Full-Text Search:
Home|Journal Papers|About CNKI|User Service|FAQ|Contact Us|中文
《Acta Armamentarii》 2018-05
Add to Favorite Get Latest Update

Application of MPI-based ISAT Accelerated Algorithm in Gaseous Detonation Numerical Simulation

WU Jin-tao;DONG Gang;LI Bao-ming;National Key Laboratory of Transient Physics,Nanjing University of Science and Technology;  
In the numerical simulation of gaseous detonation process with the detailed chemical reaction mechanism,the computation of chemical source term takes a large amount of CPU times due to the stiffness and nonlinearity of the term. To improve the computational efficiency without the loss of the computational accuracy for chemistry computation,two algorithms of in situ adaptive tabulation( ISAT) based on multi-threaded parallel integration( MPI) are proposed to replace the original direct integration( DI)method,and then applied in the computation of two-dimensional gaseous detonation process for H2/O2 medium with detailed chemical mechanism. In the parallel algorithms,two types of tabulation,namely the local tabulation for each processor and the global tabulation for all processors,are proposed,and theeffect of the tabulation mode on the computational efficiency is investigated. Further,the influences of error criterions and numerical schemes on the computational efficiency of the algorithms are also examined.The results show that both the proposed algorithms can provide satisfactory computational accuracy compared with the results of DI,and the computational efficiency of the local tabulation for each processor with the chemical speedup ratio of 2. 17 ~ 2. 43 is higher than that of the global tabulation for all processors. The proposed parallel accelerated algorithms can not only accurately describe the propagation process of two-dimensional gaseous detonation wave,but also improve the parallel computational efficiency of of chemically reactive flows.
【Fund】: 国家自然科学基金项目(11372140)
【CateGory Index】: TJ301
Download(CAJ format) Download(PDF format)
CAJViewer7.0 supports all the CNKI file formats; AdobeReader only supports the PDF format.
©2006 Tsinghua Tongfang Knowledge Network Technology Co., Ltd.(Beijing)(TTKN) All rights reserved