Font Size: a A A

Research On Parallel Algorithm For Granule Hydrodynamics Method Based On OpenMP

Posted on:2017-05-21Degree:MasterType:Thesis
Country:ChinaCandidate:C L WeiFull Text:PDF
GTID:2308330485468750Subject:Software engineering
Abstract/Summary:PDF Full Text Request
Granule hydrodynamics method is a calculation method for solving the motion process of solid and fluid coupling system. In the study of fluid dynamics for the fluid-solid coupling system, GHM can establish an analysis model effectively for the coupling system with solid granules and fluid granules, through which the instantaneous flow field distribution can be obtained conveniently and it is easily to calculate the interaction force between fluid and solid every moment of the motion process. Compared with the traditional method of fluid dynamics, GHM is a universal and convenient method under certain conditions, and it possesses unique superiority in solving problems of a solid-fluid coupling of a multiple moving solid system. However, with the increasing number of fluid particles and the computation scale increased, leading to calculating speed of serial program based on the GHM becomes slower, and the calculation efficiency is reduced.In order to improve the computational efficiency of the GHM model, this paper proposes a parallel computation scheme based on OpenMP on the multicores computers. Firstly, this paper analyzes the primary computational modules and extracts the computing modules which can be paralleled of the GHM model. Then, the OpenMP compiler directive and runtime library functions are studied, and using OpenMP to realize the parallel computing of the solving process for fluid motion equations in the GHM. Experimental results in Windows platform with 4 cores CPU show that, parallel program’s acceleration ratio reaches 2.5, which proves the OpenMP technology can improve the computational efficiency of the GHM significantly, and it has advantages of easy programming and good portability as well.
Keywords/Search Tags:OpenMP, Granule hydrodynamics method, Fluid hydrodynamics, Parallel computing, Multicore processor
PDF Full Text Request
Related items