Geodesic
Parallelization algorithm for numerical solution of the two-dimensional stationary Navier–Stokes equations with
Matematičeskoe modelirovanie, Tome 17 (2005) no. 11, pp. 63-71 
A. B. Gorshkov. Parallelization algorithm for numerical solution of the two-dimensional stationary Navier–Stokes equations with. Matematičeskoe modelirovanie, Tome 17 (2005) no. 11, pp. 63-71. http://geodesic.mathdoc.fr/item/MM_2005_17_11_a4/
@article{MM_2005_17_11_a4,
     author = {A. B. Gorshkov},
     title = {Parallelization algorithm for numerical solution of the two-dimensional stationary {Navier{\textendash}Stokes} equations with},
     journal = {Matemati\v{c}eskoe modelirovanie},
     pages = {63--71},
     year = {2005},
     volume = {17},
     number = {11},
     language = {ru},
     url = {http://geodesic.mathdoc.fr/item/MM_2005_17_11_a4/}
}
TY  - JOUR
AU  - A. B. Gorshkov
TI  - Parallelization algorithm for numerical solution of the two-dimensional stationary Navier–Stokes equations with
JO  - Matematičeskoe modelirovanie
PY  - 2005
SP  - 63
EP  - 71
VL  - 17
IS  - 11
UR  - http://geodesic.mathdoc.fr/item/MM_2005_17_11_a4/
LA  - ru
ID  - MM_2005_17_11_a4
ER  - 
%0 Journal Article
%A A. B. Gorshkov
%T Parallelization algorithm for numerical solution of the two-dimensional stationary Navier–Stokes equations with
%J Matematičeskoe modelirovanie
%D 2005
%P 63-71
%V 17
%N 11
%U http://geodesic.mathdoc.fr/item/MM_2005_17_11_a4/
%G ru
%F MM_2005_17_11_a4

Voir la notice de l'article provenant de la source Math-Net.Ru

For the case of perfect gas flow the parallelization is carried out of a computer code, developed for solution of the two-dimensional stationary Navier–Stokes equations with use of an implicit iterative scheme. The parallelization algorithm is based on the decomposition of a computation region into several parts according to the number of processors retaining the implicit type of the difference scheme in each subregion. As an example a supersonic flow past a sphere with account of its near wake is considered at $M_{\infty}=5$, $\operatorname{Re}_{0D}=10$ (laminar flow without recirculation region). A good scalability is obtained for the number of processors $N\le15$.

[1] Yoon, Seokkwan, Jameson, Anthonyn, “An LU-SSOR scheme for the Euler and Navier–Stokes equations”, AIAA 25rd Aerospace Sciences Meeting (Jan. 12–15, 1987), Reno, 1987, AIAA Paper 87-16806, 12 p | MR

[2] Gorshkov A. B., Lunev V. V., “Raschet turbulentnogo donnogo teploobmena za osesimmetrichnymi telami”, Kosmonavtika i raketostroenie, 1997, no. 10, 119–126 | MR

[3] Gorshkov A. B., “Raschet laminarnogo donnogo teploobmena za telami v vide tonkikh konusov”, Kosmonavtika i raketostroenie, 1997, no. 11, 13–20 | MR

[4] Gorshkov A. B., “Teploobmen pri sverkhzvukovom obtekanii sfery i tsilindra pri malykh chislakh Reinoldsa”, Izv. RAN MZhG, 2001, no. 1, 156–164 | MR | Zbl

[5] Vlasov V. I., Gorshkov A. B., “Sravnenie rezultatov raschetov giperzvukovogo obtekaniya zatuplennykh tel s letnym eksperimentom OREX”, Izv. RAN MZhG, 2001, no. 5, 160–168 | Zbl

[6] Pulliam T. H., Steger J. L., “Recent improvements in efficiency, accuracy and convergence for implicit approximate factorization algorithms”, AIAA 23rd Aerospace Sciences Meeting, Reno, 1985, AIAA Paper 85-0360, 37 p | MR | Zbl

[7] Gropp W., Lusk E., User's Guide for mpich, a Portable Implementation of MPI Version 1.2.1, Technical Report ANL/MCS-TM-ANL-96/6, Argonne National Laboratory, University of Chicago, 1996