Modeling Turubulence

Chapter 10.
Modeling Turbulence
This chapter provides details about the turbulence models available in
FLUENT.
Information is presented in the following sections:
Section 10.1: Introduction
Section 10.2: Choosing a Turbulence Model
Section 10.3: The Spalart-Allmaras Model
Section 10.4: The Standard, RNG, and Realizable k- Models
Section 10.5: The Standard and Shear-Stress Transport (SST) k-
Models
Section 10.6: The Reynolds Stress Model (RSM)
Section 10.7: The Large Eddy Simulation (LES) Model
Section 10.8: Near-Wall Treatments for Wall-Bounded Turbulent
Flows
Section 10.9: Grid Considerations for Turbulent Flow Simulations
Section 10.10: Problem Setup for Turbulent Flows
Section 10.11: Solution Strategies for Turbulent Flow Simulations
Section 10.12: Postprocessing for Turbulent Flows
c Fluent Inc. November 28, 2001 10-1
Modeling Turbulence
10.1 Introduction
Turbulent ows are characterized by uctuating velocity elds. These
uctuations mix transported quantities such as momentum, energy, and
species concentration, and cause the transported quantities to uctuate
as well. Since these uctuations can be of small scale and high frequency,
they are too computationally expensive to simulate directly in practical
engineering calculations. Instead, the instantaneous (exact) governing
equations can be time-averaged, ensemble-averaged, or otherwise manip-
ulated to remove the small scales, resulting in a modied set of equations
that are computationally less expensive to solve. However, the modied
equations contain additional unknown variables, and turbulence models
are needed to determine these variables in terms of known quantities.
FLUENT provides the following choices of turbulence models:
Spalart-Allmaras model
k- models
Standard k- model
Renormalization-group (RNG) k- model
Realizable k- model
k- models
Standard k- model
Shear-stress transport (SST) k- model
Reynolds stress model (RSM)
Large eddy simulation (LES) model
10-2 c Fluent Inc. November 28, 2001
10.2 Choosing a Turbulence Model
It is an unfortunate fact that no single turbulence model is universally
accepted as being superior for all classes of problems. The choice of
turbulence model will depend on considerations such as the physics en-
compassed in the ow, the established practice for a specic class of
problem, the level of accuracy required, the available computational re-
sources, and the amount of time available for the simulation. To make
the most appropriate choice of model for your application, you need to
understand the capabilities and limitations of the various options.
The purpose of this section is to give an overview of issues related to
the turbulence models provided in FLUENT. The computational eort
and cost in terms of CPU time and memory of the individual models is
discussed. While it is impossible to state categorically which model is
best for a specic application, general guidelines are presented to help
you choose the appropriate turbulence model for the ow you want to
model.
10.2.1 Rey nolds-Averaged Approach vs. LES
A complete time-dependent solution of the exact Navier-Stokes equa-
tions for high-Reynolds-number turbulent ows in complex geometries
is unlikely to be attainable for some time to come. Two alternative
methods can be employed to transform the Navier-Stokes equations in
such a way that the small-scale turbulent uctuations do not have to
be directly simulated: Reynolds averaging and ltering. Both methods
introduce additional terms in the governing equations that need to be
modeled in order to achieve closure. (Closure implies that there are a
sucient number of equations for all the unknowns.)
The Reynolds-averaged Navier-Stokes (RANS) equations represent trans-
port equations for the mean ow quantities only, with all the scales of
the turbulence being modeled. The approach of permitting a solution
for the mean ow variables greatly reduces the computational eort. If
the mean ow is steady, the governing equations will not contain time
derivatives and a steady-state solution can be obtained economically. A
computational advantage is seen even in transient situations, since the
time step will be determined by the global unsteadiness in the mean
Modeling Turbulence
ow rather than by the turbulence. The Reynolds-averaged approach is
generally adopted for practical engineering calculations, and uses models
such as Spalart-Allmaras, k- and its variants, k- and its variants, and
the RSM.
LES provides an alternative approach in which the large eddies are com-
puted in a time-dependent simulation that uses a set of ltered equa-
tions. Filtering is essentially a manipulation of the exact Navier-Stokes
equations to remove only the eddies that are smaller than the size of the
lter, which is usually taken as the mesh size. Like Reynolds averaging,
the ltering process creates additional unknown terms that must be mod-
eled in order to achieve closure. Statistics of the mean ow quantities,
which are generally of most engineering interest, are gathered during the
time-dependent simulation. The attraction of LES is that, by modeling
less of the turbulence (and solving more), the error induced by the tur-
bulence model will be reduced. One might also argue that it ought to be
easier to nd a universal model for the small scales, which tend to be
more isotropic and less aected by the macroscopic ow features than
the large eddies.
It should, however, be stressed that the application of LES to industrial
uid simulations is in its infancy. As highlighted in a recent review publi-
cation [72], typical applications to date have been for simple geometries.
This is mainly because of the large computer resources required to re-
solve the energy-containing turbulent eddies. Most successful LES has
been done using high-order spatial discretization, with great care being
taken to resolve all scales larger than the inertial subrange. The degra-
dation of accuracy in the mean ow quantities with poorly resolved LES
is not well documented. In addition, the use of wall functions with LES
is an approximation that requires further validation.
As a general guideline, therefore, it is recommended that the conven-
tional turbulence models employing the Reynolds-averaged approach be
used for practical calculations. The LES approach, described further in
Section 10.7, has been made available for you to try if you have the com-
putational resources and are willing to invest the eort. The rest of this
section will deal with the choice of models using the Reynolds-averaged
approach.
10.2.2 Rey nolds (Ensemble) Averaging
In Reynolds averaging, the solution variables in the instantaneous (ex-
act) Navier-Stokes equations are decomposed into the mean (ensemble-
averaged or time-averaged) and uctuating components. For the velocity
components:
u
i
= u
i
+ u
i
(10.2-1)
where u
i
and u
i
are the mean and uctuating velocity components (i =
1, 2, 3).
Likewise, for pressure and other scalar quantities:
=

+
(10.2-2)
where denotes a scalar such as pressure, energy, or species concentra-
tion.
Substituting expressions of this form for the ow variables into the in-
stantaneous continuity and momentum equations and taking a time (or
ensemble) average (and dropping the overbar on the mean velocity, u)
yields the ensemble-averaged momentum equations. They can be written
in Cartesian tensor form as:
t
+

x
i
(u
i
) = 0 (10.2-3)
t
(u
i
) +

x
j
(u
i
u
j
) =
p
x
i
+

x
j
_
_
u
i
x
j
+
u
j
x
i
2
3
ij
u
l
x
l
__
+

x
j
(u
i
u
j
) (10.2-4)
Equations 10.2-3 and 10.2-4 are called Reynolds-averaged Navier-Stokes
(RANS) equations. They have the same general form as the instan-
taneous Navier-Stokes equations, with the velocities and other solution
variables now representing ensemble-averaged (or time-averaged) values.
Modeling Turbulence
Additional terms now appear that represent the eects of turbulence.
These Reynolds stresses, u
i
u
j
, must be modeled in order to close
Equation 10.2-4.
For variable-density ows, Equations 10.2-3 and 10.2-4 can be interpreted
as Favre-averaged Navier-Stokes equations [91], with the velocities rep-
resenting mass-averaged values. As such, Equations 10.2-3 and 10.2-4
can be applied to density-varying ows.
10.2.3 Boussinesq Approach vs. Rey nolds Stress Transport
Models
The Reynolds-averaged approach to turbulence modeling requires that
the Reynolds stresses in Equation 10.2-4 be appropriately modeled. A
common method employs the Boussinesq hypothesis [91] to relate the
Reynolds stresses to the mean velocity gradients:
u
i
u
j
=
t
_
u
i
x
j
+
u
j
x
i
_
2
3
_
k +
t
u
i
x
i
_
ij
(10.2-5)
The Boussinesq hypothesis is used in the Spalart-Allmaras model, the
k- models, and the k- models. The advantage of this approach is the
relatively low computational cost associated with the computation of the
turbulent viscosity,
t
. In the case of the Spalart-Allmaras model, only
one additional transport equation (representing turbulent viscosity) is
solved. In the case of the k- and k- models, two additional transport
equations (for the turbulence kinetic energy, k, and either the turbulence
dissipation rate, , or the specic dissipation rate, ) are solved, and
t
is computed as a function of k and . The disadvantage of the Boussi-
nesq hypothesis as presented is that it assumes
t
is an isotropic scalar
quantity, which is not strictly true.
The alternative approach, embodied in the RSM, is to solve transport
equations for each of the terms in the Reynolds stress tensor. An addi-
tional scale-determining equation (normally for ) is also required. This
means that ve additional transport equations are required in 2D ows
and seven additional transport equations must be solved in 3D.
In many cases, models based on the Boussinesq hypothesis perform very
well, and the additional computational expense of the Reynolds stress
model is not justied. However, the RSM is clearly superior for situations
in which the anisotropy of turbulence has a dominant eect on the mean
ow. Such cases include highly swirling ows and stress-driven secondary
ows.
10.2.4 The Spalart-Allmaras Model
The Spalart-Allmaras model is a relatively simple one-equation model
that solves a modeled transport equation for the kinematic eddy (tur-
bulent) viscosity. This embodies a relatively new class of one-equation
models in which it is not necessary to calculate a length scale related to
the local shear layer thickness. The Spalart-Allmaras model was designed
specically for aerospace applications involving wall-bounded ows and
has been shown to give good results for boundary layers subjected to ad-
verse pressure gradients. It is also gaining popularity for turbomachinery
applications.
In its original form, the Spalart-Allmaras model is eectively a low-
Reynolds-number model, requiring the viscous-aected region of the
boundary layer to be properly resolved. In FLUENT, however, the Spalart-
Allmaras model has been implemented to use wall functions when the
mesh resolution is not suciently ne. This might make it the best
choice for relatively crude simulations on coarse meshes where accurate
turbulent ow computations are not critical. Furthermore, the near-wall
gradients of the transported variable in the model are much smaller than
the gradients of the transported variables in the k- or k- models. This
might make the model less sensitive to numerical error when non-layered
meshes are used near walls. See Section 5.1.2 for further discussion of
numerical error.
On a cautionary note, however, the Spalart-Allmaras model is still rel-
atively new, and no claim is made regarding its suitability to all types
of complex engineering ows. For instance, it cannot be relied on to
predict the decay of homogeneous, isotropic turbulence. Furthermore,
one-equation models are often criticized for their inability to rapidly ac-
commodate changes in length scale, such as might be necessary when
Modeling Turbulence
the ow changes abruptly from a wall-bounded to a free shear ow.
10.2.5 The Standard k- Model
The simplest complete models of turbulence are two-equation models
in which the solution of two separate transport equations allows the tur-
bulent velocity and length scales to be independently determined. The
standard k- model in FLUENT falls within this class of turbulence model
and has become the workhorse of practical engineering ow calculations
in the time since it was proposed by Launder and Spalding [128]. Ro-
bustness, economy, and reasonable accuracy for a wide range of turbulent
ows explain its popularity in industrial ow and heat transfer simula-
tions. It is a semi-empirical model, and the derivation of the model
equations relies on phenomenological considerations and empiricism.
As the strengths and weaknesses of the standard k- model have become
known, improvements have been made to the model to improve its per-
formance. Two of these variants are available in FLUENT: the RNG k-
model [272] and the realizable k- model [209].
10.2.6 The RNG k- Model
The RNG k- model was derived using a rigorous statistical technique
(called renormalization group theory). It is similar in form to the stan-
dard k- model, but includes the following renements:
The RNG model has an additional term in its equation that
signicantly improves the accuracy for rapidly strained ows.
The eect of swirl on turbulence is included in the RNG model,
enhancing accuracy for swirling ows.
The RNG theory provides an analytical formula for turbulent
Prandtl numbers, while the standard k- model uses user-specied,
constant values.
While the standard k- model is a high-Reynolds-number model,
the RNG theory provides an analytically-derived dierential for-
mula for eective viscosity that accounts for low-Reynolds-number
eects. Eective use of this feature does, however, depend on an
appropriate treatment of the near-wall region.
These features make the RNG k- model more accurate and reliable for
a wider class of ows than the standard k- model.
10.2.7 The Realizable k- Model
The realizable k- model is a relatively recent development and diers
from the standard k- model in two important ways:
The realizable k- model contains a new formulation for the tur-
bulent viscosity.
A new transport equation for the dissipation rate, , has been de-
rived from an exact equation for the transport of the mean-square
vorticity uctuation.
The term realizable means that the model satises certain mathemat-
ical constraints on the Reynolds stresses, consistent with the physics of
turbulent ows. Neither the standard k- model nor the RNG k- model
is realizable.
An immediate benet of the realizable k- model is that it more accu-
rately predicts the spreading rate of both planar and round jets. It is
also likely to provide superior performance for ows involving rotation,
boundary layers under strong adverse pressure gradients, separation, and
recirculation.
Both the realizable and RNG k- models have shown substantial im-
provements over the standard k- model where the ow features include
strong streamline curvature, vortices, and rotation. Since the model is
still relatively new, it is not clear in exactly which instances the real-
izable k- model consistently outperforms the RNG model. However,
initial studies have shown that the realizable model provides the best
performance of all the k- model versions for several validations of sep-
arated ows and ows with complex secondary ow features.
Modeling Turbulence
One limitation of the realizable k- model is that it produces non-physical
turbulent viscosities in situations when the computational domain con-
tains both rotating and stationary uid zones (e.g., multiple reference
frames, rotating sliding meshes). This is due to the fact that the realiz-
able k- model includes the eects of mean rotation in the denition of
the turbulent viscosity (see Equations 10.4-1710.4-19). This extra ro-
tation eect has been tested on single rotating reference frame systems
and showed superior behavior over the standard k- model. However, due
to the nature of this modication, its application to multiple reference
frame systems should be taken with some caution.
The standard k- model in FLUENT is based on the Wilcox k- model [267],
which incorporates modications for low-Reynolds-number eects, com-
pressibility, and shear ow spreading. The Wilcox model predicts free
shear ow spreading rates that are in close agreement with measure-
ments for far wakes, mixing layers, and plane, round, and radial jets,
and is thus applicable to wall-bounded ows and free shear ows. A
variation of the standard k- model called the SST k- model is also
available in FLUENT, and is described in Section 10.2.9.
10.2.9 The Shear-Stress Transport (SST) k- Model
The shear-stress transport (SST) k- model was developed by Menter [153]
to eectively blend the robust and accurate formulation of the k- model
in the near-wall region with the free-stream independence of the k-
model in the far eld. To achieve this, the k- model is converted into
a k- formulation. The SST k- model is similar to the standard k-
model, but includes the following renements:
The standard k- model and the transformed k- model are both
multiplied by a blending function and both models are added to-
gether. The blending function is designed to be one in the near-
wall region, which activates the standard k- model, and zero away
from the surface, which activates the transformed k- model.
The SST model incorporates a damped cross-diusion derivative
term in the equation.
The denition of the turbulent viscosity is modied to account for
the transport of the turbulent shear stress.
The modeling constants are dierent.
These features make the SST k- model more accurate and reliable for
a wider class of ows (e.g., adverse pressure gradient ows, airfoils, tran-
sonic shock waves) than the standard k- model.
10.2.10 The Rey nolds Stress Model (RSM)
The Reynolds stress model (RSM) is the most elaborate turbulence
model that FLUENT provides. Abandoning the isotropic eddy-viscosity
hypothesis, the RSM closes the Reynolds-averaged Navier-Stokes equa-
tions by solving transport equations for the Reynolds stresses, together
with an equation for the dissipation rate. This means that four addi-
tional transport equations are required in 2D ows and seven additional
transport equations must be solved in 3D.
Since the RSM accounts for the eects of streamline curvature, swirl,
rotation, and rapid changes in strain rate in a more rigorous manner
than one-equation and two-equation models, it has greater potential to
give accurate predictions for complex ows. However, the delity of
RSM predictions is still limited by the closure assumptions employed to
model various terms in the exact transport equations for the Reynolds
stresses. The modeling of the pressure-strain and dissipation-rate terms
is particularly challenging, and often considered to be responsible for
compromising the accuracy of RSM predictions.
The RSM might not always yield results that are clearly superior to the
simpler models in all classes of ows to warrant the additional compu-
tational expense. However, use of the RSM is a must when the ow
features of interest are the result of anisotropy in the Reynolds stresses.
Among the examples are cyclone ows, highly swirling ows in combus-
tors, rotating ow passages, and the stress-induced secondary ows in
ducts.
Modeling Turbulence
10.2.11 Computational Eort: CPU Time and Solution
Behavior
In terms of computation, the Spalart-Allmaras model is the least expen-
sive turbulence model of the options provided in FLUENT, since only one
turbulence transport equation is solved.
The standard k- model clearly requires more computational eort than
the Spalart-Allmaras model since an additional transport equation is
solved. The realizable k- model requires only slightly more computa-
tional eort than the standard k- model. However, due to the extra
terms and functions in the governing equations and a greater degree of
non-linearity, computations with the RNG k- model tend to take 10
15% more CPU time than with the standard k- model. Like the k-
models, the k- models are also two-equation models, and thus require
about the same computational eort.
Compared with the k- and k- models, the RSM requires additional
memory and CPU time due to the increased number of the transport
equations for Reynolds stresses. However, ecient programming in FLU-
ENT has reduced the CPU time per iteration signicantly. On average,
the RSM in FLUENT requires 5060% more CPU time per iteration com-
pared to the k- and k- models. Furthermore, 1520% more memory is
needed.
Aside from the time per iteration, the choice of turbulence model can
aect the ability of FLUENT to obtain a converged solution. For example,
the standard k- model is known to be slightly over-diusive in certain
situations, while the RNG k- model is designed such that the turbulent
viscosity is reduced in response to high rates of strain. Since diusion
has a stabilizing eect on the numerics, the RNG model is more likely
to be susceptible to instability in steady-state solutions. However, this
should not necessarily be seen as a disadvantage of the RNG model,
since these characteristics make it more responsive to important physical
instabilities such as time-dependent turbulent vortex shedding.
Similarly, the RSM may take more iterations to converge than the k-
and k- models due to the strong coupling between the Reynolds stresses
and the mean ow.
10.3 The Spalart-Allmaras Model
In turbulence models that employ the Boussinesq approach, the central
issue is how the eddy viscosity is computed. The model proposed by
Spalart and Allmaras [226] solves a transport equation for a quantity
that is a modied form of the turbulent kinematic viscosity.
10.3.1 Transport Equation for the Spalart-Allmaras Model
The transported variable in the Spalart-Allmaras model, , is identical
to the turbulent kinematic viscosity except in the near-wall (viscous-
aected) region. The transport equation for is
t
( ) +

x
i
( u
i
) =
G
+
1

_
_
x
j
_
( + )

x
j
_
+ C
b2
_

x
j
_
2
_
_
Y
+S

(10.3-1)
where G
is the production of turbulent viscosity and Y
is the destruc-
tion of turbulent viscosity that occurs in the near-wall region due to
wall blocking and viscous damping.

and C
b2
are constants and
is the molecular kinematic viscosity. S

is a user-dened source term.
Note that since the turbulence kinetic energy k is not calculated in the
Spalart-Allmaras model, it is not taken into account when estimating
the Reynolds stresses in Equation 10.2-5.
10.3.2 Modeling the Turbulent Viscosity
The turbulent viscosity,
t
, is computed from
t
= f
v1
(10.3-2)
where the viscous damping function, f
v1
, is given by
f
v1
=

3
3
+C
3
v1
(10.3-3)
and
Modeling Turbulence

(10.3-4)
10.3.3 Modeling the Turbulent Production
The production term, G
, is modeled as
G
= C
b1
S (10.3-5)
where
S S +

2
d
2
f
v2
(10.3-6)
and
f
v2
= 1

1 + f
v1
(10.3-7)
C
b1
and are constants, d is the distance from the wall, and S is a
scalar measure of the deformation tensor. By default in FLUENT, as in
the original model proposed by Spalart and Allmaras, S is based on the
magnitude of the vorticity:
S
_
2
ij
ij
(10.3-8)
where
ij
is the mean rate-of-rotation tensor and is dened by
ij
=
1
2
_
u
i
x
j
u
j
x
i
_
(10.3-9)
The justication for the default expression for S is that, for the wall-
bounded ows that were of most interest when the model was formu-
lated, turbulence is found only where vorticity is generated near walls.
However, it has since been acknowledged that one should also take into
account the eect of mean strain on the turbulence production, and a
modication to the model has been proposed [46] and incorporated into
FLUENT.
This modication combines measures of both rotation and strain tensors
in the denition of S:
S |
ij
| + C
prod
min (0, |S
ij
| |
ij
|) (10.3-10)
where
C
prod
= 2.0, |
ij
|
_
2
ij
ij
, |S
ij
|
_
2S
ij
S
ij
with the mean strain rate, S
ij
, dened as
S
ij
=
1
2
_
u
j
x
i
+
u
i
x
j
_
(10.3-11)
Including both the rotation and strain tensors reduces the production
of eddy viscosity and consequently reduces the eddy viscosity itself in
regions where the measure of vorticity exceeds that of strain rate. One
such example can be found in vortical ows, i.e., ow near the core of a
vortex subjected to a pure rotation where turbulence is known to be sup-
pressed. Including both the rotation and strain tensors more correctly
accounts for the eects of rotation on turbulence. The default option (in-
cluding the rotation tensor only) tends to overpredict the production of
eddy viscosity and hence overpredicts the eddy viscosity itself in certain
circumstances.
You can select the modied form for calculating production in the Viscous
Model panel.
10.3.4 Modeling the Turbulent Destruction
The destruction term is modeled as
Y
= C
w1
f
w
_

d
_
2
(10.3-12)
Modeling Turbulence
where
f
w
= g
_
1 + C
6
w3
g
6
+ C
6
w3
_
1/6
(10.3-13)
g = r + C
w2
_
r
6
r
_
(10.3-14)
r

S
2
d
2
(10.3-15)
C
w1
, C
w2
, and C
w3
are constants, and

S is given by Equation 10.3-6.
Note that the modication described above to include the eects of mean
strain on S will also aect the value of

S used to compute r.
10.3.5 Model Constants
The model constants C
b1
, C
b2
,

, C
v1
, C
w1
, C
w2
, C
w3
, and have the fol-
lowing default values [226]:
C
b1
= 0.1335, C
b2
= 0.622,

=
2
3
, C
v1
= 7.1
C
w1
=
C
b1
2
+
(1 + C
b2
)

, C
w2
= 0.3, C
w3
= 2.0, = 0.4187
10.3.6 Wall Boundary Conditions
At walls, the modied turbulent kinematic viscosity, , is set to zero.
When the mesh is ne enough to resolve the laminar sublayer, the wall
shear stress is obtained from the laminar stress-strain relationship:
u
u
=
u
(10.3-16)
If the mesh is too coarse to resolve the laminar sublayer, it is assumed
that the centroid of the wall-adjacent cell falls within the logarithmic
region of the boundary layer, and the law-of-the-wall is employed:
u
u
=
1
ln E
_
u
_
(10.3-17)
where u is the velocity parallel to the wall, u
is the shear velocity, y is

the distance from the wall, is the von Karm an constant (0.4187), and
E = 9.793.
10.3.7 Convective Heat and Mass Transfer Modeling
In FLUENT, turbulent heat transport is modeled using the concept of
Reynolds analogy to turbulent momentum transfer. The modeled
energy equation is thus given by the following:
t
(E) +

x
i
[u
i
(E + p)] =

x
j
_
_
k +
c
p
t
Pr
t
_
T
x
j
+ u
i
(
ij
)
e
_
+ S
h
(10.3-18)
where k, in this case, is the thermal conductivity, E is the total energy,
and (
ij
)
e
is the deviatoric stress tensor, dened as
(
ij
)
e
=
e
_
u
j
x
i
+
u
i
x
j
_
2
3
e
u
i
x
i
ij
The term involving (
ij
)
e
represents the viscous heating, and is always
computed in the coupled solvers. It is not computed by default in the
segregated solver, but it can be enabled in the Viscous Model panel. The
default value of the turbulent Prandtl number is 0.85. You can change
the value of Pr
t
in the Viscous Model panel.
Turbulent mass transfer is treated similarly, with a default turbulent
Schmidt number of 0.7. This default value can be changed in the Viscous
Model panel.
Modeling Turbulence
Wall boundary conditions for scalar transport are handled analogously
to momentum, using the appropriate law-of-the-wall.
10.4 The Standard, RNG, and Realizable k- Models
This section presents the standard, RNG, and realizable k- models. All
three models have similar forms, with transport equations for k and .
The major dierences in the models are as follows:
the method of calculating turbulent viscosity
the turbulent Prandtl numbers governing the turbulent diusion
of k and
the generation and destruction terms in the equation
The transport equations, methods of calculating turbulent viscosity, and
model constants are presented separately for each model. The features
that are essentially common to all models follow, including turbulent
production, generation due to buoyancy, accounting for the eects of
compressibility, and modeling heat and mass transfer.
The standard k- model [128] is a semi-empirical model based on model
transport equations for the turbulence kinetic energy (k) and its dissi-
pation rate (). The model transport equation for k is derived from the
exact equation, while the model transport equation for was obtained
using physical reasoning and bears little resemblance to its mathemati-
cally exact counterpart.
In the derivation of the k- model, it was assumed that the ow is fully
turbulent, and the eects of molecular viscosity are negligible. The stan-
dard k- model is therefore valid only for fully turbulent ows.
Transport Equations for the Standard k- Model
The turbulence kinetic energy, k, and its rate of dissipation, , are ob-
tained from the following transport equations:
t
(k) +

x
i
(ku
i
) =

x
j
_
_
+

t
k
_
k
x
j
_
+ G
k
+G
b
Y
M
+ S
k
(10.4-1)
and
t
() +

x
i
(u
i
) =

x
j
_
_
+

t
_

x
j
_
+ C
1
k
(G
k
+ C
3
G
b
)
C
2
2
k
+ S
(10.4-2)
In these equations, G
k
represents the generation of turbulence kinetic
energy due to the mean velocity gradients, calculated as described in
Section 10.4.4. G
b
is the generation of turbulence kinetic energy due
to buoyancy, calculated as described in Section 10.4.5. Y
M
represents
the contribution of the uctuating dilatation in compressible turbulence
to the overall dissipation rate, calculated as described in Section 10.4.6.
C
1
, C
2
, and C
3
are constants.
k
and
are the turbulent Prandtl

numbers for k and , respectively. S
k
and S
are user-dened source

terms.
Modeling the Turbulent Viscosity
The turbulent (or eddy) viscosity,
t
, is computed by combining k and
as follows:
t
= C
k
2
(10.4-3)
where C
is a constant.
Model Constants
1
, C
2
, C
,
k
, and
have the following default

values [128]:
C
1
= 1.44, C
2
= 1.92, C
= 0.09,
k
= 1.0,
= 1.3
Modeling Turbulence
These default values have been determined from experiments with air
and water for fundamental turbulent shear ows including homogeneous
shear ows and decaying isotropic grid turbulence. They have been found
to work fairly well for a wide range of wall-bounded and free shear ows.
Although the default values of the model constants are the standard ones
most widely accepted, you can change them (if needed) in the Viscous
Model panel.
10.4.2 The RNG k- Model
The RNG-based k- turbulence model is derived from the instantaneous
Navier-Stokes equations, using a mathematical technique called renor-
malization group (RNG) methods. The analytical derivation results in
a model with constants dierent from those in the standard k- model,
and additional terms and functions in the transport equations for k and
. A more comprehensive description of RNG theory and its application
to turbulence can be found in [36].
Transport Equations for the RNG k- Model
The RNG k- model has a similar form to the standard k- model:
t
(k) +

x
i
(ku
i
) =

x
j
_
e
k
x
j
_
+ G
k
+ G
b
Y
M
+S
k
(10.4-4)
and
t
() +

x
i
(u
i
) =

x
j
_
x
j
_
+
C
1
k
(G
k
+ C
3
G
b
) C
2
2
k
R
+ S
(10.4-5)
k
Section 10.4.4. G
b
is the generation of turbulence kinetic energy due
to buoyancy, calculated as described in Section 10.4.5. Y
M
represents
the contribution of the uctuating dilatation in compressible turbulence
to the overall dissipation rate, calculated as described in Section 10.4.6.
The quantities
k
and
are the inverse eective Prandtl numbers for

k and , respectively. S
k
and S
are user-dened source terms.

Modeling the Eective Viscosity
The scale elimination procedure in RNG theory results in a dierential
equation for turbulent viscosity:
d
_

2
k
_
= 1.72

3
1 + C
d (10.4-6)
where
=
e
/
C
100
Equation 10.4-6 is integrated to obtain an accurate description of how
the eective turbulent transport varies with the eective Reynolds num-
ber (or eddy scale), allowing the model to better handle low-Reynolds-
number and near-wall ows.
In the high-Reynolds-number limit, Equation 10.4-6 gives
t
= C
k
2
(10.4-7)
with C
= 0.0845, derived using RNG theory. It is interesting to note

that this value of C
is very close to the empirically-determined value of

0.09 used in the standard k- model.
In FLUENT, by default, the eective viscosity is computed using the
high-Reynolds-number form in Equation 10.4-7. However, there is an
option available that allows you to use the dierential relation given in
Equation 10.4-6 when you need to include low-Reynolds-number eects.
Modeling Turbulence
RNG Swirl Modication
Turbulence, in general, is aected by rotation or swirl in the mean ow.
The RNG model in FLUENT provides an option to account for the eects
of swirl or rotation by modifying the turbulent viscosity appropriately.
The modication takes the following functional form:
t
=
t0
f
_
s
, ,
k
_
(10.4-8)
where
t0
is the value of turbulent viscosity calculated without the swirl
modication using either Equation 10.4-6 or Equation 10.4-7. is a
characteristic swirl number evaluated within FLUENT, and
s
is a swirl
constant that assumes dierent values depending on whether the ow is
swirl-dominated or only mildly swirling. This swirl modication always
takes eect for axisymmetric, swirling ows and three-dimensional ows
when the RNG model is selected. For mildly swirling ows (the default
in FLUENT),
s
is set to 0.05 and cannot be modied. For strongly
swirling ows, however, a higher value of
s
can be used.
Calculating the Inverse Eective Prandtl Numbers
The inverse eective Prandtl numbers,
k
and
, are computed using

the following formula derived analytically by the RNG theory:
1.3929
0
1.3929
0.6321
+ 2.3929
0
+ 2.3929
0.3679
=

mol
e
(10.4-9)
where
0
= 1.0. In the high-Reynolds-number limit (
mol
/
e
1),
k
=
1.393.
The R
Term in the Equation

The main dierence between the RNG and standard k- models lies in
the additional term in the equation given by
R
=
C
3
(1 /
0
)
1 +
3
2
k
(10.4-10)
where Sk/,
0
= 4.38, = 0.012.
The eects of this term in the RNG equation can be seen more clearly
by rearranging Equation 10.4-5. Using Equation 10.4-10, the third and
fourth terms on the right-hand side of Equation 10.4-5 can be merged,
and the resulting equation can be rewritten as
t
() +

x
i
(u
i
) =

x
j
_
x
j
_
+ C
1
k
(G
k
+ C
3
G
b
) C
2
k
(10.4-11)
where C
2
is given by
C
2
C
2
+
C
3
(1 /
0
)
1 +
3
(10.4-12)
In regions where <
0
, the R term makes a positive contribution, and
C
2
becomes larger than C
2
. In the logarithmic layer, for instance, it can
be shown that 3.0, giving C
2
2.0, which is close in magnitude to
the value of C
2
in the standard k- model (1.92). As a result, for weakly
to moderately strained ows, the RNG model tends to give results largely
comparable to the standard k- model.
In regions of large strain rate ( >
0
), however, the R term makes a neg-
ative contribution, making the value of C
2
less than C
2
. In comparison
with the standard k- model, the smaller destruction of augments ,
reducing k and, eventually, the eective viscosity. As a result, in rapidly
strained ows, the RNG model yields a lower turbulent viscosity than
the standard k- model.
Thus, the RNG model is more responsive to the eects of rapid strain
and streamline curvature than the standard k- model, which explains
the superior performance of the RNG model for certain classes of ows.
Model Constants
1
and C
2
in Equation 10.4-5 have values de-
rived analytically by the RNG theory. These values, used by default in
FLUENT, are
Modeling Turbulence
C
1
= 1.42, C
2
= 1.68
10.4.3 The Realizable k- Model
In addition to the standard and RNG-based k- models described in
Sections 10.4.1 and 10.4.2, FLUENT also provides the so-called realizable
k- model [209]. The term realizable means that the model satises
certain mathematical constraints on the normal stresses, consistent with
the physics of turbulent ows. To understand this, consider combining
the Boussinesq relationship (Equation 10.2-5) and the eddy viscosity
denition (Equation 10.4-3) to obtain the following expression for the
normal Reynolds stress in an incompressible strained mean ow:
u
2
=
2
3
k 2
t
U
x
(10.4-13)
Using Equation 10.4-3 for
t

t
/, one obtains the result that the
normal stress, u
2
, which by denition is a positive quantity, becomes
negative, i.e., non-realizable, when the strain is large enough to satisfy
k
U
x
>
1
3C
3.7 (10.4-14)
Similarly, it can also be shown that the Schwarz inequality for shear
stresses (u
2
u
2
u
2
; no summation over and ) can be violated

when the mean strain rate is large. The most straightforward way to
ensure the realizability (positivity of normal stresses and Schwarz in-
equality for shear stresses) is to make C
variable by sensitizing it to the

mean ow (mean deformation) and the turbulence (k, ). The notion
of variable C
is suggested by many modelers including Reynolds [191],

and is well substantiated by experimental evidence. For example, C
is
found to be around 0.09 in the inertial sublayer of equilibrium boundary
layers, and 0.05 in a strong homogeneous shear ow.
Another weakness of the standard k- model or other traditional k-
models lies with the modeled equation for the dissipation rate (). The
well-known round-jet anomaly (named based on the nding that the
spreading rate in planar jets is predicted reasonably well, but predic-
tion of the spreading rate for axisymmetric jets is unexpectedly poor) is
considered to be mainly due to the modeled dissipation equation.
The realizable k- model proposed by Shih et al. [209] was intended
to address these deciencies of traditional k- models by adopting the
following:
a new eddy-viscosity formula involving a variable C
originally
proposed by Reynolds [191]
a new model equation for dissipation () based on the dynamic
equation of the mean-square vorticity uctuation
Transport Equations for the Realizable k- Model
The modeled transport equations for k and in the realizable k- model
are
t
(k) +

x
i
(ku
j
) =

x
i
_
_
+

t
k
_
k
x
j
_
+ G
k
+G
b
Y
M
+ S
k
(10.4-15)
and
t
() +

x
j
(u
j
) =

x
j
_
_
+

t
_

x
j
_
+ C
1
S
C
2
2
k +
+ C
1
k
C
3
G
b
+ S
(10.4-16)
where
C
1
= max
_
0.43,

+ 5
_
and
Modeling Turbulence
= S
k
k
Section 10.4.4. G
b
is the generation of turbulence kinetic energy due to
buoyancy, calculated as described in Section 10.4.5. Y
M
represents the
contribution of the uctuating dilatation in compressible turbulence to
the overall dissipation rate, calculated as described in Section 10.4.6. C
2
and C
1
are constants.
k
and
are the turbulent Prandtl numbers for

k and , respectively. S
k
and S

Note that the k equation (Equation 10.4-15) is the same as that in the
standard k- model (Equation 10.4-1) and the RNG k- model (Equa-
tion 10.4-4), except for the model constants. However, the form of the
equation is quite dierent from those in the standard and RNG-based k-
models (Equations 10.4-2 and 10.4-5). One of the noteworthy features
is that the production term in the equation (the second term on the
right-hand side of Equation 10.4-16) does not involve the production of
k; i.e., it does not contain the same G
k
term as the other k- models. It
is believed that the present form better represents the spectral energy
transfer. Another desirable feature is that the destruction term (the next
to last term on the right-hand side of Equation 10.4-16) does not have
any singularity; i.e., its denominator never vanishes, even if k vanishes
or becomes smaller than zero. This feature is contrasted with traditional
k- models, which have a singularity due to k in the denominator.
This model has been extensively validated for a wide range of ows [116,
209], including rotating homogeneous shear ows, free ows including
jets and mixing layers, channel and boundary layer ows, and separated
ows. For all these cases, the performance of the model has been found to
be substantially better than that of the standard k- model. Especially
noteworthy is the fact that the realizable k- model resolves the round-
jet anomaly; i.e., it predicts the spreading rate for axisymmetric jets as
well as that for planar jets.
Modeling the Turbulent Viscosity
As in other k- models, the eddy viscosity is computed from
t
= C
k
2
(10.4-17)
The dierence between the realizable k- model and the standard and
RNG k- models is that C
is no longer constant. It is computed from

C
=
1
A
0
+ A
s
kU
(10.4-18)
where
U

_
S
ij
S
ij
+

ij
ij
(10.4-19)
and
ij
=
ij
2
ijk
ij
=
ij
ijk
k
where
ij
is the mean rate-of-rotation tensor viewed in a rotating refer-
ence frame with the angular velocity
k
. The model constants A
0
and
A
s
are given by
A
0
= 4.04, A
s
=
6 cos
where
=
1
3
cos
1
(
6W), W =
S
ij
S
jk
S
ki
S
,

S =
_
S
ij
S
ij
S
ij
=
1
2
_
u
j
x
i
+
u
i
x
j
_
It can be seen that C
is a function of the mean strain and rotation rates,

the angular velocity of the system rotation, and the turbulence elds (k
Modeling Turbulence
and ). C
in Equation 10.4-17 can be shown to recover the standard

value of 0.09 for an inertial sublayer in an equilibrium boundary layer.
Model Constants
2
,
k
, and
have been established to ensure that

the model performs well for certain canonical ows. The model constants
are
C
1
= 1.44, C
2
= 1.9,
k
= 1.0,
= 1.2
10.4.4 Modeling Turbulent Production in the k- Models
The term G
k
, representing the production of turbulence kinetic energy,
is modeled identically for the standard, RNG, and realizable k- models.
From the exact equation for the transport of k, this term may be dened
as
G
k
= u
i
u
j
u
j
x
i
(10.4-20)
To evaluate G
k
in a manner consistent with the Boussinesq hypothesis,
G
k
=
t
S
2
(10.4-21)
where S is the modulus of the mean rate-of-strain tensor, dened as
S
_
2S
ij
S
ij
(10.4-22)
10.4.5 Eects of Buoy ancy on Turbulence in the k- Models
When a non-zero gravity eld and temperature gradient are present si-
multaneously, the k- models in FLUENT account for the generation of k
due to buoyancy (G
b
in Equations 10.4-1, 10.4-4, and 10.4-15), and the
corresponding contribution to the production of in Equations 10.4-2,
10.4-5, and 10.4-16.
The generation of turbulence due to buoyancy is given by
G
b
= g
i
t
Pr
t
T
x
i
(10.4-23)
where Pr
t
is the turbulent Prandtl number for energy and g
i
is the com-
ponent of the gravitational vector in the ith direction. For the standard
and realizable k- models, the default value of Pr
t
is 0.85. In the case
of the RNG k- model, Pr
t
= 1/, where is given by Equation 10.4-9,
but with
0
= 1/Pr = k/c
p
. The coecient of thermal expansion, , is
dened as
=
1
T
_
p
(10.4-24)
For ideal gases, Equation 10.4-23 reduces to
G
b
= g
i
t
Pr
t
x
i
(10.4-25)
It can be seen from the transport equations for k (Equations 10.4-1,
10.4-4, and 10.4-15) that turbulence kinetic energy tends to be aug-
mented (G
b
> 0) in unstable stratication. For stable stratication,
buoyancy tends to suppress the turbulence (G
b
< 0). In FLUENT, the
eects of buoyancy on the generation of k are always included when
you have both a non-zero gravity eld and a non-zero temperature (or
density) gradient.
While the buoyancy eects on the generation of k are relatively well
understood, the eect on is less clear. In FLUENT, by default, the
buoyancy eects on are neglected simply by setting G
b
to zero in the
transport equation for (Equation 10.4-2, 10.4-5, or 10.4-16).
However, you can include the buoyancy eects on in the Viscous Model
panel. In this case, the value of G
b
given by Equation 10.4-25 is used in
the transport equation for (Equation 10.4-2, 10.4-5, or 10.4-16).
The degree to which is aected by the buoyancy is determined by the
constant C
3
. In FLUENT, C
3
is not specied, but is instead calculated
according to the following relation [90]:
Modeling Turbulence
C
3
= tanh
v
u
(10.4-26)
where v is the component of the ow velocity parallel to the gravita-
tional vector and u is the component of the ow velocity perpendicular
to the gravitational vector. In this way, C
3
will become 1 for buoyant
shear layers for which the main ow direction is aligned with the direc-
tion of gravity. For buoyant shear layers that are perpendicular to the
gravitational vector, C
3
will become zero.
10.4.6 Eects of Compressibility on Turbulence in the k-
Models
For high-Mach-number ows, compressibility aects turbulence through
so-called dilatation dissipation, which is normally neglected in the
modeling of incompressible ows [267]. Neglecting the dilatation dis-
sipation fails to predict the observed decrease in spreading rate with
increasing Mach number for compressible mixing and other free shear
layers. To account for these eects in the k- models in FLUENT, the
dilatation dissipation term, Y
M
, is included in the k equation. This term
is modeled according to a proposal by Sarkar [197]:
Y
M
= 2M
2
t
(10.4-27)
where M
t
is the turbulent Mach number, dened as
M
t
=
k
a
2
(10.4-28)
where a (
RT) is the speed of sound.

This compressibility modication always takes eect when the compress-
ible form of the ideal gas law is used.
10.4.7 Convective Heat and Mass Transfer Modeling in the k-
Models
In FLUENT, turbulent heat transport is modeled using the concept of
Reynolds analogy to turbulent momentum transfer. The modeled
energy equation is thus given by the following:
t
(E) +

x
i
[u
i
(E +p)] =

x
j
_
k
e
T
x
j
+ u
i
(
ij
)
e
_
+S
h
(10.4-29)
where E is the total energy, k
e
is the eective thermal conductivity,
and (
ij
)
e
is the deviatoric stress tensor, dened as
(
ij
)
e
=
e
_
u
j
x
i
+
u
i
x
j
_
2
3
e
u
i
x
i
ij
ij
)
e
segregated solver, but it can be enabled in the Viscous Model panel.
Additional terms may appear in the energy equation, depending on the
physical models you are using. See Section 11.2.1 for more details.
For the standard and realizable k- models, the eective thermal con-
ductivity is given by
k
e
= k +
c
p
t
Pr
t
where k, in this case, is the thermal conductivity. The default value of
the turbulent Prandtl number is 0.85. You can change the value of the
turbulent Prandtl number in the Viscous Model panel.
For the RNG k- model, the eective thermal conductivity is
k
e
= c
p
e
Modeling Turbulence
where is calculated from Equation 10.4-9, but with
0
= 1/Pr = k/c
p
.
The fact that varies with
mol
/
e
, as in Equation 10.4-9, is an advan-
tage of the RNG k- model. It is consistent with experimental evidence
indicating that the turbulent Prandtl number varies with the molecular
Prandtl number and turbulence [111]. Equation 10.4-9 works well across
a very broad range of molecular Prandtl numbers, from liquid metals
(Pr 10
2
) to paran oils (Pr 10
3
), which allows heat transfer to be
calculated in low-Reynolds-number regions. Equation 10.4-9 smoothly
predicts the variation of eective Prandtl number from the molecular
value ( = 1/Pr) in the viscosity-dominated region to the fully turbu-
lent value ( = 1.393) in the fully turbulent regions of the ow.
Turbulent mass transfer is treated similarly. For the standard and re-
alizable k- models, the default turbulent Schmidt number is 0.7. This
default value can be changed in the Viscous Model panel. For the RNG
model, the eective turbulent diusivity for mass transfer is calculated
in a manner that is analogous to the method used for the heat trans-
port. The value of
0
in Equation 10.4-9 is
0
= 1/Sc, where Sc is the
molecular Schmidt number.
10.5 The Standard and SST k- Models
This section presents the standard and shear-stress transport (SST) k-
models. Both models have similar forms, with transport equations
for k and . The major ways in which the SST model diers from the
standard model are as follows:
gradual change from the standard k- model in the inner region of
the boundary layer to a high-Reynolds-number version of the k-
model in the outer part of the boundary layer
modied turbulent viscosity formulation to account for the trans-
port eects of the principal turbulent shear stress
The transport equations, methods of calculating turbulent viscosity, and
methods of calculating model constants and other terms are presented
separately for each model.
The standard k- model is an empirical model based on model transport
equations for the turbulence kinetic energy (k) and the specic dissipa-
tion rate (), which can also be thought of as the ratio of to k [267].
As the k- model has been modied over the years, production terms
have been added to both the k and equations, which have improved
the accuracy of the model for predicting free shear ows.
Transport Equations for the Standard k- Model
The turbulence kinetic energy, k, and the specic dissipation rate, , are
obtained from the following transport equations:
t
(k) +

x
i
(ku
i
) =

x
j
_
k
k
x
j
_
+G
k
Y
k
+ S
k
(10.5-1)
and
Modeling Turbulence
t
() +

x
i
(u
i
) =

x
j
_
x
j
_
+ G
+ S
(10.5-2)
k
energy due to mean velocity gradients. G
represents the generation of

.
k
and
represent the eective diusivity of k and , respectively.

Y
k
and Y
represent the dissipation of k and due to turbulence. All

of the above terms are calculated as described below. S
k
and S
are
user-dened source terms.
Modeling the Eective Diusivity
The eective diusivities for the k- model are given by
k
= +

t
k
(10.5-3)
= +

t
(10.5-4)
where
k
and
are the turbulent Prandtl numbers for k and , respec-

tively. The turbulent viscosity,
t
, is computed by combining k and
as follows:
t
=
(10.5-5)
Low-Reynolds-Number Correction
The coecient
damps the turbulent viscosity causing a low-Reynolds-

number correction. It is given by
0
+ Re
t
/R
k
1 + Re
t
/R
k
_
(10.5-6)
where
Re
t
=
k
(10.5-7)
R
k
= 6 (10.5-8)
0
=

i
3
(10.5-9)
i
= 0.072 (10.5-10)
Note that, in the high-Reynolds-number form of the k- model,
= 1.
Modeling the Turbulence Production
Production of k
The term G
k
represents the production of turbulence kinetic energy.
From the exact equation for the transport of k, this term may be dened
as
G
k
= u
i
u
j
u
j
x
i
(10.5-11)
To evaluate G
k
in a manner consistent with the Boussinesq hypothesis,
G
k
=
t
S
2
(10.5-12)
where S is the modulus of the mean rate-of-strain tensor, dened in the
same way as for the k- model (see Equation 10.4-22).
Production of
The production of is given by
G
k
G
k
(10.5-13)
where G
k
is given by Equation 10.5-11.
The coecient is given by
=

0
+ Re
t
/R
1 + Re
t
/R
_
(10.5-14)
Modeling Turbulence
where R
= 2.95.
and Re
t
are given by Equations 10.5-6 and 10.5-7,
respectively.
Note that, in the high-Reynolds-number form of the k- model, =
= 1.
Modeling the Turbulence Dissipation
Dissipation of k
The dissipation of k is given by
Y
k
=
k (10.5-15)
where
f
=
_
_
_
1
k
0
1+680
2
k
1+400
2
k
k
> 0
(10.5-16)
where
k

1
3
k
x
j
x
j
(10.5-17)
and
i
[1 +
F(M
t
)] (10.5-18)
i
=
_
4/15 + (Re
t
/R
)
4
1 + (Re
t
/R
)
4
_
(10.5-19)
= 1.5 (10.5-20)
R
= 8 (10.5-21)
= 0.09 (10.5-22)
where Re
t
is given by Equation 10.5-7.
Dissipation of
The dissipation of is given by
Y
= f

2
(10.5-23)
where
f
=
1 + 70
1 + 80
(10.5-24)
ij
jk
S
ki
(
)
3
(10.5-25)
ij
=
1
2
_
u
i
x
j
u
j
x
i
_
(10.5-26)
The strain rate tensor, S
ij
is dened in Equation 10.3-11. Also,
=
i
_
1

F(M
t
)
_
(10.5-27)
i
and F(M
t
) are dened by Equations 10.5-19 and 10.5-28, respectively.
Compressibility Correction
The compressibility function, F(M
t
), is given by
F(M
t
) =
_
0 M
t
M
t0
M
2
t
M
2
t0
M
t
> M
t0
(10.5-28)
where
M
2
t

2k
a
2
(10.5-29)
M
t0
= 0.25 (10.5-30)
a =
_
RT (10.5-31)
Note that, in the high-Reynolds-number form of the k- model,
i
=
.
In the incompressible form,
i
.
Modeling Turbulence
Model Constants
= 1,
= 0.52,
0
=
1
9
,
= 0.09,
i
= 0.072, R
= 8
R
k
= 6, R
= 2.95,
= 1.5, M
t0
= 0.25,
k
= 2.0,
= 2.0
Wall Boundary Conditions
The wall boundary conditions for the k equation in the k- models are
treated in the same way as the k equation is treated when enhanced wall
treatments are used with the k- models. This means that all boundary
conditions for wall-function meshes will correspond to the wall func-
tion approach, while for the ne meshes, the appropriate low-Reynolds-
number boundary conditions will be applied.
In FLUENT the value of at the wall is specied as
w
=
(u
)
2

+
(10.5-32)
The asymptotic value of
+
in the laminar sublayer is given by
+
= min
_
+
w
,
6
(y
+
)
2
_
(10.5-33)
where
+
w
=
_
_
_
50
k
+
s
_
2
k
+
s
< 25
100
k
+
s
k
+
s
25
(10.5-34)
where
k
+
s
= max
_
1.0,
k
s
u
_
(10.5-35)
and k
s
is the roughness height.
In the logarithmic (or turbulent) region, the value of
+
is
+
=
1
_
du
+
turb
dy
+
(10.5-36)
which leads to the value of in the wall cell as
=
u
y
(10.5-37)
Note that in the case of a wall cell being placed in the buer region,
FLUENT will blend
+
between the logarithmic and laminar sublayer
values.
10.5.2 The Shear-Stress Transport (SST) k- Model
In addition to the standard k- model described in Section 10.5.1, FLU-
ENT also provides a variation called the shear-stress transport (SST)
k- model, so named because the denition of the turbulent viscosity
is modied to account for the transport of the principal turbulent shear
stress. It is this feature that gives the SST k- model an advantage in
terms of performance over both the standard k- model and the standard
k- model. Other modications include the addition of a cross-diusion
term in the equation and a blending function to ensure that the model
equations behave appropriately in both the near-wall and far-eld zones.
Transport Equations for the SST k- Model
The SST k- model has a similar form to the standard k- model:
t
(k) +

x
i
(ku
i
) =

x
j
_
k
k
x
j
_
+ G
k
Y
k
+ S
k
(10.5-38)
and
t
() +

x
i
(u
i
) =

x
j
_
x
j
_
+ G
+D
+ S
(10.5-39)
k
energy due to mean velocity gradients, calculated as described in Sec-
tion 10.5.1. G
represents the generation of , calculated as described in

Modeling Turbulence
Section 10.5.1.
k
and
represent the eective diusivity of k and ,

respectively, which are calculated as described below. Y
k
and Y
repre-
sent the dissipation of k and due to turbulence, calculated as described
in Section 10.5.1. D
represents the cross-diusion term, calculated as

described below. S
k
and S

Modeling the Eective Diusivity
The eective diusivities for the SST k- model are given by
k
= +

t
k
(10.5-40)
= +

t
(10.5-41)
where
k
and
are the turbulent Prandtl numbers for k and , respec-

tively. The turbulent viscosity,
t
, is computed as follows:
t
=
k
1
max
_
1
,
F
2
a
1
_ (10.5-42)
where

_
2
ij
ij
(10.5-43)
k
=
1
F
1
/
k,1
+ (1 F
1
)/
k,2
(10.5-44)
=
1
F
1
/
,1
+ (1 F
1
)/
,2
(10.5-45)
ij
is the mean rate-of-rotation tensor and
is dened in Equation 10.5-6.

The blending functions, F
1
and F
2
, are given by
F
1
= tanh
_
4
1
_
(10.5-46)
1
= min
_
max
_
k
0.09y
,
500
y
2
_
,
4k
,2
D
+
y
2
_
(10.5-47)
D
+
= max
_
2
1
,2
1
k
x
j
x
j
, 10
20
_
(10.5-48)
F
2
= tanh
_
2
2
_
(10.5-49)
2
= max
_
2
k
0.09y
,
500
y
2
_
(10.5-50)
where y is the distance to the next surface and D
+
is the positive portion

of the cross-diusion term (see Equation 10.5-60).
Modeling the Turbulence Production
Production of k
The term G
k
represents the production of turbulence kinetic energy,
and is dened in the same manner as in the standard k- model. See
Section 10.5.1 for details.
Production of
The term G
represents the production of and is given by

G
t
G
k
(10.5-51)
Note that this formulation diers from the standard k- model. The
dierence between the two models also exists in the way the term
is evaluated. In the standard k- model,
is dened as a constant
(0.52). For the SST k- model,
is given by
= F
1
,1
+ (1 F
1
)
,2
(10.5-52)
where
Modeling Turbulence
,1
=

i,1
w,1
_
(10.5-53)
,2
=

i,2
w,2
_
(10.5-54)
where is 0.41.
i,1
and
i,2
are given by Equations 10.5-58 and 10.5-59,
respectively.
Modeling the Turbulence Dissipation
Dissipation of k
The term Y
k
represents the dissipation of turbulence kinetic energy, and
is dened in a similar manner as in the standard k- model (see Sec-
tion 10.5.1). The dierence is in the way the term f
is evaluated. In
the standard k- model, f
is dened as a piecewise function. For the

SST k- model, f
is a constant equal to 1. Thus,

Y
k
=
k (10.5-55)
Dissipation of
The term Y
represents the dissipation of , and is dened in a similar

manner as in the standard k- model (see Section 10.5.1). The dier-
ence is in the way the terms
i
and f
are evaluated. In the standard

k- model,
i
is dened as a constant (0.072) and f
is dened in Equa-
tion 10.5-24. For the SST k- model, f
is a constant equal to 1. Thus,

Y
k
=
2
(10.5-56)
Instead of a having a constant value,
i
is given by
i
= F
1
i,1
+ (1 F
1
)
i,2
(10.5-57)
where
10.6 The Rey nolds Stress Model (RSM)
i,1
= 0.075 (10.5-58)
i,2
= 0.0828 (10.5-59)
and F
1
is obtained from Equation 10.5-46.
Cross-Diusion Modication
The SST k- model is based on both the standard k- model and the
standard k- model. To blend these two models together, the stan-
dard k- model has been transformed into equations based on k and ,
which leads to the introduction of a cross-diusion term (D
in Equa-
tion 10.5-39). D
is dened as
D
= 2 (1 F
1
)
,2
1
k
x
j
x
j
(10.5-60)
For details about the various k- models, see Section 10.4.
Model Constants
k,1
= 1.176,
,1
= 2.0,
k,2
= 1.0,
,2
= 1.168
a
1
= 0.31,
i,1
= 0.075
i,2
= 0.0828
All additional model constants (
,
0
,
, R
, R
k
, R
, and
M
t0
) have the same values as for the standard k- model (see Sec-
tion 10.5.1).
The Reynolds stress model [75, 125, 126] involves calculation of the in-
dividual Reynolds stresses, u
i
u
j
, using dierential transport equations.
The individual Reynolds stresses are then used to obtain closure of the
Reynolds-averaged momentum equation (Equation 10.2-4).
Modeling Turbulence
The exact form of the Reynolds stress transport equations may be de-
rived by taking moments of the exact momentum equation. This is
a process wherein the exact momentum equations are multiplied by a
uctuating property, the product then being Reynolds-averaged. Unfor-
tunately, several of the terms in the exact equation are unknown and
modeling assumptions are required in order to close the equations.
In this section, the Reynolds stress transport equations are presented
together with the modeling assumptions required to attain closure.
10.6.1 The Rey nolds Stress Transport Equations
The exact transport equations for the transport of the Reynolds stresses,
u
i
u
j
, may be written as follows:
t
( u
i
u
j
)
. .
Local Time Derivative
+

x
k
(u
k
u
i
u
j
)
. .
C
ij
Convection
=

x
k
_
u
i
u
j
u
k
+p
_
kj
u
i
+
ik
u
j
_
_
. .
D
T,ij
Turbulent Diusion
+

x
k
_

x
k
(u
i
u
j
)
_
. .
D
L,ij
Molecular Diusion

_
u
i
u
k
u
j
x
k
+ u
j
u
k
u
i
x
k
_
. .
P
ij
Stress Production
(g
i
u
j
+ g
j
u
i
)
. .
G
ij
Buoyancy Production
+ p
_
u
i
x
j
+
u
j
x
i
_
. .
ij
Pressure Strain
2
u
i
x
k
u
j
x
k
. .
ij
Dissipation
2
k
_
u
j
u
ikm
+ u
i
u
jkm
_
. .
F
ij
Production by System Rotation
+ S
user
. .
User-Dened Source Term
(10.6-1)
Of the various terms in these exact equations, C
ij
, D
L,ij
, P
ij
, and F
ij
do not require any modeling. However, D
T,ij
, G
ij
,
ij
, and
ij
need to
be modeled to close the equations. The following sections describe the
modeling assumptions required to close the equation set.
10.6.2 Modeling Turbulent Diusive Transport
D
T,ij
can be modeled by the generalized gradient-diusion model of Daly
and Harlow [48]:
D
T,ij
= C
s
x
k
_
ku
k
u
i
u
j
x
_
(10.6-2)
However, this equation can result in numerical instabilities, so it has been
simplied in FLUENT to use a scalar turbulent diusivity as follows [138]:
D
T,ij
=

x
k
_
k
u
i
u
j
x
k
_
(10.6-3)
t
, is computed using Equation 10.6-27.
Lien and Leschziner [138] derived a value of
k
= 0.82 by applying the
generalized gradient-diusion model, Equation 10.6-2, to the case of a
planar homogeneous shear ow. Note that this value of
k
is dierent
from that in the standard and realizable k- models, in which
k
= 1.0.
10.6.3 Modeling the Pressure-Strain Term
Linear Pressure-Strain Model
By default in FLUENT, the pressure-strain term,
ij
, in Equation 10.6-1
is modeled according to the proposals by Gibson and Launder [75], Fu
et al. [71], and Launder [124, 125].
The classical approach to modeling
ij
uses the following decomposition:
ij
=
ij,1
+
ij,2
+
ij,w
(10.6-4)
Modeling Turbulence
where
ij,1
is the slow pressure-strain term, also known as the return-to-
isotropy term,
ij,2
is called the rapid pressure-strain term, and
ij,w
is
the wall-reection term.
The slow pressure-strain term,
ij,1
, is modeled as
ij,1
C
1

k
_
u
i
u
j

2
3
ij
k
_
(10.6-5)
with C
1
= 1.8.
The rapid pressure-strain term,
ij,2
, is modeled as
ij,2
C
2
_
(P
ij
+F
ij
+ G
ij
C
ij
)
2
3
ij
(P +GC)
_
(10.6-6)
where C
2
= 0.60, P
ij
, F
ij
, G
ij
, and C
ij
are dened as in Equation 10.6-1,
P =
1
2
P
kk
, G =
1
2
G
kk
, and C =
1
2
C
kk
.
The wall-reection term,
ij,w
, is responsible for the redistribution of
normal stresses near the wall. It tends to damp the normal stress per-
pendicular to the wall, while enhancing the stresses parallel to the wall.
This term is modeled as
ij,w
C
k
_
u
k
u
m
n
k
n
m
ij

3
2
u
i
u
k
n
j
n
k
3
2
u
j
u
k
n
i
n
k
_
k
3/2
C
d
+ C
2
_
km,2
n
k
n
m
ij

3
2
ik,2
n
j
n
k
3
2
jk,2
n
i
n
k
_
k
3/2
C
d
(10.6-7)
where C
1
= 0.5, C
2
= 0.3, n
k
is the x
k
component of the unit normal to
the wall, d is the normal distance to the wall, and C
= C
3/4
/, where
C
= 0.09 and is the von Karm an constant (= 0.4187).
ij,w
is included by default in the Reynolds stress model.
Low-Re Modications to the Linear Pressure-Strain Model
When the RSM is applied to near-wall ows using the enhanced wall
treatment described in Section 10.8.3, the pressure-strain model needs
to be modied. The modication used in FLUENT species the values
of C
1
, C
2
, C
1
, and C
2
as functions of the Reynolds stress invariants and
the turbulent Reynolds number, according to the suggestion of Launder
and Shima [127]:
C
1
= 1 + 2.58A
_
A
2
_
1 exp
_
(0.0067Re
t
)
2
__
(10.6-8)
C
2
= 0.75
A (10.6-9)
C
1
=
2
3
C
1
+ 1.67 (10.6-10)
C
2
= max
_
2
3
C
2
1
6
C
2
, 0
_
(10.6-11)
with the turbulent Reynolds number dened as Re
t
= (k
2
/). The
parameter A and tensor invariants, A
2
and A
3
, are dened as
A
_
1
9
8
(A
2
A
3
)
_
(10.6-12)
A
2
a
ik
a
ki
(10.6-13)
A
3
a
ik
a
kj
a
ji
(10.6-14)
a
ij
is the Reynolds-stress anisotropy tensor, dened as
a
ij
=
_
u
i
u
j
+
2
3
k
ij
k
_
(10.6-15)
The modications detailed above are employed only when the enhanced
wall treatment is selected in the Viscous Model panel.
Modeling Turbulence
Quadratic Pressure-Strain Model
An optional pressure-strain model proposed by Speziale, Sarkar, and
Gatski [228] is provided in FLUENT. This model has been demonstrated
to give superior performance in a range of basic shear ows, including
plane strain, rotating plane shear, and axisymmetric expansion/contrac-
tion. This improved accuracy should be benecial for a wider class of
complex engineering ows, particularly those with streamline curvature.
The quadratic pressure-strain model can be selected as an option in the
Viscous Model panel.
This model is written as follows:
ij
= (C
1
+ C
1
P) b
ij
+ C
2
_
b
ik
b
kj
1
3
b
mn
b
mn
ij
_
+
_
C
3
C
3
_
b
ij
b
ij
_
kS
ij
+C
4
k
_
b
ik
S
jk
+ b
jk
S
ik
2
3
b
mn
S
mn
ij
_
+C
5
k (b
ik
jk
+ b
jk
ik
) (10.6-16)
where b
ij
is the Reynolds-stress anisotropy tensor dened as
b
ij
=
_
u
i
u
j
+
2
3
k
ij
2k
_
(10.6-17)
The mean strain rate, S
ij
, is dened as
S
ij
=
1
2
_
u
j
x
i
+
u
i
x
j
_
(10.6-18)
The mean rate-of-rotation tensor,
ij
, is dened by
ij
=
1
2
_
u
i
x
j
u
j
x
i
_
(10.6-19)
The constants are
C
1
= 3.4, C
1
= 1.8, C
2
= 4.2, C
3
= 0.8, C
3
= 1.3, C
4
= 1.25, C
5
= 0.4
The quadratic pressure-strain model does not require a correction to
account for the wall-reection eect in order to obtain a satisfactory
solution in the logarithmic region of a turbulent boundary layer. It
should be noted, however, that the quadratic pressure-strain model is
not available when the enhanced wall treatment is selected in the Viscous
Model panel.
10.6.4 Eects of Buoy ancy on Turbulence
The production terms due to buoyancy are modeled as
G
ij
=

t
Pr
t
_
g
i
T
x
j
+ g
j
T
x
i
_
(10.6-20)
where Pr
t
is the turbulent Prandtl number for energy, with a default
value of 0.85.
Using the denition of the coecient of thermal expansion, , given by
Equation 10.4-24, the following expression is obtained for G
ij
for ideal
gases:
G
ij
=

t
Pr
t
_
g
i
x
j
+ g
j
x
i
_
(10.6-21)
10.6.5 Modeling the Turbulence Kinetic Energy
In general, when the turbulence kinetic energy is needed for modeling a
specic term, it is obtained by taking the trace of the Reynolds stress
tensor:
k =
1
2
u
i
u
i
(10.6-22)
Modeling Turbulence
As described in Section 10.6.8, an option is available in FLUENT to solve
a transport equation for the turbulence kinetic energy in order to obtain
boundary conditions for the Reynolds stresses. In this case, the following
model equation is used:
t
(k) +

x
i
(ku
i
) =

x
j
_
_
+

t
k
_
k
x
j
_
+
1
2
(P
ii
+ G
ii
) (1 + 2M
2
t
) + S
k
(10.6-23)
where
k
= 0.82 and S
k
is a user-dened source term. Equation 10.6-23 is
obtainable by contracting the modeled equation for the Reynolds stresses
(Equation 10.6-1). As one might expect, it is essentially identical to
Equation 10.4-1 used in the standard k- model.
Although Equation 10.6-23 is solved globally throughout the ow do-
main, the values of k obtained are used only for boundary conditions. In
every other case, k is obtained from Equation 10.6-22. This is a minor
point, however, since the values of k obtained with either method should
be very similar.
10.6.6 Modeling the Dissipation Rate
The dissipation tensor,
ij
, is modeled as
ij
=
2
3
ij
( + Y
M
) (10.6-24)
where Y
M
= 2M
2
t
is an additional dilatation dissipation term ac-
cording to the model by Sarkar [197]. The turbulent Mach number in
this term is dened as
M
t
=
k
a
2
(10.6-25)
where a (

RT) is the speed of sound. This compressibility modi-
cation always takes eect when the compressible form of the ideal gas
law is used.
The scalar dissipation rate, , is computed with a model transport equa-
tion similar to that used in the standard k- model:
t
() +

x
i
(u
i
) =

x
j
_
_
+

t
_

x
j
_
+
C
1
1
2
[P
ii
+ C
3
G
ii
]

k
C
2
2
k
+ S
(10.6-26)
where
= 1.0, C
1
= 1.44, C
2
= 1.92, C
3
is evaluated as a function of
the local ow direction relative to the gravitational vector, as described
in Section 10.4.5, and S
is a user-dened source term.

10.6.7 Modeling the Turbulent Viscosity
t
, is computed similarly to the k- models:
t
= C
k
2
(10.6-27)
where C
= 0.09.
10.6.8 Boundary Conditions for the Rey nolds Stresses
Whenever ow enters the domain, FLUENT requires values for individual
Reynolds stresses, u
i
u
j
, and for the turbulence dissipation rate, . These
quantities can be input directly or derived from the turbulence intensity
and characteristic length, as described in Section 10.10.2.
At walls, FLUENT computes the near-wall values of the Reynolds stresses
and from wall functions (see Section 10.8.2). FLUENT applies explicit
wall boundary conditions for the Reynolds stresses by using the log-law
and the assumption of equilibrium, disregarding convection and diusion
in the transport equations for the stresses (Equation 10.6-1). Using
a local coordinate system, where is the tangential coordinate, is
Modeling Turbulence
the normal coordinate, and is the binormal coordinate, the Reynolds
stresses at the wall-adjacent cells are computed from
u
k
= 1.098,
u
k
= 0.247,
u
k
= 0.655,
u
k
= 0.255 (10.6-28)
To obtain k, FLUENT solves the transport equation of Equation 10.6-23.
For reasons of computational convenience, the equation is solved globally,
even though the values of k thus computed are needed only near the wall;
in the far eld k is obtained directly from the normal Reynolds stresses
using Equation 10.6-22. By default, the values of the Reynolds stresses
near the wall are xed using the values computed from Equation 10.6-28,
and the transport equations in Equation 10.6-1 are solved only in the
bulk ow region.
Alternatively, the Reynolds stresses can be explicitly specied in terms
of wall-shear stress, instead of k:
u
u
2
= 5.1,
u
u
2
= 1.0,
u
u
2
= 2.3,
u
u
2
= 1.0 (10.6-29)
where u
is the friction velocity dened by u
w
/, where
w
is the
wall-shear stress. When this option is chosen, the k transport equation
is not solved.
10.6.9 Convective Heat and Mass Transfer Modeling
With the Reynolds stress model in FLUENT, turbulent heat transport is
modeled using the concept of Reynolds analogy to turbulent momentum
transfer. The modeled energy equation is thus given by the following:
t
(E) +

x
i
[u
i
(E + p)] =

x
j
_
_
k +
c
p
t
Pr
t
_
T
x
j
+ u
i
(
ij
)
e
_
+ S
h
(10.6-30)
10.7 The Large Eddy Simulation (LES) Model
where E is the total energy and (
ij
)
e
is the deviatoric stress tensor,
dened as
(
ij
)
e
=
e
_
u
j
x
i
+
u
i
x
j
_
2
3
e
u
i
x
i
ij
ij
)
e
segregated solver, but it can be enabled in the Viscous Model panel. The
default value of the turbulent Prandtl number is 0.85. You can change
the value of Pr
t
in the Viscous Model panel.
Turbulent mass transfer is treated similarly, with a default turbulent
Schmidt number of 0.7. This default value can be changed in the Viscous
Model panel.
Turbulent ows are characterized by eddies with a wide range of length
and time scales. The largest eddies are typically comparable in size
to the characteristic length of the mean ow. The smallest scales are
responsible for the dissipation of turbulence kinetic energy.
It is theoretically possible to directly resolve the whole spectrum of tur-
bulent scales using an approach known as direct numerical simulation
(DNS). DNS is not, however, feasible for practical engineering prob-
lems. To understand the large computational cost of DNS, consider that
the ratio of the large (energy-containing) scales to the small (energy-
dissipating) scales is proportional to Re
3/4
t
, where Re
t
is the turbulent
Reynolds number. Therefore, to resolve all the scales, the mesh size
in three dimensions will be proportional to Re
9/4
t
. Simple arithmetic
shows that, for high Reynolds numbers, the mesh sizes required for DNS
are prohibitive. Adding to the computational cost is the fact that the
simulation will be a transient one with very small time steps, since the
temporal resolution requirements are governed by the dissipating scales,
rather than the mean ow or the energy-containing eddies.
Modeling Turbulence
As explained in Section 10.2.1, the conventional approach to ow sim-
ulations employs the solution of the Reynolds-averaged Navier-Stokes
(RANS) equations. In the RANS approach, all the turbulent motions
are modeled, resulting in a signicant savings in computational eort.
Conceptually, large eddy simulation (LES) is situated somewhere be-
tween DNS and the RANS approach. Basically large eddies are resolved
directly in LES, while small eddies are modeled. The rationale behind
LES can be summarized as follows:
Momentum, mass, energy, and other passive scalars are trans-
ported mostly by large eddies.
Large eddies are more problem-dependent. They are dictated by
the geometries and boundary conditions of the ow involved.
Small eddies are less dependent on the geometry, tend to be more
isotropic, and are consequently more universal.
The chance of nding a universal model is much higher when only
small eddies are modeled.
Solving only for the large eddies and modeling the smaller scales results
in mesh resolution requirements that are much less restrictive than with
DNS. Typically, mesh sizes can be at least one order of magnitude smaller
than with DNS. Furthermore, the time step sizes will be proportional to
the eddy-turnover time, which is much less restrictive than with DNS.
In practical terms, however, extremely ne meshes are still required. It
is only due to the explosive increases in computer hardware performance
coupled with the availability of parallel processing that LES can even be
considered as a possibility for engineering calculations.
The following sections give details of the governing equations for LES,
present the two options for modeling the subgrid-scale stresses (necessary
to achieve closure of the governing equations), and discuss the relevant
boundary conditions.
10.7.1 Filtered Navier-Stokes Equations
The governing equations employed for LES are obtained by ltering the
time-dependent Navier-Stokes equations in either Fourier (wave-number)
space or conguration (physical) space. The ltering process eectively
lters out the eddies whose scales are smaller than the lter width or grid
spacing used in the computations. The resulting equations thus govern
the dynamics of large eddies.
A ltered variable (denoted by an overbar) is dened by
(x) =
_
D
(x
)G(x, x
)dx
(10.7-1)
where D is the uid domain, and G is the lter function that determines
the scale of the resolved eddies.
In FLUENT, the nite-volume discretization itself implicitly provides the
ltering operation:
(x) =
1
V
_
V
(x
) dx
, x
V (10.7-2)
where V is the volume of a computational cell. The lter function,
G(x, x
), implied here is then

G(x, x
)
_
1/V, x
V
0, x
otherwise
(10.7-3)
Since the application of LES to compressible ows is still in its infancy,
the theory is presented here for incompressible ows. It is assumed that
the LES model in FLUENT will be applied to essentially incompressible
(but not necessarily constant-density) ows.
Filtering the incompressible Navier-Stokes equations, one obtains
t
+

x
i
(u
i
) = 0 (10.7-4)
Modeling Turbulence
and
t
(u
i
) +

x
j
(u
i
u
j
) =

x
j
_
u
i
x
j
_
p
x
i

ij
x
j
(10.7-5)
where
ij
is the subgrid-scale stress dened by
ij
u
i
u
j
u
i
u
j
(10.7-6)
The similarity between the ltered equations, 10.7-4 through 10.7-6, and
the incompressible form of the RANS equations, Equations 10.2-3 and
10.2-4, is obvious. The major dierence is that the dependent variables
are now ltered quantities rather than mean quantities, and the expres-
sions for the turbulent stresses dier.
10.7.2 Subgrid-Scale Models
The subgrid-scale stresses resulting from the ltering operation are un-
known, and require modeling. The majority of subgrid-scale models in
use today are eddy viscosity models of the following form:
ij

1
3
kk
ij
= 2
t
S
ij
(10.7-7)
where
t
is the subgrid-scale turbulent viscosity, and S
ij
is the rate-of-
strain tensor for the resolved scale dened by
S
ij

1
2
_
u
i
x
j
+
u
j
x
i
_
(10.7-8)
FLUENT contains two models for
t
: the Smagorinsky-Lilly model and
the RNG-based subgrid-scale model.
Smagorinsky -Lilly Model
The most basic of subgrid-scale models was proposed by Smagorin-
sky [214] and further developed by Lilly [139]. In the Smagorinsky-Lilly
model, the eddy viscosity is modeled by
t
= L
2
s
(10.7-9)
where L
s
is the mixing length for subgrid scales and

_
2S
ij
S
ij
. C
s
is the Smagorinsky constant. In FLUENT, L
s
is computed using
L
s
= min
_
d, C
s
V
1/3
_
(10.7-10)
where is the von Karm an constant, d is the distance to the closest wall,
and V is the volume of the computational cell.
Lilly derived a value of 0.23 for C
s
from homogeneous isotropic turbu-
lence in the inertial subrange. However, this value was found to cause
excessive damping of large-scale uctuations in the presence of mean
shear or in transitional ows. C
s
=0.1 has been found to yield the best
results for a wide range of ows, and is the default value in FLUENT.
RNG-Based Subgrid-Scale Model
Renormalization group (RNG) theory can be used to derive a model for
the subgrid-scale eddy viscosity [271]. The RNG procedure results in an
eective subgrid viscosity,
e
= +
t
, given by
e
= [1 + H(x)]
1/3
(10.7-11)
H(x) is the Heaviside function:
H(x) =
_
x, x > 0
0, x 0
(10.7-12)
where
Modeling Turbulence
x =

2
s
3
C (10.7-13)
and
s
= (C
rng
V
1/3
)
2
_
2S
ij
S
ij
(10.7-14)
where V is the volume of the computational cell. The theory gives C
rng
=
0.157 and C = 100.
In highly turbulent regions of the ow (
t
),
e

s
, and the
RNG-based subgrid-scale model reduces to the Smagorinsky-Lilly model
with a dierent model constant. In low-Reynolds-number regions of
the ow, the argument of the ramp function becomes negative and the
eective viscosity recovers molecular viscosity. This enables the RNG-
based subgrid-scale eddy viscosity to model the low-Reynolds-number
eects encountered in transitional ows and near-wall regions.
10.7.3 Boundary Conditions for the LES Model
The stochastic components of the ow at the velocity-specied inlet
boundaries are accounted for by superposing random perturbations on
individual velocity components as
u
i
=< u
i
> +I |u| (10.7-15)
where I is the intensity of the uctuation, is a Gaussian random num-
ber satisfying = 0, and
_
= 1.
When the mesh is ne enough to resolve the laminar sublayer, the wall
shear stress is obtained from the laminar stress-strain relationship:
u
u
=
u
(10.7-16)
10.8 Near-Wall Treatments for Wall-Bounded Turbulent Flows
If the mesh is too coarse to resolve the laminar sublayer, it is assumed
that the centroid of the wall-adjacent cell falls within the logarithmic
region of the boundary layer, and the law-of-the-wall is employed:
u
u
=
1
ln E
_
u
_
(10.7-17)
where is the von Karm an constant and E = 9.793.
10.8 Near-Wall Treatments for Wall-Bounded Turbulent
Flows
10.8.1 Overview
Turbulent ows are signicantly aected by the presence of walls. Ob-
viously, the mean velocity eld is aected through the no-slip condition
that has to be satised at the wall. However, the turbulence is also
changed by the presence of the wall in non-trivial ways. Very close to
the wall, viscous damping reduces the tangential velocity uctuations,
while kinematic blocking reduces the normal uctuations. Toward the
outer part of the near-wall region, however, the turbulence is rapidly
augmented by the production of turbulence kinetic energy due to the
large gradients in mean velocity.
The near-wall modeling signicantly impacts the delity of numerical so-
lutions, inasmuch as walls are the main source of mean vorticity and tur-
bulence. After all, it is in the near-wall region that the solution variables
have large gradients, and the momentum and other scalar transports oc-
cur most vigorously. Therefore, accurate representation of the ow in
the near-wall region determines successful predictions of wall-bounded
turbulent ows.
The k- models, the RSM, and the LES model are primarily valid for
turbulent core ows (i.e., the ow in the regions somewhat far from
walls). Consideration therefore needs to be given as to how to make
these models suitable for wall-bounded ows. The Spalart-Allmaras and
k- models were designed to be applied throughout the boundary layer,
provided that the near-wall mesh resolution is sucient.
Modeling Turbulence
Numerous experiments have shown that the near-wall region can be
largely subdivided into three layers. In the innermost layer, called the
viscous sublayer, the ow is almost laminar, and the (molecular) vis-
cosity plays a dominant role in momentum and heat or mass transfer. In
the outer layer, called the fully-turbulent layer, turbulence plays a ma-
jor role. Finally, there is an interim region between the viscous sublayer
and the fully turbulent layer where the eects of molecular viscosity and
turbulence are equally important. Figure 10.8.1 illustrates these subdi-
visions of the near-wall region, plotted in semi-log coordinates.
U/U = U y/

U
/
U
2.5 ln(U y/ ) + 5.45 U/U =

viscous sublayer
buffer layer
or
blending
region
fully turbulent region
or
log-law region
y
+
y
+
5 60
ln U y/
outer layer
Upper limit
depends on
Reynolds no.
inner layer
Figure 10.8.1: Subdivisions of the Near-Wall Region
Wall Functions vs. Near-Wall Model
Traditionally, there are two approaches to modeling the near-wall region.
In one approach, the viscosity-aected inner region (viscous sublayer
and buer layer) is not resolved. Instead, semi-empirical formulas called
wall functions are used to bridge the viscosity-aected region between
the wall and the fully-turbulent region. The use of wall functions obviates
the need to modify the turbulence models to account for the presence of
the wall.
In another approach, the turbulence models are modied to enable the
viscosity-aected region to be resolved with a mesh all the way to the
wall, including the viscous sublayer. For purposes of discussion, this will
be termed the near-wall modeling approach. These two approaches
are depicted schematically in Figure 10.8.2.
Wall Function Approach Near-Wall Model Approach
buffer &
sublayer
t
u
r
b
u
l
e
n
t

c
o
r
e
wall functions.
The viscosity-affected region is not The near-wall region is resolved
used.
High-Re turbulence models can be
?
resolved, instead is bridged by the all the way down to the wall.
throughout the near-wall region.
The turbulence models ought to be valid
Figure 10.8.2: Near-Wall Treatments in FLUENT
In most high-Reynolds-number ows, the wall function approach sub-
stantially saves computational resources, because the viscosity-aected
near-wall region, in which the solution variables change most rapidly,
does not need to be resolved. The wall function approach is popular be-
cause it is economical, robust, and reasonably accurate. It is a practical
option for the near-wall treatments for industrial ow simulations.
The wall function approach, however, is inadequate in situations where
the low-Reynolds-number eects are pervasive in the ow domain in
question, and the hypotheses underlying the wall functions cease to be
valid. Such situations require near-wall models that are valid in the
viscosity-aected region and accordingly integrable all the way to the
wall.
Modeling Turbulence
FLUENT provides both the wall function approach and the near-wall
modeling approach.
Near-Wall Treatments for the Spalart-Allmaras, k-, and LES
Models
See Sections 10.3.6, 10.5.1, and 10.7.3, respectively, for a description of
the near-wall treatments applied by the Spalart-Allmaras, k-, and LES
models.
10.8.2 Wall Functions
Wall functions are a collection of semi-empirical formulas and functions
that in eect bridge or link the solution variables at the near-wall
cells and the corresponding quantities on the wall. The wall functions
comprise
laws-of-the-wall for mean velocity and temperature (or other scalars)
formulas for near-wall turbulent quantities
FLUENT oers two choices of wall function approaches:
standard wall functions
non-equilibrium wall functions
Standard Wall Functions
The standard wall functions in FLUENT are based on the proposal of
Launder and Spalding [129], and have been most widely used for indus-
trial ows. They are provided as a default option in FLUENT.
Momentum
The law-of-the-wall for mean velocity yields
U
=
1
ln(Ey
) (10.8-1)
where
U

U
P
C
1/4
k
1/2
P
w
/
(10.8-2)
y

C
1/4
k
1/2
P
y
P
(10.8-3)
and = von Karman constant (= 0.42)
E = empirical constant (= 9.81)
U
P
= mean velocity of the uid at point P
k
P
= turbulence kinetic energy at point P
y
P
= distance from point P to the wall
= dynamic viscosity of the uid
The logarithmic law for mean velocity is known to be valid for y
>
about 30 to 60. In FLUENT, the log-law is employed when y
> 11.225.
When the mesh is such that y
< 11.225 at the wall-adjacent cells, FLU-

ENT applies the laminar stress-strain relationship that can be written
as
U
= y
(10.8-4)
It should be noted that, in FLUENT, the laws-of-the-wall for mean ve-
locity and temperature are based on the wall unit, y
, rather than y
+
( u
y/). These quantities are approximately equal in equilibrium

turbulent boundary layers.
Energy
Reynolds analogy between momentum and energy transport gives a sim-
ilar logarithmic law for mean temperature. As in the law-of-the-wall for
Modeling Turbulence
mean velocity, the law-of-the-wall for temperature employed in FLUENT
comprises the following two dierent laws:
linear law for the thermal conduction sublayer where conduction is
important
logarithmic law for the turbulent region where eects of turbulence
dominate conduction
The thickness of the thermal conduction layer is, in general, dierent
from the thickness of the (momentum) viscous sublayer, and changes
from uid to uid. For example, the thickness of the thermal sublayer for
a high-Prandtl-number uid (e.g., oil) is much less than its momentum
sublayer thickness. For uids of low Prandtl numbers (e.g., liquid metal),
on the contrary, it is much larger than the momentum sublayer thickness.
In highly compressible ows, the temperature distribution in the near-
wall region can be signicantly dierent from that of low subsonic ows,
due to the heating by viscous dissipation. In FLUENT, the temperature
wall functions include the contribution from the viscous heating [246].
The law-of-the-wall implemented in FLUENT has the following composite
form:
T

(T
w
T
P
) c
p
C
1/4
k
1/2
P
q
=
_
_
Pr y
+
1
2
Pr
C
1/4
k
1/2
P
q
U
2
P
(y
< y
T
)
Pr
t
_
1
ln(Ey
) + P
_
+
1
2
C
1/4
k
1/2
P
q
_
Pr
t
U
2
P
+ (Pr Pr
t
)U
2
c
_
(y
> y
T
)
(10.8-5)
where P is computed by using the formula given by Jayatilleke [104]:
P = 9.24
_
_

t
_
3/4
1
_
_
1 + 0.28e
0.007/t
_
(10.8-6)
and
k
f
= thermal conductivity of uid
= density of uid
c
p
= specic heat of uid
q = wall heat ux
T
P
= temperature at the cell adjacent to wall
T
w
= temperature at the wall
Pr = molecular Prandtl number (c
p
/k
f
)
Pr
t
= turbulent Prandtl number (0.85 at the wall)
A = 26 (Van Driest constant)
= 0.4187 (von Karman constant)
E = 9.793 (wall function constant)
U
c
= mean velocity magnitude at y
= y
T
Note that, for the segregated solver, the terms
1
2
Pr
C
1/4
k
1/2
P
q
U
2
P
and
1
2
C
1/4
k
1/2
P
q
_
Pr
t
U
2
P
+ (Pr Pr
t
)U
2
c
_
will be included in Equation 10.8-5 only for compressible ow calcula-
tions.
The non-dimensional thermal sublayer thickness, y
T
, in Equation 10.8-5
is computed as the y
value at which the linear law and the logarithmic

law intersect, given the molecular Prandtl number of the uid being
modeled.
The procedure of applying the law-of-the-wall for temperature is as fol-
lows. Once the physical properties of the uid being modeled are speci-
ed, its molecular Prandtl number is computed. Then, given the molec-
ular Prandtl number, the thermal sublayer thickness, y
T
, is computed
from the intersection of the linear and logarithmic proles, and stored.
During the iteration, depending on the y
value at the near-wall cell,

either the linear or the logarithmic prole in Equation 10.8-5 is applied
to compute the wall temperature T
w
or heat ux q (depending on the
type of the thermal boundary conditions).
Modeling Turbulence
Species
When using wall functions for species transport, FLUENT assumes that
species transport behaves analogously to heat transfer. Similarly to
Equation 10.8-5, the law-of-the-wall for species can be expressed for con-
stant property ow with no viscous dissipation as
Y

(Y
i,w
Y
i
) C
1/4
k
1/2
P
J
i,w
=
_
Sc y
(y
< y
c
)
Sc
t
_
1
ln(Ey
) + P
c
_
(y
> y
c
)
(10.8-7)
where Y
i
is the local species mass fraction, Sc and Sc
t
are molecular and
turbulent Schmidt numbers, and J
i,w
is the diusion ux of species i
at the wall. Note that P
c
and y
c
are calculated in a similar way as P
and y
T
, with the dierence being that the Prandtl numbers are always
replaced by the corresponding Schmidt numbers.
Turbulence
In the k- models and in the RSM (if the option to obtain wall boundary
conditions from the k equation is enabled), the k equation is solved in the
whole domain including the wall-adjacent cells. The boundary condition
for k imposed at the wall is
k
n
= 0 (10.8-8)
where n is the local coordinate normal to the wall.
The production of kinetic energy, G
k
, and its dissipation rate, , at the
wall-adjacent cells, which are the source terms in the k equation, are
computed on the basis of the local equilibrium hypothesis. Under this
assumption, the production of k and its dissipation rate are assumed to
be equal in the wall-adjacent control volume.
Thus, the production of k is computed from
G
k

w
U
y
=
w
w
C
1/4
k
1/2
P
y
P
(10.8-9)
and is computed from
P
=
C
3/4
k
3/2
P
y
P
(10.8-10)
The equation is not solved at the wall-adjacent cells, but instead is
computed using Equation 10.8-10.
Note that, as shown here, the wall boundary conditions for the solution
variables, including mean velocity, temperature, species concentration,
k, and , are all taken care of by the wall functions. Therefore, you do
not need to be concerned about the boundary conditions at the walls.
The standard wall functions described so far are provided as a default
option in FLUENT. The standard wall functions work reasonably well for
a broad range of wall-bounded ows. However, they tend to become less
reliable when the ow situations depart too much from the ideal condi-
tions that are assumed in their derivation. Among others, the constant-
shear and local equilibrium hypotheses are the ones that most restrict
the universality of the standard wall functions. Accordingly, when the
near-wall ows are subjected to severe pressure gradients, and when the
ows are in strong non-equilibrium, the quality of the predictions is likely
to be compromised.
The non-equilibrium wall functions oered as an additional option can
improve the results in such situations.
Non-Equilibrium Wall Functions
In addition to the standard wall function described above (which is
the default near-wall treatment) a two-layer-based, non-equilibrium wall
function [115] is also available. The key elements in the non-equilibrium
wall functions are as follows:
Modeling Turbulence
Launder and Spaldings log-law for mean velocity is sensitized to
pressure-gradient eects.
The two-layer-based concept is adopted to compute the budget of
turbulence kinetic energy (G
k
,) in the wall-neighboring cells.
The law-of-the-wall for mean temperature or species mass fraction re-
mains the same as in the standard wall function described above.
The log-law for mean velocity sensitized to pressure gradients is
UC
1/4
k
1/2
w
/
=
1
ln
_
E
C
1/4
k
1/2
y
_
(10.8-11)
where
U = U
1
2
dp
dx
_
y
v
k
ln
_
y
y
v
_
+
y y
v
k
+
y
2
v
_
(10.8-12)
and y
v
is the physical viscous sublayer thickness, and is computed from
y
v

y
v
C
1/4
k
1/2
P
(10.8-13)
where y
v
= 11.225.
The non-equilibrium wall function employs the two-layer concept in com-
puting the budget of turbulence kinetic energy at the wall-adjacent cells,
which is needed to solve the k equation at the wall-neighboring cells. The
wall-neighboring cells are assumed to consist of a viscous sublayer and
a fully turbulent layer. The following prole assumptions for turbulence
quantities are made:
t
=
_
0, y < y
v
w
, y > y
v
k =
_
_
_
_
y
yv
_
2
k
P
, y < y
v
k
P
, y > y
v
=
_
2k
y
2
, y < y
v
k
3/2
C
y
, y > y
v
(10.8-14)
where C
= C
3/4
, and y
v
is the dimensional thickness of the viscous
sublayer, dened in Equation 10.8-13.
Using these proles, the cell-averaged production of k, G
k
, and the cell-
averaged dissipation rate, , can be computed from the volume average
of G
k
and of the wall-adjacent cells. For quadrilateral and hexahedral
cells for which the volume average can be approximated with a depth-
average,
G
k

1
y
n
_
yn
0
t
U
y
dy
=
1
y
n
2
w
C
1/4
k
1/2
P
ln
_
y
n
y
v
_
(10.8-15)
and
=
1
y
n
_
yn
0
dy
1
y
n
_
2
y
v
+
k
1/2
P
C
ln
_
y
n
y
v
_
_
k
P
(10.8-16)
where y
n
is the height of the cell (y
n
= 2y
P
). For cells with other shapes
(e.g., triangular and tetrahedral grids), the appropriate volume averages
are used.
In Equations 10.8-15 and 10.8-16, the turbulence kinetic energy budget
for the wall-neighboring cells is eectively sensitized to the proportions
of the viscous sublayer and the fully turbulent layer, which varies widely
from cell to cell in highly non-equilibrium ows. It eectively relaxes the
local equilibrium assumption (production = dissipation) that is adopted
by the standard wall function in computing the budget of the turbulence
kinetic energy at wall-neighboring cells. Thus, the non-equilibrium wall
functions, in eect, partly account for non-equilibrium eects neglected
in the standard wall function.
Modeling Turbulence
Standard Wall Functions vs. Non-Equilibrium Wall Functions
Because of the capability to partly account for the eects of pressure
gradients and departure from equilibrium, the non-equilibrium wall func-
tions are recommended for use in complex ows involving separation,
reattachment, and impingement where the mean ow and turbulence
are subjected to severe pressure gradients and change rapidly. In such
ows, improvements can be obtained, particularly in the prediction of
wall shear (skin-friction coecient) and heat transfer (Nusselt or Stanton
number).
Limitations of the Wall Function Approach
The standard wall functions give reasonably accurate predictions for
the majority of high-Reynolds-number, wall-bounded ows. The non-
equilibrium wall functions further extend the applicability of the wall
function approach by including the eects of pressure gradient and strong
non-equilibrium. However, the wall function approach becomes less reli-
able when the ow conditions depart too much from the ideal conditions
underlying the wall functions. Examples are as follows:
Pervasive low-Reynolds-number or near-wall eects (e.g., ow
through a small gap or highly viscous, low-velocity uid ow)
Massive transpiration through the wall (blowing/suction)
Severe pressure gradients leading to boundary layer separations
Strong body forces (e.g., ow near rotating disks, buoyancy-driven
ows)
High three-dimensionality in the near-wall region (e.g., Ekman spi-
ral ow, strongly skewed 3D boundary layers)
If any of the items listed above is a prevailing feature of the ow you
are modeling, and if it is considered critically important to capture that
feature for the success of your simulation, you must employ the near-
wall modeling approach combined with adequate mesh resolution in the
near-wall region. FLUENT provides the enhanced wall treatment for such
situations. This approach can be used with the three k- models and the
RSM.
10.8.3 Enhanced Wall Treatment
Enhanced wall treatment is a near-wall modeling method that com-
bines a two-layer model with enhanced wall functions. If the near-wall
mesh is ne enough to be able to resolve the laminar sublayer (typically
y
+
1), then the enhanced wall treatment will be identical to the tradi-
tional two-layer zonal model (see below for details). However, the restric-
tion that the near-wall mesh must be suciently ne everywhere might
impose too large a computational requirement. Ideally, then, one would
like to have a near-wall formulation that can be used with coarse meshes
(usually referred to as wall-function meshes) as well as ne meshes (low-
Reynolds-number meshes). In addition, excessive error should not be
incurred for intermediate meshes that are too ne for the near-wall cell
centroid to lie in the fully turbulent region, but also too coarse to prop-
erly resolve the sublayer.
To achieve the goal of having a near-wall modeling approach that will
possess the accuracy of the standard two-layer approach for ne near-wall
meshes and that, at the same time, will not signicantly reduce accuracy
for wall-function meshes, FLUENT can combine the two-layer model with
enhanced wall functions, as described in the following sections.
Two-Lay er Model for Enhanced Wall Treatment
In FLUENTs near-wall model, the viscosity-aected near-wall region is
completely resolved all the way to the viscous sublayer. The two-layer
approach is an integral part of the enhanced wall treatment and is used
to specify both and the turbulent viscosity in the near-wall cells. In
this approach, the whole domain is subdivided into a viscosity-aected
region and a fully-turbulent region. The demarcation of the two regions
is determined by a wall-distance-based, turbulent Reynolds number, Re
y
,
dened as
Modeling Turbulence
Re
y

y
(10.8-17)
where y is the normal distance from the wall at the cell centers. In
FLUENT, y is interpreted as the distance to the nearest wall:
y min
rww
r r
w
(10.8-18)
where r is the position vector at the eld point, and r
w
is the position
vector on the wall boundary.
w
is the union of all the wall boundaries
involved. This interpretation allows y to be uniquely dened in ow do-
mains of complex shape involving multiple walls. Furthermore, y dened
in this way is independent of the mesh topology used, and is denable
even on unstructured meshes.
In the fully turbulent region (Re
y
> Re
y
; Re
y
= 200), the k- models or
the RSM (described in Sections 10.4 and 10.6) are employed.
In the viscosity-aected near-wall region (Re
y
< Re
y
), the one-equation
model of Wolfstein [269] is employed. In the one-equation model, the
momentum equations and the k equation are retained as described in
Sections 10.4 and 10.6. However, the turbulent viscosity,
t
, is computed
from
t,2layer
= C
k (10.8-19)
where the length scale that appears in Equation 10.8-19 is computed
from [34]
= yc
_
1 e
Rey/A
_
(10.8-20)
The two-layer formulation for turbulent viscosity described above is used
as a part of the enhanced wall treatment, in which the two-layer denition
is smoothly blended with the high-Reynolds-number
t
denition from
the outer region, as proposed by Jongen [106]:
t,enh
=
t
+ (1
)
t,2layer
(10.8-21)
where
t
is the high-Reynolds-number denition as described in Sec-
tion 10.4 or 10.6 for the k- models or the RSM. A blending function,
, is dened in such a way that it is equal to unity far from walls and
is zero very near to walls. The blending function chosen is
=
1
2
_
1 + tanh
_
Re
y
Re
y
A
__
(10.8-22)
The constant A determines the width of the blending function. By den-
ing a width such that the value of
will be within 1% of its far-eld

value given a variation of Re
y
, the result is
A =
|Re
y
|
tanh(0.98)
(10.8-23)
Typically, Re
y
would be assigned a value that is between 5% and 20%
of Re
y
. The main purpose of the blending function
is to prevent
solution convergence from being impeded when the k- solution in the
outer layer does not match with the two-layer formulation.
The eld is computed from
=
k
3/2
(10.8-24)
The length scales that appear in Equation 10.8-24 are again computed
from Chen and Patel [34]:
= yc
_
1 e
Rey/A
_
(10.8-25)
If the whole ow domain is inside the viscosity-aected region
(Re
y
< 200), is not obtained by solving the transport equation; it
is instead obtained algebraically from Equation 10.8-24. FLUENT uses
Modeling Turbulence
a procedure for the specication that is similar to the
t
blending in
order to ensure a smooth transition between the algebraically-specied
in the inner region and the obtained from solution of the transport
equation in the outer region.
The constants in the length scale formulas, Equations 10.8-20 and 10.8-25,
are taken from [34]:
c
= C
3/4
, A
= 70, A
= 2c
(10.8-26)
Enhanced Wall Functions
To have a method that can extend its applicability throughout the near-
wall region (i.e., laminar sublayer, buer region, and fully-turbulent
outer region) it is necessary to formulate the law-of-the wall as a sin-
gle wall law for the entire wall region. FLUENT achieves this by blend-
ing linear (laminar) and logarithmic (turbulent) laws-of-the-wall using a
function suggested by Kader [108]:
u
+
= e
u
+
lam
+ e
1
u
+
turb
(10.8-27)
where the blending function is given by:
=
a(y
+
)
4
1 + by
+
(10.8-28)
c = exp
_
E
E
1.0
_
(10.8-29)
a = 0.01c (10.8-30)
b =
5
c
(10.8-31)
Similarly, the general equation for the derivative
du
+
dy
+
is
du
+
dy
+
= e
du
+
lam
dy
+
+ e
1
du
+
turb
dy
+
(10.8-32)
This approach allows the fully turbulent law to be easily modied and
extended to take into account other eects such as pressure gradients or
variable properties. This formula also guarantees the correct asymptotic
behavior for large and small values of y
+
and reasonable representation
of velocity proles in the cases where y
+
falls inside the wall buer region
(3 < y
+
< 10).
The enhanced wall functions were developed by smoothly blending an
enhanced turbulent wall law with the laminar wall law. The enhanced
turbulent law-of-the-wall for compressible ow with heat transfer and
pressure gradients has been derived by combining the approaches of
White and Cristoph [266] and Huang et al. [95]:
du
+
turb
dy
+
=
1
y
+
_
S
(1 u
+
(u
+
)
2
)
_
1/2
(10.8-33)
where
S
=
_
1 + y
+
for y
+
< y
+
s
1 + y
+
s
for y
+
y
+
s
(10.8-34)
and

w
w
u
dp
dx
=

2
(u
)
3
dp
dx
(10.8-35)

t
q
w
u
c
p
w
T
w
=

t
q
w
c
p
u
T
w
(10.8-36)

t
(u
)
2
2c
p
T
w
(10.8-37)
where y
+
s
is the location at which the log-law slope will remain xed. By
default, y
+
s
= 60. The coecient in Equation 10.8-33 represents the
inuences of pressure gradients while the coecients and represent
thermal eects. Equation 10.8-33 is an ordinary dierential equation and
FLUENT will provide an appropriate analytical solution. If , , and
all equal 0, an analytical solution would lead to the classical turbulent
logarithmic law-of-the-wall.
The laminar law-of-the-wall is determined from the following expression:
Modeling Turbulence
du
+
lam
dy
+
= 1 + y
+
(10.8-38)
Note that the above expression only includes eects of pressure gradients
through , while the eects of variable properties due to heat transfer
and compressibility on the laminar wall law are neglected. These eects
are neglected because they are thought to be of minor importance when
they occur close to the wall. Integration of Equation 10.8-38 results in
u
+
lam
= y
+
_
1 +

2
y
+
_
(10.8-39)
Enhanced thermal wall functions follow the same approach developed
for the prole of u
+
. The unied wall thermal formulation blends the
laminar and logarithmic proles according to the method of Kader [108]:
T
+
= e
T
+
lam
+ e
1
T
+
turb
(10.8-40)
where
=
a(Pr y
+
)
4
1 + bPr
3
y
+
(10.8-41)
where Pr is the molecular Prandtl number, and the coecients a and b
are dened as in Equations 10.8-30 and 10.8-31. Apart from the above
formulation for T
+
, enhanced thermal wall functions follow the same
logic as previously described for standard thermal wall functions (see
Section 10.8.2). A similar procedure is also used for species wall functions
when the enhanced wall treatment is used. See Section 10.8.2 for details
about species wall functions.
The boundary condition for turbulence kinetic energy is the same as for
standard wall functions (Equation 10.8-8). However, the production of
turbulence kinetic energy G
k
is computed using the velocity gradients
that are consistent with the enhanced law-of-the-wall (Equations 10.8-27
and 10.8-32), ensuring a formulation that is valid throughout the near-
wall region.
10.9 Grid Considerations for Turbulent Flow Simulations
Successful computations of turbulent ows require some consideration
during the mesh generation. Since turbulence (through the spatially-
varying eective viscosity) plays a dominant role in the transport of mean
momentum and other scalars for the majority of complex turbulent ows,
you must ascertain that turbulence quantities are properly resolved, if
high accuracy is required. Due to the strong interaction of the mean
ow and turbulence, the numerical results for turbulent ows tend to be
more susceptible to grid dependency than those for laminar ows.
It is therefore recommended that you resolve, with suciently ne meshes,
the regions where the mean ow changes rapidly and there are shear lay-
ers with a large mean rate of strain.
You can check the near-wall mesh by displaying or plotting the values of
y
+
, y
, and Re
y
, which are all available in the postprocessing panels. It
should be remembered that y
+
, y
, and Re
y
are not xed, geometrical
quantities. They are all solution-dependent. For example, when you
double the mesh (thereby halving the wall distance), the new y
+
does
not necessarily become half of the y
+
for the original mesh.
For the mesh in the near-wall region, dierent strategies must be used
depending on which near-wall option you are using. In Sections 10.9.1
and 10.9.2 are general guidelines for the near-wall mesh.
10.9.1 Near-Wall Mesh Guidelines for Wall Functions
The distance from the wall at the wall-adjacent cells must be determined
by considering the range over which the log-law is valid. The distance
is usually measured in the wall unit, y
+
( u
y/), or y
. Note that
y
+
and y
have comparable values when the rst cell is placed in the

log-layer.
It is known that the log-law is valid for y
+
> 30 to 60.
Although FLUENT employs the linear (laminar) law when y
+
<
11.225, using an excessively ne mesh near the walls should be
avoided, because the wall functions cease to be valid in the viscous
sublayer.
Modeling Turbulence
The upper bound of the log-layer depends on, among others, pres-
sure gradients and Reynolds number. As the Reynolds number
increases, the upper bound tends to also increase. y
+
values that
are too large are not desirable, because the wake component be-
comes substantially large above the log-layer.
A y
+
value close to the lower bound (y
+
30) is most desirable.
Using excessive stretching in the direction normal to the wall should
be avoided.
It is important to have at least a few cells inside the boundary
layer.
10.9.2 Near-Wall Mesh Guidelines for the Enhanced Wall
Treatment
Although the enhanced wall treatment is designed to extend the valid-
ity of near-wall modeling beyond the viscous sublayer, it is still recom-
mended that you construct a mesh that will fully resolve the viscosity-
aected near-wall region. In such a case, the two-layer component of
the enhanced wall treatment will be dominant and the following mesh
requirements are recommended (note that, here, the mesh requirements
are in terms of y
+
, not y
):
When the enhanced wall treatment is employed with the inten-
tion of resolving the laminar sublayer, y
+
at the wall-adjacent cell
should be on the order of y
+
= 1. However, a higher y
+
is ac-
ceptable as long as it is well inside the viscous sublayer (y
+
< 4 to
5).
You should have at least 10 cells within the viscosity-aected near-
wall region (Re
y
< 200) to be able to resolve the mean velocity
and turbulent quantities in that region.
10.9.3 Near-Wall Mesh Guidelines for the Spalart-Allmaras
Model
The Spalart-Allmaras model in its complete implementation is a low-
Reynolds-number model. This means that it is designed to be used with
meshes that properly resolve the viscous-aected region, and damping
functions have been built into the model in order to properly attenuate
the turbulent viscosity in the viscous sublayer. Therefore, to obtain the
full benet of the Spalart-Allmaras model, the near-wall mesh spacing
should be as described in Section 10.9.2 for the enhanced wall treatment.
However, as discussed in Section 10.3.6, the boundary conditions for
the Spalart-Allmaras model have been implemented so that the model
will work on coarser meshes, such as would be appropriate for the wall
function approach. If you are using a coarse mesh, you should follow the
guidelines described in Section 10.9.1.
In summary, for best results with the Spalart-Allmaras model, you should
use either a very ne near-wall mesh spacing (on the order of y
+
= 1) or
a mesh spacing such that y
+
30.
10.9.4 Near-Wall Mesh Guidelines for the k- Models
Both k- models available in FLUENT are available as low-Reynolds-
number models as well as high-Reynolds-number models. If the Transi-
tional Flows option is enabled in the Viscous Model panel, low-Reynolds-
number variants will be used, and, in that case, mesh guidelines should
be the same as for the enhanced wall treatment. However, if this option
is not active, then the mesh guidelines should be the same as for the wall
functions.
10.9.5 Near-Wall Mesh Guidelines for Large Eddy Simulation
For the LES implementation in FLUENT, the wall boundary conditions
have been implemented using a law-of-the-wall approach as described in
Section 10.7.3. This means that there are no computational restrictions
on the near-wall mesh spacing. However, for best results, it might be
necessary to use a very ne near-wall mesh spacing (on the order of
y
+
= 1).
Modeling Turbulence
10.10 Problem Setup for Turbulent Flows
When your FLUENT model includes turbulence you need to activate the
relevant model and options, and supply turbulent boundary conditions.
These inputs are described in this section.
The procedure for setting up a turbulent ow problem is described below.
(Note that this procedure includes only those steps necessary for the
turbulence model itself; you will need to set up other models, boundary
conditions, etc. as usual.)
1. To activate the turbulence model, select Spalart-Allmaras, k-epsilon,
k-omega, Reynolds Stress, or (in 3D) Large Eddy Simulation under
Model in the Viscous Model panel (Figure 10.10.1).
Dene Models Viscous...
If you choose the k-epsilon model, select Standard, RNG, or Realiz-
able under k-epsilon Model. If you choose the k-omega model, select
Standard or SST under k-omega Model.
The Large Eddy Simulation model is available only for 3D cases. !
2. If the ow involves walls, and you are using one of the k- models
or the RSM, choose one of the following options for the Near-Wall
Treatment in the Viscous Model panel:
Standard Wall Functions
Non-Equilibrium Wall Functions
Enhanced Wall Treatment
These near-wall options are described in detail in Section 10.8. By
default, the standard wall function is enabled.
The near-wall treatment for the Spalart-Allmaras, k-, and LES
models is dened automatically, as described in Sections 10.3.6,
10.5.1, and 10.7.3, respectively.
3. Enable the appropriate turbulence modeling options in the Viscous
Model panel. See Section 10.10.1 for details.
Figure 10.10.1: The Viscous Model Panel
Modeling Turbulence
4. Specify the boundary conditions for the solution variables.
Dene Boundary Conditions...
See Section 10.10.2 for details.
5. Specify the initial guess for the solution variables.
Solve Initialize Initialize...
See Section 10.10.3 for details. Note that Reynolds stresses are
automatically initialized using k, and therefore need not be initial-
ized.
10.10.1 Turbulence Options
The various options available for the turbulence models are described
in detail in Sections 10.3 through 10.7. Instructions for activating these
options are provided here.
If you choose the Spalart-Allmaras model, the following options are avail-
able:
Vorticity-based production
Strain/vorticity-based production
Viscous heating (always activated for the coupled solvers)
If you choose the standard k- model or the realizable k- model, the
following options are available:
Inclusion of buoyancy eects on
If you choose the RNG k- model, the following options are available:
Dierential viscosity model
Swirl modication
If you choose the standard k- model, the following options are available:
Transitional ows
Shear ow corrections
If you choose the shear-stress transport k- model, the following options
are available:
Transitional ows
If you choose the RSM, the following options are available:
Wall reection eects on Reynolds stresses
Wall boundary conditions for the Reynolds stresses from the k
equation
Quadratic pressure-strain model
If you choose the enhanced wall treatment (available for the k- models
and the RSM), the following options are available:
Pressure gradient eects
Thermal eects
Modeling Turbulence
If you choose the LES model, the following options are available:
Smagorinsky-Lilly model for the subgrid-scale viscosity
RNG model for the subgrid-scale viscosity
It is also possible to modify the Model Constants, but this is not nec-
essary for most applications. See Sections 10.3 through 10.7 for details
about these constants. Note that C1-PS and C2-PS are the constants C
1
and C
2
in the linear pressure-strain approximation of Equations 10.6-5
and 10.6-6, and C1-PS and C2-PS are the constants C
1
and C
2
in
Equation 10.6-7. C1-SSG-PS, C1-SSG-PS, C2-SSG-PS, C3-SSG-PS, C3-
SSG-PS, C4-SSG-PS, and C5-SSG-PS are the constants C
1
, C
1
, C
2
, C
3
,
C
3
, C
4
, and C
5
in the quadratic pressure-strain approximation of Equa-
tion 10.6-16.
Including the Viscous Heating Eects
See Sections 11.2.1 and 11.2.2 for information on including viscous heat-
ing eects in your model.
Including Turbulence Generation Due to Buoy ancy
If you specify a non-zero gravity force (in the Operating Conditions panel),
and you are modeling a non-isothermal ow, the generation of turbulent
kinetic energy due to buoyancy (G
b
in Equation 10.4-1) is, by default, al-
ways included in the k equation. However, FLUENT does not, by default,
include the buoyancy eects on .
To include the buoyancy eects on , you must turn on the Full Buoyancy
Eects option under Options in the Viscous Model panel.
This option is available for the three k- models and for the RSM.
Vorticity - and Strain/Vorticity -Based Production
For the Spalart-Allmaras model, you can choose either Vorticity-Based
Production or Strain/Vorticity-Based Production under Spalart-Allmaras
Options in the Viscous Model panel. If you choose Vorticity-Based Pro-
duction, FLUENT will use Equation 10.3-8 to compute the value of the
deformation tensor S; if you choose Strain/Vorticity-Based Production, it
will use Equation 10.3-10.
(These options will not appear unless you have activated the Spalart-
Allmaras model.)
Dierential Viscosity Modication
In the RNG turbulence model in FLUENT, you have an option to use
a dierential formula for eective viscosity
e
(Equation 10.4-6) to
account for low-Reynolds-number eects. To enable this option, turn
on Dierential Viscosity Model under RNG Options in the Viscous Model
panel.
(This option will not appear unless you have activated the RNG k-
model.)
Swirl Modication
Once you choose the RNG model, the swirl modication takes eect,
by default, for all three-dimensional ows and axisymmetric ows with
swirl. The default swirl constant (
s
in Equation 10.4-8) is set to 0.05,
which works well for weakly to moderately swirling ows. However, for
strongly swirling ows, you may need to use a larger swirl constant.
In order to change the value of the swirl constant, you must rst turn
on the Swirl Dominated Flow option under RNG Options in the Viscous
Model panel. (This option will not appear unless you have activated the
RNG k- model.)
Once you turn on the Swirl Dominated Flow option, the swirl constant
s
is increased to 0.07. You can change its value in the Swirl Factor eld
under Model Constants.
Modeling Turbulence
Transitional Flows
If either of the k- models are used, you may enable a low-Reynolds-
number correction to the turbulent viscosity by enabling the Transitional
Flows option under k-omega Options in the Viscous Model panel. By
default, this option is not enabled, and the damping coecient (
in
Equation 10.5-6) is equal to 1.
Shear Flow Corrections
In the standard k- model, you also have the option of including correc-
tions to improve the accuracy in predicting free shear ows. The Shear
Flow Corrections option under k-omega Options is enabled by default in
the Viscous Model panel, as these corrections are included in the standard
k- model [267]. When this option is enabled, FLUENT will calculate f
and f
using Equations 10.5-16 and 10.5-24, respectively. If this option

is disabled, f
and f
will be set equal to 1.

Including Pressure Gradient Eects
If the enhanced wall treatment is used, you may include the eects of
pressure gradients by enabling the Pressure Gradient Eects option under
Enhanced Wall Treatment Options. When this option is enabled, FLUENT
will include the coecient in Equation 10.8-33.
Including Thermal Eects
If the enhanced wall treatment is used, you may include thermal eects
by enabling the Thermal Eects option under Enhanced Wall Treatment
Options. When this option is enabled, FLUENT will include the coe-
cient in Equation 10.8-33. will also be included in Equation 10.8-33
when the Thermal Eects option is enabled if the ideal gas law is selected
for the uid density in the Materials panel.
Including the Wall Reection Term
If the RSM is used with the default model for pressure strain, FLUENT
will, by default, include the wall-reection eects in the pressure-strain
term. That is, FLUENT will calculate
w
ij
using Equation 10.6-7 and
include it in Equation 10.6-4. Note that wall-reection eects are not
included if you have selected the quadratic pressure-strain model.
The empirical constants and the function f used in the calculation of !
w
ij
are calibrated for simple canonical ows such as channel ows and
at-plate boundary layers involving a single wall. If the ow involves
multiple walls and the wall has signicant curvature (e.g., an axisym-
metric pipe or curvilinear duct), the inclusion of the wall-reection term
in Equation 10.6-7 may not improve the accuracy of the RSM predic-
tions. In such cases, you can disable the wall-reection eects by turning
o the Wall Reection Eects under Reynolds-Stress Options in the Viscous
Model panel.
Solving the k Equation to Obtain Wall Boundary Conditions
In the RSM, FLUENT, by default, uses the explicit setting of boundary
conditions for the Reynolds stresses near the walls, with the values com-
puted with Equation 10.6-28. k is calculated by solving the k equation
obtained by summing Equation 10.6-1 for normal stresses. To disable this
option and use the wall boundary conditions given in Equation 10.6-29,
turn o Wall B.C. from k Equation under Reynolds-Stress Options in the
Viscous Model panel. (This option will not appear unless you have acti-
vated the RSM.)
Quadratic Pressure-Strain Model
To use the quadratic pressure-strain model described in Section 10.6.3,turn
on the Quadratic Pressure-Strain Model option under Reynolds-Stress Op-
tions in the Viscous Model panel. (This option will not appear unless you
have activated the RSM.) The following options are not available when
the Quadratic Pressure-Strain Model is enabled:
Wall Reection Eects under Reynolds-Stress Options
Enhanced Wall Treatment under Near-Wall Treatment
Modeling Turbulence
Subgrid-Scale Model
If you have selected the Large Eddy Simulation model, you will be able to
choose which of the two subgrid-scale models described in Section 10.7.2
is to be used. You can choose either the Smagorinsky-Lilly or the RNG
subgrid-scale model.
(These options will not appear unless you have activated the LES model.)
Customizing the Turbulent Viscosity
If you are using the Spalart-Allmaras, k-, k-, or LES model, a user-
dened function can be used to customize the turbulent viscosity. This
option will enable you to modify
t
in the case of the Spalart-Allmaras,
k-, and k- models, and incorporate completely new subgrid models in
the case of the LES model. See the separate UDF Manual for information
about user-dened functions.
In the Viscous Model panel, under User-Dened Functions, select the ap-
propriate user-dened function in the Turbulent Viscosity drop-down list.
10.10.2 Dening Turbulence Boundary Conditions
k- Models and k- Models
When you are modeling turbulent ows in FLUENT using one of the
k- models or one of the k- models, you must provide the boundary
conditions for k and (or k and ) in addition to other mean solution
variables. The boundary conditions for k and (or k and ) at the
walls are internally taken care of by FLUENT, which obviates the need
for your inputs. The boundary condition inputs for k and (or k and )
you must supply to FLUENT are the ones at inlet boundaries (velocity
inlet, pressure inlet, etc.). In many situations, it is important to specify
correct or realistic boundary conditions at the inlets, because the inlet
turbulence can signicantly aect the downstream ow.
See Section 6.2.2 for details about specifying the boundary conditions
for k and (or k and ) at the inlets.
You may want to include the eects of the wall roughness on selected
wall boundaries. In such cases, you can specify the roughness param-
eters (roughness height and roughness constant) in the panels for the
corresponding wall boundaries (see Section 6.13.1).
The Spalart-Allmaras Model
When you are modeling turbulent ows in FLUENT using the Spalart-
Allmaras model, you must provide the boundary conditions for in
addition to other mean solution variables. The boundary conditions for
at the walls are internally taken care of by FLUENT, which obviates
the need for your inputs. The boundary condition input for you must
supply to FLUENT is the one at inlet boundaries (velocity inlet, pressure
inlet, etc.). In many situations, it is important to specify correct or
realistic boundary conditions at the inlets, because the inlet turbulence
can signicantly aect the downstream ow.
See Section 6.2.2 for details about specifying the boundary condition for
at the inlets.
You may want to include the eects of the wall roughness on selected
wall boundaries. In such cases, you can specify the roughness param-
eters (roughness height and roughness constant) in the panels for the
corresponding wall boundaries (see Section 6.13.1).
Rey nolds Stress Model
The specication of turbulent boundary conditions for the RSM is the
same as for the other turbulence models for all boundaries except at
boundaries where ow enters the domain. Additional input methods are
available for these boundaries and are described here.
When you choose to use the RSM, the default inlet boundary condition
inputs required are identical to those required when the k- model is
active. You can input the turbulence quantities using any of the tur-
bulence specication methods described in Section 6.2.2. FLUENT then
uses the specied turbulence quantities to derive the Reynolds stresses
at the inlet from the assumption of isotropy of turbulence:
Modeling Turbulence
u
2
i
=
2
3
k (i = 1, 2, 3) (10.10-1)
u
i
u
j
= 0.0 (10.10-2)
where u
2
i
is the normal Reynolds stress component in each direction.
The boundary condition for is determined in the same manner as for
the k- turbulence models (see Section 6.2.2). To use this method, you
will select K or Turbulence Intensity as the Reynolds-Stress Specication
Method in the appropriate boundary condition panel.
Alternately, you can directly specify the Reynolds stresses by selecting
Reynolds-Stress Components as the Reynolds-Stress Specication Method
in the boundary condition panel. When this option is enabled, you
should input the Reynolds stresses directly.
You can set the Reynolds stresses by using constant values, prole func-
tions of coordinates (see Section 6.25), or user-dened functions (see the
separate UDF Manual).
Large Eddy Simulation Model
It is possible to specify the magnitude of random uctuations of the
velocity components at an inlet only if the velocity inlet boundary con-
dition is selected. In this case, you must specify a Turbulence Intensity
that determines the magnitude of the random perturbations on individ-
ual mean velocity components as described in Section 10.7.3. For all
boundary types other than velocity inlets, the boundary conditions for
LES remain the same as for laminar ows.
10.10.3 Providing an Initial Guess for k and (or k and )
For ows using one of the k- models, one of the k- models, or the RSM,
the converged solutions or (for unsteady calculations) the solutions after
a suciently long time has elapsed should be independent of the initial
values for k and (or k and ). For better convergence, however, it is
benecial to use a reasonable initial guess for k and (or k and ).
Figure 10.10.2: Specifying Inlet Boundary Conditions for the Reynolds
Stresses
Modeling Turbulence
In general, it is recommended that you start from a fully-developed state
of turbulence. When you use the enhanced wall treatment for the k-
models or the RSM, it is critically important to specify fully-developed
turbulence elds. Guidelines are provided below.
If you were able to specify reasonable boundary conditions at the
inlet, it may be a good idea to compute the initial values for k and
(or k and ) in the whole domain from these boundary values.
(See Section 22.13 for details.)
For more complex ows (e.g., ows with multiple inlets with dif-
ferent conditions) it may be better to specify the initial values in
terms of turbulence intensity. 510% is enough to represent fully-
developed turbulence. k can then be computed from the turbulence
intensity and the characteristic mean velocity magnitude of your
problem (k = 1.5(Iu
avg
)
2
).
You should specify an initial guess for so that the resulting eddy
viscosity (C
k
2
) is suciently large in comparison to the molecular

viscosity. In fully-developed turbulence, the turbulent viscosity is
roughly two orders of magnitude larger than the molecular viscos-
ity. From this, you can compute .
Note that, for the RSM, Reynolds stresses are initialized automatically
using Equations 10.10-1 and 10.10-2.
10.11 Solution Strategies for Turbulent Flow Simulations
Compared to laminar ows, simulations of turbulent ows are more chal-
lenging in many ways. For the Reynolds-averaged approach, additional
equations are solved for the turbulence quantities. Since the equations
for mean quantities and the turbulent quantities (
t
, k, , , or the
Reynolds stresses) are strongly coupled in a highly non-linear fashion,
it takes more computational eort to obtain a converged turbulent solu-
tion than to obtain a converged laminar solution. The LES model, while
embodying a simpler, algebraic model for the subgrid-scale viscosity, re-
quires a transient solution on a very ne mesh.
The delity of the results for turbulent ows is largely determined by the
turbulence model being used. Here are some guidelines that can enhance
the quality of your turbulent ow simulations.
10.11.1 Mesh Generation
The following are suggestions to follow when generating the mesh for use
in your turbulent ow simulation:
Picture in your mind the ow under consideration using your phys-
ical intuition or any data for a similar ow situation, and identify
the main ow features expected in the ow you want to model.
Generate a mesh that can resolve the major features that you ex-
pect.
If the ow is wall-bounded, and the wall is expected to signicantly
aect the ow, take additional care when generating the mesh. You
should avoid using a mesh that is too ne (for the wall function ap-
proach) or too coarse (for the enhanced wall treatment approach).
See Section 10.9 for details.
10.11.2 Accuracy
The suggestions below are provided to help you obtain better accuracy
in your results:
Modeling Turbulence
Use the turbulence model that is better suited for the salient fea-
tures you expect to see in the ow (see Section 10.2).
Because the mean quantities have larger gradients in turbulent
ows than in laminar ows, it is recommended that you use high-
order schemes for the convection terms. This is especially true if
you employ a triangular or tetrahedral mesh. Note that excessive
numerical diusion adversely aects the solution accuracy, even
with the most elaborate turbulence model.
In some ow situations involving inlet boundaries, the ow down-
stream of the inlet is dictated by the boundary conditions at the
inlet. In such cases, you should exercise care to make sure that
reasonably realistic boundary values are specied.
10.11.3 Convergence
The suggestions below are provided to help you enhance convergence for
turbulent ow calculations:
Starting with excessively crude initial guesses for mean and tur-
bulence quantities may cause the solution to diverge. A safe ap-
proach is to start your calculation using conservative (small) under-
relaxation parameters and (for the coupled solvers) a conservative
Courant number, and increase them gradually as the iterations
proceed and the solution begins to settle down.
It is also helpful for faster convergence to start with reasonable ini-
tial guesses for the k and (or k and ) elds. Particularly when
the enhanced wall treatment is used, it is important to start with
a suciently developed turbulence eld, as recommended in Sec-
tion 10.10.3, to avoid the need for an excessive number of iterations
to develop the turbulence eld.
When you are using the RNG k- model, an approach that might
help you achieve better convergence is to obtain a solution with
the standard k- model before switching to the RNG model. Due
to the additional non-linearities in the RNG model, lower under-
relaxation factors and (for the coupled solvers) a lower Courant
number might also be necessary.
Note that when you use the enhanced wall treatment, you may some-
times nd during the calculation that the residual for is reported to
be zero. This happens when your ow is such that Re
y
is less than 200
in the entire ow domain, and is obtained from the algebraic formula
(Equation 10.8-24) instead of from its transport equation.
10.11.4 RSM-Specic Solution Strategies
Using the RSM creates a high degree of coupling between the momentum
equations and the turbulent stresses in the ow, and thus the calculation
can be more prone to stability and convergence diculties than with the
k- models. When you use the RSM, therefore, you may need to adopt
special solution strategies in order to obtain a converged solution. The
following strategies are generally recommended:
Begin the calculations using the standard k- model. Turn on the
RSM and use the k- solution data as a starting point for the RSM
calculation.
Use low under-relaxation factors (0.2 to 0.3) and (for the coupled
solvers) a low Courant number for highly swirling ows or highly
complex ows. In these cases, you may need to reduce the under-
relaxation factors both for the velocities and for all of the stresses.
Instructions for setting these solution parameters are provided below. If
you are applying the RSM to prediction of a highly swirling ow, you
will want to consider the solution strategies discussed in Section 8.4 as
well.
Under-Relaxation of the Rey nolds Stresses
FLUENT applies under-relaxation to the Reynolds stresses. You can set
under-relaxation factors using the Solution Controls panel.
Modeling Turbulence
Solve Controls Solution...
The default settings of 0.5 are recommended for most cases. You may be
able to increase these settings and speed up the convergence when the
RSM solution begins to converge.
Disabling Calculation Updates of the Rey nolds Stresses
In some instances, you may wish to let the current Reynolds stress eld
remain xed, skipping the solution of the Reynolds transport equations
while solving the other transport equations. You can activate/deactivate
all Reynolds stress equations in the Solution Controls panel.
Residual Reporting for the RSM
When you use the RSM for turbulence, FLUENT reports the equation
residuals for the individual Reynolds stress transport equations. You
can apply the usual convergence criteria to the Reynolds stress residuals:
normalized residuals in the range of 10
3
usually indicate a practically-
converged solution. However, you may need to apply tighter convergence
criteria (below 10
4
) to ensure full convergence.
10.11.5 LES-Specic Solution Strategies
Large eddy simulation involves running a transient solution from some
initial condition, on an appropriately ne grid, using an appropriate time
step size. The solution must be run long enough to become independent
of the initial condition and to enable the statistics of the ow eld to be
determined.
The following are suggestions to follow when running a large eddy sim-
ulation:
1. Start by running a ow simulation assuming laminar ow or using
a simple Reynolds-averaged turbulence model such as standard k-
or Spalart-Allmaras. Since this is only an initial condition, you
need run only until the ow eld is somewhat converged. This step
is optional.
2. When you enable LES, FLUENT will automatically turn on the
unsteady solver option and choose the second-order implicit for-
mulation. You will need to set the appropriate time step size and
all the needed solution parameters. (See Section 22.15.1 for guide-
lines on setting solution parameters for transient calculations in
general.) Use the central-dierencing spatial discretization scheme
for all equations.
3. Run LES until the ow becomes statistically steady. The best way
to see if the ow is fully developed and statistically steady is to
monitor forces and solution variables (e.g., velocity components or
pressure) at selected locations in the ow.
4. Zero out the initial statistics using the solve/initialize/
init-flow-statistics text command. Before you restart the so-
lution, enable Data Sampling for Time Statistics in the Iterate panel,
as described in Section 22.15.1.
5. Continue until you get statistically stable data. The duration of
the simulation can be determined beforehand by estimating the
mean ow residence time in the solution domain (L/U, where L is
the characteristic length of the solution domain and U is a charac-
teristic mean ow velocity). The simulation should be run for at
least a few mean ow residence times.
Instructions for setting the solution parameters for LES are provided
below.
Temporal Discretization
FLUENT provides both rst-order and second-order temporal discretiza-
tions. For LES, the second-order discretization is recommended.
Dene Models Solver...
Spatial Discretization
Overly diusive schemes such as the rst-order upwind or power law
scheme should be avoided, because they may unduly damp out the energy
Modeling Turbulence
of the resolved eddies. The central-dierencing scheme is recommended
for all equations when you use the LES model.
10.12 Postprocessing for Turbulent Flows
FLUENT provides postprocessing options for displaying, plotting, and
reporting various turbulence quantities, which include the main solution
variables and other auxiliary quantities.
Turbulence quantities that can be reported for the k- models are as
follows:
Turbulent Kinetic Energy (k)
Turbulence Intensity
Turbulent Dissipation Rate (Epsilon)
Production of k
Turbulent Viscosity
Eective Viscosity
Turbulent Viscosity Ratio
Eective Thermal Conductivity
Eective Prandtl Number
Wall Yplus
Wall Ystar
Turbulent Reynolds Number (Re y) (only when the enhanced wall
treatment is used for the near-wall treatment)
Turbulence quantities that can be reported for the k- models are as
follows:
Specic Dissipation Rate (Omega)
Production of k
Turbulent Viscosity
Eective Viscosity
Wall Ystar
Wall Yplus
Turbulence quantities that can be reported for the Spalart-Allmaras
model are as follows:
Modied Turbulent Viscosity
Turbulent Viscosity
Eective Viscosity
Wall Yplus
Turbulence quantities that can be reported for the RSM are as follows:
Modeling Turbulence
UU Reynolds Stress
VV Reynolds Stress
WW Reynolds Stress
UV Reynolds Stress
VW Reynolds Stress
UW Reynolds Stress
Turbulent Dissipation Rate (Epsilon)
Production of k
Turbulent Viscosity
Eective Viscosity
Wall Yplus
Wall Ystar
Turbulent Reynolds Number (Re y)
Turbulence quantities that can be reported for the LES model are as
follows:
Subgrid Turbulent Kinetic Energy
Subgrid Turbulent Viscosity
Subgrid Eective Viscosity
Subgrid Turbulent Viscosity Ratio
Wall Yplus
All of these variables can be found in the Turbulence... category of the
variable selection drop-down list that appears in postprocessing panels.
See Chapter 27 for their denitions.
10.12.1 Custom Field Functions for Turbulence
In addition to the quantities listed above, you can dene your own tur-
bulence quantities using the Custom Field Function Calculator panel.
Dene Custom Field Functions...
The following functions may be useful:
Ratio of production of k to its dissipation (G
k
/)
Ratio of the mean ow to turbulent time scale, ( Sk/)
Reynolds stresses derived from the Boussinesq formula (e.g., uv =
t
u
y
)
10.12.2 Postprocessing LES Statistics
As described in Section 10.7, LES involves the solution of a transient
ow eld, but it is the mean ow quantities that are of most engineering
interest. If you turn on the Data Sampling for Time Statistics option
in the Iterate panel, FLUENT will gather data for time statistics while
performing a large eddy simulation. You can then view both the mean
and the root-mean-square (RMS) values in FLUENT. See Section 22.15.3
for details.
Modeling Turbulence
10.12.3 Troubleshooting
You can use the postprocessing options not only for the purpose of in-
terpreting your results but also for investigating any anomalies that may
appear in the solution. For instance, you may want to plot contours of
the k eld to check if there are any regions where k is erroneously large
or small. You should see a high k region in the region where the produc-
tion of k is large. You may want to display the turbulent viscosity ratio
eld in order to see whether or not turbulence takes full eect. Usu-
ally turbulent viscosity is at least two orders of magnitude larger than
molecular viscosity for fully-developed turbulent ows modeled using the
RANS approach (i.e., not using LES). You may also want to see whether
you are using a proper near-wall mesh for the enhanced wall treatment.
In this case, you can display lled contours of Re
y
(turbulent Reynolds
number) overlaid on the mesh.

Modeling Turubulence

Uploaded by

Document Information

Original Description:

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

Modeling Turubulence

Uploaded by

Copyright:

Available Formats

Chapter 10.

is the production of turbulent viscosity and Y

is the shear velocity, y is

are the turbulent Prandtl

are user-dened source

have the following default

are the inverse eective Prandtl numbers for

are user-dened source terms.

= 0.0845, derived using RNG theory. It is interesting to note

is very close to the empirically-determined value of

, are computed using

Term in the Equation

; no summation over and ) can be violated

variable by sensitizing it to the

is suggested by many modelers including Reynolds [191],

are the turbulent Prandtl numbers for

are user-dened source terms.

is no longer constant. It is computed from

is a function of the mean strain and rotation rates,

in Equation 10.4-17 can be shown to recover the standard

have been established to ensure that

RT) is the speed of sound.

represents the generation of

represent the eective diusivity of k and , respectively.

represent the dissipation of k and due to turbulence. All

are the turbulent Prandtl numbers for k and , respec-

damps the turbulent viscosity causing a low-Reynolds-

represents the generation of , calculated as described in

represent the eective diusivity of k and ,

represents the cross-diusion term, calculated as

are user-dened source terms.

are the turbulent Prandtl numbers for k and , respec-

is dened in Equation 10.5-6.

is the positive portion

represents the production of and is given by

is evaluated. In the standard k- model,

is dened as a piecewise function. For the

is a constant equal to 1. Thus,

represents the dissipation of , and is dened in a similar

are evaluated. In the standard

is a constant equal to 1. Thus,

= 0.09 and is the von Karm an constant (= 0.4187).

is a user-dened source term.

is the friction velocity dened by u

), implied here is then

2.5 ln(U y/ ) + 5.45 U/U =

< 11.225 at the wall-adjacent cells, FLU-

y/). These quantities are approximately equal in equilibrium

value at which the linear law and the logarithmic

value at the near-wall cell,

will be within 1% of its far-eld

have comparable values when the rst cell is placed in the

using Equations 10.5-16 and 10.5-24, respectively. If this option

will be set equal to 1.

) is suciently large in comparison to the molecular

You might also like