A numerical grid model optimization algorithm for PBXs targeting voronoi modeling process

Abstract

The application provides a numerical grid model optimization algorithm for a Voronoi modeling process of a PBX, and the numerical grid model of the PBX is a two-dimensional or three-dimensional fine numerical grid model. The algorithm eliminates the large-width-to-height-ratio cells in the two-dimensional or three-dimensional fine numerical grid model by searching the smallest cell boundary in the two-dimensional or three-dimensional fine numerical grid model, so as to optimize the numerical grid model of the PBX. The algorithm can not only realize the automatic search and determination of the smallest boundary of the crystal cell, the smallest boundary of the binder cell and the smallest boundary of the whole model cell based on the established two-dimensional or three-dimensional numerical grid model, but also realize the node merging and cell deletion based on the search result, improve the grid topology quality, increase the stability and calculation efficiency of the numerical calculation, and avoid the calculation termination caused by the distortion of the poor-quality cell.

Description

Technical Field

[0001] This invention belongs to the field of numerical computation technology using computer models, and relates to numerical grid models of PBX, specifically to an optimization algorithm for numerical grid models of PBX for Voronoi modeling processes. Background Technology

[0002] Polymer-bonded explosives (PBX) are highly packed polymer-based composite materials, typically composed of high-energy elemental explosive particles, polymer binders, and related additives. Given that the elastic modulus of the explosive particles is much higher than that of the binder, it can be considered a rigid, highly packed polymer-based composite material. Due to its high energy density, low sensitivity, high strength, high detonation velocity, good physical stability, and ease of machining, PBX has been widely used in missile, rocket, and anti-tank weapon charges.

[0003] Currently, scholars have proposed mature construction procedures for establishing PBX mesoscopic two-dimensional and three-dimensional polycrystalline numerical mesh models, among which the mesoscopic model construction method based on the Voronoi polygon algorithm has been widely studied and adopted. However, in the process of generating PBX mesoscopic numerical mesh models based on random distribution in Voronoi space, due to the limitations of the algorithm itself, extremely small boundaries easily appear at a certain boundary between the crystal and the binder. This leads to the formation of meshes with extremely large aspect ratios at these locations during subsequent mesh generation. The appearance of such extremely small boundaries will cause three main drawbacks: firstly, it will lead to a sharp decrease in mesh size, a surge in the number of meshes, and a deterioration in the transition characteristics of mesh size; secondly, if deformation occurs in this region, some meshes are prone to large deformations, leading to simulation termination due to negative volumes; and thirdly, it will drastically reduce the numerical simulation calculation time step, resulting in a surge in overall computation time. Summary of the Invention

[0004] To address the shortcomings of existing technologies, the present invention aims to provide a numerical grid model optimization algorithm for PBX in the Voronoi modeling process, thereby solving the technical problem that the stability and computational efficiency of numerical calculations in existing technologies need to be further improved.

[0005] To solve the above-mentioned technical problems, the present invention adopts the following technical solution.

[0006] A numerical mesh model optimization algorithm for PBX in Voronoi modeling process is disclosed. The numerical mesh model of PBX is a two-dimensional or three-dimensional mesoscopic numerical mesh model. The algorithm optimizes the numerical mesh model of PBX by searching for the smallest cell boundary in the two-dimensional or three-dimensional mesoscopic numerical mesh model and eliminating cells with large aspect ratios inside the two-dimensional or three-dimensional mesoscopic numerical mesh model.

[0007] The method includes the following steps.

[0008] Step 1: Based on the Voronoi polygon modeling process, establish a numerical mesh model of the PBX and save it as a mesoscale mesh file in .dat format.

[0009] Step two: Using the JavaScript programming language, read the fine mesh file and create an array of all cells. ArrayE _T and the array of all nodes ArrayN _T .

[0010] Step 3: Based on the ID numbers of the crystal units and the binder units, first establish the crystal unit array. ArrayE _C and crystal unit node array ArrayN _C Then, an array of adhesive units is established. ArrayE _B and adhesive unit node array ArrayN _B .

[0011] Step 4, based on the crystal unit array ArrayE _C and crystal unit node array ArrayN _C Obtain the minimum size of the crystal unit boundary. L min_C Based on the adhesive unit array ArrayE _B and adhesive unit node array ArrayN _B Obtain the minimum size of the adhesive unit boundary. L min_B .

[0012] Step 5: Based on the minimum size of the crystal unit boundary L min_C Minimum dimension of the bond unit boundary L min_B Obtain the minimum element boundary size of the numerical mesh model of the entire PBX. L min_T .

[0013] Step 6: Based on the minimum element boundary size of the entire PBX numerical mesh model. L min_T Obtain the cell ID number to which the cell boundary belongs. E ID_minT and cell node ID array ArrayN ID_minT .

[0014] Step 7, based on the unit ID number E ID_minT and cell node ID array ArrayN ID_minT The spatial location of the unit is determined, and the unit nodes are deleted or merged.

[0015] Step 8: Renumber the merged element nodes in the mesome file and save it as a mesome file in .dat format.

[0016] Step 9: Repeat steps 2 through 8 to complete the multi-round iterative optimization of the numerical grid model of the PBX.

[0017] In step seven, the specific method for deleting and merging unit nodes is as follows:

[0018] Step 701: If the smallest grid size cell is located inside the crystal, first delete the crystal cell with the smallest grid size, and then merge the surrounding crystal cells with the smallest grid size.

[0019] Step 702: If the smallest grid size cell is located inside the adhesive, first delete the smallest grid size adhesive cell in this adhesive cell, and then merge the surrounding adhesive cells of the smallest grid size.

[0020] Step 703: If the smallest grid size unit is located at the crystal-binder interface, firstly delete the smallest grid size unit of the binder or crystal, then fuse the surrounding crystal units of the smallest grid size, and finally fuse the surrounding binder units of the smallest grid size.

[0021] Compared with the prior art, the present invention has the following technical effects.

[0022] The algorithm of this invention can not only automatically search and determine the minimum boundary of crystal unit, minimum boundary of binder unit, and minimum boundary of all unit in the established two-dimensional and three-dimensional numerical mesh model, but also realize node merging and unit deletion based on the search results, thereby improving the mesh topology quality, increasing the stability and efficiency of numerical calculation, and avoiding the termination of calculation due to distortion caused by poor quality units. Attached Figure Description

[0023] Figure 1It is a two-dimensional mesoscopic numerical mesh model of PBX constructed based on the Voronoi polygon modeling process in this invention.

[0024] Figure 2 This is the mesh partition diagram of region A, where the minimum boundary of the full model unit is located, obtained from the first round of search.

[0025] Figure 3 This is the mesh partition diagram of region A after deletion and merging.

[0026] Figure 4 It is a graph showing the size change of the minimum boundary of the unit during multiple iterations.

[0027] The specific content of the present invention will be further explained in detail below with reference to the embodiments. Detailed Implementation

[0028] It should be noted that, unless otherwise specified, all models and methods in this invention adopt models and methods known in the prior art.

[0029] In the current process of generating PBX mesoscopic numerical mesh models based on random distribution in Voronoi space, due to the limitations of the algorithm itself, it is very easy for a tiny boundary to appear at a certain boundary between the crystal and the binder, which in turn leads to the formation of a mesh with a very large aspect ratio at that point in the subsequent mesh generation process.

[0030] To increase the stability and efficiency of numerical calculations and avoid calculation termination due to distortion caused by poor-quality cells, this invention proposes a PBX mesoscopic numerical mesh optimization algorithm based on the Javascript programming language for the Voronoi polygon modeling process. On the one hand, it realizes the search for the minimum boundary size of crystal cells, the minimum boundary size of binder cells, and the minimum boundary size of the entire model in the PBX numerical mesh. On the other hand, it optimizes the mesh topology based on the mesh topology relationship and the cell and node information corresponding to the minimum size boundary.

[0031] The mesh optimization algorithm of this invention uses Javascript as the programming language. First, it searches for the minimum boundary size of crystal cells, the minimum boundary size of binder cells, and the minimum boundary size of the entire PBX numerical mesh. Second, it optimizes the mesh topology based on the mesh topological relationships and the cell and node information corresponding to the minimum boundary size. This invention provides researchers and engineers with a PBX mesoscopic numerical mesh optimization algorithm for Voronoi polygon modeling processes. It can improve mesh generation quality based on the optimization algorithm, supporting efficient and accurate numerical simulation calculations, and can be used in fields such as two-dimensional and three-dimensional numerical simulation of PBX samples.

[0032] The following are specific embodiments of the present invention. It should be noted that the present invention is not limited to the following specific embodiments. All equivalent modifications made based on the technical solutions of this application fall within the protection scope of the present invention.

[0033] Example: This embodiment presents a numerical mesh model optimization algorithm for PBXs using the Voronoi modeling process. The PBX's numerical mesh model is a two-dimensional or three-dimensional mesoscale numerical mesh model. This algorithm optimizes the PBX's numerical mesh model by searching for the smallest element boundary in the two-dimensional or three-dimensional mesoscale numerical mesh model, eliminating elements with large aspect ratios within the model. This improves mesh generation quality and supports accurate simulation of the PBX's mechanical response characteristics under static, quasi-static, and dynamic loads.

[0034] The method includes the following steps.

[0035] Step 1: Based on the Voronoi polygon modeling process, establish a numerical mesh model of the PBX and save it as a mesoscale mesh file in .dat format.

[0036] In step one, based on the crystal gradation characteristics required for modeling, a series of random circles (two-dimensional model) or spheres (three-dimensional model) corresponding to the gradation are first generated using methods such as Monte Carlo mapping. Next, Voronoi polygons (two-dimensional) or Voronoi polyhedra (three-dimensional) are generated using the center of the circles / spheres as the center points of the Voronoi polygons. Then, the Voronoi polygons or polyhedra are shrunk according to the specified crystal and binder volume fractions to generate an interparticle bonding layer, resulting in the mesoscopic geometric model of the PBX. Subsequently, the established geometric model is imported into preprocessing software for mesh generation, obtaining a PBX mesoscopic two-dimensional or three-dimensional mesoscopic numerical mesh model suitable for numerical simulation, and saved as a .dat format mesoscopic mesh file, Model.dat. Specifically, in this embodiment, the two-dimensional mesoscopic numerical mesh model of the PBX constructed based on the Voronoi polygon modeling process is as follows: Figure 1 As shown.

[0037] Step two: Using the JavaScript programming language, read the fine mesh file and create an array of all cells. ArrayE _T and the array of all nodes ArrayN _T .

[0038] In step two, the Javascript programming language is used to read the PBX mesome file Model.dat generated in step one, as shown in Table 1. The first row represents the total number of nodes in this model. N Node and total number of unitsN Elem , line 2 to line 3 N Node The +1 row contains node data, consisting of the node ID, X-coordinate, Y-coordinate, and Z-coordinate, in that order. N Node +2 lines~ N Node + N Elem The +1 row contains the cell data, consisting of the cell ID, the ID of node 1, the ID of node 2, the ID of node 3, the ID of node 4, the ID of node 5, the ID of node 6, the ID of node 7, the ID of node 8, and the cell group number. The data in the mesoscale mesh file is read and parsed using the JavaScript programming language to generate the complete cell array. ArrayE _T and the array of all nodes ArrayN _T array ArrayE _T The number of elements in the middle is the total number of units in the model. N Elem array ArrayN _T The number of elements in the middle is the total number of nodes in the model. N Node .

[0039] Table 1. Example of the PBX's fine mesh file Model.dat

[0040] Step 3: Based on the ID numbers of the crystal units and the binder units, first establish the crystal unit array. ArrayE _C and crystal unit node array ArrayN _C Then, an array of adhesive units is established. ArrayE _B and adhesive unit node array ArrayN _B .

[0041] In step three, specifically, each element in the PBX mesoscale mesh file Model.dat contains 10 data points, the first being the element ID and the tenth being the element group number. For crystalline and binder elements, to ensure accurate determination of whether an element belongs to a crystalline or binder element in subsequent numerical simulation calculations, the tenth data point (i.e., the group number) of the element in the mesh file needs to be assigned different values. Assume the group number of the crystalline element is... I C The group number of the adhesive unit isI B Based on group number I C and I B For all cell arrays in step 2 ArrayE _T The data is read and determined; if a certain unit belongs to a crystal unit, the array is... ArrayE _T The data of this cell is stored in the crystal cell array. ArrayE _C Simultaneously, based on the cell-node topology, the node data belonging to that cell is stored in the crystal cell node array. ArrayN _C If a certain unit belongs to the adhesive unit, the array will be... ArrayE _T The data of this cell is stored in the crystal cell array. ArrayE _B Simultaneously, based on the unit-node topological relationship, the node data belonging to that unit is stored in the adhesive unit node array. ArrayN _B Since a unit can only belong to either a crystal unit or a binder unit, the array... ArrayE _C The number of elements plus the array ArrayE _B The number of elements is equal to the array ArrayE _T The number of elements. However, a large number of nodes are located at the interface between crystal units and binder units, therefore they belong to both crystal units and binder units, hence the array... ArrayN _C The number of elements plus the array ArrayN _B The number of elements is greater than the array ArrayN _T The number of elements.

[0042] Step 4, based on the crystal unit array ArrayE _C and crystal unit node array ArrayN _C Obtain the minimum size of the crystal unit boundary. L min_C Based on the adhesive unit array ArrayE _B and adhesive unit node array ArrayN _B Obtain the minimum size of the adhesive unit boundary. L min_B .

[0043] Specifically, in step four, the crystal unit array... ArrayE_C Traverse each cell in the crystal, and based on the node number corresponding to each side length of the cell, retrieve the data from the crystal cell node array. ArrayN _C Read the spatial coordinates (X, Y, and Z coordinates) of each corresponding node and calculate the length of that side. First, compare all the side lengths of a crystal unit to obtain the minimum side length of that crystal unit. Then, compare the minimum side lengths of all crystal units to obtain the minimum size of the boundaries of all crystal units. L min_C .

[0044] Specifically, in step four, the adhesive unit array... ArrayE _B Iterate through each cell, and based on the node number corresponding to each side length of the cell, retrieve the data from the adhesive cell node array. ArrayN _B Read the spatial coordinates (X, Y, and Z coordinates) of each corresponding node and calculate the length of that side. First, compare all the side lengths of an adhesive element to obtain the minimum side length of that adhesive element. Then, compare the minimum side lengths of all adhesive elements to obtain the minimum dimension of the boundary of all adhesive elements. L min_B .

[0045] Step 5: Based on the minimum size of the crystal unit boundary L min_C Minimum dimension of the bond unit boundary L min_B Obtain the minimum element boundary size of the numerical mesh model of the entire PBX. L min_T .

[0046] In step five, specifically, if L min_C < L min_B It is determined that the unit boundaries of the entire model occur at the crystal unit boundaries. L min_T = L min_C .like L min_B < L min_C It was determined that the element boundaries of the entire model occurred at the binder element boundaries. L min_T = L min_B .

[0047] Step 6: Based on the minimum element boundary size of the entire PBX numerical mesh model. Lmin_T Obtain the cell ID number to which the cell boundary belongs. E ID_minT and cell node ID array ArrayN ID_minT .

[0048] In step six, specifically, the minimum size of the element boundary of the entire model is used. L min_T First, read the two node numbers corresponding to the cell boundaries stored in step four, and then process the crystal cell array. ArrayE _C Each cell is traversed, and it is determined whether the two node numbers are located at a boundary of a certain crystal cell. If so, the cell ID number to which the cell boundary belongs is obtained. E ID_minT And based on the unit-node topology, obtain the node ID array of the unit. ArrayN ID_minT If not, for the adhesive cell array ArrayE _B Each cell is traversed, and it is determined whether the two node numbers are located at a boundary of a certain adhesive cell. If so, the cell ID number to which the cell boundary belongs is obtained. E ID_minT And based on the unit-node topology, obtain the node ID array of the unit. ArrayN ID_minT .

[0049] Step 7, based on the unit ID number E ID_minT and cell node ID array ArrayN ID_minT The spatial location of the unit is determined, and the unit nodes are deleted or merged.

[0050] In step seven, the specific methods for deleting and merging unit nodes are as follows.

[0051] Step 701: If the smallest grid size cell is located inside the crystal, first delete the crystal cell with the smallest grid size, and then merge the surrounding crystal cells with the smallest grid size.

[0052] Step 702: If the smallest grid size cell is located inside the adhesive, first delete the smallest grid size adhesive cell in this adhesive cell, and then merge the surrounding adhesive cells of the smallest grid size.

[0053] Step 703: If the smallest grid size unit is located at the crystal-binder interface, firstly delete the smallest grid size unit of the binder or crystal, then fuse the surrounding crystal units of the smallest grid size, and finally fuse the surrounding binder units of the smallest grid size.

[0054] In this embodiment, the mesh partition diagram of region A, where the minimum boundary of the full model unit is located, obtained from the first round of search, is as follows: Figure 2 As shown. The mesh breakdown diagram of region A after deletion and merging is as follows. Figure 3 As shown.

[0055] Step 8: Renumber the merged element nodes in the mesome file and save it as a mesome file in .dat format.

[0056] In step eight, specifically for a two-dimensional mesoscopic numerical mesh model, if the deleted element is a triangular element, the number of elements decreases by 2, and the number of nodes decreases by 1. For a three-dimensional mesoscopic numerical mesh model, if the deleted element is a tetrahedral element, the number of elements decreases by 4, and the number of nodes decreases by 1. After the elements and nodes are numbered, save the mesh file in .dat format.

[0057] Step 9: Repeat steps 2 through 8 to complete the multi-round iterative optimization of the numerical grid model of the PBX.

[0058] In this embodiment, the specific curve of the size change of the minimum boundary of the unit during multiple iterations is shown in the figure below. Figure 4 As shown.

Claims

1. A numerical mesh model optimization algorithm for a PBX (Personalized Power Grid) in a Voronoi modeling process, wherein the numerical mesh model of the PBX is a two-dimensional or three-dimensional mesoscopic numerical mesh model, characterized in that... This algorithm optimizes the numerical mesh model of PBX by searching for the smallest element boundary in a two-dimensional or three-dimensional mesoscopic numerical mesh model, thereby eliminating elements with large aspect ratios within the model.

2. A numerical mesh model optimization algorithm for a PBX (Personalized Power Grid) in a Voronoi modeling process, wherein the numerical mesh model of the PBX is a two-dimensional or three-dimensional mesoscopic numerical mesh model, characterized in that... The method includes the following steps: Step 1: Based on the Voronoi polygon modeling process, establish the numerical mesh model of the PBX and save it as a mesoscale mesh file in .dat format. Step two: Using the JavaScript programming language, read the fine mesh file and create an array of all cells. ArrayE _T and the array of all nodes ArrayN _T ; Step 3: Based on the ID numbers of the crystal units and the binder units, first establish the crystal unit array. ArrayE _C and crystal unit node array ArrayN _C Then, an array of adhesive units is established. ArrayE _B and adhesive unit node array ArrayN _B ; Step 4, based on the crystal unit array ArrayE _C and crystal unit node array ArrayN _C Obtain the minimum size of the crystal unit boundary. L min_C Based on the adhesive unit array ArrayE _B and adhesive unit node array ArrayN _B Obtain the minimum size of the adhesive unit boundary. L min_B ; Step 5: Based on the minimum size of the crystal unit boundary L min_C Minimum dimension of the bond unit boundary L min_B Obtain the minimum element boundary size of the numerical mesh model of the entire PBX. L min_T ; Step 6: Based on the minimum element boundary size of the entire PBX numerical mesh model. L min_T Obtain the cell ID number to which the cell boundary belongs. E ID_minT and cell node ID array ArrayN ID_minT ; Step 7, based on the unit ID number E ID_minT and cell node ID array ArrayN ID_minT Determine the spatial location of the unit, and delete or merge the unit nodes; Step 8: Renumber the merged element nodes in the mesome file and save it as a mesome file in .dat format; Step 9: Repeat steps 2 through 8 to complete the multi-round iterative optimization of the numerical grid model of the PBX.

3. The numerical grid model optimization algorithm for PBX in the Voronoi modeling process as described in claim 2, characterized in that, In step seven, the specific method for deleting and merging unit nodes is as follows: Step 701: If the smallest grid size cell is located inside the crystal, first delete the crystal cell with the smallest grid size, and then merge the surrounding crystal cells with the smallest grid size. Step 702: If the smallest grid size cell is located inside the adhesive, first delete the smallest grid size adhesive cell in this adhesive cell, and then fuse the surrounding adhesive cells of the smallest grid size. Step 703: If the smallest grid size unit is located at the crystal-binder interface, firstly delete the smallest grid size unit of the binder or crystal, then fuse the surrounding crystal units of the smallest grid size, and finally fuse the surrounding binder units of the smallest grid size.

Images