Three-grid system construction method for pretreatment of variable-density grids of solid shell of large shell component
By rapidly constructing locally refined hexahedral meshes with controllable density using a three-mesh system, the problems of low mesh quality and difficulty in density control in the spinning process are solved. This achieves an efficient simulation method, shortens simulation time, and supports the rapid manufacturing of large-size thin-walled components.
Patent Information
- Application Number
- CN202511521642.0
- Authority / Receiving Office
- CN · China
- Patent Type
- Applications(China)
- Current Assignee / Owner
- Filing Date
- 2025-10-23
- Publication Date
- 2026-02-10
AI Technical Summary
Existing technologies using incremental forming processes such as spinning suffer from low computational mesh quality and difficulty in controlling mesh density, resulting in excessively long simulation times and failing to meet the high-efficiency simulation requirements for large-sized, thin-walled, complex components.
A three-grid system construction method is adopted, which uses the preprocessing of large shell components with variable density meshes. By generating a high-resolution structured storage mesh SM, combining it with the background mesh BM and the computational mesh CM, and dynamically marking the encryption attributes, a local encryption full hexahedral mesh is quickly constructed using a multi-level encryption template, so as to achieve flexible control of mesh density.
It significantly reduces the computational scale of simulation models, saves 50% of simulation time, ensures simulation accuracy, adapts to dynamic changes in the plastic deformation region, and possesses versatility and high efficiency, supporting low-energy consumption and short-cycle manufacturing of large-size thin-walled complex components.
Smart Images
Figure CN121502998A_ABST
Abstract
Description
Technical Field
[0001] This invention belongs to the field of numerical simulation of plastic forming processes, specifically involving a three-grid system construction method for preprocessing variable density meshes of solid shell components for large shell components. Background Technology
[0002] In the manufacturing of high-end equipment such as aviation and aerospace, with the increasing demands for lightweight and high reliability of key components of launch vehicles, the integral manufacturing of large-size thin-walled components using incremental forming processes such as spinning has become a key technological approach. This process overcomes the problems of high material consumption, long processing cycles, and heavy equipment loads encountered with traditional machining and stamping processes. However, when using the commonly used finite element method (FEM) to study the forming mechanism and process window of incremental forming processes such as spinning, the point / line contact form requires extremely small element sizes to ensure simulation accuracy. This not only makes the model reach a scale of hundreds of thousands or even millions of elements, but also causes the time increment step to decrease drastically to the microsecond level due to the conditional stability of the solver, accompanied by a tedious mesh generation process. As a result, a process that takes only tens of seconds can take weeks or even months to simulate. Therefore, there is an urgent need for an innovative and cost-effective simulation method to explore the process window and forming mechanism. Therefore, the research proposes that efficient simulation technology can fill the gap in simulation tools for large-sized thin-walled complex components, which is of great significance for accelerating process design and shortening the development cycle of high-end aerospace equipment.
[0003] The adaptive mesh method can effectively reduce the computational scale of the model by retaining only the fine mesh in the concentrated plastic deformation region, thereby significantly reducing the simulation time. The multi-mesh method adds a high-resolution mesh to the adaptive mesh method to store historical data during the simulation process, which avoids the loss of data accuracy caused by the periodic change of mesh density. However, the use of the multi-mesh method in incremental forming simulation also faces the following problems: (1) lack of a method for quickly constructing a high-quality fully hexahedral locally refined computational mesh; (2) difficulty in controlling the degree of refinement of the plastic deformation region in the computational mesh.
[0004] Therefore, there is an urgent need for a fast and flexible algorithm for constructing locally refined hexahedral meshes to quickly generate computational meshes with controllable density based on the location of plastic deformation. Summary of the Invention
[0005] The technical problem to be solved: To overcome the shortcomings of existing technologies, this invention provides a three-grid system construction method for preprocessing variable-density meshes of large shell components. This method aims to rapidly construct locally refined, fully hexahedral computational meshes with controllable density for rotating workpieces, significantly reducing the computational scale of incremental forming simulation models while ensuring accuracy. This promotes the development of advanced manufacturing technologies and provides technical support for low-energy, short-cycle, and lightweight integrated manufacturing of large-size, thin-walled complex components. This invention solves the problems of low computational mesh quality and difficulty in controlling mesh density in current multi-grid methods.
[0006] The technical solution of this invention is: a three-grid system construction method for preprocessing variable-density grids of solid shell components for large shell structures, comprising the following steps: Based on the geometric parameters of the target thin-walled rotating body component, a high-resolution structured hexahedral storage grid SM is generated for storing historical simulation data. By aggregating the cells in the storage grid SM at intervals, a background grid BM is generated that shares nodes with the storage grid SM and has the same topology. The background grid BM serves as a topological framework for quickly marking encrypted areas. Based on the changes in the plastic deformation zone during the simulation, the encryption attributes of the elements to be encrypted are dynamically marked in the background mesh BM; Based on the encryption attributes marked in the background grid BM, a preset multi-level encryption template is invoked. Without generating additional nodes, the nodes and cells of the storage grid SM and the background grid BM are combined to quickly generate a local encryption computing grid CM that shares nodes with the storage grid SM and the background grid BM. The storage grid SM, background grid BM, and computational grid CM together form a three-grid system that overlaps with each other and shares nodes. Through the bridging role of the background grid BM, a fast and coordinated conversion from data storage to computational grid generation is achieved to adapt to the dynamic changes of the deformation zone in incremental forming simulation and ensure simulation accuracy.
[0007] A further technical solution of the present invention is: the structured hexahedral storage grid SM is implemented using a tiling method, specifically: Based on the inner diameter, thickness and axial length of the thin-walled rotating body component, hexahedral units are sequentially laid along the circumferential, radial and axial directions to form a structured hexahedral storage grid SM; Each node and element is uniquely indexed and located by three numbers in the circumferential, radial, and axial directions.
[0008] A further technical solution of the present invention is: the paving method specifically includes: The storage grid SM is divided into sections in the circumferential, radial, and axial directions, respectively. , and Each unit, thus having in the circumferential, radial and axial directions respectively. , and One node; Each node is identified by its three numbers in the circumferential, radial, and axial directions. Index it, its overall numbering Calculated using the following formula: Coordinates of each node Calculated using the following formula: Where D is the inner diameter of the thin-walled rotating component, T is the thickness of the thin-walled rotating component, and L is the axial length of the thin-walled rotating component; each unit is identified by its three numbers in the circumferential, radial, and axial directions. Index it, its overall numbering Calculated using the following formula:
[0009] Furthermore, each unit consists of 8 connected nodes, and the global numbering of the 8 nodes is based on the unit numbering. Sure.
[0010] A further technical solution of the present invention is: the generation of the background mesh BM specifically includes: Set a cohesion interval that is related to the selected encryption template type. α ; The storage grid SM is spaced at intervals in both the circumferential and axial directions. α -1) Merge the elements, retaining all radial elements, to generate a storage grid SM with element sizes in both the circumferential and axial directions. α Background mesh BM (multiplied by 100%); The background grid BM shares all nodes with the storage grid SM, and the topology of the background grid BM is directly inherited from the storage grid SM. A further technical solution of the present invention is: the condensation interval α The value is related to the encryption template type: When using a multi-level four-part encryption template k represents the encryption level; When using a multi-level nine-point encryption template k represents the encryption level.
[0011] A further technical solution of the present invention is: the basis for dynamically marking the encryption attributes of the unit to be encrypted in the background grid BM includes at least one of the following: Distribution of field variables during simulation; The element size field is calculated from the posterior error estimate; Real-time tracking of the tool's spatial position; A pre-defined fixed encryption zone range.
[0012] A further technical solution of the present invention is: the invocation of the preset multi-level encryption template specifically includes: Provides multi-level nine-point encryption templates and / or multi-level four-point encryption templates; The multi-level nine-part encryption template achieves a compact transition from the encrypted area to the unencrypted area by nesting the single-level nine-part template and its variants, and by using mirroring operations. The multi-level four-part encryption template restores self-symmetry by merging 2x2 background grid BM units, thereby constructing an encryption template capable of multi-level transition.
[0013] A further technical solution of the present invention is that the encryption template is configured into multiple types according to the number and distribution of encryption nodes in a quadrilateral face of the background mesh BM hexahedral unit, including at least: no encryption template, convex corner encryption template, edge encryption template, diagonal encryption template, concave corner encryption template and full encryption template; For thin-walled rotating body components, the quadrilateral transition template is stretched along the thickness direction to form a corresponding hexahedral transition template. A further technical solution of the present invention is: the step of calling the preset multi-level encryption template also includes template adaptation application: When the node encryption attribute marker position of the cell to be encrypted in the background mesh BM is inconsistent with the preset orientation of the standard encryption template, the standard encryption template is rotated to match the actual encryption node distribution of the cell to be encrypted; wherein, the local numbering of the nodes within the rotated template... The rotation angle is determined based on a predefined mapping relationship. The global number of any node in the template is calculated according to the following formula:
[0014] in, Here is the current template number, and n is the template size. In a multi-level nine-point encryption template... In the multi-level four-part encryption template k represents the encryption level.
[0015] A three-grid system for preprocessing variable-density meshes in solid shell components of large shell structures includes: The storage grid generation module is used to generate a high-resolution structured hexahedral storage grid for storing historical simulation data based on the geometric parameters of the target thin-walled rotating body component. The background grid generation module is used to generate a background grid that shares nodes with the storage grid and has the same topology by aggregating the cells in the storage grid at intervals. The background grid serves as a topological framework for marking encrypted areas. The encryption attribute marking module is used to dynamically mark the encryption attributes of the unit to be encrypted in the background mesh according to the changes in the plastic deformation zone during the simulation process. The computational grid generation module is used to call a preset multi-level encryption template based on the encryption attributes marked in the background grid, and combine the nodes and cells of the storage grid and the background grid to generate a local encrypted computational grid that shares nodes with the storage grid and the background grid without generating additional nodes. The storage grid, background grid, and computing grid together form a three-grid system that overlaps with each other and shares nodes.
[0016] Beneficial effects The beneficial effects of this invention are as follows: This invention employs an overlapping three-mesh system and a multi-level refinement template, enabling the rapid construction of locally refined, fully hexahedral meshes for thin-walled solids of revolution models with controllable density. Specific effects are analyzed below: 1. In this method, the connection relationships of the element node topology in the transition zone and the densification zone mesh are known in advance, and no additional nodes besides those in the SM are generated, so mesh construction is quick and simple. By comparing the changes in element properties in the BM, the method proposed in this invention can control the mesh density by locally modifying the mesh topology during simulation. Combining this method with the multi-mesh method can save nearly 50% of the simulation time due to the significant reduction in model size.
[0017] 2. In this method, the historical variables and shape details stored in the SM (Structured Module) ensure the accuracy of the simulation. Traditional multi-mesh methods require independent parsing of the complex topology of the SM, leading to data and topology separation. Furthermore, multi-level template encryption requires multiple recursive traversals of nodes and progressive encryption, resulting in inefficient dynamic response. In contrast, the three-mesh system of this invention achieves rapid CM reconstruction through the bridging role of the BM (Model Module), solving the dual bottlenecks of fragmented data management and high latency in mesh generation in traditional methods. Moreover, since this invention does not limit the specific shape of the FEM solver and deformable body, it can be extended to rapid numerical simulation of any FEM software platform and any incremental forming process, exhibiting a certain degree of versatility. 3. This method achieves precise and flexible control over the degree of mesh refinement. The refinement operation is entirely based on the labeling of BM element properties, and the labeling basis can be field variables, posterior errors, or even real-time tracking of the mold position. This mechanism enables mesh refinement to intelligently "follow" the development of plastic deformation, dynamically concentrating computational resources in the most critical regions, thereby minimizing the computational scale while ensuring accuracy. Attached Figure Description
[0018] Figure 1 This is a schematic diagram of a grid system using the multigrid method.
[0019] Figure 2 This is a flowchart of the multi-grid simulation for incremental forming of rotating body components.
[0020] Figure 3 This is a schematic diagram of generating a structured hexahedral cylindrical mesh using the paving method.
[0021] Figure 4 This is a schematic diagram of the background grid generated by the aggregation of storage grid cells.
[0022] Figure 5 This is a schematic diagram of the construction of a multi-level nine-point encryption template.
[0023] Figure 6 This is a schematic diagram of the construction of a multi-level four-part encryption template.
[0024] Figure 7 The images show the effect of a locally encrypted mesh generated from a background mesh marked with template attributes. (a) and (c) are background meshes marked with attributes using multi-level nine-part and four-part templates, respectively, and (b) and (d) are locally encrypted computational meshes using two-level nine-part and four-part templates, respectively.
[0025] Figure 8 This is a display diagram showing the calculation and storage mesh results of single-spinning flow spinning simulation using a two-level nine-part template.
[0026] Figure 9 The data represents the time consumption for simulating flow spinning using the multigrid method. (a) represents the calculation time using the two-level nine-part template model, and (b) represents the calculation time using the two-level four-part template model. Detailed Implementation
[0027] The embodiments described below with reference to the accompanying drawings are exemplary and intended to explain the invention, and should not be construed as limiting the invention.
[0028] Based on the problems existing in the prior art, this invention proposes a three-grid system construction method for preprocessing variable-density grids of large shell components, including the following steps: Based on the geometric parameters of the target thin-walled rotating body component, a high-resolution structured hexahedral storage grid SM is generated for storing historical simulation data. By aggregating the cells in the storage grid SM at intervals, a background grid BM is generated that shares nodes with the storage grid SM and has the same topology. The background grid BM serves as a topological framework for quickly marking encrypted areas. Based on the changes in the plastic deformation zone during the simulation, the encryption attributes of the elements to be encrypted are dynamically marked in the background mesh BM; Based on the encryption attributes marked in the background grid BM, a preset multi-level encryption template is invoked. Without generating additional nodes, the nodes and cells of the storage grid SM and the background grid BM are combined to quickly generate a local encryption computing grid CM that shares nodes with the storage grid SM and the background grid BM. The storage grid SM, background grid BM, and computational grid CM together form a three-grid system that overlaps with each other and shares nodes. Through the bridging role of the background grid BM, a fast and coordinated conversion from data storage to computational grid generation is achieved to adapt to the dynamic changes of the deformation zone in incremental forming simulation and ensure simulation accuracy.
[0029] This invention also proposes a three-grid system for preprocessing variable-density grids in solid shell components of large shell structures, comprising: The storage grid generation module is used to generate a high-resolution structured hexahedral storage grid for storing historical simulation data based on the geometric parameters of the target thin-walled rotating body component. The background grid generation module is used to generate a background grid that shares nodes with the storage grid and has the same topology by aggregating the cells in the storage grid at intervals. The background grid serves as a topological framework for marking encrypted areas. The encryption attribute marking module is used to dynamically mark the encryption attributes of the unit to be encrypted in the background mesh according to the changes in the plastic deformation zone during the simulation process. The computational grid generation module is used to call a preset multi-level encryption template based on the encryption attributes marked in the background grid, and combine the nodes and cells of the storage grid and the background grid to generate a local encrypted computational grid that shares nodes with the storage grid and the background grid without generating additional nodes. The storage grid, background grid, and computing grid together form a three-grid system that overlaps with each other and shares nodes.
[0030] The key to this invention is a three-grid system consisting of an overlapping and node-sharing storage mesh (SM), a background mesh (BM), and a computation mesh (CM) (see [link to invention]). Figure 1The SM (Structured Grid) serves as a high-precision storage container for geometric and field variables. Its structured nature allows the BM (Browser Module) to be directly generated by selecting points at equal intervals. The BM inherits the geometric topology from the SM without additional computation, forming a fast geometric search framework. This process avoids the computational overhead of re-analyzing the SM topology to construct a coarse mesh, as required by traditional methods. Furthermore, the BM acts as the topology hub, and the CM (Computational Grid) construction relies entirely on the BM's topology framework and the SM's node data, without the need to insert new nodes. This eliminates the risks of data redundancy and geometric distortion, forming a closed-loop "data-structure-computation" chain. The above technical solution will be further explained below with reference to the accompanying drawings: In one embodiment, to generate locally refined meshes that meet the aforementioned objectives based on this mesh system, this invention first designs a construction method for structured SMs and BMs based on tiling and element cohesion. Then, it designs two multi-level refinement templates with controllable density for constructing meshes of different densities. Finally, it combines these templates with labeled BMs to achieve rapid generation of locally refined full hexahedral computational meshes. The locally refined mesh construction method proposed in this invention can be integrated into the multi-mesh simulation process for incremental forming of rotating body components (e.g., ...). Figure 2 (As shown); This embodiment presents a three-grid system construction method for preprocessing variable-density meshes in large shell components, including the following steps: S1: Construction of structured mesh.
[0031] S11: Construction of SM. SM is a high-resolution structured hexahedral mesh used to store historical simulation data. This mesh can be constructed using methods such as... Figure 3 The paving method shown generates hexahedral elements by sequentially paving them circumferentially, radially, and axially. If the inner diameter is... mm, thickness is mm and length is The mm cylindrical blank is divided into sections along the circumferential, radial, and axial directions during grid generation. , and If there are 1 unit, then the mesh has circumferential, radial and axial directions. , and There are several nodes, so elements and nodes can be indexed using numbers in three directions. Each node is indexed according to its number in the circumferential, radial, and axial directions. , can be done according to the formula Calculate its overall number The coordinates of this point This can be calculated using equation (2). Similarly, each unit is determined based on its numbering in the three directions. , can be done according to the formula Calculate its overall number This unit consists of 8 connected nodes as shown in Table 1. When meshing, and When selecting specific process-related parameters, it is generally necessary to ensure that the dimensions of the divided units accurately describe the contact state between the workpiece and the mold. The value must be no less than 3 to ensure the stability of the simulation. Furthermore, [the value is determined by...]. , and The generated mesh should maintain the regularity of the elements as much as possible.
[0032] (1) (2) (3) Table 1 shows the node numbers of the cells in SM.
[0033] S12: Construction of BM. The cells in SM are arranged according to spacing. BM can be obtained by condensation along the circumferential and radial directions (see Figure 4 Therefore, the mesh division in BM along the circumferential, radial, and axial directions is as follows: , and The relationship between the mesh fraction in SM and the mesh number is shown in equation (4). The value of depends on the encryption template used. If the selected template is a multi-level four-part encryption template, then... If the selected template is a multi-level nine-point encryption template, then Here Indicates the encryption level, which needs to be controlled. Values to ensure No more than or of .
[0034] (4) S2: Marking of the refinement zone in the BM. The refinement attributes of elements in the BM can be marked based on the distribution of field variables, the element size field calculated from the posterior error estimate, the tracking of the die position, and the pre-determined fixed refinement zone range. Since the BM of a body of revolution is a structured mesh, to reduce the search complexity, only the single-layer mesh on the cylindrical surface can be marked before extending the element attributes to all elements in the thickness direction. The nodes of the refinement elements in the BM will also be marked as refinement nodes.
[0035] S3: Construction of CM. Because the meshes in a three-mesh system overlap and share nodes, the enriched and unenriched regions can be directly constructed using the elements and nodes from SM and BM. However, without proper handling at the boundary between the two, issues such as... Figure 1The diagram shows a large number of suspended nodes that affect simulation accuracy. To address this, the present invention designs two sets of multi-level densification templates, which can achieve a coordinated mesh transition without generating additional nodes, and can ensure that the elements within the template maintain high quality.
[0036] S31: Template Construction. Based on the number and distribution of densified nodes in the vertices of a quadrilateral base of a BM hexahedral element, templates can be divided into six categories: undensified templates (no densified nodes, using only BM elements), convex-corner densified templates (only 1 densified node), edge densified templates (2 densified nodes on the same edge), diagonal densified templates (2 densified nodes on the diagonal), concave-corner densified templates (3 densified nodes), and fully densified templates (all densified nodes, using only SM elements). Since thin-walled revolution meshes do not require density control in the thickness direction, a usable hexahedral transition template can be formed simply by stretching the multi-level quadrilateral transition template constructed above along the thickness direction. Furthermore, the constructed template can flexibly adapt to the arbitrary shape of the densified zone caused by the complex actual processes of thin-walled revolution meshes.
[0037] S311: Multi-level nine-point encryption template (see...) Figure 5 This template achieves high-density gradient transitions through nesting of single-level nine-part encryption templates and their variants. However, the side encryption template suffers from insufficient side width due to the 2:3 aspect ratio of the single-level nine-part template. However, an additional layer of SM fine-tuning units is needed. Multi-level convex and concave angle encryption templates can be obtained by dividing the multi-level edge encryption template along the secondary diagonal and mirroring the upper left and lower right triangles respectively. The diagonal encryption template is obtained by dividing the multi-level convex angle encryption template along the main diagonal and then mirroring the lower left triangle. Compared with the traditional cross-tree method of repeatedly applying nine-part encryption templates for transition, the template designed in this invention eliminates the need for strip-shaped fine-tuning units inserted to ensure template squareness, resulting in a more compact transition.
[0038] S312: Multi-level four-part encryption template (see...) Figure 6 The lack of self-symmetry in single-level quarter templates necessitates their use in pairs, thus necessitating the addition of directly nested single-level quarter templates. The difficulty of multi-level transitions. Therefore, this invention addresses this by combining... Constructed by BM units The transition restores the template's self-symmetry, allowing for the construction of multi-level encryption templates using the same method as multi-level nine-part templates. Similarly, since the aspect ratio of paired single-level four-part templates is 1:2, this template requires two additional layers of SM fine units. The multi-level four-part template designed in this invention can quickly achieve the same encryption effect as repeatedly applying single-level four-part encryption templates.
[0039] S32: Template Usage. The location of the encryption attribute markers for each cell node in BM may be related to... Figure 5 and Figure 6 The results differ from those shown in the original image, requiring the template to be rotated appropriately before use. This is because the nodes in the template are composed of... The grid, In the nine-point template In the four-part template Therefore, the node has a local number. , .by Figure 5 and Figure 6 Using the template in the table as a reference, the local numbering of each node within the template after clockwise rotation can be obtained from Table 2. If the current template number is... Then the global number of the node in the template is .
[0040] Table 2 shows the local node numbers within the unit after clockwise rotation of the template.
[0041] S33: CM Mesh Generation. Traverse the template in the BM to extract the corresponding SM and BM elements, or generate the corresponding transition elements based on the template. Reorder the elements and their contained nodes to obtain the desired locally refined CM (see...). Figure 7 ).
[0042] To facilitate demonstration of the effectiveness of the method proposed in this invention, this invention relies on... Figure 2 The process design included a multigrid method calculation script on ABAQUS.
[0043] An embodiment of this invention is a rapid calculation model for single-spinning flow spinning of 2219 aluminum alloy cylindrical parts using ABAQUS 6.16 under the Windows 10 operating system. The geometric and process parameters of the model are listed in Table 3. This model uses a two-level densified nine-part template, with annular densified zones marked according to the spindle position during deformation. The density change of the mesh during the forming process is shown in Table 3. Figure 8 As shown, the Mises stress and equivalent plastic strain distribution (PEEQ) demonstrates that the locally refined mesh constructed using the method proposed in this invention can obtain an accurate distribution of field variables. The spin-formed ripples on the SM surface indicate that this method can preserve precise geometric details. Figure 9 (a) records that this model reduces computation time by 53% compared to the model that directly uses a fully encrypted mesh. Furthermore, replacing the transition template of this model with a two-level four-part encrypted template also yields similar results. Figure 9 The 47% time reduction is shown in (b).
[0044] Table 3 shows the geometric and process parameters of the single-spindle model.
[0045] Although embodiments of the present invention have been shown and described above, it is understood that the above embodiments are exemplary and should not be construed as limiting the present invention. Those skilled in the art can make changes, modifications, substitutions and variations to the above embodiments within the scope of the present invention without departing from the principles and spirit of the present invention.
Claims
1. A method for constructing a three-grid system for preprocessing variable-density meshes in solid shell components of large shell structures, characterized in that: Includes the following steps: Based on the geometric parameters of the target thin-walled rotating body component, a high-resolution structured hexahedral storage grid SM is generated for storing historical simulation data. By aggregating the cells in the storage grid SM at intervals, a background grid BM is generated that shares nodes with the storage grid SM and has the same topology. The background grid BM serves as a topological framework for quickly marking encrypted areas. Based on the changes in the plastic deformation zone during the simulation, the encryption attributes of the elements to be encrypted are dynamically marked in the background mesh BM; Based on the encryption attributes marked in the background grid BM, a preset multi-level encryption template is invoked. Without generating additional nodes, the nodes and cells of the storage grid SM and the background grid BM are combined to quickly generate a local encryption computing grid CM that shares nodes with the storage grid SM and the background grid BM. The storage grid SM, background grid BM, and computational grid CM together form a three-grid system that overlaps with each other and shares nodes. Through the bridging role of the background grid BM, a fast and coordinated conversion from data storage to computational grid generation is achieved to adapt to the dynamic changes of the deformation zone in incremental forming simulation and ensure simulation accuracy.
2. The three-grid system construction method for preprocessing large shell components with variable density meshes according to claim 1, characterized in that: The structured hexahedral storage grid SM is implemented using a tiling method, specifically: Based on the inner diameter, thickness and axial length of the thin-walled rotating body component, hexahedral units are sequentially laid along the circumferential, radial and axial directions to form a structured hexahedral storage grid SM; Each node and element is uniquely indexed and located by three numbers in the circumferential, radial, and axial directions.
3. The three-grid system construction method for preprocessing large shell component solid shell variable density grids according to claim 2, characterized in that: The paving method specifically includes: The storage grid SM is divided into sections in the circumferential, radial, and axial directions, respectively. , and Each unit, thus having in the circumferential, radial and axial directions respectively. , and One node; Each node is identified by its three numbers in the circumferential, radial, and axial directions. Index it, its overall numbering Calculated using the following formula: Coordinates of each node Calculated using the following formula: Where D is the inner diameter of the thin-walled rotating component, T is the thickness of the thin-walled rotating component, and L is the axial length of the thin-walled rotating component; each unit is identified by its three numbers in the circumferential, radial, and axial directions. Index it, its overall numbering Calculated using the following formula: Furthermore, each unit consists of 8 connected nodes, and the global numbering of the 8 nodes is based on the unit numbering. Sure.
4. The three-grid system construction method for preprocessing the solid shell of a large shell component with variable density mesh according to claim 1, characterized in that: The generation of the background mesh BM specifically includes: Set a cohesion interval that is related to the selected encryption template type. α ; The storage grid SM is spaced at intervals in both the circumferential and axial directions. α -1) Merge the elements, retaining all radial elements, to generate a storage grid SM with element sizes in both the circumferential and axial directions. α Background mesh BM (multiplied by 100%); The background grid BM shares all nodes with the storage grid SM, and the topology of the background grid BM is directly inherited from the storage grid SM.
5. The three-grid system construction method for preprocessing large shell component solid shell variable density grids according to claim 4, characterized in that: The condensation interval α The value is related to the encryption template type: When using a multi-level four-part encryption template k represents the encryption level; When using a multi-level nine-point encryption template k represents the encryption level.
6. The three-grid system construction method for preprocessing the solid shell of a large shell component with variable density mesh according to claim 4, characterized in that: The basis for dynamically marking the encryption attributes of the cells to be encrypted in the background grid BM includes at least one of the following: Distribution of field variables during simulation; The element size field is calculated from the posterior error estimate; Real-time tracking of the tool's spatial position; A pre-defined fixed encryption zone range.
7. The three-grid system construction method for preprocessing large shell component solid shell variable density grids according to claim 1, characterized in that: The specific steps of invoking the preset multi-level encryption template include: Provides multi-level nine-point encryption templates and / or multi-level four-point encryption templates; The multi-level nine-part encryption template achieves a compact transition from the encrypted area to the unencrypted area by nesting the single-level nine-part template and its variants, and by using mirroring operations. The multi-level four-part encryption template restores self-symmetry by merging 2x2 background grid BM units, thereby constructing an encryption template capable of multi-level transition.
8. The three-grid system construction method for preprocessing large shell component solid shell variable density grids according to claim 7, characterized in that: The encryption template is configured into multiple types based on the number and distribution of encryption nodes within a quadrilateral face of the background mesh BM hexahedral element, including at least: no encryption template, convex corner encryption template, edge encryption template, diagonal encryption template, concave corner encryption template, and fully encrypted template; For thin-walled rotating body components, the quadrilateral transition template is stretched along the thickness direction to form a corresponding hexahedral transition template.
9. The three-grid system construction method for preprocessing large shell component solid shell variable density grids according to claim 7, characterized in that: The step of calling the preset multi-level encryption template also includes template adaptation: When the node encryption attribute marker position of the cell to be encrypted in the background mesh BM is inconsistent with the preset orientation of the standard encryption template, the standard encryption template is rotated to match the actual encryption node distribution of the cell to be encrypted; wherein, the local numbering of the nodes within the rotated template... The rotation angle is determined based on a predefined mapping relationship. The global number of any node in the template is calculated according to the following formula: in, Here is the current template number, and n is the template size. In a multi-level nine-point encryption template... In the multi-level four-part encryption template k represents the encryption level.
10. A three-mesh system construction system for preprocessing variable-density meshes of solid shell components for large shell structures, used to execute the three-mesh system construction method for preprocessing variable-density meshes of solid shell components for large shell structures as described in any one of claims 1-9; characterized in that, include: The storage grid generation module is used to generate a high-resolution structured hexahedral storage grid for storing historical simulation data based on the geometric parameters of the target thin-walled rotating body component. The background grid generation module is used to generate a background grid that shares nodes with the storage grid and has the same topology by aggregating the cells in the storage grid at intervals. The background grid serves as a topological framework for marking encrypted areas. The encryption attribute marking module is used to dynamically mark the encryption attributes of the unit to be encrypted in the background mesh according to the changes in the plastic deformation zone during the simulation process. The computational grid generation module is used to call a preset multi-level encryption template based on the encryption attributes marked in the background grid, and combine the nodes and cells of the storage grid and the background grid to generate a local encrypted computational grid that shares nodes with the storage grid and the background grid without generating additional nodes. The storage grid, background grid, and computing grid together form a three-grid system that overlaps with each other and shares nodes.