A coupled Eulerian-Lagrangian method for accurately capturing the shock wave propagation process

Through the partition weighted bidirectional mapping and particle splitting method coupled to the Euler-Lagrangian method, the problems of grid distortion and numerical dissipation during simulation explosion and impact are solved, and the precise capture and tracking of shock waves is achieved. It is suitable for Numerical simulation of large deformation flow.

CN114611340BActive Publication Date: 2024-11-22BEIJING INST OF TECH
View PDF 0 Cites 0 Cited by

Patent Information

Application Number
CN202210263789.4
Authority / Receiving Office
CN · China
Patent Type
Patents(China)
Current Assignee / Owner
Filing Date
2022-03-17
Publication Date
2024-11-22
Estimated Expiration
2042-03-17

AI Technical Summary

Technical Problem

The prior art is difficult to efficiently and accurately capture and track nonlinear and non-constant shock wave fronts during simulating explosion and impact. Especially in large deformation conditions, the problems of grid distortion and numerical dissipation are prominent, and it is impossible to directly simulate large deformation flow problems. .

Method used

The coupled Euler-Lagrangian method is adopted to achieve high-precision mapping and splitting between the grid and particles through the partition weighted bidirectional mapping method and particle splitting method, avoid grid distortion, and ensure accurate capture and tracking of shock waves.

Benefits of technology

It realizes accurate capture and tracking of shock waves, solves the problem of grid distortion in large deformation situations, makes up for the tracking flow defect of Euler's method, improves the accuracy and stability of the simulation, and is suitable for numerical simulation in the fields of explosion and impact .

✦ Generated by Eureka AI based on patent content.

Smart Images

  • Figure CN114611340B_ABST
    Figure CN114611340B_ABST
Patent Text Reader

Abstract

The present invention relates to a coupled Euler-Lagrangian method for accurately capturing the shock wave propagation process, and belongs to the technical field of computational explosion mechanics. A partitioned weighted bidirectional mapping method is adopted to perform weighted processing on the shape function through the coincident volume fraction of the particle volume domain and the grid domain, which not only improves the mapping accuracy, but also eliminates the cross-grid oscillation; a virtual particle method is adopted to meet the continuity of particles on the boundary, realize the application of outflow, inflow, and periodic fluid boundaries, and ensure that the interpolation of the shape function at the boundary does not reduce the order; in view of the "numerical fracture" fracture problem, a particle splitting method is proposed to split the particles with large deformation to make up for the defect of too few particles in a single network.
Need to check novelty before this filing date? Find Prior Art

Description

Technical Field

[0001] The invention relates to a coupled Euler-Lagrangian method for accurately capturing a shock wave propagation process, and belongs to the technical field of computational explosion mechanics. Background Art

[0002] Explosion and shock issues have always been the focus of national defense construction and civil security, and in recent decades, numerical simulation has gradually developed into an important means of studying this problem. Numerically, the detonation and shock process is analyzed by solving the hyperbolic conservation law equations of the explosion problem. However, as a compressible fluid, the detonation wave has an extremely rich flow structure, which includes both extremely wide-band multi-scale structures in time and space, and discontinuous structures such as shock waves. Therefore, the difficulty of solving the problem lies in how to capture and track nonlinear, unsteady and uncertain shock wave fronts or contact discontinuities with high resolution and efficiency. Although it is difficult, it is necessary to simulate the explosion and shock process as realistically as possible, accurately capture the propagation process of the shock wave, enable people to have a visual understanding of it, and provide a scientific basis for people to formulate corresponding engineering protection and other applications.

[0003] Typical discrete formats for capturing shock wave discontinuities mainly include TVD format, ENO format, WENO format, WCNS and RKDG format, and these formats can be divided into Lagrangian method and Euler method according to the different grids used. In the Euler method, the grid is fixed in space and the object flows through the grid. It is suitable for simulating large deformation and fluid flow problems, but due to the limited grid, discontinuities are often smoothed out too much. For example, the WENO format has high accuracy and good robustness in capturing shock waves, but there are problems such as local extreme point reduction, resulting in relatively large dissipation. It still needs to be improved in simulating complex multi-scale flow problems [6-8]. In the Lagrangian method, the grid deforms with the material, which can avoid numerical dissipation caused by transport terms and can accurately describe the material interface. However, in the case of large deformation, the grid will be distorted and twisted, causing large errors, making it impossible to directly simulate large deformation flow problems. Summary of the invention

[0004] The purpose of the present invention is to provide a coupled Euler-Lagrangian method for accurately capturing the propagation process of shock waves, which can not only solve the grid distortion problem of the Lagrangian method, but also make up for the defect that the Euler method cannot accurately track the flow. The invention is applicable to the field of explosion and shock, and can handle flow-like problems with arbitrary large deformations. It can perform numerical simulations of accurate tracking for problems involving shock waves, such as the propagation of shock waves and the interaction of shock waves of different intensities, and can provide a criterion for damage assessment analysis based on the pressure at any position in the calculation area.

[0005] The purpose of the present invention is achieved through the following technical solutions.

[0006] A coupled Euler-Lagrangian method for accurately capturing the shock wave propagation process includes the following steps:

[0007] The first step is to create a simulation model for the object to be analyzed and initialize the created simulation model, including determining the size of the calculation domain, the position information and geometric size information of various materials in the calculation domain, the grid step size and coordinates, the arrangement of particles, material properties and parameters, boundary conditions and virtual particles, and the initial calculation control parameters;

[0008] The second step is to divide the mass points into A partitions according to the topological relationship between the grid and the mass points by using the partition weighted bidirectional mapping method, calculate the state equation and constitutive model corresponding to the mass points, and update the physical quantities of the mass points.

[0009] Step 2.1: Calculate the strain increment of the particle based on the grid velocity gradient

[0010]

[0011] In the formula, is the grid velocity, Δt k is the time step;

[0012] Step 2.2: Calculate the Cauchy stress of the particle based on the constitutive model and the equation of state

[0013]

[0014] And update the density of the particle

[0015]

[0016] Step 2.3, update the energy of the particle according to the state equation and artificial viscosity;

[0017] Step 2.4, determine the volume of the particle, and when the length of a particle in one direction exceeds a preset threshold, use the particle splitting method to split the particle;

[0018] In the third step, the mass points are mapped back to the grid and the grid velocity is updated;

[0019] Step 3.1: Map the physical quantity of the particle to the grid according to the topological relationship between the particle and the grid

[0020]

[0021]

[0022] Step 3.2: Update the momentum of the mesh nodes

[0023]

[0024] In the formula, is the grid stress, is the particle pressure.

[0025] The fourth step is to determine the topological relationship between the grid and the mass point to map the grid physical quantity to the mass point and calculate the velocity and displacement of the mass point.

[0026] Step 4.1: Map the quantity of the grid node to the mass point according to the partition weighted bidirectional mapping method, and update the displacement and velocity change of the mass point.

[0027]

[0028]

[0029] Step 4.2: Calculate and update the position and velocity of the particle at the new time step.

[0030]

[0031]

[0032] The fifth step is to determine the positions of all particle movements, delete the real particle points outside the calculation domain, and add the virtual particle points within the calculation domain to the set of real particle points.

[0033] Step 6: Add virtual particles again according to the boundary conditions.

[0034] According to the inflow, outflow and periodic boundary conditions, new virtual particles are added in the boundary area, and the relationship between the virtual particles and the real particles conforms to the boundary conditions.

[0035] The seventh step is to determine the topological relationship between the particle and the grid after the movement, and further divide the particle into partition B based on partition A. The physical quantity of the particle is mapped to the grid, and the oscillation of the discontinuity surface is suppressed by the smoothing function to obtain the physical quantity of the grid in the next time step.

[0036] Step 7.1, re-update the physical quantity of the mass point through the smooth function;

[0037]

[0038] Step 7.2: Map the quantity of the mass points to the grid nodes.

[0039]

[0040] Step 8: Based on the set termination conditions, output the corresponding simulation results.

[0041] Step 8.1: If the current number of calculation steps meets the output setting conditions, all simulation data will be merged and output and the serial number mark of the output file will be recorded; the variables and variable names contained in the grid and particles will be written into the file.

[0042] Step 8.2: If the end condition is not met, return to step 2 and start a new time step; if the end condition is met, end the calculation and obtain the pressure of the shock wave;

[0043] The above eight steps achieve accurate numerical simulation calculations of explosion and impact problems;

[0044] The ninth step is to provide guidance and evaluation on the damage to personnel caused by the mine gas explosion based on the obtained shock wave pressure.

[0045] In the second step, the partition weighted bidirectional mapping method is implemented as follows: at the kth time step, the grid domain is regularly distributed along the i and j directions, X represents the Xth grid node in the i direction, and Y represents the Yth grid node in the j direction. The volume domain of the particle p is assumed to be a rectangle, The structure vector defines the particle domain, and the volume domain is defined by the grid lines. and The nth A area divided is denoted as p An , the volume is denoted as V pAn , define the weight ω of each A partition in the mass point pAn for

[0046] ω pAn =V pAn / V p (1)

[0047] Where V p is the total volume of particle p, subscript p represents the particle, and subscript pAn represents the nth large partition A in the volume domain of particle p. When mapping from the grid to the particle, it is first mapped to the partitions of the particle, that is,

[0048]

[0049] In the formula, f pAn For partition p An The physical quantity, n g The number of nodes required for shape function calculation, subscript I represents the I-th mesh node, f I is the physical quantity of the I node, S IpAn =S I (x pAn ) is the shape function, x pAn For partition p An The centroid coordinates of .

[0050] The physical quantity f of the particle p is obtained by weighted calculation of each partition p ,

[0051]

[0052] Where n pA is the total number of partitions A in particle p.

[0053] Similarly, when a particle is mapped to a grid, the physical quantity of the particle is first distributed to each partition according to the weight, and then mapped from each partition to the grid.

[0054]

[0055] When the particle moves to the position of the next time step, the particle needs to carry the cut state and be cut by the grid at the new position. After the particle moves to the new position, the positional relationship between the particle domain and the grid domain is re-determined. At the k+1th time step, the large partition p An New Grid Lines and Continue to split into small partitions p Bn-m , the weight of each small partition is allocated according to the proportion of its volume in the large partition, then the formula for mapping the mass point to the grid is:

[0056]

[0057] Where n pAnB is the total number of small partitions B in the nth large partition A of the particle p, and the subscript pBn-m represents the mth small partition B in the nth large partition A of the particle p. S IpBn-m =S I (x pBn-m ) is the shape function, x pBn-m For small partition p Bn-m The centroid position of .

[0058] In step 2.4, the implementation method of the particle splitting method is as follows: the particle domain is assumed to be a rectangle, The structure vector defines the particle domain, so the length of the particle in two directions can be defined as

[0059]

[0060] When the length in any of the above directions meets the following splitting requirements, the particle splitting is performed

[0061]

[0062] Where α is the specified splitting factor, which is usually less than 1; and is the length of the grid. When the length in the direction reaches the splitting requirement, The direction is split into p1 and p2, and the distance between the particles after the split is The length of the corresponding direction becomes The volume, mass and internal energy are half of the original, and the strain, stress and density remain unchanged.

[0063] In the above-mentioned step 1.5, the implementation method of the virtual particle method is as follows: at the beginning of each time step, a boundary grid with the same spacing as the grid in the computational domain is established on the boundary area, and virtual particles are added therein. The solid circle represents the real particle, and the dotted circle represents the virtual particle, whose density, pressure, velocity and volume are all specified as fixed values ​​by the inflow boundary conditions. At the initial moment, the relative position of the virtual particle in the boundary grid is the same as the relative position of the real particle in the computational grid. It can be inferred that in each time step, the virtual and real particles moving at the same speed have the same relative positions in their respective grids. The position of the virtual particle p* can be inferred based on the position of the real particle p.

[0064]

[0065] Where B is the boundary, x B is the coordinate of the boundary, and the virtual particle positions of other boundary grids can be deduced in the same way. The virtual particle p* flows at the specified speed v0, and after a time Δt k , determine whether the virtual particle has entered the calculation area. If so, add the information of the virtual particle to the set of physical particles. The virtual particles will no longer be continuous after being moved, so all virtual particles should be deleted and returned to the initial step to add particles again.

[0066] Beneficial Effects

[0067] 1. The present invention adopts a partitioned bidirectional weighted mapping method to effectively improve the mapping accuracy, avoid the numerical fluctuation defects of the particle-type method, and maintain its advantage of tracking material flow, thereby achieving accurate capture of the shock wave discontinuity surface;

[0068] 2. The present invention proposes a split particle method, which solves the problem of excessive stretching of particles under large deformation conditions and effectively eliminates the error of "numerical fracture";

[0069] 3. The present invention combines the advantages of the Euler method in that it is easy to handle flow-like problems with arbitrary large deformations and the Lagrangian particle in that it is easy to track the discontinuity process. It can not only handle flow-like problems involving arbitrary large deformations in continuous medium mechanics, but also achieve accurate tracking of the shock wave deformation process, and can be better applied to the numerical simulation research of various explosions and shocks.

[0070] 4. To simulate the large deformation flow problem of explosion and impact, it not only avoids the Lagrangian grid distortion problem, but also makes up for the defect of the Euler method that cannot accurately track the flow, and realizes the high-precision capture of shock waves and discontinuities. It adopts the partition weighted bidirectional mapping method, and weights the shape function through the coincidence volume fraction of the particle volume domain and the grid domain, which not only improves the mapping accuracy, but also eliminates the cross-grid oscillation; the virtual particle method is used to meet the continuity of particles on the boundary, realizes the application of outflow, inflow, and periodic fluid boundaries, and ensures that the interpolation of the shape function at the boundary does not reduce the order; in response to the "numerical fracture" fracture problem, the particle splitting method is proposed to split the large deformation particles to make up for the defect of too few particles in a single network. BRIEF DESCRIPTION OF THE DRAWINGS

[0071] Figure 1 It is a flowchart of the method of the present invention;

[0072] Figure 2 A schematic diagram of two-dimensional partitioning of the pre-movement partition weighted bidirectional mapping method of the present invention;

[0073] Figure 3 It is a two-dimensional partition schematic diagram of the post-mobile partition weighted bidirectional mapping method of the present invention;

[0074] Figure 4 Schematic diagram of particle splitting according to the present invention;

[0075] Figure 5 A schematic diagram of adding virtual particles according to the present invention;

[0076] Figure 6 Simulation modeling for mine explosion examples;

[0077] Figure 7 This is the density cloud diagram of the mine explosion example at T = 0.134s;

[0078] Figure 8 This is the density cloud diagram of the mine explosion example at T = 0.383s;

[0079] Fig. 9 This is the density cloud diagram of the mine explosion example at T = 0.589s;

[0080] Fig.10 This is the density cloud diagram of the mine explosion example at T = 0.911s;

[0081] Fig.11 This is the density cloud map of the mine explosion example at T = 1.054s. DETAILED DESCRIPTION

[0082] The present invention will be described in detail below with reference to the accompanying drawings and embodiments. The technical problems solved by the technical solution of the present invention and the beneficial effects are also described. It should be noted that the described embodiments are only intended to facilitate the understanding of the present invention and do not have any limiting effect on it.

[0083] Taking the explosion problem of shock wave propagation in mine explosion as an example, the shock wave propagation process during mine explosion was numerically simulated. The results show that there is a strong discontinuity on the shock wave interface, which shows that the method of the present invention has a high tracking accuracy for the shock wave discontinuity surface; and can output the pressure at any point in the calculation area. The industry can use this pressure as a damage criterion to conduct target damage assessment and analysis on this location. The following takes the numerical simulation of the shock wave propagation process of mine explosion as an example to explain in detail the coupled Euler-Lagrangian method of the present invention for accurately capturing the shock wave propagation process. Figure 1 As shown;

[0084] Step 1: Create a simulation model for the explosion problem and initialize the created simulation model, including determining the size of the calculation domain, the position information and geometric size information of various materials in the calculation domain, the grid step size and coordinates, the arrangement of particles, material properties and parameters, boundary conditions and virtual particles, and the initial calculation control parameters;

[0085] Specifically, as a preferred embodiment of the present invention, the step 1 includes:

[0086] Step 1.1, set the calculation area to [0,4]×[0,3], set the grid size to 0.01×0.01, set the number of Lagrangian particles in a single grid to 4×4, set the material model to the ideal gas model, and the gas coefficient to 1.4;

[0087] Step 1.2, grid division is realized according to the calculation area and the grid size, and the grid division information is obtained and stored in the grid data, wherein the grid information includes the grid number and the grid coordinates;

[0088] Step 1.3: On all grids, according to the number of Lagrangian particles in a single grid, scatter Lagrangian particles belonging to ideal gas materials in each direction. Based on the principle that the influence area of ​​the Lagrangian particles is closely connected and completely covers all media in the calculation domain, determine the volume of the Lagrangian particles and enter the initial value:

[0089]

[0090] Step 1.4, set the state equation and gas coefficient of the ideal gas material;

[0091] Step 1.5, set the boundary conditions, the left boundary is the inflow boundary, the right boundary is the outflow boundary, the upper and lower boundaries are the solid wall boundaries, and add virtual particles according to the corresponding boundary conditions through the virtual particle method;

[0092] Step 1.6, set the initial calculation control parameters, set the end time to 5, the step interval for saving the results to 1000, and the time advancement technology coefficient to 0.5.

[0093] Step 2: At the beginning of a new time step, the state equation and constitutive model corresponding to the mass point are calculated according to the partition bidirectional weighted mapping method, and the physical quantities of the mass point are updated;

[0094] Specifically, as a preferred embodiment of the present invention, the step 2 includes:

[0095] Step 2.1, calculate the strain increment of the particle according to the grid velocity gradient.

[0096] Step 2.2: According to the constitutive model and the equation of state, calculate the Cauchy stress P of the particle and update the density of the particle.

[0097] Step 2.3: Further, update the energy of the particle according to the equation of state and artificial viscosity.

[0098] Step 2.4: Split the particles that are too deformed according to the particle splitting method

[0099] Step 3: Through the partitioned bidirectional weighted mapping method, the mass points are mapped back to the grid and the grid velocity is updated.

[0100] Step 3.1: Divide region A according to the topological relationship between particles and grids.

[0101] Step 3.2: Update the momentum of the grid nodes.

[0102] Step 4: According to the partitioned bidirectional weighted mapping method, the topological relationship between the grid and the mass point is determined to realize the mapping of the grid physical quantity to the mass point, and the velocity and displacement of the mass point are calculated.

[0103] Step 4.1: Map the momentum of the grid nodes to the mass points and update the velocity of the mass points.

[0104] Step 4.2: Calculate and update the position of the particle at the next time step.

[0105] Step 5, determine the positions of all particle movements, delete the real particles outside the calculation domain, and add the virtual particles in the calculation domain to the real particle set.

[0106] Step 6: Add virtual particles again using the virtual particle method according to the boundary conditions.

[0107] Step 7, determine the topological relationship between the particle and the grid after the movement, divide area A into area B according to the partition bidirectional weighted mapping method, map the particle physical quantity to the grid, suppress the discontinuous surface oscillation through the smoothing function, and obtain the grid physical quantity of the next time step.

[0108] Step 7.1: Update the physical quantities of the particle through the smooth function.

[0109] Step 7.2: Map the physical quantities of the particles to the grid nodes.

[0110] Step 8: Based on the set termination conditions, output the corresponding simulation results.

[0111] Specifically, as a preferred example of the present invention, step 8 includes:

[0112] Step 8.1: If the current number of calculation steps meets the output setting conditions, all simulation data will be merged and output and the serial number mark of the output file will be recorded, and the grid data and particle data will be saved separately.

[0113] Step 8.2: If the end condition is not met, return to step 2; if the end condition is met, end the program.

[0114] Step 9, providing guidance and evaluation on the damage to personnel caused by the mine gas explosion based on the obtained shock wave pressure.

[0115] In step 2, the implementation method of the partition weighted bidirectional mapping method is: Figure 2 It can be seen that at the kth time step, the grid domain is regularly distributed along the i and j directions. X represents the Xth grid node in the i direction, and Y represents the Yth grid node in the j direction. The volume domain of the particle p is assumed to be a rectangle, which is given by The structure vector defines the particle domain, and the volume domain is defined by the grid lines. and The nth A area divided is denoted as p An , the volume is denoted as V pAn , define the weight ω of each A partition in the mass point pAn for

[0116] ω pAn =V pAn / V p (1)

[0117] Where V p is the total volume of particle p, subscript p represents the particle, and subscript pAn represents the nth large partition A in the volume domain of particle p. When mapping from the grid to the particle, it is first mapped to the partitions of the particle, that is,

[0118]

[0119] In the formula, f pAn For partition p An The physical quantity, n g The number of nodes required for shape function calculation, subscript I represents the I-th mesh node, f I is the physical quantity of the I node, S IpAn =S I (x pAn ) is the shape function, x pAn For partition p An The centroid coordinates of .

[0120] The physical quantity f of the particle p is obtained by weighted calculation of each partition p ,

[0121]

[0122] Where n pA is the total number of partitions A in particle p.

[0123] Similarly, when a particle is mapped to a grid, the physical quantity of the particle is first distributed to each partition according to the weight, and then mapped from each partition to the grid.

[0124]

[0125] When the particle moves to the position of the next time step, the particle needs to carry the cut state and be cut by the grid at the new position, see Figure 3 After the particle moves to the new position, the positional relationship between the particle domain and the grid domain is re-determined. At the k+1th time step, the large partition p An New Grid Lines and Continue to split into small partitions p Bn-m , the weight of each small partition is allocated according to the proportion of its volume in the large partition, then the formula for mapping the mass point to the grid is:

[0126]

[0127] Where n pAnB is the total number of small partitions B in the nth large partition A of the particle p, and the subscript pBn-m represents the mth small partition B in the nth large partition A of the particle p. S IpBn-m =S I (x pBn-m ) is the shape function, x pBn-m For small partition p Bn-m The centroid position of .

[0128] In step 2.4, the implementation method of the particle splitting method is as follows, and the schematic diagram is shown in Figure 4The particle domain is assumed to be a rectangle, given by The structure vector defines the particle domain, so the length of the particle in two directions can be defined as

[0129]

[0130] When the length in any of the above directions meets the following splitting requirements, the particle splitting is performed

[0131]

[0132] Where α is the specified splitting factor, which is usually less than 1; and is the length of the grid. When the length in the direction reaches the splitting requirement, The direction is split into p1 and p2, and the distance between the particles after the split is The length of the corresponding direction becomes The volume, mass and internal energy are half of the original, and the strain, stress and density remain unchanged.

[0133] In step 1.5, the implementation method of the virtual particle method is as follows, and the schematic diagram is shown in Figure 5 . At the beginning of each time step, a boundary grid with the same spacing as the grid in the computational domain is established on the boundary area, and virtual particles are added to it. The solid circle represents the real particle, and the dashed circle represents the virtual particle, whose density, pressure, velocity and volume are all specified as fixed values ​​by the inflow boundary conditions. At the initial moment, the relative position of the virtual particle in the boundary grid is the same as the relative position of the real particle in the computational grid. It can be inferred that in each time step, the virtual and real particles moving at the same speed have the same relative positions in their respective grids. The position of the virtual particle p* can be inferred based on the position of the real particle p.

[0134]

[0135] Where B is the boundary, x B is the coordinate of the boundary, and the virtual particle positions of other boundary grids can be deduced in the same way. The virtual particle p* flows at the specified speed v0, and after a time Δt k , determine whether the virtual particle has entered the calculation area. If so, add the information of the virtual particle to the set of physical particles. The virtual particles will no longer be continuous after being moved, so all virtual particles should be deleted and returned to the initial step to add particles again.

[0136] exist Figure 6The figure shows the modeling of the mine tunnel at the initial moment. On the left is a shock wave with a Mach number of 6. In the middle and lower part, there is a wall with a length of 0.6 meters and a width of 0.2 meters. After that, the density cloud map of each time step can be recorded. The cloud maps at several moments are presented below for analysis. Figure 7 The density cloud diagram at T = 0.134s is shown in Figure 1. At this time, the shock wave bypasses the wall to form a diffraction shock wave. Figure 8 The density cloud map at T = 0.383s is shown in the figure. The shock waves of the flow converge at the center to form a converging flow. The reflection of the shock waves can also be clearly seen on the solid wall at the lower end. Figure 9-11 Under the effect of shock wave reflection from the lower wall, the converging flow began to move upward and generated more eddies, which clearly showed the movement of shock wave in the mine and provided a force criterion for damage analysis during mine explosion.

[0137] Through the above embodiments, it can be seen that the partitioned coupled Euler-Lagrangian method for accurately capturing shock waves disclosed in the present invention can well track the discontinuity course of shock waves and record the pressure of each point in the calculation domain, effectively realizing the accurate numerical simulation calculation of explosion and impact problems, and providing a powerful damage criterion for target damage assessment and analysis.

[0138] The specific description above further illustrates the purpose, technical solutions and beneficial effects of the invention in detail. It should be understood that the above description is only a specific embodiment of the present invention and is not intended to limit the scope of protection of the present invention. Any modifications, equivalent substitutions, improvements, etc. made within the spirit and principles of the present invention should be included in the scope of protection of the present invention.

Claims

1. A coupled Eulerian-Lagrangian method to accurately capture the shock wave propagation process, Features The steps include: The first step is to create a simulation model for the object to be analyzed and initialize the created simulation model. The initialization settings include determining the size of the calculation domain, the position information and geometric size information of various materials in the calculation domain, the grid step size and coordinates, the arrangement of particles, material properties and parameters, boundary conditions and virtual particles, and the initial calculation control parameters. In the second step, according to the simulation model created in the first step, at the beginning of each time step, the mass points are divided into A partitions using the partition weighted bidirectional mapping method according to the topological relationship between the grid and the mass points, the state equation and constitutive model corresponding to the mass points are calculated, and the physical quantities of the mass points are updated; In the second step, the partition weighted bidirectional mapping method is implemented as follows: at the kth time step, the grid domain is regularly distributed along the i and j directions, X represents the Xth grid node in the i direction, Y represents the Yth grid node in the j direction, and the volume domain of the particle p is assumed to be a rectangle, The structure vector defines the particle domain, and the volume domain is defined by the grid lines. and The nth A area divided is denoted as p An , the volume is denoted as V pAn , define the weight ω of each A partition in the mass point pAn for ω pAn =V pAn / V p (1) Where V p is the total volume of particle p, subscript p represents the particle, subscript pAn represents the nth large partition A in the volume domain of particle p. When mapping from the grid to the particle, it is first mapped to the partitions of the particle, that is, In the formula, For partition p An The physical quantity, n g The number of nodes required for shape function calculation, subscript I represents the I-th mesh node, f I is the physical quantity of the I node, S IpAn =S I (x pAn ) is the shape function, xp A n is partition p A The centroid coordinates of n; The physical quantity f of the particle p is obtained by weighted calculation of each partition p , Where n pA is the total number of partitions A in the particle p; When a particle is mapped to a grid, the physical quantity of the particle is first assigned to each partition according to the weight, and then mapped from each partition to the grid. When the particle moves to the position of the next time step, the particle needs to carry the cut state and be cut by the grid at the new position. After the particle moves to the new position, the positional relationship between the particle domain and the grid domain is re-determined. At the k+1th time step, the large partition p An New Grid Lines and Continue to split into small partitions p Bn-m , the weight of each small partition is allocated according to the proportion of its volume in the large partition, then the formula for mapping the mass point to the grid is: Where n pAnB is the total number of small partitions B in the nth large partition A of the particle p, and the subscript pBn-m represents the mth small partition B in the nth large partition A of the particle p. S IpBn-m =S I (x pBn-m ) is the shape function, x pBn-m For small partition p Bn-m The centroid position of The third step is to map the physical quantity of the particle updated in the second step to the grid and update the grid velocity quantity; The fourth step is to map the grid velocity quantity updated in the third step to the mass point, and calculate the velocity and displacement of the mass point; The fifth step is to determine the positions of all the mass points in the simulation model created in the first step, delete the real mass points outside the calculation domain, and add the virtual mass points in the calculation domain to the real mass point set; Step 6: Add virtual particles again according to the boundary conditions; Step 7: Based on partition A, the particles are further divided into partition B, the physical quantities of the particles are mapped to the grid, the oscillation of the discontinuity surface is suppressed by the smoothing function, and the physical quantities of the grid at the next time step are obtained; Step 8, according to the set termination condition, the pressure of the shock wave is obtained; The ninth step is to provide guidance and evaluation on the damage to personnel caused by the mine gas explosion based on the obtained shock wave pressure.

2. A coupled Euler-Lagrangian method for accurately capturing the shock wave propagation process according to claim 1, Features: In the second step, the method for updating the physical quantity of the particle is: Step 2.1: Calculate the strain increment of the particle based on the grid velocity gradient In the formula, is the grid velocity, Δt k is the time step; Step 2.2: Calculate the Cauchy stress of the particle based on the constitutive model and the equation of state And update the density of the particle Step 2.3, update the energy of the particle according to the state equation and artificial viscosity; Step 2.4: Determine the volume of the particle. When the length of one direction of the particle exceeds a preset threshold, the particle is split using the particle splitting method.

3. A coupled Euler-Lagrangian method for accurately capturing the shock wave propagation process according to claim 2, Features: In the third step, the method for updating the grid velocity is: Step 3.1, according to the topological relationship between the particle and the grid, map the physical quantity of the particle to the grid; Step 3.2, update the momentum of the grid nodes; In the formula, is the grid stress, is the particle pressure.

4. A coupled Euler-Lagrangian method for accurately capturing the shock wave propagation process according to claim 3, Features: In the fourth step, the method for calculating the velocity and displacement of the particle is: Step 4.1, map the quantity of the grid node to the mass point according to the partition weighted bidirectional mapping method, and update the displacement and velocity change of the mass point; Step 4.2, calculate and update the position and velocity of the particle in the time step; 5. A coupled Euler-Lagrangian method for accurately capturing the shock wave propagation process according to claim 4, Features: In the sixth step, according to the boundary conditions, the method of adding virtual particles again is: According to the inflow, outflow and periodic boundary conditions, new virtual particles are added in the boundary area, and the relationship between the virtual particles and the real particles conforms to the boundary conditions.

6. A coupled Euler-Lagrangian method for accurately capturing the shock wave propagation process according to claim 5, Features: In the seventh step, the physical quantity of the particle is updated again by a smooth function; The method of mapping the mass point quantity to the grid node is:

7. A coupled Euler-Lagrangian method for accurately capturing the shock wave propagation process according to claim 6, Features: In the eighth step, if the number of steps currently calculated meets the output setting conditions, all simulation data are merged and output and the serial number mark of the output file is recorded; the variables and variable names contained in the grid and the particle are written into the file; If the end condition is not met, return to the second step and start a new time step; If the end condition is met, the calculation ends and the pressure of the shock wave is obtained.

8. A coupled Euler-Lagrangian method for accurately capturing the shock wave propagation process according to claim 7, Features: In step 2.4, the particle splitting method is implemented as follows: the particle domain is assumed to be a rectangle, The structure vector defines the particle domain, so the length of the particle in two directions can be defined as When the length in any of the above directions meets the following splitting requirements, the particle splitting is performed Where α is the specified splitting factor, which is less than 1; and is the length of the grid. When the particle p is When the length in the direction reaches the splitting requirement, The direction is split in half into p 1 and p 2 , the distance between the particles after splitting is 0.5 The length of the corresponding direction becomes 0.5 The volume, mass and internal energy are half of the original, and the strain, stress and density remain unchanged.

9. A coupled Euler-Lagrangian method for accurately capturing the shock wave propagation process according to claim 8, Features: In the sixth step, the implementation method of the virtual particles is as follows: at the beginning of each time step, a boundary grid with the same spacing as the grid in the calculation domain is established on the boundary area, and virtual particles are added therein. The solid circle represents the real particle, and the dotted circle represents the virtual particle. Its density, pressure, velocity and volume are all specified as fixed values ​​by the inflow boundary conditions. At the initial moment, the relative position of the virtual particle in the boundary grid is the same as the relative position of the real particle in the calculation grid. It can be inferred that in each time step, the virtual and real particles moving at the same speed have the same relative positions in their respective grids. The position of the virtual particle p* can be inferred based on the position of the real particle p. Where B is the boundary, x B is the coordinate of the boundary. Similarly, the position of the virtual particles of other boundary grids can be deduced. The virtual particle p* follows the specified speed v 0 Flow, after time Δt k , determine whether the virtual particle has entered the calculation area. If so, add the information of the virtual particle to the set of physical particles. After the movement, the virtual particles will no longer be continuous, so all virtual particles must be deleted and return to the initial step to add the particles again.