Large eddy simulation of rotating turbulent channel flow with a new dynamic global-coefficient nonlinear subgrid stress model

Zixuan Yang, Guixiang Cui, Chunxiao Xu, Zhaoshun Zhang

Research output: Contribution to journalArticlepeer-review

20 Scopus citations

Abstract

In this paper, a new dynamic global-coefficient nonlinear subgrid scale (SGS) model is proposed for large eddy simulation (LES) of rotating turbulent channel flow. The basic model is a nonlinear model with a tensorial polynomial relation between the SGS stress and the resolved strain rate tensor. A new dynamic procedure is proposed to determine the model coefficients of the nonlinear model. The new dynamic method is derived from the globally averaged transport equation of the Reynolds shear stress, on which the rotation has strong and direct effects. The new dynamic nonlinear SGS model is examined in rotating turbulent channel at Re = umh/ν = 7000, Ro = 2Ωh/um = 0.3 and 0.6, where Reynolds number Re and Rotation number Ro are defined by bulk mean velocity um, half channel width h, kinematic viscosity ν and angular velocity of spanwise rotation Ω. The statistical results obtained from the new model agree well with those from direct numerical simulation (DNS). The new model also successfully predicts the major structures in rotating turbulent channel flow, such as Taylor-Görtler vortices and streaks.

Original languageEnglish (US)
Pages (from-to)1-20
Number of pages20
JournalJournal of Turbulence
Volume13
DOIs
StatePublished - 2012

Bibliographical note

Funding Information:
The work is supported by Natural Science Foundation of China (NSFC Grant 11132005, 10925210) and Tsinghua National Laboratory for Information Science and Technology.

Keywords

  • Globally dynamic method
  • Nonlinear SGS model
  • Rotating turbulent channel flow
  • Source file coding
  • Submission instructions

Fingerprint

Dive into the research topics of 'Large eddy simulation of rotating turbulent channel flow with a new dynamic global-coefficient nonlinear subgrid stress model'. Together they form a unique fingerprint.

Cite this