This paper summarizes our progress in the application and improvement of a high order discontinuous Galerkin (DG) method for scale resolving fluid dynamics simulations towards robust and flexible industrial applications. We report the results obtained on the Cray XC40 Hazel Hen cluster at HLRS and show code performance. We present three application cases and developments: An implicit time integration scheme for split-form DG schemes allows us to solve stiff problems with increased efficiency, which will open up new classes of problems for simulations with FLEXI. We follow this by discussing a Large Eddy Simulation (LES) of a compressible turbulent boundary layer and provide comparison to DNS data. Lastly, we demonstrate how to extend the high order scheme with a consistent and conservative sliding mesh interface, and present results of a 1.5 stage turbine simulation with wall-resolved LES.
%0 Book Section
%1 beck2021increasing
%A Beck, Andrea
%A Gao, Min
%A Kempf, Daniel
%A Kopper, Patrick
%A Krais, Nico
%A Kurz, Marius
%A Zeifang, Jonas
%A Munz, Claus-Dieter
%B High Performance Computing in Science and Engineering'20
%D 2021
%I Springer
%K EXC2075 fr:simtech pn1
%P 343--358
%R 10.1007/978-3-030-80602-6_22
%T Increasing the Flexibility of the High Order Discontinuous Galerkin Framework FLEXI Towards Large Scale Industrial Applications
%U http://dx.doi.org/10.1007/978-3-030-80602-6_22
%X This paper summarizes our progress in the application and improvement of a high order discontinuous Galerkin (DG) method for scale resolving fluid dynamics simulations towards robust and flexible industrial applications. We report the results obtained on the Cray XC40 Hazel Hen cluster at HLRS and show code performance. We present three application cases and developments: An implicit time integration scheme for split-form DG schemes allows us to solve stiff problems with increased efficiency, which will open up new classes of problems for simulations with FLEXI. We follow this by discussing a Large Eddy Simulation (LES) of a compressible turbulent boundary layer and provide comparison to DNS data. Lastly, we demonstrate how to extend the high order scheme with a consistent and conservative sliding mesh interface, and present results of a 1.5 stage turbine simulation with wall-resolved LES.
%@ print 978-3-030-80601-9, online 978-3-030-80602-6
@incollection{beck2021increasing,
abstract = {This paper summarizes our progress in the application and improvement of a high order discontinuous Galerkin (DG) method for scale resolving fluid dynamics simulations towards robust and flexible industrial applications. We report the results obtained on the Cray XC40 Hazel Hen cluster at HLRS and show code performance. We present three application cases and developments: An implicit time integration scheme for split-form DG schemes allows us to solve stiff problems with increased efficiency, which will open up new classes of problems for simulations with FLEXI. We follow this by discussing a Large Eddy Simulation (LES) of a compressible turbulent boundary layer and provide comparison to DNS data. Lastly, we demonstrate how to extend the high order scheme with a consistent and conservative sliding mesh interface, and present results of a 1.5 stage turbine simulation with wall-resolved LES.},
added-at = {2022-01-24T12:28:31.000+0100},
author = {Beck, Andrea and Gao, Min and Kempf, Daniel and Kopper, Patrick and Krais, Nico and Kurz, Marius and Zeifang, Jonas and Munz, Claus-Dieter},
biburl = {https://puma.ub.uni-stuttgart.de/bibtex/2cb61853b47ec5acaa03f90775c79f7f5/simtechpuma},
booktitle = {High Performance Computing in Science and Engineering'20},
doi = {10.1007/978-3-030-80602-6_22},
interhash = {fed0afed5924813c29a05f2427fdb154},
intrahash = {cb61853b47ec5acaa03f90775c79f7f5},
isbn = {print 978-3-030-80601-9, online 978-3-030-80602-6},
keywords = {EXC2075 fr:simtech pn1},
pages = {343--358},
publisher = {Springer},
timestamp = {2022-01-27T10:38:50.000+0100},
title = {Increasing the Flexibility of the High Order Discontinuous Galerkin Framework FLEXI Towards Large Scale Industrial Applications},
url = {http://dx.doi.org/10.1007/978-3-030-80602-6_22},
year = 2021
}