t
+ (v) = S
m
(8.21)
Equation 8.21 is the general form of the mass conservation equation and is valid for
incompressible as well as compressible ows. The source S
m
is the mass added to the
continuous phase from the dispersed second phase (e.g., due to vaporization of liquid
droplets) and any userdened sources.
For 2D axisymmetric geometries, the continuity equation is given by
t
+
x
(v
x
) +
r
(v
r
) +
v
r
r
= S
m
(8.22)
where x is the axial coordinate, r is the radial coordinate, v
x
is the axial velocity, and v
r
is the radial velocity.
Momentum Conservation Equations
Conservation of momentum in an inertial (nonaccelerating) reference frame is described
by [11]
t
(v) + (vv) = p + () + g +
F (8.23)
c Fluent Inc. January 28, 2003 83
Modeling Basic Fluid Flow
where p is the static pressure, is the stress tensor (described below), and g and
F are
the gravitational body force and external body forces (e.g., that arise from interaction
with the dispersed phase), respectively.
F also contains other modeldependent source
terms such as porousmedia and userdened sources.
The stress tensor is given by
=
_
(v +v
T
)
2
3
vI
_
(8.24)
where is the molecular viscosity, I is the unit tensor, and the second term on the right
hand side is the eect of volume dilation.
For 2D axisymmetric geometries, the axial and radial momentum conservation equations
are given by
t
(v
x
) +
1
r
x
(rv
x
v
x
) +
1
r
r
(rv
r
v
x
) =
p
x
+
1
r
x
_
r
_
2
v
x
x
2
3
( v)
__
+
1
r
r
_
r
_
v
x
r
+
v
r
x
__
+ F
x
(8.25)
and
t
(v
r
) +
1
r
x
(rv
x
v
r
) +
1
r
r
(rv
r
v
r
) =
p
r
+
1
r
x
_
r
_
v
r
x
+
v
x
r
__
+
1
r
r
_
r
_
2
v
r
r
2
3
( v)
__
2
v
r
r
2
+
2
3
r
( v) +
v
2
z
r
+ F
r
(8.26)
where
v =
v
x
x
+
v
r
r
+
v
r
r
(8.27)
and v
z
is the swirl velocity. (See Section 8.4 for information about modeling axisymmetric
swirl.)
84 c Fluent Inc. January 28, 2003
8.3 Periodic Flows
8.3 Periodic Flows
Periodic ow occurs when the physical geometry of interest and the expected pattern of
the ow/thermal solution have a periodically repeating nature. Two types of periodic ow
can be modeled in FLUENT. In the rst type, no pressure drop occurs across the periodic
planes. (Note to FLUENT 4 users: This type of periodic ow is modeled using a cyclic
boundary in FLUENT 4.) In the second type, a pressure drop occurs across translationally
periodic boundaries, resulting in fullydeveloped or streamwiseperiodic ow. (In
FLUENT 4, this type of periodic ow is modeled using a periodic boundary.)
This section discusses streamwiseperiodic ow. A description of nopressuredrop pe
riodic ow is provided in Section 6.15, and a description of streamwiseperiodic heat
transfer is provided in Section 11.4.
Information about streamwiseperiodic ow is presented in the following sections:
Section 8.3.1: Overview and Limitations
Section 8.3.2: Theory
Section 8.3.3: User Inputs for the Segregated Solver
Section 8.3.4: User Inputs for the Coupled Solvers
Section 8.3.5: Monitoring the Value of the Pressure Gradient
Section 8.3.6: Postprocessing for StreamwisePeriodic Flows
8.3.1 Overview and Limitations
Overview
FLUENT provides the ability to calculate streamwiseperiodicor fullydeveloped
uid ow. These ows are encountered in a variety of applications, including ows in
compact heat exchanger channels and ows across tube banks. In such ow congura
tions, the geometry varies in a repeating manner along the direction of the ow, leading
to a periodic fullydeveloped ow regime in which the ow pattern repeats in succes
sive cycles. Other examples of streamwiseperiodic ows include fullydeveloped ow in
pipes and ducts. These periodic conditions are achieved after a sucient entrance length,
which depends on the ow Reynolds number and geometric conguration.
Streamwiseperiodic ow conditions exist when the ow pattern repeats over some length
L, with a constant pressure drop across each repeating module along the streamwise
direction. Figure 8.3.1 depicts one example of a periodically repeating ow of this type
which has been modeled by including a single representative module.
c Fluent Inc. January 28, 2003 85
Modeling Basic Fluid Flow
Velocity Vectors Colored By Velocity Magnitude (m/s)
3.57e03
3.33e03
3.09e03
2.86e03
2.62e03
2.38e03
2.14e03
1.90e03
1.67e03
1.43e03
1.19e03
9.53e04
7.15e04
4.77e04
2.39e04
1.01e06
Figure 8.3.1: Example of Periodic Flow in a 2D Heat Exchanger Geometry
Constraints on the Use of StreamwisePeriodic Flow
The following constraints apply to modeling streamwiseperiodic ow:
The ow must be incompressible.
The geometry must be translationally periodic.
If one of the coupled solvers is used, you can specify only the pressure jump; for the
segregated solver, you can specify either the pressure jump or the mass ow rate.
No net mass addition through inlets/exits or extra source terms is allowed.
Species can be modeled only if inlets/exits (without net mass addition) are included
in the problem. Reacting ows are not permitted.
Discrete phase and multiphase modeling are not allowed.
8.3.2 Theory
Denition of the Periodic Velocity
The assumption of periodicity implies that the velocity components repeat themselves in
space as follows:
86 c Fluent Inc. January 28, 2003
8.3 Periodic Flows
u(r) = u(r +
L) = u(r + 2
L) =
v(r) = v(r +
L) = v(r + 2
L) = (8.31)
w(r) = w(r +
L) = w(r + 2
L) =
where r is the position vector and
L is the periodic length vector of the domain considered
(see Figure 8.3.2).
L
A B C
u
B
u
A
u
C
= =
v
B
v
A
v
C
= = p
B
p
A
p
C
=
p
B
 
p
B
p
A
p
C
= =
Figure 8.3.2: Example of a Periodic Geometry
Denition of the StreamwisePeriodic Pressure
For viscous ows, the pressure is not periodic in the sense of Equation 8.31. Instead,
the pressure drop between modules is periodic:
p = p(r) p(r +
L) = p(r +
L) p(r + 2
L) = (8.32)
If one of the coupled solvers is used, p is specied as a constant value. For the segregated
solver, the local pressure gradient can be decomposed into two parts: the gradient of a
periodic component, p(r), and the gradient of a linearlyvarying component,
L

L
:
p(r) =
L

L
+ p(r) (8.33)
where p(r) is the periodic pressure and r is the linearlyvarying component of the
pressure. The periodic pressure is the pressure left over after subtracting out the linearly
varying pressure. The linearlyvarying component of the pressure results in a force acting
on the uid in the momentum equations. Because the value of is not known a priori,
it must be iterated on until the mass ow rate that you have dened is achieved in the
computational model. This correction of occurs in the pressure correction step of the
SIMPLE, SIMPLEC, or PISO algorithm where the value of is updated based on the
dierence between the desired mass ow rate and the actual one. You have some control
over the number of subiterations used to update , as described in Section 8.3.3.
c Fluent Inc. January 28, 2003 87
Modeling Basic Fluid Flow
8.3.3 User Inputs for the Segregated Solver
If you are using the segregated solver, in order to calculate a spatially periodic ow eld
with a specied mass ow rate or pressure derivative, you must rst create a grid with
translationally periodic boundaries that are parallel to each other and equal in size. You
can specify translational periodicity in the Periodic panel, as described in Section 6.15.
(If you need to create periodic boundaries, see Section 5.7.5.)
Once the grid has been read into FLUENT, you will complete the following inputs in the
Periodicity Conditions panel (Figure 8.3.3):
Dene Periodic Conditions...
Figure 8.3.3: The Periodicity Conditions Panel
1. Select either the specied mass ow rate (Specify Mass Flow) option or the specied
pressure gradient (Specify Pressure Gradient) option. For most problems, the mass
ow rate across the periodic boundary will be a known quantity; for others, the
mass ow rate will be unknown, but the pressure gradient ( in Equation 8.33)
will be a known quantity.
2. Specify the mass ow rate and/or the pressure gradient ( in Equation 8.33):
If you selected the Specify Mass Flow option, enter the desired value for the
Mass Flow Rate. You can also specify an initial guess for the Pressure Gradient,
but this is not required.
For axisymmetric problems, the mass ow rate is per 2 radians. !
If you selected the Specify Pressure Gradient option, enter the desired value for
Pressure Gradient.
88 c Fluent Inc. January 28, 2003
8.3 Periodic Flows
3. Dene the ow direction by setting the X,Y,Z (or X,Y in 2D) point under Flow
Direction. The ow will move in the direction of the vector pointing from the
origin to the specied point. The direction vector must be parallel to the periodic
translation direction or its opposite.
4. If you chose in step 1 to specify the mass ow rate, set the parameters used for the
calculation of . These parameters are described in detail below.
After completing these inputs, you can solve the periodic velocity eld to convergence.
Setting Parameters for the Calculation of
If you choose to specify the mass ow rate, FLUENT will need to calculate the appropriate
value of the pressure gradient . You can control this calculation by specifying the
Relaxation Factor and the Number of Iterations, and by supplying an initial guess for .
All of these inputs are entered in the Periodicity Conditions panel.
The Number of Iterations sets the number of subiterations performed on the correction
of in the pressure correction equation. Because the value of is not known a priori,
it must be iterated on until the Mass Flow Rate that you have dened is achieved in
the computational model. This correction of occurs in the pressure correction step of
the SIMPLE, SIMPLEC, or PISO algorithm. A correction to the current value of is
calculated based on the dierence between the desired mass ow rate and the actual one.
The subiterations referred to here are performed within the pressure correction step to
improve the correction for before the pressure correction equation is solved for the
resulting pressure (and velocity) correction values. The default value of 2 subiterations
should suce in most problems, but can be increased to help speed convergence. The
Relaxation Factor is an underrelaxation factor that controls convergence of this iteration
process.
You can also speed up convergence of the periodic calculation by supplying an initial guess
for in the Pressure Gradient eld. Note that the current value of will be displayed in
this eld if you have performed any calculations. To update the Pressure Gradient eld
with the current value at any time, click on the Update button.
8.3.4 User Inputs for the Coupled Solvers
If you are using one of the coupled solvers, in order to calculate a spatially periodic ow
eld with a specied pressure jump, you must rst create a grid with translationally
periodic boundaries that are parallel to each other and equal in size. (If you need to
create periodic boundaries, see Section 5.7.5.)
c Fluent Inc. January 28, 2003 89
Modeling Basic Fluid Flow
Then, follow the steps below:
1. In the Periodic panel (Figure 8.3.4), which is opened from the Boundary Conditions
panel, indicate that the periodicity is Translational (the default).
Dene Boundary Conditions...
Figure 8.3.4: The Periodic Panel
2. Also in the Periodic panel, set the Periodic Pressure Jump (p in Equation 8.32).
After completing these inputs, you can solve the periodic velocity eld to convergence.
8.3.5 Monitoring the Value of the Pressure Gradient
If you have specied the mass ow rate, you can monitor the value of the pressure
gradient during the calculation using the Statistic Monitors panel. Select per/prgrad
as the variable to be monitored. See Section 24.16.2 for details about using this feature.
8.3.6 Postprocessing for StreamwisePeriodic Flows
For streamwiseperiodic ows, the velocity eld should be completely periodic. If a
coupled solver is used to compute the periodic ow, the pressure eld reported will be the
actual pressure p (which is not periodic). If the segregated solver is used, the pressure
eld reported will be the periodic pressure eld p(r) of Equation 8.33. Figure 8.3.5
displays the periodic pressure eld in the geometry of Figure 8.3.1.
If you specied a mass ow rate and had FLUENT calculate the pressure gradient, you
can check the pressure gradient in the streamwise direction () by looking at the current
value for Pressure Gradient in the Periodicity Conditions panel.
810 c Fluent Inc. January 28, 2003
8.4 Swirling and Rotating Flows
Contours of Static Pressure (pascal)
1.68e03
1.29e03
8.98e04
5.07e04
1.16e04
2.74e04
6.65e04
1.06e03
1.45e03
1.84e03
2.23e03
2.62e03
3.01e03
3.40e03
3.79e03
4.18e03
Figure 8.3.5: Periodic Pressure Field Predicted for Flow in a 2D Heat Exchanger Geom
etry
8.4 Swirling and Rotating Flows
Many important engineering ows involve swirl or rotation and FLUENT is wellequipped
to model such ows. Swirling ows are common in combustion, with swirl introduced in
burners and combustors in order to increase residence time and stabilize the ow pattern.
Rotating ows are also encountered in turbomachinery, mixing tanks, and a variety of
other applications.
Information about rotating and swirling ows is provided in the following subsections:
Section 8.4.1: Overview of Swirling and Rotating Flows
Section 8.4.2: Physics of Swirling and Rotating Flows
Section 8.4.3: Turbulence Modeling in Swirling Flows
Section 8.4.4: Grid Setup for Swirling and Rotating Flows
Section 8.4.5: Modeling Axisymmetric Flows with Swirl or Rotation
c Fluent Inc. January 28, 2003 811
Modeling Basic Fluid Flow
When you begin the analysis of a rotating or swirling ow, it is essential that you classify
your problem into one of the following ve categories of ow:
axisymmetric ows with swirl or rotation
fully threedimensional swirling or rotating ows
ows requiring a rotating reference frame
ows requiring multiple rotating reference frames or mixing planes
ows requiring sliding meshes
Modeling and solution procedures for the rst two categories are presented in this section.
The remaining three, which all involve moving zones, are discussed in Chapter 9.
8.4.1 Overview of Swirling and Rotating Flows
Axisymmetric Flows with Swirl or Rotation
Your problem may be axisymmetric with respect to geometry and ow conditions but
still include swirl or rotation. In this case, you can model the ow in 2D (i.e., solve
the axisymmetric problem) and include the prediction of the circumferential (or swirl)
velocity. It is important to note that while the assumption of axisymmetry implies
that there are no circumferential gradients in the ow, there may still be nonzero swirl
velocities.
Momentum Conservation Equation for Swirl Velocity
The tangential momentum equation for 2D swirling ows may be written as
t
(w)+
1
r
x
(ruw)+
1
r
r
(rvw) =
1
r
x
_
r
w
x
_
+
1
r
2
r
_
r
3
r
_
w
r
_
_
vw
r
(8.41)
where x is the axial coordinate, r is the radial coordinate, u is the axial velocity, v is the
radial velocity, and w is the swirl velocity.
ThreeDimensional Swirling Flows
When there are geometric changes and/or ow gradients in the circumferential direction,
your swirling ow prediction requires a threedimensional model. If you are planning
a 3D FLUENT model that includes swirl or rotation, you should be aware of the setup
constraints listed in Section 8.4.4. In addition, you may wish to consider simplications
812 c Fluent Inc. January 28, 2003
8.4 Swirling and Rotating Flows
to the problem which might reduce it to an equivalent axisymmetric problem, especially
for your initial modeling eort. Because of the complexity of swirling ows, an initial
2D study, in which you can quickly determine the eects of various modeling and design
choices, can be very benecial.
For 3D problems involving swirl or rotation, there are no special inputs required during !
the problem setup and no special solution procedures. Note, however, that you may want
to use the cylindrical coordinate system for dening velocityinlet boundary condition
inputs, as described in Section 6.4.1. Also, you may nd the gradual increase of the
rotational speed (set as a wall or inlet boundary condition) helpful during the solution
process. This is described for axisymmetric swirling ows in Section 8.4.5.
Flows Requiring a Rotating Reference Frame
If your ow involves a rotating boundary which moves through the uid (e.g., an impeller
blade or a grooved or notched surface), you will need to use a rotating reference frame to
model the problem. Such applications are described in detail in Section 9.2. If you have
more than one rotating boundary (e.g., several impellers in a row), you can use multiple
reference frames (described in Section 9.3) or mixing planes (described in Section 9.4).
8.4.2 Physics of Swirling and Rotating Flows
In swirling ows, conservation of angular momentum (rw or r
2
= constant) tends to
create a free vortex ow, in which the circumferential velocity, w, increases sharply as the
radius, r, decreases (with w nally decaying to zero near r = 0 as viscous forces begin
to dominate). A tornado is one example of a free vortex. Figure 8.4.1 depicts the radial
distribution of w in a typical free vortex.
axis
r
Figure 8.4.1: Typical Radial Distribution of w in a Free Vortex
It can be shown that for an ideal free vortex ow, the centrifugal forces created by the
circumferential motion are in equilibrium with the radial pressure gradient:
p
r
=
w
2
r
(8.42)
As the distribution of angular momentum in a nonideal vortex evolves, the form of this
radial pressure gradient also changes, driving radial and axial ows in response to the
highly nonuniform pressures that result. Thus, as you compute the distribution of swirl
in your FLUENT model, you will also notice changes in the static pressure distribution
c Fluent Inc. January 28, 2003 813
Modeling Basic Fluid Flow
and corresponding changes in the axial and radial ow velocities. It is this high degree
of coupling between the swirl and the pressure eld that makes the modeling of swirling
ows complex.
In ows that are driven by wall rotation, the motion of the wall tends to impart a forced
vortex motion to the uid, wherein w/r or is constant. An important characteristic
of such ows is the tendency of uid with high angular momentum (e.g., the ow near
the wall) to be ung radially outward (Figure 8.4.2). This is often referred to as radial
pumping, since the rotating wall is pumping the uid radially outward.
axis of rotation
Contours of Stream Function (kg/s)
7.69e03
6.92e03
6.15e03
5.38e03
4.62e03
3.85e03
3.08e03
2.31e03
1.54e03
7.69e04
0.00e+00
Figure 8.4.2: Stream Function Contours for Rotating Flow in a Cavity (Geometry of
Figure 8.4.3)
8.4.3 Turbulence Modeling in Swirling Flows
If you are modeling turbulent ow with a signicant amount of swirl (e.g., cyclone ows,
swirling jets), you should consider using one of FLUENTs advanced turbulence models:
the RNG k model, realizable k model, or Reynolds stress model. The appropriate
choice depends on the strength of the swirl, which can be gauged by the swirl number.
The swirl number is dened as the ratio of the axial ux of angular momentum to the
axial ux of axial momentum:
S =
_
rwv d
R
_
uv d
A
(8.43)
814 c Fluent Inc. January 28, 2003
8.4 Swirling and Rotating Flows
where
R is the hydraulic radius.
For ows with weak to moderate swirl (S < 0.5), both the RNG k model and the
realizable k model yield appreciable improvements over the standard k model. See
Sections 10.4.2, 10.4.3, and 10.10.1 for details about these models.
For highly swirling ows (S > 0.5), the Reynolds stress model (RSM) is strongly recom
mended. The eects of strong turbulence anisotropy can be modeled rigorously only by
the secondmoment closure adopted in the RSM. See Sections 10.6 and 10.10 for details
about this model.
For swirling ows encountered in devices such as cyclone separators and swirl combustors,
nearwall turbulence modeling is quite often a secondary issue at most. The delity of the
predictions in these cases is mainly determined by the accuracy of the turbulence model
in the core region. However, in cases where walls actively participate in the generation
of swirl (i.e., where the secondary ows and vortical ows are generated by pressure
gradients), nonequilibrium wall functions can often improve the predictions since they
use a law of the wall for mean velocity sensitized to pressure gradients. See Section 10.8
for additional details about nearwall treatments for turbulence.
8.4.4 Grid Setup for Swirling and Rotating Flows
CoordinateSystem Restrictions
Recall that for an axisymmetric problem, the axis of rotation must be the x axis and the
grid must lie on or above the y = 0 line.
Grid Sensitivity in Swirling and Rotating Flows
In addition to the setup constraint described above, you should be aware of the need
for sucient resolution in your grid when solving ows that include swirl or rotation.
Typically, rotating boundary layers may be very thin, and your FLUENT model will
require a very ne grid near a rotating wall. In addition, swirling ows will often involve
steep gradients in the circumferential velocity (e.g., near the centerline of a freevortex
type ow), and thus require a ne grid for accurate resolution.
8.4.5 Modeling Axisymmetric Flows with Swirl or Rotation
As discussed in Section 8.4.1, you can solve a 2D axisymmetric problem that includes
the prediction of the circumferential or swirl velocity. The assumption of axisymmetry
implies that there are no circumferential gradients in the ow, but that there may be
nonzero circumferential velocities. Examples of axisymmetric ows involving swirl or
rotation are depicted in Figures 8.4.3 and 8.4.4.
c Fluent Inc. January 28, 2003 815
Modeling Basic Fluid Flow
Region to
be modeled
Rotating Cover
x
y
Figure 8.4.3: Rotating Flow in a Cavity
Region to be modeled
t
+ (v) = S
m
(8.61)
Equation 8.61 is the general form of the mass conservation equation and is valid for
incompressible as well as compressible ows. The source S
m
is the mass added to the
continuous phase from the dispersed second phase (e.g., due to vaporization of liquid
droplets) and any userdened sources.
For 2D axisymmetric geometries, the continuity equation is given by
t
+
x
(v
x
) +
r
(v
r
) +
v
r
r
= S
m
(8.62)
where x is the axial coordinate, r is the radial coordinate, v
x
is the axial velocity, and v
r
is the radial velocity.
Momentum Conservation Equations
Conservation of momentum is described by
t
(v) + (vv) = p + g +
F (8.63)
where p is the static pressure and g and
F are the gravitational body force and external
body forces (e.g., forces that arise from interaction with the dispersed phase), respectively.
F also contains other modeldependent source terms such as porousmedia and user
dened sources.
For 2D axisymmetric geometries, the axial and radial momentum conservation equations
are given by
t
(v
x
) +
1
r
x
(rv
x
v
x
) +
1
r
r
(rv
r
v
x
) =
p
x
+ F
x
(8.64)
and
t
(v
r
) +
1
r
x
(rv
x
v
r
) +
1
r
r
(rv
r
v
r
) =
p
r
+ F
r
(8.65)
where
c Fluent Inc. January 28, 2003 829
Modeling Basic Fluid Flow
v =
v
x
x
+
v
r
r
+
v
r
r
(8.66)
Energy Conservation Equation
Conservation of energy is described by
t
(E) + (v(E + p)) =
_
_
j
h
j
J
j
_
_
+ S
h
(8.67)
8.6.2 Setting Up an Inviscid Flow Model
For inviscid ow problems, you will need to perform the following steps during the prob
lem setup procedure. (Only those steps relevant specically to the setup of inviscid ow
are listed here. You will need to set up the rest of the problem as usual.)
1. Activate the calculation of inviscid ow by selecting Inviscid in the Viscous Model
panel.
Dene Models Viscous...
2. Set boundary conditions and ow properties.
Dene Boundary Conditions...
Dene Materials...
3. Solve the problem and examine the results.
8.6.3 Solution Strategies for Inviscid Flows
Since inviscid ow problems will usually involve highspeed ow, you may have to reduce
the underrelaxation factors for momentum (if you are using the segregated solver) or
reduce the Courant number (if you are using the coupled solver), in order to get the
solution started. Once the ow is started and the residuals are monotonically decreasing,
you can start increasing the underrelaxation factors or Courant number back up to the
default values.
Modications to the underrelaxation factors and the Courant number can be made in
the Solution Controls panel.
Solve Controls Solution...
The solution strategies for compressible ows apply also to inviscid ows. See Sec
tion 8.5.5 for details.
830 c Fluent Inc. January 28, 2003
8.6 Inviscid Flows
8.6.4 Postprocessing for Inviscid Flows
If you are interested in the lift and drag forces acting on your model, you can use the
Force Reports panel to compute them.
Report Forces...
See Section 28.3 for details.
c Fluent Inc. January 28, 2003 831
Modeling Basic Fluid Flow
832 c Fluent Inc. January 28, 2003
Bien plus que des documents.
Découvrez tout ce que Scribd a à offrir, dont les livres et les livres audio des principaux éditeurs.
Annulez à tout moment.