000873125 001__ 873125
000873125 005__ 20210130004404.0
000873125 0247_ $$2arXiv$$aarXiv:1911.11380
000873125 0247_ $$2Handle$$a2128/24059
000873125 0247_ $$2altmetric$$aaltmetric:71224905
000873125 037__ $$aFZJ-2020-00570
000873125 1001_ $$0P:(DE-HGF)0$$aBode, Mathis$$b0$$eCorresponding author
000873125 245__ $$aUsing Physics-Informed Super-Resolution Generative Adversarial Networks for Subgrid Modeling in Turbulent Reactive Flows
000873125 260__ $$c2019
000873125 3367_ $$0PUB:(DE-HGF)25$$2PUB:(DE-HGF)$$aPreprint$$bpreprint$$mpreprint$$s1580209300_29129
000873125 3367_ $$2ORCID$$aWORKING_PAPER
000873125 3367_ $$028$$2EndNote$$aElectronic Article
000873125 3367_ $$2DRIVER$$apreprint
000873125 3367_ $$2BibTeX$$aARTICLE
000873125 3367_ $$2DataCite$$aOutput Types/Working Paper
000873125 500__ $$aSubmitted to Combustion Symposium 2020
000873125 520__ $$aTurbulence is still one of the main challenges for accurately predicting reactive flows. Therefore, the development of new turbulence closures which can be applied to combustion problems is essential. Data-driven modeling has become very popular in many fields over the last years as large, often extensively labeled, datasets became available and training of large neural networks became possible on GPUs speeding up the learning process tremendously. However, the successful application of deep neural networks in fluid dynamics, for example for subgrid modeling in the context of large-eddy simulations (LESs), is still challenging. Reasons for this are the large amount of degrees of freedom in realistic flows, the high requirements with respect to accuracy and error robustness, as well as open questions, such as the generalization capability of trained neural networks in such high-dimensional, physics-constrained scenarios. This work presents a novel subgrid modeling approach based on a generative adversarial network (GAN), which is trained with unsupervised deep learning (DL) using adversarial and physics-informed losses. A two-step training method is used to improve the generalization capability, especially extrapolation, of the network. The novel approach gives good results in a priori as well as a posteriori tests with decaying turbulence including turbulent mixing. The applicability of the network in complex combustion scenarios is furthermore discussed by employing it to a reactive LES of the Spray A case defined by the Engine Combustion Network (ECN).
000873125 536__ $$0G:(DE-HGF)POF3-512$$a512 - Data-Intensive Science and Federated Computing (POF3-512)$$cPOF3-512$$fPOF III$$x0
000873125 536__ $$0G:(DE-Juel1)jhpc55_20180501$$aUsing deep learning to predict statistics of turbulent flows at high Reynolds numbers (jhpc55_20180501)$$cjhpc55_20180501$$fUsing deep learning to predict statistics of turbulent flows at high Reynolds numbers$$x1
000873125 536__ $$0G:(DE-Juel1)sbda006_20151101$$aDirect Numerical Simulations of Fluid Turbulence at High Reynolds Numbers (sbda006_20151101)$$csbda006_20151101$$fDirect Numerical Simulations of Fluid Turbulence at High Reynolds Numbers$$x2
000873125 588__ $$aDataset connected to arXivarXiv
000873125 7001_ $$0P:(DE-HGF)0$$aGauding, Michael$$b1
000873125 7001_ $$0P:(DE-HGF)0$$aLian, Zeyu$$b2
000873125 7001_ $$0P:(DE-HGF)0$$aDenker, Dominik$$b3
000873125 7001_ $$0P:(DE-HGF)0$$aDavidovic, Marco$$b4
000873125 7001_ $$0P:(DE-HGF)0$$aKleinheinz, Konstantin$$b5
000873125 7001_ $$0P:(DE-Juel1)158080$$aJitsev, Jenia$$b6$$ufzj
000873125 7001_ $$0P:(DE-HGF)0$$aPitsch, Heinz$$b7
000873125 8564_ $$uhttps://arxiv.org/abs/1911.11380
000873125 8564_ $$uhttps://juser.fz-juelich.de/record/873125/files/1911.11380.pdf$$yOpenAccess
000873125 8564_ $$uhttps://juser.fz-juelich.de/record/873125/files/1911.11380.pdf?subformat=pdfa$$xpdfa$$yOpenAccess
000873125 909CO $$ooai:juser.fz-juelich.de:873125$$pdnbdelivery$$pdriver$$pVDB$$popen_access$$popenaire
000873125 9101_ $$0I:(DE-588b)36225-6$$6P:(DE-HGF)0$$aRWTH Aachen$$b0$$kRWTH
000873125 9101_ $$0I:(DE-588b)5008462-8$$6P:(DE-Juel1)158080$$aForschungszentrum Jülich$$b6$$kFZJ
000873125 9131_ $$0G:(DE-HGF)POF3-512$$1G:(DE-HGF)POF3-510$$2G:(DE-HGF)POF3-500$$3G:(DE-HGF)POF3$$4G:(DE-HGF)POF$$aDE-HGF$$bKey Technologies$$lSupercomputing & Big Data$$vData-Intensive Science and Federated Computing$$x0
000873125 9141_ $$y2019
000873125 915__ $$0StatID:(DE-HGF)0510$$2StatID$$aOpenAccess
000873125 920__ $$lyes
000873125 9201_ $$0I:(DE-Juel1)JSC-20090406$$kJSC$$lJülich Supercomputing Center$$x0
000873125 980__ $$apreprint
000873125 980__ $$aVDB
000873125 980__ $$aUNRESTRICTED
000873125 980__ $$aI:(DE-Juel1)JSC-20090406
000873125 9801_ $$aFullTexts