Machine learning methods for development of data-driven turbulence models

Sergey N. Yakovenko, Omid Razizadeh

Research output: Chapter in Book/Report/Conference proceedingConference contributionResearchpeer-review

1 Citation (Scopus)

Abstract

The implementation of the machine learning methods of convolutional neural network combined with support vector machines to enhance RANS closure models is presented. The RANS models are not universal and accurate, however they are computationally affordable. Finding a way to improve the model predictability will be an advantage, and machine learning algorithms based on available high-fidelity data sets for canonical flow cases obtained from DNS and measurements can be helpful for this. The application of these algorithms for a fully-developed turbulent channel flow between parallel walls, with periodic hills and for other cases is considered.

Original languageEnglish
Title of host publicationHigh-Energy Processes in Condensed Matter, HEPCM 2020
Subtitle of host publicationProceedings of the XXVII Conference on High-Energy Processes in Condensed Matter, Dedicated to the 90th Anniversary of the Birth of RI Soloukhin
EditorsVasily M. Fomin
PublisherAmerican Institute of Physics Inc.
Number of pages5
ISBN (Electronic)9780735440180
DOIs
Publication statusPublished - 26 Oct 2020
Event27th Conference on High-Energy Processes in Condensed Matter, HEPCM 2020 - Novosibirsk, Russian Federation
Duration: 29 Jun 20203 Jul 2020

Publication series

NameAIP Conference Proceedings
Volume2288
ISSN (Print)0094-243X
ISSN (Electronic)1551-7616

Conference

Conference27th Conference on High-Energy Processes in Condensed Matter, HEPCM 2020
CountryRussian Federation
CityNovosibirsk
Period29.06.202003.07.2020

Keywords

  • SIMULATION
  • FLOW

Fingerprint Dive into the research topics of 'Machine learning methods for development of data-driven turbulence models'. Together they form a unique fingerprint.

Cite this