CN114004174A - Efficient host unit searching method suitable for multiple sets of complex grid coupling CFD (computational fluid dynamics) calculation - Google Patents

Efficient host unit searching method suitable for multiple sets of complex grid coupling CFD (computational fluid dynamics) calculation Download PDF

Info

Publication number
CN114004174A
CN114004174A CN202111272625.XA CN202111272625A CN114004174A CN 114004174 A CN114004174 A CN 114004174A CN 202111272625 A CN202111272625 A CN 202111272625A CN 114004174 A CN114004174 A CN 114004174A
Authority
CN
China
Prior art keywords
unit
grid
plane
search
node
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
Application number
CN202111272625.XA
Other languages
Chinese (zh)
Other versions
CN114004174B (en
Inventor
金奕星
吴乘胜
王星
王建春
何术龙
Current Assignee (The listed assignees may be inaccurate. Google has not performed a legal analysis and makes no representation or warranty as to the accuracy of the list.)
Csic Orlando Wuxi Software Technology Co ltd
Original Assignee
Csic Orlando Wuxi Software Technology Co ltd
Priority date (The priority date 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 date listed.)
Filing date
Publication date
Application filed by Csic Orlando Wuxi Software Technology Co ltd filed Critical Csic Orlando Wuxi Software Technology Co ltd
Priority to CN202111272625.XA priority Critical patent/CN114004174B/en
Publication of CN114004174A publication Critical patent/CN114004174A/en
Application granted granted Critical
Publication of CN114004174B publication Critical patent/CN114004174B/en
Active legal-status Critical Current
Anticipated expiration legal-status Critical

Links

Images

Classifications

    • 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
    • G06F—ELECTRIC DIGITAL DATA PROCESSING
    • G06F2119/00—Details relating to the type or aim of the analysis or the optimisation
    • G06F2119/14—Force analysis or force optimisation, e.g. static or dynamic forces
    • Y—GENERAL TAGGING OF NEW TECHNOLOGICAL DEVELOPMENTS; GENERAL TAGGING OF CROSS-SECTIONAL TECHNOLOGIES SPANNING OVER SEVERAL SECTIONS OF THE IPC; TECHNICAL SUBJECTS COVERED BY FORMER USPC CROSS-REFERENCE ART COLLECTIONS [XRACs] AND DIGESTS
    • Y02—TECHNOLOGIES OR APPLICATIONS FOR MITIGATION OR ADAPTATION AGAINST CLIMATE CHANGE
    • Y02T—CLIMATE CHANGE MITIGATION TECHNOLOGIES RELATED TO TRANSPORTATION
    • Y02T90/00—Enabling technologies or technologies with a potential or indirect contribution to GHG emissions mitigation

Abstract

The invention provides an efficient host unit searching method suitable for multiple sets of complex grid coupling CFD calculation, which comprises the following steps: s1, proposing a concept of barycentric coordinates, determining the position relation of the nodes and the units according to the barycentric coordinates of the current nodes and the current judging units, and giving a searching direction; s2, when the gravity center coordinates of the nodes and the units are calculated, and under the condition that more than 3 grid points exist in one plane of any polyhedral unit, dyeing and marking the nodes in the plane, and ensuring that each time the calculation is carried out, 3 points determine that at least one grid point in one plane is not adopted before; s3, for the grid unit with abnormal and complex individual unit form, when the barycentric coordinates of the nodes and the units are calculated, dyeing and marking the nodes in the plane and outside the plane of the current unit; and S4, optimizing the search starting unit of the node which does not start to be searched according to the searched host unit information, and improving the searching efficiency. The invention can greatly improve the robustness and the search efficiency of the host unit search in the complex computational grid.

Description

Efficient host unit searching method suitable for multiple sets of complex grid coupling CFD (computational fluid dynamics) calculation
Technical Field
The invention relates to the field of computational fluid mechanics, in particular to an efficient host unit searching method suitable for multiple sets of complex grid coupling CFD (computational fluid dynamics) calculation.
Background
With the development of the CFD technology, the problem of the fluid mechanics numerical simulation becomes more and more complex, and in the numerical simulation of the multi-body relative motion problem, a multi-reference coordinate system method, a sliding grid method, an overlapping grid method and other sets of grid coupling computing technologies are often adopted, wherein the host unit search operation of the computing grid is performed.
For the above-mentioned several common CFD calculation methods for processing multi-body relative motion, the correctness of host unit search is crucial, and only if the correct host unit is found out from the interpolation point, the transfer of the flow field between the calculation domains can be correctly performed, thereby realizing multi-calculation-domain grid coupling calculation. In addition, robustness of the host cell search is also very important, especially for non-stationary computing methods such as the sliding grid method and the overlapping grid method. In the process of non-stationary calculation, the grid has relative motion, after each motion of the grid, host cell search needs to be performed again, the relative motion of the grid increases the complexity of the situation of connecting cells between domains, increases the search difficulty, and increases the frequency of host cell search, and as long as one host cell search fails, the whole calculation is interrupted, which is not acceptable for non-stationary calculation. It follows that robustness and correctness of the host cell search are equally important.
In the CFD calculation, as the geometric model becomes more complex, the difficulty in generating a high-quality computational grid becomes greater and greater, and the adopted grid division manner becomes more and more diverse, which undoubtedly puts higher requirements on the robustness and the accuracy of the host cell search. At present, in a complex unit form computational grid, especially in any polyhedral grid adopting a cutting unit method combined with a paste prism layer, the traditional host unit searching method has the problem that a host unit cannot be searched or is searched circularly in the searching process. The problem of host unit search robustness is unacceptable, and a diversified grid unit form capable of processing complex computational grids is urgently needed, and meanwhile, the host unit search robustness and the host unit search efficiency are good.
Disclosure of Invention
The invention aims to at least solve the technical problems in the prior art, and particularly creatively provides an efficient host unit searching method suitable for multiple sets of complex grid coupling CFD (computational fluid dynamics) calculation.
In order to achieve the above object, the present invention provides an efficient host cell search method suitable for multiple sets of complex mesh coupled CFD calculations, comprising the following steps:
s1, determining the position relation of the nodes and the units by calculating the barycentric coordinates of the nodes and the units, and giving a searching direction; in the host unit searching process, if the unit is repeatedly searched, the repeated searching times of the unit is increased by one; judging whether the node is in the cell, if so, finding the position relation between the host cell and the cell, otherwise, executing the step S2;
s2, judging whether the repetition times of the unit on the search path in the search process is less than m times, if so, executing a step S1, and if not, executing a step S3;
s3, judging whether the repeated searching times of the unit is more than n times, if so, executing the step S4, otherwise, adopting a dyeing marking method when calculating the gravity center coordinates of the node and the unit, and adding one to the repeated searching times of the unit every time of searching; judging whether the node is in the cell, if so, finding the position relation between the host cell and the cell, otherwise, executing the step S3;
s4, for the grid unit with abnormal and complex unit form, adopting a double-dyeing marking method when calculating the gravity center coordinates of the nodes and the units, and not only dyeing and marking the in-plane nodes of the current unit, but also dyeing and marking the out-of-plane nodes; adding one to the repeated searching times of the unit every time of searching; judging whether the node is in the cell, if so, finding the position relation between the host cell and the cell, otherwise, executing the step S4;
and S5, optimizing the search starting unit of the node which does not start to search according to the searched host unit information, namely, using the searched host unit as the search starting unit of the node connected with the node, thereby improving the search efficiency.
Where m is equal to {10,20, 30, 40}, n is equal to 2m, and the larger the value of m and n is, the more search time is consumed.
A grid point generally refers to a grid point constituting a grid cell, and a node refers to an interpolation point required to find a host cell.
Further, the calculating barycentric coordinates of the nodes and the units comprises:
s1-1, using the ith out-of-plane lattice point of the K unit as a vertex, and taking three lattice points in the ith plane of the K unit as vectors
Figure BDA0003329205070000031
Wherein
Figure BDA0003329205070000032
Representing a vector
Figure BDA0003329205070000033
A vector mixture product of;
s1-2, using the node as the vertex, and taking the three lattice points in the ith surface of the K unit as vectors
Figure BDA0003329205070000034
Then
Figure BDA0003329205070000035
S1-3, determining barycentric coordinates of the space relation between the nodes and the cells, and determining the search direction according to the barycentric coordinates, wherein the barycentric coordinates are calculated according to the following formula:
Figure BDA0003329205070000036
wherein K represents the current cell;
for a two-dimensional cell, βiThat is nodes and units are related toThe barycentric coordinates of edge i; a. theKRepresenting the area formed by the ith edge of the K unit and any grid point of the K unit outside the ith edge;
Figure BDA0003329205070000037
representing the area formed by the node and the ith side of the K unit;
for three-dimensional cells, betaiThat is, the barycentric coordinates of the nodes and cells with respect to plane i; a. theKRepresenting that the ith surface of the K unit and any grid point of the out-of-plane K unit form the volume of a geometric body;
Figure BDA0003329205070000038
representing the volume of the geometry formed by the node and the ith surface of the K unit.
The method specifically comprises the following steps: taking the outer grid points of the plane of the K unit i as vertexes, and taking three grid points in the plane as vectors
Figure BDA0003329205070000039
Then
Figure BDA00033292050700000310
In the same way to obtain
Figure BDA00033292050700000311
Taking the current node as a vertex, and taking three lattice points in the plane as vectors
Figure BDA00033292050700000312
Then
Figure BDA00033292050700000313
Then beta is obtained by calculationiBeta thus calculatediIs signed, and the positive and negative of the signed value represent the position relationship between the current node and the current unit K.
Wherein
Figure BDA0003329205070000041
Representing a vector
Figure BDA0003329205070000042
The vector mixture product of (2).
For AKAnd
Figure BDA0003329205070000043
the calculation of (a) is obtained by spatially resolving the geometry, using the geometric meaning of the vector-mixed product.
In practical application, a three-dimensional unit is taken as a main part, the barycentric coordinate is not a true coordinate, and only a concept is provided for expressing the position relationship between each surface of the node and each surface of the unit, and the position relationship between the node and the unit is determined through the barycentric coordinate. Judging whether the node is in the cell according to the sign of the barycentric coordinate, wherein the positive and negative represent the position relation between the current node and the current cell, and if the barycentric coordinate values are all regular, finding the host cell of the current node; if only one barycentric coordinate is negative, the direction of the next search is revealed.
Further, the staining labeling method in S3 includes:
in any polyhedral mesh, some faces of the mesh cells contain mesh points more than 3, and theoretically, the mesh points of the same cell face are in the same plane, but due to data errors of mesh generation software or mesh output data errors, the mesh points are not in the same plane mathematically. If the node for which the host cell is being sought is near the grid surface, the result of barycentric coordinates is distorted due to slight distortion of the grid surface, so that the search direction points back and forth between two or more cells, and finally the host cell of the node cannot be determined.
In any polyhedral grid, the grid unit has a plurality of faces, and the shape of each face cannot be predicted. The dyeing marking method can solve the problem:
determining a grid cell plane from optional 3 grid points in the ith plane of the K cell, and sorting the selected 3 grid points according to local numbers, for example, the optional grid point is P0,P4,P2Then is P after sorting according to local number0,P2,P4Determine the groupIf the combination has the mark, the combination is reselected if the combination has the mark, and if the combination does not have the mark, the combination is marked with group (P)0,P2,P4)＝color。
Further, at least one of the 3 grid points is not previously used.
Further, the exceptionally complex grid cell in S4 includes:
cells shaped other than the conventional mesh cell types in fluid computational meshes-tetrahedrons, hexahedrons, triangular prisms, rectangular pyramids, or more than 10 faces per mesh cell, or more than 5 mesh points within a cell boundary.
Further, the double staining labeling method in S4 includes:
the grid cell plane is determined from optionally 3 grid points of the in-plane grid points, such as P0,P4,P2Optionally an out-of-plane grid point, e.g. P5Arranging the out-of-plane lattice points at the first position, arranging the in-plane lattice points at the second position to the fourth position after sorting according to the local numbers, judging whether the combination has a mark, if so, reselecting, and if not, dyeing and marking the combination, such as group (P)5,P0,P2,P4)＝color。
Further, comprising: at least one of the 3 grid points is previously unused or the out-of-plane grid point is previously unselected.
Further, the S5 includes:
and using the node information of which the host unit search is completed as a search starting unit of the node connected with the node.
In summary, due to the adoption of the technical scheme, the invention has the beneficial effects that:
1. the adjacent unit searching method based on the barycentric coordinate concept carries out host unit searching, and adds a dyeing marking method to cope with the complex grid unit form possibly appearing in any polyhedral grid in the barycentric coordinate calculation process, so that the robustness of host unit searching in the complex calculation grid can be greatly improved.
2. The invention optimizes the initial unit searched by other node host units by using the node information and the unit information which are searched by the host unit, can greatly shorten the host unit search path and improve the host unit search efficiency.
Additional aspects and advantages of the invention will be set forth in part in the description which follows and, in part, will be obvious from the description, or may be learned by practice of the invention.
Drawings
The above and/or additional aspects and advantages of the present invention will become apparent and readily appreciated from the following description of the embodiments, taken in conjunction with the accompanying drawings of which:
FIG. 1 is a diagram showing the positional relationship between nodes and cells represented by barycentric coordinate symbols in the present invention.
FIG. 2 is a schematic diagram illustrating the determination of the searching direction of the host cell according to the barycentric coordinates in the present invention.
FIG. 3 is a schematic view of an in-plane grid point dyed marking in accordance with the present invention.
FIG. 4 is a schematic view of a dual-colored marker of an in-plane grid point and an out-of-plane grid point in the present invention.
FIG. 5 is a diagram of the present invention for optimizing search starting units using found host units.
FIG. 6 is a view of the interface and its vicinity of the computational grid in an example of the present invention.
FIG. 7 shows the results of a host cell search performed without the present invention in an example of the present invention.
FIG. 8 shows the results of a host cell search performed by the present invention in an example of the present invention.
FIG. 9 is a flow chart of the host cell search of the present invention.
Detailed Description
Reference will now be made in detail to embodiments of the present invention, examples of which are illustrated in the accompanying drawings, wherein like or similar reference numerals refer to the same or similar elements or elements having the same or similar function throughout. The embodiments described below with reference to the accompanying drawings are illustrative only for the purpose of explaining the present invention, and are not to be construed as limiting the present invention.
The method is based on the searching of adjacent units based on the barycentric coordinate concept, combines a searching unit dyeing marking method, assists in judging the searching path of a host unit, and assists in judging by a double dyeing marking method under the condition that the form of a grid unit is abnormal and complicated; in addition, according to the host unit information which is searched, the search starting unit of the node which is not searched is optimized, and the search efficiency is improved. The method can process the diversified grid unit form of the complex computational grid, and has good search robustness and search efficiency.
The technical scheme of the invention is as follows:
s1, proposing a concept of barycentric coordinates, determining the position relation of the nodes and the units according to the barycentric coordinates of the current nodes and the current judging units, and giving a searching direction;
and judging the position relation of the nodes and the units by calculating barycentric coordinates of each surface of the nodes and the units. Determining a searching direction through the barycentric coordinates, searching a host unit, and counting the repeated searching times of a certain unit if repeated searching of the unit occurs in the searching process. If the repetition times of the units on the search path are less than 10 times in the search process, the dyeing marks are not needed for auxiliary search; finally, the position relation between the node and the unit is obtained, namely whether the node is in the unit or not; according to the fact that the number of times of repeated searching of a unit is used as a judgment standard in the searching process, the repeated searching is only the case, and basically the case that more than 3 grid points exist on one surface in any polyhedron is the case.
S2, when the gravity center coordinates of the nodes and the units are calculated, and under the condition that more than 3 grid points exist in one plane of any polyhedral unit, dyeing and marking the nodes in the plane, and ensuring that each time the calculation is carried out, 3 points determine that at least one grid point in one plane is not adopted before;
s3, for the grid unit with abnormal and complex individual unit form, when the gravity center coordinates of the nodes and the units are calculated, a double-dyeing marking method is adopted, and not only the in-plane nodes of the current units are dyed and marked, but also the out-of-plane nodes are dyed and marked;
and S4, optimizing the search starting unit of the node which does not start to be searched according to the searched host unit information, and improving the searching efficiency.
The S1 determines barycentric coordinates of a spatial relationship between the nodes and the cells by introducing a concept of barycentric coordinates, and determines the search direction according to the barycentric coordinates, wherein the barycentric coordinates are calculated as follows:
Figure BDA0003329205070000071
wherein beta isiI.e. the coordinates of the center of gravity of the node and the cell with respect to plane i, K representing the current cell and i representing the ith plane of the K cell. A. theKAny grid point of the representing unit K on the ith surface and the K unit out of the surface forms the volume of the geometrical body,
Figure BDA0003329205070000072
representing the volume of the geometry formed by the current node and the ith face of the K cell. For AKAnd
Figure BDA0003329205070000073
the calculation of (a) is obtained by spatially resolving the geometry, using the geometric meaning of the vector-mixed product. Taking the outer grid points of the plane of the K unit i as vertexes, and taking three grid points in the plane as vectors
Figure BDA0003329205070000074
Then
Figure BDA0003329205070000075
In the same way to obtain
Figure BDA0003329205070000076
Vector to three lattice points in plane
Figure BDA0003329205070000077
Then
Figure BDA0003329205070000078
Is recalculated to obtainβiBeta thus calculatediIs signed, and the positive and negative of the signed value represent the position relationship between the current node and the current unit K.
Taking a two-dimensional triangular unit as an example, the possible situation that the barycentric coordinates of the nodes and the unit exist is shown in figure 1, and in the two-dimensional situation, P0,P1,P2The method is a grid point of a current unit, a traditional method does not relate to a dyeing marking method, the dyeing marking method is used for marking an inner grid point of the opposite side, and a double-dyeing marking method is used for marking the inner grid point of the opposite side and an outer grid point of the opposite side at the same time. The line is the boundary of the cell in the two-dimensional case, i.e. the line between the grid points. When the node and the line outer grid point are on the same side of the line, the corresponding barycentric coordinate of the line is positive, otherwise, the barycentric coordinate of the node and the unit in fig. 1 is negative, and there are 7 possible situations. After the barycentric coordinates of the nodes and the units are obtained through calculation, the position relation of the nodes and the units can be judged, and the searching direction is determined. Also taking a two-dimensional triangle unit as an example, if all barycentric coordinates of a node with respect to the unit K are positive, the node is located within the unit K, meaning that a host unit is found; if only one barycentric coordinate is negative, the direction of next search, beta, is revealediThe adjacent cell of the negative face i is the next search cell; if the two barycentric coordinates are negative, then randomly selecting a betaiThe adjacent cell of the negative face i is the next search cell as shown in fig. 2. The node performs one-time barycentric coordinate calculation on all the surfaces of the unit, so that the position relation between the node and the unit can be judged.
The S2 is an improvement of the S1 host cell search method, and for the host cell search of most interpolation points, the host cell can be found without using the improvement method. For a simple grid, an improvement method is not needed, and the improvement is mainly aimed at the condition that a node host unit cannot be determined by a common method. The prior art is adjacent unit search, various forms exist on the path judgment method of the adjacent unit search, and the adjacent unit search based on barycentric coordinates is also one of the characteristics of the text. The key of the host cell search in S1 is the calculation of barycentric coordinates, which are obtained from the cell plane information and the lattice point information, and the mathematical principle of determining a plane using 3 points is used for determining the plane. In any polyhedral mesh, some faces of the mesh cells contain mesh points more than 3, and theoretically, the mesh points of the same cell face are in the same plane, but due to data errors of mesh generation software or mesh output data errors, the mesh points are not in the same plane mathematically. If the node for which the host cell is being sought is near the grid surface, the result of barycentric coordinates is distorted due to slight distortion of the grid surface, so that the search direction points back and forth between two or more cells, and finally the host cell of the node cannot be determined.
In view of the above, the present invention provides for the dye marking of grid points within a plane, ensuring that at least one of the 3 grid points used to define the plane has not been previously selected. Taking the plane in fig. 3 as an example:
s1, determining the number of grid points contained in the surface of all grid units, and the global number and the local number of each grid point, wherein the surface in figure 3 contains 5 grid points, P0,P1,P2,P3,P4(ii) a Is the lattice point local number within the cell. P0,P1,P2,P3,P4Are all lattice points in one plane of the cell. Numbering here is not a sequential requirement. P is the out-of-plane lattice point.
S2, determining a plane from any 3 grid points in the plane, sorting the selected 3 grid points according to local numbers, and dyeing and marking the sorted 3 grid point combinations, for example, in fig. 3, where any selected grid point is P0,P4,P2Then is P after sorting according to local number0,P2,P4Judging whether the combination has a mark or not, if so, reselecting, and if not, marking the combination group (P)0,P2,P4) Color; the selected nodes form a combination, and are reselected only if the combination is repeated. Some node duplication within the combination is not a consideration.
S3, calculating barycentric coordinates for the combination of the 3 grid points selected in the S2;
and S4, judging the searching direction according to the barycentric coordinates calculated in S3, if the host cell can not be determined and one cell is frequently searched, returning to the 2 nd step of determining the plane lattice point combination by the cell.
Similar to S2, the S3 is an improvement of the search method for the S1 host cell, and for a grid cell with an abnormally complex individual grid cell form that may exist in a computational grid, it may not be possible to solve the problem by only performing in-plane grid point staining, and it is necessary to perform staining labeling on an in-plane grid point and an out-of-plane grid point at the same time. The concept of dyeing the mark is similar to S2, taking fig. 4 as an example, randomly selecting 3 grid points in the plane, for example, P0,P4,P2，P0,P4,P2Is a local number, which is only used in the process of calculating barycentric coordinates. Optionally one out-of-plane grid point, e.g. P5In FIG. 4, with respect to plane P0P1P2P3P4In other words, P5,P6,P7,P8,P9Are all out-of-plane lattice points; arranging the out-of-plane grid points at the first position, sequencing the in-plane grid points from small to large according to the local numbers of the unit grid points, arranging the in-plane grid points at the second position to the fourth position, judging whether the combination has a mark, reselecting if the combination has the mark, and dyeing and marking the combination if the combination does not have the mark, such as group (P)5,P0,P2,P4) Color. The rest of the operations in the host cell search process are similar to those in S2, including the operations of dyeing marks, barycentric coordinate calculation, search direction judgment, and grid point reselection combination. P5,P0,P2,P4Are local numbers, are ordered in the combination, and are ordered more easily and quickly when compared to the combination. No ordering is required in the calculation of the barycentric coordinates.
And S4, optimizing the search starting unit of the node which does not start to be searched according to the host unit information which is searched, and improving the searching efficiency. The host unit search will have a starting unit from which to start the search. The invention uses the node information of the searched host unit, takes the host unit of the node as the searching initial unit of the node connected with the node, takes the figure 5 as an example, the host unit of the found node P is K, and sets the initial units of all Q nodes connected with the P as K, thereby obviously shortening the searching path and improving the searching efficiency.
The present invention will be described in detail with reference to specific examples.
Aiming at a certain oil tanker ship type, carrying out self-propulsion numerical simulation of a ship body with a tail rudder, an energy-saving device and a propeller, wherein a rotating domain comprises the propeller and adopts a tetrahedral unstructured grid form; the static field includes the hull and other accessories, in the form of a cartesian hybrid grid incorporating a conformal prismatic layer by the cut-cell method. The flow field coupling calculation between the rotating domain and the static domain needs to use the host unit search technology of the invention to transfer the flow field information through node and host unit information interpolation.
Fig. 6 shows a computational grid view of an interface and a region near the interface, and it can be seen that grid unit forms on two sides of the interface are different, grid forms of an external computational domain are extremely complex, whether a surface grid of the interface or a body grid around the interface is near a propeller hub, various polyhedral grids are filled, and grid unit forms are diversified, which is a great challenge for searching a host unit.
In the ship type self-navigation numerical simulation process, a conventional host unit is firstly adopted for searching. Since the computational grid is too complex, when the host cell search proceeds to node 4552, there is a case where the host cell cannot be determined, and the search program points back and forth to the search among 3 cells with cell numbers 743134, 749640, 743138, as shown in fig. 7, presenting the possible problems mentioned earlier herein. This is unacceptable because the host cell of the node cannot be determined, which may result in the entire numerical simulation being unable to be performed.
Then, the invention is adopted to search the host unit of the same set of computational grids, as shown in figure 8. Because the invention carries out special optimization and transformation aiming at the host cell searching method of the complex cell form computational grid, in the host cell searching process, the host cell with the serial number of 743138 is successfully found at the 4552 node, and respective host cells are successfully found in the cell searching of the subsequent nodes.
In addition, the host unit search starting unit optimization operation of S4 is added, so that the search path is greatly shortened, most nodes can find the host unit only by 1-step search, and the search efficiency can be effectively improved.
TABLE 1 host cell search time
Method Time consuming(s)
Common search method 7.4
Searching method of the invention 1.9
In conclusion, the comparison of the search results shows that compared with the common host unit search method, the method has the capability of processing the host unit search in the complex unit form calculation grid, simultaneously has better search efficiency and has very strong practical value.
The host unit search flow chart of the invention is shown in FIG. 9:
S-A, starting;
S-B, Node ═ 1; node represents the Node currently carrying out host unit search;
S-C, starting from the starting unit;
S-D, calculating barycentric coordinates of each surface of the unit;
and S-E, judging whether the node is in the unit, if not, executing the step S-G, and if so, finishing the host unit search of the node. And updating the search starting unit of the node connected with the node by using the node host unit information. After updating the initial unit of the connected node, the current node host unit search is completed completely, and then step S-C is executed to start the host unit search of the next node;
and performing the next step;
S-F, judging whether the Node is Nnodes, if not, executing the step S-K, and if so, executing the step S-L; nnodes represents the total number of interface interpolation points needing to be searched by the host unit;
S-G, determining a next searching unit by the barycentric coordinates;
S-H, if the number of times N of repeated searching of the unit is more than 10 and less than or equal to 20, executing the step S-I, if N is more than 20, executing the step S-J, if N is less than or equal to 10, executing the step S-D,
S-I, dyeing and marking the inner grid of the face, and then executing the step S-D;
S-J, adopting a double-dyeing marking method to dye and mark the inner lattice points and the outer lattice points of the face, and then executing the step S-D;
S-K, Node +1, and then performing step S-C;
S-L, searching the host unit, and ending.
In summary, it is necessary to perform a host cell search on all interface interpolation points, where the search is performed in the volume grid cells in the region near the interface, and an adjacent cell search method based on barycentric coordinates is adopted. In the process of searching the host unit of the node, the problem of repeated searching of the same unit may occur, which indicates that the searching is complicated, and whether to use the dyeing mark for auxiliary searching and whether to use the double dyeing marks for auxiliary searching are judged according to the times of repeated searching of the same unit. And after the host unit search of the node is completed, taking the host unit of the node as a search starting unit of the node connected with the node.
While embodiments of the invention have been shown and described, it will be understood by those of ordinary skill in the art that: various changes, modifications, substitutions and alterations can be made to the embodiments without departing from the principles and spirit of the invention, the scope of which is defined by the claims and their equivalents.

Claims (8)

1. An efficient host cell search method suitable for multiple sets of complex grid coupling CFD calculation is characterized by comprising the following steps:
s1, determining the position relation of the nodes and the units by calculating the barycentric coordinates of the nodes and the units, and giving a searching direction; in the host unit searching process, if the unit is repeatedly searched, the repeated searching times of the unit is increased by one; judging whether the node is in the cell, if so, finding the host cell, otherwise, executing the step S2;
s2, judging whether the repetition times of the unit on the search path in the search process is less than m times, if so, executing a step S1, and if not, executing a step S3;
s3, judging whether the repeated searching times of the unit is more than n times, if so, executing the step S4, otherwise, adopting a dyeing marking method when calculating the gravity center coordinates of the node and the unit, and adding one to the repeated searching times of the unit every time of searching; judging whether the node is in the cell, if so, finding the host cell, otherwise, executing the step S3;
s4, for the grid unit with abnormal and complex unit form, a double-dyeing marking method is adopted when the gravity center coordinates of the nodes and the units are calculated, and the repeated searching times of the units are increased by one every time the units are searched; judging whether the node is in the cell, if so, finding the host cell, otherwise, executing the step S4;
and S5, optimizing the search starting unit of the node which does not start to search according to the host unit information which completes the search.
2. The method of claim 1, wherein the computing barycentric coordinates of nodes and cells comprises:
s1-1, taking the ith out-of-plane lattice point of the K unit as a vertex, and making three lattice points in the ith plane of the K unit(Vector)
Figure FDA0003329205060000011
Figure FDA0003329205060000012
Wherein
Figure FDA0003329205060000013
Representing a vector
Figure FDA0003329205060000014
A vector mixture product of;
s1-2, using the node as the vertex, and taking the three lattice points in the ith surface of the K unit as vectors
Figure FDA0003329205060000015
Then
Figure FDA0003329205060000016
S1-3, determining barycentric coordinates of the space relation between the nodes and the cells, and determining the search direction according to the barycentric coordinates, wherein the barycentric coordinates are calculated according to the following formula:
Figure FDA0003329205060000021
wherein K represents the current cell;
βithat is, the barycentric coordinates of the nodes and cells with respect to plane i;
AKrepresenting that the ith surface of the K unit and any grid point of the out-of-plane K unit form the volume of a geometric body;
Figure FDA0003329205060000022
representing the volume of the geometry formed by the node and the ith surface of the K unit.
3. The method of claim 1, wherein the staining labeling method in S3 comprises:
and optionally selecting 3 grid points from the grid points in the ith plane of the K unit to determine a grid unit plane, sequencing the selected 3 grid points according to local numbers, judging whether the combination has a mark or not, reselecting if the combination has the mark, and marking the combination if the combination does not have the mark.
4. The method as claimed in claim 3, wherein at least one of the 3 grid points is previously unused.
5. The method of claim 1, wherein the abnormal complex grid cells in S4 include:
the cells having shapes other than tetrahedrons, hexahedrons, triangular prisms, rectangular pyramids, or a grid cell having more than 10 faces, or more than 5 grid points within a cell boundary.
6. The method of claim 1, wherein the double-staining labeling method in S4 comprises:
and (3) optionally selecting 3 grid points from the in-plane grid points to determine a grid unit plane, optionally selecting one out-of-plane grid point, arranging the out-of-plane grid point at the first position, arranging the in-plane grid points at the second position to the fourth position after sorting the in-plane grid points according to the local numbers, judging whether the combination has a mark or not, reselecting if the combination exists, and dyeing and marking the combination if the combination does not exist.
7. The method of claim 6, wherein the method for efficient host cell search for multiple sets of complex mesh coupled CFD calculations comprises: at least one of the 3 grid points is previously unused or the out-of-plane grid point is previously unselected.
8. The method of claim 1, wherein the S5 includes:
and using the node information of which the host unit search is completed as a search starting unit of the node connected with the node.
CN202111272625.XA 2021-10-29 2021-10-29 Efficient host unit searching method suitable for multiple sets of complex grid coupling CFD (computational fluid dynamics) calculation Active CN114004174B (en)

Priority Applications (1)

Application Number Priority Date Filing Date Title
CN202111272625.XA CN114004174B (en) 2021-10-29 2021-10-29 Efficient host unit searching method suitable for multiple sets of complex grid coupling CFD (computational fluid dynamics) calculation

Applications Claiming Priority (1)

Application Number Priority Date Filing Date Title
CN202111272625.XA CN114004174B (en) 2021-10-29 2021-10-29 Efficient host unit searching method suitable for multiple sets of complex grid coupling CFD (computational fluid dynamics) calculation

Publications (2)

Publication Number Publication Date
CN114004174A true CN114004174A (en) 2022-02-01
CN114004174B CN114004174B (en) 2023-08-25

Family

ID=79925265

Family Applications (1)

Application Number Title Priority Date Filing Date
CN202111272625.XA Active CN114004174B (en) 2021-10-29 2021-10-29 Efficient host unit searching method suitable for multiple sets of complex grid coupling CFD (computational fluid dynamics) calculation

Country Status (1)

Country Link
CN (1) CN114004174B (en)

Citations (8)

* Cited by examiner, † Cited by third party
Publication number Priority date Publication date Assignee Title
US5245550A (en) * 1991-01-25 1993-09-14 Hitachi, Ltd. Apparatus for wire routing of VLSI
JP2002073130A (en) * 2000-06-13 2002-03-12 Yaskawa Electric Corp Planning method for gross motion path of robot and its controller
CN101120545A (en) * 2005-02-14 2008-02-06 艾利森电话股份有限公司 Method and nodes for handling multicast messages
WO2013160467A1 (en) * 2012-04-27 2013-10-31 Ugolin Nicolas Gilbert Gtl process and reactor employing a mobile phase and plasma
CN104408773A (en) * 2014-12-02 2015-03-11 北京大学工学院南京研究院 Method for interpolating structured grid non-matching interface
CN105631068A (en) * 2014-10-31 2016-06-01 北京临近空间飞行器系统工程研究所 Grid boundary condition processing method of unstructured grid CFD computation
CN112100937A (en) * 2020-09-11 2020-12-18 南京航空航天大学 Efficient overlapped grid parallel assembly method based on wall surface distance
CN112364442A (en) * 2020-12-22 2021-02-12 北京机电工程研究所 Bidirectional intelligent selection fluid-solid coupling analysis method

Patent Citations (8)

* Cited by examiner, † Cited by third party
Publication number Priority date Publication date Assignee Title
US5245550A (en) * 1991-01-25 1993-09-14 Hitachi, Ltd. Apparatus for wire routing of VLSI
JP2002073130A (en) * 2000-06-13 2002-03-12 Yaskawa Electric Corp Planning method for gross motion path of robot and its controller
CN101120545A (en) * 2005-02-14 2008-02-06 艾利森电话股份有限公司 Method and nodes for handling multicast messages
WO2013160467A1 (en) * 2012-04-27 2013-10-31 Ugolin Nicolas Gilbert Gtl process and reactor employing a mobile phase and plasma
CN105631068A (en) * 2014-10-31 2016-06-01 北京临近空间飞行器系统工程研究所 Grid boundary condition processing method of unstructured grid CFD computation
CN104408773A (en) * 2014-12-02 2015-03-11 北京大学工学院南京研究院 Method for interpolating structured grid non-matching interface
CN112100937A (en) * 2020-09-11 2020-12-18 南京航空航天大学 Efficient overlapped grid parallel assembly method based on wall surface distance
CN112364442A (en) * 2020-12-22 2021-02-12 北京机电工程研究所 Bidirectional intelligent selection fluid-solid coupling analysis method

Non-Patent Citations (4)

* Cited by examiner, † Cited by third party
Title
YAN WU等: "A directed search strategy for evolutionary dynamic multiobjective optimization:", SOFT COMPUTING, vol. 19, pages 3221 - 3235, XP035535841, DOI: 10.1007/s00500-014-1477-4 *
余培汛;杨海;潘凯;韩啸;: "结合并行策略与距离减缩法的搜索算法及其在CAA中的应用", 计算力学学报, vol. 35, no. 02, pages 255 - 260 *
金奕星: "基于多参考系的螺旋桨敞水CFD求解器开发研究", 第十一届全国流体力学学术会议, pages 369 *
陈乐乐等: "软式加油过程中软管鞭打现象数值模拟研究", 南京航空航天大学空气动力学系, vol. 46, no. 6, pages 56 - 59 *

Also Published As

Publication number Publication date
CN114004174B (en) 2023-08-25

Similar Documents

Publication Publication Date Title
CN108459909B (en) Multi-body separation grid overlapping method and system suitable for parallel processing
CN104484508B (en) Optimizing method for noncontact three-dimensional matching detection of complex curved-surface part
US20100063784A1 (en) System and method for fitting feature elements using a point-cloud of an object
CN105654483A (en) Three-dimensional point cloud full-automatic registration method
CN104036102B (en) Calculation method and device for product assembly deviation
CN109741374A (en) Point cloud registering rotation transformation methods, point cloud registration method, equipment and readable storage medium storing program for executing
CN110489778A (en) Pattern dividing method, laser ablation control system towards laser ablation processing
CN112439601B (en) Spraying robot automatic trajectory planning method for outer vertical surface of large ship
CN112613126B (en) Wall distance calculation method and device applied to appearance of aircraft
CN103838907A (en) Curved surface cutting trajectory obtaining method based on STL model
CN104809760A (en) Depth-first strategy based method for automatically constructing geo-spatial three-dimensional outline
CN105785919B (en) Point five axis of cloud is without global interference generating tool axis vector feasible zone computational methods
CN107665496B (en) Three-dimensional attitude registration method
CN104504758B (en) Artificial tooth is preced with surface curve generation method
Wang et al. Accelerating advancing layer viscous mesh generation for 3D complex configurations
CN114004174B (en) Efficient host unit searching method suitable for multiple sets of complex grid coupling CFD (computational fluid dynamics) calculation
Cheng et al. Determining the topology of real algebraic surfaces
Livesu et al. Loopy cuts: Surface-field aware block decomposition for Hex-meshing
Huang et al. A matching algorithm between precursory 3D process model and 2D working procedure drawing based on subgraph isomorphism
CN114186301A (en) Part nesting data determining method, device, equipment and storage medium
CN110348163B (en) Fitting method for continuous tangent circular arc of tire pattern boundary
CN107507279A (en) Triangle network generating method based on quick Convex Hull Technology
Dahu et al. A fast surface reconstruction algorithm based on Delaunay
CN112184911B (en) Non-popular grid model slicing processing method
CN108062433A (en) The gradient curved surface layered approach of point cloud model is remanufactured based on increasing material

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
CB02 Change of applicant information

Address after: Building T1, Xuelang Town, No. 99 Qingshu Road, Wuxi Economic Development Zone, Jiangsu Province, 214100

Applicant after: China Shipbuilding Orlando Wuxi Software Technology Co.,Ltd.

Address before: 214021 Xintiandi leisure Plaza, Wuxi, Jiangsu 1-403

Applicant before: CSIC Orlando Wuxi Software Technology Co.,Ltd.

CB02 Change of applicant information
GR01 Patent grant
GR01 Patent grant