Growth type anisotropic adaptive grid generation method and device based on orthogonal field

Abstract

The invention relates to a growth type anisotropic adaptive grid generation method and device based on an orthogonal field. The method comprises the following steps: taking an obtained initial surface grid of a boundary layer of the quasi-two-dimensional complex geometric shape of the aircraft as an initial array surface, and generating a layer grid unit by adopting an improved leading edge layer approach method based on an orthogonal field; adjusting the grid units of the layer by adopting an anisotropic self-adaptive strategy; and continuously generating a next layer of grid units, and carrying out anisotropic adaptive adjustment until the number of the grid layers reaches a preset value position to obtain a final surface grid of the boundary layer of the quasi-two-dimensional complex geometric shape of the aircraft. The method solves the problems that the quality of grids generated by complex shapes such as convex and concave ridge lines and angular points of typical complex configurations on a quasi-two-dimensional aircraft is poor, and the grids are easy to intersect, improves the automation degree of grid generation, improves the robustness and adaptability to the complex shapes, and improves the accuracy of grid generation. And the growth type anisotropic adaptive grid with good orthogonality, good continuity and reasonable distribution is generated.

Description

Technical Field

[0001] The present application relates to the technical field of computational fluid dynamics, and in particular to a method and device for generating a growth-type anisotropic adaptive grid based on an orthogonal field. Background Art

[0002] Mesh generation is a pre-processing step in computational fluid dynamics (CFD), accounting for about 60% of the entire CFD calculation cycle, and plays an important role in CFD. The geometric configurations of aircraft studied by CFD in the aerospace field are becoming increasingly complex, and mesh generation technology has gradually become one of the key factors restricting CFD. On the one hand, in numerical calculations, in addition to the use of high-precision numerical algorithms, the quality of mesh generation is also crucial to the calculation accuracy, efficiency and even calculation stability. High-precision and high-resolution discrete formats also place higher requirements on mesh quality. On the other hand, in the face of complex configurations, semi-automatic or manual mesh generation is often required, and the quality of the generated mesh depends on the experience and knowledge of the mesh engineer, which consumes a lot of manpower and material resources.

[0003] The basic requirements for mesh generation are orthogonality, continuity, isotropy and reasonable distribution. Orthogonality is mainly for quadrilateral and hexahedral meshes, which means that the mesh edges or mesh faces satisfy the orthogonal relationship; continuity refers to the gradual transition of meshes in the research domain, reducing the errors caused by spatial drastic changes in interpolation or integration in numerical calculations; isotropy means that the shape of the mesh unit is as close to a regular polygon as possible, mainly for mesh units far away from the wall. Mesh units near the wall often use anisotropic units to reduce the computational cost and capture the details in the boundary layer as much as possible; reasonable distribution means generating dense meshes in areas where the physical field changes drastically in the flow field, and generating sparse meshes in areas where the changes are gentle.

[0004] The computational grid consists of surface grid, boundary layer grid and space grid. The boundary layer grid is the key and difficulty of grid generation technology. The existing boundary layer grid generation method has poor robustness, and the algorithm is prone to failure in complex geometric shapes, making it difficult to generate high-quality units.

[0005] At present, almost all boundary layer mesh generation methods are advanced layer technology (ALT) and its variants. When advanced layer technology generates meshes on concave and convex feature geometric surfaces, it is easy for mesh units to self-intersect and intersect; it is easy to generate low-quality units in narrow and long areas; and it may generate mesh units with negative volumes in complex corner areas. In the face of the above special geometric areas, special processing is often required. Common methods include multi-normal technology, normal fusion and compression units to improve the intersection problem. However, the above processing methods may affect the layered structure of the boundary layer mesh, and at the same time cause the orthogonality of local units to decrease, reducing the unit quality.

[0006] In the existing method, a three-dimensional tensor-guided anisotropic volume harmonic field is constructed, and the boundary layer mesh is generated layer by layer based on the field, and the quality of the boundary layer mesh is further improved by using the Laplace smoothing optimization method. This method adjusts the harmonic field by adding a local tensor, and the mesh generation is more flexible and the mesh intersection problem is solved. However, the field needs to create a background mesh, and the background mesh has low controllability and flexibility. The mesh near complex geometric shapes inevitably has the problem of excessive deformation, and the volume harmonic field calculation is complex and has the problem of low computational efficiency.

[0007] The existing methods have poor mesh quality and mesh intersections easily for complex shapes such as convex and concave ridges and corner points of typical complex configurations on quasi-two-dimensional aircraft. Summary of the invention

[0008] Based on this, it is necessary to provide a method and device for generating a growing anisotropic adaptive grid based on an orthogonal field to address the above technical problems.

[0009] A growing anisotropic adaptive grid generation method based on orthogonal fields, the method comprising: Obtain the initial surface mesh of the boundary layer of the local complex quasi-two-dimensional geometric shape of the aircraft.

[0010] The initial surface grid is used as the initial front face, and the improved frontal advancement method based on orthogonal field is used to generate the local grid unit; the improved frontal advancement method based on orthogonal field means: applying the orthogonal field model to the frontal advancement method, taking the direction of electric field intensity as the front face advancement direction, and controlling the advancement distance by electric potential.

[0011] An anisotropic adaptive strategy is used to adjust the grid units of this layer to obtain the adjusted grid units of this layer; the anisotropic adaptive strategy is used to determine the number of normal splitting of each grid unit after the grid units of this layer are generated, and perform normal splitting on the units of this layer while meeting the level difference requirements; after all units are normal split, perform array uniformity detection from the top unit of this layer, perform circumferential twin block discrimination and circumferential merging operations on densely distributed grids, and perform circumferential splitting operations on sparsely distributed grids; repeat this process until the array distribution is uniform and the distribution adjustment of the grid units of this layer is completed.

[0012] Continue to generate the next layer of grid units and perform anisotropic adaptive adjustments until the number of grid layers reaches the preset value, completing the grid generation of the complex quasi-two-dimensional geometric shape of the local aircraft.

[0013] A growing anisotropic adaptive grid generation device based on orthogonal fields, the device comprising: The initial surface mesh acquisition module is used to obtain the initial surface mesh of the local complex quasi-two-dimensional geometric shape of the aircraft.

[0014] The current layer grid unit generation module is used to use the initial surface grid or the upper array surface of the generated grid as the initial array surface, and use the improved frontier layer-by-layer method based on orthogonal fields to generate the current layer grid units, continue to generate the next layer of grid units, and perform anisotropic adaptive adjustment until the number of grid layers meets the requirements, and finally obtain the boundary layer grid and spatial external field grid of the local complex quasi-two-dimensional geometric shape of the aircraft; the improved frontier layer-by-layer method based on orthogonal fields means: applying the orthogonal field model to the frontier layer-by-layer method, using the direction of the electric field intensity as the array propulsion direction, determining the propulsion distance by the electric potential, and connecting to form the current layer grid according to the geometric topological relationship.

[0015] The growing anisotropic adaptive mesh adjustment module is used to adjust the mesh units of this layer by adopting an anisotropic adaptive strategy to obtain the adjusted mesh units of this layer; the anisotropic adaptive strategy is used to determine the number of normal splitting of each mesh unit after the mesh units of this layer are generated, and perform normal splitting on the units of this layer while meeting the level difference requirements; after all the units are normally split, perform array uniformity detection from the top unit of this layer, perform circumferential twin block discrimination and circumferential merging operations on densely distributed meshes, and perform circumferential splitting operations on sparsely distributed meshes; repeat this process until the array distribution is uniform and the distribution adjustment of the mesh units of this layer is completed; continue to generate the next layer of mesh units and perform anisotropic adaptive adjustment until the number of mesh layers reaches the preset value, and complete the generation of the complex quasi-two-dimensional geometric mesh of the local aircraft.

[0016] The above-mentioned method and device for generating a growing anisotropic adaptive grid based on an orthogonal field include: using the initial surface grid of the boundary layer of the aircraft's quasi-two-dimensional complex geometric shape as the initial array, and using the improved frontal layer method based on the orthogonal field to generate the grid units of this layer; using an anisotropic adaptive strategy to adjust the grid units of this layer to obtain the adjusted grid units of this layer; continuing to generate the next layer of grid units, and performing anisotropic adaptive adjustment until the number of grid layers reaches the preset value position, and obtaining the final surface grid of the boundary layer of the aircraft's quasi-two-dimensional complex geometric shape. This method solves the problem of poor grid quality and easy intersection of grids generated by complex shapes such as convex and concave ridges and corners of typical complex configurations on quasi-two-dimensional aircraft, improves the degree of grid generation automation, robustness and adaptability to complex shapes, and generates growing anisotropic adaptive grids with good orthogonality, good continuity and reasonable distribution. BRIEF DESCRIPTION OF THE DRAWINGS

[0017] Figure 1 It is a schematic diagram of a flow chart of a growth-type anisotropic adaptive grid generation method based on an orthogonal field in one embodiment; Figure 2 is a flow chart of a process of generating a growth-type anisotropic adaptive grid based on an orthogonal field in another embodiment; Figure 3 A schematic diagram of an improved frontal layering method based on an orthogonal field in another embodiment; Figure 4 This is a schematic diagram of anisotropic merging in another embodiment. Figure 4 (a) is circumferential merging, Figure 4 (b) normal merging; Figure 5 It is a schematic diagram of levels in another embodiment, Figure 5 (a) is a schematic diagram of the circumferential level. Figure 5 (b) is a schematic diagram of the normal level; Figure 6 A schematic diagram of a suspended node in another embodiment; Figure 7 This is a schematic diagram of grid unit scale calculation in another embodiment. Figure 7 (a) is a schematic diagram of circumferential scale calculation. Figure 7 (b) is a schematic diagram of normal scale calculation; Figure 8 A schematic diagram of an orthogonal field in another embodiment; Fig. 9 It is a schematic diagram of meshing the "F" shape in another embodiment. Fig. 9 (a) is a local enlarged schematic diagram. Fig. 9 (b) is the overall schematic diagram of the grid; Fig.10 It is a structural block diagram of a growth-type anisotropic adaptive grid generation device based on orthogonal fields in one embodiment. DETAILED DESCRIPTION

[0018] In order to make the purpose, technical solution and advantages of the present application more clearly understood, the present application is further described in detail below in conjunction with the accompanying drawings and embodiments. It should be understood that the specific embodiments described herein are only used to explain the present application and are not used to limit the present application.

[0019] In one embodiment, Figure 1 , Figure 2 As shown, a growth-type anisotropic adaptive grid generation method based on orthogonal fields is provided, and the method comprises the following steps: Step 100: Obtain an initial surface mesh of the boundary layer of the local complex quasi-two-dimensional geometric shape of the aircraft.

[0020] Specifically, the initial surface mesh is a surface mesh of a local complex quasi-two-dimensional geometric shape on the aircraft generated by computer mesh generation software.

[0021] Step 102: Using the initial surface grid as the initial front face, and using the improved frontal layering method based on orthogonal fields to generate the grid units of this layer; the improved frontal layering method based on orthogonal fields means: applying the orthogonal field model to the frontal layering method, using the direction of the electric field intensity as the front face advancement direction, and using the electric potential to control the advancement distance.

[0022] Specifically, an orthogonal field refers to a field in which field lines and contour lines satisfy an orthogonal relationship. Field lines and contour lines are used as grid line clusters to meet the orthogonality requirements of grid generation. The electrostatic field is a conservative field, which satisfies the negative gradient of the electric potential and the direction of the field strength is the direction in which the electric potential decreases fastest. In the passive region, the electrostatic field is a harmonic field with the good characteristic that the field lines and equipotential lines are mutually orthogonal. It is proposed to use the electrostatic field as an orthogonal field, and the field lines and equipotential lines will not self-intersect. The grid generated based on the orthogonal field has the characteristics of good orthogonality and no self-intersection.

[0023] This application models the orthogonal field model, improves the traditional frontier layering method, combines the orthogonal field with the frontier layering method, uses the electric field intensity as the growth normal, and the field lines and equipotential lines as the grid edges to generate grids layer by layer. It can generate anisotropic boundary layer grids and isotropic spatial external field grids, thereby improving the efficiency and automation of grid generation.

[0024] In order to solve the problems of low quality, poor robustness and low automation of boundary layer mesh generation for quasi-two-dimensional complex geometric shapes of aircraft, this application builds an orthogonal field model, improves the traditional frontier layering method, combines the orthogonal field and the frontier layering method, uses the electric field intensity as the growth normal, and the field lines and equipotential lines as the mesh edges to generate meshes layer by layer, which can generate anisotropic boundary layer meshes and isotropic spatial external field meshes, improving mesh generation efficiency and automation. An improved frontier layering method based on orthogonal fields is proposed, and a set of automatic, efficient, robust and easy-to-use growth-type anisotropic adaptive mesh automatic generation technology is developed. In the process of layer-by-layer generation, the mesh units are anisotropically merged and split at the same time to ensure the orthogonality and continuity of the generated mesh, solve the problem of mesh intersection during the generation process, and realize the integrated generation of boundary layer meshes and spatial external field meshes.

[0025] Step 104: Use an anisotropic adaptive strategy to adjust the grid units of this layer to obtain adjusted grid units of this layer; the anisotropic adaptive strategy is used to determine the number of normal splitting of each grid unit after the grid units of this layer are generated, perform normal splitting on the units of this layer, and meet the level difference requirements at the same time; after all units are normally split, perform array uniformity detection from the top unit of this layer, perform circumferential twin block discrimination and circumferential merging operations on densely distributed grids, and perform circumferential splitting operations on sparsely distributed grids; repeat this process until the array distribution is uniform and the distribution adjustment of the grid units of this layer is completed.

[0026] Specifically, adaptive mesh (Adaptive Refinement Mesh, ARM) technology is divided into computational adaptation and geometric adaptation. Computational adaptation refers to the autonomous mesh optimization technology based on flow characteristics. It requires real-time acquisition of physical field gradient information, dynamic mesh adjustment, and multiple iterations to achieve consistent convergence between the mesh and the flow field. It is the main adaptive technology. Geometric adaptation refers to the technology that guides mesh adjustment based on geometric features. During the mesh generation process, the mesh distribution is adjusted according to the geometric information, and multiple adjustments are made to meet the basic requirements of mesh generation.

[0027] There are two main adaptive methods: r-refinement and h-refinement. The r-refinement technique refines a specific area by moving grid nodes, and the number of overall grid nodes and grid cells before and after adaptation remains unchanged. The h-refinement technique directly refines a specific area locally, and adapts to changes in the number of overall grid nodes and grid cells before and after.

[0028] Due to the concave and convex features of geometric surfaces, the field lines of the orthogonal field inevitably have the problem of field line concentration and dispersion. The present invention provides a geometric anisotropic adaptive technology, which uses H-type adaptive technology to adjust the above-mentioned grid based on the orthogonal field, anisotropically merge and split the grid, improve the continuity and distribution rationality of the grid, and improve the grid quality.

[0029] After generating the grid cells of this layer, the grid node distribution on the array surface of this layer is tested for uniformity. The uniformity of the array surface is reflected by the top edge length of the grid cells. Identify the dense and sparse distribution areas of the grid nodes, call the circumferential merging and circumferential splitting modules to adjust the grid distribution, and improve the continuity and rationality of the grid distribution.

[0030] The orthogonal field model does not need to generate background grids. It combines the frontier layer-by-layer method to directly generate boundary layer and spatial external field grids layer by layer. There is also a geometric anisotropy adaptive strategy to further improve the quality of the generated grid, and improve the efficiency, robustness and automation of grid generation.

[0031] Step 106: Continue to generate the next layer of grid units and perform anisotropic adaptive adjustment until the number of grid layers reaches the preset value position, completing the grid generation of the local complex quasi-two-dimensional geometric shape of the aircraft.

[0032] Specifically, the overall idea of ​​this method is to automatically generate boundary layer grids and spatial external field grids given a surface grid.

[0033] In the above-mentioned growing anisotropic adaptive grid generation method based on orthogonal field, the method includes: using the initial surface grid of the boundary layer of the quasi-two-dimensional complex geometric shape of the aircraft as the initial array, and using the improved front layer method based on orthogonal field to generate the grid units of this layer; using the anisotropic adaptive strategy to adjust the grid units of this layer to obtain the adjusted grid units of this layer; continuing to generate the next layer of grid units, and performing anisotropic adaptive adjustment until the number of grid layers reaches the preset value position, and obtaining the final surface grid of the boundary layer of the quasi-two-dimensional complex geometric shape of the aircraft. This method solves the problem of poor quality of grid generation and easy intersection of grids for complex shapes such as convex and concave ridges and corners of typical complex configurations on quasi-two-dimensional aircraft, improves the degree of grid generation automation, robustness and adaptability to complex shapes, and generates growing anisotropic adaptive grids with good orthogonality, good continuity and reasonable distribution.

[0034] In one embodiment, the specific process of the improved frontal advancement method based on the orthogonal field in step 102 includes: considering the initial surface grid as the initial front, and calculating the front advancement distance as: (1) in, z is the number of layers, is the growth rate, h For the z Layer front advancement distance, h 0 is the initial front advancement distance.

[0035] The direction of the electric field strength is calculated at the grid nodes on the initial array, and the direction of the electric field strength is used as the array advancement direction. The array advancement distance is advanced according to the field lines, and the equipotential surface is generated with the lowest electric potential as the next array. The grid generation of this layer is completed according to the geometric topological relationship; and the same process is repeated until the specified number of grid layers is generated.

[0036] Specifically, the frontal layering method is improved. In the traditional frontal layering method, the frontal advancement direction is generally determined based on local geometric information, and the frontal advancement distance is specified. This paper applies the orthogonal field model to the frontal layering method, using the direction of the electric field intensity as the frontal advancement direction and the electric potential to determine the advancement distance.

[0037] The specific process of the improved frontier progression method based on orthogonal field is as follows: The initial surface mesh is regarded as the initial front face, and the front face advancement distance is calculated according to formula (1): h ; Calculate the direction of the electric field intensity at the grid nodes on the array surface, and advance the array surface by the distance calculated above according to the field lines h , the equipotential surface generated by the lowest potential is used as the next front, and the grid generation of this layer is completed according to the geometric topological relationship. The subsequent process is deduced in this way until the specified number of grid layers is generated. The schematic diagram of the improved frontal layering method based on orthogonal field is shown in Figure 3 shown.

[0038] In one of the embodiments, the specific process of the anisotropic adaptive strategy in step 104 includes: judging the normal splitting conditions of all grid cells in this layer, and performing normal splitting on the cells that meet the conditions; after normal splitting, performing a uniformity test on the array surface of this layer; starting from the top grid cell of this layer, twin block judgment is performed, and circumferential merging is performed on the twin blocks that meet the circumferential merging conditions; after circumferential merging, starting from the top grid cell of this layer, circumferential splitting operations are performed on the grid cells that meet the circumferential splitting conditions, until all grid cells in this layer meet the expected grid scale requirements of this layer, and the grid generation and improvement of this layer are completed, wherein both the merging and splitting operations meet the level requirements.

[0039] Specifically, due to the uneven distribution of field lines caused by the complex geometric surface shape, the field lines tend to be concentrated in the concave feature area of ​​the wall and tend to be dispersed in the convex feature area, which in turn causes the generated grid units to be dense in the concave feature area and sparse in the convex feature area. It is necessary to formulate corresponding geometric anisotropy adaptive strategies to further adjust the distribution of grid units and lay a good foundation for the generation of the next layer of grid.

[0040] In one embodiment, a normal splitting condition is judged for all grid cells in the current layer, and normal splitting is performed on the cells that meet the condition, including: after the grid cells in the current layer are generated, the normal splitting times of each grid cell in the current layer are calculated by a normal splitting times calculation formula; wherein the normal splitting times calculation formula is: ; in, SpNum is the number of normal splits, Indicates rounding down. h i for i The element normal scale, h under for i The normal scale of the neighboring cells under the cell; The normal classification conditions of all grid cells in this layer are judged, and the cells that meet the conditions are normally split. After the normal splitting, the difference between the normal level of this unit and the normal level of the adjacent left and right grid cells does not exceed 1; the grid nodes of the calculated grid unit scale are used as new splitting points through the electric potential.

[0041] Specifically, the anisotropic splitting operation is for one mesh unit, and is divided into circumferential splitting and normal splitting according to the splitting direction. The two have different priorities, with normal splitting being the main one near the wall and circumferential splitting being the main one far from the wall.

[0042] After splitting, the requirement that the level difference between this unit and the adjacent grid units should not exceed 1 should be met. Specifically, after circumferential splitting, the circumferential level difference between this unit and the adjacent upper and lower grid units should not exceed 1; after normal splitting, the normal level difference between this unit and the adjacent left and right grid units should not exceed 1.

[0043] The splitting condition is used to determine whether the grid unit is to be split. The splitting condition is also divided into normal splitting condition and circumferential splitting condition. The splitting condition design should consider the scale of the grid unit itself and the adjacent grid units. The scale of the unit after splitting should reach the expected grid unit scale of this layer as much as possible, while meeting the continuity requirements of grid generation.

[0044] After the unit splits, new grid nodes are generated. The method of obtaining the newly added grid nodes is similar to the calculation unit scale. The grid nodes of the calculation unit scale are used as new splitting points through electric potential.

[0045] In one of the embodiments, twin block identification is performed starting from the top grid unit of the current layer, and the twin blocks that meet the circumferential merging conditions are circumferentially merged, including: starting from the top grid unit of the current layer, a pair of adjacent grid units that meet the level requirements are regarded as a twin block; the level requirement of the twin block means that within the twin block, the normal levels of the two grid units of the circumferential twin block should be the same, and the circumferential levels of the two grid units of the normal twin block should be the same, and the level requirement can ensure that adjacent grid cells contain a common grid edge; outside the twin block, the circumferential level of the circumferential twin block is not less than the circumferential level of the upper and lower adjacent grid units, and the normal level of the normal twin block is not less than the normal level of the left and right adjacent grid units; the twin blocks that meet the circumferential merging conditions are circumferentially merged.

[0046] Specifically, a twin block refers to a pair of adjacent grid units with level requirements, which are also divided into circumferential twin blocks and normal twin blocks. Only twin blocks can perform anisotropic merge operations. The level requirements of twin blocks refer to that within a twin block, the normal levels of the two grid units of the circumferential twin block should be the same, and the circumferential levels of the two grid units of the normal twin block should be the same. The level requirements can ensure that adjacent grid units contain a common grid edge; outside a twin block, the circumferential level of a circumferential twin block is not less than the circumferential level of the upper and lower adjacent grid units, and the normal level of a normal twin block is not less than the normal level of the left and right adjacent grid units.

[0047] The anisotropic merging operation is for the twin block unit. The twin block contains two mesh units. According to the merging direction, it is divided into circumferential merging and normal merging. The merging conditions are also divided into circumferential merging conditions and normal merging conditions. The merging condition design should consider the scale of the twin block and the adjacent mesh units. The unit scale after merging should reach the expected mesh unit scale of this layer as much as possible, and at the same time meet the mesh generation continuity requirements.

[0048] The merge and split operations also need to consider the order of merge and split, and the merge and split conditions have unequal characteristics to avoid falling into an infinite loop of merge and split.

[0049] The schematic diagram of anisotropic merging is as follows Figure 4 As shown, Figure 4 (a) is circumferential merging, Figure 4 (b) is normal merging.

[0050] In one of the embodiments, the level is used to regulate anisotropic splitting and merging operations; in the corresponding direction, the grid unit is anisotropically split once, and the level in this direction increases by 1; the twin blocks are merged once, and the level in this direction decreases by 1; the level difference of adjacent units is limited to 1 to ensure that the number of dangling nodes after anisotropic splitting or merging does not exceed 1.

[0051] Specifically, the level diagram is as follows Figure 5 As shown, Figure 5 (a) is a schematic diagram of the circumferential level. Figure 5 (b) is a schematic diagram of the normal level; each mesh unit has levels in the circumferential direction (parallel to the wall direction) and the normal direction (perpendicular to the wall direction), which are used to standardize anisotropic splitting and merging operations. In the corresponding direction, the mesh unit is anisotropically split once, and the level in that direction increases by 1; the twin blocks are merged once, and the level in that direction decreases by 1. The level difference between adjacent units is limited to 1, that is, the absolute value of the difference in normal level between unit i and the left and right adjacent mesh units is less than or equal to 1, and the absolute value of the difference in circumferential level between unit i and the upper and lower adjacent mesh units is less than or equal to 1, to ensure that the number of dangling nodes after anisotropic splitting or merging does not exceed 1. The schematic diagram of dangling nodes is shown in Figure 6 As shown in formula (2): express i The circumferential level of the unit, express i The circumferential level of the upper adjacent unit of the unit, express i The circumferential level of the unit's lower neighbor, express i The normal level of the element, express i The normal level of the cell's left neighbor, express i The normal level of the cell's right neighbor; (2) In one embodiment, the grid unit scale includes a circumferential scale and a normal scale; the process of calculating the grid unit scale includes: recording the lower left grid node in the grid unit as point O, the upper left grid node as point P, the lower right grid node as point N, and the upper left grid node as point M; calculating the grid unit scale based on the orthogonal field, i In the circumferential scale: in grid units i The average potential of nodes O and P is calculated along the field line to calculate the left point, and the average potential of points N and M is calculated along the field line to calculate the right point; in the calculation grid unit i Normal scale: in grid units i The corresponding midpoint of the surface charged line segment is extrapolated along the field line to the equipotential surface where points O and N are located as the lower point for normal scale calculation, and then extrapolated from the newly added split lower point along the field line to the equipotential surface where points P and M are located as the upper point for normal scale.

[0052] Specifically, the schematic diagram of grid unit scale calculation is as follows: Figure 7 As shown, Figure 7 (a) is a schematic diagram of circumferential scale calculation. Figure 7 (b) is a schematic diagram of normal scale calculation. The grid unit scale refers to the size of a grid unit, which is also divided into circumferential scale and normal scale. The grid unit scale is calculated based on the orthogonal field. When calculating the circumferential scale of grid unit i, the average potential of nodes O and P of this unit is used to deduce the left point of the circumferential scale calculation along the field line, and the average potential of points N and M is used to deduce the right point of the circumferential scale calculation along the field line; when calculating the normal scale of grid unit i, the midpoint of the surface charged line segment corresponding to this unit is extrapolated along the field line to the equipotential surface where points O and N are located as the lower point of the normal scale calculation, and then the newly added split lower point is extrapolated along the field line to the equipotential surface where points P and M are located as the upper point of the normal scale.

[0053] In one of the embodiments, the orthogonal field model construction process in step 102 includes: the orthogonal field adopts an electrostatic field model, sources are arranged on the grid edges of the surface grid, each grid edge is regarded as a uniformly charged line segment, and the electric potential and electric field strength of the entire surface grid at any point in space can be analytically solved according to the electric field superposition principle; a field line and equipotential line tracking model is constructed, in which the height or electric potential is specified as needed, the field strength direction at the current position is used as the propulsion direction, and the fourth-order Runge-Kutta method is used for extrapolation calculation; in the equipotential line tracking model, with the help of field lines, the field lines are extrapolated to the specified potential position based on the field line tracking model.

[0054] Specifically, the purpose of constructing the field line and equipotential line tracking model is to ensure that the field lines and equipotential lines are grid line clusters. The field line and equipotential line tracking model can be extrapolated along the field lines or equipotential lines.

[0055] In the equipotential line tracing model, the equipotential lines cannot be directly extrapolated along the equipotential lines. Instead, they need to be extrapolated from the field lines to the specified potential position based on the field line tracing model.

[0056] In one embodiment, the grid data structure adopts the data structure of grid nodes and cells, and records the grid nodes and cells respectively in the form of one-dimensional arrays, and the edge relationship is implicit in the data structure of the points.

[0057] Specifically, the grid data structure is the basis of grid generation. A good grid data structure takes up less memory and has higher generation efficiency. The topology and algorithm aspects need to be considered during design. Since the present invention adopts an orthogonal field model, it is necessary to add field-related parameters to the grid data structure. For quasi-two-dimensional quadrilateral grids, it is proposed to adopt the data structure of grid nodes and cells, and record the grid nodes and cells in the form of a one-dimensional array respectively. The edge relationship is implicit in the data structure of the point and does not need to be specially recorded.

[0058] In one embodiment, Figure 2 As shown in the figure, the overall process of growing anisotropic adaptive mesh generation includes: First, input the surface mesh, calculate the array advancement height, generate equipotential surfaces with minimum electric potential, and connect the meshes according to the geometric topology to generate the mesh framework of this layer. Secondly, call the anisotropic adaptive strategy to further improve the quality of the generated mesh units. The normal splitting conditions of all mesh units in this layer are judged, and the units that meet the conditions are normally split. After the normal splitting, the array uniformity test is performed on the array surface of this layer, and the twin block judgment starts from the top mesh unit of this layer, and the twin blocks that meet the circumferential merging conditions are circumferentially merged. After the circumferential merger, the mesh units that meet the circumferential splitting conditions are circumferentially split starting from the top mesh unit of this layer until all mesh units in this layer meet the standard scale requirements. The above anisotropic merging and splitting operations should meet the level requirements. Finally, complete the generation and improvement of the mesh of this layer, and generate the next layer of mesh in a similar way until the number of mesh layers meets the requirements and the mesh generation is completed.

[0059] The schematic diagram of the orthogonal field is as follows Figure 8 As shown in the figure, the meshing of the “F” shape is as follows Fig. 9 As shown, Fig. 9 (a) is a local enlarged schematic diagram. Fig. 9 (b) is a schematic diagram of the overall grid. The "F" shape contains special geometric features such as pits, external sharp corners, and internal concave corners, which can better reflect the typical complex geometric features such as pits and sharp corners on the aircraft, so it is used as an example.

[0060] It should be understood that although Figure 1The steps in the flowchart are shown in sequence as indicated by the arrows, but these steps are not necessarily executed in the order indicated by the arrows. Unless otherwise specified in this document, there is no strict order restriction for the execution of these steps, and these steps can be executed in other orders. Moreover, Figure 1 At least part of the steps may include multiple sub-steps or multiple stages. These sub-steps or stages are not necessarily executed at the same time, but can be executed at different times. The execution order of these sub-steps or stages is not necessarily sequential, but can be executed in turn or alternately with other steps or at least part of the sub-steps or stages of other steps.

[0061] In one embodiment, Fig.10 As shown, a growth-type anisotropic adaptive grid generation device based on an orthogonal field is provided, comprising: an initial surface grid acquisition module, a current layer grid unit generation module and a growth-type anisotropic adaptive grid adjustment module, wherein: An initial surface mesh acquisition module is used to acquire the initial surface mesh of the local complex quasi-two-dimensional geometric shape of the aircraft; The current layer grid unit generation module is used to use the initial surface grid or the upper array surface of the generated grid as the initial array surface, use the improved frontier layer-by-layer method based on the orthogonal field to generate the current layer grid unit, continue to generate the next layer of grid units, and perform anisotropic adaptive adjustment until the number of grid layers meets the requirements, and finally obtain the boundary layer grid and the spatial external field grid of the local complex quasi-two-dimensional geometric shape of the aircraft; the improved frontier layer-by-layer method based on the orthogonal field means: applying the orthogonal field model to the frontier layer-by-layer method, using the direction of the electric field intensity as the array propulsion direction, determining the propulsion distance by the electric potential, and connecting to form the current layer grid according to the geometric topological relationship; The growing anisotropic adaptive mesh adjustment module is used to adjust the mesh units of this layer by adopting an anisotropic adaptive strategy to obtain the adjusted mesh units of this layer; the anisotropic adaptive strategy is used to determine the number of normal splitting of each mesh unit after the mesh units of this layer are generated, and perform normal splitting on the units of this layer while meeting the level difference requirements; after all the units are normally split, perform array uniformity detection from the top unit of this layer, perform circumferential twin block discrimination and circumferential merging operations on densely distributed meshes, and perform circumferential splitting operations on sparsely distributed meshes; repeat this process until the array distribution is uniform and the distribution adjustment of the mesh units of this layer is completed; continue to generate the next layer of mesh units and perform anisotropic adaptive adjustment until the number of mesh layers reaches the preset value, and complete the generation of the complex quasi-two-dimensional geometric mesh of the local aircraft.

[0062] In one of the embodiments, the specific process of the improved frontal layer advancement method based on the orthogonal field in the current layer grid unit generation module includes: considering the initial surface grid or the upper array surface of the generated grid as the initial array surface, and using formula (1) to calculate the array surface advancement distance; calculating the direction of the electric field intensity at the grid nodes on the initial array surface, using the electric field intensity direction as the array surface advancement direction, advancing the array surface advancement distance according to the field lines, generating an equipotential surface with the lowest electric potential as the next array surface, and completing the current layer grid generation according to the geometric topological relationship; and subsequently repeating the process until the specified number of grid layers is generated.

[0063] In one of the embodiments, the specific process of the anisotropic adaptive strategy in the growing anisotropic adaptive grid generation module includes: judging the normal splitting conditions of all grid cells in this layer, and performing normal splitting on the cells that meet the conditions; after normal splitting, performing a uniformity test on the array surface of this layer; starting from the top grid cell of this layer, twin block judgment is performed, and circumferential merging is performed on the twin blocks that meet the circumferential merging conditions; after circumferential merging, starting from the top grid cell of this layer, circumferential splitting operations are performed on the grid cells that meet the circumferential splitting conditions, until all grid cells in this layer meet the expected grid scale requirements of this layer, and the grid generation and improvement of this layer are completed, wherein both the merging and splitting operations meet the level requirements.

[0064] It is worth noting that the expected grid scale of the first layer is a preset value, and the expected grid scales of subsequent layers are obtained through the grid continuity requirement, that is, the circumferential scale of the grid unit is determined according to the circumferential scale of the adjacent grid units to prevent drastic changes in the grid.

[0065] In one embodiment, the growing anisotropic adaptive grid generation module is also used to calculate the normal splitting times of each grid unit in the current layer through the normal splitting times calculation formula after the grid units in the current layer are generated; wherein the normal splitting times calculation formula is: ; in, SpNum is the number of normal splits, Indicates rounding down. h i for i The element normal scale, h under for i Normal scale of the neighboring cells below the cell.

[0066] The normal classification conditions of all grid cells in this layer are judged, and the cells that meet the conditions are normally split. After the normal splitting, the difference between the normal level of this unit and the normal level of the adjacent left and right grid cells does not exceed 1; the grid nodes of the calculated grid unit scale are used as new splitting points through the electric potential.

[0067] In one of the embodiments, the growing anisotropic adaptive grid generation module is also used to treat a pair of adjacent grid cells that meet the level requirements as a twin block starting from the top grid cell of the current layer; the level requirement of the twin block means that within the twin block, the normal levels of the two grid cells of the circumferential twin block should be the same, and the circumferential levels of the two grid cells of the normal twin block should be the same. The level requirement can ensure that adjacent grid cells contain a common grid edge; outside the twin block, the circumferential level of the circumferential twin block is not less than the circumferential level of the upper and lower adjacent grid cells, and the normal level of the normal twin block is not less than the normal level of the left and right adjacent grid cells; the twin blocks that meet the circumferential merging conditions are circumferentially merged.

[0068] In one of the embodiments, the level in the growing anisotropic adaptive grid generation module is used to standardize anisotropic splitting and merging operations; in the corresponding direction, the grid unit is anisotropically split once, and the level in this direction increases by 1; the twin blocks are merged once, and the level in this direction decreases by 1; the level difference of adjacent units is limited to 1 to ensure that the number of dangling nodes after anisotropic splitting or merging does not exceed 1.

[0069] In one embodiment, the grid cell scale includes a circumferential scale and a normal scale; the process of calculating the grid cell scale in the growing anisotropic adaptive grid generation module includes: recording the lower left grid node in the grid cell as point O, the upper left grid node as point P, the lower right grid node as point N, and the upper left grid node as point M; calculating the grid cell scale based on the orthogonal field, and calculating the grid cell scale. i In the circumferential scale: in grid units i The average potential of nodes O and P is calculated along the field line to calculate the left point, and the average potential of points N and M is calculated along the field line to calculate the right point; in the calculation grid unit i Normal scale: in grid units i The corresponding midpoint of the surface charged line segment is extrapolated along the field line to the equipotential surface where points O and N are located as the lower point for normal scale calculation, and then extrapolated from the newly added split lower point along the field line to the equipotential surface where points P and M are located as the upper point for normal scale.

[0070] In one of the embodiments, the orthogonal field model construction process in the grid unit generation module of this layer includes: the orthogonal field adopts the electrostatic field model, the source is arranged on the grid edge of the surface grid, each grid edge is regarded as a uniformly charged line segment, and the electric potential and electric field strength of the entire surface grid to any point in the space can be analytically solved according to the electric field superposition principle; a field line and equipotential line tracking model is constructed, in which the height or electric potential is specified as needed, the field strength direction at the current position is used as the propulsion direction, and the fourth-order Runge-Kutta method is used for extrapolation calculation; in the equipotential line tracking model, with the help of field lines, the field lines are extrapolated to the specified potential position based on the field line tracking model.

[0071] In one embodiment, the mesh data structure in the initial surface mesh acquisition module adopts a data structure of mesh nodes and cells, and the mesh nodes and cells are recorded in a one-dimensional array form, and the edge relationship is implicit in the data structure of the point.

[0072] The specific definition of the growth-type anisotropic adaptive grid generation device based on orthogonal fields can be found in the definition of the growth-type anisotropic adaptive grid generation method based on orthogonal fields above, which will not be repeated here. Each module in the above-mentioned growth-type anisotropic adaptive grid generation device based on orthogonal fields can be implemented in whole or in part by software, hardware and a combination thereof. The above-mentioned modules can be embedded in or independent of the processor in the computer device in the form of hardware, or can be stored in the memory in the computer device in the form of software, so that the processor can call and execute the operations corresponding to the above modules.

[0073] The technical features of the above embodiments may be combined arbitrarily. To make the description concise, not all possible combinations of the technical features in the above embodiments are described. However, as long as there is no contradiction in the combination of these technical features, they should be considered to be within the scope of this specification.

[0074] The above-described embodiments only express several implementation methods of the present application, and the descriptions thereof are relatively specific and detailed, but they cannot be understood as limiting the scope of the present application. It should be pointed out that, for a person of ordinary skill in the art, several variations and improvements can be made without departing from the concept of the present application, and these all belong to the protection scope of the present application. Therefore, the protection scope of the present application shall be subject to the attached claims.

Claims

1. A growth-type anisotropic adaptive grid generation method based on orthogonal fields, characterized in that: The method comprises: Obtain the initial surface mesh of the boundary layer of the local complex quasi-two-dimensional geometric shape of the aircraft; The initial surface grid is used as the initial front face, and the improved frontal layering method based on the orthogonal field is used to generate the grid unit of the current layer; the improved frontal layering method based on the orthogonal field refers to: applying the orthogonal field model to the frontal layering method, using the direction of the electric field intensity as the front face advancement direction, and controlling the advancement distance by the electric potential; The anisotropic adaptive strategy is used to adjust the grid units of the current layer to obtain the adjusted grid units of the current layer; the anisotropic adaptive strategy is used to determine the number of normal splitting of each grid unit after the grid units of the current layer are generated, and perform normal splitting on the units of the current layer while meeting the level difference requirements; after all the units are normally split, the array uniformity detection is performed from the top unit of the current layer, the circumferential twin block discrimination and circumferential merging operation are performed on the densely distributed grids, and the circumferential splitting operation is performed on the sparsely distributed grids; this process is repeated until the array distribution is uniform and the distribution adjustment of the grid units of the current layer is completed; Continue to generate the next layer of grid units and perform anisotropic adaptive adjustments until the number of grid layers reaches the preset value, completing the grid generation of the complex quasi-two-dimensional geometric shape of the local aircraft.

2. The method for generating anisotropic adaptive mesh based on orthogonal field according to claim 1, characterized in that: The specific process of the improved frontier progression method based on orthogonal fields includes: Considering the initial surface mesh as the initial front, the front advancement distance is calculated as: in z is the number of layers, is the growth rate, h For the z Layer front advancement distance, h 0 is the initial front advancement distance; The direction of the electric field intensity is calculated at the grid nodes on the initial array, and the direction of the electric field intensity is used as the array advancement direction. The array advancement distance is advanced according to the field lines, and the equipotential surface is generated with the lowest electric potential as the next array, and the grid generation of this layer is completed according to the geometric topological relationship. The process is deduced in this way until the specified number of grid layers is generated.

3. The method for generating anisotropic adaptive mesh based on orthogonal field according to claim 1, characterized in that: The specific process of anisotropic adaptive strategy includes: The normal splitting conditions are judged for all grid cells in this layer, and normal splitting is performed on cells that meet the conditions; After normal splitting, the uniformity of the array surface on this layer is tested; The twin blocks are identified starting from the top grid unit of this layer, and the twin blocks that meet the circumferential merging conditions are merged circumferentially; After the circumferential merging, the circumferential splitting operation is performed on the grid cells that meet the circumferential splitting conditions starting from the top grid cell of this layer until all grid cells of this layer reach the expected grid scale requirements of this layer, completing the grid generation and improvement of this layer. Among them, both the merging and splitting operations meet the level requirements.

4. The method for generating anisotropic adaptive mesh based on orthogonal field according to claim 3, characterized in that: The normal splitting conditions are judged for all grid cells in this layer, and normal splitting is performed on cells that meet the conditions, including: After the grid cells of this layer are generated, the normal splitting times of each grid cell of this layer are calculated by the normal splitting times calculation formula; the normal splitting times calculation formula is: in, SpNum is the number of normal splits, Indicates rounding down. h i for i The element normal scale, h under for i The normal scale of the neighboring cells under the cell; The normal classification conditions of all grid cells in this layer are judged, and the cells that meet the conditions are normally split. After normal splitting, the difference between the normal level of this unit and the normal level of the adjacent left and right grid cells does not exceed 1; The grid nodes at the scale of the calculated grid cells are used as new splitting points through the electric potential.

5. The method for generating anisotropic adaptive mesh based on orthogonal field according to claim 3, characterized in that: The twin blocks are identified starting from the top grid unit of this layer, and the twin blocks that meet the circumferential merging conditions are merged circumferentially, including: Starting from the top grid unit of this layer, a pair of adjacent grid units that meet the level requirements are regarded as a twin block; the level requirement of the twin block means that within the twin block, the normal levels of the two grid units of the circumferential twin block should be the same, and the circumferential levels of the two grid units of the normal twin block should be the same. The level requirement can ensure that the adjacent grid units contain a common grid edge; outside the twin block, the circumferential level of the circumferential twin block is not less than the circumferential level of the upper and lower adjacent grid units, and the normal level of the normal twin block is not less than the normal level of the left and right adjacent grid units; The twin blocks that meet the circumferential merging conditions are merged circumferentially.

6. The method for generating anisotropic adaptive mesh based on orthogonal field according to claim 3, characterized in that: Level is used to regulate anisotropic splitting and merging operations. In the corresponding direction, if the grid cell is anisotropically split once, the level in that direction increases by 1; if the twin blocks are merged once, the level in that direction decreases by 1. The level difference between adjacent units is limited to 1 to ensure that the number of dangling nodes does not exceed 1 after anisotropic splitting or merging.

7. The method for generating anisotropic adaptive mesh based on orthogonal field according to claim 3, characterized in that: The grid cell scale includes the circumferential scale and the normal scale; The process of calculating the grid cell size includes: The lower left grid node in the grid unit is recorded as point O, the upper left grid node is recorded as point P, the lower right grid node is recorded as point N, and the upper left grid node is recorded as point M; Based on the orthogonal field calculation grid unit scale, in the calculation grid unit i Circumferential scale: in grid units i The left point is calculated by taking the average potential of points O and P along the field line and extending the circumferential scale, and the right point is calculated by taking the average potential of points N and M along the field line and extending the circumferential scale; In the calculation grid cell i Normal scale: in grid units i The corresponding midpoint of the surface charged line segment is extrapolated along the field line to the equipotential surface where points O and N are located as the lower point for normal scale calculation, and then extrapolated from the newly added split lower point along the field line to the equipotential surface where points P and M are located as the upper point for normal scale.

8. The method for generating anisotropic adaptive mesh based on orthogonal field according to claim 1, characterized in that: The orthogonal field model construction process includes: the orthogonal field adopts the electrostatic field model, the source is arranged on the grid edge of the surface grid, and each grid edge is regarded as a uniformly charged line segment. According to the electric field superposition principle, the electric potential and electric field strength of the entire surface grid at any point in the space can be analytically solved; the field line and equipotential line tracking model is constructed. In the field line tracking model, the height or electric potential is specified as needed, and the field strength direction at the current position is used as the propulsion direction, and the fourth-order Runge-Kutta method is used for extrapolation calculation; in the equipotential line tracking model, with the help of field lines, the field line tracking model is used to extrapolate from the field lines to the specified potential position.

9. The method for generating anisotropic adaptive mesh based on orthogonal field according to any one of claims 1 to 8, characterized in that: The grid data structure uses the data structure of grid nodes and cells, and records the grid nodes and cells respectively in the form of one-dimensional arrays. The edge relationship is implicit in the data structure of the point.

10. A growth-type anisotropic adaptive grid generation device based on orthogonal fields, characterized in that: The device comprises: An initial surface mesh acquisition module is used to acquire the initial surface mesh of the local complex quasi-two-dimensional geometric shape of the aircraft; The current layer grid unit generation module is used to use the initial surface grid or the upper array surface of the generated grid as the initial array surface, use the improved frontier layer-by-layer method based on the orthogonal field to generate the current layer grid unit, continue to generate the next layer of grid units, and perform anisotropic adaptive adjustment until the number of grid layers meets the requirements, and finally obtain the boundary layer grid and the spatial external field grid of the local complex quasi-two-dimensional geometric shape of the aircraft; the improved frontier layer-by-layer method based on the orthogonal field refers to: applying the orthogonal field model to the frontier layer-by-layer method, using the direction of the electric field intensity as the array propulsion direction, determining the propulsion distance by the electric potential, and connecting to form the current layer grid according to the geometric topological relationship; A growing anisotropic adaptive mesh adjustment module is used to adjust the mesh units of this layer by adopting an anisotropic adaptive strategy to obtain the adjusted mesh units of this layer; the anisotropic adaptive strategy is used to determine the number of normal splitting of each mesh unit after the mesh units of this layer are generated, and perform normal splitting on the units of this layer while meeting the level difference requirements; after all the units are normally split, perform array uniformity detection from the top unit of this layer, perform circumferential twin block discrimination and circumferential merging operations on densely distributed meshes, and perform circumferential splitting operations on sparsely distributed meshes; repeat this process until the array distribution is uniform and the distribution adjustment of the mesh units of this layer is completed; continue to generate the mesh units of the next layer and perform anisotropic adaptive adjustment until the number of mesh layers reaches the preset value position, and complete the generation of the mesh of the local complex quasi-two-dimensional geometric shape of the aircraft.

Images