Lecture # 05 Turbulence Modelling

Lecture # 05 Turbulence Modelling
Turbulence Modeling
19 May 2010
1
JetWings
Outline
1. Characteristics of turbulence
2. Approaches to predicting turbulent flows
3. Reynolds averaging
4. RANS equations and unknowns
5. The Reynolds-Stress Equations
6. The closure problem of turbulence
7. Characteristics of wall-bound turbulent flows
8. Turbulence models and ranges of applicability
7.1. RANS
7.2. LES/DES
7.3. DNS
9. Example: diffuser
19 May 2010
2
JetWings
Characteristics of turbulence
• Randomness and fluctuation: u = U + u′
• Nonlinearity: Reynolds stresses from the nonlinear

convective terms
• Diffusion: enhanced diffusion of momentum, energy
etc.
• Vorticity/eddies/energy cascade: vortex stretching
19 May 2010
3
JetWings
Characteristics of turbulence
•Dissipation: occurs at smallest scales

•Three-dimensional: fluctuations are always 3D
•Coherent structures: responsible for a large part of
the mixing
•A broad range of length and time scales: making
DNS very difficult
19 May 2010
4
JetWings
Approaches to predicting turbulent flows

• AFD, EFD and CFD:
– AFD: No analytical solutions exist
– EFD: Expensive, time-consuming, and sometimes impossible
(e.g. fluctuating pressure within a flow)
– CFD: Promising, the need for turbulence modeling
• Another classification scheme for the approaches
– The use of correlations: CD = f ( Re )
– Integral equations: reduce PDE to ODE for simple cases
– One-point closure: RANS equations + turbulent models
– Two-point closure: rarely used, FFT of Two-point equations
– LES: solve for large eddies while model small eddies
– DNS: solve NS equations directly without any model
19 May 2010
5
JetWings
Deeper insights on RANS/URANS/LES
RANS
URANS
LES
19 May 2010
6
JetWings
Reynolds Averaging
• Time averaging: for stationary turbulence
• Spatial averaging: for homogenous
turbulence
• Ensemble averaging: for any turbulence
• Phase averaging: for turbulence with
periodic motion
19 May 2010
7
JetWings
RANS equations and unknowns

• RANS equation
∂U i ∂U i ∂P ∂
ρ
∂t
+ ρU j
∂x j
=− +
∂xi ∂x j
(
µ2 S ji − ρ u′j ui ′ )
• RANS equation in conservative form
∂U i ∂ ∂P ∂
ρ
∂t
+ρ
∂x j
( ′ ′
U jU i + u j ui = −) +
∂ xi ∂ x j
( µ2 S ji )
• Numbers of unknowns and equations

– Unknowns: 10 = P (1) + U (3) + ( −ui′u j′)(6)
– Equations: 4 = Continuity (1) + Momentum (3)
19 May 2010
8
JetWings
The Reynolds-Stress Equation

•Derivation: Taking moments of the NS equation.
Multiply the NS equation by a fluctuating property and time
average the product. Using this procedure, one can derive a
differential equation for the Reynolds-stress tensor.
∂τ ij ∂τ ij ∂U j ∂U i ∂ui′ ∂u′j
+ Uk = −τ ik − τ jk + 2ν
∂t ∂xk ∂xk ∂xk ∂xk ∂xk
 ui′ ∂p′ u ′j ∂p ′ ∂  ∂τ ij 
+ + + ν + ui′u j′uk ′
 ρ ∂x j ρ ∂xi  ∂xk  ∂xk 
 
•NEW Equations: 6 = 6 equation for the Reynolds stress tensor
•NEW Unknowns: 22 = 6 + 6 + 10
 ui′ ∂p′ u ′j ∂p ′
 +  → 6 unkowns
 ρ ∂x j ρ ∂xi 
 
∂ui′ ∂u′j
2ν → 6 unkowns ui′u j′uk ′→ 10 unkowns
∂xk ∂xk
19 May 2010 9
9
JetWings
The closure problem of turbulence

• Because of the non-linearity of the Navier-
Stokes equation, as we take higher and higher
moments, we generate additional unknowns at
each level.
• In essence, Reynolds averaging is a brutal
simplification that loses much of the information
contained in the Navier-Stokes equation.
• The function of turbulence modeling is to devise
approximations for the unknown correlations in
terms of flow properties that are known so that a
sufficient number of equations exist.
• In making such approximations, we close the
system.
19 May 2010
10
JetWings
Turbulent models (RANS)

• Boussinesq eddy-viscosity approximation
• Algebraic (zero-equation) models
– Mixing length
– Cebeci-Smith Model
– Baldwin-Lomax Model
• One-equation models
– Baldwin-Barth model
– Spalart-Allmaras model
• Two-equation models
– k-ε model
– k-ω model
• Four-equation (v2f) models
• Reynolds-stress (seven-equation) models
19 May 2010
11
JetWings
Turbulent models (RANS)

• Boussinesq eddy-viscosity approximation
 ∂U i ∂U j  2
−ui′u j′ = ν T  +  − δ ij k
 ∂x ∂xi
 j  3
• Dimensional analysis shows:ν T = Cµ qL , where q is a

turbulence velocity scale and L is a turbulence length
1
scale. Usually q = 2k where k = ui′ui ′ is the turbulent
2
kinetic energy. Models that do not provide a length
scale are called incomplete.
19 May 2010
12
JetWings
Turbulent models (0-eqn RANS)

dU
• Mixing length model: ν T = lmix
2
dy
• Assume lmix = αδ ( x ) for free shear flow, then

─ α=0.180 for far wake
─ α=0.071 for mixing layer
─ α=0.098 for plane jet
─ α=0.080 for round jet
• Comments:
─Reliable only for free shear flows with different α values
─Not applicable to wall-bounded flows
19 May 2010
13
JetWings

Cebeci-Smith Model (Two-layer model)
ν Ti , y ≤ ym
νT =  Where ym is the smallest value of y for ν T = ν T
ν To y > ym
i o
12
 ∂U  2  ∂V 2  +
A0+
Inner layer: ν Ti = lmix 
2
 +   lmix = κ y 1 − e− y 
 ∂y   ∂x    
Outer layer: ν T = αU eδ v* FKleb ( y; δ )
0
−1 2
 dP dx 
Closure coefficients: κ = 0.40 α = 0.0168 A+ = 26 1 + y
−1  ρUτ2 
  y 
6
FKleb ( y; δ ) = 1 + 5.5   
δ
δ v* = ∫ (1 − U U e ) dy
Comments:   δ   0
Three key modifications to mixing length model;

Applicable to wall-bounded flows;
Applicable to 2D flows only;
Not reliable for separated flows;
δ , δ v* , U e difficult to determine in some cases;
19 May 2010
14
JetWings

Baldwin-Lomax Model (Two-layer model)
ν Ti , y ≤ ym
νT =  Where ym is the smallest value of y for ν T = ν T
ν To y > ym
i o
+
A0+
Inner layer: ν T = l ω lmix 2
= κ y 1 − e− y 
i mix
 
Outer layer: ν T = α Ccp Fwake FKleb ( y; ymax CKleb )
0
Closure coefficients: κ = 0.40 α = 0.0168 A+ = 26 Ccp = 1.6 C Kleb = 0.3 Cwk = 1

−1 12
  y  
6
 ∂V ∂U  2  ∂W ∂V 2  ∂U ∂W  2 
FKleb ( y; ymax CKleb ) = 1 + 5.5    ω =  −  + −  + −  
  ymax CKleb    ∂x ∂y   ∂y ∂z   ∂z ∂x  
Comments:
Applicable to 3D flows;
Not reliable for separated flows;
No need to determine δ , δ v , U e
*
19 May 2010
15
JetWings

Baldwin-Barth model
Kinematic eddy viscosity : ν T = Cµν R%T D1D2
Turbulence Reynolds number :

(
∂ ν R%T ) + U ∂ (ν R% ) = ( C ( )
∂ 2 ν R%T 1 (
∂ν T ∂ ν R%T )
ε 2 f 2 − Cε 1 ) ν R%T P+ ν( + ν T σ ε ) −
T
∂t
j
∂x j ∂x ∂xk σ
k ε ∂xk ∂xk
Closure coefficients and auxiliary relations:

+ Cµ
Cε 1 = 1.2 Cε 2 = 2.0 C µ = 0.09 A0+ = 26 A2 = 10 κ = 0.41
1
= ( Cε 1 − Cε 2 )
σε κ 2
 ∂U ∂U j  ∂U 2 ∂U ∂U  − y + A2+
P = ν T  i
+ 
i
− k k
 D1 = 1 − e
− y+ A0+ D2 = 1 − e
 ∂x ∂xi  ∂x j 3 ∂xk ∂xk 
 j
Cε 1  Cε 1   1   1  D2 − y + A0+ D +
A2+ 
f2 = + 1 −  + D D
1 2 ⋅  D D +  +e + +1 e− y 
Cε 2  Cε 2   κ y +
1 2
  D1 D2  A0 A2  
19 May 2010
16
JetWings

Spalart-Allmaras model
Kinematic eddy viscosity : ν T = ν% f v1
Eddy viscosity equation:

∂ν% ∂ν% ν %  1 ∂  ∂ν%  cb1 ν∂ % ν∂ %
2
+U j = cb1S%ν% − cw1 f w   +  (ν + ν% ) +
∂t ∂x j  d  σ ∂xk  ∂xk  σ ∂xk ∂xk
Closure coefficients and auxiliary relations:

cb1 (1 + cb 2 )
cb1 = 0.1355 cb 2 = 0.622 cv1 = 7.1 σ = 2 3 cw1 = +
κ2 σ
ν% ν%
cw 2 = 0.3 cw 2 = 2.0 κ = 0.41 χ = g = r + cw 2 ( r − r )6 r=
ν S%κ 2 d 2
χ3 χ 16
 1 + cw6 3  ν%
f v1 = fv 2 = 1 − fw = g  6 6  S% = S + 2 2 f v 2 S = 2Ωij Ωij
χ +c
3 3
1 + χ f v1 κ d
v1  g + cw 3 
19 May 2010
17
JetWings

Comments on one-equation models:
1.One-equation models based on turbulence kinetic energy are incomplete as
they relate the turbulence length scales to some typical flow dimension. They
are rarely used.
2.One-equation models based on an equation for the eddy viscosity are
complete such as Baldwin-Barth model and Spalart-Allmaras model.
3.They circumvent the need to specify a dissipation length by expressing the
decay, or dissipation, of the eddy viscosity in terms of spatial gradients.
4 Spalart-Allmaras model can predicts better results than Baldwin-Barth
model, and much better results for separated flow than Baldwin-Barth model
and algebraic models.
5 Also most of DES simulations are based on the Spalart-Allmaras model.
19 May 2010
18
JetWings

k-ε model: ν T = Cµ k 2 ε
∂k ∂k ∂U i ∂  ∂ k
+U j = τ ij −ε + (ν + ν T σ k ) 
∂t ∂x j ∂x j ∂x j  ∂ x j 

∂ε ∂ε ε ∂U i ε 2 ∂  ∂ε 
+U j = Cε 1 τ ij − Cε 2 + (ν +ν T σ ε ) 
∂t ∂x j k ∂ xj k ∂ x j  ∂ x j 
,
Cε 1 = 1.44 Cε 2 = 1.92 Cµ = 0.09 σ k = 1.0 σ ε = 1.3 ω = ε ( Cµ k ) l = Cµ k 3 2 ε
νT = k ω
,
k-ω model:
∂k ∂k ∂U i ∂  ∂ k
= τ ij − β * kω + (ν + σ ν T )
,
+U j *

∂t ∂x j ∂x j ∂ xj  ∂ x j 
,
∂ω ∂ω ω ∂U i ∂  ω
∂ 
+U j = α τ ij − βω 2 + (ν + σν T ) 
∂t ∂x j k ∂x j ∂x j  ∂ x j 
α=
13
β = β0 fβ β * = β 0* f β σ=
1 1
β0 =
9 1 + 70 χω
* σ* = fβ =
25 2 2 125 1 + 80 χω
1, χk ≤ 0
Ωij Ω jk Ski  1 ∂k ∂ω
χω = β 0* =
9 fβ * = 1 + 680 χ k2 χk =
(β ω) 1 + 400 χ 2 χ k > 0 ω 3 ∂x j ∂x j
* 3
0 100
 k
ε = β *ω k
19 May 2010 l = k1 2 ω
19
JetWings

Comments on two-equation models:
1. Two-equation models are complete;
2. k-ε and k-ω models are the most widely used two-
equation models and a lot of versions exist. For example, a
,
popular variant of k-ω model introduced by Menter has
been used in our research code CFDSHIP-IOWA. There
are also a lot of low-Reynolds-number versions with
different damping functions.
3. k-ω model shows better results than k-ε model for
flows with adverse pressure gradient and separated flows
as well as better numerical stability.
19 May 2010
20
JetWings

v2f-kε model: ν T = Cµ v 2T
Dv 2 v2 ∂  ∂v 2  C1  v 2 2  P
= kf − ε + (ν + ν t )  L ∇ f − f =  −  − C2 k
2 2
Dt k ∂x j  ∂x j  T  k 3  k
∂  νt ∂ k  Dε ε ε 2 ∂  νt  ∂ε 
Dk
= P −ε + = − +  ν + 
ν +  
C P C
  ε 1 ε 2
k ∂x j  σε ∂ x j 
, Dt ∂x j  σ k  ∂x 
j 
Dt k
C1 = 0.4 C2 = 0.3 CL = 0.3 Cη = 70 Cε 2 = 1.9 σ ε = 1.0

, 32  k  ν 3  
14
 k ν 
4
Cε 1 = 1.3 + 0.25 1 + ( C L d 2 L )  L = max CL
2
T = max  , 6 
, Cη   
 
 ε ε  
ε  ε  
, 1− n
v2f-kω model: ν T = Cµ k v n 2
ω
,
Dv 2 ε ∂  ∂v2  1  v2 2  1 v ′2 P
= kf − 6v 2 + (ν +ν t )  L ∇ − f=
2 2
( C1− 1)  − − 5 − C2 k
Dt k ∂x j  ∂x j  T  k 3 T k k
1− n
Dk ∂U i ∂  νt ∂ k  Dω ω ∂U i  2 ∂  ν ω∂ 
= τ ij − β *ω k + 2 v
ν +   = α τ ij − βω   + ν  + t  
Dt ∂x j ∂x j  σ k  ∂x j  Dt k ∂x j k  ∂ x 
  σ ∂
 x 
  j ω j
α = 0.45977 β * = 0.09 β = 3 40 σ k = 1.0 σ ω = 1.5

 k ν   k 3 2  ν 3  
14
T = max  , 6  L = max CL , Cη   


 ε ε   ε  ε  
19 May 2010
21
JetWings

Comments on four-equation models:
1. For two-equation models a major problem is that it is hard to
specify the proper conditions to be applied near walls.
2. Durbin suggested that the problem is that the Reynolds number is
low near a wall and that the impermeability condition (zero normal
, velocity) is far more important. That is the motivation for the equation
for the normal velocity fluctuation.
3. It was found that the model also required a damping function f,
hence the name v2f model.
4. They appear to give improved results at essentially the same cost
as the k-ε and k-ω models especially for separated flows. Hopefully the
v2f-kω model can have better numerical stability than v2f-kε model as
their counterparts behave in two-equation models.
19 May 2010
22
JetWings

Some of the most noteworthy types of applications for which models based
on the Boussinesq approximation fail are:
1. Flows with sudden changes in mean strain rate
2. Flow over curved surfaces
3. Flow in ducts with secondary motions
4. Flow in rotating fluids
5. Three-dimensional flows
6. Flows with boundary-layer separation
In Reynolds-stress models, the equations for the Reynolds stress

tensor are modeled and solved along with the ε-equation:
∂τ ij ∂τ ij ∂U j ∂U i ∂ui′ ∂u′j  ui ′ ∂p′ u j′ ∂p ′ ∂  ∂τ ij 

+Uk = −τ ik − τ jk + 2ν + + + ν + u ′ ′ ′
i j k 
u u
∂t ∂xk ∂xk ∂xk ∂xk ∂xk  ρ ∂x j ρ ∂xi ∂ xk ∂ xk 
19 May 2010
23
JetWings

These are some versions of Reynolds stress models:
1. LRR rapid pressure-strain model
2. Lumley pressure-strain model
3. SSG pressure-strain model
4. Wilcox stress- model
Comments on Reynolds stress models:

1.Reynolds stress models require the solution of seven additional
PDEs and those equation are even harder to solve than the two-equation
models.
2. Although Reynolds stress models have greater potential to
represent turbulent flow more correctly, their success so far has been
moderate.
3. There is a lot of current research in this field, and new models
are often proposed. Which model is best for which kind of flow is not clear
due to the fact that in many attempts to answer this question numerical
errors were too large to allow clear conclusions to be reached.
19 May 2010
24
JetWings
Turbulent models (LES)
• Large scale motions are generally much more energetic than the small scale
ones.
• The size and strength of large scale motions make them to be the most
effective transporters of the conserved properties.
• LES treats the large eddies more exactly than the small ones may make
sense
• LES is 3D, time dependent and expensive but much less costly than DNS.
19 May 2010
25
JetWings
Turbulent models (LES, Scale-similarity model and

Dynamic model)
• Dynamic model:
1. filtered LES solutions can be filtered again using a filter
broader than the previous filter to obtain a very large scale
field.
2. An effective subgrid-scale field can be obtained by
subtraction of the two fields.
3. model parameter can then be computed.
Advantages: 1. model parameter computed at every spatial
grid point and every time step from the results of LES
2. Self-consistent subgrid-scale model
3. Automatically change the parameter near the wall and in
shear flows
Disadvantages: backscatter (eddy viscosity<0) may cause
instability.
19 May 2010
26
JetWings
Turbulent models (DES)

• Massively separated flows at high Re usually involve
both large and small scale vortical structures and very
thin turbulent boundary layer near the wall
• RANS approaches are efficient inside the boundary
layer but predict very excessive diffusion in the separated
regions
• LES is accurate in the separated regions but is unaffordable
for resolving thin near-wall turbulent boundary layers at
industrial Reynolds numbers
• Motivation for DES: combination of LES and RANS. RANS
inside attached boundary layer and LES in the separated
regions
19 May 2010
27
JetWings
DES Formulation
• Modification to RANS models was straightforward by
substituting the length scale d w , which is the distance to the
closest wall, with the new DES length scale, l% defined as:
l% = min(d w , CDES ∆) ∆ = max(δ x , δ y , δ z )
• where C is the DES constant, ∆ is the grid spacing and is
DES
based on the largest dimensions of the local grid cell, and δ x , δ y , δ z

are the grid spacing in x, y and z coordinates respectively
• Applying the above modification will result in S-A Based,
standard k-ε (or k-ω) based and Menter’s SST based DES
models, etc.
= ρβ *kω = ρ k 3 2 / lk −ω lk −ω = k 1 2 ( β *ω )
k
DRANS
32
D k
=ρ k l% = min(lk −ω , CDES ∆)
DES
l%
19 May 2010
28
JetWings
Resolved/Modeled/Total Reynolds stress

(DES)
TKE
Modeled
Resolved
Total
Modeled Reynolds stress: u 'v ' =−ν t  ∂u + ∂v 

19 May 2010  ∂y ∂x 
29
JetWings
Resolved/Modeled/Total Reynolds
stress (URANS)
Modeled Total
Resolved
19 May 2010
30
JetWings
Deep insight into RANS/DES (Point oscillation on free

surface with power spectral analysis)
EFD DES URANS
-5/3
19 May 2010 2HZ

31
JetWings
Direct Numerical Simulation (DNS)

• DNS is to solve the Navier-Stokes equation directly without
averaging or approximation other than numerical
discretizations whose errors can be estimated (V&V) and
controlled.
• The domain of DNS must be at least as large as the physical
domain or the largest turbulent eddy (scale L)
• The size of the grid must be no larger than a viscously
determined scale, Kolmogoroff scale, η
• The number of grid points in each direction must be at least L/
η
• The computational cost is proportional to
Re 3L/ 4 ≈ (0.01 Re )
34
• Provide detailed information on flow field

• Due to the computational cost, DNS is more likely to be a
research tool, not a design tool.
19 May 2010
32
JetWings
Examples (Diffuser)
• Mean velocity
predicted by V2f
agreed very well
with EFD data,
particular the
separation region
is captured.
• K-ε model fails to
predict the
separation caused
by adverse
pressure gradient.
19 May 2010
33
JetWings
Examples (Diffuser)
v2f
k-ε
• TKE predicted by V2f agreed better with EFD data

than k- ε model, particular the asymmetric
distribution.
• Right column is for the skin friction coefficient on 2x/H
the lower wall, from which the separation and
19reattachment
May 2010 point can be found.
34
JetWings
References
1. J. H. Ferziger, M. Peric, “Computational Methods for Fluid

Dynamics,” 3rd edition, Springer, 2002.
2. D. C. Wilcox, “Turbulence Modeling for CFD”, 1998.
3. S.B.Pope, “Turbulent Flows,” Cambridge, 2000
4. P. A. Durbin and B. A. Pettersson Reif, “Statistical Theory and
Modeling for Turbulent Flows,” John Wiley & Sons, LTD, 2001.
5. P.A. Davidson, “Turbulence: An Introduction for Scientists
and Engineers,” Oxford, 2004.
19 May 2010
35
JetWings

Lecture # 05 Turbulence Modelling

Uploaded by

Document Information

Copyright

Available Formats

Share this document

Share or Embed Document

Sharing Options

Did you find this document useful?

Is this content inappropriate?

Copyright:

Available Formats

Lecture # 05 Turbulence Modelling

Uploaded by

Copyright:

Available Formats

Lecture # 05 Turbulence Modelling

• Nonlinearity: Reynolds stresses from the nonlinear

•Dissipation: occurs at smallest scales

Approaches to predicting turbulent flows

Deeper insights on RANS/URANS/LES

RANS equations and unknowns

• Numbers of unknowns and equations

The Reynolds-Stress Equation

The closure problem of turbulence

Turbulent models (RANS)

Turbulent models (RANS)

• Dimensional analysis shows:ν T = Cµ qL , where q is a

Turbulent models (0-eqn RANS)

• Assume lmix = αδ ( x ) for free shear flow, then

─Not applicable to wall-bounded flows

Turbulent models (0-eqn RANS)

Three key modifications to mixing length model;

Turbulent models (0-eqn RANS)

Closure coefficients: κ = 0.40 α = 0.0168 A+ = 26 Ccp = 1.6 C Kleb = 0.3 Cwk = 1

Turbulent models (1-eqn RANS)

Turbulence Reynolds number :

Closure coefficients and auxiliary relations:

Turbulent models (1-eqn RANS)

Eddy viscosity equation:

Closure coefficients and auxiliary relations:

Turbulent models (1-eqn RANS)

Turbulent models (2-eqn RANS)

Turbulent models (2-eqn RANS)

Turbulent models (4-eqn RANS)

C1 = 0.4 C2 = 0.3 CL = 0.3 Cη = 70 Cε 2 = 1.9 σ ε = 1.0

α = 0.45977 β * = 0.09 β = 3 40 σ k = 1.0 σ ω = 1.5

T = max  , 6  L = max CL , Cη   

Turbulent models (4-eqn RANS)

Turbulent models (7-eqn RANS)

In Reynolds-stress models, the equations for the Reynolds stress

∂τ ij ∂τ ij ∂U j ∂U i ∂ui′ ∂u′j  ui ′ ∂p′ u j′ ∂p ′ ∂  ∂τ ij 

Turbulent models (7-eqn RANS)

Comments on Reynolds stress models:

Turbulent models (LES)

Turbulent models (LES, Scale-similarity model and

Turbulent models (DES)

based on the largest dimensions of the local grid cell, and δ x , δ y , δ z

Resolved/Modeled/Total Reynolds stress

Modeled Reynolds stress: u 'v ' =−ν t  ∂u + ∂v 

Deep insight into RANS/DES (Point oscillation on free

19 May 2010 2HZ

Direct Numerical Simulation (DNS)

• Provide detailed information on flow field

• TKE predicted by V2f agreed better with EFD data

1. J. H. Ferziger, M. Peric, “Computational Methods for Fluid

You might also like