Premium
URANS computations for an oscillatory non‐isothermal triple‐jet using the k – ε and second moment closure turbulence models
Author(s) -
Nishimura M.,
Kimura N.
Publication year - 2003
Publication title -
international journal for numerical methods in fluids
Language(s) - English
Resource type - Journals
SCImago Journal Rank - 0.938
H-Index - 112
eISSN - 1097-0363
pISSN - 0271-2091
DOI - 10.1002/fld.510
Subject(s) - turbulence , mechanics , prandtl number , turbulence kinetic energy , jet (fluid) , physics , reynolds averaged navier–stokes equations , k epsilon turbulence model , computation , reynolds number , turbulence modeling , computational fluid dynamics , turbulent prandtl number , boundary value problem , thermodynamics , mathematics , heat transfer , nusselt number , algorithm , quantum mechanics
L ow R eynolds number turbulence s tress and heat f lux equation m odels (LRSFM) have been developed to enhance predictive capabilities. A new method is proposed for providing the wall boundary condition for dissipation rate of turbulent kinetic energy, ε , to improve the model capability upon application of coarse meshes for practical use. The proposed method shows good agreement with accepted correlations and experimental data for flows with various Reynolds and Prandtl numbers including transitional regimes. Also, a mesh width about 5 times or larger than that used in existing models is applicable by using the present boundary condition. The present method thus enhanced computational efficiency in applying the complex turbulence model, LRSFM, to predictions of complicated flows. U nsteady R eynolds a veraged N avier– S tokes (URANS) computations are conducted for an oscillatory non‐isothermal quasi‐planar triple‐jet. Comparisons are made between an experiment and predictions with the LRSFM and the standard k – ε model. A water test facility with three vertical jets, the cold in between two hot jets, simulates temperature fluctuations anticipated at the outlet of a liquid metal fast reactor core. The LRSFM shows good agreement with the experiment, with respect to mean profiles and the oscillatory motion of the flow, while the k – ε model under‐predicts the mixing due to the oscillation, such that a transverse mean temperature difference remains far downstream. Copyright © 2003 John Wiley & Sons, Ltd.