High-precision capture method of strong impact material interface based on particle level set
By combining the semi-Lagrangian method and finite volume method based on particle level set, combined with escape particle projection method, the problem of low interface capture accuracy and calculation efficiency of strong impact multiphase fluids is solved, and high-precision and efficient interface capture are achieved Effect.
Patent Information
- Application Number
- CN202310528792.9
- Authority / Receiving Office
- CN · China
- Patent Type
- Patents(China)
- Current Assignee / Owner
- Filing Date
- 2023-05-11
- Publication Date
- 2024-11-26
- Estimated Expiration
- 2043-05-11
AI Technical Summary
When the prior art captures the material interface of strong impact compressible multiphase fluid, especially complex topological changes and mass loss problems, it is difficult to solve, resulting in low interface capture accuracy and computational efficiency.
Using a method based on particle level set, the Level-Set equation and the particle transfer equation are combined, and discrete solution is performed using the semi-Lagrangian method and the finite volume method, and the interface correction is performed by combining the escape particle projection method to improve the accuracy of interface capture and Conservation.
It realizes higher precision interface capture, reduces computing resource consumption, has stronger applicability and lower limitations, and can effectively deal with severe deformation and complex topological changes in the interface of strong impact substances.
Smart Images

Figure CN116595745B_ABST
Abstract
Description
Technical Field
[0001] The present application relates to the field of multiphase flow numerical simulation, and in particular to a high-precision capture method of strong impact material interfaces based on particle level sets. Background Art
[0002] The capture and processing of material interfaces are the core issues of multiphase flow numerical simulation, especially for compressible multiphase fluids under strong impact extreme conditions such as underwater explosions. The interface may undergo quite complex topological changes, such as fragmentation, coalescence and flipping, which greatly increases the difficulty of numerical simulation. The current material interface description methods can be divided into Lagrangian method and Euler method. The Lagrangian method directly tracks the movement of fluid particles, can accurately track interface changes, and handle discontinuities without numerical dissipation, but it is difficult to handle complex interface shape changes, and the calculation efficiency is relatively low. At the same time, there are problems such as tension instability and particle aggregation. The Euler method reconstructs or directly captures the interface position and shape by constructing a transport equation with an additional variable. It can easily handle complex interface topological changes and has high calculation efficiency, so it is widely used in interface description. Therefore, for the simulation of compressible multiphase fluids with strong impact such as underwater explosions, the Euler method is very suitable for description.
[0003] In the framework of the Euler description method, there are two main types of simulation methods: 1) The material interface is regarded as a diffusion zone of finite thickness, mainly introducing a mixing model to describe the interface, such as the γ-model, but this type of simulation method will lead to non-physical mixing or false pressure oscillations. (2) The material interface is regarded as a clear interface with zero thickness. The most representative simulation methods of this type are the volume fraction (VOF) method and the level set (LS) method. Among them, the VOF method reconstructs the interface according to the volume fraction values of different grid cells in the flow field, which has good conservation properties, but the overall accuracy is not high, it is difficult to accurately obtain the interface position and it is difficult to generalize to high-order accuracy and three-dimensional conditions. The LS method defines a signed distance function to the interface in the flow field, and obtains the motion state of the interface by solving the transport equation of the distance function. This method can obtain a smooth interface and accurate interface parameters, such as interface normal, curvature, etc., but there is a significant quality loss problem in the reinitialization process, which seriously affects the effect of interface capture.
[0004] In comparison, the LS method has outstanding advantages in interface capture, but the key issue is to solve the mass loss. There are currently three different solutions. The first is to improve the reinitialization process to reduce the mass loss. This method is not obvious in optimization. The second is to couple the LS method with other Euler methods, such as the CLSVOF method, which not only improves the curvature calculation of the VOF method, but also improves the mass loss of the LS method. However, it is currently mainly used for incompressible fluids, with limited application and high computational complexity. The third method is to combine the LS method with the Lagrangian method, such as the particle level set (PLS) method, which introduces Lagrangian particles into the LS method and uses escaped particles to correct the interface. It is a high-precision interface capture method. Compared with other methods, the PLS method has the advantages of significantly improving interface accuracy and conservation, and can be used to accurately describe the interface of two-phase flow with drastic changes. However, in the specific implementation process, the accuracy of the material interface captured by the traditional PLS method is often insufficient. Summary of the invention
[0005] In view of the above problems and technical requirements, the applicant has proposed a high-precision capture method of strong impact material interface based on particle level set. The technical solution of this application is as follows:
[0006] A high-precision capture method for a strong impact material interface based on a particle level set, the high-precision capture method for a strong impact material interface comprising:
[0007] The flow field calculation domain of strong impact compressible multiphase flow is meshed to obtain several mesh units;
[0008] Determine the time step t according to the constructed Level-Set equation 0 The strong impact material interface Γ(t 0 ) passes through as interface meshes, and initializes particle information of particles in each interface mesh;
[0009] At any time step t n , according to the time step t n The particle information of n The particles are attracted to the strong impact material interface Γ(t n ), n is a parameter with a starting value of 0;
[0010] The semi-Lagrangian method is used for the first-order discretization of the time dimension, and the semi-Lagrangian method or the finite volume method is used for the discretization of the space dimension. The Level-Set equation and the particle transport equation are solved to obtain the grid points of each grid unit at the time step t. n+1 The distance indicator function value of , and the time step t n+1 Particle information;
[0011] According to each grid point at time step t n+1 The distance indicator function value and time step t n+1 The particle information of the escaped particle is used to determine the escaped particle, and the projection of the escaped particle on the normal line of the grid point of the grid unit is obtained, and the grid point is corrected at the time step t n+1 The distance indicator function value of ;
[0012] Reinitialize the distance indicator function value of the flow field calculation domain and get the time step t n+1 The strong impact material interface Γ(t n+1 );
[0013] Let n = n + 1, and execute again according to the time step t n The particle information of n The particles are attracted to the strong impact material interface Γ(t n ) until all time steps are solved and the strong impact material interface of the strong impact compressible multiphase flow is captured.
[0014] The beneficial technical effects of this application are:
[0015] The present application discloses a high-precision capture method for strong impact material interfaces based on particle level sets. The method uses a projection method to determine whether the distance indicator function value at a grid point needs to be updated, thereby reducing the interference caused by particle correction and capturing strong impact material interfaces with higher precision.
[0016] In addition, when solving the Level-Set equation, the semi-Lagrangian method is used to discretize the first-order accuracy in time and space. It is unconditionally stable, reduces the calculation time, and greatly reduces the consumption of resources. It can also be applied to the finite volume method. Compared with traditional methods, it has stronger applicability and fewer limitations.
[0017] This method takes into account the severe deformation of the interface of the strong impact material, and distributes the added particles according to the curvature of the interface of the strong impact material when adding particles, so as to add particles in the interface area that needs the most correction. This method greatly improves the utilization rate of particles, reduces resource consumption, and improves computing efficiency. BRIEF DESCRIPTION OF THE DRAWINGS
[0018] Figure 1 This is a method flow chart of a method for high-precision capture of a strong impact material interface according to an embodiment of the present application.
[0019] Figure 2 It is a schematic diagram of the relationship between an escaping particle and the four grid points of the grid unit in which it is located in one embodiment of the present application.
[0020] Figure 3 The diagram is a schematic diagram of a strong impact material interface captured by the method of the present application in an experimental example.
[0021] Figure 4 yes Figure 3 It is a schematic diagram of another strong impact material interface captured by the method of the present application in the experimental example.
[0022] Figure 5 It is a schematic diagram of a strong impact material interface and pressure contour map captured by the method of the present application in another experimental example. DETAILED DESCRIPTION
[0023] The specific implementation of the present application is further described below in conjunction with the accompanying drawings.
[0024] This application discloses a high-precision capture method for strong impact material interface based on particle level set, please refer to Figure 1 As shown in the flowchart, the high-precision capture method of the strong impact material interface includes the following steps:
[0025] Step S1, meshing the flow field calculation domain of the strong impact compressible multiphase flow to obtain a plurality of mesh units.
[0026] Step S2, determine the time step t according to the constructed Level-Set equation 0 The strong impact material interface Γ(t 0 ) passes through as the interface mesh, and initializes the particle information of the particles in each interface mesh.
[0027] Since the particle level set is based on the LS method, we first need to construct the Level-Set equation X is the position parameter, t n is a time parameter and represents the nth time step, where n is a parameter with an initial value of 0. This is because the LS method treats the material interface moving with time as a function The zero isosurface of Satisfies certain equations. At each time step t n , only ask for The distance indicator function value of Then we know the time step t n The position of the strong impact material interface Γ(tn), Ω represents the interface Γ(t n ) area. The constructed Level-Set equation is:
[0028]
[0029] In the above formula, speed
[0030] At the initial time step t 0 , the initialized distance indicator function value
[0031]
[0032] Among them, Ω=Ω1∪Ω2, Ω1 and Ω2 represent the areas on both sides of the strong impact material interface. d(X,Γ(t 0 )) represents the distance from the point at coordinate X to Γ(t 0 ) distance, at the strong impact material interface, Indicates the area where a fluid is located. Indicates the area where another fluid is located.
[0033] According to the constructed Level-Set equation, the distance indicator function value of the grid point of the grid unit The interface mesh can be determined by calculation: if the distance indicator function value of each grid point of a grid unit If the signs of are different, the grid unit is determined to be an interface grid.
[0034] At the initial time step t 0 , arrange the particles into the interface grid, arrange the particles along the local coordinates of the interface grid, and arrange 4 particles in each direction. The particle information of each particle includes the particle position X P , particles in X P Flow rate at Particle Symbols P and the particle radius r P , where when the distance indicator function value of the particle is positive, the particle is a positive particle, and the particle symbol s P =1; when the distance indicator function value of the particle is negative, the particle is a negative particle, and the particle symbol s P =-1. Particle radius r P According to the PLS method, the following settings are made:
[0035]
[0036] Among them, r min =0.1Δh,r max =0.5Δh, Δh is the minimum value of the grid cell spacing, is position X P The distance indicator function value of the particle at can be obtained by interpolation based on the distance indicator function value of the grid point. Different particles can overlap each other.
[0037] Step 3: At any time step t n , according to the time step tn The particle information of n The particles are attracted to the strong impact material interface Γ(t n )superior.
[0038] After initialization, the particles are attracted to the strong impact material interface along the shortest path. Since the shortest path is along the normal direction of the material interface, the position of the particles after being attracted is updated as follows:
[0039]
[0040] Among them, X new is position X P The updated position of the particle at , λ is the attraction coefficient and λ=1, is the ideal Level-Set function value 0 (indicating that the particles are eventually drawn to the interface), n(X P ) is the position X P The particles are attracted through several iterations, and the particles that are not attracted to the strong impact material interface are deleted. The particles attracted to the strong impact material interface will be transported with the flow and update their positions.
[0041] Step 4: Solve the Level-Set equation and the particle transport equation to obtain the grid points of each grid unit at time step t n+1 The distance indicator function value of , and the time step t n+1 Particle information.
[0042] (1) In one method, a semi-Lagrangian method is used for first-order discretization of the time dimension, and a semi-Lagrangian method is used for first-order discretization of the space dimension, including:
[0043] Perform the first-order discretization of the spatial dimension of the flow field calculation domain and determine the coordinates G of the grid point in the i-th row and j-th column i,j =(iΔx,jΔy), and the velocity at coordinate Gi,j Δx is the unit step in the row direction, Δy is the unit step in the column direction, and u i,j is the velocity along the row direction, v i,j is the velocity along the column direction, i and j are parameters.
[0044] Discretize the time step in the first order of time dimension and determine any time step t n =n·Δt, Δt is the unit step of the time dimension.
[0045] Then the Level-Set equation is solved according to the semi-Lagrangian method to obtain the coordinate G i,j The grid points at time step t n+1 The distance indicator function value of is:
[0046]
[0047] in, Represents the coordinate G r+1,s+1 At time step t n The distance indicator function value of Represents the coordinate G r,s+1 At time step t n The distance indicator function value of Represents the coordinate G r+1,s At time step t n The distance indicator function value of Represents the coordinate G r,s At time step t n The distance indicator function value of ;
[0048]
[0049] The velocity of each particle can be determined by interpolating the velocity of each grid point. Then solve the particle transport equation The particle position after the particle is updated with the flow can be determined. By interpolating the distance indicator function value of each grid point, the particle symbol s after the particle is updated with the flow can be obtained. P and the particle radius r P , thus obtaining the updated time step t n+1 Particle information.
[0050] The method is unconditionally stable, so the size of the time step is not controlled by the stability-based CFL condition, which helps to reduce the computational time overhead and improve computational efficiency.
[0051] (2) In another method, the semi-Lagrangian method is used for the first-order discretization of the time dimension, and the finite volume method is used for the discretization of the space dimension, including:
[0052] The constructed Level-Set equation is The gradient of the spatially discretized Level-Set equation based on the finite volume method is: The semi-discrete form of the Level-Set equation is obtained as
[0053] Among them, A f is the area of the calculation unit surface, V cell is the volume of the calculation unit, n f is the unit external normal of the calculation unit surface, ρ is the medium density, u is the flow velocity, is the distance indicator function value, and faces is the set of calculation unit faces.
[0054] The remaining time discretization methods and subsequent solution methods are similar to the first method and will not be described in detail in this embodiment.
[0055] Step 5: According to the time step t of each grid point n+1 The distance indicator function value and time step t n+1 The particle information of the escaped particle is used to determine the escaped particle, and the projection of the escaped particle on the normal line of the grid point of the grid unit is obtained, and the grid point is corrected at the time step t n+1 The distance indicator function value.
[0056] Interface correction is a key step to accurately capture the interface of strongly impacted compressible multiphase fluids. It can effectively reduce the error caused by numerical dissipation in the LS method and improve the interface accuracy. During the calculation process, some particles drift through the interface of strongly impacted materials. When the drift exceeds the particle radius, they are called escaped particles. When the particle sign of an escaped particle before drift is positive and the particle sign after drift is negative, the escaped particle is an escaped positive particle. When the particle sign of an escaped particle before drift is negative and the particle sign after drift is positive, the escaped particle is an escaped negative particle.
[0057] In the traditional interface correction method, the coordinate X P , particle radius r P , particle symbol s P The correction value of the escaping particle to the nearby grid point with coordinate X is For each escaping positive particle, to reconstruct domain, calculate the correction value of the grid point containing the grid cell of the escaping positive particle For each grid point, there may be multiple escaped particles around the grid point, so the distance indicator function value of the grid point is compared with the correction value of each escaped positive particle calculated for it. Compare and use the distance indicator function value of the grid point Corrected to The same is true for escaping negative particles. After calculating the correction value of each escaping negative particle to it, Then, the distance indicator function value of the grid point is Corrected to use and The minimum reconstruction interface in , has:
[0058] However, when the interface of a strong impact material is severely deformed, some areas will lack a sufficient number of particles, while some areas will accumulate a large number of particles. If the interface is corrected using this traditional method, the interface correction will be inaccurate.
[0059] like Figure 2As shown in the figure, assuming that the particle center of an escaping particle is P, the coordinates of the four grid points of the grid unit where the escaping particle is located are G i,j+1 , G i+1,j+1 , G i,j and G i+1,j According to the traditional interface correction method, the escaping particle coordinates G i,j+1 Correction value of grid point D The escaping particle coordinates G i+1,j Correction value of grid point A After comparison, we can get the distance indicator function value of grid point D Corrected to The distance indicator function value of grid point A Keep the original value. In this scene, for this escaping particle, the distance from the grid point to the particle center P is |XX P |Always larger than the particle radius r of the escaping particle P , so the coordinate X P , particle radius r P , particle symbol s P The correction value of the escaping positive particle to the grid point with coordinate X near the grid point is:
[0060]
[0061] Coordinate X P , particle radius r P , particle symbol s P The correction value of the escaping negative particle for the grid point with coordinate X near the grid point is:
[0062]
[0063] in, is the distance indicator function value of the grid point with coordinate X.
[0064] According to the above formula, the coordinate G i,j+1 The modified distance indicator function value of the grid point D is Coordinate G i+1,j The distance indicator function value of the grid point A is Coordinate G i+1,j+1 The distance indicator function value of the grid point I is This proves that correction should not be made in the above manner at this time.
[0065] In order to solve this problem, the present application makes corrections based on the projection of the escaped particle on the normal line of the grid point of the grid unit where it is located. This includes: determining each grid point of the grid unit where the escaped particle is located, for any grid point, determining the intersection of the line connecting the grid point and the particle center of the escaped particle on the boundary of the escaped particle, and calculating the coordinates of the projection point of the intersection on the normal line of the grid point. The obtained coordinates are:
[0066] X′=X+γproj n (X P -X);
[0067] in, proj n (X P -X) means (X P -X) is projected onto the normal n of the grid point with coordinate X, is the distance indicator function value based on the grid point with coordinate X Determine the normal direction.
[0068] When the coordinates of the projection point are located in the grid cell where the escaping particle is located, the particle information of the escaping particle is used to correct the grid point at time step t. n+1 The distance indicator function value of . Including:
[0069] When the escaped particle is an escaped positive particle, the correction value of the escaped positive particle to the nearby grid points is determined as according to For the grid point at time step t n+1 The distance indicator function value of make corrections;
[0070] When the escaped particle is an escaped negative particle, the correction value of the escaped negative particle to the nearby grid point is determined as And follow For the grid point at time step t n+1 The distance indicator function value of Make corrections.
[0071] For example, Figure 2 For an escaping particle with a particle center of P, determine the intersection point C of the line connecting the grid point D and the particle center P of the escaping particle on the boundary of the escaping particle, and calculate the normal n of the intersection point C at the grid point D. i,j+1 The coordinates of the projection point on the grid are determined to be located in the grid unit, and the correction is performed according to the above method. For grid point I, the intersection point of the line connecting grid point I and the particle center P of the escaping particle on the boundary of the escaping particle is E, and the normal n of the intersection point E at grid point I is calculated. i+1,j+1The coordinates of the projection point H on the grid are determined not to be located in the grid unit, so the escaping particle is not used to correct the grid point I. Through the above method, the interference caused by particle correction can be reduced, which is conducive to capturing a more accurate distance indicator function value. Through the above interface correction method, the captured strong impact material interface can be corrected from the dotted line to the solid line curve.
[0072] Step 6: After solving and updating the distance indicator function values at different positions, the signed distance property of the Level-Set equation may no longer be satisfied. Therefore, the distance indicator function value of the flow field calculation domain is Reinitialize to satisfy The particle reinitialization method is:
[0073]
[0074] Where τ is the virtual time, and solving the equation for Δt r It has nothing to do with the Δt of the original equation. sign is the sign function.
[0075] After reinitialization is complete, We can get the time step t n+1 The strong impact material interface Γ(t n+1 ).
[0076] Step 7, let n=n+1, and execute the above steps 3 to 6 again until all time steps are solved and the strong impact material interface of the strong impact compressible multiphase flow is captured.
[0077] In addition, as mentioned above, when the interface of a strong impact compressible multiphase fluid is severely deformed, some areas will lack a sufficient number of particles, while some areas will have accumulated particles, resulting in the traditional particle reinitialization method requiring the number of particles to be arranged far exceeding the number of particles that can be used to correct the interface in the calculation, wasting a lot of resources and having very low calculation efficiency. Since the position and number of initialized particles will significantly affect the effect of interface correction, in order to improve the utilization efficiency of particles and use limited particles to correct more important local interfaces, the present application corrects the particles in the flow field calculation domain every predetermined time step, or, when it is determined that the deformation degree of the strong impact material interface meets predetermined requirements, corrects the particles in the flow field calculation domain. The deformation degree of the strong impact material interface can be determined by the change in curvature or the change in the number of particles in the grid unit. Among them, the method of correcting particles in the flow field calculation domain includes:
[0078] On the one hand, according to the current time step t mThe captured strong impact material interface Γ(tm) interface curvature in different local areas, the strong impact material interface Γ(t m ) is added to the grid cells that pass through, and more particles are added to the local area with higher interface curvature, that is, particles are added to the area with larger curvature, thereby improving the capture accuracy of the interface. In one embodiment, determining The local area is formed by the coordinates H, and particles are added in this local area. m )| is the time step t m The strong impact material interface Γ(t m ) is the absolute value of the curvature at coordinate H, is the time step t m The strong impact material interface Γ(t m ) is the average value of the absolute value of the curvature at different positions, is the coordinate H at time step t m The distance indicator function is the absolute value of the value, Δh is the minimum value of the grid unit spacing, and C is a constant. The larger the constant C is, the more the particle initialization position is biased towards the local part with large curvature of the strong impact material interface, the smaller the number of particles required, and the higher the calculation efficiency. However, if the constant C is too large, it will affect the correction accuracy of the strong impact material interface, so it is necessary to take a suitable constant C.
[0079] On the other hand, the strong impact material interface Γ(t m ) are deleted. When the strong impact material interface changes, some grid cells are no longer penetrated by the strong impact material interface, and the particles remaining in these grid cells need to be deleted. The method of deleting particles is: generate virtual particles I1 at a certain distance from the particle along the normal direction of the particle, and generate virtual particles I2 at the same distance in the opposite direction. Compare the signs of the distance indicator function values at I1 and I2. If they are different, the particle will not be deleted, otherwise it will be deleted.
[0080] In a two-dimensional single vortex application example, assuming that the flow field calculation domain is X×Y∈[0,1]×[0,1], at the initial moment, a circular interface with a radius of 0.15 is located at (0.50, 0.75) in the flow field calculation domain, and the flow field velocity field is:
[0081]
[0082] The interface of this two-dimensional single vortex will experience the maximum tensile deformation at t=4.0, and the interface will return to its initial circular state at t=8.0.
[0083] When the method of the present application is used for interface capture, the strong impact material interface captured at t = 4.0 is as follows Figure 3As shown in Figure 2, the strong impact material interface captured at t = 8.0 is as follows: Figure 4 As shown. By comparison, it can be seen that the strong impact material interface captured at t = 4.0 produced the maximum tensile interface deformation, in which the slender filament features were also clearly captured. The strong impact material interface captured at t = 8.0 returned to its original circular state and maintained a good circular shape. The method of the present application has high capture accuracy and mass conservation for complex topological change interfaces.
[0084] In another application example of the explosion process of a single cavitation bubble in an infinite domain, assuming that the flow field calculation domain is X×Y∈[1.0×1.0], the radius of the explosion cavitation bubble is 0.5 and is located at the coordinate origin of the calculation domain. When the method of this application is used for interface capture, at t=1.80×10 -4 The schematic diagram of the strong impact material interface (indicated by Interface in the figure) and the pressure contour map captured at s is shown in the figure below. Figure 5 As shown, due to the extreme compression of the internal gas, the interface of the explosion bubble moves outward rapidly, thereby generating a main shock wave propagating outward and a subsequent series of pressure waves. As can be seen from the figure, both the distribution of pressure contours and the interface movement of the explosion bubble are very accurate, without any obvious numerical oscillation phenomenon. It can be seen that the method of the present application can capture interface changes very accurately and clearly, and combined with an effective interface boundary processing algorithm (such as the modified virtual fluid method), it can accurately and stably realize the capture and processing of the material interface of compressible multiphase fluids under strong impact extreme conditions such as underwater explosions.
[0085] The above is only a preferred embodiment of the present application, and the present application is not limited to the above embodiments. It is understood that other improvements and changes directly derived or associated by those skilled in the art without departing from the spirit and concept of the present application should be considered to be included in the protection scope of the present application.
Claims
1. A high-precision capture method for strong impact material interface based on particle level set, characterized in that: The high-precision capture method of the strong impact material interface comprises: The flow field calculation domain of strong impact compressible multiphase flow is meshed to obtain several mesh units; Determine the time step t according to the constructed Level-Set equation 0 The strong impact material interface Γ(t 0 ) passes through as interface meshes, and initializes particle information of particles in each interface mesh; At any time step t n , according to the time step t n The particle information of n The particles are attracted to the strong impact material interface Γ(t n ), n is a parameter with a starting value of 0; The semi-Lagrangian method is used for the first-order discretization of the time dimension, and the semi-Lagrangian method or the finite volume method is used for the discretization of the space dimension. The Level-Set equation and the particle transport equation are solved to obtain the grid points of each grid unit at the time step t. n+1 The distance indicator function value of , and the time step t n+1 Particle information; According to each grid point at time step t n+1 The distance indicator function value and time step t n+1 The particle information of the escaped particle is used to determine the escaped particle, and the projection of the escaped particle on the normal line of the grid point of the grid unit is obtained, and the grid point is corrected at the time step t n+1 The distance indicator function value of ; Reinitialize the distance indicator function value of the flow field calculation domain and get the time step t n+1 The strong impact material interface Γ(t n+1 ); Let n = n + 1, and execute again the above process according to time step t n The particle information of n The particles are attracted to the strong impact material interface Γ(t n ) until all time steps are solved and the strong impact material interface of the strong impact compressible multiphase flow is captured.
2. The high-precision capture method of strong impact material interface according to claim 1 is characterized in that: Correction grid point at time step t n+1 Methods for distance indicator function values include: Determine each grid point of the grid unit where the escaped particle is located, for any grid point, determine the intersection of the line connecting the grid point and the particle center of the escaped particle on the boundary of the escaped particle, calculate the coordinates of the projection point of the intersection point on the normal line of the grid point, and when the coordinates of the projection point are within the grid unit where the escaped particle is located, use the particle information of the escaped particle to correct the grid point at time step t n+1 The distance indicator function value.
3. The high-precision capture method of strong impact material interface according to claim 2 is characterized in that: For coordinate X P , radius r P The grid point with coordinates X in the grid unit where the escaped particle is located, the coordinates of the projection point of the intersection of the line connecting the grid point and the particle center of the escaped particle on the boundary of the escaped particle on the normal line of the grid point are determined as: X′=X+γproj n (X P -X); in, proj n (X P -X) means (X P -X) is projected onto the normal n of the grid point with coordinate X, is the distance indicator function value based on the grid point with coordinate X Determine the normal direction.
4. The high-precision capture method of strong impact material interface according to claim 3 is characterized in that: Correction grid point at time step t n+1 Methods for distance indicator function values include: When the escaped particle is an escaped positive particle, the correction value of the escaped positive particle to the nearby grid points is determined as And follow For the grid point at time step t n+1 The distance indicator function value of make corrections; When the escaped particle is an escaped negative particle, the correction value of the escaped negative particle to the nearby grid point is determined as And follow For the grid point at time step t n+1 The distance indicator function value of make corrections; Among them, s P is the particle symbol. When the distance indicator function value of the particle is positive, the particle is a positive particle, and the particle symbol s P =1; when the distance indicator function value of the particle is negative, the particle is a negative particle, and the particle symbol s P =-1.
5. The high-precision capture method of strong impact material interface according to claim 1, characterized in that: The high-precision capture method for the strong impact material interface also includes: The particles in the flow field calculation domain are modified every predetermined time step, or the particles in the flow field calculation domain are modified when it is determined that the deformation degree of the strong impact material interface meets the predetermined requirement.
6. The high-precision capture method of strong impact material interface according to claim 5, characterized in that: The method for correcting particles in the flow field calculation domain includes: According to the current time step t m The captured strong impact material interface Γ(t m ) interface curvature in different local areas, the interface Γ(t m ) is added to the grid cells that pass through. The local area with higher interface curvature has more particles added, and the current time step t m The particles in the grid cells that have not passed through the captured strong impact material interface Γ(tm) are deleted.
7. The high-precision capture method of strong impact material interface according to claim 6, characterized in that: Methods for adding particles to the flow field computational domain when reinitializing particles include: Determine satisfaction A local area formed by coordinates H, and adding particles in the local area; Among them, |κ(H,t m )| is the time step t m The strong impact material interface Γ(t m ) is the absolute value of the curvature at coordinate H, is the strong impact material interface Γ(t m ) is the average value of the absolute value of the curvature at different positions, is the coordinate H at time step t m The distance indicator function is the absolute value of the value, Δh is the minimum value of the spacing of the grid cells, and C is a constant.
8. The high-precision capture method of strong impact material interface according to claim 1, characterized in that: Methods for using the semi-Lagrangian method to discretize the time dimension and the space dimension and solve the Level-Set equation include: The flow field calculation domain is discretized in the first order of spatial dimension to determine the coordinates G of the grid point in the i-th row and j-th column. i,j =(iΔx,jΔy), and the velocity at coordinate Gi,j Δx is the unit step in the row direction, Δy is the unit step in the column direction, and u i,j is the velocity along the row direction, v i,j is the velocity along the column direction, i and j are parameters; Discretize the time step in the first order of time dimension and determine any time step t n =n·Δt, Δt is the unit step of the time dimension; Determine the coordinate G i,j The grid points at time step t n+1 The distance indicator function value of is: in, Represents the coordinate G r+1,s+1 At time step t n The distance indicator function value of Represents the coordinate G r,s+1 At time step t n The distance indicator function value of Represents the coordinate G r+1,s At time step t n The distance indicator function value of Represents the coordinate G r,s At time step t n The distance indicator function value of ; 9. The high-precision capture method of strong impact material interface according to claim 1, characterized in that: Methods for discretizing spatial dimensions using the finite volume method include: The constructed Level-Set equation is The gradient of the spatially discretized Level-Set equation based on the finite volume method is: The semi-discrete form of the Level-Set equation is obtained as Among them, A f is the area of the calculation unit surface, V cell is the volume of the calculation unit, n f is the unit external normal of the calculation unit surface, ρ is the medium density, u is the flow velocity, is the distance indicator function value, and faces is the set of calculation unit faces.
Citation Information
Patent Citations
Two-phase flow interface capture calculation method based on single-layer particle level set
CN110147575A
Multiphase flow calculation method
CN111241742A
Interface capturing method of free surface using least-square weighted residual method
KR1020130059464A
