothing process for the extracted free surface mesh data specifically includes: The extracted free surface mesh data is smoothed using one or more combinations of Laplacian smoothing, curvature-based smoothing, or Humphrey Classes smoothing. During the smoothing process, volume conservation constraints are set to prevent the overall shape of the free surface from shifting.

8. The post-processing display method for improving the visualization quality of particle simulation results as described in claim 1, characterized in that, The step of importing the processed free-face mesh data into the visualization rendering engine for 3D free-face display specifically includes: The processed free surface mesh data is imported into the visualization rendering engine, and rendering parameters such as lighting model settings, material shading, and transparency adjustment are configured to realize the display of three-dimensional free surfaces. The lighting model is either the Phong lighting model or the PBR physical rendering model.

9. A post-processing display method for improving the visualization quality of particle simulation results as described in claim 1 or 8, characterized in that, When importing the processed free surface mesh data into the visualization rendering engine for three-dimensional free surface display, post-processing elements are also superimposed to achieve three-dimensional visualization display with multi-element fusion. The post-processing elements include one or more of particle position, velocity vector, pressure cloud map or floating body structure.