![Sample our Earth Sciences journals, sign in here to start your access, latest two full volumes FREE to you for 14 days]({"type":"image" , "src" : "/sda/5930/TOC-Subject-Sample-2021-Earth-Sciences.jpg"})
Abstract
The paper presents the results of the application of large-eddy simulation (LES) to turbulent channel flow with a varying pressure gradient obtained by an appropriately specified shape of one of the walls. The main objective of the paper is to assess various subgrid scale (SGS) models implemented in two different codes as well as to assess the sensitivity of the predictive accuracy to grid resolution. Additionally, the role of SGS viscosity, controlled by a constant parameter of the SGS model, was investigated. The simulations were performed with inlet conditions corresponding to two Reynolds numbers: 
and 
. The consistency and the accuracy of simulations are evaluated using direct numerical simulation (DNS) results. It is demonstrated that all SGS models require a comparable minimum grid refinement in order to capture accurately the recirculation region. Such a test case with a reversal flow, where the turbulence transport is dictated by the dynamics of the large-scale eddies, is well suited to demonstrate the predictive features of the LES technique.
Acknowledgements
This work was supported by CISIT (International Campus on Safety and Inter-modality in Transport) and WALLTURB (A European synergy for the assessment of wall turbulence), which is funded by the EC under the 6th framework program (CONTRACT: AST4-CT-2005-516008). The DNS was performed through two successive DEISA Extreme Computing Initiatives (DEISA is Distributed European Infrastructure for Supercomputing Applications). This work was granted access to the HPC resources of IDRIS (Institut du Developpement et des Ressources en Informatique Scientifique) under the allocation 2010-021741 made by GENCI (Grand Equipement National de Calcul Intensif) and to the HPC resources of Lille University.