Insights into second-moment closure modelling performance for stratified wakes from DNS ensembles
ORAL
Abstract
Wake flows in the environment, usually characterized by high Re and Fr, require large computational resources, prohibiting the use of DNS/LES. RANS based models are more practical, but the inherently complex dynamics render RANS eddy-viscosity modelling inappropriate. RANS Second-Moment Closure (SMC) modelling is more suitable, and here we study their performance for stratified towed and momentumless wakes. 11-equation based modelling is used, and a range of sub-model complexity is employed for diffusion, pressure strain/scrambling, and dissipation terms. These sub-models are evaluated in terms of how well DNS represents them in comparison to exact Reynolds averaged terms. Statistically converged DNS results for direct comparison are obtained by an ensemble average of 100 realizations. These highlight important inconsistencies in model performance, such as non-negligible pressure-diffusion and dissipation anisotropy, as presented by the authors in an earlier study of stratified shear layers. Some mean flow elements are captured well for stratified wakes, but the flow dynamics are inaccurately predicted. Simple corrections to model shortcomings are being studied and tested.
–
Presenters
-
Naman Jain
Pennsylvania State University
Authors
-
Naman Jain
Pennsylvania State University
-
Xinyi Huang
Pennsylvania State University
-
Xiang Yang
Pennsylvania State University
-
Robert F Kunz
Pennsylvania State University, Penn State, Pennsylvania State University Department of Mechanical Engineering