On the prediction capabilities of two SGS models for large-eddy simulations of turbulent incompressible wall-bounded flows in OpenFOAM

Cogent Engineering

Carlos Duque-Daza¹  Jesus Ramirez-Pastran¹ 

[1] GNUM Research Group, Universidad Nacional de Colombia, Bogota, Colombi;

关键词: turbulence;    large-eddy simulations;    sub-grid scale models;    OpenFOAM;   

DOI  :  10.1080/23311916.2019.1679067

来源: Taylor & Francis

【 摘 要 】

A comprehensive evaluation of the prediction capabilities of two Sub-Grid-Scale (SGS) models, for Large-Eddy Simulations (LES) of incompressible turbulent flows, is performed by studying two cases of flows through wall-bounded domains: a standard smooth-wall channel and a channel featuring fixed geometrical perturbations in its bottom wall. The SGS models were computationally implemented in OpenFOAM’s framework and numerical results have been evaluated using available DNS data of turbulent flow in the standard channel flow case at four different friction Reynolds numbers, i.e. Reτ =180, 395, 550 and 950. A number of well-known features were employed to judge about the reliability of the selected SGS models in the standard channel case, including: mean flow velocity profiles, ensemble averages of first and second-order turbulent statistics, and profiles of TKE. One of the SGS models was employed to explore the second case, a flow with high levels of strain and formation of large-scale vortical structures induced by cavities and/or ribs-like protuberances in one of the boundaries. Our numerical experiments show that both SGS models exhibit good quantitative and qualitatively agreement with most of the important flow features in both wall-bounded flows considered, although the attained levels of accuracy exhibited a mesh-resolution dependency. In the non-smooth-wall case, the selected SGS model was able to successfully capture, in accurate manner, effects as turbulence modulation and changes of the skin friction coefficient. It is clear that the implementations of the SGS models were successful in predicting the turbulent behaviour of the selected flow cases.

【 授权许可】

CC BY