A high-precision interface processing method suitable for multi-resolution particle method
By using a high-precision interface processing method of multi-resolution particle method in fluid flow simulation, setting a virtual grid and correcting the particle velocity and position, the problems of insufficient fluid flow simulation accuracy and low calculation efficiency in the prior art are solved, and stable and accurate calculation results and efficient calculation processes are achieved.
Patent Information
- Application Number
- CN202410698523.1
- Authority / Receiving Office
- CN · China
- Patent Type
- Patents(China)
- Current Assignee / Owner
- Filing Date
- 2024-05-31
- Publication Date
- 2025-04-18
- Estimated Expiration
- 2044-05-31
AI Technical Summary
The prior art has insufficient accuracy in fluid flow simulation, especially in complex geometric boundaries and large displacement problems, and the stability and calculation efficiency of the gridless particle method are low.
A high-precision interface processing method using the multi-resolution particle method is used to set virtual grids on both sides of the interface, and the fluid flow is calculated through the virtual grid and fluid particle information to correct the particle velocity and position.
It realizes stable and accurate calculation results, simplifies the calculation process, improves overall efficiency, and can meet the needs of different application scenarios.
Smart Images
Figure CN118569052B_ABST
Abstract
Description
Technical field:
[0001] The invention relates to a high-precision interface processing method suitable for a multi-resolution particle method. Background technology:
[0002] Fluid flow is ubiquitous in nature and industry. In fields such as oceanography, aerospace, and biomedicine, accurate and efficient numerical simulation of fluid flow phenomena is a challenging task and is of great significance to the advancement of science and technology.
[0003] To date, researchers have applied many numerical methods to fluid flow simulation, such as grid-based methods and meshless particle methods. Although the method based solely on the Euler grid has the advantages of maintaining grid regularity and being applicable to large distortions, its accuracy on complex geometric boundaries and moving boundaries is limited. In the method based solely on the Lagrangian grid, the grid moves with the flow of the fluid, but this method is only suitable for small displacement problems, and the grid may be distorted when the deformation is too large.
[0004] Therefore, for problems with large displacements and complex boundary conditions, the accuracy of the pure Euler grid method and the pure Lagrangian grid method is not enough. In this case, the researchers proposed the arbitrary Lagrangian-Euler method, which uses both Euler description and Lagrangian description to overcome the shortcomings of the above two methods. However, in order to ensure the accuracy of the simulation, the above method inevitably requires re-gridding, which takes a lot of time; and for the meshless particle method, although it has great advantages in fluid flow simulation, stability and accuracy are its biggest defects, which cannot meet different application scenarios. At the same time, the overall calculation amount is too large and the calculation efficiency is low. Summary of the invention:
[0005] The embodiment of the present invention provides a high-precision interface processing method suitable for a multi-resolution particle method. The method is reasonably designed, and is aimed at the interface processing method of high and low precision areas of the multi-resolution moving semi-implicit particle method. Virtual grids are set on both sides of the interface, which effectively avoids the direct mutual calculation of particles of different sizes and causes calculation instability, thereby being able to obtain stable and accurate calculation results, and correct the particle speed and position according to the calculation results. The method is simple in calculation and has high overall efficiency, can meet different application requirements of different application scenarios, and solves the problems existing in the prior art.
[0006] The technical solution adopted by the present invention to solve the above technical problems is:
[0007] A high-precision interface processing method suitable for a multi-resolution particle method, the processing method comprising the following steps:
[0008] S1, measure the flow field and physical parameters of the fluid;
[0009] S2, set the initial distribution of particles;
[0010] S3, determine the range of high and low precision areas and clarify the interface information;
[0011] S4, setting virtual grids on both sides of the interface;
[0012] S5, calculating the fluid flow using the virtual grid and fluid particle information to obtain corresponding calculation results;
[0013] S6, correcting the particle velocity and particle position according to the obtained calculation results.
[0014] The physical parameters of the flow field include the size and boundary of the flow field, and the physical parameters of the fluid include the density and viscosity of the fluid.
[0015] The virtual grid and fluid particle information are used to calculate the fluid flow and obtain the corresponding calculation results, including the following steps:
[0016] S5.1, read the large particle physics information and small particle physics information in the fluid model, read the virtual grid, transition particles and wall boundary information;
[0017] S5.2, search for small particles, small virtual grids and wall boundaries around them, and exclude large particles, large virtual grids and transition particles;
[0018] Search for large particles, large virtual grids, transition particles and wall boundaries around them, and exclude small particles and small virtual grids;
[0019] Search for the transition particle, search the large particles and large virtual grids around it, and exclude the small particles and small virtual grids;
[0020] S5.3, marking the transition particles according to the position, marking the particles belonging to each virtual grid according to the distance, and marking each virtual grid according to the Part ID;
[0021] S5.4, large particles, small particles and transition particles and their corresponding kernel functions, particle number density, pressure gradient terms and velocity viscosity terms are calculated separately.
[0022] For large particles, the kernel function, particle number density, pressure gradient term and velocity viscosity term are expressed as:
[0023]
[0024] Among them, j is the large particle or transition particle around the large particle; in addition, the particle gap is l j = l r = lc ; The particle influence radius is re j =re r =re c ; j pressure P j =P c or P j =P r ; j's position: r j =r c or j =r r ; j's speed: v j =v c or v j = =v r ; Among them, l r 、re r , P r 、r r and v r are the physical information of transition particles; l c 、re c , P c 、r c and v c They are the physical information of large particles respectively.
[0025] For small particles, the kernel function, particle number density, pressure gradient term and velocity viscosity term are expressed as:
[0026]
[0027] Among them, j is a small particle or a small virtual grid around a small particle; the particle gap of j is l j = l f ; j's influence radius re j =re f ; When j is a small virtual grid, the pressure P of j j =P gc =P r +rgabs(r gc -r r ); j's position: r j =r gc ; j's speed: v j =v gc =v r； Among them, l f 、re f , P f 、r f and v f They are the physical information of small particles respectively.
[0028] For transition particles, the kernel function, particle number density, pressure gradient term and velocity viscosity term are expressed as:
[0029]
[0030] Among them, j is a large particle, transition particle or large virtual grid around a small particle; the particle gap of j is l j = l r = l c ; j's influence radius re j =re c ; j pressure P j =P gc =P r +rgabs(r gc -r r ); j's position: r j =r gc ; j's speed: v j =v gc =v r .
[0031] The present invention adopts the above structure, and provides a data basis for the entire method process by measuring the physical parameters of the flow field and the fluid; by setting the initial distribution of particles, determining the range of high and low precision areas, clarifying the interface information and setting virtual grids on both sides of the interface to obtain the corresponding processing model; the fluid flow state is calculated through the virtual grid and fluid particle information, and the particle velocity and particle position are corrected according to the obtained calculation results, which has the advantages of stability, efficiency, simplicity and practicality. Description of the drawings:
[0032] Figure 1 It is a schematic diagram of the process of the present invention.
[0033] Figure 2 It is a schematic diagram of the processing method of the present invention. Specific implementation method:
[0034] In order to clearly illustrate the technical features of the present solution, the present invention is described in detail below through specific implementation methods and in conjunction with the accompanying drawings.
[0035] like Figure 1-2 As shown in , a high-precision interface processing method suitable for multi-resolution particle method, the processing method comprises the following steps:
[0036] S1, measure the flow field and physical parameters of the fluid;
[0037] S2, set the initial distribution of particles;
[0038] S3, determine the range of high and low precision areas and clarify the interface information;
[0039] S4, setting virtual grids on both sides of the interface;
[0040] S5, calculating the fluid flow using the virtual grid and fluid particle information to obtain corresponding calculation results;
[0041] S6, correcting the particle velocity and particle position according to the obtained calculation results.
[0042] The physical parameters of the flow field include the size and boundary of the flow field, and the physical parameters of the fluid include the density and viscosity of the fluid.
[0043] The virtual grid and fluid particle information are used to calculate the fluid flow and obtain the corresponding calculation results, including the following steps:
[0044] S5.1, read the large particle physics information and small particle physics information in the fluid model, read the virtual grid, transition particles and wall boundary information;
[0045] S5.2, search for small particles, small virtual grids and wall boundaries around them, and exclude large particles, large virtual grids and transition particles;
[0046] Search for large particles, large virtual grids, transition particles and wall boundaries around them, and exclude small particles and small virtual grids;
[0047] Search for the transition particle, search the large particles and large virtual grids around it, and exclude the small particles and small virtual grids;
[0048] S5.3, marking the transition particles according to the position, marking the particles belonging to each virtual grid according to the distance, and marking each virtual grid according to the Part ID;
[0049] S5.4, large particles, small particles and transition particles and their corresponding kernel functions, particle number density, pressure gradient terms and velocity viscosity terms are calculated separately.
[0050] For large particles, the kernel function, particle number density, pressure gradient term and velocity viscosity term are expressed as:
[0051]
[0052] Among them, j is the large particle or transition particle around the large particle; in addition, the particle gap is l j = l r = l c ; The particle influence radius is re j =re r =re c ; j pressure P j =P c or P j =P r; j's position: r j =r c or j =r r ; j's speed: v j =v c or v j = =v r ; Among them, l r 、re r , P r 、r r and v r are the physical information of transition particles; l c 、re c , P c 、r c and v c They are the physical information of large particles respectively.
[0053] For small particles, the kernel function, particle number density, pressure gradient term and velocity viscosity term are expressed as:
[0054]
[0055] Among them, j is a small particle or a small virtual grid around a small particle; the particle gap of j is l j = l f ; j's influence radius re j =re f ; When j is a small virtual grid, the pressure P of j j =P gc =P r +rgabs(r gc -r r ); j's position: r j =r gc ; j's speed: v j =v gc =v r； Among them, l f 、re f , P f 、r f and v f They are the physical information of small particles respectively.
[0056] For transition particles, the kernel function, particle number density, pressure gradient term and velocity viscosity term are expressed as:
[0057]
[0058] Among them, j is a large particle, transition particle or large virtual grid around a small particle; the particle gap of j is l j = l r = lc ; j's influence radius re j =re c ; j pressure P j =P gc =P r +rgabs(r gc -r r ); j's position: r j =r gc ; j's speed: v j =v gc =v r .
[0059] The working principle of a high-precision interface processing method suitable for a multi-resolution particle method in an embodiment of the present invention is as follows: based on an established high-precision interface model, in conjunction with corresponding calculation formulas, a method for processing the interface of high and low precision areas of a multi-resolution moving semi-implicit particle method is implemented, and virtual grids are set on both sides of the interface to effectively avoid calculation instability caused by direct calculation of particles of different sizes, thereby being able to obtain stable and accurate calculation results, and correct the particle speed and position according to the calculation results. The calculation is simple and the overall efficiency is high, which can meet the different application requirements of different application scenarios.
[0060] Furthermore, the Lagrangian meshless particle method has attracted the attention of many researchers because it can naturally handle large displacement problems and complex boundary conditions; in the meshless particle method, the fluid domain is discretized into a series of particles, and each particle carries corresponding dynamic information, such as mass, velocity, acceleration, etc.
[0061] The movement of particles is calculated by neighboring particles covered by kernel functions. The SPH (smoothed particle method) method and the MPS (moving particle method) method are the two most popular meshless particle methods in fluid flow simulation. SPH is a method based on integral representation of quantity and spatial derivatives. It was originally a compressible inviscid flow calculation method developed using a fully explicit algorithm, while the MPS method is based on Taylor series expansion and was originally a incompressible fluid calculation method developed using a semi-implicit algorithm.
[0062] Although the meshless particle method has great advantages in fluid flow simulation, there are still many problems. First of all, stability and accuracy are the most prominent problems of the meshless particle method. In order to improve the uniformity of particle distribution, particle movement technology is used, and methods such as artificial viscosity terms and correction matrices are considered; in addition, an accurate wall boundary model is also very important for simulation accuracy, because the wall boundary not only provides sufficient support for the fluid domain, but sometimes also needs to provide velocity and displacement conditions.
[0063] In the multi-resolution particle method, accurate calculation of the interface between high- and low-precision regions is a prerequisite for stable calculation. Therefore, a high-precision interface processing method suitable for the multi-resolution particle method is proposed in the present application.
[0064] In the overall scheme, the processing method includes the following steps: measuring the physical parameters of the flow field and the fluid; setting the initial distribution of particles; determining the range of high- and low-precision areas and clarifying the interface information; setting virtual grids on both sides of the interface; using the virtual grids and fluid particle information to calculate the fluid flow and obtain the corresponding calculation results; and correcting the particle velocity and particle position based on the obtained calculation results.
[0065] Specifically, the physical parameters of the flow field include the size and boundary of the flow field, and the physical parameters of the fluid include the density and viscosity of the fluid.
[0066] Preferably, the fluid flow is calculated using virtual grids and fluid particle information, and obtaining the corresponding calculation results includes the following steps: reading the physical information of large particles and small particles in the fluid model, and reading the virtual grids, transition particles and wall boundary information; searching for small particles, searching the small particles, small virtual grids and wall boundaries around them, and excluding large particles, large virtual grids and transition particles; searching for large particles, searching the large particles, large virtual grids, transition particles and wall boundaries around them, and excluding small particles and small virtual grids; searching for transition particles, searching the large particles and large virtual grids around them to exclude small particles and small virtual grids; marking the transition particles according to the position, marking the particles belonging to each virtual grid according to the distance, and marking each virtual grid according to the Part ID; respectively calculating the large particles, small particles and transition particles and the corresponding kernel functions, particle number density, pressure gradient terms and velocity viscosity terms.
[0067] Generally speaking, in fluid calculations, for large particles, the kernel function, particle number density, pressure gradient term, and velocity viscosity term are expressed as:
[0068]
[0069]
[0070] Among them, j is the large particle or transition particle around the large particle; in addition, the particle gap is l j = l r = l c ; The particle influence radius is re j =re r =re c ; j pressure P j =P c or P j =P r ; j's position: rj =r c or j =r r ; j's speed: v j =v c or v j = =v r ; Among them, l r 、re r , P r 、r r and v r are the physical information of transition particles; l c 、re c , P c 、r c and v c They are the physical information of large particles respectively.
[0071] For small particles, the kernel function, particle number density, pressure gradient term and velocity viscosity term are expressed as:
[0072]
[0073] Among them, j is a small particle or a small virtual grid around a small particle; the particle gap of j is l j = l f ; j's influence radius re j =re f ; When j is a small virtual grid, the pressure P of j j =P gc =P r +rgabs(r gc -r r ); j's position: r j =r gc ; j's speed: v j =v gc =v r； Among them, l f 、re f , P f 、r f and v f They are the physical information of small particles respectively.
[0074] For transition particles, the kernel function, particle number density, pressure gradient term and velocity viscosity term are expressed as:
[0075]
[0076] Among them, j is a large particle, transition particle or large virtual grid around a small particle; the particle gap of j is l j = l r = l c; j's influence radius re j =re c ; j pressure P j =P gc =P r +rgabs(r gc -r r ); j's position: r j =r gc ; j's speed: v j =v gc =v r .
[0077] The core innovation of this application is to propose an accurate interface processing method. By setting up virtual grids on both sides of the interface, it effectively avoids the direct mutual calculation of particles of different sizes, which may cause calculation instability.
[0078] It should be noted that the present application can be applied to a variety of different application scenarios and has obvious practicality and accuracy compared with existing conventional technologies.
[0079] In summary, a high-precision interface processing method suitable for a multi-resolution particle method in an embodiment of the present invention is based on an established high-precision interface model, in conjunction with corresponding calculation formulas, and is aimed at the interface processing method of high and low precision areas of a multi-resolution moving semi-implicit particle method. Virtual grids are set on both sides of the interface to effectively avoid calculation instability caused by direct calculation of particles of different sizes, thereby being able to obtain stable and accurate calculation results, and correct the particle speed and position according to the calculation results. The calculation is simple and the overall efficiency is high, which can meet the different application requirements of different application scenarios.
[0080] The above specific implementation manner cannot be used as a limitation on the protection scope of the present invention. For those skilled in the art, any substitution, improvement or change made to the implementation manner of the present invention falls within the protection scope of the present invention.
[0081] The matters not described in detail in the present invention are all known technologies to those skilled in the art.
Claims
1. A high-precision interface processing method suitable for multi-resolution particle method, characterized in that: The processing method comprises the following steps: S1, measure the flow field and physical parameters of the fluid; S2, set the initial distribution of particles; S3, determine the range of high and low precision areas and clarify the interface information; S4, setting virtual grids on both sides of the interface; S5, calculating the fluid flow using the virtual grid and fluid particle information to obtain corresponding calculation results; S6, correcting the particle velocity and particle position according to the obtained calculation results; The virtual grid and fluid particle information are used to calculate the fluid flow and obtain the corresponding calculation results, including the following steps: S5.1, read the large particle physics information and small particle physics information in the fluid model, read the virtual grid, transition particles and wall boundary information; S5.2, search for small particles, small virtual grids and wall boundaries around them, and exclude large particles, large virtual grids and transition particles; Search for large particles, large virtual grids, transition particles and wall boundaries around them, and exclude small particles and small virtual grids; Search for the transition particle, search the large particles and large virtual grids around it, and exclude the small particles and small virtual grids; S5.3, marking the transition particles according to the position, marking the particles belonging to each virtual grid according to the distance, and marking each virtual grid according to the Part ID; S5.4, large particles, small particles and transition particles and their corresponding kernel functions, particle number density, pressure gradient terms and velocity viscosity terms are calculated separately.
2. A high-precision interface processing method suitable for multi-resolution particle method according to claim 1, characterized in that: The physical parameters of the flow field include the size and boundary of the flow field, and the physical parameters of the fluid include the density and viscosity of the fluid.
3. A high-precision interface processing method suitable for multi-resolution particle method according to claim 1, characterized in that: For large particles, the kernel function, particle number density, pressure gradient term and velocity viscosity term are expressed as: Among them, j1 is the large particle or transition particle around the large particle; in addition, the particle gap is l j1 = l r = l c ; The particle influence radius is re j1 =re r =re c ; Pressure P of j1 j1 =P c or P j1 =P r ; Position of j1: r j1 =r c or j1 =r r ; Speed of j1: v j1 =v c or v j1 =v r ; Among them, l r 、re r , P r 、r r and v r are the physical information of transition particles; l c 、re c , P c 、r c and v c They are the physical information of large particles respectively.
4. A high-precision interface processing method suitable for multi-resolution particle method according to claim 1, characterized in that: For small particles, the kernel function, particle number density, pressure gradient term and velocity viscosity term are expressed as: Among them, j2 is a small particle or a small virtual grid around the small particle; the particle gap of j2 is l j = l f ; The influence radius of j2 re j =re f ; When j2 is a small virtual grid, the pressure P of j2 j =P gc =P r +rgabs(r gc -r r ); j2 position: r j2 =r gc ; Speed of j2: v j2 =v gc =v r； Among them, l f 、re f , P f 、r f and v f They are the physical information of small particles respectively.
5. The high-precision interface processing method suitable for multi-resolution particle method according to claim 1, characterized in that: For transition particles, the kernel function, particle number density, pressure gradient term and velocity viscosity term are expressed as: Among them, j3 is a large particle, transition particle or large virtual grid around the small particle; the particle gap of j3 is l j = l r = l c ; The influence radius of j3 re j3 =re c ; Pressure P of j3 j3 =P gc =P r +rg abs(r gc -r r ); j3 position: r j3 =r gc ; Speed of j3: v j3 =v gc =v r .