You are on page 1of 1

Large eddy simulation of rotating turbulence

Hao Lu, Christopher Rutland, Leslie Smith Sponsors: NSF, SCREMS


Introduction Forced rotating turbulence – by DNS A-posteriori test of decaying flows Conclusions
Turbulence subjected to system rotation provides a simple Force low energy Decay at 1283. And  New models more accurate
data by Gaussian filter the initial data initial Reλ=86, Roω3=0.41
configuration for studying the characteristic features of both
white-noise scheme into 323 for LES  Energy decay rate
anisotropic turbulent flows and model performances in
anisotropic turbulent flows. It is of great importance in Energy Evolution 3D Spectra @t=0.3  Energy spectra

m
engineering and geophysics. The most important

tr u
-3
1.0 k

c
SGS energy production, dissipation, backscatter

pe

Ene
application in the first case is the development and the

yS
rg
rgy

e
0.01
design of turbo-machinery. Here one has to take into

En
 New models able to duplicate important features of forced

Spe
Without Model Without Model
account the detailed properties of the turbulent fluids, which rotating turbulence

ctru

Kr/Kr,ini
0.9
SSM

m
pass through the device and are rotated (e.g., by the motion SSM
1E-3
Too dissipative at

E(k)
of the turbine blades). A detailed understanding of rotational
large scales via SM  Quasi 2D flow at large scales
effects on flow characteristics is essential for an advanced 0.8
MixSCDSM
1E-4
 Cyclonical vortices
layout of these machines. Second, the whole field of MixGCDSM MixSCDSM 3
128 DNS
geophysics is crucially determined by planetary rotation,
sm MixGCDSM SM
Reverse energy transfers from small scales to large
Development of cyclonic two-dimensional coherent structures appearing in rotating 1E-5

which influences both atmospheric flows and oceanic flows turbulence as indicated by iso-surfaces of vorticity, contours of kinetic energy and
0.7 3
128 DNS scales
velocity vectors: (a) initial very low energy level isotropic turbulence; (b) final state (at
0.0 0.1 0.2 0.3 0.4 1 10 Computational efficient (less CPU cost) using
and affects global climate as well as short-term weather Time k
MixGCDSM (spectral code, on Intel P4 3.4GHz
normalized time 3.88) of large scale forced rotating case; (c) final state (at normalized
forecasting. Understanding the fundamental processes in time 3.68) of small scale forced rotating case.
Ω=1[rad/s], υ=0.002 Linux Cluster, 1cpu, 323)
these fluid layer forms the basis for a detailed analysis of TSM : TMixGCDSM : TMixSCDSM : TDynSM =
complex phenomena such as the development of climate A-posteriori test of intermediate scale forced rotating 0.24 : 0.33 : 0.59 : 1.00
anomalies (such as El Niño), the formation of hurricanes Scatter plot of τ 11 by SCDSM
and tidal waves, the spreading of pollutants, and the turbulence using MixGCDSM Time sequence
Literature cited
16

14

oceanic circulation of nutrients. 12


slope= β=1.003
Assume: b    a   . Ideally, expect   1,   1
[1] C. Cambon, N. N. Mansour, and K. D. Squires.
10
 ab    a  b 
Regression:  
Many features are well known. The trends toward 2D [1, 2],
SCDSM

 a 2    a 2
11

6
Anisotropic structure of homogeneous turbulence subjected
τ

the cyclonic/anti-cyclonic asymmetry in favor of cyclones 4


Correlation:  
 ab    a  b 
to uniform rotation. Proceeding of the Summer Program,
variance is
 a    a 2  b2    b 2
2
2

[3], and the influence of the background rotation on the


described by: ρ=0.999
0

energy transfers [2], are among the most challenging issues


0 2 4 6
DNS
τ11
8 10 12 14 16 pages 397–420, 1994.
for rotating turbulence. 3 17.20
2.306
3 17.20
2.306
[2] C. Cambon, N. N. Mansour, and F. S. Godeferd. Energy
Scatter plot analysis, and
transfer in rotating turbulence. J. Fluid Mech., 337:303–332,
0.3093 0.3093
0.04147 0.04147
Researchers have used SGS models to simulate rotating correlation/regression
analysis. The correlation
2
0.005561
7.457E-42
0.005561
7.457E-4

turbulence. For rotating turbulence, however, there still coefficient can represent
1.000E-4
DNS
1.000E-4
SCDSM
1997.
τ11 τ11
the variance between the 1 1

exists a need for a better understanding of SGS models [4] modeled and the exact [3] L. M. Smith and Y. Lee. On near resonances and
terms on the scatter plot 0 0

because anisotropic characteristics may influence LES symmetry breaking in forced rotating flows at moderate
y

y
and on the PDF diagram.
The regression coefficient
modeling and comparative studies of model performance in can represent the contour
-1 -1
Rossby number. J. Fluid Mech., 535:111–142, 2005.
rotating turbulence are insufficient. level ratio between the
modeled and the exact
-2 -2

terms, the slope of the [4] C. G. Speziale. Subgrid scale stress models for the
regression (scatter) line. -3 -3
large-eddy simulation of rotating turbulent flows. Geophys.
Objectives -3 -2 -1 0
x
1 2 3 -3 -2

Comparison of contour plots of SGS stress τ11 (left) and similarity type consistent dynamic
-1 0
x
1 2 3

Astrophys. Fluid Dynamics, 33:199–222,1985.


structure modeled stress τ11SCDSM (right) at z=0 layer. Flow is the small scale forced case ( (c) at
 Direct numerical simulation (DNS) of rotating turbulence: slide 3). Cutoff wave-number: k=11.6.
[5] E. Pomraning and C. J. Rutland. Dynamic one-equation
We have small scale forced cases, large scale forced nonviscosity large eddy simulation model. AIAA Journal,
cases and decaying cases. DNS provides data for LES
model development. A-priori test A-posteriori test of large scale forced rotating turbulence 40(4):689–701, April 2002.
[Traditional models]
 Developing sub-grid scale models: Model has the 1.0
SSM: Scale Similarity Model
using MixGCDSM Time sequence Acknowledgement
capability to capture small-scale turbulence properties, GM: Gradient Model
β

and anisotropic features of rotating turbulence.


0.8 This material is based upon work supported by the National
DSM: Dynamic Structure Model
Science Foundation (NSF) under Grant No. 0500056 and
[New models]
33

0.6
by the NSF Scientific Computing Research Environments in
Development of new models DSM
GCDSM
GCDSM: Gradient type Consistent DSM
the Mathematical Sciences (SCREMS) under Grant No.
0.4 SCDSM SCDSM: Similarity type Consistent DSM
 Dynamic structure model [5] SSM DMS-0532085. We would like to thank Dr. Yun-Liang Wang
GM
 Lij  for helpful discussions and the Engine Research Center at
 ijDSM   2k sgs , where Lij are modified Leonard terms.
0.2
1.0
L
 mm 
the University of Wisconsin - Madison for providing
β

0 20 40 60 80 100 120 140


 Consistent dynamic structure models for rotating flow k= π
/∆ computing resources.
0.8
 Gij  [Case description]
 ij
GCDSM
 2ksgs , where Gij are gradient terms. Rotating flow forced at large For further information
3

G
 mm  scales with rotation rate of 12 0.6 DSM

 ij  [rad/s].
GCDSM
SCDSM
Please contact haolu@wisc.edu. More information on this
 
SCDSM
2k sgs , where ij are modeling of Lij  Cij . SSM and related projects can be obtained at
ij [Tested quantities]
  mm 
0.4
GM
 Mixed modeling,  ij  u  S ij , u  Ck  k sgs  
2 ' 3 4
Regression coefficients of τ 33 http://homepages.cae.wisc/~luh
and ∂τ 3i /∂x i 0.2
0 20 40 60 80 100 120 140
k= π
/∆

University of Wisconsin –– Engine Research Center

You might also like