Looking for breakthrough ideas for innovation challenges? Try Patsnap Eureka!

Efficient decomposition parallel method for time domain finite element regions

A technology of domain decomposition and finite element, applied in the direction of instrumentation, design optimization/simulation, calculation, etc., can solve problems such as the limitation of solution matrix scale, loss of high parallelism, and rapid drop in efficiency, so as to save simulation calculation time and improve solution Effect of problem size, increased capacity

Inactive Publication Date: 2018-02-13
NANJING UNIV OF SCI & TECH
View PDF2 Cites 13 Cited by
  • Summary
  • Abstract
  • Description
  • Claims
  • Application Information

AI Technical Summary

Problems solved by technology

The traditional UMFPACK solver cannot solve sparse matrices in parallel, and there are certain restrictions on the size of the solution matrix. Although the parallel MUMPS solver can solve in parallel, as the number of processes increases, the efficiency drops rapidly and loses high parallelism. characteristics

Method used

the structure of the environmentally friendly knitted fabric provided by the present invention; figure 2 Flow chart of the yarn wrapping machine for environmentally friendly knitted fabrics and storage devices; image 3 Is the parameter map of the yarn covering machine
View more

Image

Smart Image Click on the blue labels to locate them in the text.
Viewing Examples
Smart Image
  • Efficient decomposition parallel method for time domain finite element regions
  • Efficient decomposition parallel method for time domain finite element regions
  • Efficient decomposition parallel method for time domain finite element regions

Examples

Experimental program
Comparison scheme
Effect test

Embodiment 1

[0067] In order to verify the correctness and effectiveness of the present invention, a metal resonant cavity is analyzed below.

[0068] one such as figure 2 The metal resonant cavity shown has a size of 1m*1m*1m, the source is added at 0.2m in the z direction, the electric field direction of the source is in the Y direction, the observation frequency is 0-0.4GHz, and the observation point is set at 0.8m in the z direction. Using a 0.05m tetrahedron to dissect, a total of about 30,000 individuals are dissected, and the unknown quantity is about 180,000. The time step is 1000 steps, the time step length is 8.33ps, and each process in the buffer zone expands outward by a distance of 0.075m. image 3 The time-domain waveforms calculated by using 16 processes for the time-domain finite element domain decomposition parallel method are compared with the time-domain waveforms calculated by the traditional time-domain finite element using the parallel MUMPS solver, which verifies t...

the structure of the environmentally friendly knitted fabric provided by the present invention; figure 2 Flow chart of the yarn wrapping machine for environmentally friendly knitted fabrics and storage devices; image 3 Is the parameter map of the yarn covering machine
Login to View More

PUM

No PUM Login to View More

Abstract

The invention discloses an efficient decomposition parallel method for time domain finite element regions. The method comprises the following steps of establishing a solution model, conducting discretization, and obtaining tetrahedral node information and unit information of the model; determining the total number of processes to make sub-regions correspond to process numbers in a one-to-one mode,distributing units into all the sub-regions, determining the thickness of a buffering area according to the time step length, and establishing numbering indexes of each sub-region and a global calculation region and communication indexes between every adjacent sub-regions; from a first-order Maxwell's curl equation with electric field intensity and magnetic flux as unknown quantities, adopting aGalerkin method to test the two sides of the equation, using a primary function for expansion, and obtaining a final iterative formula to fill a matrix of each solution sub-region; adopting an iterative formula of time domain finite elements in CN difference schemes to conduct time iteration, and obtaining the electric field value and the magnetic flux in the space. By means of the method, the calculation time is effectively saved, the capability of solving large-scale electromagnetic problems is improved, and the method has great significance in practical application.

Description

technical field [0001] The invention belongs to the technical field of large-scale parallel computing, in particular to an efficient time-domain finite element domain decomposition parallel method. Background technique [0002] With the rapid development of computer hardware technology and the increase of computing scale, single-core computing can no longer meet the current needs, and multi-core large-scale parallel computing has become the current mainstream direction. The traditional UMFPACK solver cannot solve sparse matrices in parallel, and there are certain restrictions on the size of the solution matrix. Although the parallel MUMPS solver can solve in parallel, as the number of processes increases, the efficiency drops rapidly and loses high parallelism. characteristics. Contents of the invention [0003] The purpose of the present invention is to provide an efficient time-domain finite element domain decomposition parallel method to improve the ability to solve la...

Claims

the structure of the environmentally friendly knitted fabric provided by the present invention; figure 2 Flow chart of the yarn wrapping machine for environmentally friendly knitted fabrics and storage devices; image 3 Is the parameter map of the yarn covering machine
Login to View More

Application Information

Patent Timeline
no application Login to View More
IPC IPC(8): G06F17/50
CPCG06F30/23
Inventor 丁大志汪晨昊陈如山樊振宏
Owner NANJING UNIV OF SCI & TECH
Who we serve
  • R&D Engineer
  • R&D Manager
  • IP Professional
Why Patsnap Eureka
  • Industry Leading Data Capabilities
  • Powerful AI technology
  • Patent DNA Extraction
Social media
Patsnap Eureka Blog
Learn More
PatSnap group products