CN110427711A - MD the and KMC coupling process of reactor critical material Parallel Simulation - Google Patents
MD the and KMC coupling process of reactor critical material Parallel Simulation Download PDFInfo
- Publication number
- CN110427711A CN110427711A CN201910724903.7A CN201910724903A CN110427711A CN 110427711 A CN110427711 A CN 110427711A CN 201910724903 A CN201910724903 A CN 201910724903A CN 110427711 A CN110427711 A CN 110427711A
- Authority
- CN
- China
- Prior art keywords
- simulation
- module
- kmc
- dynamics
- defect
- Prior art date
- Legal status (The legal status is an assumption and is not a legal conclusion. Google has not performed a legal analysis and makes no representation as to the accuracy of the status listed.)
- Pending
Links
Abstract
The present invention relates to a kind of MD and KMC coupling process of reactor critical material Parallel Simulation, this method carries out the simulation of molecular dynamics cascade collision by MD module;After MD module complete call, defect statistics and analysis are carried out for the state of atom after the completion of MD module simulation;It sends the initial imperfection information of cascade collision to KMC module and carries out KMC simulation;Finally KMC module simulation is waited to terminate, export the defect evolution structure of long period or defect information is passed into Rate Theory (RT) module, carries out the defect evolution analysis of longer time.The present invention improves the efficiency of coupled simulation by close coupled system, by the design of calling interface and unified atom and lattice-site representation method and task division mode, facilitates program interaction.
Description
Technical field
The present invention relates to nuclear material irradiation damage computer modeling technique, and in particular to a kind of reactor critical material is parallel
The molecular dynamics (MD) of simulation and the coupling process of dynamics Monte Carlo (KMC).
Background technique
The dynamic behavior evolution of material is Multiscal process, the materials such as pressure vessel, involucrum especially in nuclear reactor
Material.Under the high temperature of reactor, high pressure, high radiation environment, due to the effect by high-energy neutron, generated by cascade collision initial
Defect;Defect nucleation growth process and subsequent fault in material long-time evolutionary process are arrived again, and material property to the end loses
Effect.Whole process is related to different time scales and space scale.In multi-dimension analogy method, usually using molecular dynamics side
Defect under method (Molecular dynamics, MD) Lai Jinhang atomic scale generates the simulation of process;Since MD method cannot
Carry out the limitation of long-time Evolution Simulation, it is also necessary to use dynamics Monte Carlo (Kinetic Monte Carlo, KMC)
Method carries out the simulation of long-time defect evolution and eliminates the space correlation row of defect.KMC method is by focus from the " former of MD
Son " has been promoted " system ", " the atomic motion track " of MD is roughened as " transition of system configuration ", thus by the time of simulation
Span is increased to the scale of configuration transition from the scale of atomic vibration.MD method can simulate the cascade collision mistake in the short period
Journey counts initial imperfection, as the input of KMC method, with the simulation for supporting the long-time of KMC method to develop.
On the other hand, since the limitation of computing capability and storage capacity, single machine or small-scale cluster are unable to complete expected rule
The simulation calculus of mould, the introducing of High Performance Computing allow to simulate to more on a large scale and for more time.Simultaneously as high
The introducing of performance computing technique needs to consider the problems such as task is divided, communicated, but also parallel MD method and parallel KMC method
Coupled simulation realize it is more complicated.Current existing correlative study mostly uses the loose coupling mode of reading and writing of files, this side
Formula coupled modes, which introduce a large amount of IO, reduces duration performance, along with intermediate individually defect analysis process, so that simulating
Journey is not coherent enough.
Summary of the invention
It is a kind of for parallel MD and parallel KMC method it is an object of the invention in view of the drawbacks of the prior art, provide
Close coupling method is used for reactor critical material Parallel Simulation, so as to convenience, is combined using MD-KMC method to high-performance
Carry out fault in material simulation.
Technical scheme is as follows: a kind of MD and KMC coupling process of reactor critical material Parallel Simulation, packet
It includes:
(1) it determines simulated domain range, and generates the required call parameters of parallel molecular dynamics simulation;
(2) molecular dynamics (MD) module is called to carry out the simulation of molecular dynamics cascade collision;
(3) defect statistical analysis is carried out for the state of atom after the completion of molecular dynamics module simulation;
(4) defect statistical analysis result is sent to dynamics together with call parameters needed for dynamics Monte Carlo simulation
Monte Carlo (KMC) module carries out Parallel Dynamics Monte Carlo simulation;
(5) it after dynamics Monte Carlo simulation, exports the defect evolution structure of long period or passes defect information
Rate Theory (RT) module is passed, the defect evolution analysis of longer time is carried out.
Further, MD the and KMC coupling process of reactor critical material Parallel Simulation as described above, in step (1) point
Call parameters needed for subdynamics simulation include: simulated time step, first lattice atoms PKA to be offed normal by neutron bombardment
Energy size, space size.
Further, MD the and KMC coupling process of reactor critical material Parallel Simulation as described above, step make in (3)
Defect statistical analysis is carried out with Wigner-Seitz primitive unit cell method, the atomic coordinates in analyzing molecules dynamics simulation result, system
Count the coordinate position and quantity of polyatom cluster, vacancy, the input as dynamics Monte Carlo module.
Further, MD the and KMC coupling process of reactor critical material Parallel Simulation as described above moves in step (4)
Call parameters needed for mechanics Monte Carlo simulation include: time span, random number seed, space size.
Further, MD the and KMC coupling process of reactor critical material Parallel Simulation as described above, in step (2)
The simulation of molecular dynamics cascade collision uses unified lattice-site and atomicity with the dynamics Monte Carlo simulation in step (4)
According to structure representation mode, in task division, be all made of the mode of space division, and guarantee to be divided into molecular dynamics module with
Lattice-site id in the module of dynamics Monte Carlo on each parallel computation unit is consistent, to make to be divided into molecular dynamics
Module and the lattice-site being divided into the module of dynamics Monte Carlo can set up one-to-one mapping relations.
Further, MD the and KMC coupling process of reactor critical material Parallel Simulation as described above, wherein described
The concrete form of unified lattice-site and atomic data structure representation mode are as follows: set three-dimensional simulation area size as x*y*z (x, y,
Z is integer, and unit is lattice constant), then three-dimensional array array [z] [y] [2*x] is constructed, to save lattice-site or atom
Information;Wherein, it sets the length of first dimension of array to twice of the first dimension lattice in the space points of simulated domain,
Remaining dimension remains unchanged.
Further, MD the and KMC coupling process of reactor critical material Parallel Simulation as described above, wherein pass through tune
With the design of interface, dynamics to is sent defect statistical analysis result using unified data interaction structure in step (4)
Monte Carlo module.
Beneficial effects of the present invention are as follows: MD the and KMC coupling of reactor critical material Parallel Simulation provided by the present invention
Conjunction method realizes the close coupled system of parallel MD and parallel KMC, by the non-loose coupling mode of file interaction, improves coupling
The quasi- efficiency of molding, facilitates program interaction, breaches 1012The simulation scale of atom, second rank.Pass through unified data interaction
Structure ensure that efficient, the convenient transmitting of data.
Detailed description of the invention
Fig. 1 is the program architecture figure of the MD and KMC coupling process of reactor critical material Parallel Simulation provided by the invention.
Fig. 2 is the atom storage mode and lattice-site storage mode schematic diagram that MD module and KMC module are used uniformly.
Fig. 3 is the flow chart of MD-KMC coupled simulation, is supported including individual MD and independent KMC simulation.
Specific embodiment
The present invention is described in detail below with reference to the accompanying drawings and embodiments.
The molecular dynamics (MD) of reactor critical material Parallel Simulation provided by the present invention and dynamics Monte Carlo
(KMC) coupling process is coupled MD program and KMC program by a coupling procedure, practical when carrying out program compiling
On be that the compiling of three programs is become to a program, centre call each program be substantially function/module call (based on this,
MD program and KMC program before coupling are hereafter referred to MD module, KMC module), specific coupling framework is as shown in Figure 1.Couple journey
Sequence calls MD module to carry out the simulation of cascade collision process, generates initial imperfection, then carries out the statistics and analysis of defect, and will
It counts the initial imperfection completed and KMC module is passed to by unified data-interface, KMC module is called to carry out lacking for long period
Fall into Evolution Simulation.
Coupling procedure by its input parameter produce MD module needed for parameter (such as simulated time step, first hit by neutron
Hit energy size, the space size etc. of the lattice atoms PKA to off normal), call MD module to be simulated;In MD module complete call
Afterwards, coupling procedure carries out defect statistics and analysis for the state of atom (mainly atomic coordinates) after the completion of MD module simulation;
Meanwhile coupling procedure can also generate KMC module needed for parameter (such as time span, random number seed, space size), with lack
Falling into the output of statistical analysis module --- the initial imperfection information of cascade collision is transmitted to KMC module together and carries out KMC simulation;Finally
It waits KMC module simulation to terminate, exports the defect evolution structure of long period or defect information is passed into Rate Theory (RT)
Procedures/modules carry out the defect evolution analysis of longer time.It should be pointed out that the initial imperfection information transmitting of cascade collision
When to KMC module, need to design a kind of unified data interaction structure, to guarantee efficient, the convenient transmitting of data.
Meanwhile the coupled modes between the characteristics of taking into account concurrent program and concurrent program, MD module are used with KMC module
Identical atom and lattice-site representation method and task division mode.It is specific as follows: for the metal material lattice of BCC structure,
Using lattice-site shown in Fig. 2 and atomic data representation mode in MD and KMC;In task division, it is all made of space
The mode of division, and guarantee each parallel computation unit (being herein MPI process) being divided into MD module and KMC module
On lattice-site id it is consistent, that is, an a pair can be set up with the lattice-site that is divided into KMC module by being divided into MD module
The mapping relations answered.For BCC lattice structure, unified lattice-site is specifically expressed as follows with atomic data representation mode:
As shown in Fig. 2, setting three-dimensional simulation area size as x*y*z (x, y, z are integer, and unit is lattice constant), then three-dimensional is constructed
Array array [2*x] [y] [z], to save lattice-site or atom information;Wherein, it sets the length of first dimension of array to
The first dimension lattice in the space points of simulated domain twice (because of BCC lattice dot matrix feature: each atom in 8 angles, in body
One atom of the heart;When choosing an axis work storage projection, the storage of body central atom needs another opening space), remaining dimension is protected
It holds constant.For convenience of displaying, Fig. 2 shows only the storage of the atom under two-dimensional case and lattice-site storage mode.Such as [4] [2]
Indicate that first dimension index of the lattice-site in array is 4,2-d index is 2.The two-dimensional case can regard BCC dot matrix as in x-
The projection of the face y, [0] [0], [0] [1], [2] [0], [2] [1] can regard atom on 8 angles of BCC dot matrix as and project in the face x-y, [1] [0]
It is projected for body central atom in the face x-y.It is projected using x-axis as storage, then needs to open up one 2 times of memory space more, with storage
Body central atom.
Defect statistical analysis module of the invention mainly uses Wigner-Seitz primitive unit cell method, analyzes in MD analog result
Atomic coordinates, statistics polyatom cluster, vacancy coordinate position and quantity, using the input as KMC module.
The specific steps of coupling process of the present invention are described below with reference to Fig. 3:
1) according to input parameter (such as simulated domain size, process number), carry out task divides to obtain the mould of each process
Quasi- regional scope;
2) call parameters (such as PKA energy size, simulation of parallel molecular dynamics (MD) module are generated according to input parameter
Time step, space size etc.);
3) MD module is called, each process generates atom information, wherein each according to oneself space size and process number
The storage mode of atomic data on lattice-site is as shown in Figure 2;
4) setting input parameter, calls MD module, executes formal molecular dynamics cascade collision simulation;
5) molecular dynamics module simulation terminates, and each atomic coordinates information according at the end of uses Wigner-
The statistical analysis of Seitz primitive unit cell method progress defect;
6) defect of statistics is passed into dynamics Monte Carlo (KMC) module by unified data-interface, as KMC
The input of module;
7) call parameters of parallel Monte Carlo module are produced (such as simulation duration, random number seed, sky according to input parameter
Between size etc.);
8) KMC module is called, each process generates lattice-site information according to oneself space size and process number, wherein
The storage mode of each lattice-site is as shown in Figure 2;
9) setting input parameter, calls KMC module, reads the initial imperfection that passes over of MD module, execute it is formal and
Row Monte Carlo simulation;
10) MD-KMC coupled simulation result is exported.
Obviously, various changes and modifications can be made to the invention without departing from essence of the invention by those skilled in the art
Mind and range.If in this way, belonging to the model of the claims in the present invention and its equivalent technology to these modifications and changes of the present invention
Within enclosing, then the present invention is also intended to include these modifications and variations.
Claims (7)
1. a kind of MD and KMC coupling process of reactor critical material Parallel Simulation, comprising:
(1) it determines simulated domain range, and generates the required call parameters of parallel molecular dynamics simulation;
(2) molecular dynamics module is called to carry out the simulation of molecular dynamics cascade collision;
(3) defect statistical analysis is carried out for the state of atom after the completion of molecular dynamics module simulation;
(4) defect statistical analysis result is sent to dynamics together with call parameters needed for dynamics Monte Carlo simulation cover spy
Carlow module carries out Parallel Dynamics Monte Carlo simulation;
(5) it after dynamics Monte Carlo simulation, exports the defect evolution structure of long period or passes to defect information
Rate Theory module carries out the defect evolution analysis of longer time.
2. MD the and KMC coupling process of reactor critical material Parallel Simulation as described in claim 1, it is characterised in that: step
Suddenly call parameters needed for molecular dynamics simulation include: simulated time step, first dot matrix offed normal by neutron bombardment in (1)
The energy size of atom PKA, space size.
3. MD the and KMC coupling process of reactor critical material Parallel Simulation as described in claim 1, it is characterised in that: step
Suddenly defect statistical analysis is carried out using Wigner-Seitz primitive unit cell method in (3), the original in analyzing molecules dynamics simulation result
Subcoordinate counts the coordinate position and quantity of polyatom cluster, vacancy, the input as dynamics Monte Carlo module.
4. MD the and KMC coupling process of reactor critical material Parallel Simulation as described in claim 1, it is characterised in that: step
Suddenly call parameters needed for dynamics Monte Carlo simulation include: time span, random number seed, space size in (4).
5. MD the and KMC coupling process of reactor critical material Parallel Simulation as described in claim 1, it is characterised in that: step
Suddenly the molecular dynamics cascade collision simulation in (2) uses unified lattice with the dynamics Monte Carlo simulation in step (4)
Point and atomic data structure representation mode are all made of the mode of space division in task division, and guarantee to be divided into Molecule Motion
Mechanics module is consistent with the lattice-site id on each parallel computation unit in the module of dynamics Monte Carlo, to make to be divided into
Molecular dynamics module and the lattice-site being divided into the module of dynamics Monte Carlo can set up one-to-one mapping and close
System.
6. MD the and KMC coupling process of reactor critical material Parallel Simulation as claimed in claim 5, it is characterised in that: institute
The concrete form of the unified lattice-site and atomic data structure representation mode stated are as follows: set three-dimensional simulation area size as x*y*z,
X, y, z is integer, and unit is lattice constant, then constructs three-dimensional array array [z] [y] [2*x], to save lattice-site or
Atom information;Wherein, the length of first dimension of array is set to first, the space dimension lattice points of simulated domain
Twice, remaining dimension remains unchanged.
7. MD the and KMC coupling process of reactor critical material Parallel Simulation as described in claim 1, it is characterised in that: logical
The design for crossing calling interface, the unified data interaction structure of use sends defect statistical analysis result to dynamic in step (4)
Mechanics Monte Carlo module.
Priority Applications (1)
Application Number | Priority Date | Filing Date | Title |
---|---|---|---|
CN201910724903.7A CN110427711A (en) | 2019-08-07 | 2019-08-07 | MD the and KMC coupling process of reactor critical material Parallel Simulation |
Applications Claiming Priority (1)
Application Number | Priority Date | Filing Date | Title |
---|---|---|---|
CN201910724903.7A CN110427711A (en) | 2019-08-07 | 2019-08-07 | MD the and KMC coupling process of reactor critical material Parallel Simulation |
Publications (1)
Publication Number | Publication Date |
---|---|
CN110427711A true CN110427711A (en) | 2019-11-08 |
Family
ID=68413052
Family Applications (1)
Application Number | Title | Priority Date | Filing Date |
---|---|---|---|
CN201910724903.7A Pending CN110427711A (en) | 2019-08-07 | 2019-08-07 | MD the and KMC coupling process of reactor critical material Parallel Simulation |
Country Status (1)
Country | Link |
---|---|
CN (1) | CN110427711A (en) |
Cited By (3)
Publication number | Priority date | Publication date | Assignee | Title |
---|---|---|---|---|
CN111695243A (en) * | 2020-05-20 | 2020-09-22 | 北京科技大学 | Communication method for synchronous parallel spatial resolution random cluster dynamics |
CN112885414A (en) * | 2021-03-31 | 2021-06-01 | 中国核动力研究设计院 | Irradiation damage simulation system and method based on rate theory and cluster dynamics |
CN116167272A (en) * | 2023-02-21 | 2023-05-26 | 北京科技大学 | Atomic scale MD-KMC parallel simulation unified modeling method and system |
Citations (5)
Publication number | Priority date | Publication date | Assignee | Title |
---|---|---|---|---|
US20060069515A1 (en) * | 2004-09-30 | 2006-03-30 | International Business Machines Corporation | Method and system for protein folding trajectory analysis using patterned clusters |
CN104048997A (en) * | 2014-05-27 | 2014-09-17 | 天津大学 | Oil-water phase content measurement method based on multivariate phase space complex network and verification method thereof |
CN104268120A (en) * | 2014-09-26 | 2015-01-07 | 安徽大学 | Monte carlo simulation parallel computing method for kinetic process of polymerization reaction |
WO2015021540A1 (en) * | 2013-08-15 | 2015-02-19 | Zymeworks Inc. | Systems and methods for in silico evaluation of polymers |
CN106407005A (en) * | 2016-08-31 | 2017-02-15 | 中国科学院计算机网络信息中心 | Concurrent process merging method and system based on multi-scale coupling |
-
2019
- 2019-08-07 CN CN201910724903.7A patent/CN110427711A/en active Pending
Patent Citations (5)
Publication number | Priority date | Publication date | Assignee | Title |
---|---|---|---|---|
US20060069515A1 (en) * | 2004-09-30 | 2006-03-30 | International Business Machines Corporation | Method and system for protein folding trajectory analysis using patterned clusters |
WO2015021540A1 (en) * | 2013-08-15 | 2015-02-19 | Zymeworks Inc. | Systems and methods for in silico evaluation of polymers |
CN104048997A (en) * | 2014-05-27 | 2014-09-17 | 天津大学 | Oil-water phase content measurement method based on multivariate phase space complex network and verification method thereof |
CN104268120A (en) * | 2014-09-26 | 2015-01-07 | 安徽大学 | Monte carlo simulation parallel computing method for kinetic process of polymerization reaction |
CN106407005A (en) * | 2016-08-31 | 2017-02-15 | 中国科学院计算机网络信息中心 | Concurrent process merging method and system based on multi-scale coupling |
Non-Patent Citations (2)
Title |
---|
GABRIEL KABBE 等: "A Coupled Molecular Dynamics/Kinetic Monte Carlo Approach for Protonation Dynamics in Extended Systems", 《JOURNAL OF CHEMICAL THEORY AND COMPUTATION》 * |
李建江 等: "MD与KMC的耦合模拟研究与实现", 《电子科技大学学报》 * |
Cited By (5)
Publication number | Priority date | Publication date | Assignee | Title |
---|---|---|---|---|
CN111695243A (en) * | 2020-05-20 | 2020-09-22 | 北京科技大学 | Communication method for synchronous parallel spatial resolution random cluster dynamics |
CN111695243B (en) * | 2020-05-20 | 2023-05-12 | 北京科技大学 | Communication method for synchronous parallel space resolution random cluster dynamics |
CN112885414A (en) * | 2021-03-31 | 2021-06-01 | 中国核动力研究设计院 | Irradiation damage simulation system and method based on rate theory and cluster dynamics |
CN116167272A (en) * | 2023-02-21 | 2023-05-26 | 北京科技大学 | Atomic scale MD-KMC parallel simulation unified modeling method and system |
CN116167272B (en) * | 2023-02-21 | 2024-01-26 | 北京科技大学 | Atomic scale MD-KMC parallel simulation unified modeling method and system |
Similar Documents
Publication | Publication Date | Title |
---|---|---|
CN110427711A (en) | MD the and KMC coupling process of reactor critical material Parallel Simulation | |
Brown et al. | Monte Carlo methods for radiation transport analysis on vector computers | |
Mallinson et al. | Cloverleaf: Preparing hydrodynamics codes for exascale | |
CN100576221C (en) | A kind of fault injection system of verifying anti-single particle effect capability and method thereof | |
JP6605610B2 (en) | Semiconductor device | |
CN108427861A (en) | A method of material periodicities polycrystalline structure is built based on mpt kits | |
TW200424941A (en) | Memory rewind and reconstruction for hardware emulator | |
Dayan et al. | EagleTree: Exploring the design space of SSD-based algorithms | |
Yu et al. | GPGPU-MiniBench: accelerating GPGPU micro-architecture simulation | |
Brown | Accelerating advection for atmospheric modelling on Xilinx and Intel FPGAs | |
Abreu et al. | PIC codes in new processors: A full relativistic PIC code in CUDA-enabled hardware with direct visualization | |
Moor et al. | libFAUDES—An open source C++ library for discrete event systems | |
CN110032489A (en) | The processor device of performance information is collected by the recurrence based on command set | |
Almadi et al. | New methods and tools for design of tests memory | |
Vlassopoulos et al. | An FPGA design for the stochastic Greenberg-Hastings cellular automata | |
Ho et al. | Multi-agent simulation on multiple GPUs | |
JPH07140289A (en) | Neutron transport simulator | |
Todorov et al. | Automated construction of a cycle-approximate transaction level model of a memory controller | |
Levesque | Application development for titan-a multi-petaflop hybrid-multicore MPP system | |
Lee et al. | Toward Heterogeneous Virtual Platforms For Early SW Development | |
Saboori et al. | Hybrid prototyping of multicore embedded systems | |
Conti et al. | Performance evaluation of digital pixel readout chip architecture operating at very high rate through a reusable UVM simulation framework | |
Navaridas et al. | Realistic evaluation of interconnection networks using synthetic traffic | |
US20230074456A1 (en) | Trace buffer data management | |
Sirakoulis et al. | Cellular Automata and Power Consumption. |
Legal Events
Date | Code | Title | Description |
---|---|---|---|
PB01 | Publication | ||
PB01 | Publication | ||
SE01 | Entry into force of request for substantive examination | ||
SE01 | Entry into force of request for substantive examination |