WO2017084105A1 - 一种数值模拟等离子体放电的系统及方法 - Google Patents
一种数值模拟等离子体放电的系统及方法 Download PDFInfo
- Publication number
- WO2017084105A1 WO2017084105A1 PCT/CN2015/095219 CN2015095219W WO2017084105A1 WO 2017084105 A1 WO2017084105 A1 WO 2017084105A1 CN 2015095219 W CN2015095219 W CN 2015095219W WO 2017084105 A1 WO2017084105 A1 WO 2017084105A1
- Authority
- WO
- WIPO (PCT)
- Prior art keywords
- model
- mesh
- grid
- equation
- result
- Prior art date
Links
Images
Classifications
-
- G—PHYSICS
- G06—COMPUTING; CALCULATING OR COUNTING
- G06F—ELECTRIC DIGITAL DATA PROCESSING
- G06F30/00—Computer-aided design [CAD]
Definitions
- the invention relates to a numerical calculation simulation system and a method, in particular to a calculation simulation and design of a fluid field and a thermal field coupling involving a fluid, such as a numerical simulation of a plasma discharge process.
- Computational Fluid Dynamics can handle more and more complex problems.
- Computational fluid dynamics, theoretical fluid mechanics, and experimental fluid mechanics are the three main methods of fluid mechanics research work, complement each other.
- Theoretical analysis provides the basis for experimental and computational research; experiments provide data for numerical studies and verify calculation results; numerical calculations are experiments in a special sense. Numerical calculations have greater freedom and flexibility to perform experiments where "physical experiments" are impossible or difficult.
- Computational fluid dynamics is the use of numerical methods to solve the governing equations of fluid mechanics in a computer to predict the flow of the flow field.
- computational fluid dynamics mainly involves fluid non-viscous flow and viscous flow.
- Non-viscous flow includes low velocity flow, transonic flow, supersonic flow, etc.
- viscous flow includes turbulent flow, boundary layer flow, and the like.
- fluid mechanics covers all aspects of the basic industry. In addition to aerospace, weather forecasting and oil and gas exploration, it also includes aerodynamic optimization of automotive models, analysis and reduction of airborne noise sources for fans and moving objects, thermal convection and heat. The influence of transmission on the device and the environment, the flow field and thermal field of the neutral gas in the process chamber, the medium-pressure discharge plasma process, etc. With the birth of new industries, the application of fluid mechanics also extends to micro-flow and multi-phase flow. , non-Newtonian flow and other complex fluids.
- the numerical calculation simulation system includes a computer readable storage medium, the storage medium storing an executable module, the storage The medium includes a data receiving module, the data receiving module is capable of receiving data, and a grid processing module capable of performing meshing processing on the received data to generate a grid processing result; and calculating a solving module,
- the calculation solution module can find the result according to the grid processing Solving the calculation equation to generate the calculation result of the flow field and the thermal field; and the result analysis module, the result analysis module can further analyze and generate the analysis result of the generated flow field and the thermal field calculation result.
- a processor capable of executing the executable module of the computer readable storage medium storage.
- the numerical calculation simulation system further includes a database capable of storing the received data, the grid processing result, the calculation result, and the analysis result.
- the grid processing module of the numerical calculation simulation system further includes a mesh generation unit, a mesh division unit, a node attribute marking unit, a wall surface distance calculation unit, and a mesh optimization unit.
- the computational solution module of the numerical calculation simulation system further includes a collection unit, an initialization unit, a solver unit, a choke unit, a boundary condition unit, and a result output unit.
- the mesh generated by the mesh generation unit of the numerical calculation simulation system includes a surface mesh, a tetrahedral mesh, a hexahedral mesh, a prismatic mesh (boundary layer mesh), and a tetrahedron.
- a surface mesh a tetrahedral mesh, a hexahedral mesh, a prismatic mesh (boundary layer mesh), and a tetrahedron.
- the ball fill method grid of the numerical calculation simulation system is based on the Delaunay split insertion technique.
- the node attributes of the node attribute marking unit of the numerical computing simulation system include a fluid node, a solid node, and a boundary node.
- the turbulence unit solving calculation model of the numerical calculation simulation system includes a zero equation model, an equation model, a Spalart-Allmaras model, a k-epsilon model, a k-omega model, a RNG k-epsilon model, Realizable k-epsilon model, RSM model, ASM model, SGS model, BGK model, MRT-LBM model, SRT-LBM model, lattice Boltzmann model, incompressible lattice Boltzmann model, thermal lattice Boltzmann Model, a lattice Boltzmann model of a non-uniform grid.
- the result of the result analysis module of the numerical calculation simulation system further includes the speed, the pressure, the density, the temperature, the macroscopic physical quantity of the overall flow field, the macroscopic physical quantity of the specified section, the aerodynamic force, the aerodynamic moment, Computational domain geometric model and mesh analysis, vector graphics (such as speed Vector lines), contour maps, filled contour maps (cloud maps), XY scatter plots, particle trace plots, simulated flow effects, image processing functions.
- the numerical calculation simulation method includes receiving data; performing meshing processing according to the data; The sub-processed grid solves the flow field and thermal field calculation results of the control equation; further extracts the calculation results of the flow field and the thermal field for analysis.
- the received data of the numerical calculation method includes a mesh structure, a material property, a gas characteristic parameter, a geometric file, a control parameter, and a mesh file.
- control parameters of the numerical calculation method include a mesh type, a governing equation, a model, a calculation precision, and a residual convergence precision.
- the mesh type of the numerical calculation method includes a body-fitted mesh, a partitioned mesh, a Cartesian mesh, an adaptive right-angle mesh, a multi-grid, and a ball filling method mesh combined with a fast Delaunay. , structured grids, unstructured grids, hybrid grids.
- the numerical calculation method model includes a zero equation model, an equation model, a Spalart-Allmaras model, a k-epsilon model, a k-omega model, a RNG k-epsilon model, a Realizable k-epsilon model, RSM model, ASM model, SGS model, BGK model, MRT-LBM model, SRT-LBM model, lattice Boltzmann model, incompressible lattice Boltzmann model, thermal lattice Boltzmann model, non-uniform grid Lattice Boltzmann model.
- the received data of the numerical calculation method includes user input data and non-user input data
- the non-user input data source includes a server and a communication terminal.
- Figure 1 is a schematic diagram of an exemplary system configuration of a numerical computing simulation system.
- Figure 2 shows a block diagram of a numerical calculation simulation system.
- Figure 3 shows a numerical simulation simulation flow chart.
- Figure 4 shows the structure of the grid processing module.
- Figure 6 shows a flow chart of a computational solution for a numerical simulation system.
- Figure 7-a shows the results of a numerical calculation simulation system for solving the problem of a flow around a cylinder with a Reynolds number of 20.
- Figure 7-b shows the results of a numerical calculation simulation system that solves the problem of a flow around a cylinder with a Reynolds number of 100.
- Figure 8-b shows a streamlined flow diagram of the discharge of working gas argon at a flow rate of 2000 sccm.
- Figure 8-c shows a streamlined flow diagram of the discharge of working gas argon at a flow rate of 3000 sccm.
- FIG. 9 is a schematic diagram showing the structure of a device for implementing a numerical calculation simulation system and method.
- Plasma generation methods include, but are not limited to, glow discharge, corona discharge, dielectric barrier discharge, radio frequency discharge, arc discharge, jet discharge, atmospheric pressure glow discharge, sub-atmospheric glow discharge, DC arc discharge, AC power frequency discharge, One or more combinations of high frequency induction discharge, low pressure discharge, combustion method, and the like.
- Application scenarios of different embodiments of the present invention may include, but are not limited to, simulations of interactions between vehicles and vehicles such as automobiles, trains, high-speed rails, ships, aircrafts, etc., and may include, but are not limited to, laser, electromagnetic, radio frequency, wave optics, ray optics.
- One or more physical field couplings such as MEMS, semiconductor, heat transfer, structural mechanics, acoustics, microfluidics, molecular flow, pipeline flow, etc., may also include but are not limited to multiphase flow, porous medium flow, suspended particle flow, reaction A combination of one or more fields, such as flow, magnetohydrodynamics, and biomechanics.
- the above description of the applicable fields is merely a specific example and should not be considered as the only feasible implementation.
- the communication between the client system 103 and the numerical computing simulation system 101 includes but is not limited to one-to-one, one-to-many, many-to-one, many-to-many, and the like.
- the client system 103 can run a web browser that can interact with a network connection system running on the numerical computing simulation system 101.
- the client system 103 can be used to receive data entered by the user, or can be displayed to the user in accordance with an exemplary user interface.
- the displayed content can be an interactive interface or a non-interactive interface.
- the interactive interface may be a display of various conditions input to the internal processing of the numerical calculation simulation system 101, or may be a result of displaying the analysis.
- Client system 103 can be concurrently directed to include, but is not limited to, one or more users. For example, different client systems 103 simultaneously display the results of the same numerical calculation simulation system 101 calculation analysis. As another example, the same client system 103 displays different numerical calculation simulation systems 101 to calculate the results of the analysis. As another example, one client system 103 enters conditions that need to be processed, and another client system 103 displays the results of the numerical calculation simulation system 101 to calculate the analysis.
- Figure 2 shows a block diagram of a numerical calculation simulation system.
- the numerical calculation simulation system 101 can include, but is not limited to, one or more data receiving modules 201, one or more grid processing modules 202, one or more computational solving modules 203, one or more results analysis modules 204, one or more System database 205. Some or all of the modules of the numerical computing simulation system 101 may be coupled to the network 102. The modules of the numerical calculation simulation system 101 may be centralized or distributed. One or more modules of the numerical computing simulation system 101 can be local or remote.
- the data receiving module 201 can be configured to receive a data file.
- the data file may include, but is not limited to, one or more combinations of mesh structures, material properties, gas property parameters, geometry files, control parameters, processed mesh files, and the like.
- the manner in which the geometry files are obtained may be direct (eg, directly fetching data from one or more client systems 103 over the network 102) or indirectly (eg, through the grid processing module 202, the computational solution module 203, the results analysis module 204). , the system database 205 to obtain).
- the grid processing module 202 can be used for grid processing of geometric files as well as for grid optimization. Grid processing includes, but is not limited to, one or more combinations of body-fitted meshes, partitioned meshes, adaptive right-angled meshes, multiple meshes, and ball filling methods in conjunction with fast Delaunay.
- the basic idea of multigrid is to iterate a number of steps on the coarse mesh to get a more accurate result, and then use this result as the initial value to perform iteration on the fine mesh.
- Ball filling method combined fast
- the specific implementation of Delaunay insertion to form a three-dimensional unstructured mesh may be a ball-packing arrangement in a three-dimensional geometric region, and adaptively distributing the appropriate nodes according to the geometric features and spatial relationships of the geometric model. In order to obtain the ideal node distribution, and then use the fast Delaunay insertion technology to efficiently generate the unstructured grid. If the grid size exceeds the computing power of the single machine, you can also consider the grid generation method using parallel technology.
- the splitting algorithm divides the whole structure into several sub-areas, then independently performs meshing in each sub-area, and finally completes the sub-area boundary mesh repair to obtain the mesh of the overall structure.
- the optimization of the mesh can be used for one or more combinations including, but not limited to, a body-fitted mesh, a partitioned mesh, an adaptive right-angle mesh, a multi-grid, a quality optimization of a structured three-dimensional mesh, and the like.
- Grid optimization includes, but is not limited to, one or more combinations of geometric optimization (smoothing), topology optimization, and the like.
- Geometric optimization can move the node position and keep the node connection relationship unchanged.
- Topology optimization optimizes the connection between nodes.
- the mesh optimization uses a combination of geometric optimization and topology optimization.
- the analysis of macroscopic physical quantities can include, but is not limited to, one or more combinations of speed, pressure, density, temperature, aerodynamic torque, and the like.
- the system database 205 is primarily used to store data received from the client system 103 and various data generated in the operation of the numerical computing simulation system 101. System database 205 can be local or remote. The connection or communication between the system database and other modules of the system can be wired or wireless.
- the data receiving module 201 can be configured to receive a data file.
- the data file may include, but is not limited to, one or more combinations of mesh structures, material properties, gas property parameters, geometry files, control parameters, processed mesh files, and the like.
- the manner in which the geometry files are obtained may be direct (eg, directly fetching data from one or more client systems 103 over the network 102) or indirectly (eg, through a grid processing module) 202.
- the calculation solution module 203, the result analysis module 204, and the system database 205 are obtained.
- the data receiving module 201 can receive the geometric file, and the geometric file can be unprocessed or processed after processing.
- the data receiving module 201 can receive the processed grid file.
- the grid processing module 202 can transmit the data to the results analysis module 204.
- the calculation solution module 203 can receive the request sent by the grid processing module 202, and can also access the system database 205 according to the request to obtain the required data. After the required data is acquired, the calculation solution module 203 can transmit the data to the grid processing module 204.
- the calculation solution module 203 can receive the request sent by the result analysis module 204, and can also access the system database 205 according to the request to obtain the required data. After the required data is acquired, the calculation solution module 203 can send the data transmission result to the analysis module 204.
- the result analysis module 204 can receive the request sent by the grid processing module 202, and can also access the system database 205 according to the request to obtain the required data.
- the grid processing module 202 can be used for mesh processing of geometric files and/or to confirm the meshing as required, or to optimize the mesh structure.
- Grid processing includes, but is not limited to, body-fitted mesh, partitioned mesh, Cartesian mesh, adaptive right-angle mesh, multi-grid, ball filling method combined with fast Delaunay, structured mesh, unstructured mesh, etc. One or more combinations.
- the basic idea of multigrid is to iterate a number of steps on the coarse mesh to get a more accurate result, and then use this result as the initial value to perform iteration on the fine mesh.
- the specific implementation of the ball filling method combined with the rapid Delaunay insertion to form a three-dimensional unstructured mesh may be a ball-packing arrangement in the three-dimensional geometric region, which can be adaptively distributed according to the geometric features and spatial relationships of the geometric model.
- a suitable node is obtained to obtain an ideal node distribution, and then a fast Delaunay insertion technique is used to efficiently generate an unstructured grid. If the grid size exceeds the computing power of a single machine, a grid generation method using parallel technology may also be considered.
- the region splitting algorithm can be applied to divide the whole structure into several sub-regions, and then the mesh segmentation is performed independently in each sub-region, and finally the sub-region boundary mesh repair is completed to obtain a mesh of the overall structure.
- Geometric optimization can move the node position and keep the node connection relationship unchanged.
- Topology optimization optimizes the connection between nodes.
- the mesh optimization uses a combination of geometric optimization and topology optimization.
- Geometric optimization may employ one or more combinations including, but not limited to, mesh quality smoothing techniques based on interpolation errors, unit deformation techniques, and the like.
- the unit deformation technique can move all node positions of a unit at the same time.
- the geometric optimization may employ a mesh quality smoothing technique based on interpolation error combined with a fast mesh quality smoothing algorithm of the unit deformation technique.
- Topology optimization can use the small polyhedral optimal split SPR technology based on the frontier propulsion strategy.
- Grid processing module 202 can obtain the required data by sending a request to client system 103. After obtaining the required data, the grid processing module 202 may perform the next processing or store the data in the system database 205. The grid processing module 202 can also retrieve the data stored in the system database 205 by sending a request to the system database 205. Alternatively, the system database 205 can also send a request directly to the client system 103, and the acquired data can be stored in the system data. Library 205.
- the client system 103 can be a server, a communication terminal, or the like. Further, the server may be a web server, a file server, a system database server, an FTP server, an application server, a proxy server, etc., or any combination of the above.
- the communication terminal may be a mobile phone, a personal computer, a wearable device, a tablet computer, a smart TV, or the like, or various combinations of the above communication terminals.
- the data acquired by the grid processing module 202 may include, but is not limited to, one or more combinations of grid structure, material properties, gas property parameters, geometry files, control parameters, processed grid files, and the like.
- the grid processing module 202 directly receives the unprocessed geometry file input by the client system 103. After the grid processing module 202 determines that the geometry file is not processed, the geometry file is subjected to meshing and the like, and the processed geometry is processed. The file is sent to the calculation solution module 203.
- the three-dimensional unstructured mesh rapid deformation module rapidly performs mesh deformation according to the structural optimization parameter change amount, and obtains the optimized unstructured mesh.
- the 3D unstructured mesh rapid deformation module can change the number of nodes of the mesh, the number of cells, and/or the topology connection relationship between nodes.
- the boundary point of the original computational grid can generate a background grid, and then the nodes of each computational grid are positioned, that is, the relative position of each computational node in the background grid is determined; then the structural boundary is determined according to the change of the optimized design parameters.
- the position change amount (deformation amount) of the point, and thus the optimized modified (deformed) background mesh; and then the new position of the node is calculated by the relative position of each computing node in the background mesh, that is, optimized Modified unstructured grid.
- the computational solution module 203 can be used to solve the physical quantities of the computational flow field.
- the governing equations for solving the calculations may be based on the Euler equations, or based on the N-S equations (Navier-Stokes Equations), or based on the lattice Boltzmann equation.
- Discretization methods for solving calculations include, but are not limited to, one or more combinations of finite difference method, finite volume method, finite element method, boundary element method, spectral method, lattice Boltzmann method, meshless method, and the like.
- the fluid for calculating the flow field may be non-viscous or viscous, may be a compressible fluid or an incompressible fluid, may be laminar or turbulent, and may be either a steady flow or an unsteady flow. Can be based on the physical properties of the fluid being simulated, Correspondingly select the corresponding control equation and simulation method.
- the Euler equations or lattice Boltzmann equations can be used for the flow field calculation of inviscid fluids.
- the N-S equations or Boltzmann equations can be used for the flow field calculation of viscous fluids.
- a medium pressure discharge plasma process determines the distribution of flow and thermal fields in the process chamber by numerical differential techniques to solve the N-S equations of the neutral gas.
- the low pressure discharge plasma process eg, plasma etch and PVD process
- the low pressure discharge plasma process is solved by the LBM and Monte Carlo methods to determine the distribution of the flow field and the thermal field.
- Numerical simulation methods for turbulence include, but are not limited to, Direct Numerical Simulation (DNS), Large Eddy Simulation (LES), and Reynolds-averaged Navier-Stokes equations (RANS).
- DBS Direct Numerical Simulation
- LES Large Eddy Simulation
- RANS Reynolds-averaged Navier-Stokes equations
- the turbulence model for solving the calculation may include, but is not limited to, a zero equation model, an equation model, a Spalart-Allmaras model, a k-epsilon model, a k-omega model, a RNG (Renormalization group) k-epsilon model, a Realizable k-epsilon model, and a Reynolds model.
- the governing equation for solving the calculation is the N-S equations and the discrete method is the finite volume method.
- the governing equation for solving the computation is the Boltzmann-BGK equation, the discrete method is a discrete time and space method, and the discrete velocity model is the DmQn model.
- the continuous Boltzmann equation is as follows,
- ⁇ (f' 1 f' 2 -f 1 f 2 )g d ⁇ d ⁇ 2 is a collision term.
- the BGK model can simplify the Boltzmann equation collision term, in which the collision process is a process of changing the distribution function f to the equilibrium distribution function f (eq) .
- the Boltzmann-BGK equation is as follows.
- ⁇ is the relaxation time of the distribution function and f (eq) is the Maxwell-Boltzmann local equilibrium distribution function.
- the discrete velocity model DmQn model can represent the equilibrium distribution function as follows.
- ⁇ i is a weight coefficient and may be a function of particle velocity; For the plaid sound speed.
- the discrete velocity model DmQn model may include, but is not limited to, one or more combinations of D1Q3, D1Q5, D2Q7, D2Q9, D3Q15, D3Q19, and the like. In one embodiment, the discrete velocity model is D2Q9. In another embodiment, the discrete velocity model is D3Q19.
- the calculation results of the analysis may also include, but are not limited to, aerodynamic forces, aerodynamic moments, geometric models of the computational domain and mesh analysis, vector graphics (such as velocity vector lines), contour maps, filled contours, etc. One or more combinations of graphs (cloud maps), XY scatter plots, particle trajectories, simulated flow effects, image processing functions, and the like.
- the result analysis module 204 analyzes the calculated result form to include, but is not limited to, one or more combinations of text, pictures, animations, and the like.
- the type of calculation result may include, but is not limited to, one or more combinations of txt, ASCII, MIME, and the like.
- System database 205 or other storage devices within the system generally refer to all media that can have read/write capabilities.
- the system database 205 or other storage devices in the system may be internal to the system or external devices of the system.
- the connection between the system database 205 or other storage devices in the system may be wired, It can be wireless.
- System database 205 or other storage devices within the system may include, but are not limited to, one or more combinations of hierarchical databases, networked databases, and relational databases.
- the system database 205 or other storage devices within the system may digitize the information and store it in a storage device that utilizes electrical, magnetic or optical means.
- System database 205 or other storage devices within the system can be used to store various information such as programs and data.
- the system database 205 or other storage devices in the system may be devices that store information by means of electrical energy, such as various memories, random access memory (RAM), read only memory (ROM), and the like.
- the system database 205 or other storage devices within the system may be devices that store information using magnetic energy, such as hard disks, floppy disks, magnetic tapes, magnetic core memories, magnetic bubble memories, USB flash drives, flash memories, and the like.
- System database 205 or other storage devices within the system may be devices that optically store information, such as CDs or DVDs.
- the system database 205 or other storage devices within the system may be devices that store information using magneto-optical means, such as magneto-optical disks.
- the access method of the system database 205 or other storage devices in the system may be one or more combinations of random storage, serial access storage, read-only storage, and the like.
- the system database 205 or other storage devices within the system may be non-persistent memory or permanent memory.
- the storage device mentioned above is a few examples, and the storage device that the system can use is not limited thereto.
- the information sent by the system database 205 may be data obtained directly from the client system 103, or may be processed and analyzed.
- the information processed by the analysis may be the information stored in the system database 205 after being processed by the grid processing module 202, or may be processed by the solution calculation module 203, or may be the information stored by the result analysis module 204.
- the system database 205 or other storage devices in the system may be local, remote, or on a cloud server.
- the information of the system database 205 or other storage devices in the system and other modules may be wired or wireless, and may be direct or indirect, and may be performed simultaneously or sequentially.
- the period can also be aperiodic or the like.
- the data receiving module 201, the grid processing module 202, the calculation solving module 203, the result analyzing module 204, and the system database 205 may be different modules embodied in one system, or may be one module to implement the above two or two.
- the calculation solution module 203 can directly receive the geometric file input by the data receiving module 201, the calculation solution module 203 can process the grid and can also solve the equation and analyze the result, and calculate the solution module and realize the grid processing.
- the functions of module 202 and result analysis module 204, and similar variations are still within the scope of the claims of the present invention.
- the analysis result of the result analysis module 205 can be directly displayed on the client system 103, or can be stored in the system database 205. When the display result needs to be viewed, it can be directly read from the system database.
- the analysis result of the result analysis module 205 is directly displayed on the client system 103, it may be displayed on the same client system 103 or on different client systems 103.
- Figure 3 shows a numerical simulation simulation flow chart.
- Data is received from client system 103 in step 301.
- Client system 103 can include, but is not limited to, one or more combinations of servers, communication terminals.
- the server may be a web server, a file server, a database server, an FTP server, an application server, a proxy server, etc., or any combination of the above.
- the communication terminal may be a mobile phone, a personal computer, a wearable device, a tablet computer, a smart TV, or the like, or any combination of the above communication terminals.
- Step 301 receives the instruction and data and proceeds to step 302 for operation.
- Grid processing and/or mesh optimization after data reception including but not limited to body-fitted mesh, partitioned mesh, Cartesian mesh, adaptive right-angle mesh, multi-grid, combined with fast Delaunay One or more combinations of ball filling methods, structured grids, unstructured grids, hybrid grids, and the like.
- the structured grid can be a single block or a multiple block.
- Multi-grid grids can be either spliced grids or nested grids.
- Cartesian grids can be single or multiple.
- the grid processing uses a structured grid. Further preferably, the grid processing uses a multi-block grid. Still more preferably, the mesh processing uses a Cartesian grid.
- the operations of the mesh processing include, but are not limited to, one or more combinations of mesh generation, mesh partitioning, node attribute marking, and wall distance calculation.
- Processing grid step 302 may be performed by grid processing module 202.
- Grid generation can include, but is not limited to, surface mesh, tetrahedral mesh, hexahedral mesh, prismatic mesh (boundary layer mesh), tetrahedral and hexahedral hybrid mesh, global Cartesian mesh, etc. Group Hehe. Grid generation allows geometric processing of the input geometry file data to be used for calculations.
- the mesh partition can generate complex and balanced multi-sub-grids according to the generated grid, and form data communication between the grids of each area.
- Grid partitioning can include, but is not limited to, partitioning according to the shape characteristics, determining the mesh topology (for example, a fine mesh is used where the flow field changes drastically, a sparse mesh is used for the far field region), a boundary curve mesh is generated, and a boundary surface mesh is generated.
- determining the mesh topology for example, a fine mesh is used where the flow field changes drastically, a sparse mesh is used for the far field region
- a boundary curve mesh is generated
- a boundary surface mesh is generated.
- the processing equation of the mesh encryption technology can be based on the lattice Boltzmann equation or the discrete-based velocity Boltzmann equation.
- Grid encryption techniques based on lattice Boltzmann equations may include, but are not limited to, interpolation, Taylor expansion, least squares, multi-block or multi-grid, ball filling method combined with fast Delaunay, regional split lattice Boltzmann, etc. One or more combinations.
- the mesh encryption technique based on the discrete velocity Boltzmann equation may include, but is not limited to, one or more combinations of finite difference, finite volume, finite element lattice Boltzmann method, and the like.
- Node attribute tags can distinguish between, but not limited to, fluid nodes, solid nodes, and boundary nodes.
- the node attribute marking method can be based on the relationship between the defined node and the solid wall boundary.
- Attributes whose nodes are outside the boundary can be marked as fluid nodes, attributes whose nodes are on the boundary can be marked as boundary nodes, and attributes whose nodes are within the boundary can be marked as solid nodes.
- the wall distance calculation can be used to calculate the distance between the boundary node and the solid wall boundary. For example, when you need to calculate surface boundary conditions, you need to use the result of the wall distance.
- the results of processing the grid may be stored in system database 205 or may be stored in an internal storage device of grid processing module 202.
- the mesh deformation can be changed according to the parameters of the rapid deformation module.
- the three-dimensional unstructured mesh rapid deformation module rapidly performs mesh deformation according to the structural optimization parameter change amount, and obtains the optimized unstructured mesh.
- the 3D unstructured mesh rapid deformation module can change the number of nodes of the mesh, the number of cells, and/or the topology connection relationship between nodes.
- the boundary point of the original computational grid can generate a background grid, and then the nodes of each computational grid are positioned, that is, the relative position of each computational node in the background grid is determined; then the structural boundary is determined according to the change of the optimized design parameters.
- the steps of grid optimization can be implemented by grid processing module 202.
- the results of the grid optimization may be stored in the system database 205 or may be stored in the internal storage device of the grid processing module 202.
- the flow field solution calculation can be performed (step 303).
- Flow field solution calculations include, but are not limited to, grid data and parameter input, initialization calculations, One or more combinations of selector unit selection, turbulence model selection and implementation, boundary condition correction, and result output.
- the solution calculation of the flow field can be done by the computational solution module 203.
- the calculation mode for grid data and parameter input selection can be serial or parallel. In the parallel computing mode, each process reads the sub-grid data corresponding to the grid, regardless of the primary and secondary.
- Parameter inputs include, but are not limited to, one or more combinations of determining model, pressure, density, temperature, velocity, boundary conditions, equilibrium distribution functions, and the like.
- the turbulence model for solving the calculation may include, but is not limited to, a zero equation model, an equation model, a Spalart-Allmaras model, a k-epsilon model, a k-omega model, a RNG k-epsilon model, a Realizable k-epsilon model, an RSM model, an ASM model.
- the solver unit can be determined according to different governing equations.
- the governing equation can be the Euler equation, the N-S equation, or the lattice Boltzmann equation.
- the solver unit of the lattice Boltzmann equation can transform the distribution function of the velocity space into the moment space, and combine the calculated model to correct the viscosity, collide in the moment space, and distribute the distribution.
- the function converts back to the velocity space and completes the migration of the distribution function.
- the boundary condition types may include, but are not limited to, one or more combinations of speed boundary conditions, pressure boundary conditions, periodic boundary conditions, standard bounce format boundary conditions, modified bounce format boundary conditions, half-step bounce boundary conditions, surface boundary conditions, and the like. .
- the boundary conditions are processed, but not limited to one or more of solid wall bounce and mass momentum combination method, solid wall bounce plus adjustment step method, hydrodynamic condition method, virtual solid wall method, solid wall bounce method, interpolation method, etc. combination.
- the resulting output can be a direct result of the calculated distribution function on each node, or it can be the result of converting the distribution function on each node into a macroscopic parameter.
- the result of the solution calculation is extracted and input to an analysis system (such as a result analysis module), and the result is further analyzed (step 304), and the further analysis result can output a graph, an animation, a data file, a result data table including but not limited to the visualization.
- an analysis system such as a result analysis module
- the further analysis result can output a graph, an animation, a data file, a result data table including but not limited to the visualization.
- the numerical calculation simulation system may also directly perform the solution calculation on the received processed grid file (step 303), and then analyze the result of the solution calculation (step 304). Alternatively, the numerical calculation simulation system may also directly analyze the received solution calculation result (step 304).
- FIG. 4 shows the structure of the grid processing module.
- the grid processing module 202 can include, but is not limited to, one or more grid generation units 401, one or more grid partition units 402, one or more node attribute marking units 403, one or more wall distance calculation units 404, One or more mesh optimization units 405.
- Each unit in the grid processing module 202 may be independent or a unit may be merged into one unit.
- the various units in the grid processing module 202 can be local or remote.
- the grid generating unit 401 can process the input data of the data receiving module 201, and generate a grid suitable for calculation according to the definitions of the grid structure, the geometric file, and the control parameters.
- the generated mesh may include, but is not limited to, a surface mesh, a tetrahedral mesh, a hexahedral mesh, a prismatic mesh (boundary layer mesh), a tetrahedral and hexahedral hybrid mesh, a global Cartesian mesh, or the like.
- the input data is a three-dimensional geometric figure
- the mesh generation unit 401 can process the geometrical image to generate a tetrahedral mesh.
- the results of the grid generation may be stored in the system database 205, or may be stored in the storage device of the grid generation unit 401, or may be stored in other storage devices in the system.
- the mesh partitioning unit 402 can generate a complex and balanced multi-block sub-grid through the partition according to the generated grid to form data communication between the grids of the respective regions.
- the mesh partitioning unit 402 can also determine a plurality of sub-grids that are complexly balanced after the partitioning, and form data communication between the grids of the respective regions.
- Grid partitioning can include, but is not limited to, partitioning according to the shape characteristics, determining the mesh topology (for example, a fine mesh is used where the flow field changes drastically, a sparse mesh is used for the far field region), a boundary curve mesh is generated, and a boundary surface mesh is generated.
- the node attribute tagging unit 403 can perform attribute tagging on the mesh boundary.
- the marked node attributes may include, but are not limited to, fluid nodes, solid nodes, and boundary nodes.
- the node attribute tag can be marked after the mesh partition, or it can be marked before the mesh partition.
- the method of node attribute tagging can be based on the relationship between the node and the solid wall boundary. For example, attributes whose nodes are outside the boundary can be marked as fluid nodes, attributes whose nodes are on the boundary can be marked as boundary nodes, and attributes whose nodes are within the boundary can be marked as solid nodes.
- the node attribute tagging unit 403 performs attribute tagging on the generated mesh.
- the result of the node attribute tag may be stored in the system database 205, or may be stored in the storage device of the grid generating unit 401, or may be stored in other storage devices in the system.
- the calculation method of the wall distance may include, but is not limited to, one or more combinations of a Possion equation method, an Eikonal equation method, a Hamilton-Jacobi equation method, and the like.
- the results of the wall distance calculations may be stored in the system database 205 or may be stored in the internal storage device of the grid processing module 202.
- the mesh optimization unit 405 can be used for one or more combinations including, but not limited to, a body-fitted mesh, a partitioned mesh, an adaptive right-angle mesh, a multi-grid, a quality optimization of a structured three-dimensional mesh, and the like.
- Grid optimization includes, but is not limited to, one or more combinations of geometric optimization (smoothing), topology optimization, and the like.
- Geometric optimization can move the node position and keep the node connection relationship unchanged.
- Topology optimization optimizes the connection between nodes.
- the mesh optimization uses a combination of geometric optimization and topology optimization.
- Geometric optimization may employ one or more combinations including, but not limited to, mesh quality smoothing techniques based on interpolation errors, unit deformation techniques, and the like.
- the 3D unstructured mesh rapid deformation module can change the number of nodes, the number of cells and/or the topology connection between nodes without changing the mesh. relationship.
- the boundary point of the original computational grid can generate a background grid, and then the nodes of each computational grid are positioned, that is, the relative position of each computational node in the background grid is determined; then the structural boundary is determined according to the change of the optimized design parameters.
- the position change amount (deformation amount) of the point, and thus the optimized modified (deformed) background mesh; and then the new position of the node is calculated by the relative position of each computing node in the background mesh, that is, optimized Modified unstructured grid.
- the initializing unit 502 can take the macro physical quantity input by the collecting unit 501 as an initial condition.
- the operations that the initialization unit 502 can perform include, but are not limited to, initializing a parallel process number, initializing a computation area corresponding to each process, initializing a macroscopic physical quantity of the flow field, initializing a distribution function, and the like.
- the physical quantities initialized include, but are not limited to, one or more combinations of pressure, density, temperature, speed, and the like.
- the solver unit 503 can determine different solver units based on the governing equations.
- the governing equation can be the Euler equation, the N-S equation, or the lattice Boltzmann equation.
- the turbulence unit 504 can correct the relaxation time by solving the calculated model.
- the choke unit 504 can correct the solution calculation model of the solver unit 503.
- the numerical simulation method for solving the calculation can be a direct numerical simulation or an indirect numerical simulation. Indirect numerical simulations may include, but are not limited to, one or more combinations of large eddy simulation methods, Reynolds averaging methods, statistical averaging methods, and the like.
- the turbulence model for solving the calculation may include, but is not limited to, a zero equation model, an equation model, a Spalart-Allmaras model, a k-epsilon model, a k-omega model, a RNG k-epsilon model, a Realizable k-epsilon model, an RSM model, an ASM model.
- the boundary condition unit 505 makes a correction to the distribution function on the boundary.
- the distribution function can be either migrated without collision or after collision migration.
- the boundary condition types may include, but are not limited to, one or more combinations of speed boundary conditions, pressure boundary conditions, periodic boundary conditions, standard bounce format boundary conditions, modified bounce format boundary conditions, half-step bounce boundary conditions, surface boundary conditions, and the like. .
- the boundary conditions are processed, but not limited to one or more of solid wall bounce and mass momentum combination method, solid wall bounce plus adjustment step method, hydrodynamic condition method, virtual solid wall method, solid wall bounce method, interpolation method, etc. combination.
- the format used by the boundary condition unit 505 may be a non-equilibrium extrapolation pressure format.
- Step 601 collects data.
- the received data may be derived from the mesh generation module 202, may be derived from the data receiving module 201, or may be derived from the system database 205.
- Step 601 can be done by collection unit 501.
- the collected data may include, but is not limited to, one or more combinations of grid data, parameters, calculation modes, models, boundary conditions, accuracy requirements, and the like.
- Grid data can be processed or unprocessed.
- the calculation mode can be serial or parallel. In the parallel computing mode, each process reads the sub-grid data corresponding to the grid, regardless of the primary and secondary.
- Parameter inputs include, but are not limited to, one or more combinations of determining governing equations, models, pressure, density, temperature, velocity, boundary conditions, equilibrium distribution functions, step requirements, and the like.
- the governing equation chosen is the Boltzmann-BGK equation and the model is DmQn.
- Step 602 initializes the flow field based on the collected data.
- the content of the initialization flow field may include, but is not limited to, one or more combinations of initializing the parallel process number, initializing the calculation area corresponding to each process, initializing the macroscopic physical quantity of the flow field, and initializing the distribution function.
- the physical quantities initialized include, but are not limited to, one or more combinations of pressure, density, temperature, speed, and the like.
- Step 602 can be completed by initialization unit 502.
- the governing equation is a lattice Boltzmann equation, which can be used to solve computational models including but not limited to BGK models, MRT-LBM models, SRT-LBM models, lattice Boltzmann models (eg DmQn models), incompressible lattices Boltz One or more combinations of a Mann model, a hot lattice Boltzmann model, a lattice Boltzmann model of a non-uniform grid, and the like.
- the equilibrium distribution function of the Boltzmann equation can be determined by a discrete velocity model.
- the discrete velocity model is a DmQn model.
- step 604 further, the viscosity can be corrected by solving the calculation model, and the relaxation time of the distribution function is further determined by the viscosity.
- Step 604 can This is done by the solver unit 503 and the choke unit 504.
- the relaxation time reflects the rate at which the distribution function approaches equilibrium due to the collision, and it can also be corrected by the turbulence model.
- Step 605 processes the boundary conditions and corrects the calculated distribution function.
- Step 605 can be completed by boundary condition unit 505.
- the distribution function can be either migrated without collision or after collision migration.
- the boundary condition types may include, but are not limited to, one or more combinations of speed boundary conditions, pressure boundary conditions, periodic boundary conditions, standard bounce format boundary conditions, modified bounce format boundary conditions, half-step bounce boundary conditions, surface boundary conditions, and the like.
- the boundary conditions are processed, but not limited to one or more of solid wall bounce and mass momentum combination method, solid wall bounce plus adjustment step method, hydrodynamic condition method, virtual solid wall method, solid wall bounce method, interpolation method, etc. combination.
- Step 606 updates the distribution function on each node after the calculation to a macro physical quantity.
- Macroscopic physical quantities include, but are not limited to, one or more combinations of speed, pressure, density, and the like.
- Step 606 can be completed by result output unit 506.
- Step 607 judges that if the solution calculation is a constant problem, the process proceeds to step 609.
- Step 609 continues to determine whether the calculated macro physical quantity satisfies the criterion of the steady condition. If the output result of the entering step 610 is satisfied, if the continuous solution calculation is not satisfied, the process proceeds to step 604.
- Step 607 determines if the solution calculation is an unsteady problem, then proceeds to step 608.
- Step 608 continues to determine whether the calculation has reached the set step size requirement. If the result is required to proceed to step 610, if not, the continuation solution calculation proceeds to step 604.
- the above steps are performed by the solver unit 503, the choke unit 504, the boundary condition unit 505, and the result output unit 506.
- the vorticity of the flow field near the front surface of the cylinder is the largest, and the vorticity is small in other regions, and the upper half is negative and the lower half is positive.
- the coefficient of drag and the coefficient of lift of the cylinder are calculated by the momentum exchange method and stabilized after 15,000 steps.
- Figure 7-b shows the flow around a cylinder with a Reynolds number of 100. An asymmetric vortex is formed in the area behind the cylinder, and the positive and negative values of the vorticity are also asymmetrical. The coefficient of drag and the coefficient of lift of the cylinder then oscillate periodically.
- Figure 8-a shows a chamber structure diagram of a neutral gas discharge plasma.
- Neutral gases include, but are not limited to, one or more combinations of argon, helium, neon, xenon, xenon, oxygen, nitrogen, hydrogen, methane, carbon monoxide, inert gases, and the like.
- argon is selected as the working gas
- the gas pressure is 1 Torr
- the current is 10 A
- the frequency is 13.56 MHz.
- oxygen is selected as the discharge gas
- the gas pressure is 1 Torr
- the current is 10 A
- the frequency is 13.56 MHz.
- helium is selected as the working gas
- the gas pressure is 1 Torr
- the current is 10 A
- the frequency is 13.56 MHz.
- the flow rate of the working gas may be from 500 sccm to 9000 sccm.
- the flow rate of the working gas may be 2000 sccm.
- the flow rate of the working gas may be 3000 sccm. More preferably, the flow rate of the working gas may be 4000 sccm.
- the neutral gas discharge plasma calculation preferably employs a non-viscous model, and the solution algorithm is preferably a steady state simple algorithm.
- the axial and radial grid sizes of the chambers of the neutral gas discharge plasma were both 0.2 cm.
- the boundary condition at the inlet is a constant flow rate
- the boundary condition at the outlet is a constant air pressure.
- the calculated residual convergence accuracy setting can be 10 -9 to 10 -1 .
- the residual convergence accuracy may be 10 -6 .
- Figure 8-b shows a streamlined flow diagram of the discharge of working gas argon at a flow rate of 2000 sccm.
- Figure 8-c shows a streamlined flow diagram of the discharge of working gas argon at a flow rate of 3000 sccm.
- Figure 8-d shows a streamlined flow diagram of the discharge of working gas oxygen at a flow rate of 3000 sccm.
- FIG. 9 is a schematic diagram showing the structure of a device for implementing a numerical calculation simulation system and method.
- the computer 900 can be a general purpose computer or a special purpose computer.
- Computer 900 can be used to implement any of the components described above that can provide functional services.
- the numerical computing simulation system 101 can be implemented by one or more of a combination of hardware, software programs, etc. of a computer device (eg, computer 900).
- computer 900 can include a COM port 950 and a network connection to facilitate data communication.
- Computer 900 can also include a central processing unit (CPU) 920, which can be one or more processors to execute program instructions.
- the computing platform in the example may also include an internal communication bus 910 and different forms of programs, data storage devices.
- the numerical computational simulation system of the present invention can selectively solve different control equations, either user-defined or system-defined.
- the numerical calculation simulation system of the present invention can be run as a browser plug-in, or can be run as a program or a web version.
Landscapes
- Engineering & Computer Science (AREA)
- Physics & Mathematics (AREA)
- Theoretical Computer Science (AREA)
- Computer Hardware Design (AREA)
- Evolutionary Computation (AREA)
- Geometry (AREA)
- General Engineering & Computer Science (AREA)
- General Physics & Mathematics (AREA)
- Management, Administration, Business Operations System, And Electronic Commerce (AREA)
Abstract
一种模拟等离子体放电的数值计算仿真系统(101),该数值计算仿真系统(101)可以包括计算机可读的存储媒介,存储媒介可以包括数据接收模块(201),其可以接收数据,网格处理模块(202)可以对接收的数据进行网格剖分处理生成处理结果,计算求解模块(203)可以根据网格处理结果求解计算控制方程生成流场与热场的计算结果,结果分析模块(204)可以对生成的流场与热场计算结果进一步分析生成分析结果,处理器可以执行所述计算机可读的存储媒介存储的可执行模块。一种模拟等离子体放电的数值计算仿真方法,包括接收数据,根据接收的数据进行网格剖分处理,求解计算控制方程得到网格剖分处理后的流场和热场计算结果,进一步分析提取的计算结果等。
Description
本发明涉及一种数值计算仿真系统及方法,尤其适用于涉及流体的流场和热场耦合的计算仿真和设计等,如等离子体放电过程的数值模拟等。
随着计算机的发展和数值计算方法的进步,计算流体力学(Computational Fluid Dynamic,简称CFD)已经可以处理越来越多的复杂问题。计算流体力学、理论流体力学、实验流体力学是流体力学研究工作的三种主要手段,相辅相成。理论分析为实验和计算研究提供依据;实验为数值研究提供数据并验证计算结果;数值计算则是特殊意义下的实验。数值计算具有更大的自由度和灵活性,可以进行“物理实验”不可能或很难进行的实验。计算流体力学是使用数值方法在计算机中对流体力学的控制方程进行求解,从而预测流场的流动。其中,计算流体力学主要涉及流体无粘绕流和粘性流动。无粘绕流包括低速流、跨声速流、超声速流等;粘性流动包括湍流、边界层流动等。
流体力学的应用领域遍及基础工业的各个方面,除了航空航天、气象预报和油气开采等领域外,还包括汽车车型空气动力学优化、风扇及运动物体空气噪音源的分析和降低、热对流及热传输对于器件乃至环境的影响、工艺腔室内的中性气体的流场和热场、中等气压放电等离子体工艺等,随着新兴工业的诞生,流体力学的应用也扩展到微流、多相流、非牛顿流以及其他复杂流体。
发明内容
本发明一方面是关于一种模拟等离子体放电的数值计算仿真系统,根据其中一个实施例,该数值计算仿真系统包括一种计算机可读的存储媒介,所述存储媒介存储可执行模块,该存储媒介包括数据接收模块,所述数据接收模块能够接收数据;网格处理模块,所述网格处理模块能够对接收的数据进行网格剖分处理,生成网格处理结果;计算求解模块,所述计算求解模块能够根据网格处理结果求
解计算控制方程生成流场与热场的计算结果;结果分析模块,所述结果分析模块能够对生成的流场与热场计算结果进一步分析生成分析结果。一个处理器,所述处理器能够执行所述计算机可读的存储媒介存储的可执行模块。
根据本发明的另一个实施例,该数值计算仿真系统进一步包括一个数据库,所述数据库能够储存所述的接收数据、网格处理结果、计算结果、分析结果。
根据本发明的另一个实施例,该数值计算仿真系统的网格处理模块进一步包括网格生成单元、网格分区单元、节点属性标记单元、壁面距离计算单元、网格优化单元。
根据本发明的另一个实施例,该数值计算仿真系统的控制方程包括Euler方程、N-S方程组、格子玻尔兹曼方程。
根据本发明的另一个实施例,该数值计算仿真系统的计算求解模块进一步包括收集单元、初始化单元、求解器单元、湍流单元、边界条件单元、结果输出单元。
根据本发明的另一个实施例,该数值计算仿真系统的网格生成单元生成的网格包括表面网格、四面体网格、六面体网格、棱柱体网格(边界层网格)、四面体与六面体混合网格、笛卡尔网格、球填充法网格。
根据本发明的另一个实施例,该数值计算仿真系统的球填充法网格基于Delaunay剖分插入技术。
根据本发明的另一个实施例,该数值计算仿真系统的节点属性标记单元的节点属性包括流体节点、固体节点和边界节点。
根据本发明的另一个实施例,该数值计算仿真系统的湍流单元求解计算模型包括零方程模型、一方程模型、Spalart-Allmaras模型、k-epsilon模型、k-omega模型、RNG k-epsilon模型、Realizable k-epsilon模型、RSM模型、ASM模型、SGS模型、BGK模型、MRT-LBM模型、SRT-LBM模型、格子玻尔兹曼模型、不可压格子玻尔兹曼模型、热格子玻尔兹曼模型、非均匀网格的格子玻尔兹曼模型。
根据本发明的另一个实施例,该数值计算仿真系统的结果分析模块进一步分析的结果包括速度、压力、密度、温度、整体流场的宏观物理量、指定截面的宏观物理量、气动力、气动力矩、计算域的几何模型及网格分析、矢量图(如速度
矢量线)、等值线图、填充型的等值线图(云图)、XY散点图、粒子轨迹图、模拟流动效果、图像处理功能。
本发明另一方面是关于一种模拟等离子体放电的数值计算仿真方法,根据其中一个实施例,该数值计算仿真方法,包括接收数据;根据所述数据进行网格剖分处理;根据所述剖分处理的网格求解计算控制方程的流场和热场计算结果;进一步提取所述流场和热场的计算结果进行分析。
根据本发明的另一个实施例,该数值计算方法的接收数据包括网格结构、材料属性、气体特性参数、几何文件、控制参数、网格文件。
根据本发明的另一个实施例,该数值计算方法的控制参数包括网格类型、控制方程、模型、计算精度、残差收敛精度。
根据本发明的另一个实施例,该数值计算方法的网格类型包括贴体网格、分区网格、笛卡尔网格、自适应直角网格、多重网格、结合快速Delaunay的球充填法网格、结构化网格、非结构化网格、杂交网格。
根据本发明的另一个实施例,该数值计算方法的控制方程包括Euler方程、N-S方程、格子玻尔兹曼方程。
根据本发明的另一个实施例,该数值计算方法模型包括零方程模型、一方程模型、Spalart-Allmaras模型、k-epsilon模型、k-omega模型、RNG k-epsilon模型、Realizable k-epsilon模型、RSM模型、ASM模型、SGS模型、BGK模型、MRT-LBM模型、SRT-LBM模型、格子玻尔兹曼模型、不可压格子玻尔兹曼模型、热格子玻尔兹曼模型、非均匀网格的格子玻尔兹曼模型。
根据本发明的另一个实施例,该数值计算方法的接收数据包括用户输入数据和非用户输入数据,所述的非用户输入数据来源包括服务器、通信终端。
为了更清楚地说明本发明实施例的技术方案,下面将对实施例描述中所需要使用的附图作简单的介绍。显而易见地,下面描述中的附图仅仅是本发明的一些实施例,对于本领域的普通技术人员来讲,在不付出创造性劳动的前提下,还可以根据这些附图将本发明应用于其它类似情景。除非从语言环境中显而易见或另做说明,图中相同标号代表相同结构和操作。
图1所示的是数值计算仿真系统的一种示例系统配置的示意图。
图2所示的是数值计算仿真系统的模块示意图。
图3所示的是数值计算仿真流程图。
图4所示的是网格处理模块的结构示意图。
图5所示的是计算求解模块的结构示意图。
图6所示的是数值计算仿真系统的一种计算求解流程图。
图7-a所示的是数值计算仿真系统解决雷诺数为20的圆柱绕流问题结果示意图。
图7-b所示的是数值计算仿真系统解决雷诺数为100的圆柱绕流问题结果示意图。
图8-a所示的是中性气体放电等离子体的腔室结构图。
图8-b所示的是工作气体氩气在流速2000sccm时放电的流场仿真流线图。
图8-c所示的是工作气体氩气在流速3000sccm时放电的流场仿真流线图。
图8-d所示的是工作气体氧气在流速3000sccm时放电的流场仿真流线图。
图9所示的是实现数值计算仿真系统及方法的设备结构示意图。
详细描述
如本说明书和权利要求书中所示,除非上下文明确提示例外情形,“一”、“一个”、“一种”和/或“该”等词并非特指单数,也可包括复数。一般说来,术语“包括”与“包含”仅提示包括已明确标识的步骤和元素,而这些步骤和元素不构成一个排它性的罗列,方法或者设备也可能包含其它的步骤或元素。
本说明书描述的“数值计算模拟”、“数值计算仿真”、“数值仿真”、“数值模拟”、“仿真计算”、“计算仿真”、“数值计算”、“数值分析”等是可互换的,是指研究并解决数学问题的数值近似解方法,在计算机等计算求解设备上使用的求解数学问题的方法。计算流体力学是指使用数值方法在计算机等计算求解设备中对流体力学的控制方程进行求解,从而预测流场的流动。计算流体力学求解的领域包括但不限于工程设计(如汽车、列车、航空、航天、核工业、建筑业等),流体交互的设备(如泵、化学装置、通风系统、食品的冷冻冷藏等),计算机图形学(如模拟动画或游戏的流体)、等离子体放电技术(如半导体集成电路及其他微电子设备制造工艺、有机和无机材料的表面功能化处理、等离子体
制备有机和无机纳米颗粒、等离子体灭菌、照明及显示等)等。本说明书描述的等离子体包括但不限于高温等离子体、低温等离子体(热等离子体、冷等离子体)、平衡等离子体、非平衡等离子体等一种或多种组合。等离子体的产生方法包括但不限于辉光放电、电晕放电、介质阻挡放电、射频放电、电弧放电、射流放电、大气压辉光放电、次大气压辉光放电、直流弧光放电、交流工频放电、高频感应放电、低气压放电、燃烧法等一种或多种组合。
本发明的不同实施例应用场景可以包括但不限于汽车、火车、高铁、船舶、飞行器等交通运输工具与流体相互作用的模拟,还可以包括但不限于激光、电磁、射频、波动光学、射线光学、MEMS、半导体、传热、结构力学、声学、微流体、分子流、管道流等一种或多种物理场耦合,还可以包括但不限于多相流、多孔介质流、悬浮粒子流、反应流、磁流体力学和生物力学等一种或多种领域的组合。以上对适用领域的描述仅仅是具体的示例,不应被视为是唯一可行的实施方案。显然,对于本领域的专业人员来说,在了解一种数值计算仿真系统及方法的基本原理后,可能在不背离这一原理的情况下,对实施上述方法和系统的应用领域形式和细节上的各种修正和改变,但是这些修正和改变仍在以上描述的范围之内。
为了更清楚地说明本发明实施例的技术方案,下面将对实施例描述中所需要使用的附图作简单的介绍。显而易见地,下面描述中的附图仅仅是本发明的一些实施例,对于本领域的普通技术人员来讲,在不付出创造性劳动的前提下,还可以根据这些附图将本发明应用于其它类似情景。除非从语言环境中显而易见或另做说明,图中相同标号代表相同结构和操作。
图1所示的是数值计算仿真系统的一种示例系统配置的示意图。示例系统配置100可以包含但不限于一个或多个数值计算仿真系统101、一个或多个网络系统102和一个或多个客户端系统103。数值计算仿真系统101可以是一个服务器,也可以是一个服务器群组。一个服务器群组可以是集中式的,例如数据中心。一个服务器群组也可以是分布式的,例如一个分布式系统。数值计算仿真系统101可以是本地的,也可以是远程的。网络102可以是单一网络,也可以是多种网络组合的。网络102可以包括但不限于局域网、广域网、公用网络、专用网络、无线局域网、虚拟网络、都市城域网、公用开关电话网络等一种或多种组合。网络102可以包括多种网络接入点,如有线或无线接入点、基站或网络交换点,通
过以上接入点使数据源连接网络102并通过网络发送数据。客户端系统103可以包括但不限于手机103-1、笔记本电脑103-2、台式机103-3等一种或多种组合。客户端系统103可以通过网络系统102与数值计算仿真系统101通信。客户端系统103与数值计算仿真系统101的通信包括但不限一对一、一对多、多对一、多对多等。在一个实施例中,客户端系统103可以运行网络浏览器,网络浏览器可以与运行在数值计算仿真系统101上的网络连接系统进行交互。客户端系统103可以用于接收用户输入的数据,也可以根据示例的用户界面向用户显示。显示的内容可以是交互式界面,也可以是非交互式界面。交互式界面可以是显示数值计算仿真系统101内部处理的各种条件输入,也可以是显示分析后的结果。客户端系统103可以同时面向包括但不限于一个或多个用户。例如,不同客户端系统103同时显示同一个数值计算仿真系统101计算分析的结果。又例如,同一个客户端系统103显示不同的数值计算仿真系统101计算分析的结果。再例如,一个客户端系统103输入需要处理的条件,另一个客户端系统103显示数值计算仿真系统101计算分析的结果。
图2所示的是数值计算仿真系统的模块示意图。数值计算仿真系统101可以包含但不限于一个或多个数据接收模块201、一个或多个网格处理模块202、一个或多个计算求解模块203、一个或多个结果分析模块204、一个或多个系统数据库205。数值计算仿真系统101的模块部分或全部可以与网络102连接。数值计算仿真系统101的模块可以是集中式的也可以是分布式的。数值计算仿真系统101的一个或多个模块可以是本地的也可以是远程的。数据接收模块201可以用于接收数据文件。数据文件可以包括但不限于网格结构、材料属性、气体特性参数、几何文件、控制参数、已处理后的网格文件等一种或多种组合。几何文件的获取方式可以是直接的(例如直接通过网络102从一个或多个客户端系统103获取数据)也可以是间接的(例如通过网格处理模块202、计算求解模块203、结果分析模块204、系统数据库205来获取)。网格处理模块202可以用于几何文件的网格处理,也可以用于网格的优化。网格处理包括但不限于贴体网格、分区网格、自适应直角网格、多重网格、结合快速Delaunay的球充填法等一种或多种组合。多重网格的基本思路是在粗网格上迭代若干步后,得到一个比较精确地结果,然后再用这个结果为初始值,进行细网格上的迭代。球充填法结合快速
Delaunay插入形成三维非结构网格的具体实现方案可以是在三维几何区域内用球充填法(ball-packing)布点,可以依据几何模型的几何特征和空间关系自适应地分布疏密合适的节点,从而获得理想的节点分布,然后采用快速Delaunay插入技术高效率地生成非结构网格,如果网格规模超出单机的计算能力,则还可以考虑采用并行技术的网格生成方法,此时可应用区域分裂算法将整个结构分为若干个子区域,然后在每个子区域内独立进行网格剖分,最后完成子区域边界网格修复,得到整体结构的网格。网格的优化可以用于包括但不限于贴体网格、分区网格、自适应直角网格、多重网格、结构三维网格的质量优化等一种或多种组合。网格优化包括但不限于几何优化(修匀)、拓扑优化等一种或多种组合。几何优化可以移动节点位置,保持节点连接关系不变。拓扑优化可以优化节点之间的连接关系。优选地,网格优化采用几何优化和拓扑优化相结合。几何优化可以采用包括但不限于基于插值误差的网格质量修匀技术、单元变形技术等一种或多种组合。单元变形技术可以同时移动一个单元的所有节点位置。优选地,几何优化可以采用基于插值误差的网格质量修匀技术结合单元变形技术的快速网格质量修匀算法。拓扑优化可以采用基于前沿推进策略的小多面体最优剖分SPR技术。
计算求解模块203可以用于求解网格处理后的流场情况。求解流场的方法包括但不限于有限差分法、有限体积法、有限元法、谱方法、格子玻尔兹曼法、无网格法等一种或多种组合。计算求解可以是并行计算,也可以是串行计算。并行计算可以是数据并行的,也可以是任务并行的。结果分析模块204可以用于分析计算结果。分析的计算结果可以是整体流场的宏观物理量,也可以是指定截面的流场宏观物理量。宏观物理量的分析可以包括但不限于速度、压力、密度、温度、气动力矩等一种或多种组合。系统数据库205主要用于存储从客户端系统103接收的数据和数值计算仿真系统101工作中产生的各种数据。系统数据库205可以是本地的,也可以是远程的。系统数据库与系统其他模块间的连接或通信可以是有线的,也可以是无线的。
数据接收模块201可以用于接收数据文件。数据文件可以包括但不限于网格结构、材料属性、气体特性参数、几何文件、控制参数、已处理后的网格文件等一种或多种组合。几何文件的获取方式可以是直接的(例如直接通过网络102从一个或多个客户端系统103获取数据)也可以是间接的(例如通过网格处理模块
202、计算求解模块203、结果分析模块204、系统数据库205来获取)。例如,数据接收模块201可以接收几何文件,几何文件可以是未经处理的,也可以是处理后不符合要求再处理的。又例如,数据接收模块201可以接收处理后的网格文件,进一步地,数据接收模块201分析该网格文件符合要求,若符合要求可以直接进入计算求解模块203计算流场情况,若不符合要求可以进入网格处理模块202进行网格再处理。数据接收模块201可以接收客户端系统103输入的数据,也可以接收系统数据库205输入的数据。数据可以是原始数据,也可以是经过其他模块(如网格处理模块202、计算求解模块203、结果分析模块204、系统数据库205)处理后的数据。
网格处理模块202可以接收数据接收模块201输入的数据,也可以直接接收客户端系统103输入的数据。网格处理模块202也可以将接收到的数据传输给计算求解模块203以进行计算和处理。网格处理模块202也可以将接收到的数据传输到结果分析模块204。网格处理模块202可以接收计算求解模块203发送的请求,也可以按照该请求访问系统数据库205,以获取需要的数据。需要的数据被获取之后,网格处理模块202可以将该数据传输给计算求解模块203。网格处理模块202可以接收结果分析模块204发送的请求,也可以按照该请求访问系统数据库205,以获取需要的数据。需要的数据被获取之后,网格处理模块202可以将该数据传输到结果分析模块204。计算求解模块203可以接收网格处理模块202发送的请求,也可以按照该请求访问系统数据库205,以获取需要的数据。需要的数据被获取之后,计算求解模块203可以将该数据传输给网格处理模块204。计算求解模块203可以接收结果分析模块204发送的请求,也可以按照该请求访问系统数据库205,以获取需要的数据。需要的数据被获取之后,计算求解模块203可以将该数据传输结果给分析模块204。结果分析模块204可以接收网格处理模块202发送的请求,也可以按照该请求访问系统数据库205,以获取需要的数据。需要的数据被获取之后,结果分析模块204可以将该数据传输给网格处理模块202。结果分析模块204可以接收计算求解模块203发送的请求,也可以按照该请求访问系统数据库205,以获取需要的数据。需要的数据被获取之后,结果分析模块204可以将该数据传输给计算求解模块203。
网格处理模块202可以用于几何文件的网格处理和/或根据要求确认划分网格,也可以用于优化网格结构。网格处理包括但不限于贴体网格、分区网格、笛卡尔网格、自适应直角网格、多重网格、结合快速Delaunay的球充填法、结构化网格、非结构化网格等一种或多种组合。多重网格的基本思路是在粗网格上迭代若干步后,得到一个比较精确地结果,然后再用这个结果为初始值,进行细网格上的迭代。球充填法结合快速Delaunay插入形成三维非结构网格的具体实现方案可以是在三维几何区域内用球充填法(ball-packing)布点,可以依据几何模型的几何特征和空间关系自适应地分布疏密合适的节点,从而获得理想的节点分布,然后采用快速Delaunay插入技术高效率地生成非结构网格,如果网格规模超出单机的计算能力,则还可以考虑采用并行技术的网格生成方法,此时可应用区域分裂算法将整个结构分为若干个子区域,然后在每个子区域内独立进行网格剖分,最后完成子区域边界网格修复,得到整体结构的网格。结构化网格是网格区域内所有的内部点都具有相同的毗邻单元,可以是四边形的,也可以是六面体的。非结构化网格是网格区域内的内部点不具有相同的毗邻单元,可以是三角形的,也可以是四面体的。网格处理模块202处理的几何图形可以是二维的,也可以是三维的。网格的优化可以用于包括但不限于贴体网格、分区网格、自适应直角网格、多重网格、结构三维网格的质量优化等一种或多种组合。网格优化包括但不限于几何优化(修匀)、拓扑优化等一种或多种组合。几何优化可以移动节点位置,保持节点连接关系不变。拓扑优化可以优化节点之间的连接关系。优选地,网格优化采用几何优化和拓扑优化相结合。几何优化可以采用包括但不限于基于插值误差的网格质量修匀技术、单元变形技术等一种或多种组合。单元变形技术可以同时移动一个单元的所有节点位置。优选地,几何优化可以采用基于插值误差的网格质量修匀技术结合单元变形技术的快速网格质量修匀算法。拓扑优化可以采用基于前沿推进策略的小多面体最优剖分SPR技术。
网格处理模块202可以通过向客户端系统103发送请求,以获取需要的数据。网格处理模块202在获取需要的数据后,可以将所获得的数据进行下一步处理或者存储在系统数据库205中。网格处理模块202也可以通过向系统数据库205发送请求,以获取存储在系统数据库205中的数据。可选择地,系统数据库205也可以直接向客户端系统103发送请求,获取的数据可以被存储在系统数据
库205中。客户端系统103可以是服务器、通信终端等。进一步地,服务器可以是web服务器、文件服务器、系统数据库服务器、FTP服务器、应用程序服务器、代理服务器等,或者上述服务器的任意组合。通信终端可以是手机、个人电脑、可穿戴设备、平板电脑、智能电视等,或者上述通信终端的各种组合。网格处理模块202获取的数据可以包括但不限于网格结构、材料属性、气体特性参数、几何文件、控制参数、已处理后的网格文件等一种或多种组合。例如,网格处理模块202直接接收客户端系统103输入的未处理的几何文件,网格处理模块202判断为未处理几何文件后,对该几何文件进行网格划分等处理,将处理后的几何文件发送给计算求解模块203。又例如,网格处理模块202接收客户端系统103输入的几何文件,网格处理模块202判断为已处理几何文件后,直接将该几何文件发送给计算求解模块203。再例如,网格处理模块202接收客户端系统103输入的几何文件,网格处理模块202判断为已处理几何文件后需要网格优化,网格处理模块202会对该几何文件进一步进行网格优化,再将优化后的几何文件发送给计算求解模块203。网格处理模块202在结构优化中可以进行结构重分析、重新生成,也可以根据快速变形模块的参数改变了进行网格变形。优选地,三维非结构网格快速变形模块,根据结构优化设计参数的改变量,快速地进行网格变形,得到优化后的非结构网格。三维非结构网格快速变形模块可以不改变网格的节点数量、单元数量和/或节点间的拓扑连接关系。原计算网格的边界点可以生成背景网格,其次对每个计算网格的节点进行定位,即确定每个计算节点在背景网格中的相对位置;然后根据优化设计参数的变化确定结构边界点的位置变化量(变形量),并由此获得经优化修改(变形)后的背景网格;再由每个计算节点在背景网格中的相对位置计算节点的新位置,即得到经过优化修改后的非结构网格。
计算求解模块203可以用于求解计算流场的物理量。求解计算的控制方程可以是基于Euler方程组,也可以是基于N-S方程组(Navier-Stokes Equations)的,也可以是基于格子玻尔兹曼方程的。求解计算的离散化方法包括但不限于有限差分法、有限体积法、有限元法、边界元方法、谱方法、格子玻尔兹曼法、无网格法等一种或多种组合。求解计算流场的流体可以是无粘的也可以是有粘性的,可以是可压缩流体也可以是不可压缩流体,可以是层流也可以是湍流,可以是定常流动也可以是非定常流动。可以根据所模拟流体的物理特性,
相应地选择相应的控制方程以及模拟方法。例如,针对无粘流体的流场计算可以选用Euler方程组或者格子玻尔兹曼方程,针对有粘性流体的流场计算可以选用N-S方程组或者玻尔兹曼方程。优选地,中等气压放电等离子体工艺(例如,PECVD工艺),通过数值差分技术求解中性气体的N-S方程组确定工艺腔室中流场和热场的分布。进一步优选地,低气压放电等离子体工艺(例如,等离子体刻蚀及PVD工艺),通过LBM求解确定流场和热场的分布。更优选地,低气压放电等离子体工艺,通过LBM和Monte Carlo方法求解确定流场和热场的分布。针对湍流的数值模拟方法包括但不限于直接数值模拟(Direct Numerical Simulation,简称DNS)、大涡模拟法(Large Eddy Simulation,简称LES)、雷诺平均法(Reynolds-averaged Navier-Stokes equations,简称RANS)、分离涡流模拟(Detached Eddy Simulation,简称DES)等一种或多种组合。求解计算的湍流模型可以包括但不限于零方程模型、一方程模型、Spalart-Allmaras模型、k-epsilon模型、k-omega模型、RNG(Renormalization group)k-epsilon模型、Realizable k-epsilon模型、Reynolds应力方程模型(Reynolds S tress equation Model,简称RSM)、代数应力方程模型(Algebraic Stress equation Model,简称ASM)、亚格子尺度模型(Subgrid-scale Model,简称SGS)、BGK(Bhatnagar-Gross-Krook)模型、MRT-LBM(Multi-relaxation-time Lattice Boltzmann Method)模型、SRT-LBM(Single-relaxation-time Lattice Boltzmann Method)模型、格子玻尔兹曼模型(例如DmQn模型)等一种或多种组合。在一个实施例中,求解计算的控制方程是N-S方程组,离散方法是有限体积法。在另一个实施例中,求解计算的控制方程是Boltzmann-BGK方程,离散方法是对时间和空间离散的方法,而采用的离散速度模型是DmQn模型。连续的Boltzmann方程式如下所示,
其中,∫(f′1f′2-f1f2)g dΩdξ2为碰撞项。BGK模型可以简化Boltzmann方程碰撞项,其中碰撞的过程是改变分布函数f使其趋向平衡态分布函数f(eq)的过程,Boltzmann-BGK方程式如下所示,
其中,λ为分布函数的弛豫时间,f(eq)是Maxwell-Boltzmann局部平衡分布函数,
其中,R表示玻尔兹曼常数,T表示温度。离散速度模型DmQn模型可以把平衡分布函数表示如下,
其中,ωi为权系数,可以是粒子速度的函数;为格子声速。离散速度模型DmQn模型可以包括但不限于D1Q3、D1Q5、D2Q7、D2Q9、D3Q15、D3Q19等一种或多种组合。在一个实施例中,离散速度模型是D2Q9。在另一个实施例中,离散速度模型是D3Q19。
结果分析模块204可以用于分析计算结果。分析的计算结果可以包括但不限于速度、压力、密度、温度、整体流场的宏观物理量、指定截面的宏观物理量等宏观物理量。分析的计算结果也可以包括,但不限于,诸微观物理量,如分布函数,以及平衡分布函数的局部及全局表现,局部粘性系数的计算与比对,局部声速的计算与比对。同时,分析的计算结果也可以包括,但不限于,气动力、气动力矩、计算域的几何模型及网格分析、矢量图(如速度矢量线)、等值线图、填充型的等值线图(云图)、XY散点图、粒子轨迹图、模拟流动效果、图像处理功能等一种或多种组合。结果分析模块204分析计算的结果形式可以包括但不限于文字、图片、动画等一种或多种组合。计算结果的类型可以包括但不限于txt、ASCII、MIME等一种或多种组合。计算结果的文件格式可以包括但不限于csv、pdf、doc、epub、mobi、caj、kdh、nh、bmp、jpg、png、jpeg、tiff、gif、mng、xpm、psd、psp、ufo、xcf、pcx、ppm、ps、eps、ai、fh、swf、fla、wmf、svg、dxf、cgm、ai等一种或多种组合。结果分析模块204可以是对计算求解后的结果进行分析,也可以是对网格处理后的结果直接分析,也可以是对系统数据库205中已存储的结果进行分析。
系统数据库205或系统内的其他存储设备泛指所有可以具有读/写功能的媒介。系统数据库205或系统内其他存储设备可以是系统内部的,也可以是系统的外接设备。系统数据库205或系统内其他存储设备的连接方式可以是有线的,也
可以是无线的。系统数据库205或系统内其他存储设备可以包括但不限于层次式数据库、网络式数据库和关系式数据库等其中一种或多种组合。系统数据库205或系统内其他存储设备可以将信息数字化后再以利用电、磁或光学等方式的存储设备加以存储。系统数据库205或系统内其他存储设备可以用来存放各种信息例如程序和数据等。系统数据库205或系统内其他存储设备可以是利用电能方式存储信息的设备,例如各种存储器、随机存取存储器(Random Access Memory,RAM)、只读存储器(Read Only Memory,ROM)等。系统数据库205或系统内其他存储设备可以是利用磁能方式存储信息的设备,例如硬盘、软盘、磁带、磁芯存储器、磁泡存储器、U盘、闪存等。系统数据库205或系统内其他存储设备可以是利用光学方式存储信息的设备,例如CD或DVD等。系统数据库205或系统内其他存储设备可以是利用磁光方式存储信息的设备,例如磁光盘等。系统数据库205或系统内其他存储设备的存取方式可以是随机存储、串行访问存储、只读存储等一种或多种组合。系统数据库205或系统内其他存储设备可以是非永久记忆存储器,也可以是永久记忆存储器。以上提及的存储设备是列举了一些例子,该系统可以使用的存储设备并不局限于此。
系统数据库205发送的信息可以是直接从客户端系统103获取的数据,也可以是经过处理分析后的信息。经过处理分析的信息,可以是经过网格处理模块202处理后储存在系统数据库205的信息,也可以是经过求解计算模块203处理后的,也可以是经过结果分析模块204储存的信息。系统数据库205或系统内其他存储设备可以是本地的,也可以是远程的,也可以是云服务器上的。系统数据库205或系统内其他存储设备与其他模块信息传递的方式可以是有线的也可以是无线的,可以是直接的也可以是间接的,可以是同时进行的也可以是顺序进行的,可以是周期的也可以是非周期的等。
显然,对于本领域的专业人员来说,在了解数值计算仿真系统及方法的原理后,可能在不背离这一原理的情况下,对各个模块进行任意组合,或者构成子系统与其它模块连接,对实施上述方法和系统的应用领域形式和细节上的各种修正和改变,但是这些修正和改变仍在以上描述的范围之内。例如,数据接收模块201、网格处理模块202、计算求解模块203、结果分析模块204和系统数据库205可以是体现在一个系统中的不同模块,也可以是一个模块实现上述的两个或两个以
上模块的功能,例如,计算求解模块203可以直接接收数据接收模块201输入的几何文件,该计算求解模块203可以处理网格同时也可以求解方程并分析结果,计算求解模块同时实现了网格处理模块202和结果分析模块204的功能,类似的变形仍在本发明的权利要求保护范围之内。又例如,结果分析模块205的分析结果可以直接显示在客户端系统103,也可以储存在系统数据库205中,需要查看该显示结果时,可以从系统数据库中直接读取。再例如,结果分析模块205的分析结果直接显示在客户端系统103时,可以在同一个客户端系统103显示,也可以在不同的客户端系统103显示。
图3所示的是数值计算仿真流程图。数据在步骤301从客户端系统103中接收。客户端系统103可以包括但不限于服务器、通信终端一种或多种组合。进一步地,服务器可以是web服务器、文件服务器、数据库服务器、FTP服务器、应用程序服务器、代理服务器器等,或者上述服务器的任意组合。通信终端可以是手机、个人电脑、可穿戴设备、平板电脑、智能电视等,或则上述通信终端的任意组合。步骤301获取的数据可以包括但不限于网格结构、材料属性、气体特性参数、几何文件、控制参数、已处理后的网格文件等一种或多种组合。接收的文件格式可以包括但不限于txt、ASCII、MIME等一种或多种组合。控制参数可以包括但不限于网格类型、控制方程、模型、计算精度等一种或多种组合。步骤301可以由数据接收模块201完成。
步骤301接收指令和数据后进入步骤302进行操作。数据接收后进行划分网格处理和/或网格优化,网格处理包括但不限于贴体网格、分区网格、笛卡尔网格、自适应直角网格、多重网格、结合快速Delaunay的球充填法、结构化网格、非结构化网格、杂交网格等一种或多种组合。结构化网格可以是单块网格,也可以是多块网格。多块网格可以是拼接网格,也可以是嵌套网格。笛卡尔网格可以是单重的,也可以是多重的。优选地,网格处理使用结构化网格。进一步优选地,网格处理使用多块网格。更进一步优选地,网格处理使用笛卡尔网格。网格处理的操作包括但不限于网格生成、网格分区、节点属性标记和壁面距离计算等一种或多种组合。处理网格步骤302可以由网格处理模块202完成。
网格生成可以包括但不限于表面网格、四面体网格、六面体网格、棱柱体网格(边界层网格)、四面体与六面体混合网格、全局笛卡尔网格等一种或多种组
合。网格生成可以对输入的几何文件数据做适合计算使用的几何处理。网格分区可以根据已经生成的网格,经过分区生成复杂均衡的多块子网格,形成各区网格之间的数据通讯。网格分区可以包括但不限于根据外形特点分区、确定网格拓扑(例如流场变化剧烈的地方采用细密网格,远场区域选用稀疏网格)、边界曲线网格生成、边界曲面网格生成、物面网格生成、空间网格生成等一种或多种组合。网格加密技术的处理方程可以是基于格子玻尔兹曼方程的,也可以是基于离散后的速度玻尔兹曼方程。基于格子玻尔兹曼方程的网格加密技术可以包括但不限于插值、泰勒展开、最小二乘格、多块或多重网格、结合快速Delaunay的球充填法、区域分裂格子玻尔兹曼等一种或多种组合。基于离散速度玻尔兹曼方程的网格加密技术可以包括但不限于有限差分、有限体积、有限元格子玻尔兹曼法等一种或多种组合。节点属性标记可以区分包括但不限于流体节点、固体节点和边界节点。节点属性标记方法可以是根据定义节点与固体壁面边界的关系。节点位于边界外的属性可以标记为流体节点,节点位于边界上的属性可以标记为边界节点,节点位于边界内的属性可以标记为固体节点。壁面距离计算可以用来计算边界节点与固体壁面边界的距离。例如,当需要计算曲面边界条件是,需要使用壁面距离的结果。处理网格的结果可以存储在系统数据库205,也可以存储在网格处理模块202的内部存储设备。
网格优化中可以进行结构重分析、重新生成,也可以根据快速变形模块的参数改变了进行网格变形。优选地,三维非结构网格快速变形模块,根据结构优化设计参数的改变量,快速地进行网格变形,得到优化后的非结构网格。三维非结构网格快速变形模块可以不改变网格的节点数量、单元数量和/或节点间的拓扑连接关系。原计算网格的边界点可以生成背景网格,其次对每个计算网格的节点进行定位,即确定每个计算节点在背景网格中的相对位置;然后根据优化设计参数的变化确定结构边界点的位置变化量(变形量),并由此获得经优化修改(变形)后的背景网格;再由每个计算节点在背景网格中的相对位置计算节点的新位置,即得到经过优化修改后的非结构网格。网格优化的步骤可以由网格处理模块202实现。网格优化的结果可以存储在系统数据库205,也可以存储在网格处理模块202的内部存储设备。系统获取网格处理后的结果后,可以进行流场求解计算(步骤303)。流场求解计算包括但不限于网格数据与参数输入、初始化计算、
求解器单元的选择确定、湍流模型选择与实施、边界条件修正和结果输出等一种或多种组合。流场的求解计算可以由计算求解模块203完成。网格数据与参数输入选择的计算模式可以是串行的,也可以是并行的。并行计算模式下,各个进程不分主次,分别读入格子对应的子网格数据。参数输入包括但不限于确定模型、压力、密度、温度、速度、边界条件、平衡态分布函数等一种或多种组合。求解计算的湍流模型可以包括但不限于零方程模型、一方程模型、Spalart-Allmaras模型、k-epsilon模型、k-omega模型、RNG k-epsilon模型、Realizable k-epsilon模型、RSM模型、ASM模型、SGS模型、BGK模型、MRT-LBM模型、SRT-LBM模型、格子玻尔兹曼模型(例如DmQn模型)、不可压格子玻尔兹曼模型、热格子玻尔兹曼模型、非均匀网格的格子玻尔兹曼模型等一种或多种组合。求解器单元可以根据不同的控制方程来确定。控制方程可以是Euler方程,也可以是N-S方程,也可以是格子玻尔兹曼方程。作为本披露的具体实施例之一,格子玻尔兹曼方程的求解器单元可以将速度空间的分布函数转换到矩空间,结合求解计算的模型对粘性进行修正,在矩空间进行碰撞,将分布函数转换回速度空间,完成分布函数的迁移。边界条件类型可以包括但不限于速度边界条件、压力边界条件、周期性边界条件、标准反弹格式边界条件、修正反弹格式边界条件、半步式反弹边界条件、曲面边界条件等一种或多种组合。边界条件的处理方法包括但不限于固体壁面反弹与质量动量结合法、固体壁面反弹加调整分步法、水动力条件法、虚拟固体壁面法、固体壁面反弹法、插值法等一种或多种组合。结果输出可以是计算得到的各个节点上分布函数的直接结果,也可以是将各个节点上的分布函数转化为宏观参数的结果。
求解计算的结果提取后输入到分析系统(如结果分析模块),对结果进行进一步的分析(步骤304),进一步的分析结果可以输出包括但不限于可视化的图、动画、数据文件、结果数据表等一种或多种组合。需要注意的是,上述对数值计算仿真系统流程的描述只是为了便于理解发明,不应被视为是本发明唯一可行的实施例。数值计算仿真系统也可以直接将接收的已处理网格文件进行求解计算(步骤303),然后对求解计算的结果进行分析(步骤304)。可选择地,数值计算仿真系统也可以直接对接收的求解计算结果进行分析(步骤304)。
图4所示的是网格处理模块的结构示意图。网格处理模块202可以包含但不限于一个或多个网格生成单元401、一个或多个网格分区单元402、一个或多个节点属性标记单元403、一个或多个壁面距离计算单元404、一个或多个网格优化单元405。网格处理模块202中的各个单元可以是独立的,也可以是部分单元合并一个单元的。网格处理模块202中的各个单元可以是本地的,也可以是远程的。网格处理模块202中,网格生成单元401可以对数据接收模块201的输入数据进行处理,根据网格结构、几何文件和控制参数的定义,生成适合计算使用的网格。生成的网格可以包括但不限于表面网格、四面体网格、六面体网格、棱柱体网格(边界层网格)、四面体与六面体混合网格、全局笛卡尔网格等一种或多种组合。例如,输入的数据是三维几何图形,网格生成单元401可以对该几何图形处理后生成四面体网格。网格生成的结果可以存储在系统数据库205中,也可以存储在网格生成单元401的存储设备中,也可以存储在系统内其他存储设备。
网格分区单元402可以根据已经生成的网格,经过分区生成复杂均衡的多块子网格,形成各区网格之间的数据通讯。网格分区单元402也可以确定分区后生成复杂均衡的多块子网格,形成各区网格之间的数据通讯。网格分区可以包括但不限于根据外形特点分区、确定网格拓扑(例如流场变化剧烈的地方采用细密网格,远场区域选用稀疏网格)、边界曲线网格生成、边界曲面网格生成、物面网格生成、空间网格生成等一种或多种组合。网格加密技术的处理方程可以是基于格子玻尔兹曼方程的,也可以是基于离散后的速度玻尔兹曼方程。基于格子玻尔兹曼方程的网格加密技术可以包括但不限于插值、泰勒展开、最小二乘格、多块或多重网格、结合快速Delaunay的球充填法、区域分裂格子玻尔兹曼等一种或多种组合。基于离散速度玻尔兹曼方程的网格加密技术可以包括但不限于有限差分、有限体积、有限元格子玻尔兹曼法等一种或多种组合。例如,网格加密方法是多块网格,该方法基于标准格子玻尔兹曼方程,计算区域由几块网格组成,不同块内采用不同分辨率的网格,不同块之间通过边界连接。网格分区的结果可以存储在系统数据库205中,也可以存储在网格生成单元401的存储设备中,也可以存储在系统内其他存储设备。
节点属性标记单元403可以对网格边界进行属性标记。标记的节点属性可以包括但不限于流体节点、固体节点和边界节点。节点属性标记可以是在网格分区后标记,也可以是在网格分区前标记。节点属性标记的方法可以是根据节点与固体壁面边界的关系。例如,节点位于边界外的属性可以标记为流体节点,节点位于边界上的属性可以标记为边界节点,节点位于边界内的属性可以标记为固体节点。例如,节点属性标记单元403对生成后的网格进行属性标记。节点属性标记的结果可以存储在系统数据库205中,也可以存储在网格生成单元401的存储设备中,也可以存储在系统内其他存储设备。
壁面距离计算单元404可以用来计算标记后的节点与固体壁面边界的距离。壁面距离计算可以是针对湍流计算的,也可以是针对层流计算的。优选地,湍流计算时,湍流在近壁面演变为层流,使用壁面计算单元404计算节点与固体壁面边界的距离。标记后的节点包括但不限于流体节点、固体节点、边界节点等一种或多种组合。例如,当需要计算曲面边界条件是,需要使用壁面距离的结果。壁面距离的计算方法可以包括但不限于Possion方程方法、Eikonal方程方法、Hamilton-Jacobi方程方法等一种或多种组合。壁面距离计算的结果可以存储在系统数据库205,也可以存储在网格处理模块202的内部存储设备。
网格优化单元405可以用于包括但不限于贴体网格、分区网格、自适应直角网格、多重网格、结构三维网格的质量优化等一种或多种组合。网格优化包括但不限于几何优化(修匀)、拓扑优化等一种或多种组合。几何优化可以移动节点位置,保持节点连接关系不变。拓扑优化可以优化节点之间的连接关系。优选地,网格优化采用几何优化和拓扑优化相结合。几何优化可以采用包括但不限于基于插值误差的网格质量修匀技术、单元变形技术等一种或多种组合。单元变形技术可以同时移动一个单元的所有节点位置。优选地,几何优化可以采用基于插值误差的网格质量修匀技术结合单元变形技术的快速网格质量修匀算法。拓扑优化可以采用基于前沿推进策略的小多面体最优剖分SPR技术。网格优化单元405在结构优化中可以进行结构重分析、重新生成,也可以根据快速变形模块的参数改变了进行网格变形。优选地,三维非结构网格快速变形模块,根据结构优化设计参数的改变量,快速地进行网格变形,得到优化后的非结构网格。三维非结构网格快速变形模块可以不改变网格的节点数量、单元数量和/或节点间的拓扑连接
关系。原计算网格的边界点可以生成背景网格,其次对每个计算网格的节点进行定位,即确定每个计算节点在背景网格中的相对位置;然后根据优化设计参数的变化确定结构边界点的位置变化量(变形量),并由此获得经优化修改(变形)后的背景网格;再由每个计算节点在背景网格中的相对位置计算节点的新位置,即得到经过优化修改后的非结构网格。
需要注意的是,以上对网格处理模块202的描述仅仅是本发明的具体实施例,不应被视为是唯一的实施例。显然,对于本领域的专业人员来说,在了解本发明内容和原理后,都可能在不背离本发明原理、结构的情况下,进行形式和细节上的各种修正和改变,但是这些修正和改变仍在本发明的权利要求保护范围之内。
图5所示的是计算求解模块的结构示意图。计算求解模块203可以包括一个或多个收集单元501、一个或多个初始化单元502、一个或多个求解器单元503、一个或多个湍流单元504、一个或多个边界条件单元505、一个或多个结果输出单元506。计算求解模块203中的各个单元可以是独立的,也可以是部分单元合并一个单元的。计算求解模块203中的各个单元可以是本地的,也可以是远程的。收集单元501接收计算求解的数据输入。输入的数据可以是来源于网格生成模块202的,也可以是来源于数据接收模块201的,也可以是来源于系统数据库205的。输入的数据可以包括但不限于网格数据、参数、计算模式、模型、边界条件、精度要求等一种或多种组合。网格数据可以是处理过的,也可以是没有处理过的。计算模式可以是串行的,也可以是并行的。并行计算模式下,各个进程不分主次,分别读入格子对应的子网格数据。参数输入包括但不限于确定模型、压力、密度、温度、速度、边界条件、平衡态分布函数等一种或多种组合。
初始化单元502可以把收集单元501输入的宏观物理量作为初始条件。初始化单元502可以执行的操作包括但不限于初始化并行进程编号、初始化每个进程对应的计算区域、初始化流场的宏观物理量、初始化分布函数等一种或多种组合。初始化的物理量包括但不限于压力、密度、温度、速度等一种或多种的组合。求解器单元503可以根据控制方程确定不同的求解器单元。控制方程可以是Euler方程,也可以是N-S方程,也可以是格子玻尔兹曼方程。作为本披露的一个具体实施例之一,基于N-S方程的求解器单元可以借助基函数来近似待求的流动变
量,将该近似关系代入控制方程形成离散方程组,完成数值求解。基于格子玻尔兹曼方程的求解器单元可以将速度空间的分布函数转换到矩空间,结合求解计算的模型对粘性进行修正,在矩空间进行碰撞,将分布函数转换回速度空间,完成分布函数的迁移。
湍流单元504可以通过求解计算的模型修正弛豫时间。湍流单元504可以对求解器单元503的求解计算模型进行修正。求解计算的数值模拟方法可以是直接数值模拟,也可以是非直接数值模拟。非直接数值模拟可以包括但不限于大涡模拟方法、Reynolds平均法、统计平均法等一种或多种组合。求解计算的湍流模型可以包括但不限于零方程模型、一方程模型、Spalart-Allmaras模型、k-epsilon模型、k-omega模型、RNG k-epsilon模型、Realizable k-epsilon模型、RSM模型、ASM模型、SGS模型、BGK模型、MRT-LBM模型、SRT-LBM模型、格子玻尔兹曼模型(例如DmQn模型)、不可压格子玻尔兹曼模型、热格子玻尔兹曼模型、非均匀网格的格子玻尔兹曼模型等一种或多种组合。例如,数值仿真高雷诺数流动时,求解计算的模型可以加入大涡模拟方法。
边界条件单元505对边界上的分布函数做出修正。分布函数可以是未经过碰撞迁移的,也可以是碰撞迁移后的。边界条件类型可以包括但不限于速度边界条件、压力边界条件、周期性边界条件、标准反弹格式边界条件、修正反弹格式边界条件、半步式反弹边界条件、曲面边界条件等一种或多种组合。边界条件的处理方法包括但不限于固体壁面反弹与质量动量结合法、固体壁面反弹加调整分步法、水动力条件法、虚拟固体壁面法、固体壁面反弹法、插值法等一种或多种组合。例如,压力边界条件时,边界条件单元505使用的格式可以是非平衡态外推压力格式。
结果输出单元506可以输出通过求解计算后得到的结果。输出的结果可以是计算得到的各个节点上分布函数的直接结果,也可以是将各个节点上的分布函数转化为宏观参数的结果。转化为宏观参数的结果可以包括但不限于速度、压力、密度、温度、整体流场的宏观物理量、指定截面的宏观物理量、气动力、气动力矩、计算域的几何模型及网格分析、矢量图(如速度矢量线)、等值线图、填充型的等值线图(云图)、XY散点图、粒子轨迹图等一种或多种组合。
需要注意的是,以上对计算求解模块203的描述仅仅是具体的示例,不应被视为是唯一可行的实施方案。显然,对于本领域的专业人员来说,在了解所需要的信息的基本原理后,可能在不背离这一原理的情况下,对所需要的信息的内容进行各种修正和改变,但是这些修正和改变仍在以上描述的范围之内。
图6所示的是数值计算仿真系统的一种计算求解流程图。步骤601收集数据。接收的数据可以是来源于网格生成模块202的,也可以是来源于数据接收模块201的,也可以是来源于系统数据库205的。步骤601可以由收集单元501完成。收集的数据可以包括但不限于网格数据、参数、计算模式、模型、边界条件、精度要求等一种或多种组合。网格数据可以是处理过的,也可以是没有处理过的。计算模式可以是串行的,也可以是并行的。并行计算模式下,各个进程不分主次,分别读入格子对应的子网格数据。参数输入包括但不限于确定控制方程、模型、压力、密度、温度、速度、边界条件、平衡态分布函数、步长要求等一种或多种组合。例如,选择的控制方程是Boltzmann-BGK方程,模型是DmQn。步骤602根据收集的数据初始化流场。初始化流场的内容可以包括但不限于初始化并行进程编号、初始化每个进程对应的计算区域、初始化流场的宏观物理量、初始化分布函数等一种或多种组合。初始化的物理量包括但不限于压力、密度、温度、速度等一种或多种的组合。步骤602可以由初始化单元502完成。
步骤603中根据初始化流场和/参数来确定求解器单元。步骤603可以由求解器单元503完成。进一步地,步骤604可以求解控制方程。控制方程可以是Euler方程,也可以是N-S方程,也可以是格子玻尔兹曼方程。控制方程是N-S方程,其求解计算模型可以包括但不限于零方程模型、一方程模型、Spalart-Allmaras模型、k-epsilon模型、k-omega模型、RNG k-epsilon模型、Realizable k-epsilon模型、RSM模型、ASM模型、SGS模型等一种或多种组合。控制方程是格子玻尔兹曼方程,其求解计算模型可以包括但不限于BGK模型、MRT-LBM模型、SRT-LBM模型、格子玻尔兹曼模型(例如DmQn模型)、不可压格子玻尔兹曼模型、热格子玻尔兹曼模型、非均匀网格的格子玻尔兹曼模型等一种或多种组合。进一步地,玻尔兹曼方程的平衡分布函数可以通过离散速度模型确定。优选地,离散速度模型是DmQn模型。步骤604中,更进一步地,可以通过求解计算模型来修正粘性,通过粘性进一步确定分布函数的弛豫时间。步骤604可以
由求解器单元503和湍流单元504完成。例如,弛豫时间反映分布函数因碰撞趋近平衡的速率,同时也可以利用湍流模型对其进行修正。
步骤605处理边界条件,对求解计算的分布函数做出修正。步骤605可以由边界条件单元505完成。分布函数可以是未经过碰撞迁移的,也可以是碰撞迁移后的。边界条件类型可以包括但不限于速度边界条件、压力边界条件、周期性边界条件、标准反弹格式边界条件、修正反弹格式边界条件、半步式反弹边界条件、曲面边界条件等一种或多种组合。边界条件的处理方法包括但不限于固体壁面反弹与质量动量结合法、固体壁面反弹加调整分步法、水动力条件法、虚拟固体壁面法、固体壁面反弹法、插值法等一种或多种组合。步骤606对计算后各个节点上的分布函数更新为宏观物理量。宏观物理量包括但不限于速度、压力、密度等一种或多种组合。步骤606可以由结果输出单元506完成。步骤607判断求解计算的如果是定常问题则进入步骤609,步骤609继续判断计算的宏观物理量是否满足定常条件的判据,如果满足进入步骤610输出结果,如果不满足继续求解计算进入步骤604。步骤607判断求解计算的如果是非定常问题则进入步骤608,步骤608继续判断计算是否达到设置的步长要求,如果达到要求进入步骤610输出结果,如果不满足继续求解计算进入步骤604。上述步骤由求解器单元503、湍流单元504、边界条件单元505和结果输出单元506完成。
需要注意的是,上述对数值计算仿真系统的计算求解流程的描述只是为了便于理解发明,不应被视为唯一的实施例。显然,对于本领域的专业人员来说,在了解本发明的内容和原理之后,都可能在不背离本发明原理和结构的情况下,进行形式和细节上的各种修正和改变,但这些修正和改变仍在本发明的权利要求保护范围之内。例如,收集的数据中已经确认求解计算的问题是定常问题,步骤607就不是必须的,可以直接进入步骤609。又例如,收集的数据中控制方程是N-S方程,步骤603可以确定为相应求解N-S方程的求解器单元,或者将N-S方程转化为格子玻尔兹曼方程使用格子玻尔兹曼方程的求解器单元来求解。
图7-a和图7-b所示的是数值计算仿真系统解决圆柱绕流问题的结果示意图。二维圆柱绕流计算采用格子玻尔兹曼控制方程和D2Q9的多松弛时间模型(MRT),边界条件使用线性插值格式的曲线边界条件。计算的流场长度和宽度分别为22D和4.1D。其中,D为圆柱的直径。圆柱中心在流场中距离入口和下
边界是2D。入口边界速度按抛物线分布,出口采用压力边界,管道上下边界采用速度边界。图7-a是雷诺数为20的圆柱绕流结果。圆柱前表面附近流场的涡量最大,其他区域涡量都很小,而且上半部分为负值,下半部分为正值。圆柱的阻力系数和升力系数采用动量交换法计算得到,在15000步后达到稳定。图7-b是雷诺数为100的圆柱绕流结果。圆柱后面区域形成了非对称涡,涡量的正负值也不对称。圆柱的阻力系数和升力系数随即出现周期性的震荡。
图8-a所示的是中性气体放电等离子体的腔室结构图。中性气体包括但不限于氩气、氦气、氖气、氙气、氪气、氧气、氮气、氢气、甲烷、一氧化碳、惰性气体等一种或多种组合。在一个实施例中,选择氩气作为工作气体,气压是1Torr,电流是10A,频率是13.56MHz。在另一个实施例中,选择氧气作为放电气体,气压是1Torr,电流是10A,频率是13.56MHz。在另一个实施例中,选择氦气作为工作气体,气压是1Torr,电流是10A,频率是13.56MHz。工作气体的流速可以是500sccm~9000sccm。优选地,工作气体的流速可以是2000sccm。进一步优选地,工作气体的流速可以是3000sccm。更优选地,工作气体的流速可以是4000sccm。在一个实施例中,中性气体放电等离子体计算优选地采用无粘模型,求解算法优选地是稳态simple算法。中性气体放电等离子体的腔室的轴向和径向的网格大小均为0.2cm。入口处边界条件是恒定流速,出口处边界条件是恒定气压。计算的残差收敛精度设置可以是10-9~10-1。优选地,残差收敛精度可以是10-6。图8-b所示的是工作气体氩气在流速2000sccm时放电的流场仿真流线图。图8-c所示的是工作气体氩气在流速3000sccm时放电的流场仿真流线图。图8-d所示的是工作气体氧气在流速3000sccm时放电的流场仿真流线图。
图9所示的是实现数值计算仿真系统及方法的设备结构示意图。计算机900可以是通用的计算机,也可以是特殊目的使用计算机。计算机900可以用来实现上述描述的可以提供功能服务的任何组件。例如,数值计算仿真系统101可以通过计算机设备(例如计算机900)的硬件、软件程序等一种或多种结合后实现。例如,计算机900可以包括COM端口950和网络连接以方便数据通信。计算机900也可以包括一个中央处理器(CPU)920,可以是一个或多个处理器来执行程序指令。示例中的计算平台还可以包括一个内部通信总线910和不同形式的程序、数据存储设备。不同形式的程序、数据存储设备可以利用电、磁或光学等方
式的存储设备加以存储。例如,磁盘970、只读存储器(ROM)930、随机存取存储器(RAM)940等可以通过计算机处理和/或通信的数据文件,同样地,也可以是通过CPU执行的程序指令。计算机900还可以包括一个输入/输出(I/O)组件960,可以用于计算机和其他组件(如用户界面)之间的输入/输出。计算机900还可以通过网络接收程序和数据。
以上的描述仅仅是本发明实现的设备结构的具体实施例,不应被视为是唯一的实施例。显然,对于本领域的专业人员来说,在了解本发明内容和原理后,都可能在不背离本发明原理、结构的情况下,进行形式和细节上的各种修正和改变,但是这些修正和改变仍在本发明的权利要求保护范围之内。例如,本发明的数值计算仿真系统可以选择性地求解不同控制方程,可以是用户自定义的,也可以是系统定义的。本发明的数值计算仿真系统可以作为浏览器插件运行,也可以作为程序运行,也可以是网页版运行。
Claims (17)
- 一种模拟等离子体放电的数值计算仿真系统,包括:一种计算机可读的存储媒介,所述存储媒介存储可执行模块,包括:数据接收模块,所述数据接收模块能够接收数据;网格处理模块,所述网格处理模块能够对接收的数据进行网格剖分处理,生成网格处理结果;计算求解模块,所述计算求解模块能够根据网格处理结果和接收的数据选择适用的求解计算控制方程生成流场与热场的计算结果;结果分析模块,所述结果分析模块能够对生成的流场与热场计算结果进一步分析生成分析结果;一个处理器,所述处理器能够执行所述计算机可读的存储媒介存储的可执行模块。
- 根据权利要求1所述的数值计算仿真系统,其特征在于,进一步包括一个数据库,所述数据库能够储存所述的接收数据、网格处理结果、计算结果、分析结果。
- 根据权利要求1所述的数值计算仿真系统,其特征在于,所述的网格处理模块进一步包括网格生成单元、网格分区单元、节点属性标记单元、壁面距离计算单元、网格优化单元。
- 根据权利要求1所述的数值计算仿真系统,其特征在于,所述的控制方程包括Euler方程、N-S方程组、格子玻尔兹曼方程。
- 根据权利要求1所述的数值计算仿真系统,其特征在于,所述的计算求解模块进一步包括收集单元、初始化单元、求解器单元、湍流单元、边界条件单元、结果输出单元。
- 根据权利要求3所述的数值计算仿真系统,其特征在于,所述的网格生成单元生成的网格包括表面网格、四面体网格、六面体网格、棱柱体网格(边界层网格)、四面体与六面体混合网格、笛卡尔网格、球填充法网格。
- 根据权利要求6所述的数值计算仿真系统,其特征在于,所述的球填充法网格基于Delaunay剖分插入技术。
- 根据权利要求3所述的数值计算仿真系统,其特征在于,所述的节点属性标记单元的节点属性包括流体节点、固体节点和边界节点。
- 根据权利要求5所述的数值计算仿真系统,其特征在于,所述的湍流单元求解计算模型包括零方程模型、一方程模型、Spalart-Allmaras模型、k-epsilon模型、k-omega模型、RNG k-epsilon模型、Realizable k-epsilon模型、RSM模型、ASM模型、SGS模型、BGK模型、MRT-LBM模型、SRT-LBM模型、格子玻尔兹曼模型、不可压格子玻尔兹曼模型、热格子玻尔兹曼模型、非均匀网格的格子玻尔兹曼模型。
- 根据权利要求1所述的数值计算仿真系统,其特征在于,所述的结果分析模块进一步分析的结果包括速度、压力、密度、温度、整体流场的宏观物理量、指定截面的宏观物理量、气动力、气动力矩、计算域的几何模型及网格分析、矢量图(如速度矢量线)、等值线图、填充型的等值线图(云图)、XY散点图、粒子轨迹图、模拟流动效果、图像处理功能。
- 一种模拟等离子体放电的数值计算仿真方法,包括:接收数据;根据所述数据进行网格剖分处理;根据所述剖分处理的网格求解计算控制方程的流场和热场计算结果;进一步提取所述流场和热场的计算结果进行分析。
- 根据权利要求11所述的数值计算仿真方法,其特征在于,所述的接收数据包括网格结构、材料属性、气体特性参数、几何文件、控制参数、网格文件。
- 根据权利要求12所述的数值计算仿真方法,其特征在于,所述的控制参数包括网格类型、控制方程、模型、计算精度、残差收敛精度。
- 根据权利要求13所述的数值计算仿真方法,其特征在于,所述的网格类型包括贴体网格、分区网格、笛卡尔网格、自适应直角网格、多重网格、结合快速Delaunay的球充填法网格、结构化网格、非结构化网格、杂交网格。
- 根据权利要求14所述的数值计算仿真方法,其特征在于,所述的控制方程包括Euler方程、N-S方程、格子玻尔兹曼方程。
- 根据权利要求14所述的数值计算仿真方法,其特征在于,所述的模型包括零方程模型、一方程模型、Spalart-Allmaras模型、k-epsilon模型、k-omega 模型、RNG k-epsilon模型、Realizable k-epsilon模型、RSM模型、ASM模型、SGS模型、BGK模型、MRT-LBM模型、SRT-LBM模型、格子玻尔兹曼模型、不可压格子玻尔兹曼模型、热格子玻尔兹曼模型、非均匀网格的格子玻尔兹曼模型。
- 根据权利要求11所述的数值计算仿真方法,其特征在于,所述的接收数据包括用户输入数据和非用户输入数据,所述的非用户输入数据来源包括服务器、通信终端。
Priority Applications (1)
Application Number | Priority Date | Filing Date | Title |
---|---|---|---|
PCT/CN2015/095219 WO2017084105A1 (zh) | 2015-11-20 | 2015-11-20 | 一种数值模拟等离子体放电的系统及方法 |
Applications Claiming Priority (1)
Application Number | Priority Date | Filing Date | Title |
---|---|---|---|
PCT/CN2015/095219 WO2017084105A1 (zh) | 2015-11-20 | 2015-11-20 | 一种数值模拟等离子体放电的系统及方法 |
Publications (1)
Publication Number | Publication Date |
---|---|
WO2017084105A1 true WO2017084105A1 (zh) | 2017-05-26 |
Family
ID=58717267
Family Applications (1)
Application Number | Title | Priority Date | Filing Date |
---|---|---|---|
PCT/CN2015/095219 WO2017084105A1 (zh) | 2015-11-20 | 2015-11-20 | 一种数值模拟等离子体放电的系统及方法 |
Country Status (1)
Country | Link |
---|---|
WO (1) | WO2017084105A1 (zh) |
Cited By (30)
Publication number | Priority date | Publication date | Assignee | Title |
---|---|---|---|---|
CN108763794A (zh) * | 2018-06-01 | 2018-11-06 | 北京航空航天大学 | 一种用于电路板产品热致失效及可靠性评估的热分析方法 |
CN109271651A (zh) * | 2018-07-09 | 2019-01-25 | 广东工业大学 | 一种基于激光诱导的格子玻尔兹曼气-液两相流的仿真方法 |
CN109829184A (zh) * | 2018-12-06 | 2019-05-31 | 大族激光科技产业集团股份有限公司 | 3d打印系统吸气口管道结构及其优化设计方法、装置 |
CN110414053A (zh) * | 2019-06-27 | 2019-11-05 | 西安空间无线电技术研究所 | 一种快速确定部件微放电阈值的时域数值模拟方法 |
CN110414156A (zh) * | 2019-07-31 | 2019-11-05 | 武汉理工大学 | 一种四边简支板相对辐射声阻抗的确定方法 |
CN110765685A (zh) * | 2019-10-18 | 2020-02-07 | 南方电网科学研究院有限责任公司 | 一种电抗器多物理场耦合的仿真方法、装置及存储介质 |
CN110795898A (zh) * | 2019-10-18 | 2020-02-14 | 南方电网科学研究院有限责任公司 | 一种穿墙套管多物理场耦合的仿真方法、装置及存储介质 |
CN111027252A (zh) * | 2019-12-17 | 2020-04-17 | 西安科技大学 | 煤矿综掘面粉尘、瓦斯和风流参数化仿真模拟方法 |
CN111159941A (zh) * | 2019-12-25 | 2020-05-15 | 同济大学 | 一种用于汽车液力变矩器内流场瞬态数值模拟的方法 |
CN111259514A (zh) * | 2019-12-26 | 2020-06-09 | 兰州空间技术物理研究所 | 霍尔推力器的全流程数值仿真系统及使用该系统的全流程数值仿真方法 |
CN111581791A (zh) * | 2020-04-22 | 2020-08-25 | 中国海洋石油集团有限公司 | 一种电阻率正演仿真方法、装置及计算机可读存储介质 |
CN111611698A (zh) * | 2020-05-13 | 2020-09-01 | 广东省智能制造研究所 | 一种超薄体传热仿真方法 |
CN111813067A (zh) * | 2020-07-21 | 2020-10-23 | 秦皇岛玻璃工业研究设计院有限公司 | 基于玻璃锡槽数学模型对玻璃生产环节进行优化的方法 |
CN112035995A (zh) * | 2019-07-19 | 2020-12-04 | 交通运输部天津水运工程科学研究所 | 基于gpu计算技术的非结构网格潮汐潮流数值模拟方法 |
CN112765843A (zh) * | 2020-12-31 | 2021-05-07 | 三一重型装备有限公司 | 除尘器的结构设计方法及装置、存储介质、计算机设备 |
CN112966467A (zh) * | 2021-03-17 | 2021-06-15 | 武汉大学 | 一种基于湿法化学蚀刻联合仿真的柔性pcb板结构设计方法 |
CN113111553A (zh) * | 2021-04-09 | 2021-07-13 | 西北工业大学 | 一种基于插值变形网格的大变形运动数值模拟方法 |
CN113420379A (zh) * | 2021-06-29 | 2021-09-21 | 西北工业大学 | 一种从cfl3d计算结果中提取物面时均压力分布的方法 |
CN113792367A (zh) * | 2021-09-07 | 2021-12-14 | 清华大学 | 基于PySWMM的排水系统多来源入流入渗和出渗量动态估算方法 |
CN114155135A (zh) * | 2021-12-02 | 2022-03-08 | 西北核技术研究所 | 基于gpu集群的相对论性波尔兹曼方程计算方法、存储介质及设备 |
CN114585950A (zh) * | 2019-10-28 | 2022-06-03 | 西门子歌美飒可再生能源创新与技术有限公司 | 用于对风力涡轮机具有影响的风现象的计算机实现的预报的方法 |
CN114996858A (zh) * | 2022-07-14 | 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 | 飞行器仿真方法、装置、终端设备和存储介质 |
CN115017788A (zh) * | 2022-06-22 | 2022-09-06 | 东南大学 | 一种中性气体容性耦合等离子体放电的数值模拟方法 |
CN115587552A (zh) * | 2022-12-12 | 2023-01-10 | 中国空气动力研究与发展中心计算空气动力研究所 | 网格优化方法、装置、终端设备及存储介质 |
CN115906718A (zh) * | 2023-03-09 | 2023-04-04 | 陕西空天信息技术有限公司 | 一种旋转机械cfd系统 |
CN115906714A (zh) * | 2023-02-22 | 2023-04-04 | 拉普拉斯(无锡)半导体科技有限公司 | 一种板式pecvd反应腔气流模拟方法及其模拟设备 |
CN116611370A (zh) * | 2023-07-19 | 2023-08-18 | 东方空间技术(山东)有限公司 | 一种发射台导流模型的仿真分析方法、装置及计算设备 |
CN117350137A (zh) * | 2023-12-05 | 2024-01-05 | 山东理工大学 | 一种放电等离子体通道瞬态特性的有限元仿真方法 |
CN117973100A (zh) * | 2024-04-02 | 2024-05-03 | 上海华模科技有限公司 | 一种基于数值模拟的百叶格栅叶片外形设计方法 |
CN111199118B (zh) * | 2019-12-13 | 2024-05-03 | 同济大学 | 非齐平双涵喷口喷流噪声计算方法、计算装置及存储介质 |
Citations (3)
Publication number | Priority date | Publication date | Assignee | Title |
---|---|---|---|---|
CN102034006A (zh) * | 2010-12-16 | 2011-04-27 | 上海奕洁汽车科技有限公司 | 一种基于有限元法的蓄电池热管理分析及优化方法 |
CN103617367A (zh) * | 2013-12-06 | 2014-03-05 | 三峡大学 | 电磁场-流场-温度场耦合计算中的异型网格映射方法 |
CN104679936A (zh) * | 2014-12-05 | 2015-06-03 | 国家电网公司 | 永磁饱和型故障限流器磁流热耦合建模方法 |
-
2015
- 2015-11-20 WO PCT/CN2015/095219 patent/WO2017084105A1/zh active Application Filing
Patent Citations (3)
Publication number | Priority date | Publication date | Assignee | Title |
---|---|---|---|---|
CN102034006A (zh) * | 2010-12-16 | 2011-04-27 | 上海奕洁汽车科技有限公司 | 一种基于有限元法的蓄电池热管理分析及优化方法 |
CN103617367A (zh) * | 2013-12-06 | 2014-03-05 | 三峡大学 | 电磁场-流场-温度场耦合计算中的异型网格映射方法 |
CN104679936A (zh) * | 2014-12-05 | 2015-06-03 | 国家电网公司 | 永磁饱和型故障限流器磁流热耦合建模方法 |
Non-Patent Citations (2)
Title |
---|
CHEN, HUANMIN G ET AL.: "Numerical Simulation of Keyholing Puddle during Plasma Arc Welding", JOURNAL OF NANCHANG INSTITUTE OF AERONAUTICAL TECHNOLOGY, vol. 13, no. 2, 30 June 1999 (1999-06-30) * |
LEI, YUCHENG ET AL.: "Simulation of Molten Pool for Vertical Keyhole Plasma Arc Welding in Aluminium Alloy", TRANSACTIONS OF THE CHINA WELDING INSTITUTION, vol. 24, no. 1, 28 February 2003 (2003-02-28) * |
Cited By (46)
Publication number | Priority date | Publication date | Assignee | Title |
---|---|---|---|---|
CN108763794A (zh) * | 2018-06-01 | 2018-11-06 | 北京航空航天大学 | 一种用于电路板产品热致失效及可靠性评估的热分析方法 |
CN109271651A (zh) * | 2018-07-09 | 2019-01-25 | 广东工业大学 | 一种基于激光诱导的格子玻尔兹曼气-液两相流的仿真方法 |
CN109271651B (zh) * | 2018-07-09 | 2023-01-20 | 广东工业大学 | 一种基于激光诱导的格子玻尔兹曼气-液两相流的仿真方法 |
CN109829184A (zh) * | 2018-12-06 | 2019-05-31 | 大族激光科技产业集团股份有限公司 | 3d打印系统吸气口管道结构及其优化设计方法、装置 |
CN110414053A (zh) * | 2019-06-27 | 2019-11-05 | 西安空间无线电技术研究所 | 一种快速确定部件微放电阈值的时域数值模拟方法 |
CN112035995A (zh) * | 2019-07-19 | 2020-12-04 | 交通运输部天津水运工程科学研究所 | 基于gpu计算技术的非结构网格潮汐潮流数值模拟方法 |
CN110414156A (zh) * | 2019-07-31 | 2019-11-05 | 武汉理工大学 | 一种四边简支板相对辐射声阻抗的确定方法 |
CN110414156B (zh) * | 2019-07-31 | 2023-06-16 | 武汉理工大学 | 一种四边简支板相对辐射声阻抗的确定方法 |
CN110765685A (zh) * | 2019-10-18 | 2020-02-07 | 南方电网科学研究院有限责任公司 | 一种电抗器多物理场耦合的仿真方法、装置及存储介质 |
CN110795898A (zh) * | 2019-10-18 | 2020-02-14 | 南方电网科学研究院有限责任公司 | 一种穿墙套管多物理场耦合的仿真方法、装置及存储介质 |
CN114585950A (zh) * | 2019-10-28 | 2022-06-03 | 西门子歌美飒可再生能源创新与技术有限公司 | 用于对风力涡轮机具有影响的风现象的计算机实现的预报的方法 |
CN111199118B (zh) * | 2019-12-13 | 2024-05-03 | 同济大学 | 非齐平双涵喷口喷流噪声计算方法、计算装置及存储介质 |
CN111027252A (zh) * | 2019-12-17 | 2020-04-17 | 西安科技大学 | 煤矿综掘面粉尘、瓦斯和风流参数化仿真模拟方法 |
CN111027252B (zh) * | 2019-12-17 | 2023-04-11 | 西安科技大学 | 煤矿综掘面粉尘、瓦斯和风流参数化仿真模拟方法 |
CN111159941A (zh) * | 2019-12-25 | 2020-05-15 | 同济大学 | 一种用于汽车液力变矩器内流场瞬态数值模拟的方法 |
CN111159941B (zh) * | 2019-12-25 | 2024-03-22 | 同济大学 | 一种用于汽车液力变矩器内流场瞬态数值模拟的方法 |
CN111259514B (zh) * | 2019-12-26 | 2022-11-25 | 兰州空间技术物理研究所 | 霍尔推力器的全流程数值仿真系统及使用该系统的全流程数值仿真方法 |
CN111259514A (zh) * | 2019-12-26 | 2020-06-09 | 兰州空间技术物理研究所 | 霍尔推力器的全流程数值仿真系统及使用该系统的全流程数值仿真方法 |
CN111581791A (zh) * | 2020-04-22 | 2020-08-25 | 中国海洋石油集团有限公司 | 一种电阻率正演仿真方法、装置及计算机可读存储介质 |
CN111581791B (zh) * | 2020-04-22 | 2024-05-31 | 中国海洋石油集团有限公司 | 一种电阻率正演仿真方法、装置及计算机可读存储介质 |
CN111611698A (zh) * | 2020-05-13 | 2020-09-01 | 广东省智能制造研究所 | 一种超薄体传热仿真方法 |
CN111611698B (zh) * | 2020-05-13 | 2023-03-24 | 广东省智能制造研究所 | 一种超薄体传热仿真方法 |
CN111813067A (zh) * | 2020-07-21 | 2020-10-23 | 秦皇岛玻璃工业研究设计院有限公司 | 基于玻璃锡槽数学模型对玻璃生产环节进行优化的方法 |
CN111813067B (zh) * | 2020-07-21 | 2024-02-02 | 秦皇岛玻璃工业研究设计院有限公司 | 基于玻璃锡槽数学模型对玻璃生产环节进行优化的方法 |
CN112765843B (zh) * | 2020-12-31 | 2024-05-14 | 三一重型装备有限公司 | 除尘器的结构设计方法及装置、存储介质、计算机设备 |
CN112765843A (zh) * | 2020-12-31 | 2021-05-07 | 三一重型装备有限公司 | 除尘器的结构设计方法及装置、存储介质、计算机设备 |
CN112966467B (zh) * | 2021-03-17 | 2022-04-29 | 武汉大学 | 一种基于湿法化学蚀刻联合仿真的柔性pcb板结构设计方法 |
CN112966467A (zh) * | 2021-03-17 | 2021-06-15 | 武汉大学 | 一种基于湿法化学蚀刻联合仿真的柔性pcb板结构设计方法 |
CN113111553A (zh) * | 2021-04-09 | 2021-07-13 | 西北工业大学 | 一种基于插值变形网格的大变形运动数值模拟方法 |
CN113111553B (zh) * | 2021-04-09 | 2023-08-29 | 西北工业大学 | 一种基于插值变形网格的大变形运动数值模拟方法 |
CN113420379A (zh) * | 2021-06-29 | 2021-09-21 | 西北工业大学 | 一种从cfl3d计算结果中提取物面时均压力分布的方法 |
CN113792367B (zh) * | 2021-09-07 | 2022-08-05 | 清华大学 | 基于PySWMM的排水系统多来源入流入渗和出渗量动态估算方法 |
CN113792367A (zh) * | 2021-09-07 | 2021-12-14 | 清华大学 | 基于PySWMM的排水系统多来源入流入渗和出渗量动态估算方法 |
CN114155135A (zh) * | 2021-12-02 | 2022-03-08 | 西北核技术研究所 | 基于gpu集群的相对论性波尔兹曼方程计算方法、存储介质及设备 |
CN115017788A (zh) * | 2022-06-22 | 2022-09-06 | 东南大学 | 一种中性气体容性耦合等离子体放电的数值模拟方法 |
CN114996858A (zh) * | 2022-07-14 | 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 | 飞行器仿真方法、装置、终端设备和存储介质 |
CN115587552B (zh) * | 2022-12-12 | 2023-03-28 | 中国空气动力研究与发展中心计算空气动力研究所 | 网格优化方法、装置、终端设备及存储介质 |
CN115587552A (zh) * | 2022-12-12 | 2023-01-10 | 中国空气动力研究与发展中心计算空气动力研究所 | 网格优化方法、装置、终端设备及存储介质 |
CN115906714A (zh) * | 2023-02-22 | 2023-04-04 | 拉普拉斯(无锡)半导体科技有限公司 | 一种板式pecvd反应腔气流模拟方法及其模拟设备 |
CN115906718B (zh) * | 2023-03-09 | 2023-05-23 | 陕西空天信息技术有限公司 | 一种旋转机械cfd系统 |
CN115906718A (zh) * | 2023-03-09 | 2023-04-04 | 陕西空天信息技术有限公司 | 一种旋转机械cfd系统 |
CN116611370A (zh) * | 2023-07-19 | 2023-08-18 | 东方空间技术(山东)有限公司 | 一种发射台导流模型的仿真分析方法、装置及计算设备 |
CN116611370B (zh) * | 2023-07-19 | 2023-09-19 | 东方空间技术(山东)有限公司 | 一种发射台导流模型的仿真分析方法、装置及计算设备 |
CN117350137B (zh) * | 2023-12-05 | 2024-03-01 | 山东理工大学 | 一种放电等离子体通道瞬态特性的有限元仿真方法 |
CN117350137A (zh) * | 2023-12-05 | 2024-01-05 | 山东理工大学 | 一种放电等离子体通道瞬态特性的有限元仿真方法 |
CN117973100A (zh) * | 2024-04-02 | 2024-05-03 | 上海华模科技有限公司 | 一种基于数值模拟的百叶格栅叶片外形设计方法 |
Similar Documents
Publication | Publication Date | Title |
---|---|---|
WO2017084105A1 (zh) | 一种数值模拟等离子体放电的系统及方法 | |
WO2017084106A1 (zh) | 一种数值模拟飞行器流场的系统及方法 | |
Mohammadzadeh et al. | A parallel DSMC investigation of monatomic/diatomic gas flows in a micro/nano cavity | |
Luo et al. | A reduced FVE formulation based on POD method and error analysis for two-dimensional viscoelastic problem | |
Ceze et al. | Drag prediction using adaptive discontinuous finite elements | |
Pandya et al. | Assessment of USM3D Hierarchical Adaptive Nonlinear Method Preconditioners for Three-Dimensional Cases | |
McDonald et al. | Application of Gaussian moment closure to microscale flows with moving embedded boundaries | |
Park et al. | Hybrid grid generation for viscous flow analysis | |
Zhan et al. | Meshfree method based on discrete gas-kinetic scheme to simulate incompressible/compressible flows | |
Palha et al. | A hybrid Eulerian-Lagrangian flow solver | |
Selimefendigil | Numerical analysis and pod based interpolation of mixed convection heat transfer in horizontal channel with cavity heated from below | |
Borisov et al. | Parallel implementation of an implicit scheme based on the LU-SGS method for 3D turbulent flows | |
My-Ha et al. | Real-time optimization using proper orthogonal decomposition: Free surface shape prediction due to underwater bubble dynamics | |
Emre Tekaslan et al. | Implementation of multidisciplinary and multifidelity uncertainty quantification methods for sonic boom prediction | |
Yildirim et al. | A parallel implementation of fluid–solid interaction solver using an immersed boundary method | |
Yang et al. | A simple gas kinetic scheme for simulation of 3D incompressible thermal flows | |
Zheng et al. | Block-based adaptive mesh refinement finite-volume scheme for hybrid multi-block meshes | |
Sandberg | Direct numerical simulations for flow and noise studies | |
Mallikarjun et al. | Hybrid Navier–Stokes–Direct Simulation Monte Carlo Automatic Mesh Optimization for Hypersonics | |
Mallikarjun et al. | Direct Simulation Monte Carlo Methods for Hypersonic Flows with Automatic Mesh Optimization | |
Kwakkel et al. | An efficient multiple marker front-capturing method for two-phase flows | |
Chen et al. | Shape optimization of fluid cooling channel based on Darcy reduced-order isogeometric analysis | |
Laroche et al. | Development of an aircraft aero-icing suite using chapel programming language | |
Li et al. | Novel hybrid hard sphere model for direct simulation Monte Carlo computations | |
Tensuda et al. | Multi-dimensional validation of a maximum-entropy-based interpolative moment closure |
Legal Events
Date | Code | Title | Description |
---|---|---|---|
121 | Ep: the epo has been informed by wipo that ep was designated in this application |
Ref document number: 15908601 Country of ref document: EP Kind code of ref document: A1 |
|
NENP | Non-entry into the national phase |
Ref country code: DE |
|
122 | Ep: pct application non-entry in european phase |
Ref document number: 15908601 Country of ref document: EP Kind code of ref document: A1 |