CN111489447A - Right-angle grid adaptive modeling method suitable for lattice Boltzmann method - Google Patents
Right-angle grid adaptive modeling method suitable for lattice Boltzmann method Download PDFInfo
- Publication number
- CN111489447A CN111489447A CN202010288618.8A CN202010288618A CN111489447A CN 111489447 A CN111489447 A CN 111489447A CN 202010288618 A CN202010288618 A CN 202010288618A CN 111489447 A CN111489447 A CN 111489447A
- Authority
- CN
- China
- Prior art keywords
- grid
- flow field
- adaptive
- object plane
- mesh
- Prior art date
- Legal status (The legal status is an assumption and is not a legal conclusion. Google has not performed a legal analysis and makes no representation as to the accuracy of the status listed.)
- Granted
Links
Images
Classifications
-
- G—PHYSICS
- G06—COMPUTING; CALCULATING OR COUNTING
- G06T—IMAGE DATA PROCESSING OR GENERATION, IN GENERAL
- G06T17/00—Three dimensional [3D] modelling, e.g. data description of 3D objects
- G06T17/20—Finite element generation, e.g. wire-frame surface description, tesselation
- G06T17/205—Re-meshing
-
- G—PHYSICS
- G06—COMPUTING; CALCULATING OR COUNTING
- G06F—ELECTRIC DIGITAL DATA PROCESSING
- G06F30/00—Computer-aided design [CAD]
- G06F30/20—Design optimisation, verification or simulation
- G06F30/23—Design optimisation, verification or simulation using finite element methods [FEM] or finite difference methods [FDM]
-
- G—PHYSICS
- G06—COMPUTING; CALCULATING OR COUNTING
- G06F—ELECTRIC DIGITAL DATA PROCESSING
- G06F30/00—Computer-aided design [CAD]
- G06F30/20—Design optimisation, verification or simulation
- G06F30/28—Design optimisation, verification or simulation using fluid dynamics, e.g. using Navier-Stokes equations or computational fluid dynamics [CFD]
-
- G—PHYSICS
- G06—COMPUTING; CALCULATING OR COUNTING
- G06F—ELECTRIC DIGITAL DATA PROCESSING
- G06F2113/00—Details relating to the application field
- G06F2113/08—Fluids
-
- G—PHYSICS
- G06—COMPUTING; CALCULATING OR COUNTING
- G06T—IMAGE DATA PROCESSING OR GENERATION, IN GENERAL
- G06T2210/00—Indexing scheme for image generation or computer graphics
- G06T2210/24—Fluid dynamics
Landscapes
- Engineering & Computer Science (AREA)
- Physics & Mathematics (AREA)
- Theoretical Computer Science (AREA)
- General Physics & Mathematics (AREA)
- Geometry (AREA)
- Computer Hardware Design (AREA)
- Evolutionary Computation (AREA)
- General Engineering & Computer Science (AREA)
- Computer Graphics (AREA)
- Software Systems (AREA)
- Algebra (AREA)
- Computing Systems (AREA)
- Fluid Mechanics (AREA)
- Mathematical Analysis (AREA)
- Mathematical Optimization (AREA)
- Mathematical Physics (AREA)
- Pure & Applied Mathematics (AREA)
- Management, Administration, Business Operations System, And Electronic Commerce (AREA)
Abstract
The embodiment of the invention discloses a right-angle grid adaptive modeling method suitable for a lattice Boltzmann method, which comprises the following steps: s101: dividing the root grid cell into right-angle sub-grid cells to obtain an initial grid comprising an initial object plane grid and an initial flow field grid; s102: carrying out right-angle grid self-adaptive processing on the initial object plane grid to obtain a final self-adaptive object plane grid, and taking the final self-adaptive object plane grid and the initial flow field grid as an initial calculation grid; s103: based on the current computational grid and by adopting a lattice Boltzmann method, performing flow field solution to obtain a current flow field computation result, and obtaining a flow field computation adaptive parameter based on the current flow field computation result; s104: judging whether all the flow field calculation adaptive parameters are smaller than a first set threshold value; s105: if so, taking the current flow field calculation result as a final flow field calculation result, otherwise, performing one-time right-angle grid self-adaption processing on the current flow field grid, taking the current flow field grid and the final self-adaption object plane grid as a new calculation grid, and returning to S103.
Description
Technical Field
The invention relates to the field of computational fluid mechanics, in particular to a right-angle grid adaptive modeling method suitable for a lattice Boltzmann method.
Background
However, the Method needs to adopt a set of Cartesian grid calculation, and mature Cartesian grid generation software suitable for the L BM Method is not seen at present, and only appears in large commercial software Powerflow.
The lattice Boltzmann method brings along with various advantages some of its inherent disadvantages, such as: the calculation amount is large, the grid division is symmetrical and uniform (regular hexagon and square), the calculation accuracy is not adjustable, and the like, so that the application of the lattice Boltzmann method in many aspects is limited. Therefore, the grid technology is a hot spot for studying the lattice Boltzmann method in academic circles at home and abroad.
The traditional lattice Boltzmann method uses a standard uniform grid. While the traditional lattice Boltzmann method based on a standard uniform grid exhibits many advantages, it also introduces some inherent limitations. One very significant problem arises in the simulation of complex boundary detours. In this type of problem, the motion of the area near the physical boundary changes dramatically, requiring the use of an encrypted mesh; but in the far field area, the flow changes are smooth, and the flow field can be well described by sparse grids. The lattice Boltzmann method under the standard grid has a dilemma that cannot be realized at all.
The problems presented above limit the ability of the lattice Boltzmann method to handle problems with far field conditions such as streaming. In recent years researchers have proposed some lattice Boltzmann methods based on non-uniform grid generation. However, most of these methods require second-order interpolation or special handling of sparse grid interfaces, which increases the complexity of the lattice Boltzmann method and may lose locality of computation or introduce additional errors.
Disclosure of Invention
In order to solve the technical problems, embodiments of the present invention are expected to provide a right-angle mesh adaptive modeling method suitable for a lattice Boltzmann method, which is adaptive to a geometric shape, and can better describe an object shape and adjust density distribution of a mesh near an object plane; the flow field can be more accurately simulated according to the flow change by self-adaptation of the flow field, the spatial grid distribution during flow field calculation is adjusted, the resolution of numerical simulation is improved, and thus clearer flow field details can be obtained.
The technical scheme of the invention is realized as follows:
the embodiment of the invention provides a right-angle grid self-adaptive modeling method suitable for a lattice Boltzmann method, which comprises the following steps:
s101: dividing a root grid cell covering the whole flow field area of a specially-shaped object into a plurality of right-angle sub-grid cells with given sizes to obtain an initial grid, wherein the initial grid comprises an initial object plane grid intersected with an object plane of the object and an initial flow field grid which is not intersected with the object plane of the object and is positioned outside the object;
s102: performing right-angle mesh self-adaptive processing on the initial object plane mesh based on the geometric shape of the object to obtain a final self-adaptive object plane mesh, and taking the final self-adaptive object plane mesh and the initial flow field mesh together as an initial calculation mesh;
s103: based on the current calculation grid and by adopting a lattice Boltzmann method, performing flow field solution to obtain a current flow field calculation result, and obtaining pressure gradients between all two adjacent grids in the current flow field grid based on the current flow field calculation result as a flow field calculation adaptive parameter;
s104: judging whether all the flow field calculation adaptive parameters are smaller than a first set threshold value;
s105: if so, taking the current flow field calculation result as a final flow field calculation result, otherwise, performing one-time right-angle grid adaptive processing on the current flow field grid to obtain a temporary adaptive flow field grid, taking the temporary adaptive flow field grid and the final adaptive object plane grid together as a new calculation grid, and returning to S103.
The embodiment of the invention provides a right-angle grid self-adaptive modeling method suitable for a grid Boltzmann method, a self-adaptive object plane grid means that a complex appearance is simulated through an uneven grid, the density distribution of grids near an object plane is adjusted, a self-adaptive computing grid means that a complex flow field is simulated through the uneven grid, the motion characteristic of fluid in the flow field is better described, the space grid distribution during flow field computing is adjusted, and clearer flow field details can be obtained, so that the problem that the complex appearance and the complex flow field are difficult to simulate by the even grid in the traditional grid Boltzmann method is solved, the number of grids used in the grid Boltzmann method is greatly reduced while the precision is ensured, and meanwhile, high-precision numerical value simulation of complex flows such as shock waves, vortexes and the like is easily carried out by combining a high-order computing format. In addition, the orthogonal grid self-adaptation enables the distribution function on grid points to evolve according to the standard grid Boltzmann method in each level of grid in the calculation process. Therefore, the calculation of the whole flow field can be realized only by performing data coupling on the virtual boundaries of the grids of different levels.
Drawings
Fig. 1 is a schematic diagram of a rectangular grid adaptive modeling method suitable for a lattice Boltzmann method according to an embodiment of the present invention;
fig. 2 is a schematic diagram of a rectangular grid adaptive processing performed on an initial object plane grid according to an embodiment of the present invention;
fig. 3 is a schematic diagram of a two-dimensional grid cell adaptive encryption and coarsening process according to an embodiment of the present invention;
fig. 4 is a schematic diagram of performing a rectangular grid adaptive processing on a current flow field grid according to an embodiment of the present invention;
FIG. 5 is a diagram illustrating neighboring grid points of a non-boundary grid point according to an embodiment of the present invention;
fig. 6 is a schematic diagram of a two-dimensional grid quadtree according to an embodiment of the present invention;
FIG. 7 is a schematic diagram of a three-dimensional mesh octree according to an embodiment of the present invention;
fig. 8 is a schematic diagram of a two-dimensional mesh unit subdivision process in a tree structure according to an embodiment of the present invention;
fig. 9 is a schematic diagram of a process for generating a two-dimensional grid in a tree structure according to an embodiment of the present invention;
fig. 10 is a schematic diagram of a two-dimensional full-tree adaptive mesh encryption according to an embodiment of the present invention;
FIG. 11 is a schematic diagram of a full-treeing three-dimensional mesh unit subdivision process according to an embodiment of the present invention;
FIG. 12 shows a NACA23012 airfoil computational grid map and a grid partial magnification schematic;
FIG. 13 shows a schematic of the computational streamlines of the NACA23012 airfoil;
FIG. 14 shows a computational grid diagram and a partially enlarged schematic diagram of an airfoil of the NACA23012 with light ice.
Detailed Description
The technical solutions in the embodiments of the present invention will be clearly and completely described below with reference to the drawings in the embodiments of the present invention.
Referring to fig. 1, it shows a right-angle mesh adaptive modeling method suitable for the lattice Boltzmann method provided by an embodiment of the present invention, where the method includes:
s101: dividing a root grid cell covering the whole flow field area of a specific-shaped object into a plurality of right-angle sub-grid cells with given sizes to obtain an initial grid, wherein the initial grid comprises an initial object plane grid intersecting with an object plane of the object and an initial flow field grid which does not intersect with the object plane of the object and is positioned outside the object, and the specific-shaped object can be an airplane wing for example;
s102: performing right-angle mesh self-adaptive processing on the initial object plane mesh based on the geometric shape of the object to obtain a final self-adaptive object plane mesh, and taking the final self-adaptive object plane mesh and the initial flow field mesh together as an initial calculation mesh;
s103: based on the current calculation grid and by adopting a lattice Boltzmann method, performing flow field solution to obtain a current flow field calculation result, wherein the flow field solution can be based on that the flow field calculation converges to a certain degree, and based on the current flow field calculation result, obtaining the pressure gradient between all two adjacent grids in the current flow field grid as a flow field calculation adaptive parameter;
s104: judging whether all the flow field calculation adaptive parameters are smaller than a first set threshold value;
s105: if so, taking the current flow field calculation result as a final flow field calculation result, otherwise, performing one-time right-angle grid adaptive processing on the current flow field grid to obtain a temporary adaptive flow field grid, taking the temporary adaptive flow field grid and the final adaptive object plane grid together as a new calculation grid, and returning to S103.
In the method, the self-adaptive object plane grid means that a complex shape is simulated through an uneven grid, the density distribution of the grid near the object plane is adjusted, the self-adaptive calculation grid means that a complex flow field is simulated through the uneven grid, the motion characteristics of fluid in the flow field are better described, the space grid distribution during flow field calculation is adjusted, and clearer flow field details can be obtained, so that the problem that the complex shape and the complex flow field are difficult to simulate through the even grid in the traditional grid Boltzmann method is solved, the number of the grids used in the grid Boltzmann method is greatly reduced while the accuracy is ensured, and meanwhile, the high-accuracy numerical simulation of complex flow such as shock waves, vortexes and the like is easily carried out by combining a high-order calculation format. In addition, the orthogonal grid self-adaptation enables the distribution function on grid points to evolve according to the standard grid Boltzmann method in each level of grid in the calculation process. Therefore, the calculation of the whole flow field can be realized only by performing data coupling on virtual boundaries of grids of different levels.
For the above method, in a preferred embodiment of the present invention, referring to fig. 2, the performing a rectangular mesh adaptive process on the initial object plane mesh based on the geometric shape of the object includes:
s1021: calculating a cosine value of an included angle between external normal unit vectors of specific planes contained by all two adjacent grids in the object plane grid to be self-adapted as a geometric shape self-adapting parameter, wherein the specific planes are planes opposite to the object plane of the object;
s1022: judging whether the geometric shape adaptive parameter is larger than a second set threshold value, namely judging whether the geometric shape adaptive parameter exists
(i ≠ j), wherein,
the external normal unit vectors represent any two different specific planes, and Constant is the second set threshold;
s1023: if so, taking the two corresponding adjacent grids as the final self-adaptive object plane grid, or combining the two corresponding adjacent grids into a single grid to realize coarsening and taking the single grid as the final self-adaptive object plane grid, and if not, performing encryption processing on the two corresponding adjacent grids once and returning to the step S1021.
It can be understood that, for the above-mentioned initial object plane mesh, all meshes are "object plane meshes to be adapted", two adjacent meshes after being subjected to the encryption processing once are also "object plane meshes to be adapted", and a mesh serving as a final object plane mesh does not belong to the "object plane mesh to be adapted".
The self-adaptive process of the object plane grid mainly comprises the step of encrypting the object plane grid, so that the grids near the object plane are reasonably distributed, and the convergence of flow field calculation is ensured.
The coarsening process can be regarded as the "inverse" process of the encryption process, i.e. the process of merging a plurality of child grids (the parent grids of which are all the same grid) to generate a parent grid. After the encryption process is realized, the current grid is a 'failure' parent grid, the subsequent flow field iterative computation is not participated, and the generated new sub-grid is an effective grid. The coarsening process is just the opposite, all the child grids participating in the coarsening become 'failure' grids, and the merged parent grid becomes an effective grid.
In the case where the computing hardware conditions allow, such as the two-dimensional case shown in fig. 3, the basic data information of the "dead" grid can be retained without deletion to save computing time, so as to facilitate the "activation" of the current parent grid cell during the coarsening. In other words, in this case, the encryption process may be regarded as a "disabling" process after subdivision, and the coarsening process may be regarded as a "activating" process after merging, so that the reconstruction time of the mesh data information in the adaptive process can be effectively saved. Of course, in the three-dimensional case, the data storage becomes huge by saving the data information of all the parent grids and the child grids, which is not allowed by the computer hardware condition, and at this time, the deletion and the update of the grid data information must be completed in the self-adapting process. In the present invention, this process is performed after the coarsening is completed, i.e. the encrypted parent mesh data information is retained, and the coarsened child mesh data information is deleted.
By adopting an encryption technology based on an object plane, the grid distribution is more efficient, the calculated amount is further reduced, the adaptive encryption grid has better adaptability to the complex shape change after icing, and the details of the streaming flow can be more carefully described. The accurate detailed description of the real ice shape by the grid cells near the ice shape can be improved. The frozen unsteady flow field involves complex phenomena such as flow separation, vortex motion and interference, boundary layer reattachment flow and the like, which lead to significant unsteady mechanics problems. Therefore, the unsteady dynamic response characteristics need to be fully studied for the aerodynamic performance influence after icing.
Considering the characteristics of a complex-configuration streaming flow field after icing, the higher requirements of a numerical calculation method on high efficiency, high precision and high resolution are met, the complexity of a calculation model and the numerical calculation workload are obviously increased, and the whole calculation time cost is also obviously increased, so that an adaptive technology and a corresponding algorithm research based on L BM are required to be developed.
For the above method, in a preferred embodiment of the present invention, referring to fig. 4, the performing a rectangular grid adaptive process on the current flow field grid includes:
s1051: taking two adjacent grids with the flow field calculation adaptive parameters smaller than the first set threshold value as the temporary adaptive flow field grids;
s1052: and performing encryption processing on two adjacent grids with the flow field calculation adaptive parameters not less than the first set threshold value once to serve as the temporary adaptive flow field grid.
Specifically, the first set threshold may be given by:
Vs＝Vm·ω
where Vs is a first set threshold, Vm is the maximum (or average) of the adaptive parameters at the center of the face unit separating the two arbitrary adjacent meshes, and ω is a scale factor (the maximum is typically 5%).
According to the transformation of the Gaussian theorem, the gradient of the physical quantity Q at the center of the grid cell 0
Wherein V0Is the grid cell 0 volume, (Q)0-m)interpRepresents the amount of Q interpolated on the grid plane cell between grid cell 0 and the adjacent grid m, and
representing the outer normal area vector of this grid face cell relative to grid cell 0.
Due to the continuous approach to reasonable grid distribution of space and object plane, the adaptive encryption of the grid to the flow field can make the calculation show good development trend, and the resolution of the numerical simulation can be greatly improved.
In addition, for the newly added trellis, the initial field value is obtained from the old field value.
On the other hand, when using the central finite volume method combined with orthogonal grid adaptation, the grid is considered as the object of investigation. For example, the center point coordinates of the grid cell, the type of the grid cell, and the sequence numbers of the neighboring grid cells need to be known. The lattice Boltzmann method is very different from this. In the evolution process of the distribution function of the method, the collision step only occurs at the current grid point and is not connected with the adjacent grid points, wherein the grid point of the grid refers to the intersection point between the orthogonal planes which divide the root grid unit into a plurality of right-angle grid units; the transportation step occurs between the current grid point and the adjacent grid points, and the distribution function on any grid point will migrate to the grid point adjacent to the point, which requires that the object under investigation is placed on the grid point of the flow field calculation. Each grid lattice needs to contain appropriate information to successfully implement numerical simulation.
Thus, for the above method, in a preferred embodiment of the present invention, the performing a flow field solution based on a current computation grid and by using a lattice Boltzmann method includes:
giving the serial number of each grid point in the current computational grid;
giving a type label of each lattice point in the current computational grid, wherein the far-field boundary lattice points have a first type label, the virtual boundary lattice points between grids with different self-adaption times have a second type label, the lattice points opposite to the object plane have a third type label, and other lattice points have a fourth type label;
and acquiring the serial numbers of the adjacent grid points of each grid point with the fourth type label in the current computing grid in different directions, wherein the different directions comprise an orthogonal direction, a plane diagonal direction and a space diagonal direction.
Taking the grid point P with the fourth type label in the two-dimensional grid in fig. 5 as an example, we need to know the sequence numbers of the adjacent points in the directions of 1-8: A. b, C, D, E, F, G and H.
With regard to the foregoing method, in a preferred embodiment of the present invention, the numbers of adjacent lattice points in the diagonal direction of the plane and in the diagonal direction of the space are obtained by transfer in two orthogonal directions in the plane and transfer in three orthogonal directions in the space, respectively, based on orthogonal characteristics of a rectangular grid.
Referring to fig. 5, for any non-boundary point P, we only give the sequence numbers of four adjacent points in the 1-4 directions: A. b, C and D. Now, considering point a, the serial number of the adjacent point in the 2 direction is E, and E is also the adjacent point in the P point 5 direction. Through the transfer of the point A, we can successfully get the adjacent point in the direction of the point P5. Similarly, the adjacent points in the directions of the P points 6-8 can also be obtained smoothly.
Therefore, the occupation of the memory during calculation can be reduced as much as possible so as to improve the calculation efficiency.
As can be seen from the above, the embedding depths required to be used in the object plane boundary processing may not be the same in the 1-8 directions, so for the lattice points included in the grid cut from the object plane, we need to give the adjacent points in the 1-8 directions and the embedding depth coefficients in each direction.
Thus, for the above method, in a preferred embodiment of the present invention, the method further includes obtaining the embedding depth coefficients corresponding to the different directions of each lattice point with the third-type label in the current computational mesh.
That is, we divide the grid in the flow field into a grid that is cut from the object plane and a grid that is not cut from the object plane. In practical calculation, since most grid points in the flow field belong to non-boundary points, the above design further improves the calculation efficiency to a certain extent.
With regard to the above method, in a preferred embodiment of the present invention, the method further comprises giving coordinates of each lattice point in the current computational grid, a distribution function in the different directions (nine directions in the two-dimensional computation and nineteen directions in the three-dimensional computation), a distribution function of equilibrium states in the different directions (nine directions in the two-dimensional computation and nineteen directions in the three-dimensional computation), a macro density, a macro velocity, a macro pressure. Thereby obtaining a complete data structure.
On the other hand, in a preferred embodiment of the present invention, for the above encryption processing, the encryption processing subdivides the mesh using a cross-tree data structure, records a mesh generation process, and stores information of mesh generation.
The data structure plays an important role in the right-angle grid encryption process and the flow field calculation process, and by means of organization and management of grid data, the grid encryption and data conversion are facilitated, and the transmission, searching and searching among data are accelerated. The cross tree data structure is adopted, so that almost all grid data can be obtained through the cross tree structure, and the mutual relations between grids and between grid surfaces and grid surfaces can also be obtained through logical search and search of the cross tree data by adopting methods of point-to-point comparison and surface-to-surface matching. Due to the inherent non-constant characteristic of the lattice Boltzmann method, a memory is very needed in the calculation process, so that the fact that the grid data storage can efficiently utilize the physical memory becomes particularly important, and the grid data storage adopts a cross-tree data structure, records grid generation and stores grid generation information, so that the memory can be efficiently used by utilizing a dynamic storage method.
For a two-dimensional rectangular grid, the basic tree data structure is a quadtree, i.e., a parent grid cell is divided into four child grid cells by two straight lines passing through the center of the parent grid cell and perpendicular to each other, as shown in FIG. 6; for a three-dimensional rectangular grid, the basic tree data structure is an octree, i.e., one parent grid cell is divided into eight child grid cells by three planes passing through the center of the parent grid cell and perpendicular to each other, as shown in FIG. 7.
For a single rectangular grid cell, fig. 8 shows the subdivision of the grid in a two-dimensional grid generation under the technical framework of a quadtree data structure.
The process of generating the grid by adopting the cross-tree data structure is that starting from one or more root grid units covering the whole calculation area, according to certain conditions and judgment, a plurality of sub-nodes are generated by one node, and a huge grid unit structure is generated by continuous circulation. The utilization of the data of the cross trees accelerates the progress of grid generation, reduces the time of grid generation and enables the relation search between grids to be convenient and fast.
Referring to fig. 9, a process for generating a grid system is shown. As shown in fig. 9, the left row is the Depth value (Depth) of the right-angle mesh, which represents the size of the current mesh unit and the number of subdivision layers, and for the "root" mesh, the Depth value is 0; on the right is a mesh generation graph (Grids) representing the overall process of mesh subdivision and generation from the "root" mesh; in the middle is a Tree data structure graph (Tree), which represents the growth process of the grid under the Tree structure. Obviously, the grid cell of the child node is not regenerated, i.e., the grid cell required for the calculation.
The quadtree or octree meshes are isotropic, are not efficient near the object plane, and have the disadvantage of a large number of meshes. Accordingly, with respect to the above-described fork tree data structure, in a preferred embodiment of the present invention, the fork tree data structure is a full fork tree data structure selected from any one of a binary tree, a quadtree, and an octree. If the full-fork tree is adopted, the number of new grids generated by grid encryption is greatly reduced.
For example, as shown in fig. 10, for a simple two-dimensional geometry, if a binary tree and quadtree mixed full-tree data structure is adopted, a relatively ideal discrete effect of the computation domain can be achieved.
As shown in fig. 11, in three-dimensional mesh generation, based on a full-tree, all possible subdivisions of one "parent" mesh cell, the first three "child" meshes are binary tree subdivisions, the next three "child" meshes are quadtree subdivisions, and the last "child" mesh is octree. In the actual mesh generation, one of them may be selected according to specific needs. After such subdivision, the mesh has anisotropic characteristics.
For the above-described cross-tree data structure, in a preferred embodiment of the present invention, the cross-tree data structure employs a four-level data pattern including point cell data, line cell data, plane cell data, and mesh cell data.
Data are organized and managed through the four-level data mode, data input and data output are conveniently achieved on the basis of the data structure of the cross tree, and therefore grid generation is convenient and fast. The data level of the point unit is lowest, the point unit is defined by a space position coordinate, and the position of the point unit in a physical space is specified; the grid cell data level is the highest and is the target of the whole generation process. The high level data consists of low level data, each cell having its own attributes. For example, a grid cell is composed of a plurality of face cells, and meanwhile, the grid cell has attributes such as a center point coordinate, a grid type, and a sequence number of an adjacent grid cell; the grid surface has a directional attribute, and the sequence numbers of grids on two sides of the grid surface are sequentially given.
The establishment of the four-level data mode not only greatly simplifies the organization and management of data, but also lays a good foundation for the smooth implementation of flow field calculation and self-adaptive algorithm. In the three-dimensional case, the data elements for points, lines, planes and grids are defined by the structural format, and the four data are closely related, interdependent. The grid consists of faces, the faces consist of lines, and the lines consist of points. Of these, the structures of the faces and meshes are the most complex. In mesh generation, the definition of its specific composition is important because they directly face the flow field calculation requirements. In the two-dimensional case, the mesh is a plane, and the four-level data will be degraded to three levels.
Fig. 12 to 14 show a mesh generation diagram and a calculation result diagram. Specifically, fig. 12 shows a NACA23012 airfoil computational grid map (top) and a grid partial magnification schematic (bottom); FIG. 13 shows a schematic of the computational streamlines of the NACA23012 airfoil; fig. 14 shows a calculated grid diagram (top) and a partially enlarged schematic diagram (bottom) of an airfoil illuminated ice-ice model of NACA 23012.
It should be noted that: the technical schemes described in the embodiments of the present invention can be combined arbitrarily without conflict.
The above description is only for the specific embodiment of the present invention, but the scope of the present invention is not limited thereto, and any person skilled in the art can easily conceive of the changes or substitutions within the technical scope of the present invention, and all the changes or substitutions should be covered within the scope of the present invention. Therefore, the protection scope of the present invention shall be subject to the protection scope of the appended claims.
Claims (10)
1. A right-angle mesh adaptive modeling method suitable for a lattice Boltzmann method is characterized by comprising the following steps:
s101: dividing a root grid cell covering the whole flow field area of a specific-shaped object into a plurality of right-angle sub-grid cells of given sizes to obtain an initial grid, wherein the initial grid comprises an initial object plane grid intersected with an object plane of the object and an initial flow field grid which is not intersected with the object plane of the object and is positioned outside the object;
s102: performing right-angle mesh self-adaptive processing on the initial object plane mesh based on the geometric shape of the object to obtain a final self-adaptive object plane mesh, and taking the final self-adaptive object plane mesh and the initial flow field mesh together as an initial calculation mesh;
s103: based on the current calculation grid and by adopting a lattice Boltzmann method, performing flow field solution to obtain a current flow field calculation result, and based on the current flow field calculation result, obtaining pressure gradients between all two adjacent grids in the current flow field grid as a flow field calculation adaptive parameter;
s104: judging whether all the flow field calculation adaptive parameters are smaller than a first set threshold value;
s105: if so, taking the current flow field calculation result as a final flow field calculation result, otherwise, performing one-time right-angle grid adaptive processing on the current flow field grid to obtain a temporary adaptive flow field grid, taking the temporary adaptive flow field grid and the final adaptive object plane grid together as a new calculation grid, and returning to S103.
2. The method of claim 1, wherein the performing a rectangular mesh adaptation on the initial object plane mesh based on the geometric shape of the object comprises:
s1021: calculating a cosine value of an included angle between external normal unit vectors of specific planes contained by each of two adjacent grids of an object plane grid to be self-adapted to serve as a geometric shape self-adaption parameter, wherein the specific plane is a plane opposite to an object plane of the object;
s1022: judging whether the geometric shape adaptive parameter is larger than a second set threshold value or not;
s1023: and if so, taking the two corresponding adjacent grids as the final self-adaptive object plane grid, or combining the two corresponding adjacent grids into a single grid to realize coarsening and taking the single grid as the final self-adaptive object plane grid, otherwise, performing encryption processing on the two corresponding adjacent grids for one time and returning to the step S1021.
3. The method of claim 1, wherein performing a rectangular grid adaptation on the current flow field grid comprises:
s1051: taking two adjacent grids with the flow field calculation adaptive parameters smaller than the first set threshold value as the temporary adaptive flow field grids;
s1052: and performing encryption processing on two adjacent grids with the flow field calculation adaptive parameters not less than the first set threshold value once to serve as the temporary adaptive flow field grid.
4. The method according to claim 1, wherein the performing the flow field solution based on the current computational grid and using a lattice Boltzmann method comprises:
giving the serial number of each grid point in the current computational grid;
giving a type label of each lattice point in the current computational grid, wherein the far-field boundary lattice points have a first type label, the virtual boundary lattice points between grids with different self-adaption times have a second type label, the lattice points opposite to the object plane have a third type label, and other lattice points have a fourth type label;
and acquiring the serial numbers of the adjacent grid points of each grid point with the fourth type label in the current computing grid in different directions, wherein the different directions comprise an orthogonal direction, a plane diagonal direction and a space diagonal direction.
5. The method according to claim 4, wherein the sequence numbers of the adjacent lattice points in the diagonal direction of the plane and in the diagonal direction of the space are obtained by two times of orthogonal direction transmission in the plane and three times of orthogonal direction transmission in the space respectively based on orthogonal characteristics of the rectangular grid.
6. The method according to claim 4, further comprising obtaining the embedding depth coefficients corresponding to the different directions for each grid point with the third type label in the current computational grid.
7. The method of claim 4, further comprising giving coordinates of each grid point in the current computational grid, a distribution function in the different direction, an equilibrium distribution function in the different direction, a macro density, a macro velocity, a macro pressure.
8. The method according to claim 1 or 2, wherein the encryption process adopts a cross-tree data structure to subdivide the mesh, records the generation process of the mesh and stores the information of the mesh generation.
9. The method of claim 8, wherein the tree data structure is a full tree data structure selected from any one of a binary tree, a quadtree, and an octree.
10. The method of claim 8, wherein the cross-tree data structure employs a four-level data pattern comprising point cell data, line cell data, plane cell data, and grid cell data.
Priority Applications (1)
| Application Number | Priority Date | Filing Date | Title |
|---|---|---|---|
| CN202010288618.8A CN111489447B (en) | 2020-04-14 | 2020-04-14 | Right-angle grid adaptive modeling method suitable for lattice Boltzmann method |
Applications Claiming Priority (1)
| Application Number | Priority Date | Filing Date | Title |
|---|---|---|---|
| CN202010288618.8A CN111489447B (en) | 2020-04-14 | 2020-04-14 | Right-angle grid adaptive modeling method suitable for lattice Boltzmann method |
Publications (2)
| Publication Number | Publication Date |
|---|---|
| CN111489447A true CN111489447A (en) | 2020-08-04 |
| CN111489447B CN111489447B (en) | 2022-04-29 |
Family
ID=71797821
Family Applications (1)
| Application Number | Title | Priority Date | Filing Date |
|---|---|---|---|
| CN202010288618.8A Active CN111489447B (en) | 2020-04-14 | 2020-04-14 | Right-angle grid adaptive modeling method suitable for lattice Boltzmann method |
Country Status (1)
| Country | Link |
|---|---|
| CN (1) | CN111489447B (en) |
Cited By (4)
| Publication number | Priority date | Publication date | Assignee | Title |
|---|---|---|---|---|
| CN112992294A (en) * | 2021-04-19 | 2021-06-18 | 中国空气动力研究与发展中心计算空气动力研究所 | Porous medium LBM calculation grid generation method |
| CN113505443A (en) * | 2021-09-09 | 2021-10-15 | 南京航空航天大学 | Self-adaptive Cartesian grid generation method for three-dimensional streaming problem with any shape |
| CN114282381A (en) * | 2021-12-28 | 2022-04-05 | 中国航天空气动力技术研究院 | Method and system for automatically generating non-uniform right-angle grid |
| CN115796083A (en) * | 2023-02-17 | 2023-03-14 | 中国空气动力研究与发展中心计算空气动力研究所 | Helicopter flow field simulation method, device and equipment and readable storage medium |
Citations (6)
| Publication number | Priority date | Publication date | Assignee | Title |
|---|---|---|---|---|
| JP2013037434A (en) * | 2011-08-04 | 2013-02-21 | Hitachi Ltd | Fluid analysis method, fluid analysis device, and fluid analysis program |
| CN103345580A (en) * | 2013-07-02 | 2013-10-09 | 上海大学 | Parallel CFD method based on lattice Boltzmann method |
| CN108388742A (en) * | 2018-03-05 | 2018-08-10 | 北京空间技术研制试验中心 | A kind of simulating analysis of spacecraft separation with pressure |
| CN108710734A (en) * | 2018-05-04 | 2018-10-26 | 南京特雷西能源科技有限公司 | Method for numerical simulation and device based on mesh adaption encryption and coarsening technique |
| CN108984874A (en) * | 2018-07-02 | 2018-12-11 | 中国航空发动机研究院 | Obtain the method for numerical simulation in the flow field of potential flows |
| EP3528146A1 (en) * | 2018-02-20 | 2019-08-21 | Dassault Systemes Simulia Corp. | Lattice boltzmann based solver for high speed flows |
-
2020
- 2020-04-14 CN CN202010288618.8A patent/CN111489447B/en active Active
Patent Citations (6)
| Publication number | Priority date | Publication date | Assignee | Title |
|---|---|---|---|---|
| JP2013037434A (en) * | 2011-08-04 | 2013-02-21 | Hitachi Ltd | Fluid analysis method, fluid analysis device, and fluid analysis program |
| CN103345580A (en) * | 2013-07-02 | 2013-10-09 | 上海大学 | Parallel CFD method based on lattice Boltzmann method |
| EP3528146A1 (en) * | 2018-02-20 | 2019-08-21 | Dassault Systemes Simulia Corp. | Lattice boltzmann based solver for high speed flows |
| CN108388742A (en) * | 2018-03-05 | 2018-08-10 | 北京空间技术研制试验中心 | A kind of simulating analysis of spacecraft separation with pressure |
| CN108710734A (en) * | 2018-05-04 | 2018-10-26 | 南京特雷西能源科技有限公司 | Method for numerical simulation and device based on mesh adaption encryption and coarsening technique |
| CN108984874A (en) * | 2018-07-02 | 2018-12-11 | 中国航空发动机研究院 | Obtain the method for numerical simulation in the flow field of potential flows |
Non-Patent Citations (2)
| Title |
|---|
| SONGBAICHENG等: "CFD analysis of flow field in a 5 × 5 rod bundle with multi-grid", 《ANNALS OF NUCLEAR ENERGY》 * |
| 卓丛山等: "高雷诺数下的翼型绕流LBM数值模拟", 《航空学报》 * |
Cited By (5)
| Publication number | Priority date | Publication date | Assignee | Title |
|---|---|---|---|---|
| CN112992294A (en) * | 2021-04-19 | 2021-06-18 | 中国空气动力研究与发展中心计算空气动力研究所 | Porous medium LBM calculation grid generation method |
| CN113505443A (en) * | 2021-09-09 | 2021-10-15 | 南京航空航天大学 | Self-adaptive Cartesian grid generation method for three-dimensional streaming problem with any shape |
| CN113505443B (en) * | 2021-09-09 | 2021-12-14 | 南京航空航天大学 | Self-adaptive Cartesian grid generation method for three-dimensional streaming problem with any shape |
| CN114282381A (en) * | 2021-12-28 | 2022-04-05 | 中国航天空气动力技术研究院 | Method and system for automatically generating non-uniform right-angle grid |
| CN115796083A (en) * | 2023-02-17 | 2023-03-14 | 中国空气动力研究与发展中心计算空气动力研究所 | Helicopter flow field simulation method, device and equipment and readable storage medium |
Also Published As
| Publication number | Publication date |
|---|---|
| CN111489447B (en) | 2022-04-29 |
Similar Documents
| Publication | Publication Date | Title |
|---|---|---|
| CN111489447B (en) | Right-angle grid adaptive modeling method suitable for lattice Boltzmann method | |
| CN110069800B (en) | Three-dimensional structure topology optimization design method and equipment with smooth boundary expression | |
| CN113505443B (en) | Self-adaptive Cartesian grid generation method for three-dimensional streaming problem with any shape | |
| CN111768502A (en) | Non-structural grid two-dimensional flood simulation system based on GPU acceleration technology | |
| US9971335B2 (en) | Hybrid dynamic tree data structure and accessibility mapping for computer numerical controlled machining path planning | |
| CN101976468A (en) | Method and system for visualizing multiresolution dynamic landform | |
| CN106485766A (en) | A kind of parallel constructing method of constraint Delaunay triangulation network | |
| Brix et al. | Adaptive multiresolution methods: Practical issues on data structures, implementation and parallelization | |
| CN104361185A (en) | Wire distribution space automatic generating method for cable virtual design | |
| JP4778558B2 (en) | High-speed processing method of coupled equations by finite element method and boundary element method | |
| CN115345988A (en) | Secondary error measurement edge folding BIM lightweight method based on vertex importance | |
| Mueller‐Roemer et al. | Ternary sparse matrix representation for volumetric mesh subdivision and processing on GPUs | |
| CN116071519A (en) | Image processing method and device for generating grid model based on harmonic mapping | |
| CN117473655B (en) | Aircraft simulation driving design method and device based on edge collapse grid optimization | |
| CN113342999B (en) | Variable-resolution-ratio point cloud simplification method based on multi-layer skip sequence tree structure | |
| Ali et al. | Visualization of large time-varying vector data | |
| CN108875132B (en) | Topological optimization method of stress constraint type truss continuum structure under multiple working conditions | |
| CN113191016A (en) | Body expression model-based multi-material product modeling and analyzing integrated method | |
| CN113178011A (en) | Cut grid THINC method for solving VOF convection equation | |
| CN117010241A (en) | Automatic design method for field-guided lattice structure | |
| CN115222870B (en) | Unstructured grid implicit LUSGS (Lusgs) balanced coloring method, equipment and medium | |
| CN114119882B (en) | Efficient nested grid host unit searching method in aircraft dynamic flow field analysis | |
| CN110349265B (en) | Tetrahedral topological mesh generation method and electronic equipment | |
| Duan et al. | High order FR/CPR method for overset meshes | |
| Alexa | Polycover: Shape approximating with discrete surface orientation |
Legal Events
| Date | Code | Title | Description |
|---|---|---|---|
| PB01 | Publication | ||
| PB01 | Publication | ||
| SE01 | Entry into force of request for substantive examination | ||
| SE01 | Entry into force of request for substantive examination | ||
| GR01 | Patent grant | ||
| GR01 | Patent grant |