Finite element program parallelization method based on RAS peer-to-peer interaction

Abstract

The invention discloses a finite element program parallelization method based on RAS peer-to-peer interaction, and the method comprises the following steps: 1, initializing an MPI parallel environment, and obtaining a process number and a total process number; 2, reading a data file; 3, establishing a local-to-overall mapping relation; 4, determining a start line IStart and a stop line Iend of a storage stiffness matrix of the process, and pre-distributing a distributed stiffness matrix; 5, assembling a distributed overall stiffness matrix k according to an unknown quantity eqmap mapping relation; 6, solving the finite element equation set by using a parallel solver; 7, obtaining a process result according to unknown eqmap mapping, and emitting solutions required by other processes; 8, post-processing, and terminating the parallel environment. According to the method, on the basis of realizing parallelization by utilizing the MPI function library, the finite element parallel computing efficiency is remarkably improved under the condition that an existing finite element program is changed to be minimum.

Description

technical field [0001] The invention belongs to the field of mechanical engineering design and calculation, and in particular relates to a finite element program parallelization method based on RAS peer-to-peer interaction. Background technique [0002] As a numerical method applied to solve engineering and mathematical physics, the finite element method is widely used in typical problems in engineering fields such as structural analysis. By discretizing the specific research area of ​​the structure into several units of specific shapes, the partial differential equation that originally described the physical phenomenon can be transformed into a certain form of matrix equation, and the calculation results of specific discrete points can be used to represent the calculation of physical quantities at any point in the structure As a result, the problem is simplified. Using the finite element method to analyze engineering problems can provide guidance for the research of proble...

AI Technical Summary

Problems solved by technology

At present, there are few ways to realize the finite element parallel algorithm, and only some commercial software realize this function, and there are many limitations: 1) Commercial software does not open its source code, which increases the limitation of solving the problem, and the setting of its parameters Need to keep trying, and the convergence is not easy to control; 2) Do not use users to independently develop efficient constitutive models and algorithms, and cannot update in time for emerging engineering problems; 3) Commercial software is generally expensive, and each Upgrades require re-authorization

Images