### Skapa referens, olika format (klipp och klistra)

**Harvard**

Javadi, A. och Nilsson, H. (2013) *Advanced numerical prediction of strongly swirling turbulent flows*.

** BibTeX **

@conference{

Javadi2013,

author={Javadi, Ardalan and Nilsson, Håkan},

title={Advanced numerical prediction of strongly swirling turbulent flows},

booktitle={5th International Workshop on Cavitation and Dynamic Problems in Hydraulic Machinery},

abstract={The strongly swirling turbulent flow through an abrupt expansio
n is investigated using highly resolved LES and hybrid RANS-
LES, to shed more light on the stagnation region and the helical vortex breakdown. The vortex breakdown in an abrupt
expansion resembles the so-called vortex rope occurring in hydro
power draft tubes. It is know
n that the large-scale helical
vortex structures can be captured by re
gular RANS turbulence models. However, the spurious suppression of the small-scale
structures should be avoided using less diffusive methods [1]. The present work compares LES and hybrid RANS-LES results
with the experimental measurement of Dellenback [2]. The com
putations are conducted using a general non-orthogonal, block-
structured, finite-volume method with a fully collocated storage available in the OpenFOAM CFD code. The dynamics of the
flow is studied at two Reynolds numbers,
Re
=6.0×10
4
and
Re
=10.0×10
4
to study the effect of high Reynolds turbulent flow
with almost constant high swirl number (
S
r
=1.16 and
S
r
=1.23, respectively). The delayed detached eddy simulation (DDES)
Spalart-Allmaras model [3] and the dynamic one-equati
on LES model are used to predict the coarse (8.2×10
6
cells) and fine
(12×10
6
cells) discretized computational domain. The averaged velocity field, pressure field and the root mean square of the
velocity fluctuations are captured and investigated qualitatively.
Fig. 1 shows the instantaneous pressure iso-surface for the two operating conditions. The flow with lower Reynolds number gives much weaker outburst although the frequency of the
structures seems to be constant for the plateau swirl number. },

year={2013},

keywords={Swirling flow, Turbulence, LES, Hybrid RANS-LES, OpenFOAM},

}

** RefWorks **

RT Conference Proceedings

SR Print

ID 236205

A1 Javadi, Ardalan

A1 Nilsson, Håkan

T1 Advanced numerical prediction of strongly swirling turbulent flows

YR 2013

T2 5th International Workshop on Cavitation and Dynamic Problems in Hydraulic Machinery

AB The strongly swirling turbulent flow through an abrupt expansio
n is investigated using highly resolved LES and hybrid RANS-
LES, to shed more light on the stagnation region and the helical vortex breakdown. The vortex breakdown in an abrupt
expansion resembles the so-called vortex rope occurring in hydro
power draft tubes. It is know
n that the large-scale helical
vortex structures can be captured by re
gular RANS turbulence models. However, the spurious suppression of the small-scale
structures should be avoided using less diffusive methods [1]. The present work compares LES and hybrid RANS-LES results
with the experimental measurement of Dellenback [2]. The com
putations are conducted using a general non-orthogonal, block-
structured, finite-volume method with a fully collocated storage available in the OpenFOAM CFD code. The dynamics of the
flow is studied at two Reynolds numbers,
Re
=6.0×10
4
and
Re
=10.0×10
4
to study the effect of high Reynolds turbulent flow
with almost constant high swirl number (
S
r
=1.16 and
S
r
=1.23, respectively). The delayed detached eddy simulation (DDES)
Spalart-Allmaras model [3] and the dynamic one-equati
on LES model are used to predict the coarse (8.2×10
6
cells) and fine
(12×10
6
cells) discretized computational domain. The averaged velocity field, pressure field and the root mean square of the
velocity fluctuations are captured and investigated qualitatively.
Fig. 1 shows the instantaneous pressure iso-surface for the two operating conditions. The flow with lower Reynolds number gives much weaker outburst although the frequency of the
structures seems to be constant for the plateau swirl number.

LA eng

OL 30