Scalability of an Eulerian-Lagrangian large-eddy simulation solver with hybrid MPI/OpenMP parallelisation

Eulerian-Lagrangian approaches capable of accurately reproducing complex fluid flows are becoming more and more popular due to the increasing availability and capacity of High Performance Computing facilities. However, the parallelisation of the Lagrangian part of such methods is challenging when a...

Descripción completa

Detalles Bibliográficos
Autores: Ouro, Pablo, Fraga, Bruño, López Novoa, Unai, Stoesser, Thorsten
Tipo de recurso: artículo
Fecha de publicación:2019
País:España
Institución:Universidad del País Vasco
Repositorio:Addi. Archivo Digital para la Docencia y la Investigación
OAI Identifier:oai:addi.ehu.eus:10810/69493
Acceso en línea:http://hdl.handle.net/10810/69493
Access Level:acceso abierto
Palabra clave:hybrid MPI/openMP
Eulerian-Lagrangian
large-eddy simulation
immersed boundary method
high performance computing
id ES_7e198ffc97fc4231949965332cb8a2e7
oai_identifier_str oai:addi.ehu.eus:10810/69493
network_acronym_str ES
network_name_str España
repository_id_str
spelling Scalability of an Eulerian-Lagrangian large-eddy simulation solver with hybrid MPI/OpenMP parallelisationOuro, PabloFraga, BruñoLópez Novoa, UnaiStoesser, Thorstenhybrid MPI/openMPEulerian-Lagrangianlarge-eddy simulationimmersed boundary methodhigh performance computingEulerian-Lagrangian approaches capable of accurately reproducing complex fluid flows are becoming more and more popular due to the increasing availability and capacity of High Performance Computing facilities. However, the parallelisation of the Lagrangian part of such methods is challenging when a large number of Lagrangian markers are employed. In this study, a hybrid MPI/OpenMP parallelisation strategy is presented and implemented in a finite difference based large-eddy simulation code featuring the immersed boundary method which generally employs a large number of Lagrangian markers. A master-scattering-gathering strategy is used to deal with the handling of the Lagrangian markers and OpenMP is employed to distribute their computational load across several CPU threads. A classical domain-decomposition-based MPI approach is used to carry out the Eulerian, fixed-mesh fluid calculations. The results demonstrate that by using an effective combination of MPI and OpenMP the code can outperform a pure MPI parallelisation approach by up to 20%. Outcomes from this paper are of interest to various Eulerian-Lagrangian applications including the immersed boundary method, discrete element method or Lagrangian particle tracking.This research was partially funded by EPSRC under the grant EP/K502819/1. The simulations were carried out in the facilities of the Supercomputing Wales, a project partly funded by the European Regional Development Fund (ERDF) via Welsh Government.Elsevier202420242019info:eu-repo/semantics/articleapplication/pdfhttp://hdl.handle.net/10810/69493reponame:Addi. Archivo Digital para la Docencia y la Investigacióninstname:Universidad del País VascoIngléshttps://www.sciencedirect.com/science/article/pii/S0045793018307424info:eu-repo/semantics/openAccesshttp://creativecommons.org/licenses/by/4.0/© 2018 The Authors. Published by Elsevier Ltd. This is an open access article under the CC BY license. (http://creativecommons.org/licenses/by/4.0/)oai:addi.ehu.eus:10810/694932026-06-18T09:23:17Z
dc.title.none.fl_str_mv Scalability of an Eulerian-Lagrangian large-eddy simulation solver with hybrid MPI/OpenMP parallelisation
title Scalability of an Eulerian-Lagrangian large-eddy simulation solver with hybrid MPI/OpenMP parallelisation
spellingShingle Scalability of an Eulerian-Lagrangian large-eddy simulation solver with hybrid MPI/OpenMP parallelisation
Ouro, Pablo
hybrid MPI/openMP
Eulerian-Lagrangian
large-eddy simulation
immersed boundary method
high performance computing
title_short Scalability of an Eulerian-Lagrangian large-eddy simulation solver with hybrid MPI/OpenMP parallelisation
title_full Scalability of an Eulerian-Lagrangian large-eddy simulation solver with hybrid MPI/OpenMP parallelisation
title_fullStr Scalability of an Eulerian-Lagrangian large-eddy simulation solver with hybrid MPI/OpenMP parallelisation
title_full_unstemmed Scalability of an Eulerian-Lagrangian large-eddy simulation solver with hybrid MPI/OpenMP parallelisation
title_sort Scalability of an Eulerian-Lagrangian large-eddy simulation solver with hybrid MPI/OpenMP parallelisation
dc.creator.none.fl_str_mv Ouro, Pablo
Fraga, Bruño
López Novoa, Unai
Stoesser, Thorsten
author Ouro, Pablo
author_facet Ouro, Pablo
Fraga, Bruño
López Novoa, Unai
Stoesser, Thorsten
author_role author
author2 Fraga, Bruño
López Novoa, Unai
Stoesser, Thorsten
author2_role author
author
author
dc.subject.none.fl_str_mv hybrid MPI/openMP
Eulerian-Lagrangian
large-eddy simulation
immersed boundary method
high performance computing
topic hybrid MPI/openMP
Eulerian-Lagrangian
large-eddy simulation
immersed boundary method
high performance computing
description Eulerian-Lagrangian approaches capable of accurately reproducing complex fluid flows are becoming more and more popular due to the increasing availability and capacity of High Performance Computing facilities. However, the parallelisation of the Lagrangian part of such methods is challenging when a large number of Lagrangian markers are employed. In this study, a hybrid MPI/OpenMP parallelisation strategy is presented and implemented in a finite difference based large-eddy simulation code featuring the immersed boundary method which generally employs a large number of Lagrangian markers. A master-scattering-gathering strategy is used to deal with the handling of the Lagrangian markers and OpenMP is employed to distribute their computational load across several CPU threads. A classical domain-decomposition-based MPI approach is used to carry out the Eulerian, fixed-mesh fluid calculations. The results demonstrate that by using an effective combination of MPI and OpenMP the code can outperform a pure MPI parallelisation approach by up to 20%. Outcomes from this paper are of interest to various Eulerian-Lagrangian applications including the immersed boundary method, discrete element method or Lagrangian particle tracking.
publishDate 2019
dc.date.none.fl_str_mv 2019
2024
2024
dc.type.none.fl_str_mv info:eu-repo/semantics/article
format article
dc.identifier.none.fl_str_mv http://hdl.handle.net/10810/69493
url http://hdl.handle.net/10810/69493
dc.language.none.fl_str_mv Inglés
language_invalid_str_mv Inglés
dc.relation.none.fl_str_mv https://www.sciencedirect.com/science/article/pii/S0045793018307424
dc.rights.none.fl_str_mv info:eu-repo/semantics/openAccess
http://creativecommons.org/licenses/by/4.0/
eu_rights_str_mv openAccess
rights_invalid_str_mv http://creativecommons.org/licenses/by/4.0/
dc.format.none.fl_str_mv application/pdf
dc.publisher.none.fl_str_mv Elsevier
publisher.none.fl_str_mv Elsevier
dc.source.none.fl_str_mv reponame:Addi. Archivo Digital para la Docencia y la Investigación
instname:Universidad del País Vasco
instname_str Universidad del País Vasco
reponame_str Addi. Archivo Digital para la Docencia y la Investigación
collection Addi. Archivo Digital para la Docencia y la Investigación
repository.name.fl_str_mv
repository.mail.fl_str_mv
_version_ 1869411712190382080
score 15,811543

Similares en LA Referencia