CN107391950A - A kind of turbulent flow wall is apart from quick calculation method - Google Patents
A kind of turbulent flow wall is apart from quick calculation method Download PDFInfo
- Publication number
- CN107391950A CN107391950A CN201710655506.XA CN201710655506A CN107391950A CN 107391950 A CN107391950 A CN 107391950A CN 201710655506 A CN201710655506 A CN 201710655506A CN 107391950 A CN107391950 A CN 107391950A
- Authority
- CN
- China
- Prior art keywords
- adt
- object plane
- trees
- layer
- plane unit
- 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.)
- Pending
Links
Classifications
-
- G—PHYSICS
- G16—INFORMATION AND COMMUNICATION TECHNOLOGY [ICT] SPECIALLY ADAPTED FOR SPECIFIC APPLICATION FIELDS
- G16Z—INFORMATION AND COMMUNICATION TECHNOLOGY [ICT] SPECIALLY ADAPTED FOR SPECIFIC APPLICATION FIELDS, NOT OTHERWISE PROVIDED FOR
- G16Z99/00—Subject matter not provided for in other main groups of this subclass
Landscapes
- Management, Administration, Business Operations System, And Electronic Commerce (AREA)
Abstract
The invention discloses a kind of turbulent flow wall apart from quick calculation method, comprise the following steps：Parallel architecture is established, the grids of whole computational fields is subjected to subregion, the corresponding process of each blockette, and each process is had the backup of a global object plane unit；Establish ADT trees by different level；The N number of object plane blockette nearest from spatial point is found by first layer ADT trees, then finds the M object plane unit nearest from spatial point by second layer ADT trees in this N number of blockette, M object plane unit is finally traveled through and nearest wall distance is calculated.The positive effect of the present invention is：Even if grid amount increases, it can also ensure that the grid amount of each process is unlikely to very big by way of increasing into number of passes；Can realize it is quick search, i.e. the object plane unit according to the cartesian coordinate fast positioning of spatial point near it；A large amount of units are just filtered in first layer search procedure, so as to accelerate second layer ADT to search.
Description
Technical field
The present invention relates to a kind of turbulent flow wall apart from quick calculation method.
Background technology
In the industrial circles such as aerospace flight vehicle design, fluid machinery, bullet train/automobile, wind energy and Wind Engineering, stream
Mechanics are the trunk subjects for carrying out product design.Fluid Mechanics Computation (CFD) is hydrodynamics, computational mathematics, computer
Cross discipline, emulated using computer convection body kinetics equation, so as to obtain spatial flow, for product design provide according to
According to.With the development of computer, CFD is just playing an increasingly important role in the design process of industrial products.The stream of fluid
Dynamic state is divided into laminar flow, the class of turbulent flow two, and the Real Flow Field in nature is essentially all turbulent flow.At present, CFD technical modellings
During turbulent flow, typically all emulated using turbulence model.Conventional turbulence model (such as SA, SST) is required for calculating wall
Identity distance from.Here, " wall distance " is defined as in space any point to the beeline of object plane.
Traditional wall distance calculating method mainly has two kinds：Direct traversal, box method.Direct traversal, Gu Mingsi
Justice, it is exactly each face unit (such as triangle, quadrangle) on direct traversal object plane, then calculates spatial point to the face unit
Distance.Box method, it is that object plane is divided into by several boxes according to the cartesian coordinate scope of object plane, then puts face unit
In the box where it, calculate wall apart from when, first find the several boxes nearest from spatial point, then travel through these boxes again
Object plane unit in son, and then calculate beeline.
In in general engineer applied, the computational efficiency of box method is more much higher than direct traversal.But as problem is advised
The increase (increase for being presented as computing unit amount) of mould, the computational efficiency of box method also decline rapidly, and main cause is：For big
Scale turbulence problem of modelling, if the acquirement of object plane box number is very few, then the face unit number in each box is a lot, substantially connects
Direct traversal is bordering on, so as to which efficiency declines to a great extent；If box number is excessive, due to being needed during nearest box is calculated
Box is ranked up according to distance, and after box number is significantly increased, the sequence of box causes time overhead to be significantly increased,
Efficiency can be caused to decline to a great extent.
The content of the invention
In order to overcome the disadvantages mentioned above of prior art, the present invention proposes a kind of turbulent flow wall apart from quick calculation method.
The technical solution adopted for the present invention to solve the technical problems is：A kind of turbulent flow wall apart from quick calculation method,
Comprise the following steps：
Step 1: establishing parallel architecture, the grid of whole computational fields is subjected to subregion, corresponding one of each blockette enters
Journey, and each process is had the backup of a global object plane unit；
Step 2: establish ADT trees, including first layer ADT trees and second layer ADT trees by different level；
Step 3: calculating wall distance, the N number of object plane nearest from spatial point point is found by first layer ADT trees first
Block, then finds the M object plane unit nearest from spatial point in this N number of blockette by second layer ADT trees, last time
Go through M object plane unit and nearest wall distance is calculated..
Compared with prior art, the positive effect of the present invention is：
1st,, can also be increasing into by way of number of passes even if grid amount increases using the parallel architecture of grid block subregion
Ensure that the grid amount of each process is unlikely to very big；
2nd, by ADT data structures, quick lookup can be realized, i.e., its is attached according to the cartesian coordinate fast positioning of spatial point
Near object plane unit；
3rd, by way of being layered ADT trees, a large amount of units have just been filtered in first layer search procedure, so as to accelerate second
Layer ADT is searched.
Brief description of the drawings
Examples of the present invention will be described by way of reference to the accompanying drawings, wherein：
Fig. 1 is flow chart of the turbulent flow wall apart from quick calculation method of the present invention.
Embodiment
A kind of turbulent flow wall is apart from quick calculation method, as shown in figure 1, comprising the following steps：
Step 1, parallel architecture is established, the parallel architecture of establishing comprises the following steps：
Step 101, the subregion software METIS that increases income is called, be n blockette Z={ Z by the mesh generation of computational fields1、Z2、
Z3······Zn, the corresponding process P={ P of each block1、P2、P3······Pn}；Here blockette
Number n are determined by initial mesh scale and calculator memory, typically make the number of meshes after scoring area in each subregion less than 100
Ten thousand.
Step 102, by each blockette ZiOn object plane unit send to corresponding server processes Pi；
All object plane cell S={ s that step 103, server processes will be collected into1、s2、s3······smGroup
Fill as whole object plane S；Here, m≤n.
Step 104, server processes will include all object plane cell S={ s1、s2、s3······smIt is whole
Object plane is sent to each process PiSo that each process PiThere is the backup of a global object plane cell S；
Step 2, ADT trees by different level are established, ADT trees comprise the following steps by different level for the foundation：
Step 201, the root node for establishing ADT data：In each process PiOn, according to the cartesian coordinate model of whole object plane
Enclose and establish the root node that hypercube forms ADT data；
Step 202, establish first layer ADT trees：By each process PiOn object plane subregion according to partition boundaries Descartes
Coordinate range establishes ADT data；
Step 203, establish second layer ADT trees：By each object plane unit siAccording to object plane unit siCoordinate insert one by one
Node is formed in the ADT data established into step 202；
Step 3, wall distance is calculated, the calculating wall distance comprises the following steps：
Step 301, the N number of object plane blockette Z nearest from spatial point is found by first layer ADT treessub={ z1、
z2···zN, wherein N ＜ n, Zsub∈Z。
It is quickly found out in step 302, the N number of object plane blockette found in step 301 by second layer ADT trees from sky
Between put M nearest object plane cell Ssub={ s1、s2···sM, wherein M ＜ m, Ssub∈S。
M object plane cell S in step 303, traversal step 302sub={ s1、s2···sMAnd be calculated nearest
Wall distance d.
Following table is the present embodiment and the efficiency comparison of conventional method：
| Time consumption | |
| Direct traversal | About 3 days |
| Box method is searched | About 4 hours |
| The inventive method | About 15 minutes |
Claims (4)
1. a kind of turbulent flow wall is apart from quick calculation method, it is characterised in that：Comprise the following steps：
Step 1: establishing parallel architecture, the grid of whole computational fields is subjected to subregion, each blockette corresponds to a process, and
Each process is set to have the backup of a global object plane unit；
Step 2: establish ADT trees, including first layer ADT trees and second layer ADT trees by different level；
Step 3: calculating wall distance, the N number of object plane blockette nearest from spatial point is found by first layer ADT trees first,
Then the M object plane unit nearest from spatial point is found by second layer ADT trees in this N number of blockette, finally travels through M
Simultaneously nearest wall distance is calculated in object plane unit.
2. a kind of turbulent flow wall according to claim 1 is apart from quick calculation method, it is characterised in that：It is right described in step 1
Number of meshes after grid progress subregion in each subregion is less than 1,000,000.
3. a kind of turbulent flow wall according to claim 1 is apart from quick calculation method, it is characterised in that：It is complete described in step 1
The object plane unit on corresponding blockette that office's object plane unit is collected into by each server processes forms.
4. a kind of turbulent flow wall according to claim 1 is apart from quick calculation method, it is characterised in that：Built described in step 2
The method of vertical ADT trees by different level is：
(1) root node of ADT data is established：In each process, established according to the cartesian coordinate scope of whole object plane super vertical
Cube forms the root node of ADT data；
(2) first layer ADT trees are established：Object plane subregion in each process is established according to the cartesian coordinate scope of partition boundaries
ADT data；
(3) second layer ADT trees are established：Each object plane unit is inserted into (2) step one by one according to the coordinate of the object plane unit to build
Node is formed in vertical ADT data.
Priority Applications (1)
| Application Number | Priority Date | Filing Date | Title |
|---|---|---|---|
| CN201710655506.XA CN107391950A (en) | 2017-08-03 | 2017-08-03 | A kind of turbulent flow wall is apart from quick calculation method |
Applications Claiming Priority (1)
| Application Number | Priority Date | Filing Date | Title |
|---|---|---|---|
| CN201710655506.XA CN107391950A (en) | 2017-08-03 | 2017-08-03 | A kind of turbulent flow wall is apart from quick calculation method |
Publications (1)
| Publication Number | Publication Date |
|---|---|
| CN107391950A true CN107391950A (en) | 2017-11-24 |
Family
ID=60343530
Family Applications (1)
| Application Number | Title | Priority Date | Filing Date |
|---|---|---|---|
| CN201710655506.XA Pending CN107391950A (en) | 2017-08-03 | 2017-08-03 | A kind of turbulent flow wall is apart from quick calculation method |
Country Status (1)
| Country | Link |
|---|---|
| CN (1) | CN107391950A (en) |
Cited By (4)
| Publication number | Priority date | Publication date | Assignee | Title |
|---|---|---|---|---|
| CN114722745A (en) * | 2022-06-10 | 2022-07-08 | 中国空气动力研究与发展中心计算空气动力研究所 | Turbulent flow wall surface distance calculation method and device, computer equipment and storage medium |
| CN114756974A (en) * | 2022-06-13 | 2022-07-15 | 中国空气动力研究与发展中心计算空气动力研究所 | Wall distance calculation method considering object plane normal information |
| CN114943167A (en) * | 2022-05-31 | 2022-08-26 | 中国空气动力研究与发展中心计算空气动力研究所 | Method, system, medium and equipment for calculating wall surface distance of structural grid |
| CN115357849A (en) * | 2022-10-24 | 2022-11-18 | 中国空气动力研究与发展中心计算空气动力研究所 | Method and device for calculating wall surface distance under Cartesian grid |
-
2017
- 2017-08-03 CN CN201710655506.XA patent/CN107391950A/en active Pending
Cited By (5)
| Publication number | Priority date | Publication date | Assignee | Title |
|---|---|---|---|---|
| CN114943167A (en) * | 2022-05-31 | 2022-08-26 | 中国空气动力研究与发展中心计算空气动力研究所 | Method, system, medium and equipment for calculating wall surface distance of structural grid |
| CN114722745A (en) * | 2022-06-10 | 2022-07-08 | 中国空气动力研究与发展中心计算空气动力研究所 | Turbulent flow wall surface distance calculation method and device, computer equipment and storage medium |
| CN114756974A (en) * | 2022-06-13 | 2022-07-15 | 中国空气动力研究与发展中心计算空气动力研究所 | Wall distance calculation method considering object plane normal information |
| CN114756974B (en) * | 2022-06-13 | 2022-09-02 | 中国空气动力研究与发展中心计算空气动力研究所 | Wall surface distance calculation method considering object surface normal information |
| CN115357849A (en) * | 2022-10-24 | 2022-11-18 | 中国空气动力研究与发展中心计算空气动力研究所 | Method and device for calculating wall surface distance under Cartesian grid |
Similar Documents
| Publication | Publication Date | Title |
|---|---|---|
| CN107391950A (en) | A kind of turbulent flow wall is apart from quick calculation method | |
| CN103970960B (en) | The element-free Galerkin structural topological optimization method accelerated parallel based on GPU | |
| Zheng et al. | Fault diagnosis system of bridge crane equipment based on fault tree and Bayesian network | |
| CN106407719A (en) | Optimization method for rapid convergent robot dynamic parameter identification trajectory | |
| CN114611437B (en) | Method and device for establishing aircraft pneumatic model database based on CFD technology | |
| CN106503365B (en) | A kind of sector search method for SPH algorithm | |
| He et al. | Aerodynamic optimisation of a high-speed train head shape using an advanced hybrid surrogate-based nonlinear model representation method | |
| CN110110475B (en) | Extended Kalman filtering method based on online learning fading factors | |
| Chang et al. | A parallel implicit hole-cutting method based on background mesh for unstructured Chimera grid | |
| CN101261651A (en) | Main beam section pneumatic self excitation force whole-course numerical modeling method | |
| CN114756974A (en) | Wall distance calculation method considering object plane normal information | |
| CN105224726B (en) | The method that structured grid Dynamic mesh is used for unstrctured grid flow field calculation device | |
| CN103984832B (en) | A kind of aluminium cell Electric Field Simulation analysis method | |
| CN112035995A (en) | Nonstructural grid tidal current numerical simulation method based on GPU (graphics processing Unit) computing technology | |
| CN117394313A (en) | Power system transient stability evaluation method, system, chip and equipment | |
| CN112380786A (en) | Air vane thermal environment modeling method, equipment and storage medium | |
| CN107391871A (en) | A kind of space lattice deformation method based on parallelization RBF | |
| CN115859780A (en) | Power equipment vibration simulation method and system based on physical constraint neural network | |
| CN115257692A (en) | Visual traffic information-based PHEV energy management method and system | |
| CN114638048A (en) | Three-dimensional spray pipe flow field rapid prediction and sensitivity parameter analysis method and device | |
| He | Design of an actively controlled aerodynamic wing to increase high-speed vehicle safety | |
| CN109447131B (en) | Similar high-dimensional target information identification method and system | |
| Al-Jarro et al. | DeepSim-HiPAC: Deep learning high performance approximate calculation for interactive design and prototyping | |
| CN111754421A (en) | Improved guide filtering three-dimensional scattered point cloud quick fairing method | |
| Chen et al. | Computer-aided front and rear wings aerodynamic design of a formula SAE racing car |
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 | ||
| RJ01 | Rejection of invention patent application after publication | ||
| RJ01 | Rejection of invention patent application after publication |
Application publication date: 20171124 |